Technical Assistance Consultant’s Report

Project Number: 47163-001 June 2016

Regional: Development and Dissemination of Climate-Resilient Rice Varieties for Water-Short Areas of South Asia and Southeast Asia (Financed by the Climate Change Fund and the Government of Finland)

Prepared by:

International Rice Research Institute and partner organizations in Bangladesh, India, Nepal, Pakistan, and Philippines

Los Baños, Laguna For: Asian Development Bank

This consultant’s report does not necessarily reflect the views of ADB or the Government concerned, and ADB and the Government cannot be held liable for its contents.

2

Paper No. Description Pages

1 Water scarcity in rice cultivation: current scenario, possible solutions, and likely impact Shalabh Dixit, Arvind Kumar, and Hans Woldring

3-26

2 Developing aerobic rice varieties at IRRI Dule Zhao and Arvind Kumar

27-44

3 Aerobic rice perspectives in India: progress and challenges S.K. Pradhan, A.K. Mall, A. Ghosh, S. Singh, P. Samal, S.K. Dash, O.N. Singh, and A. Kumar

45-56

4 Prospects of aerobic rice in water-limited bunded uplands and shallow lowlands of eastern India N.P. Mandal, M. Variar, and A. Kumar

57-70

5 Varietal improvement and weed management for aerobic rice cultivation in the drought-prone eastern region of India B.N. Singh, Krishna Prasad, A.K. Singh, and Pramod Kumar

71-104

6 Rice varietal improvement and management practices under aerobic and alternate wetting and drying conditions in Nepal: progress and challenges R.B. Yadaw, R.K. Mahato, A. Kumar, B.P. Tripathi, and S.N. Sah

105-143

7 Aerobic rice perspectives in Bangladesh: progress and challenges H.U. Ahmed, K.M. Iftekharuddaula, M. Maniruzzaman, M. Shahidul Islam, and A.K.M. Zakaria

144-174

8 The alternate wetting and drying system of rice cultivation in Bangladesh: progress and challenges H.U. Ahmed, K.M. Iftekharuddaula, M. Maniruzzaman, M. Shahidul Islam, and A.K.M. Zakaria

175-210

9 Aerobic and alternate wetting and drying rice production systems in Pakistan: progress and challenges Riaz A. Mann, Muhammad Ijaz, and Shahbaz Hussain

211-243

10 Suitable management strategies for the aerobic and alternate wetting and drying systems of rice cultivation in Odisha A. Ghosh, S.K. Pradhan, O.N. Singh, and P. Samal

244-270

11 Weed management strategies in dry-seeded rice systems B.S. Chauhan

271-293

12 Adoption and dissemination of alternate wetting and drying technology in pump irrigation system areas in Bangladesh Florencia G. Palis, Rubenito M. Lampayan, Ekkehard Kürschner, and Bas Bouman

294-307

13 Farmers’ participatory research and adoption of aerobic rice in the Philippines Rubenito M. Lampayan, Florencia G. Palis, Junel B. Soriano, and B.A.M. Bouman

308-328

14 Developing and disseminating alternate wetting and drying water-saving technology in the Philippines Rubenito M. Lampayan, B.A.M. Bouman, Florencia G. Palis, and Rica Joy Flor

329-351

15

The rice root-knot nematode Meloidogyne graminicola: a new challenge for water-saving rice production systems in Asia Dirk De Waele, Zin Thu Zar Maung, Pa Pa Win, Pyone Pyone Ki, Yi Yi Myint, Luzviminda Fernandez, Teodora Cabasan, Judith Galeng, Bas Bouman, Casiana Vera-Cruz, and Arvind Kumar

352-367

16 Traits for dry direct-seeded rice Nitika Sandhu and Arvind Kumar

368-400

17 Alleviating soil sickness in tropical aerobic rice: a role for abiotic and biotic interactions C.G.B. Banaay, C. Kreye, M. Hofte, A. Kumar, D. De Waele, V.C. Cuevas, C.M. Vera Cruz, and B.A.M. Bouman

401-419

3

Paper 1 Water scarcity in rice cultivation: current scenario, possible solutions, and likely impact Shalabh Dixit, Arvind Kumar, and Hans Woldring

Rice is the staple food crop for around 50% of the world’s population. Grown on around 160 million hectares globally, rice is cultivated across a wide range of ecosystems varying in topography, soil type, and water regime. Conventionally, rice is cultivated by transplanting seedlings in puddled soil with standing water in the field throughout the crop growing period. Although water stress at any stage of the crop leads to heavy crop losses, it affects rice cultivation in specific ways in irrigated and rainfed rice areas. Increasing labor and fuel costs and decreasing irrigation water availability in irrigated areas lead to extra expenditure and input demands to maintain yield. In rainfed areas, farmers are hardly able to sustain losses due to low yield from less or erratic rainfall. Recent trends in climate change have suggested an increase in water scarcity in rice-growing areas due to less or unpredictable rainfall. Moreover, the increasing demand for freshwater from growing industries and residential areas is leading to a sharp decline in available irrigation water. These situations require the development and testing of water-saving technologies suitable to specific rice-growing ecosystems. Although the conventional methods of rice cultivation through transplanting of rice seedlings in puddled soil lead to heavy water losses through evapotranspiration and seepage, technologies such as direct seeding and alternate wetting and drying provide a suitable alternative to reduce water losses/requirements without compromising yield. However, these practices may be more advantageous if they are coupled with suitable varieties. Large-scale testing and deployment of varieties specifically suitable to target environments and growing systems may lead to increased productivity per unit water used in irrigated areas and assured crop yields in rainfed areas prone to drought.

Rice is the only cereal that can withstand water submergence, and this helps to explain the linkage between rice and water. For many centuries, rice has been grown in a variety of ecosystems, including rainfed uplands and lowlands, deep water, tidal wetlands, and irrigated areas, around the world. In 2011, a total of about 690 million tons of rice was produced from about 160 million hectares of planted area (www.ricestat.irri.org). Each of these ecosystems possesses its own specific features. It is also observed that certain factors affect the rice crop across all these ecosystems. However, the effect of such factors can vary according to ecosystems. The affinity of the rice crop for water is universally known. Rice requires two to three times more water than other cereals (Barker et al 1998, Tuong et al 2005). It is also reported that about 50% of the diverted fresh water in Asia is used to irrigate rice fields (Barker et al 1998). Recent predictions of climate change suggest a further increase in water deficit in the coming years (Wassmann et al 2009). Some 90% of the world’s rice is grown and consumed in Asia (www.asiarice.org/sections/learnrice/riceislife.html) and a large proportion of the total rice production is contributed by small and marginal farm holdings. A majority of these holdings belong to farmers who grow rice with limited inputs under an unpredictable supply of irrigation water. Among the factors that affect rice cultivation in these areas, the growing scarcity of water is the biggest limitation across many rice ecosystems.

4

Among the rice-growing ecosystems, the two major contributors to global rice yield are the irrigated and rainfed rice ecosystems and both are affected in specific ways by the growing water crisis. Particularly in rainfed areas, where the crop may suffer from flood and drought within the same season, the proper management of available water and good agronomic practices are essential. Even within the same geographic locations, the rainfed ecosystem can be divided into upland and lowland ecosystems, which are highly variable in terms of micro-climate and water regimes. Conventional rice production ecosystems (such as puddled transplanted) require an average of 2,500 liters of water to produce 1 kg of rough rice (Bouman et al 2007). A majority of the farmers in both irrigated and rainfed rice ecosystems grow rice in puddled transplanted conditions regardless of the topography and availability of irrigation water. Apart from this, a large area in rainfed rice ecosystems is covered with varieties developed for irrigated conditions. Although these varieties require an adequate supply of irrigation water and high fertilizer inputs, farmers grow them with low inputs because of the unpredictable supply of water. Such a situation leads to low productivity, poverty, and social problems such as large-scale migration in the off-season. With increasing population, expanding residential requirements, increased water demand for industry and household use, and the looming water crisis due to climate change, it is clear that this scenario is going to be more intense not only in rainfed areas but also in irrigated rice-growing areas. The highly diverse micro-climate, climatic variation, and season-specific availability of water in these ecosystems require extensive research for a clear demarcation of specific challenges within each ecosystem and the development and dissemination of site-specific water-saving technologies suited to these ecosystems. The water shortage in rice cultivation Rice is the major user of freshwater resources in the world and it accounts for the withdrawal of 24−30% of the total fresh water and consumption of 34−43% of the total irrigation water of the world. In a majority of the rice growing-areas, cultivation is done in puddled transplanted conditions. Seasonal water inputs for puddled transplanted rice vary from 660 to 5,280 mm depending on growing season, climatic conditions, soil type, and hydrological conditions (Tuong and Bouman 2003). Although the evapo-transpirational losses of water in rice are similar to those of wheat (Kumar and Ladha 2011), it is also reported that the higher water requirements of rice are due to puddling and seepage and percolation losses associated with continuous flooding (Hafeez et al 2007).

The freshwater availability of the world has declined in the past. Water scarcity can mainly be classified into physical and economic water scarcity. Physical water scarcity can be defined as a situation in which the available water resources are insufficient to meet all demand, including minimum environmental flow requirements, whereas economic water scarcity occurs when investments needed to keep up with growing water demand are constrained by financial, human, or institutional capacity (Molden et al 2007). Figure 1 presents the areas of physical and economic water scarcity around the world. Almost all of Africa and a large proportion of Asia, except for southern China and Southeast Asian countries, suffer from either physical or economic water scarcity. With 90% of the world’s rice being grown and consumed in Asia and the

5

large water requirements of conventional rice cultivation systems, it is evident that this water scarcity is going to affect rice production severely on this continent. In the case of Latin America, the conditions remain less severe in terms of water scarcity; however, the growing industrial sector and increasing population in the Latin American countries require water-saving technologies for rice cultivation. Table 1 presents the per capita water availability in the major rice-growing countries in Asia. A decline of 37% to 76% in per capita water availability between 1950 and 2005 occurred and water availability is expected to decline further (Kumar and Ladha 2011). The booming industrial sector of Asia and expanding residential areas due to increasing population make this situation more acute and it is expected that the physical and economic scarcity of water are going to increase further, leading to a sharp decline in agriculture’s share in the freshwater resources in Asia.

Agriculture’s water requirement is fulfilled by the withdrawal of water from rivers, reservoirs, lakes, and aquifers (blue water); the direct use of rainwater stored in the soil (green water); or through the use of underground water resources. Figure 2 shows the percentage of total freshwater withdrawals by agriculture, industry, and municipal sectors around the world. Agriculture accounts for 71% to 86% of the amounts of water withdrawn from different freshwater resources in Asia, Africa, Latin America, and Oceania, which are the major rice-producing areas of the world. On the one hand, it is evident that the agricultural sector is the major user of freshwater resources in rice-growing areas; on the other hand, it is also seen that the share of agriculture in these freshwater resources has declined and is likely to continue declining (Figure 3).

Rainfed rice areas suffer more severely with a shortage of water. Farmers rely on rainwater for field operations such as field preparation, transplanting, and fertilizer application and fluctuations in rainfall lead to a yield penalty not only because of a shortage of water in the fields but also because of the effect on crop management practices. Apart from this, several insects (such as Stenchaetothrips biformis and Nephotettix virescens) and diseases (such as blast and brown spot) are highly related to water deficit and affect the rice crop severely in years of low or erratic rainfall. Recent trends in climate change also suggest an increase in the incidence of water deficit in rice-growing areas. It is evident from the statistics that the conventional rice cultivation systems in both irrigated and rainfed rice-growing areas need to change with the changing water availability and climate change to maintain sustainable rice production in these ecosystems. This scenario requires a detailed understanding of rice-growing ecosystems and the development and dissemination of suitable water-saving technologies specific to target sites. The irrigated rice ecosystem Since the Green Revolution, the irrigated rice ecosystem has been the major contributor of rice production in the world. Rice has been grown under puddled transplanted conditions in these areas with high input supplies. Puddling helps to reduce weeds, facilitates easy seedling establishment, and creates anaerobic conditions to enhance nutrient availability (Sanchez 1973). Irrigated rice ecosystems are marked by relatively larger farm size where the availability of irrigation water leads to an assured and timely application of inputs, leading to high yield. But irrigated rice ecosystems now face a paradox related to water use for cultivation practices. On the one hand, an assured

6

supply of water gives farmers greater flexibility for planning seasonal crop management activities while on the other hand it is also reported that the conventional rice-growing techniques in these areas, including puddling and transplanting, lead to heavy water losses.

In the past half century, there have been massive investments in large-scale public surface irrigation infrastructure (Molden et al 2007). While the world’s cultivated land increased by about 13% from 1961 to 2003, the equipped irrigated area almost doubled from 10% to 18% of cultivated area (Molden et al 2007). However, with increasing competition for fresh water due to decreasing rainfall and groundwater levels and the rapidly increasing industrial sector and growing residential requirements in developing countries, the availability of assured irrigation throughout the season is doubtful. It is estimated that, by 2025, 2 million hectares of Asia’s dry-season rice and 13 million hectares of Asia’s wet-season rice may experience physical water scarcity (Tuong and Bouman 2003).

In addition, rapid economic growth and the booming industrial sector in Asia have increased the demand for labor in the nonagricultural sectors (Kumar and Ladha 2011). Figure 4 shows the trends of available agricultural labor and total population engaged in agriculture from 1960 to 2020 for the world as well as the major rice-growing continents. It has been seen that, in the past 50 years, the percentage of population engaged in agriculture as well as the availability of agricultural labor have declined across the world. Although 58.2% of the total population across the world was engaged in agriculture in 1961, the number declined to 37.6% in 2011 (Figure 5). It is also seen that the population employed as agricultural labor declined from 27.6% in 1961 to 18.8% in 2011 (Figs. 4 and 5). A similar trend has been seen in the past 50 years in Asia, where the percentage of the population engaged in agriculture declined from 77.7% in 1961 to 46.4% in 2011 and agricultural labor decreased from 38.1% in 1961 to 24.4% in 2011 (Figs. 4 and 5).

Increasing fuel costs (Figure 6) require higher investments in irrigation through groundwater. Rice cultivation in puddled transplanted conditions is highly labor- and water-intensive. Agricultural operations such as field preparation and transplanting require much labor and a high amount of water, which are becoming increasingly scarce and expensive, making rice production less profitable (Kumar and Ladha 2011). This situation requires rice-growing technologies that can support rice yield within these areas under a decreasing availability of irrigation water. The rainfed rice ecosystem The rainfed rice ecosystem is another important rice-growing ecosystem. Roughly 46% of the total rice area is rainfed. Unlike the irrigated ecosystem, rice is predominantly grown by small landholders in these areas. Rainfed rice ecosystems are the most unpredictable rice-growing ecosystems. The uneven distribution of rainfall across rice-growing seasons may lead to abiotic stresses such as water deficit and flood within the same area in one growing season or in different seasons. In addition, several biotic stresses coupled with these abiotic stresses can cause a strong reduction in yield. In Asia alone, at least 23 million hectares of rice area are rainfed and subjected to water deficit, causing rice yields to be highly unstable (Pandey and Bhandari 2008). Based on topographical sequences, rainfed rice-growing areas can broadly be classified into

7

rainfed upland and rainfed lowland ecosystems. Both these ecosystems vary greatly in soil type, water-holding capacity, and micro-climate. Being dependent solely on rainwater for cultivation, farmers grow rice with limited inputs, leading to a further reduction in yield. More ironically, a majority of these areas are covered with high-yielding varieties developed specifically for irrigated conditions. Although these varieties are highly input responsive under favorable irrigated conditions, most of them are also highly susceptible to stresses such as water deficit. Extensive cultivation of these varieties, low inputs, and untimely crop management activities due to dependence on rainfall lead to extremely low productivity of rice in these areas.

An analysis of rainfall patterns in rainfed areas of Asia has suggested that the rainfed rice-growing areas of eastern India, northeastern Thailand, and southern China are among the most prone to water deficit (Pandey and Bhandari 2008). The unavailability of irrigation water limits rice cultivation to only one season, leading to social problems such as poverty and migration. Another aspect that leads to low productivity in these areas is the increasing cost of agricultural inputs. Figure 7 shows the price trends of commonly used fertilizers in rice cultivation. In the past five years, the prices of these commodities have fluctuated and have increased manifold. Although a large proportion of the farmers in these areas are below the poverty line, the increase in the prices of commodities leads to a further decrease in applied inputs. This scenario leads to a vicious circle in which low productivity and poverty go hand in hand, giving rise to each other. The situation in these areas requires the development of high- yielding, water-deficit-tolerant, and fertilizer-responsive varieties in particular, which can provide farmers with assured yield in both normal and water-deficit years. Crop management practices that can lead to an increase in water-use efficiency in these areas coupled with suitable varieties may give farmers increased flexibility to grow their crop in the case of fluctuating rainfall. Water-saving technologies Rice can be established by transplanting seedlings in puddled fields or by direct seeding in dry or puddled fields (Kumar and Ladha 2011). Although the intensive water and labor requirements in transplanting of rice in puddled fields are well known, technologies such as dry and wet direct seeding and alternate wetting and drying (AWD) could be an option to produce rice in both irrigated and rainfed rice ecosystems. Direct seeding in particular is being extensively researched and tested in rice-growing areas to develop a suitable alternative to transplanted rice in irrigated as well as rainfed conditions. It often happens that basic prototype technologies undergo modifications based on local needs to maximize benefits (Kumar and Ladha 2011). Several water-saving technologies such as alternate wetting and drying (Li 2001; Tabbal et al 2002), ground-cover systems (Shan et al 2002), the system of rice intensification (SRI, Stoop et al 2002), aerobic rice (Bouman et al 2002), and raised beds (Singh et al 2002) have been proposed to be followed for efficient water use. Water-saving technologies such as AWD have high water productivity, with 15% to 20% water savings without any compromise on yield. However, the water requirement of this production system is also very high as land preparation consists of soaking, followed by wet plowing or puddling of saturated soil, and this requires a large amount of water. In this chapter, aerobic rice, dry direct-

8

seeded rice (aerobic-anaerobic rice), and alternate wetting and drying water-saving technologies have been discussed in detail (Table 2). (i) Dry direct seeding (aerobic rice) Dry direct seeding or aerobic rice refers to a cultivation system in which rice is seeded in well-tilled leveled fields with uniform slope under unpuddled conditions. The crop is cultivated under aerobic conditions with no standing water throughout the season. In this system, rice can be established using different systems such as broadcasting, drilling, or dibbling in a well-prepared field, and direct seeding with zero tillage using a mechanical seed drill or raised beds (Kumar and Ladha 2011). Traditionally, this method has been practiced in rainfed upland and rainfed shallow lowland areas of Asia (Rao et al 2007). However; the water savings related to this method compared with conventional rice-growing practices have made this method increasingly popular in irrigated areas where water is becoming scarce (Kumar and Ladha 2011). Under favorable irrigated conditions, rice is drilled in well-prepared fields, which leads to a savings of seed, and the crop is cultivated with efficient weed control and uniform irrigation throughout the crop season.

The aerobic rice system uses less water than conventional flooded rice (Tuong et al 2005) through the use of rice varieties capable of responding well to reduced water inputs in nonpuddled and nonsaturated soils (Atlin et al 2006, Peng et al 2006). In uplands, this technology is mostly followed in upland rainfed areas or irrigated areas with light soil that have poor water-holding capacity. Favorable upland areas where the land is flat and irrigation/rainfall is sufficient to frequently bring the soil water content close to field capacity are also suitable for aerobic rice cultivation. Examples include the Cerrado region of Brazil and newly formed terraces in the hills of Yunnan, China, where farmers are achieving aerobic rice yields of 3 to 4 t ha−1. In terraces, quite often, soils on upper slopes or terraces in undulating rainfed lowland areas are relatively coarse-textured and well-drained, so that ponding of water occurs briefly or not at all during the growing season. These upper fields have also been proposed as a target domain for aerobic rice cultivation with the use of varieties possessing moderate to high drought tolerance. Water-short irrigated lowland areas, where farmers do not have sufficient water to keep rice fields flooded for a substantial period of time, are also the target domain for aerobic rice. A good example is the North China Plain, where aerobic rice is grown on about 80,000 hectares with supplemental irrigation (Bouman 2007). Rice plants under anaerobic conditions release methane into the atmosphere through roots and stems. Its concentration in the atmosphere has more than doubled during the last 200 years. A continued increase in atmospheric methane concentrations at the current rate is likely to contribute more to future climate change than any other gas except carbon dioxide. Aerobic rice cultivation will curb methane production and save water without affecting productivity. However, under aerobic conditions, emissions of nitrous oxide, another greenhouse gas with a higher global warming potential than methane, are produced. The submergence of rice soils helps maintain soil organic matter (SOM), even with intensive rice cropping and removal of crop residues (Pampolino et al 2008). This maintenance of SOM ensures that carbon remains sequestered in the soil. Soil submergence also promotes biological nitrogen fixation (BNF) (Buresh and Haefele

9

2010), and submerged soils can sustain an indigenous nitrogen supply (INS) for rice as evidenced by long-term stable yields in minus-N plots in long-term experiments. Paddy soils historically cultivated with rice monoculture using puddling and soil submergence can be susceptible to a loss of built-up SOM and INS when converted to the production of aerobic rice. As a general principle, fertilizer N, P, and K requirements for a given target yield could be higher for rice grown on aerobic soil than on submerged soil. A higher need for fertilizer N can arise from lower INS due to lower BNF and possibly lower net N mineralization in aerobic soil. A higher need for fertilizer P can arise from the reduced availability of soil P in aerobic soil. A higher need for fertilizer K can arise from reduced input of K with irrigation water due to water savings with aerobic rice. Soil aeration increases Zn availability on acid soils, but it can decrease Zn and Fe availability on high-pH soils, leading to a need for Fe and Zn fertilization for dry-seeded aerobic rice (Malik and Yadav 2008). Aerobic rice has shown a yield advantage on lighter soils with high seepage and percolation rates. In the temperate region of China, yield up to 5.7 t ha−1 has been reported with aerobic rice on a sandy loam soil at a research station near Beijing (Bouman and van Laar 2006), whereas in on-farm trials yield ranged from 4.5 to 6.5 t ha−1 (Bouman et al 2002). The driving force behind aerobic rice is water (Castañeda et al 2004) and labor economy (Belder et al 2004, 2005, Huaqi et al 2002). Castañeda et al (2004) reported a savings of 73% in land preparation and 56% during crop growth. Aerobic rice yield depends on the effective use of herbicides and other biocides, such as nematicides, and an adequate supply of plant nutrients. Yield obtained with aerobic rice varieties varies from 4.5 to 6.5 t ha−1, which is about double or triple that obtained with traditional upland rainfed varieties but 20% to 30% lower than that obtained with lowland varieties grown under flooded conditions (Farooq et al 2009, Prasad 2011). Recent results from northern China and the Philippines indicate that aerobic rice yield is about 40% lower than that of flooded lowland systems. In aerobic rice experiments at the International Rice Research Institute (IRRI), the yield of aerobic rice gradually declined over time compared with that of a continuously flooded control (George et al 2002, Peng et al 2006). In Brazil also, rice yield declined after 2 years of consecutive upland cultivation and after 5 years of monoculture, and rice yield was only 1.5 t ha−1 compared with 4.3 t ha−1 after 3 years of soybean (Fageria and Baligar 2003, Pinheiro et al 2006).

Nishizawa et al (1971) introduced the term “soil sickness” for the combined effect of allelopathy (Nishio and Kusano 1975), nutrient depletion, buildup of soil-borne diseases and pests (Ventura et al 1981), and soil structure degradation. Recently, Das et al (2011) have reported newly developed aerobic rice genotypes to be superior to traditional upland cultivars in terms of resistance to rice root-knot nematode and they identified a few lines tolerant of nematodes. For sustainability of the aerobic system, breeding efforts must be adopted for sustainable long-term yield through the development of varieties that do not show a yield decline under continuous aerobic rice cultivation or follow appropriate crop rotation practices that allow soil to maintain its fertility and prevent soil-borne diseases, including the development of nematodes. (ii) Dry direct seeding (aerobic-anaerobic)

10

Crop establishment under direct seeding based on sowing methods varies from broadcasting manually or mechanically (using an airplane or power sprayer) to line sowing manually, using a bullock-driven seed drill, and using a mechanized tractor-derived seed drill. Soil saturation conditions vary from remaining totally aerobic as happens in upland rainfed areas to irrigated light soils with poor water-holding capacity, to a combination of aerobic-anaerobic conditions as occur in rainfed lowland, and to continuous anaerobic conditions a few days after emergence as may happen in irrigated areas with an assured water supply or favorable rainfed areas with high and well-distributed rainfall.

The rising scarcity of water and labor is the major driver compelling farmers to shift from flooded transplanted rice to direct seeding. In addition to the scarcity of water, labor availability for agriculture is declining sharply. Under the current socioeconomic environment, most young people are not interested in undertaking tedious farm operations such as transplanting. The availability of new herbicides for weed control provides a viable solution to one of the important constraints, weed problems in direct seeding. Precise land leveling, suitable varieties, mechanized seeding, precise water management, efficient weed control, efficient nutrient management, and mechanized harvesting and threshing through a combine are key to the success of the expansion of direct seeding in the United States, Malaysia, and Sri Lanka.

Mechanized seeding through tractor-driven seed drills has not only facilitated sowing at optimum soil depth of 2 to 3 cm but has also reduced the seed rate from 80 to 200 kg ha−1 to 20 to 25 kg ha−1. The lower seed rate has also helped to overcome spikelet sterility and lodging problems (Kumar and Ladha 2011). Savings of 12% to 35% in irrigation water and up to a 60% savings in labor under direct-seeded rice (DSR) as compared with transplanted rice have been observed. The yield under DSR has been reported to be similar to that of transplanted rice except for an 8% to 28.5% decline reported in India and Pakistan (Kumar and Ladha 2011). Under DSR, a yield decline takes place under conditions of unavailability of assured water for irrigation that leads to exposure of the crop to mild or moderate water-deficient conditions. DSR compared with transplanted rice has been reported to have a lower cost of production by US$22 to $80 ha−1 and this resulted in higher economic returns of $30−50 ha−1 compared with transplanted rice (Kumar and Ladha 2011). (iii) Wet direct seeding Wet direct seeding refers to the seeding of pregerminated rice seeds on the surface of a puddled rice field (aerobic wet direct seeding) or drilling into the puddled soil (anaerobic wet direct seeding) through broadcasting or line sowing using a drum seeder or anaerobic seeder with a furrow opener and closer (Balasubramanian and Hill 2002, Kumar and Ladha 2011). This system has traditionally been followed in eastern India, where seeds are broadcast in puddled fields coupled with beushening and thinning of seedlings to control weeds and ensure proper spacing between seedlings. The method can be highly useful in both rainfed and irrigated lowland conditions through the use of suitable varieties and cultivation practices. Wet seeding has been practiced on a large scale in Vietnam and Sri Lanka. Wet direct-seeded fields can have fewer weed problems than dry direct-seeded fields because puddling and flooding kill a number of

11

weeds but wet direct seeding has lower water savings than dry direct seeding because of the amount of water required for flooding the field for puddling. (iv) Alternate wetting and drying The use of irrigation water for producing lowland rice on puddled paddy soils can potentially be decreased by lowering the depth of standing water and by allowing the soil surface to dry before the next application of irrigation water. The practice of withholding irrigation until several days after the disappearance of ponded water is referred to as alternate wetting and drying (AWD) (Bouman et al 2007). Even without ponded water, rice roots can access the water in the subsurface soil, which remains saturated. Initially, AWD was designed for irrigated rice when surface water went down to 20 cm. Using this method, fields were sometimes exposed to mild drought stress, resulting in a yield reduction in AWD as compared to transplanted rice. Now, safe AWD that includes irrigating the field when water reaches 15-cm depth and keeping the field flooded for 10 days after transplanting or 20 days after direct seeding as well as 1 week before flowering to 1 week after flowering is recommended. Safe AWD has been reported to yield similar to transplanted rice and it reduces water input by 15−30% (Bouman et al 2007, Tuong 2009). AWD results in periodic soil aeration, but the extent and duration of soil drying when implemented at safe levels, which do not result in a loss of rice yield, are unlikely to have much effect on soil organic matter and plant availability of macronutrients in the soil. No adjustments in management practices for fertilizer N, P, and K are proposed with safe AWD. Erratic rainfall in rainfed rice ecosystems and the periodic unavailability of irrigation water in irrigated rice ecosystems can prolong the duration of soil drying, resulting in a soil water deficit, leading to a loss in rice yield. As a general principle, as soil drying becomes more prolonged and severe, the availability of soil P to rice tends to decrease and the availability of zinc in acid soils tends to increase (Dobermann and Fairhurst 2000). Fertilizer rates should be adjusted to the anticipated water-limited grain yield of rice (Haefele and Bouman 2009). The better root growth in rice under AWD reported by Zhang et al (2009) and Banoc et al (2000) can help overcome mild drought stress in the case of unavailability of water for irrigation in rainfed lowland or due to a delay in irrigation in irrigated systems. Breeding rice for water-saving conditions In practice, there are not many examples of systematic breeding programs for different water-saving technologies except for aerobic rice. Also, for aerobic rice, in most cases, the varieties grown for uplands or new varieties developed for upland conditions are grown under aerobic rainfed conditions. These varieties, because of their semitall height, low yield potential, and lodging susceptibility under high-input management, are not fit to be grown under aerobic irrigated conditions. High yield under aerobic conditions can be achieved by developing rice varieties that combine the traits of upland varieties (tolerance of water deficit, early vigor, and weed competitiveness) and lowland varieties (high yield potential and good grain quality).

In China, China Agricultural University (CAU), China Academy of Agricultural Sciences (CAAS), the Liaoning Province Academy of Agricultural Sciences (LPAAS), and the Dandong Academy of Agricultural Sciences (DAAS) were the institutes starting

12

a breeding program for aerobic rice in the 1980s. In the 1990s, varieties such as Qinai, Hedda 77-2, Zhougyuan 1 and 2, and Han 72 were released. However, the large-scale adoption of the released varieties due to susceptibility to rice blast, weak ability to emerge through the soil surface, and low vegetative vigor was not successful (Huaqi et al 2002, Prasad 2011). Later, at CAU, a strong genetic recombination of lowland and upland rice varieties started in 1984. This included upland varieties from Yunnan Province of China, Thailand, and Laos and lowland varieties from China and Japan. The breeding program at CAU grew the early-generation material in lowland environments for the selection of appropriate plant architecture and late-generation material in aerobic environments for selection for tolerance of water deficit, aerobic adaptation, and higher grain yield. This led to the development of new-generation aerobic rice varieties Han Dao (HD) 297 from Mujiao 78-595 (a lowland variety (LV)) crossed to Khaoman (an upland variety (UV)), HD 277 from Quiguang (LV) crossed to Ban Li 1 (UV), and HD 502 from Quiguang (LV) crossed to Hangkelaoshuya (UV). In addition, LPAAS released Han 58 and DAAS released Danjing 5. These new varieties have stronger tolerance of water deficit, reduced plant height, increased lodging resistance, erect upper leaves, stronger resistance to blast, higher grain yield, and better grain quality. HD 277 and H 58 are currently the most extensively grown aerobic rice varieties in China (Huaqi et al 2002, Prasad 2011). In 2000, the China National Aerobic Rice Network estimated that about 190,000 ha in China were planted to these new aerobic rice varieties, of which some 80,000 ha were found in the water-short North China Plain (Wang et al 2002). Thus, aerobic rice in China is grown in both irrigated areas where water has become too scarce or expensive to grow flooded rice and rainfed areas where rainfall is insufficient for flooded rice production, but sufficient for upland crops.

In Brazil, the National Research Centre for Rice and Beans (CNPAF) of Empresa Brasileira de Pesquisa Agropecuaria (EMBRAPA) located in Goiâs initiated work on breeding aerobic rice in 1975 (Pinheiro et al 2006). Phase I (1975 to 1985) of the aerobic rice breeding program concentrated on drought tolerance, blast resistance, and yield stability targeting exclusively the unfavorable savanna region. In phase II (1985 to 1990), the breeding strategy expanded to include selection for high yield potential targeting the favorable savanna area. Emphasis was given to grain quality, yield potential, and blast resistance in phase III (1990 onward). The strong priority for blast resistance continues in the aerobic rice breeding program due to the high incidence of blast in the savanna system (Prabhu et al 1999, Prabhu and Filippi 2002). Maravilha and Primavera were the first aerobic rice varieties (released in 1996) that combined good grain quality and desirable aerobic rice characteristics (Pinheiro 1999). Primavera received a preference rating very close to that of BR IRGA 409, considered commercially the most competitive irrigated variety (Guimaraes et al 2001). Recently released aerobic rice varieties in Brazil are Talento and Soberana (Pinheiro et al 2006). These varieties are japonica-indica derivatives and their yield potential is 5 t ha−1. A latest addition to rice varieties for the uplands of Brazil is ANCambara (Santana 2010). The high-input rice production system in Brazil has led to the development of input-responsive cultivars combined with management practices that reduce the risk of production in aerobic soils, including a shift to an assured irrigation supply that reduces incidences of water deficit.

13

IRRI, Philippines, initiated its own program for developing aerobic rice varieties with a focus on tropical and subtropical regions. A multinational varietal testing program including indica, japonica, aus, and intermediate types was launched in 1986 with a focus on genotype- environment interactions and blast resistance (Mackill et al 1996, Lafitte et al 2002). The first generation of tropical aerobic rice varieties developed by IRRI included IR55423-01 (named Apo) and UPLRI-5 from the Philippines, B6144-MR-6-0-0 from Indonesia, and CT6510-24-1-2 from Colombia. These varieties were derived from crosses between indica and tropical japonica parents. In the Philippines, Bouman et al (2005) and Peng et al (2006) reported yields up to 6 t ha−1. Recently, IRRI-developed breeding line IR74371-54-1-1 was released as variety Sahod Ulan 1 in 2009 and IR79913-B-176-B-4 was released in 2011 as Katihan 1 in the Philippines. Also, two breeding lines from IRRI, IR79971-B-191-B-B and IR79971-B-227-B-B, have been released as varieties Inpago Lipi GO1 and Inpago Lipi GO2, respectively, in Indonesia for the rainfed upland ecosystem. These varieties have yielded 5.5 t ha−1 or more under the aerobic system in the Philippines and Indonesia. In India, the work on developing varieties suitable for aerobic rice started in 2005 and is generally restricted to screening available varieties (Prasad 2011). At the Indian Agricultural Research Institute (IARI), New Delhi, Apo, IR55419-04, IR7437-46-1-1 (IRRI varieties), Pusa 834, Pusa RH 10 (IARI), and Pro-Agro 6111 (a commercial hybrid) yielded above 4 t ha−1 under aerobic conditions (Singh and Chinnusamy 2007). Recently, the All India Coordinated Rice Improvement Project (AICRIP) also initiated a systematic evaluation of aerobic rice lines all over India to identify suitable aerobic rice varieties for different water-short regions of the country. In 2007, India officially released for cultivation its first aerobic rice variety, MAS 946-1, followed by MAS 26 (2008), at the University of Agricultural Sciences, Bangalore (Gandhi et al 2012). Recently, the Central Rice Research Institute-Cuttack released IR55423-05 (Apo) as CR Dhan 200. Under aerobic conditions, moderate to severe water deficit conditions due to the absence of rainfall in rainfed areas and mild drought in irrigated areas because of 1 or 2 days’ delay in irrigation to maintain saturation may lead to a yield reduction. For rainfed aerobic conditions, semitall height, high early vigor, weed competitiveness, tolerance of moderate to severe water deficit conditions, moderate to high yielding ability, a strong sturdy culm, resistance to/tolerance of blast and brown spot, and tolerance of nematode (Meloidogyne graminicloa) are some of the characteristics required to develop suitable varieties. For aerobic irrigated conditions, semidwarf height, mild water deficit tolerance, high yielding ability, strong sturdy culm, high input responsiveness, and tolerance of blast, brown spot, and nematode (M. graminicola) are some of the characteristics required to develop suitable rice varieties. For weed control, preemergence weedicide plus one hand weeding in aerobic rainfed conditions and preemergence plus postemergence weedicides in aerobic irrigated conditions may be adopted. Recent results at IRRI have indicated that roots can play a very important role in efficient nutrient uptake under aerobic rainfed as well as aerobic irrigated conditions. Earlier, Kawata and Ishihara (1959) also reported the development of root hairs and nodal roots in aerobic soils under drying conditions. More lateral roots and root hairs in the upper 30-cm soil depth may provide better access of plants to not only water but also to nutrients. To get yield similar to that of the transplanted system under the aerobic system, nutrient availability has to be increased. As root traits are hard to be used

14

directly in breeding programs, the identification of root traits enhancing the ability of aerobic rice varieties to uptake N, P, and K, as well as Zn and Fe, under reduced nutrient availability in aerobic conditions, the identification of the QTLs related to such increased nutrient uptake through root traits and the introgression of identified QTLs in varieties following a marker-assisted backcross approach could be an appropriate strategy to bring a further increase in yield of aerobic rice varieties. Generally, rice varieties bred for puddled transplanting are used in direct-seeding (aerobic-anaerobic) conditions. The lack of suitable varieties is a major constraint to achieving the maximum potential of direct seeding. Varieties suitable for direct seeding should possess anaerobic germination, high seedling vigor, erect leaves with low specific leaf area and high chlorophyll content for high crop growth during the reproductive stage along with high remobilization ability for higher spikelet fertility, strong thick and sturdy culm with long heavy panicles positioned at lower height for lodging resistance, high yield potential, and high nutrient-use efficiency under DSR (Kumar and Ladha 2011). Again, for dry direct seeding, nutrient uptake can be increased by marker-assisted breeding for root traits: a higher number of nodal roots and root hairs as suggested for aerobic conditions. For alternate wetting and drying, mostly varieties grown under transplanted conditions can be effectively grown. However, it seems that root traits with an increased number of lateral roots and root hairs in 0 to 15-cm soil depth can have increased access of water and nutrients to plants to avoid any yield decrease because of a decrease in irrigation. Impact of water-saving technologies on rice cultivation The large-scale adoption of water savings will help address the increasing problems of water scarcity and make rice cultivation as well as rice production sustainable and help feed the growing population. Further, water-saving technologies will slow down ground-level water depletion as a result of a decrease in the frequency of irrigation and water saved in rivers, and in-flows will to a certain extent help recharge groundwater levels. Overall, water savings will enhance the food and income security of the rural and urban population, improve the environmental sustainability of rice production systems in water-short areas, reduce the use of gasoline and electricity, and thus contribute to adopting effective measures to offset climate change. References Atlin GN, Lafitte HR, Tao D, Laza M, Amante M, Courtois B. 2006. Developing rice

cultivars for high-fertility upland systems in the Asian tropics. Field Crops Res. 97:43-52.

Balasubramanian V, Hill JE. 2002. Direct seeding of rice in Asia: emerging issues and strategic research needs for the 21st century. In: Pandey S, Mortimer M, Wade L, Tuong TP, Lopez K, Hardy B, editors. Direct seeding: research strategies and opportunities. Los Baños (Philippines): International Rice Research Institute. p 15-39.

Banoc DM, Yamauchi A, Kamoshita A, Wade LJ, Pardales JR. 2000. Dry matter production and root-system development of rice cultivars under fluctuating moisture. Plant Prod. Sci. 3:197-207.

15

Barker R, Dawe D, Tuong TP, Bhuiyan SI, Guerra LC. 1998. The outlook for water resources in the year 2020: challenges for research on water management in rice production. In: Assessment and orientation towards the 21st century. Proceedings of the 19th Session of the International Rice Commission, 7-9 September 1998, Cairo, Egypt. FAO. p 96-109.

Belder P, Bouman BAM, Cabangon R, Lu G, Quilang EJP, Li Y, Spiertz JHJ, Tuong TP. 2004. Effect of water saving irrigation on rice yield and water use in typical lowland conditions in Asia. Agric. Water Manage. 65:193-210.

Belder P, Bouman BAM, Spiertz JHJ, Peng S, Castañeda AR, Visperas RM. 2005. Crop performance, nitrogen and water use in flooded and aerobic rice. Plant Soil 273:167-182.

Bouman BAM, Lampayan RM, Tuong TP. 2007. Water management in irrigated rice: coping with water scarcity. Los Baños (Philippines): International Rice Research Institute. 54 p.

Bouman BAM, van Laar HH. 2006. Description and evaluation of the rice growth model ORYZA2000 under nitrogen limited conditions. Agric. Syst. 87:249-273.

Bouman BAM, Wang H, Yang X, Zhao JF, Wang CG. 2002. Aerobic rice (Han Dao): a new way of growing rice in water-short areas. In: Proceedings of the 12th

International Soil Conservation Organization Conference, 26-31 May 2002, Beijing, China. p 175-181.

Bouman BAM, Xiaoguang Y, Wang H, Wang Z, Zhao J, Chen B. 2005. Performance of aerobic rice varieties under irrigated conditions in North China. Field Crops Res. 97(1):53-61.

Bouman BAM. 2007. The aerobic rice reality. Rice Today. July-September. 38 p. Buresh RJ, Haefele SM. 2010. Changes in paddy soils under transition to water-saving

and diversified cropping systems. In: Soil solutions for a changing world. 19th World Congress of Soil Science, 1-6 August 2010, Brisbane, Australia. Published on DVD.

Castañeda AR, Bouman BAM, Peng S, Visperas RM. 2004. Mitigating water scarcity through an aerobic system of rice production: new directions for a diverse planet. Proceedings of the 4th International Crop Science Congress, Brisbane, Australia, 26 September-1 October 2004. www.cropscience.org.au.

Das K, Zhao D, De Waele D, Tiwari RKS, Srivastava DK, Kumar A. 2011. Reactions of traditional upland and aerobic rice genotypes to rice root knot nematode (Meloidogyne graminicola). J. Plant Breed. Crop Sci. 3(7):131-137.

Dobermann A, Fairhurst T. 2000. Rice: nutrient disorders and nutrient management. Potash and Phosphate Institute, and Potash and Phosphate Institute of Canada: Singapore and International Rice Research Institute, Los Baños, Philippines. 191 p.

Fageria NK, Baligar VC. 2003. Upland rice and allelopathy. Commun. Soil Sci. Plant Anal. 34:1313-1329.

FAO. 2006. The state of food insecurity in the world. Rome, Italy. www.fao.org/icatalog/inter. e.htm.

Farooq M, Kobayashi N, Wahid A, Ito O, Basar SMA. 2009. Strategies for producing more rice with less water. Adv. Agron. 101:351-388.

16

Gandhi RV, Rudresh N, Shivamurty SM, Hittalmani S. 2012. Performance and adoption of new aerobic rice variety MAS 946-1 (Sharada) in southern Karnataka. Karnataka J. Agric. Sci. 25 (1):5-8.

George T, Magabanua R, Garrity DP, Tubana BS, Quiton J. 2002. Rapid yield loss of rice cropped successively in aerobic soil. Agron. J. 84:981-989.

Guimaraes EP, Viera NRA, Pinheiro B da. 2001. Breeding for specialty rice in Latin America: status and perspectives. In: Chaudhary RC, Tran VD, editors. Specialty rices of the world: breeding, production and marketing. Rome: FAO. p 317-322.

Haefele SM, Bouman BAM. 2009. Drought-prone rainfed lowland rice in Asia: limitations and management options. In: Serraj R, Bennett J, Hardy B, editors. Drought frontiers in rice: crop improvement for increased rainfed production. Singapore and Los Baños (Philippines): International Rice Research Institute. World Scientific. p 211-232.

Hafeez MM, Bouman BAM, Van de Giesen N, Vlek P. 2007. Scale effects on water use and water productivity in a rice-based irrigation system (UPRIIS) in the Philippines. Agric. Water Manage. 92:81-89.

Huaqi W, Bouman BAM, Zhao D, Changgui W, Moya PF. 2002. Aerobic rice in northern China: opportunities and challenges. In: Bouman BAM, Hengsdijk H, Hardy B, Bindraban PS, Tuong TP, Ladha JK, editors. Water-wise rice production. Los Baños (Philippines): International Rice Research Institute. p 143-154.

Kawata S, Ishihara S. 1959. Study on the root hairs in rice plants. Proc. Crop Sci. Soc. Jpn. 29:341-348.

Kumar V, Ladha JK. 2011. Direct seeding of rice: recent developments and future research needs. In: Donald LS, editor. Adv. Agron. Newark, Del (USA). p 297-413.

Lafitte HR, Courtois B, Arraudeau M. 2002. Genetic improvement of rice in aerobic systems: progress from yield to genes. Field Crops Res. 75:171-190.

Li YH. 2001. Research and practice of water-saving irrigation for rice in China. In: Barker R, Li Y, Tuong TP, editors. Water-saving irrigation for rice. Proceedings of an International Workshop, 23-25 March 2001, Wuhan, China. Colombo (Sri Lanka): International Water Management Institute. p 135-144.

Mackill DJ, Coffman WR, Garrity DP. 1996. Rainfed lowland rice improvement. Los Baños (Philippines): International Rice Research Institute. 242 p.

Malik RK, Yadav A. 2008. Direct-seeded rice in the Indo-Gangetic Plain: progress, problems and opportunities. In: Humphreys E, Roth CH, editors. Permanent beds and rice-residue management for rice–wheat systems in the Indo-Gangetic Plain. Proceedings of a workshop held in Ludhiana, India, 7-9 September 2006. ACIAR, Canberra. p 133-143.

Molden D, Frenken K, Barker R, de Fraiture C, Mati B, Svendsen M, Sadoff C, Finlayson CM, Attapatu S, Giordano M, Inocencio A, Lannerstad M, Manning N, Molle F, Smedema B, Vallée D. 2007. Trends in water and agriculture development. In: Molden D, editor. Water for food, water for life: a comprehensive assessment of water management in agriculture. Colombo (Sri Lanka) and London: International Water Management Institute and Earthscan. p 57-89.

17

Nishio M, Kusano S. 1975. Effect of root residues on the growth of upland rice. Soil Sci. Plant Nutr. 21:391-395.

Nishizawa T, Ohshima Y, Kurihara H. 1971. Survey of the nematode population in the experimental fields of successive or rotative plantation. Proc. Kanto-Tosan Plant Protect. Soc. 18:121-122.

Pampolino MF, Laureles EV, Gines HC, Buresh RJ. 2008. Soil carbon and nitrogen changes in long-term continuous lowland rice cropping. Soil Sci. Soc. Am. J. 72:798-807.

Pandey S, Bhandari H. 2008. Drought economic costs and research implications. In: Serraj R, Bennett J, Hardy B, editors. Drought frontiers in rice: crop improvement for increased rainfed production. Singapore and Los Baños (Philippines): International Rice Research Institute. World Scientific. p 3-17.

Peng S, Bouman BAM, Visperas RM, Castañeda A, Nie L, Park HK. 2006. Comparison between aerobic and flooded rice in the tropics: agronomic performance in an eight-season experiment. Field Crops Res. 96:252-259.

Pinheiro B da S, Castro E da M de, Guimaraes CM. 2006. Sustainability and profitability of aerobic rice production in Brazil. Field Crops Res. 97:34-42.

Pinheiro B da S. 1999. Caracteristicas morfologicas da planta relacionadas a produtividade. In: Viera NRA, Santos AB, de Ana dos Santos, Guimeraes E P, editors. A Cultura do Arroz no Brasil. Embrapa Arroz e Feijao, Santo Antonio de Goias, Brazil. p 116-147.

Prabhu AS, Filippi MC. 2002. Ocorrencia do mal-do pe causado por Gaeumanomyces gamins var. graminis, uma nova enfermidade em arroz no Brasil. Fitopatol. Bras. p 27.

Prabhu AS, Fillippi MC, Ribeiro AS. 1999. Doencas e seu controle. In: Viera NRA, Santos AB, de Ana dos Sant, editors. A Cultura do Arroz no Brasil. Embrapa Arroz e Feijao, Santo Antonio de Goias, Brasil. p 267–307.

Prasad R. 2011. Aerobic rice systems. Adv. Agron. 111:207-247. Rao AN, Johnson DE, Sivaprasad B, Ladha JK, Mortimer AM. 2007. Weed

management in direct-seeded rice. Adv. Agron. 93:153-255. Sanchez PA. 1973. Puddling tropical soils: effects on water losses. Soil Sci. 115:303-

308. Santana MR. 2010. Shift to rice. Rice Today 9(3):16. Shan L, Dittert K, Tao H, Kreye C, Xu Y, Shen Q, Fan X, Sattelmacher B. 2002. The

ground cover rice production system (GCRPS) - a successful new approach to save water and increase nitrogen fertilizer efficiency? In: Bouman BAM, Hengsdijk H, Hardy B, Bindraban PS, Tuong TP, Ladha JK, editors. Water-wise rice production. Los Baños (Philippines): IRRI. p 187-196.

Shiklomanov. 2000. Water and food, water for life insights from the comprehensive assessment of water management in agriculture. International Water Management Institute, Stockholm. World water week, 2006. International Water Resource Association Water International Vol. 25 (1), March 2000.

Singh AK, Chinnusamy V. 2007. Aerobic rice: a success story. Indian Farming 57(8):7-10.

Singh AK, Choudury BU, Bouman BAM. 2002. The effect of rice establishment techniques and water management on crop-water relations. In: Bouman BAM,

18

Hengsdijk H, Hardy B, Bindraban PS, Tuong TP, Ladha JK, editors. Water-wise rice production. Los Baños (Philippines): International Rice Research Institute. p 237-244.

Stoop W, Uphoff N, Kassam A. 2002. A review of agricultural research issues raised by the system of rice intensification (SRI) from Madagascar: opportunities for improving farming systems for resource-poor farmers. Agric. Syst. 71:249-274.

Tabbal DF, Bouman BAM, Bhuiyan SI, Sibayan EB, Sattar MA. 2002. On-farm strategies for reducing water input in irrigated rice: case studies in the Philippines. Agric. Water Manage. 56(2):93-112.

Tuong TP, Bouman BAM, Mortimer M. 2005. More rice, less water: integrated approaches for increasing water productivity in irrigated rice-based systems in Asia. Plant Prod. Sci. 8:231-241.

Tuong TP, Bouman BAM. 2003. Rice production in water scarce environments. In: Kijne JW, Barker R, Molden D, editors. Water productivity in agriculture: limits and opportunities for improvement. Wallingford (UK): CAB International. p 53-67.

Tuong TP. 2009. Promoting AWD in Bangladesh. Ripple 4(3):1-2. Ventura W, Watanabe I, Castillo M B, de la Cruz A. 1981. Involvement of nematodes in

the soil sickness of a dryland rice based cropping system. Soil Sci. Plant Nutr. 27:305-315.

Wang H, Bouman BAM, Zhao D, Wang C, Moya PF. 2002. Aerobic rice in northern China: opportunities and challenges. In: Bouman BAM, Hengsdijk H, Hardy B, Bindraban PS, Tuong TP, Ladha JK, editors. Water-wise rice production. Proceedings of the International Workshop on Water-Wise Rice Production, 8–11 April. Los Baños (Philippines): International Rice Research Institute. p 143-154.

Wassmann R, Jagadish SVK, Sumfleth K, Pathak H, Howell G, Ismail A, Serraj R, Redoña E, Singh RK, Heuer S. 2009. Regional vulnerability of climate change impacts on Asian rice production and scope for adaptation. Adv. Agron. 102:91-133.

Zhang H, Xue Y, Wang Z, Yang J, Zhang J. 2009. An alternate and moderate soil drying regime improves root and shoot growth in rice. Crop Sci. 49:2246-2260.

Notes Authors’ addresses: S. Dixit and A. Kumar, International Rice Research Institute, Los

Baños, Philippines; H. Woldring, Asian Development Bank, Manila Philippines. Figure 1. Areas of physical and economic water scarcity across the world. Figure 2. Percentage of total freshwater withdrawals by agriculture, industry, and

municipal sectors around the world. Sources: FAO 2006, Molden et al 2007. Figure 3. Increasing sectoral competition for blue-water withdrawals for human uses.

Sources: Molden et al (2007), Shiklomanov (2000). Figure 4. Percentage of total population working as agricultural labor across the world

and on major rice-growing continents: Asia, Africa, and Latin America. Source: http://ricestat.irri.org:8080/wrs/.

19

Figure 5. Percentage of total population engaged in agriculture across the world and on major rice-growing continents: Asia, Africa, and Latin America.

Source: http://ricestat.irri.org:8080/wrs/. Figure 6. Trends of prices of crude oil across the world from 1967 to 2011. Source: http://ricestat.irri.org:8080/wrs/. Figure 7. Price trends of commonly used fertilizers in rice cultivation around the world

from 1967 to 2011. Source: http://ricestat.irri.org:8080/wrs/.

20

Table 1. Trends of per capita water availability in the major rice-growing countries of the world.

Country m3

1950 1995 2000 2005 2010 2015 2020 2025 2050 Bangladesh 56,411 19,936 16,744 15,393 14,335 13,452 12,703 12,086 10,593 China 5,047 2,295 2,210 2,134 2,068 2,006 1,956 1,927 1,976 India 5,831 2,244 2,000 1,844 1,717 1,611 1,525 1,457 1,292 Indonesia 31,809 12,813 12,325 11,541 10,881 10,361 9,952 9,609 8,781 Japan 6,541 4,374 4,317 4,292 4,307 4,348 4,423 4,528 5,381 Malaysia 74,632 22,642 19,593 17,790 16,336 15,179 14,242 13,503 11,497 Nepal 21,623 7,923 6,958 6,245 5,695 5,230 4,820 4,470 3,467 Pakistan 11,844 3,435 3,159 2,822 2,533 2,277 2,069 1,900 1,396 Philippines 15,390 4,761 4,158 3,778 3,450 3,175 2,945 2,754 2,210 South Korea 3,247 1,472 1,424 1,390 1,363 1,345 1,336 1,336 1,500 Sri Lanka 5,626 2,410 2,302 2,212 2,117 2,041 1,990 1,961 1,990 Thailand 8,946 3,073 2,871 2,714 2,627 2,559 2,505 2,465 2,440 Vietnam 12,553 5,095 4,780 4,472 4,223 4,015 3,836 3,684 3,367

Source: Kumar and Ladha 2011.

21

Table 2. Direct-seeding techniques in rice cultivation.

Method Tillage Seedbed conditions

Seed environment Seeding methods

Dry direct seeding Dry tillage, reduced tillage, zero tillage

Unpuddled

Aerobic Broadcasting, dibbling, drilling

Wet direct seeding

Dry and wet tillage Puddled Aerobic, anaerobic

Broadcasting, line sowing, Drilling

Alternate wetting and drying

Dry and wet tillage Puddled/unpuddled

Aerobic, anaerobic

Drilling, transplanting, broadcasting

Water seeding Dry and wet tillage Standing water

Anaerobic Broadcasting

Modified from: Kumar and Ladha 2011.

22

Figure 1. Areas of physical and economic water scarcity across the world. Source: Molden et al (2007).

23

Figure 2. Percentage of total freshwater withdrawals by agriculture, industry, and municipal sectors around the world.

Sources: FAO (2006), Molden et al (2007).

70

32

39

68

71

72

81

86

20

53

48

9

10

10

11

4

10

15

13

23

19

18

7

10

World

Europe

North America

Caribbean

Latin America

Oceania

Asia

Africa

Agriculture Industry Municipalities

24

Figure 3. Increasing sectoral competition for blue-water withdrawals for human uses. Sources: Molden et al (2007), Shiklomanov (2000).

25

Figure 4. Percentage of total population working as agricultural labor across the world and on major rice-growing continents: Asia, Africa, and Latin America.

Source: http://ricestat.irri.org:8080/wrs/.

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

40.0

45.0

Asia Africa Latin America World

Perc

enta

ge o

f to

tal p

op

ula

tio

n

26

Figure 5. Percentage of total population engaged in agriculture across the world and on major rice-growing continents: Asia, Africa, and Latin America.

Source: http://ricestat.irri.org:8080/wrs/.

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

Asia Africa Latin America World

Perc

enta

ge o

f to

tal p

op

ula

tio

n

27

Figure 6. Trends of prices of crude oil across the world from 1967 to 2011. Source: http://ricestat.irri.org:8080/wrs/.

28

Figure 7. Price trends of commonly used fertilizers in rice cultivation around the world from 1967 to 2011.

Source: http://ricestat.irri.org:8080/wrs/.

29

Technical Assistance Consultant’s Report

Paper 2 Developing aerobic rice varieties at IRRI Dule Zhao and Arvind Kumar

Aerobic rice cultivation is a strategy to cope with water scarcity while ensuring rice food security. Aerobic rice varieties are required to be input-responsive, drought-tolerant, weed-competitive, and location-specific disease-, insect-, and nematode-resistant to achieve high yield under aerobic soil conditions. Breeding strategies and selection protocols for drought tolerance, weed competitiveness, pest resistance, and good grain quality of aerobic rice at the International Rice Research Institute (IRRI) are described. Over years of breeding efforts targeting tropical regions since 2000, IRRI has identified a number of donors with desired traits of aerobic rice, and developed several elite aerobic rice lines with yield potential of 4 to 5 t ha−1. These lines have been disseminated to a number of Asian countries for testing and adoption. To date, six IRRI-developed aerobic rice lines (IR74371-46-1-1, IR74371-54-1-1, IR74371-70-1-1, IR79971-B-191-B-B, IR79971-B-227-B-B, and IR79913-B-176-B-4) have been released as either upland or rainfed lowland rice varieties in five Asian countries.

Among four traditionally classified rice ecosystems—irrigated, rainfed lowland, upland, and flood-prone rice—irrigated rice, usually referred to as lowland rice, occupies 55% of the world rice area while producing 75% of world rice (Dobermann and Fairhurst 2000). However, lowland rice production is being threatened by a growing water scarcity worldwide (Tuong and Bouman 2003). By 2025, a “physical water scarcity” is projected for more than 2 million ha of dry-season lowland rice and 13 million ha of wet-season lowland rice in Asia, and an “economic water scarcity” is expected to hamper most of Asia’s 22 million ha of dry-season lowland rice (Tuong and Bouman 2003). Some areas that are currently planted to lowland rice are likely to experience interruptions of water supply, or even to revert to primarily rainfed production. The increasing water scarcity highlights the need to improve the water productivity of rice, and its ability to tolerate periods of water shortage, to ensure adequate food for future generations.

Aerobic rice is a promising approach for dealing with the emerging water shortage and maintaining sustainable rice production. Aerobic rice is a relatively new production system in which specially developed varieties are direct seeded and grown in nonpuddled and nonflooded aerobic soils. The crop can be either irrigated or rainfed, but, in either case, the “aerobic soil” is kept moist throughout the growing season. In a rainfed system, aerobic rice relies completely on rainfall; thus, the soil moisture content often drops below 70% of field capacity for some time, but, on the other hand, the soil may be flooded by rainfall for some time in a growing season too. Aerobic rice is targeted to water-short areas where there is irrigation, but where the water supply is insufficient for growing lowland rice, and to rainfed areas where rainfall is sufficient to frequently bring the soil water content close to field capacity. These areas can be uplands, upper slopes, and shallow lowlands in a toposequence (Bouman 2007). Aerobic rice varieties are designed to be aerobic-soil-adapted, input-responsive, and weed-competitive to attain high yield under aerobic soil conditions. They are targeted to combine drought-adapted characteristics of traditional upland rice varieties with high-yielding characteristics of lowland rice varieties. They are better adapted to intensified management with moderate input use than traditional upland rice varieties, which have a low harvest index and low yield potential, and are

30

prone to lodging when fertilized (Atlin et al 2006). They are well adapted to aerobic soil conditions and significantly outyield lowland rice varieties under such conditions (Zhao et al 2010). Therefore, aerobic rice varieties are different from either traditional upland rice or lowland rice. They are thus defined as a new type of rice germplasm. Breeders in China and Brazil have developed locally adapted aerobic rice varieties yielding 4 to 6 t ha−1 (Wang et al 2002, Pinheiro et al 2006) with limited water supply (600 mm of rainfall and irrigation water in northern China), resulting in almost doubled water productivity and 50% water savings relative to conventional lowland rice (Bouman et al 2002, Wang et al 2002). The water savings of the aerobic rice system relative to the lowland rice system mainly result from decreased losses of seepage, percolation, and evaporation because of no ponded water layer in rice fields, and from zero water consumption for land preparation because aerobic rice fields are prepared dry (Bouman 2001, Bouman et al 2005). A study in Laos showed that, under rainfed conditions, aerobic rice varieties developed for the tropics were more responsive to N, and 86% higher yielding than traditional upland rice varieties across five sites and three N rates (Saito et al 2007). In farmers’ fields in Yunnan, China, and in Luzon, Philippines, where traditional upland rice varieties are grown, aerobic rice varieties outyielded traditional checks by 30% to 50% with moderate N application under rainfed aerobic conditions (Atlin et al 2006). In West Africa, aerobic rice varieties nearly doubled the yield of traditional varieties in aerobic soils with either high or low fertility (Saito and Futakuchi 2009). These reports indicate that aerobic rice production is both a strategy for improving water productivity in water-short irrigated areas and a route to agricultural intensification and increased yield in rainfed areas. Constraints to aerobic rice In designing and developing aerobic rice varieties, constraints to aerobic rice production systems in the target population of environments must be considered so as to overcome or decrease the negative effects of those constraints by genetic improvements. The major identified constraints to aerobic rice production are drought, weeds, nutrient availability, and aerobic soil-associated pests and diseases. (i) Drought Drought is never a limiting factor to yield when an aerobic rice variety is grown under irrigated conditions, but it is a severe constraint under rainfed conditions in tropical regions. In rainfed areas, severe drought often occurs due to erratic rainfall at any growing stage in a rainy season. Drought at the reproductive stage is particularly damaging to rice yield: even short periods of drought just before or during flowering can cause severe yield losses (O’Toole 1982). Therefore, aerobic rice varieties targeted at rainfed environments require not only high yield potential under favorable aerobic soil condition but also tolerance of severe drought stress at flowering (Atlin et al 2008). Obviously, drought tolerance is an important attribute that needs breeding for in a breeding program. (ii) Weeds Weed pressure in aerobic rice is heavy because aerobic rice is direct seeded, and thus rice seedlings have no head-start over weeds as transplanted rice seedlings do, and because aerobic rice fields lack a standing-water layer to suppress weed growth (De Datta and Baltazar 1996). Weeds in direct-seeded upland rice, if not properly controlled, may result in yield losses of 30% to 100% (De Datta and Llagas 1984, Oerke and Dehne 2004). Hand-weeding is a common practice used to control weeds in aerobic rice in Asian countries, but it is labor-intensive (Roder and

31

Keobulapha 1997). Herbicides are mostly effective in controlling weeds. However, intensive and repeated use of herbicide causes problems of environment pollution and the development of resistant weed biotypes, which have aroused increasing concern (WeedScience.org 2012). To decrease the labor input and use of herbicides in weed management, it is particularly necessary to breed for weed competitiveness in aerobic rice.

(iii) Nutrient availability In lowland rice, standing water benefits the rice crop by its positive effects on soil pH; organic matter buildup; availability of phosphorus (P), iron (Fe), and zinc (Zn); and biological nitrogen fixation (Kirk 2004). In aerobic rice, however, the benefits of flooding largely disappear. For example, the availability of indigenous N (Haefele et al 2008), Fe, and Zn (Sharma et al 2002, Singh et al 2002, Choudhury et al 2007, Kreye et al 2009a) in aerobic soil decreases, even to a deficiency level in some cases. Nutrient deficiencies are indicated to be one cause of yield failure or yield decline after continuous cropping of aerobic rice (Nie et al 2008, Kreye et al 2009 a, b). These studies indicate that nutrient availability is much lower in aerobic soil than in flooded soil. Breeding for high nutrient-use efficiency of aerobic rice may be worthwhile. (iv) Diseases and pests Aerobic soil favors occurrences of some soil-borne diseases and pests such as root-knot nematodes, root aphids, and fungi (Sharma et al 2002, Singh et al 2002). There are indications that root-knot nematodes (Kreye et al 2009a) and fungi (Nishio and Kusano 1973) are biotic causes of yield decline in aerobic rice. Root-knot nematode (Meloidogyne graminicola) has been identified as one of most predominant nematode species associated with rice under rainfed or aerobic conditions in Asian countries such as Laos, Vietnam, Thailand, China, India, Bangladesh, Myanmar, and the Philippines. It stunts crop growth by decreasing plant height, root biomass, and tiller number and may result in a yield loss of up to 75% (Villanueva et al 1992) under aerobic soil conditions, damage that is never observed in irrigated lowland rice. Breeding for resistance to/tolerance of these aerobic-soil-associated diseases and pests of aerobic rice seems necessary. Rice cultivation under aerobic conditions also favors the development of rice blast (caused by Magnaporthe grisea) and bacterial blight (caused by Xanthomonas oryzae pv. oryzae) and breeding program target incorporating resistance against these two diseases. Traits for varietal improvement of aerobic rice In IRRI’s aerobic rice breeding programs targeting tropical regions, drought tolerance, weed competitiveness, disease/pest resistance/tolerance, yield potential, and grain quality are the main breeding goals. Breeding for high nutrient-use efficiency has not yet been incorporated into its breeding strategies. However, given a certain fertilizer rate in a field screening, nutrient-use efficiency of genotypes may be largely reflected by grain yield under aerobic soil conditions. Besides the main breeding goals, the following agronomic and grain quality traits are addressed in selection: nonlodging culm, intermediate plant height, erect plant type, short-medium duration, high head rice, and low chalkiness. (i) Culm A thick and sturdy stem provides strong support to large panicles and contributes to less lodging.Thus, a strong stem is an important trait to consider in selection. (ii) Plant height

32

Tall plants are prone to lodging; the correlation between plant height and lodging before harvest can be as much as 0.52 under favorable aerobic soil conditions (Zhao et al 2010). A study at IRRI with 117 aerobic rice genotypes tested over 4 years shows that, if a plant is taller than 110 cm, the chances of severe lodging increase significantly. However, it is not rare that genotypes with a height of 120 to 125 cm have no such lodging problems (perhaps because their stems are relatively stronger) (Figure 1). In selection, a height of less than 120 cm under favorable aerobic soil conditions is acceptable as long as there is not much lodging. It should be noted that the plant height of a genotype varies with water management and soil fertility. Under water-stressed conditions, a 10-cm decrease in height is often seen (Zhao et al 2010). (iii) Plant type Erect plant type permits a crop to be planted more densely and to use radiation more efficiently. A study by Zhao et al (2006b) with 40 aerobic and upland varieties over 3 years at IRRI shows that all top-yielding aerobic rice genotypes have erect stands, though erect plants do not necessarily mean high yield. They also found that erect plants tend to be more weed-suppressive, though they concluded that fast growth at the vegetative stage rather than plant type is critical to strong weed-suppressive ability. Therefore, an erect plant type is preferred. (iv) Growth duration Varieties with short growth duration are required in some rainfed systems to ensure that the rice crop matures by the end of the rainy season, and in some crop rotation systems to ensure timely planting of an after-rice crop. In these situations, 90- to 100-day varieties are often needed. However, it may be difficult to develop a very early maturing genotype with high yield potential. A study over 4 years at IRRI revealed that 28 best-yielding (>3.5 t ha−1) genotypes out of 117 being tested flower at about 70 to 80 days after seeding; in other words, the high-yielding genotypes have short to medium growth duration, ranging from 100 to 110 days (Figure 2). The study also found that growth duration is negatively correlated with grain yield (Zhao et al 2010). Obviously, long growth duration is not a desired trait of aerobic rice. In breeding, it may be wise not to select a genotype with growth duration longer than 120 days in consideration of both yield potential and farming needs. It should be noted that flowering time is usually delayed more under stress than under nonstress (Zhao et al 2010). The extent of delay is an indication of the drought tolerance level: the more the delay, the less the drought tolerance (Lafitte and Courtois 2002). (v) Grain quality Low chalkiness and high head rice are valued traits for marketing in most cases (chalkiness in parboiled rice is not important), thus they are the most important quality traits to consider in breeding. The selection criterion at IRRI for chalkiness is less than 10%, and for head rice greater than 50%. Different countries/regions have contrasting preferences for other quality traits such as grain shape, amylose content, gelatinization temperature, and gel consistency. Therefore, selection for these quality traits should be region-based.

Breeding for drought tolerance To combine the characteristics of drought tolerance with those of high yield potential in a conventional breeding program, crossing traditional upland varieties with high-yielding modern lowland varieties is usually practiced by breeders, and has proven

33

successful in China and Brazil (Wang et al 2002; Pinheiro et al 2006). Breeders in these two countries have developed a number of japonica aerobic rice varieties using this strategy. In Africa, breeders have also successfully developed some aerobic rice varieties, called NERICA rice, by interspecifically crossing domesticated O. glaberrima with upland O. sativa (Jones et al 1997) and these have been adopted in more than 17 African countries (Harsch 2004). However, empirical breeding practice at IRRI proves that crossing two aerobic rice genotypes (or improved upland rice) or crossing an aerobic rice genotype with a lowland genotype offers a better opportunity to create elite recombinants than crossing a traditional upland rice genotype with a modern lowland one (Zhao et al 2010). A strong evidence is that 22 out of the 26 most elite aerobic genotypes IRRI has developed over the years are derived from crosses of aerobic × aerobic or aerobic × lowland (only four genotypes are derived from two crosses of traditional upland × aerobic, Nok/IR74371-46-1-1 and Thadokkham1/IR74371-46-1-1 (Zhao et al 2010). These aerobic parents used as donors for drought tolerance may not necessarily have as strong a drought tolerance as traditional upland rice varieties, but perhaps because of their more desired agronomic traits of vigor, tillering, and yielding ability (Zhao et al 2006b, 2010),they have a better chance to derive elite recombinants when crossed to another aerobic or lowland genotype. Drought-tolerant donors that are intensively used in IRRI’s aerobic rice breeding programs are presented in Table 1. In addition to conventional breeding, backcross combined with marker-assisted selection (MAS) methods are being employed for introgressing and/or pyramiding major QTLs contributing to drought tolerance into a high-yielding but drought-susceptible background (Li and Gao 2008). IRRI’s Aerobic and Rainfed Rice Breeding Group has identified a major QTL (qDTY12.1) conferring drought tolerance (Bernier et al 2007), and recently introgressed it into Vandana, a popular upland variety from Jharkhand and Odisha provinces of India. The new version of Vandana with introgressed qDTY12.1 shows a yield gain of 0.5 t ha−1 under severe drought. This successful case implies that MAS may be intensively used in developing aerobic rice varieties in the future. A screening protocol is critical for identifying elite recombinants from a large number of progenies. The best strategy of breeding for drought tolerance is to directly select for grain yield under drought stress (Venuprasad et al 2007). Empirical breeding practice in Brazil showed that selection under favorable aerobic soil conditions only, without screening for tolerance of drought stress, resulted in weak tolerance of drought stress (Pinheiro et al 2006). Similarly, Kumar et al (2008) found that, when the reduction in yield under drought is more than 60%, as often occurs in rainfed rice production in the South Asian tropics, selection in favorable environments alone will not result in a yield gain under drought. Zhao et al (2010) studied indirect selection efficiencies of agronomic traits for yield under favorable and severely stressed aerobic soil conditions and found that indirect selection of a trait under one stress level for grain yield is ineffective in the other (Zhao et al 2010). These studies clearly indicate that breeding lines need to be screened under both favorable aerobic soil conditions, referred to as “nonstress,” and severe reproductive-stage drought stress conditions, referred to as “stress.” This two-stress-level screening protocol has been employed by IRRI since it began an aerobic rice breeding program in 2000, and it proved successful in developing aerobic rice varieties (Zhao et al 2010). The two stress levels are described as the following: (i) Nonstress

34

The field is irrigated to saturate the root zone when the soil water tension at 20 cm of soil depth reaches −20 kPa (monitored using tensiometers installed in the field), throughout the growing season. Usually, two to three irrigations per week are needed during the dry season (DS) to keep the aerobic soil under such nonstress conditions at IRRI. (ii) Stress The field is irrigated to saturate the root zone when the soil water tension at 20 cm of soil depth reaches −50 kPa. Usually, one irrigation per week is needed during the DS to achieve this stress level during the reproductive stage. The stress is imposed from 5 to 6 weeks after seeding to maturity. Under such a drought stress, growth and yield are more affected in drought-susceptible lines than in drought-tolerant ones. The two stress levels must be arranged in adjacent fields with equivalent soil fertility. Both fields should not be next to lowland fields to avoid the effect of water seepage. These stress levels are easy to control in the DS because little rainfall occurs in the DS. In the wet season (WS), however, frequent rainfall may result in ponded water in the screening fields and failure to achieve the desired stress levels if genotypes are grown in bunded and leveled fields. Therefore, screening in the WS should be conducted on a slight slope with well-working drains, or ideally in leveled fields with an installed rainout shelter. If it is impossible to realize the stress level in the WS because of rainfall, the stress trial can be conducted in the DS. The two-stress-level screening protocol is used to screen advanced breeding lines tested in replicated trials. Under nonstress conditions, selection is mainly based on grain yield, plant height, seedling vigor (i.e., visually rated crop biomass at the vegetative stage), disease resistance/tolerance, lodging resistance, and growth duration, while under stress conditions selection takes place mainly on grain yield, plant greenness (less senescence or dead leaves at physiological maturity is an indication of more drought tolerance), and flowering time delayed compared with that under nonstress. Lines that perform well under both stress levels are selected. In selection, plot yield measurements within stress levels should be standardized (i.e., divided by their within-trial standard deviation) before analysis for such selection, so that means from nonstress trials, which may be three- to fivefold higher than means in stress trials, do not overwhelm the information from the stress trials; selection on the basis of raw means over stress levels would be heavily weighted in favor of performance under nonstress conditions. Selection on the basis of mean performance over stress levels may not, however, be appropriate in situations in which nonstress and stress trials are conducted in different seasons, say, a nonstress trial in the WS and a stress trial in the DS. In this case, selection is completed in two steps, first on yield and other traits under nonstress and second on yield under stress where only the lines selected in the first step are tested. In screening for drought tolerance in replicated trials, breeding lines need to be grouped based on their growth duration so as to facilitate proper water and fertilizer application management. Genotypes with contrasting duration enter the reproductive stage on different dates after seeding, but stress is imposed on the crop at the panicle initiation stage. Usually, genotypes with growth duration of less than 110 days are put into the “early group” and those with greater than 110 days into the “medium group.” Each group is arranged in a replicated trial. At the pedigree stage (usually F3 and F4 generations), breeding lines are usually screened under nonstressed aerobic soil conditions only for selection on agronomic traits and resistance to diseases in trials without replication (Figure 3).

35

Breeding for weed competitiveness Breeding for weed competitiveness has been neglected for many years since the 1960s. One reason was the “trade-off between weed competitiveness and yield potential,” a conclusion drawn from studies comparing modern lowland rice varieties that are short, fertilizer-responsive, and lodging-resistant but slow-growing at the vegetative stage with traditional varieties that are tall, droopy, fertilizer-unresponsive, and lodging-prone but fast-growing at the vegetative stage (Jennings and Aquino 1968, Jennings and Jesus 1968, Jennings and Herrera 1968). However, later studies with lowland rice (Ni et al 2000, Gibson et al 2003) and upland rice (Garrity et al 1992, Fofana and Rauber 2000) showed that yield potential and weed competitiveness may not be conflicting. More recent studies (Zhao et al 2006a,b) with 40 upland and aerobic rice varieties revealed that weed competitiveness is compatible with yielding ability and heritable, and that it is feasible to genetically improve weed competitiveness and grain yield simultaneously. Two strategies, direct selection and indirect selection, can be used for breeding for weed competitiveness. Direct selection, however, entails growing rice genotypes with weeds and measuring weed biomass as a selection criterion. The measurement is laborious, and usually has a large variance resulting in less reliable selection. Indirect selection is to select for weed competitiveness on crop traits without the presence of weeds. Recent studies showed that seedling vigor, plant height, tillering ability at the vegetative stage, and grain yield under weed-free conditions are closely correlated with weed competitiveness, and that indirect selection on these traits for weed competitiveness is effective (Zhao et al 2006b). By step-wise regression analysis, researchers found that selection on grain yield and seedling vigor at 4 weeks after seeding under weed-free conditions can effectively improve the crop’s yielding ability and weed-suppressive ability simultaneously (Zhao et al 2006b). The two weed-free traits, grain yield (YieldWeed-free) and vigor (VigorWeed-

free), together explained 89% of the variation in yield (YieldWeedy) and 48% in weed biomass (Weed) in weedy environments (eqs. (i) and (ii)): (i) YieldWeedy = 0.01 + YieldWeed-free + 0.06 VigorWeed-free R2 = 0.89 (ii) Weed = 244.84 – 5.92 YieldWeed-free – 10.47 VigorWeed-free R2 = 0.48 The two indirect selection criteria may serve as one selection index with the regression coefficients as their weights. The use of such a selection index, however, would require the harvest of all lines in a pedigree nursery, requiring a great deal of labor. In IRRI’s breeding practice, the method of independent culling levels (Bernardo 2002) is being used for indirect selection for both yielding ability and weed-suppressive ability. Within one season, the first selection is conducted on seedling vigor at 4 weeks after seeding, followed by second selection based on grain yield. In this case, the harvest of entries that exhibit high vigor only is needed, thus saving labor and allowing breeders to manage larger breeding populations. Moreover, since only vigor rating and grain yield measurement are required, and these data can be taken from a relatively small plot (even a single row) for each genotype, selection for weed competitiveness can be started from as early as the F3 generation as F3 seed is usually enough for such small plots. Indirect selection for weed competitiveness is conducted under nonstressed aerobic conditions.

36

With a few years of study with a wide collection of upland and aerobic rice varieties, IRRI has identified a number of genotypes that are vigorous, weed-suppressive, and mostly high-yielding (Table 1). These genotypes can be used as donors in breeding for weed competitiveness. It should be noted that most of these genotypes are also identified as drought-tolerant donors, which makes it easy to breed for drought tolerance and weed competitiveness simultaneously.

Breeding for resistance to/tolerance of nematodes and diseases Despite widely observed damage from root-knot nematode to upland or aerobic rice a long time ago, efforts in breeding for resistance to/tolerance of this pest have rarely been made. We screened 30 rice genotypes (26 aerobic and 4 lowland) in an indoor growth chamber and raised bed both in 2008 at IRRI, and found the following: 1. There is a large genetic variation in resistance to M. graminicola among the tested genotypes (Table 2). The resistance is independent from tolerance and heritable; indicating that genetic improvement in resistance to/tolerance of nematode is feasible. 2. Three O. glaberrima genotypes (TOG 5674, TOG 7235, and CG 14) are highly resistant to M. graminicola, while five O. sativa genotypes (IR78910-23-1-3-4, WAB 638-1, IRAT 216, IR78877-208-B-1-2, and IR72) are partially resistant to M. graminicola. These genotypes are useful donors for developing nematode-resistant aerobic rice varieties. Among these potential donors, CG 14, IR78910-23-1-3-4, and IR78877-208-B-1-2 are relatively high-yielding with or without the presence of the nematode; they thus can be directly used in aerobic rice production, especially in nematode-rich regions. 3. Both indoor growth chamber and raised-bed screening are effective in identifying nematode-resistant genotypes. Raised-bed screening, however, is useful in identifying nematode-tolerant genotypes because it permits a grain harvest from both non-inoculated and inoculated raised beds. Screening for nematode resistance/tolerance can also be performed at hot spots, but field screening requires uniform distribution of nematodes in the entire experimental field to collect reliable data, which is usually difficult to achieve.

With the identified donors ready for use, a separate breeding program for resistance to/tolerance of nematodes can be immediately started. Molecular studies seem to be needed to identify the genes or QTLs responsible for the resistance/tolerance, and to transfer the genes and QTLs into elite aerobic varieties by using marker-assisted selection. Blast and bacterial leaf blight (BB) are two important diseases in aerobic rice. Selection for resistance to such diseases is embedded in IRRI’s breeding procedure as follows: screening F2 populations for disease resistance in a disease nursery in which border plants susceptible to blast are inoculated with blast disease and then transplanting F2 plants showing blast resistance from the disease nursery into a main field under flooded conditions for individual plant selection on plant height, plant type, panicle size, grain shape, growth duration, and grain yield. The F2 populations in the main field are then inoculated with BB inocula for screening for BB tolerance. In order to create a disease environment for screenings afterward, main fields for F3 and F4 nurseries should include susceptible-variety rows inoculated with blast and BB, usually with one susceptible-variety row every 20 rows of plants being tested.

37

Elite aerobic rice genotypes for tropical regions With about 10 years of breeding efforts, IRRI has developed a number of aerobic rice genotypes for tropical regions. After four years of testing (2005 to 2008) under both nonstressed and stressed aerobic soil conditions at the IRRI experimental research station in the Philippines, 26 elite aerobic rice genotypes that yielded 3.4 to 4.3 t ha−1 under moderate fertilized and nonstressed aerobic soil conditions, and 1.2 to 2.0 t ha−1 under severe drought stress at the reproductive stage were identified (Zhao et al 2010). They outyielded Apo, the best first-generation aerobic rice variety developed by IRRI in earlier years, by 10% to 36% under nonstressed conditions and by 12% to 91% under stress. They performed as well as strongly drought-tolerant and upland-adapted varieties IR71524-44-1-1, IR71525-19-1-1, and Vandana under stress, but much better than them in nonstressed aerobic soils. In other words, the 26 genotypes have a higher yield potential and much better drought tolerance than the best first-generation aerobic rice variety Apo. Therefore, we define them as second-generation aerobic rice varieties for tropical regions. These second-generation aerobic rice varieties were vigorous and had a medium plant height (100 to 118 cm under nonstressed conditions), early flowering (69 to 79 days under nonstress), short to medium duration (99 to 113 days under nonstress), and relatively high harvest index (> 0.30 under nonstress). Some of the best lines are IR82098-B-B23-B, IR81423-B-B-111-3, IR81063-B-94-U-3-2, IR79913-B-176-B-4, IR82098-B-B-62-B, IR74371-46-1-1, and IR81063-B-94-U-3-1. IRRI-developed elite aerobic rice genotypes have been disseminated through various platforms or projects to Cambodia, Bangladesh, India, Indonesia, Iran, Iraq, Laos, Malaysia, Nepal, Pakistan, the Philippines, and Vietnam for testing and adoption in water-short irrigated regions or rainfed drought-prone areas. A large-scale and intensive testing of these lines was recently conducted in Bangladesh, India, Nepal, Pakistan, and the Philippines under the ADB-supported project “Developing and disseminating water‐saving rice technologies in South Asia” (ADB RETA 6276), and a number of locally adapted lines have been identified in each country for aerobic, alternate wetting and drying, or rainfed drought-prone ecosystems. Some of them are IR74371-54-1-1 in the Philippines; IR74371-70-1-1 in India; IR74371-46-1-1, IR74371-54-1-1, IR74371-70-1-1, IR79913-B-176-B-4, and IR81449-B-B-109-3 in Nepal; and IR79971-B-149-2-3 in Bangladesh, with a yield of 4−5 t ha−1 under favorable aerobic soil conditions or alternate wetting and drying (Bangladesh). Recently, some of the IRRI-developed second-generation aerobic rice germplasm has been officially released for upland or rainfed ecosystems in several Asian countries (Table 3). This fact further proves that aerobic rice germplasm can be useful not only in irrigated aerobic or alternate wetting and drying systems but also in rainfed upland and rainfed lowland systems. Conclusions Aerobic rice breeding programs at IRRI aim to develop varieties that combine drought-tolerance and weed-competitive traits with traits for high yield to achieve high yield under favorable aerobic soil conditions, and relatively higher yield when experiencing severe drought stress. To select such genotypes, advanced breeding lines need to be screened under both nonstressed and stressed aerobic soil conditions, and only the ones performing well under both stress levels are advanced. Weed competitiveness is selected for under nonstressed and weed-free aerobic conditions by indirect selection on seedling vigor at 4 weeks after seeding and grain

38

yield starting from the F3 generation. Screening for resistance to/tolerance of root-knot nematode M. graminicola is carried out at the F5 in a nursery including two raised beds with and without inoculation of M. graminicola. Screening for resistance to/tolerance of blast and bacterial leaf blight begins in the F2 and continues in F3 and F4 pedigree nurseries. Selection for grain quality is conducted after harvesting F4 plants based mainly on chalkiness and head rice. Other selections for strong culm, intermediate plant height (100 to 125 cm), short to medium growth duration (100 to 120 days), and tolerance of lodging are carried out in every cycle of the breeding process. The aerobic rice breeding program at IRRI has identified several donors for drought tolerance, weed competitiveness, and nematode resistance. IRRI has also developed several elite aerobic rice genotypes that outyielded significantly the best first-generation aerobic rice variety Apo under both nonstressed aerobic soil conditions and severe drought stress. These genotypes have been disseminated to several Asian countries for testing and varietal release. To date, six IRRI-bred aerobic rice lines have been released in five Asian countries for upland or rainfed lowland systems, indicating that this aerobic rice germplasm is useful in water-short irrigated and drought-prone rainfed areas in the tropics. References Atlin GN, Lafitte HR, Tao D, Laza M, Amante M, Courtois B. 2006. Developing rice

cultivars for high-fertility upland systems in the Asian tropics. Field Crops Res. 97:43-52.

Atlin GN, Venuprasad R, Bernier J, Zhao DL, Virk P, Kumar A. 2008. Rice germplasm development for drought-prone environments: progress made in breeding and genetic analysis at the International Rice Research Institute. In: Serraj J, Bennett J, Hardy B, editors. Drought frontiers in rice: crop improvement for increased rainfed production. Singapore: World Scientific Publishing and Los Baños (Philippines): International Rice Research Institute. p 35-59.

Bernardo R. 2002. Selection for multiple traits. In: Bernardo R, editor. Breeding for quantitative traits in plants. Woodbury, Minn. (USA): Stemma Press. p 261-274.

Bernier J, Kumar A, Ramaiah V, Spaner D, Atlin G N. 2007. A large-effect QTL for grain yield under reproductive-stage drought stress in upland rice. Crop Sci. 47:507-518.

Bouman BAM. 2007. The aerobic rice reality. Rice Today 6:38. Bouman BAM. 2001. Water-efficient management strategies in rice production. Int.

Rice Res. Notes 16(2):17-22. Bouman BAM, Wang H, Yang X, Zhao J, Wang C. 2002. Aerobic rice Han Dao: a

new way of growing rice in water-short areas. In: Proceedings of the 12th International Soil Conservation Organization Conference, 26-31 May 2002, Beijing, China. Tsinghua University Press. p 175-181.

Bouman BAM, Peng S, Castañeda AR, Visperas RM. 2005. Yield and water use of irrigated tropical aerobic rice systems. Agric. Water Manage. 74:87-105.

Chaudhury BU, Bouman BAM, Singh AK. 2007. Yield and water productivity of rice–wheat on raised beds at New Delhi, India. Field Crops Res. 100:229-239.

39

De Datta SK, Baltazar AM. 1996. Weed control technology as a component of rice production systems. In: Auld BA, Kim KU, editors. Weed management in rice. FAO Plant Production and Protection Paper 139. Rome: FAO. p 27-52.

De Datta SK, Llagas MA. 1984. Weed problems and weed control in upland rice in tropical Asia. In: An overview of upland rice research. Los Baños (Philippines): International Rice Research Institute. p 321-342.

Dobermann A, Fairhurst T. 2000. Rice ecosystems. In: Rice: nutrient disorders and nutrient management. 1st ed. Los Baños (Philippines): Potash & Phosphate Institute/Potash & Phosphate Institute of Canada, Singapore; International Rice Research Institute. p 2-5.

Fofana B, Rauber R. 2000. Weed suppression ability of upland rice under low-input conditions in West Africa. Weed Res. 40:271-280.

Garrity DP, Movillon M, Moody K. 1992. Differential weed suppression ability in upland rice cultivars. Agron. J. 84:586-591.

Gibson KD, Fischer AJ, Foin TC, Hill JE. 2003. Crop traits related to weed suppression in water-seeded rice (Oryza sativa L.). Weed Sci. 51:87-93.

Harsch E. 2004. Farmers embrace African ‘miracle’ rice: high-yielding ‘Nerica’ varieties to combat hunger and rural poverty. Afr. Recovery 17:10.

Haefele SM, Jabbar SMA, Siopongco JDLC, Tirol-Padre A, Amarante ST, Sta Cruz PC, Cosico WC. 2008. Nitrogen use efficiency in selected rice (Oryza sativa L.) genotypes under different water regimes and nitrogen levels. Field Crops Res. 107:137-146.

Jennings PR, Jesus Jr JD. 1968. Studies on competition in rice. I. Competition in mixtures of varieties. Evolution 22:119-124.

Jennings PR, Herrera RM. 1968. Studies on competition in rice. II. Competition in segregating populations. Evolution 22:332-336.

Jennings PR, Aquino RC. 1968. Studies on competition in rice. III. The mechanism of competition among phenotypes. Evolution. 22:529-542.

Jones MP, Dingkuhn M, Aluko GK, Semon M. 1997. Interspecific O. sativa. L. x O. glaberrima Steud. progenies in upland rice improvement. Euphytica 94:237-246.

Kirk G. 2004. The biochemistry of submerged soils. Chichester (UK): John Wiley and Sons. 291 p.

Kreye C, Bouman BAM, Reversat G, Fernandez L, Vera Cruz C, Elazegui F, Faronilo JE, Llorca L. 2009a. Biotic and abiotic causes of yield failure in tropical aerobic rice. Field Crops Res. 112:97-106.

Kreye C, Bouman BAM, Faronilo JE, Llorca L. 2009b. Causes for soil sickness affecting early plant growth in aerobic rice. Field Crops Res. 114:182-187.

Kumar A, Bernier J, Verulkar S, Lafitte HR, Atlin GN. 2008. Breeding for drought tolerance: direct selection for yield, response to selection and use of drought-tolerant donors in upland and lowland-adapted populations. Field Crops Res. 107:221-231.

Lafitte HR, Courtois B. 2002. Interpreting cultivar x environment interactions for yield in upland rice: assigning value to drought-adaptive traits. Crop Sci. 42:1409-1420.

Li ZK, Gao YM. 2008. Melecular breeding for drought-tolerant rice (Oryza sativa L.): progress and perspectives. In: Serraj J, Bennett J, Hardy B, editors. Drought frontiers in rice: crop improvement for increased rainfed production. Singapore:

40

World Scientific Publishing and Los Baños (Philippines): International Rice Research Institute. p 91-112.

Ni H, Moody K, Robles RP, Paller EC, Lales JS. 2000. Oryza sativa plant traits conferring competitive ability against weeds. Weed Sci. 48:200-204.

Nie L, Peng S, Bouman BAM, Huang J, Cui K, Visperas RM, Xiang J. 2008. Alleviating soil sickness caused by aerobic monocropping: responses of aerobic rice to nutrient supply. Field Crops Res. 107:129-136.

Nishio M, Kusano S. 1973. Fungi associated with roots of continuously cropped upland rice. Soil Sci. Plant Nutr. 19:205-217.

Oerke EC, Dehne HW. 2004. Safeguarding production-losses in major crops and the role of crop protection. Crop Prod. 23:275-285.

O’Toole JC. 1982. Adaptation of rice to drought-prone environments. In: Drought resistance in crops with emphasis on rice. Los Baños (Philippines): International Rice Research Institute. p 195-213.

Pinheiro B da S, Castro E da M de, Guimarães CM. 2006. Sustainability and profitability of aerobic rice production in Brazil. Field Crops Res. 97:34-42.

Roder W, Keobulapha B. 1997. Weeds in slash-and-burn rice fields in northern Laos. Weed Res. 37:111-119.

Saito K, Futakuchi K. 2009. Performance of diverse upland rice cultivars in low and high soil fertility conditions in West Africa. Field Crops Res. 111:243-250.

Saito K, Atlin GN, Linquist B, Phanthaboon K, Shiraiwa T, Horie T. 2007. Performance of traditional and improved upland rice cultivars under nonfertilized and fertilized conditions in northern Laos. Crop Sci. 47:2473-2481.

Sharma PK, Bhushan L, Ladha JK, Naresh RK, Gupta RK, Balasubramanian BV, Bouman BAM. 2002. Crop-water relations in rice-wheat cropping under different tillage systems and water-management practices in a marginally sodic, medium-textured soil. In: Water-wise rice production. Los Baños (Philippines): International Rice Research Institute. p 223-235.

Singh AK, Choudhury BU, Bouman BAM. 2002. Effects of rice establishment methods on crop performance, water use, and mineral nitrogen. In: Water-wise rice production. Los Baños (Philippines): International Rice Research Institute. p 237-246.

Tuong TP, Bouman BAM. 2003. Rice production in water-scarce environments. In: Kijne JW, Barker R, Molden D, editors. Water productivity in agriculture: limits and opportunities for improvement. CABI Publishing. p 53-67.

Venuprasad R, Lafitte HR, Atlin GN. 2007. Response to direct selection for grain yield under drought stress in rice. Crop Sci. 47:285-293.

Villanueva M, Prot JC, Matias M. 1992. Plant parasitic nematodes associated with upland rice in the Philippines. J. Plant Protect. Trop. 9:143-149. Wang HQ, Bouman BAM, Zhao DL, Wang C Moya PF. 2002. Aerobic rice in

northern China: opportunities and challenges. In: Bouman BAM, Hengsdijk H, Hardy B, Bindraban PS, Tuong TP, Ladha, JK, editors. Water-wise rice production. Proceedings of the International Workshop on Water-wise Rice Production, 8-11 April 2002, Los Baños, Philippines. Los Baños (Philippines): International Rice Research Institute. p 143-154.

WeedScience.org. 2012. The international survey of herbicide resistant weeds. www.weedscience.com.

41

Zhao DL, Atlin GN, Amante M, Sta Cruz Ma Teresa, Kumar A. 2010. Developing aerobic rice cultivars for water-short irrigated and drought-prone rainfed areas in the tropics. Crop Sci. 50:2268-2276.

Zhao DL, Atlin GN, Bastiaans L, Spiertz JHJ. 2006a. Cultivar weed-competitiveness in aerobic rice: heritability, correlated traits, and the potential for indirect selection. Crop Sci. 46:372-380.

Zhao DL, Atlin GN, Bastiaans L, Spiertz JHJ. 2006b. Developing selection protocols for weed competitiveness in aerobic rice. Field Crops Res. 97:272-285.

Notes Authors’ address: International Rice Research Institute, Los Baños, Philippines.

Figure 1. Relationship between plant height and lodging of 117 aerobic rice genotypes tested in nonstressed aerobic soil conditions in 2005-08 at IRRI, Philippines.

Figure 2. Relationship between days to flowering and grain yield of 117 aerobic rice genotypes tested in nonstressed aerobic soil conditions in 2005-08 at IRRI, Philippines.

Figure 3. Breeding procedures and selection protocols.

42

Table 1. Donors for drought tolerance and weed competitiveness identified through testing at IRRI.

Variety Drought tolerance level

Weed suppression level

Vigor Plant height

Duration Yield potential

Apo Moderate High High Medium Medium High Aus 196 High High Medium Medium Early Low B6144F-MR-6-0-0

Moderate High High Medium Medium High

C22 Moderate High High Medium Medium High CT 6510-24-1-2 High High High Medium Medium High IR55419-04 Moderate High High Medium Medium Medium IR70358-84-1-1 Moderate High High Medium Early Low IR70360-38-1-B-1 Moderate High Medium Medium Medium Low IR71524-44-1-1 High Low Low Medium Medium Low IR71525-19-1-1 High Low Low Medium Medium Low IR74371-3-1-1 Moderate Moderate Medium Medium Medium High IR74371-46-1-1 Moderate Moderate Medium Medium Medium High IR78878-53-2-2-2 Moderate Moderate Medium Medium Medium High UPL Ri-7 Moderate High High Medium Medium High Vandana High High High Medium Early Low Way Rarem Moderate High High Medium Medium Medium

43

Table 2. Grain yield of rice genotypes with and without the presence of root-knot nematode M. graminicola under aerobic conditions in raised bed, and the nematode population measured at harvest from raised bed, and at 8 weeks after seeding from indoor growth chamber, respectively, 2008, IRRI, Philippines.

Genotype

Ecotype

Species

Grain yield plant−1 Nematode (g−1 roots) without nematode (g)

With nematode (g)

Reduction (%)

Raised bed × 1,000

Growth chamber × 1,000

TOG 5674 Lowland O. glaberrima 0.6 1.0 −69.5 0.3 0.2 TOG 7235 Lowland O. glaberrima 0.7 0.7 0 0.4 0.4 Morobereken Upland O. sativa 3.7 2.4 34.3 0.5 21.9 CG 14 Aerobic O. glaberrima 13.6 9.1* 33.1 1.6 0.2 IR78910-23-1-3-4 Aerobic O. sativa 11.2 9.4 15.7 1.6 6.1 WAB 638-1 Aerobic O. sativa 4.0 1.0 74.4 1.6 6.1 IR60080-46A Aerobic O. sativa 5.2 3.0 41.6 2.1 15.4 IRAT 216 Aerobic O. sativa 2.8 2.7 5.0 2.1 6.0 IR78877-208-B-1-2 Aerobic O. sativa 11.4 11.5 −0.6 2.2 9.2 IR72 Lowland O. sativa 10.6 7.8 26.3 2.3 0.5 IR80508-B-194-3-B Aerobic O. sativa 13.3 8.0** 40.0 2.5 13.2 CT6510-24-1-2 Aerobic O. sativa 7.2 6.8 5.5 2.7 12.5 IR71525-19-1-1 Aerobic O. sativa 9.2 5.0 45.2 2.8 13.2 IR78878-53-2-2-2 Aerobic O. sativa 7.8 5.8 26.2 3.0 22.0 UPLRi-7 Aerobic O. sativa 6.8 5.8 13.7 3.1 12.9 WAB450-24-2-3-P-38-1-HB

Aerobic O. sativa 8.2 7.7 6.5 3.1 12.7

Way Rarem Aerobic O. sativa 12.0 5.1** 57.1 3.1 10.5 B 6144F-MR-6 Aerobic O. sativa 13.1 4.2** 67.8 3.9 21.9 Dinorado Upland O. sativa 6.7 0.8** 87.8 4.0 15.3 Aus 257 Upland O. sativa 10.8 8.2 24.0 4.4 6.4 Apo Aerobic O. sativa 8.2 5.8 28.8 4.6 12.7 WAB 880 SG 42 Aerobic O. sativa 9.6 6.8 29.3 4.6 10.2 Palawan Upland O. sativa 5.3 2.5 51.8 4.9 26.3 Vandana Upland O. sativa 9.1 6.8 25.6 4.9 6.6

IR78877-163-B-1-1 Aerobic O. sativa 8.1 6.2 22.8 5.6 9.8

IR64 Lowland O. sativa 5.5 2.2 60.6 5.7 11.3

IR80508-B-57-3-B Aerobic O. sativa 10.9 9.1 16.0 5.8 14.3

Aus 196 Upland O. sativa 6.8 5.1 24.3 6.5 13.1

Azucena Upland O. sativa 1.9 1.4 26.1 6.9 24.0

UPLRi-5 Aerobic O. sativa 9.4 3.2** 66.2 8.3 17.6

LSD 0.05 4.5 3.4 3.2 10.4

* and ** indicate significant difference between grain yield with and without the presence of nematode at 5% and 1% significance levels, respectively.

44

Table 3. Recently released IRRI-developed aerobic rice lines in Asian countries. Line Parentage Variety Ecosystem Yea

r Country

IR74371-54-1-1 IR55419-4*2/Way Rarem

Sahod Ulan 1

Rainfed lowland

2009

Philippines

IR74371-70-1-1 IR55419-4*2/Way Rarem

Sahbhagi Dhan

Rainfed lowland

2010

India

IR74371-46-1-1 IR55419-4*2/Way Rarem

Sookha dhan 1

Rainfed lowland

2011

Nepal

IR74371-54-1-1 IR55419-4*2/Way Rarem

Sookha dhan 2

Rainfed lowland

2011

Nepal

IR74371-70-1-1 IR55419-4*2/Way Rarem

BRRI dhan 56

Rainfed lowland

2011

Bangladesh

IR74371-70-1-1 IR55419-4*2/Way Rarem

Sookha dhan 3

Rainfed lowland

2011

Nepal

IR79971-B-191-B-B

Vandana/Way Rarem

Inpago LIPI GO1

Upland 2011

Indonesia

IR79971-B-227-B-B

Vandana/Way Rarem

Inpago LIPI GO 2

Upland 2011

Indonesia

IR79913-B-176-B-4

IR55419-04/Way Rarem

Katihan 1 Upland 2011

Philippines

45

Figure 1. Relationship between plant height and lodging of 117 aerobic rice genotypes tested in nonstressed aerobic soil conditions in 2005-08 at IRRI, Philippines.

0

10

20

30

40

50

60

70

80

80 90 100 110 120 130 140

Plant height (cm)

Lodg

ing

(%)

46

Figure 2. Relationship between days to flowering and grain yield of 117 aerobic rice genotypes tested in nonstressed aerobic soil conditions in 2005-08 at IRRI, Philippines.

1.5

2.0

2.5

3.0

3.5

4.0

4.5

50 55 60 65 70 75 80 85 90 95

Days to flowering (d)

Yiel

d (t

ha-1

)

47

Figure 3. Breeding procedures and selection protocols.

Main field

lowland

F4

F3

F6

Pedigree

plant row

aerobic

Pedigree

plant row

aerobic

Nematode

Nursery aerobic

OYT

stress

OYT*

aerobic

AYT

aerobic

AYT

stress

F5

Seedling selection for

blast resistance

Plant selection on plant height, plant type, culm strength, BB resistance, duration, panicle size, grain shape

Line selection on plant height, plant type, culm strength, vigor,

duration, panicle size, grain shape, disease resistance, and aerobic

adaptation. Harvest 3 plants from each selected line.

plants from each selected line

Family selection on plant height, plant type, stem strength, vigor, duration,

panicle size, grain shape, disease resistance, aerobic adaptation, yield,

chalkiness, and head rice

Line selection on

vigor, lodging, yield

for yield potential and

weed competitiveness

Line selection on

yield, greenness, and

flowering delay for

drought tolerance

Line selection for

resistance to/tolerance

of nematode

Line selection on vigor,

lodging, yield for yield

potential and weed

competitiveness

Line selection on yield,

greenness, and flowering

delay for drought tolerance

Lab grain quality

measurement

After harvest

Transplant

* OYT indicates observational yield trial; AYT indicates advanced yield trial.

Disease

nursery

F2

48

Paper 3 Aerobic rice perspectives in India: progress and challenges S.K. Pradhan, A.K. Mall, A. Ghosh, S. Singh, P. Samal, S.K. Dash, O.N. Singh, and A. Kumar

Rice is a semi-aquatic plant and it grows well under lowland flooded anaerobic conditions. The increasing scarcity of water has threatened the sustainability of the irrigated rice production system and hence the food security and livelihood of rice farmers. With the looming problem of water scarcity, farmers and researchers are looking for alternatives to reduce water use in rice production without compromising yield. Aerobic rice is a new production system in which rice is grown in nonpuddled, nonflooded soil with the use of external inputs such as supplementary irrigation and fertilizer with the aim to obtain high yield. The savings in labor input has further changed the perceptions of people toward adopting aerobic rice. Its adoption is facilitated by the availability of efficient herbicides for weed control and suitable rice varieties for aerobic conditions. Yield obtained with aerobic rice varieties varied from 3.5 to 6.0 t ha−1, which is almost double that obtained with upland rainfed varieties, and 25% to 30% less than that obtained with irrigated lowland varieties grown under flooded conditions. A yield decline under aerobic conditions was observed when aerobic rice was continuously grown and the decline was greater in the dry season than in the wet season. To improve aerobic rice yield, the focus should be on combining the drought-tolerance characteristics of upland varieties with the high yield of lowland varieties and growth-limiting factors such as increased availability of N, P, Zn, and Fe; efficient weed control measures; and tolerance of root-knot nematodes. At the Central Rice Research Institute (CRRI), Cuttack, the research carried out under the Asian Development Bank-supported project “Developing and disseminating water-saving rice technologies in South Asia” led to the release of two high-yielding aerobic varieties, CR Dhan 200 and Anagha, for cultivation by farmers, and 27 aerobic rice lines are being tested in the All India Coordinated Rice Improvement Project (AICRIP). Some of these lines are likely to be released in the future as aerobic rice varieties for cultivation in different water-short areas.

Rice is the major food for more than 60% of the population in India. Rice occupies around 44 million hectares of cultivable land and it plays a vital role in the nation’s food security. India’s rice production nearly tripled between 1960 and 2010, with a compound growth rate of 2.53%. India is expected to surpass its demand by 2030 if rice production grows at 1.34% per annum. But, it will have a deficit of around 2.5 million tons if the present growth rate of 1.14% continues up to 2030. Irrigated lowland rice is grown under flooded conditions with the use of one-third of the world’s freshwater supplies. This crop consumes a majority of the freshwater resources of the country. A lot of water (2,295 mm) is used during puddling, transplanting, and irrigation (Ghosh et al 2010). It is estimated that 3,000 to 5,000 liters of water are consumed to produce 1 kg of rice through the transplanted irrigated production system. The water balance of a rice field consists of inflows by irrigation, rainfall, and capillary rise and outflows by transpiration, evaporation, runoff, seepage, and percolation. Water is a scarce commodity and its use is expected to increase in the future at an alarming rate. There will be a 10% to 15% reduction in the water available for agriculture by 2025 due to the ever-increasing demand from nonagricultural sectors such as domestic use, industry, etc. This drastic reduction will have far-reaching consequences for rice production and productivity. In India, groundwater extraction for irrigation has risen by more than

49

100-fold since 1950 and in many states this is leading to a lowering of the water table (Kukal 2004). Groundwater tables have dropped on average by 0.5 to 0.7 m per year in the Indian states of Punjab, Haryana, Rajasthan, Maharashtra, Karnataka, and northern Gujarat, and by about 1 m per year in Tamil Nadu and southern India. The drying up of wells has led to mass migrations and socioeconomic problems. The unsustainable use of irrigation water for rice production is a major socioeconomic, environmental, and health concern for the region. Aerobic rice is grown under a nonflooded and nonpuddled, dry direct-seeded sown condition without sacrificing much potential yield. There is a savings of 40% to 45% water, along with labor, nutrients, and other inputs compared with irrigated transplanted rice. Lowland rice cultivation is the major source of methane (CH4) emissions, contributing 48% of the total greenhouse gases emitted by agricultural sources. The absence of standing water drastically reduces emissions of methane to the atmosphere. Saving water is possible due to the more efficient and better use of soil moisture by the rice crop vis-à-vis flooded conditions. By developing aerobic rice varieties with an efficient root system that enables increased uptake of water and nutrients, rice yield under aerobic conditions can be further increased. Weeds are the greatest yield-limiting constraint to rice when grown under aerobic conditions, contributing about 50% of yield losses, followed by susceptibility to N deficiency, pests, and diseases (Balasubramanian and Hill 2002, Kreye et al 2009). Unpredictable climatic change due to global warming needs a relook at various alternatives to develop aerobic rice varieties with increased tolerance of drought. Aerobic rice is a promising water-saving method that improves crop water-use efficiency, but in the tropics it suffers from low and declining yield. Aerobic rice: a new system of water-saving technology To reduce unnecessary water flows from rice fields, several water-saving technologies have been developed to decrease seepage, percolation, and evaporation. Aerobic rice (Bouman and Tuong 2001), saturated soil culture (Borrell et al 1997), alternate wetting and drying (Li 2001, Tabbal et al 2002), ground-cover systems (Lin et al 2002), and the system of rice intensification (Stoop et al 2002) are some of these technologies. However, except for aerobic rice, in other methods, the fields are still kept flooded for some period, keeping the water requirement and losses high. The cultivation of suitable high-yielding rice varieties in direct-sown, nonpuddled aerobic soils under supplementary irrigation and fertilizer to achieve high yield is a new water-saving technique called aerobic rice. Aerobic rice is responsive to high inputs, can be rainfed or irrigated, and tolerates occasional flooding (Bouman and Toung 2001). The ecosystem for this type of rice is intermediate between upland and irrigated or favorable shallow rainfed lowlands. The area where water is not sufficiently available to grow lowland transplanted rice, but sufficiently available for upland crops, can be used for aerobic rice. Therefore, aerobic rice should combine the features of upland rice (drought tolerance) and modern high-yielding rice varieties of the irrigated ecosystem (Lafitte et al 2002, Atlin et al 2006). Accordingly, vigorous seedlings, rapid biomass development, deep roots with more lateral roots, erect leaves, input responsiveness, lodging resistance, and high harvest index even under moderate stress are the traits that aerobic rice varieties should have. Excessive water is being used for paddy cultivation, which can be minimized through the concept of deficit irrigation. Deficit irrigation suggests the application of

50

less than optimum amounts of water to the crop so that no wastage of water occurs for several reasons. This is expected to save a considerable amount of water drawn from reservoirs and have a savings in electricity or fossil fuel, etc. A water deficit for a crop could occur at a particular stage or at different stages during crop growth. While breeding for drought tolerance, it is a bit difficult to estimate the exact intensity, timing, or severity of stress that a crop could encounter. Thus, the ability to withstand stress at any stage of crop growth would be an invaluable asset to aerobic rice that has to contend with the unpredictability of drought in rainfed areas. Several institutes across the globe have attempted to breed for drought tolerance. Several traits have been found to contribute to enhanced drought tolerance in crop plants. For the rainfed ecosystem, aerobic rice is a concept of growing rice with appropriate genotypes suited for water-short areas. Flooded irrigated rice requires around 40% more water than aerobic rice because of cultivation practices (Bouman 2009) and 48% (570 mm) of the applied irrigation water (1,180 mm) is lost through evapotranspiration (ET). The remaining part of water loss is due to runoff and infiltration. Water represents a major and necessary production cost of rice growers (Brown et al 1978). Rice is a dietary staple food crop for half the world’s population, with annual production of around 400 million tons, for which 1.2 × 1015 L of water are required globally. An increasing proportion of water used in rice cultivation comes from unsustainable groundwater sources as the practice of irrigated dry-season paddy rice cultivation is spreading due to the increased crop yield that results from better regulation of water application and more favorable climatic conditions (e.g., more response of applied inputs, favorable temperature, and higher light intensity) in the dry season. India leads the world in total water withdrawal for irrigation, where irrigation withdrawals represent 80% to 90% of all water use in India. Around 60% of irrigated food production depends on irrigation from groundwater (Shah et al 2000). Groundwater use for irrigation increased almost 113-fold between 1950 and 1984, resulting in a decline in groundwater in 15 states in India (Sampat 2000). Recent research (Adegoke et al 2003, Kabat et al 2004) has identified dramatic changes in local and regional hydrology and weather patterns because of agricultural conversion and expanded crop irrigation. The drying up of wells has led to mass migrations, and a sudden shift of rural-urban population, leading to socioeconomic problems. The unsustainable use of irrigation water for rice production is a major socioeconomic, environmental, and health concern for the country. For this reason, breeding for aerobic rice with low water input should be a priority. Aerobic cultivars have been produced with impressive results in local environments but with lower yield than flooded rice and their productivity tends to decline over time. The reasons for this decline in yield are still unclear, but are likely to be related to altered plant growth due to root-shoot signaling, impaired ability to acquire nutrients, or a greater impact of biological constraints such as weeds and nematodes. Population growth and climate change are likely to increase the pressure on water resources, while the latter will make drought and heat stress more problematic for rice production. The global challenge of sustainable rice cultivation therefore requires the amount of water used for irrigated rice to be reduced while maintaining yield and production without having any negative effect on soils properties. Progress in the development of aerobic rice varieties

51

Breeding for aerobic rice started systematically in 2007 by using first-generation aerobic rice germplasm and other exotic materials as donors, hybridized with popular varieties under the Asian Development Bank (ADB)-supported project “Developing and disseminating water-saving rice technologies in South Asia.” Large variability for aerobic traits was generated with several populations at CRRI, Cuttack, and subsequently promising genotypes were selected using pedigree breeding methods. In addition, many segregating populations, fixed lines, and varieties were introduced from IRRI, Philippines, to select superior lines in Cuttack. In a short time, promising lines were identified and evaluated under the AICRIP in India. Based on the grain yield, variety Apo, earlier released in the Philippines, was found suitable in Odisha, and it was released as CR Dhan 200 (CR 2624-IR55423-01; IET 21214) by the “Orissa State Sub-Committee on Crop Standards, Notification, and Release of Varieties” during 2012. At the University of Agricultural Sciences (UAS), Bangalore, research on developing aerobic rice varieties was in progress and Anagha (ARB 6) was released for cultivation in aerobic environments in Karnataka, India. At CRRI, promising aerobic lines were nominated starting in 2007 to the national AICRIP trials for evaluation of materials at various target locations in different states of the country. Anagha, an elite line for aerobic conditions, exhibits a fair degree of drought tolerance and productivity. This line, with the popular name of Anagha, is a derivative of Budda (a local aus accession) and IR64, a mega-variety from IRRI, Philippines. Pedigree breeding was followed along with farmers’ participatory varietal selection for selection of the variety. The process of development started with hybridization in 2000 and many stable lines were named as BI lines (Budda/IR64) in station trials and ARB (Aerobic Rice Bangalore) in national and international coordinated trials. Anagha can be grown just like maize and is a boon to farmers who face the challenge of diminishing water resources. After direct seeding, irrigation should be given at intervals of 5 to 7 days throughout the season. Irrigation can be skipped in the event of rainfall. Six derivatives of Budda/IR64 lines were found good for aerobic conditions. These lines were almost fixed for various traits and were uniform with respect to phenotype during evaluation in station, multilocation, and national trials between the 2005 and 2008 wet seasons. The lines have manifested a very high degree of drought tolerance at rainfed severely stressed sites such as Raipur and Hazaribagh. The grain yield potential of these lines is 7.0 t ha−1 on average in station trials at UAS, Bangalore. The average yield in farmers’ fields was from 3.0 to 5.0 t ha−1 (Shashidhar 2012). CR Dhan 200 was entered into the AICRIP testing program as IET 21214 during the 2008 WS (wet season) in IVT-IME-Aerobic. It consistently outperformed the national, regional, and local checks under national testing from 2008 to 2010. This promising line, after 3 years of national testing, exhibited stable yield and other desirable traits in zones II, III, IV, and V of the country. It is found to be promising in Chhattisgarh and Jharkhand states under aerobic conditions (zone III). The elite line is promising in Tamil Nadu in zone V. It is promising in Punjab in zone II and in Gujarat in zone IV. It exhibited 13.2% and 17−20% higher grain yield over national and regional checks pooled over 3 years of testing across different zones in the country. It recorded a yield advantage of 9.0%, 7.7%, and 14.6% over national, regional, and local check varieties, respectively, in zone III; 9.9% over the national check and 20.4% over the regional check in zone V; 26.0% over the national check and 5.8% over the regional check in zone II; and 9.7% over the national check and

52

18.2% over the regional check in zone IV during 3 years of testing. Maturity duration of the culture is 115 to 120 days. It has semidwarf plant height (under aerobic conditions) and intermediate height in irrigated areas, and nonlodging plant type with an average of 272 panicles m−2. It has short bold grain, moderate tillering (7 to 10 tillers), and 1,000-grain weight of 24 g. The line is moderately resistant to leaf blast, neck blast, brown spot, whorl maggot, gall midge biotype 6, and leaffolder attack. The variety has a high response to fertilizer application compared with check varieties. It possesses good hulling, good milling, white kernel, medium slender grain, no grain chalkiness, and desirable alkali spreading value (AICRIP 2008, 2009, 2010). (i) Aerobic rice lines of mid-early duration For the upland and midland toposequence of northeastern India, including Assam, which is considered favorable upland area, as well as for the shallow lowland ecosystem of eastern India, where drought is a severe constraint, aerobic breeding lines of 110 to 120 days’ duration can be grown to replace the presently grown varieties. In on-station trials at CRRI-Cuttack and on-farm advanced yield trials for early-duration aerobic rice lines carried out in Tangi-Chowduar block of Cuttack District, genotypes CR Dhan 200 and B-6144F-MR-6-0-0 showed a stable performance over years (Table 1). In addition, IR78875-131-B-1-4, IR74371-3-1-1, and IR80021-B-86-3-4 yielded more than 4.0 t ha−1 under aerobic conditions. Among the three checks, Anjali was superior in yield to Annada and Naveen. (ii) Aerobic rice varieties for medium-late duration For most of the irrigated areas in Punjab, Haryana, and Andhra Pradesh as well as for the rainfed lowland ecosystem in eastern India, high-yielding aerobic rice varieties of 125 to 135 days’ duration will be appropriate to replace the existing cultivated varieties. Under on-farm trials (Kochila Nuagaon of Cuttack district), there was a decrease in yield compared with on-station yield. However, three lines (IR70213-10-CPA-4-2-3-2, CR 749-20-2-16-1, and CR 691-475) exhibited significantly higher yield (> 4.5 t ha−1) under aerobic conditions in on-station experiments (Table 2). Among three checks, Naveen was superior in yield to Shatabdi and IR36. (iii) Tolerance of aerobic rice genotypes of nematodes Root-knot nematode (Meloidogyne graminicola) infestation in rice roots is one of the causes of a yield decline in aerobic conditions. Aerobic rice lines were evaluated for tolerance against root-knot nematode in the net house at the Division of Plant Protection, CRRI, Cuttack, during the wet season of 2011. Two lines (IR84895-B-127-28-1-1-1 and IR84899-B-183-6-1-1-3) were observed to be highly tolerant while 27 lines showed moderate tolerance of root-knot nematode infestation (Table 3). Root gall caused by root-knot nematode was also sporadic. Spiral nematode (Helicotylenchus crenacauda) populations were recorded to be nearer the threshhold limit (1 nematode g−1 soil). (iv) Disease resistance of aerobic rice genotypes Rice cultivation under aerobic conditions appears to favor the infestation of rice blast (caused by Magnaporthe grisea) and tolerance of rice blast is an important component of aerobic rice breeding. Bacterial blight (caused by Xanthomonas oryzae pv. oryzae) is another important disease infecting rice cultivars in both irrigated and rainfed ecosystems. Aerobic rice breeding lines were evaluated for tolerance of blast and bacterial leaf blight. Thirty-four genotypes were found to be tolerant, with a score of 1 (Table 4), and 87 genotypes showed a score of 3 for blast

53

disease. For bacterial blight, many entries were found to be moderately resistant (score of 5). Challenges for aerobic rice production The following are the challenges to be focused on through research for the adoption of aerobic rice technology in India: (i) Yield decline Results at CRRI, Cuttack, showed that aerobic conditions significantly inhibited the structural development of roots and caused significant variation in biochemical root traits, accounting for a higher concentration of hydrogen peroxide (24.6%) and proline (20%) and a lower concentration of total soluble protein (20%), which resulted in a 17% yield decline over semi-aerobic conditions. Applying supplementary irrigation in semi-aerobic conditions achieved 21% more grain yield than in aerobic conditions because of 21%, 8.3%, and 10.4% more root biomass, root volume, and root/shoot ratio, respectively. The interaction of genotypes with soil water conditions revealed that, in aerobic conditions, better performance could be expected in CR Dhan 200 (4.0 t ha–1) and IR74371-3-1-1 (3.80 t ha–1) with less yield decline (7.0 to 9.5%); supplementary irrigation enhanced their grain yield (4.3 to 4.4 t ha–1) in semi-aerobic conditions. Therefore, a detrimental impact of physical and biochemical root traits leading to a yield decline in aerobic rice could be alleviated with supplementary irrigation at the critical growth stages in semi-aerobic conditions without compromising water productivity (Ghosh et al 2012). (ii) Suitable varieties Most of the varieties that farmers currently cultivate are not suitable for cultivation in aerobic conditions. There is a need to breed varieties for maximum water input-use efficiency of the plant along with less yield penalty in the aerobic ecosystem. A shift from continuously flooded to aerobic conditions may have profound effects on sustainability (e.g., soil-borne pests and weed dynamics) and environmental parameters (e.g., nitrate leaching and herbicide use, etc.). To make aerobic rice successful, suitable site-specific varieties should be developed. More research is needed to synthesize suitable varieties with wider adaptation to suit the aerobic rice ecosystem along with knowledge on crop-soil-water management recommendations.

(iii) Weeds Weeds are perceived to be the most severe constraint to aerobic rice production. Manual weeding is highly labor-intensive. Herbicides have been proven effective in many cases but intensive use may cause environmental contamination and the development of herbicide resistance. The use of weed-competitive varieties to suppress weeds might substantially reduce herbicide use and labor cost. Weed competitiveness, early vigor (a visual seedling biomass rating), and yield under both weedy and weed-free conditions of aerobic rice recorded a strong association (Zhao et al 2006). Breeding aerobic rice cultivars combining high yield, early vigor, and strong weed competitiveness is therefore critical for aerobic rice systems. Moreover, the adoption of weed-competitive cultivars will decrease environmental pollution and the development of herbicide-resistant biotypes by reducing herbicide application. Weed-competitive cultivars are reported to be a low-cost and safe tool for integrated weed management (Pester et al 1999, Fischer et al 2001, Gibson and Fischer 2004). (iv) Nutrient dynamics

54

Although flooding has beneficial effects on soil acidity (pH), soil organic matter buildup, availability of phosphorus, iron, and biological nitrogen fixation, etc., these beneficial effects may decrease with aerobic rice cultivation. Major nutrient and micronutrient dynamics and their bioavailability are expected to change due to a shift to aerobic cultivation. Since the concept of aerobic rice is new, relatively few insights exist into water, N, P, Fe, and Zn dynamics and their interactions. There are reports of soil sickness after some years of aerobic rice cultivation. Hence, a thorough understanding of these processes would lead to the development of sustainable management of irrigation and nutrients to optimize yield and resource-use efficiency. Conclusions Aerobic rice is an important water-saving technology for both irrigated and rainfed conditions. Research at CRRI, Cuttack, has identified several promising aerobic rice lines. Two aerobic rice varieties, CR Dhan 200 and Anagha, have been released for cultivation by farmers and several aerobic rice lines are in the pipeline. Although, among all water-saving rice technologies, aerobic rice provides the highest yield per unit of water used, yield under aerobic conditions is still lower than in flooded irrigated conditions. Further, the cultivation of aerobic rice continuously leads to a decline in productivity with time. The reasons for this are still unclear, but are likely to be related to altered plant growth due to root-shoot signaling, impaired ability to acquire nutrients, or a greater impact of biological constraints (particularly weeds and nematodes). Questions on the sustainability of aerobic rice production arise due to the likely depletion of soil nutrients, the role of biotic factors in the yield decline, the identification of root architecture that maximizes water and nutrient acquisition, etc. Growing rice with water-saving techniques such as aerobic rice has great potential in India in general and in eastern India in particular as there is a looming water crisis. This situation requires the development of aerobic rice varieties with higher potential than those presently available to disseminate this technology in areas with high yield potential. The yield decline due to continuous cultivation of aerobic rice and nutrient dynamics under this system has to be investigated extensively.

References Adegoke JO, Pielke RA, Eastman J, Mahmood R, Hubbard KG. 2003. Impact of

irrigation on midsummer surface fluxes and temperature under dry synoptic conditions: a regional atmospheric model study of the U.S. High Plains. Monthly Weather Rev. 131:556-564.

AICRIP (All India Coordinated Rice Improvement Project). 2008. Progress report 2008, volume 1: varietal improvement. Directorate of Rice Research, Rajendranagar, Hyderabad, India.

AICRIP (All India Coordinated Rice Improvement Project). 2009. Progress report 2009, volume 1: varietal improvement. Directorate of Rice Research, Rajendranagar, Hyderabad, India.

AICRIP (All India Coordinated Rice Improvement Project). 2010. Progress report 2010, volume 1: varietal improvement. Directorate of Rice Research, Rajendranagar, Hyderabad, India.

55

Atlin GN, Lafitte HR, Tao D, Laza M, Amante M, Coutois B. 2006. Developing rice cultivars for high fertility upland systems in the Asian tropics. Field Crops Res. 97:43-52.

Balasubramanian V, Hill JE. 2002. Direct seeding of rice in Asia: emerging issues and strategic research needs for the 21st century. In: Tuong TP, Lopez K, Hardy B, editors. Direct seeding: research strategies and opportunities. Los Baños (Philippines): International Rice Research Institute. p 15-42.

Borrell A, Garside A, Fukai S. 1997. Improving efficiency of water use for irrigated rice in a semi arid tropical environment. Field Crops Res. 52:231-248.

Bouman BAM, Tuong TP 2001. Field water management to save water and increase its productivity in irrigated rice. Agric. Water Manage. 49:11-30.

Bouman BAM, Kropff MJ, Tuong TP, Wopereis MCS, ten Berge HFM, van Laar HH. 2001. ORYZA2000: modeling lowland rice. Los Baños (Philippines): International Rice Research Institute, and Wageningen University and Research Centre. 235 p.

Bouman BAM. 2009. How much water does rice use? Rice Today 8(1):28-29. Brown KW, Turner FT, Thomas JC, Deuel LE, Keener ME. 1978. Water balance of

flooded rice paddies. Agric. Water Manage. 1:277-291. Fischer AJ, Ramirez H, Gibson KD, Pinheiro BDS. 2001. Competitiveness of semi

dwarf upland rice cultivars against palisade grass (Brachiaria brizantha) and signal grass (Brachiaria decumbens). Agron. J. 93:967-973.

Ghosh A, Dey R, Singh ON. 2012. Improved management alleviating impact of water stress on yield decline of tropical aerobic rice. Agron. J. 104:584-588.

Ghosh A, Singh ON, Rao KS. 2010. Improving irrigation management in dry season rice cultivation for optimum crop and water productivity in non traditional rice ecologies. Irrig. Drainage 60(2):174-178.

Gibson KD, Fischer AJ. 2004. Competitiveness of rice cultivars as a tool for crop-based weed management. In: Inderjit, editor. Weed biology and management. Dordrecht (Netherlands): Kluwer Academic Pulishers. p 517-537.

Kabat P, Claussen M, Dirmeyer PA., Gash JHC, Bravo de Guenni L, Meybeck M. Pielke R. Vorosmarty CJ, Hutjes RWA, Lutkemeier S. 2004. Vegetation, water, humans and the climate: a new perspective on an interactive system. Springer. 566 p.

Kreye C, Bouman BAM, Reversat G, Fernandez L, Vera Cruz C, Elazegui F, Faronilo JE, Llorca L. 2009. Biotic and abiotic causes of yield failure in tropical aerobic rice. Field Crops Res. 112:97-106.

Kukal SS. 2004. Water-saving irrigation scheduling to rice (Oryza sativa) in Indo-Gangetic plains of India. In: Huang G, Pereira LS, editors. Land and water management: decision tools and practices. vols 1 and 2. China Agricultural University, Beijing. p 83-87.

Lafitte RH, Courtois B, Arraudeau M. 2002. Genetic improvement of rice in aerobic systems: progress from yield to gene. Field Crops Res. 75:171-190.

Li Y. 2001. Research and practice of water-saving irrigation for rice in China. In: Barker R, Li Y, Tuong TP, editors. Water-saving irrigation for rice. Proceedings of the International Workshop, 23-25 March 2001, Wuhan, China. Colombo (Sri Lanka): International Water Management Institute. p 135-144.

56

Lin S, Dittert K, Sattelmacher B. 2002. The Ground Cover Rice Production System (GCRPS): a successful new approach to save water and increase nitrogen fertilizer efficiency? In: Bouman BAM, Hengsdijk H, Hardy B, Bindraban PS, Tuong TP, Ladha JK, editors. Water-wise Rice Production. Proceedings of the International Workshop on Water-wise Rice Production, 8-11 April 2002, Los Baños, Philippines. Los Baños (Philippines): International Rice Research Institute. p 187-196.

Pester TA, Burnside OC, Orf JH. 1999. Increasing crop competitiveness to weeds through crop breeding. J. Crop Prod. 2:59-76.

Sampat P. 2000. Deep trouble: the hidden threat of groundwater pollution. In: Peterson J, editor. World Watch Institute, Washington, D.C. 2000.Valuing groundwater, Natl. Acad. Press, Washington, D.C. p 10-14.

Shah T, Molden D, Sakthivadivel R, Seckler D. 2000. The global groundwater situation: overview of opportunity and challenges. Colombo (Sri Lanka): International Water Management Institute.

Shashidhar, HE 2012. An eco-friendly aerobic rice BI 33 (Anagha). Available at www.aerobicrice.in.

Stoop W, Uphoff N, Kassam A. 2002. A review of agricultural research issues raised by the system of rice intensification (SRI) from Madagascar: opportunities for improving farming systems for resource-poor farmers. Agric. Syst. 71:249-274.

Tabbal DF, Bouman BAM, Bhuiyan SI, Sibayan EB, Sattar MA 2002. On-farm strategies for reducing water input in irrigated rice: case studies in the Philippines. Agric. Water Manage. 56:93-112.

Zhao DL, Atlin GN, Bastiaans L, Spiertz JHJ. 2006. Cultivar weed-competitiveness in aerobic rice: heritability, correlated traits, and the potential for indirect selection. Crop Sci. 46:372-380.

Notes Authors’ addresses: S.K. Pradhan, A.K. Mall, A. Ghosh, S. Singh, P. Samal, S.K.

Dash, and O.N. Singh, Central Rice Research Institute (CRRI), Cuttack 753 006, Odisha, India; A. Kumar, International Rice Research Institute (IRRI), Los Baños, Philippines.

57

Table 1. Grain yield (t ha−1) of aerobic rice lines of mid-early duration at CRRI, Cuttack (on-station and on-farm), in farmers’ fields.

Designation On-station

2007 WS On-station 2008 DS

On-station 2008 WS

On-station pooled

On-farm 2008 WS

CR Dhan 200 4.60 5.13 4.50 4.74 4.03 B-6144F-MR-6-0-0

4.47 4.90 4.37 4.58 4.07

IR72667-16-B-B-3 3.33 4.30 3.63 3.75 2.91 IR78875-131-B-1-1

3.63 4.13 3.70 3.82 3.21

IR78875-131-B-1-4

4.37 4.17 4.75 4.43 3.95

IR55419-04 3.90 4.30 3.70 3.97 3.35 IR78877-181-B-1-2

3.87 3.97 3.77 3.87 3.18

IR78875-53-2-2-2 3.70 4.07 3.90 3.82 3.13 IR78875-53-2-2-4 4.00 4.07 3.77 3.95 3.58 IR79899-B-179-2-3

4.03 4.17 3.50 3.90 3.26

IR79906-B-192-2-1

3.77 3.90 3.97 3.88 3.18

IR79906-3-192-2-3

3.63 4.03 3.73 3.80 3.01

IR79956-B-60-2-3 3.83 4.13 3.53 3.83 3.18 IR79906-B-5-3-3 3.87 4.10 3.87 3.95 3.05 IR80013-B-141-4 3.83 4.23 3.60 3.89 3.16 IR80312-6-B-2-B 3.63 4.27 3.73 3.88 3.11 IR74371-3-1-1 4.17 4.67 4.20 4.35 3.65 IR80021-B-86-3-4 4.67 4.03 3.67 4.12 3.95 JD 12 3.70 3.28 3.73 3.57 3.18 WR 3-2-6-1 3.73 4.27 3.60 3.87 3.21 CR 691-58 3.63 3.83 3.87 3.78 3.05 VLD 16 3.80 3.33 3.73 3.62 3.16 Anjali (check 1) 3.50 3.63 3.50 3.54 3.01 Annada (check 2) 2.55 3.13 2.87 2.85 2.14 Naveen (check 3) 2.63 2.17 2.77 2.52 2.15 Mean 3.79 4.01 3.75 3.85 3.23 CV (%) 9.65 7.81 7.21 − 12.3 LSD0.05 0.41 0.32 0.26 − 0.39

WS = wet season, DS = dry season.

58

Table 2. Grain yield (in t ha−1) of aerobic genotypes of medium duration on-station (CRRI, Cuttack) and on-farm.

Designation 2007 WS 2008

DS 2008 WS On-

station pooled

On-farm 2008 WS

IR70213-10-CPA-4-2-3-2 5.13 5.87 5.50 5.50 4.20 IR71700-247-1-1-2 4.12 5.45 4.40 4.66 3.29 IR72176-140-1-2-2-3 4.12 4.47 3.53 4.04 3.09 IR77080-B-4-2-2 4.03 3.90 3.13 3.69 2.94 IR77298-14-1-2 3.90 3.93 5.33 4.39 3.14 IR78877-208-B-1-4 3.53 3.30 4.53 3.79 2.82 IR70215-65-CPA-2-UBN-2-B-1-1

3.83 3.37 3.40 3.53 3.12

IR72875-94-3-3-2 4.33 3.67 3.80 3.93 3.40 IR74963-262-5-1-3-3 4.27 4.17 3.63 4.02 3.67 PTB 39 4.33 4.16 3.40 3.97 3.59 GR 7 4.17 4.15 4.27 4.20 3.27 CR 749-20-2-18-15 3.53 3.33 3.67 3.51 2.89 CR 749-20-2-16-1 4.77 4.33 4.73 4.61 3.92 CR 749-20-2-15-5 3.67 3.60 3.50 3.59 2.87 CR 681-380 3.80 3.33 3.90 3.68 2.92 CR 691-475 5.41 4.90 5.53 5.28 4.52 Daya 4.17 4.00 3.43 3.87 3.27 CRK -9 3.80 3.33 4.37 3.83 2.87 CRK -26 3.67 3.23 3.27 3.39 3.07 CRK 131-243-1 3.83 3.27 3.27 3.46 3.29 CR 2340-3 4.67 3.80 3.40 3.96 3.64 PTB-45 3.87 3.80 3.70 3.79 4.05 Naveen (check 1) 3.34 3.07 3.19 3.20 3.07 Satabdi (check 2) 3.06 3.27 3.08 3.14 3.02 IR36 (check 3) 3.21 3.20 3.17 3.19 2.74 Mean 4.02 3.88 3.89 3.93 3.31 CV (%) 9.89 8.39 9.15 − 11.68 LSD0.05 0.18 0.14 0.16 − 0.32 WS = wet season, DS = dry season.

59

Table 3. Aerobic rice varieties tolerant of M. graminicola. Variety GI R/S IR84895-B-127-28-1-1-1 0.6 R IR84899-B-183-6-1-1-3 0.8 R IR84882-B-B-123-46-1-1-1 1.6 MR IR84898-B-168-15-1-1-1 1.0 MR IR84899-B-183-1-1-1-1 1.6 MR IR84882-B-121-5-1-1-1 1.3 MR IR84887-B-157-27-1-1-1 1.6 MR IR84896-B-159-11-1-1-1 1.3 MR IR84887-B-53-33-1-1-1 1.0 MR IR84887-B-153-33-1-1-3 1.6 MR IR84899-B-182-2-1-1-1 1.0 MR IR84899-B-182-3-1-1-1 1.3 MR IR84899-B-183-6-1-1-4 1.3 MR IR84899-B-184-16-1-1-1 1.3 MR IR84899-B-185-8-1-1-3 1.3 MR IR84899-B-185-9-1-1-1 1.6 MR IR84898-B-185-15-1-1-1 1.6 MR IR83920-B-B-277-2-1-1-2 1.0 MR IR83927-B-B-279-3-1-1-3 1.6 MR IR83929-B-B-291-2-1-1-1 1.6 MR IR83929-B-B-291-3-1-1-1 1.3 MR IR83929-B-B-291-3-1-1-2 1.3 MR IR83929-B-B-291-4-1-1-3 1.3 MR IR83377-B-B-83-1 1.5 MR IR83376-B-B-85-1 1.3 MR IR84895-B-125-12-1-1-1 1.0 MR IR84899-B-179-1-1-1-3 1.0 MR IR84899-B-179-1-1-1-4 1.0 MR IR84899-B-183-13-1-1-2 1.0 MR R = resistant, S = susceptible, MR = moderately resistant, GI = gall infestation.

60

Table 4. Aerobic rice lines showing high tolerance (score of 1) of blast disease.

S. N Designation 1 IR84898-B-184-19-1-1 2 IR84898-B-185-10-1-1 3 IR84898-B-185-15-1-1 4 IR84898-B-185-16-1-1 5 IR83929-B-B-291-3-1-1 6 IR83927-B-B-279-3-1-1 7 IR84899-B-185-22-1-1 8 IR84895-B-124-46-1-1 9 IR84887-B-155-40-1-1 10 IR77298-14-1-2-130-3 11 IR84899-B-181-8-1-1 12 IR83927-B-B-278-5-1-1 13 IR84895-B-125-22-1-1 14 IR84895-B-183-1-1-1 15 IR84896-159-12-1-1 16 IR84887-B-156-17-1-1-1 17 IR84887-B-156-31-1-1-2 18 IR84887-B-157-34-1-1-1 19 IR84899-B-179-13-1-1-1 20 IR84899-B-179-20-1-1-1 21 IR84899-B-183-6-1-1-3 22 IR84899-B-184-16-1-1-3 23 IR84899-B-185-8-1-1-3 24 IR84899-B-185-10-1-1-1 25 IR83377-B-B-100-3-48-1 26 IR83372-B-B-94-4-62-3 27 IR83376-B-B-143-4-76-1 28 IR83377-B-B-85-4-83-3 29 IR83752-B-B-12-3-134-1 30 IR83383-B-B-129-1-127-2 31 IR82589-B-B-124-2-190-1 32 IR83376-B-B-10-4-24-4 33 IR83383-B-B-129-4-78-1 34 IR84887-B-154-13-1-1

61

Paper 4 Prospects of aerobic rice in water-limited bunded uplands and shallow lowlands of eastern India

N.P. Mandal, M. Variar, and A. Kumar

Input-responsive aerobic rice grown in nonflooded, direct-seeded conditions has the potential to substantially reduce water use in rice cultivation. Because of erratic distribution of rain, eastern India shows a low and unstable productivity pattern. This situation warrants appropriate technological and policy interventions to properly manage the risk of less water availability for rice production, or else its continuation will endanger food security at large. Reduced labor cost, relative insulation from erratic rainfall, better opportunities for mechanized seeding and crop management, and possibilities of early harvest leading to the successful establishment of a sequence crop with residual moisture are added advantages of the aerobic system even though relatively higher weed intensity under nonpuddled conditions necessitates a slightly higher investment in weed management. Genotypes with duration ranging from 105 to 115 days, early vigor, and lodging resistance with intermediate plant stature and superior agronomic performance under moderate water stress have proved effective in harnessing the productive potential of the water-limited environment. Network evaluation of a large number of aerobic adapted breeding lines in eastern India during the last 5 years led to the identification and release of Sahbhagi dhan for Jharkhand and Odisha. With the advent of varieties such as Sahbhagi dhan and other shorter-duration varieties having the ability to provide high yield under water-short conditions, the increase in direct-seeded aerobic rice area that will replace the high-water-demanding, transplanted flooded system of rice cultivation looks to be an upcoming reality. With additional efforts for mechanized crop establishment and weed management, farmers would benefit from higher returns for their investment.

Rice is life for almost half of the global population and it is the backbone of the food and livelihood security system (FAO 2009).It is the staple food of more than 50% of the world’s population, especially for most of the people of Southeast Asia. In India, rice is the most important cereal food crop, occupying 23.3% of the gross cropped area of the country. It plays a vital role in the national food grain supply. Rice contributes 43% of India’s total food grain production and 46% of its total cereal production. It is an important staple food for about 65% of the population and means of income and employment generation for more than 50 million households. Of the total supply of 2,394 kcal per person per day energy in the average Indian diet, more than 75% comes from rice (FAO 2006).The total rice area in the country is about 42.3 million hectares. Of this, 26.8 million hectares (63.3%) are in the eastern Indian states of West Bengal, Bihar, Assam, eastern Uttar Pradesh, Chhattisgarh, Jharkhand, and Odisha (Singh and Hossain 2000).In 2011-12, for the first time, India produced 102 million tons of rice. The biggest change has come from eastern India, starting from eastern Uttar Pradesh, Bihar, Jharkhand, West Bengal, Assam, and Odisha. In eastern India, rainfed rice is grown on approximately 16.2 million ha consisting of both rainfed upland and lowland (Singh and Singh 2000). Eastern India faces two major challenges, water shortage and food in security, because of its rain-dependent agriculture, where a significant proportion of the population is still in absolute poverty. Although the area receives approximately 1,000 mm of rainfall during the monsoon season, erratic distribution and shortages at the crop’s

62

reproductive stage (Widawsky and O’Toole 1990) strongly decrease productivity. The region has been classified as physically and economically “water scarce” and it faces water deficits that limit food production (Serageldin 2004). Water scarcity in this region frequently affects about 1.6 million ha and the estimated average loss in rice production during years with less rainfall for eastern India is 5.4 million tons (Pandey et al 2007). Eastern India has a complex distribution of different ecosystems, mainly dominated by rainfed lowland and considerable upland area. For the rainfed lowland ecosystem, about 10.6 million ha have shallow water depth (0 to 30 cm), of which 5.8 million ha are prone to water deficit (Singh and Hossain 2000). A unique feature of the rainfed lowland system is the alternation between anaerobic and aerobic soil conditions, which has significant consequences for root growth, nutrient availability, and weed competition (Wade et al 1998).

With the rapid climate change affecting rainfall patterns, making them more erratic on a global scale, most of the time transplanting is not feasible in this land situation. Out of 6.0 million ha under rainfed upland, about 2 million ha are favorable, where varieties of 105 to 110 days’ duration can be grown. The distribution of fields within the subecosystem in different topography leads to differences in overall agronomic management and types of varieties to be grown. Irrigation infrastructure in eastern India is very poor, at 21.3%, and, in the absence of fresh investments, the rainfed ecosystem will remain important in the near future. Crop diversification is also not very successful in these land situations because, in a normal rainfall year, water stagnation in the root zone can severely damage the diversified cereal and pulse crops (maize, cowpea, soybean, etc.).Traditional rice cultivation in lowland through transplanting is quite labor intensive because of poor mechanization. Eastern India had surplus human labor engaged in rice cultivation earlier. However, with the changes in socioeconomic conditions, the introduction of many government social welfare schemes, and job availability in rising industries, human labor is becoming scarce and costly for profitable rice cultivation. To cope with this looming water and labor shortage and resultant unstable productivity in rainfed rice, monetary losses, social insecurity, and its ramifications for poor farm families, changes in cultivation practices are needed that address ubiquitous abiotic stress factors. Cultivation of rice aerobically can be a profitable alternative to counter the challenges posed by the erratic behavior of rainfall patterns and lower labor requirements, thus reducing the cost of cultivation and improving productivity. What is aerobic rice? The simplest way of defining aerobic rice is growing rice like irrigated upland crops, such as wheat or maize. The production of direct-seeded rice without permanent standing water in nonpuddled fields is referred to as aerobic rice. The cultivation of aerobic rice aims at growing rice without puddling and flooding under nonsaturated soil conditions like other upland crops. This is considered a new system of rice cultivation for saving water (Bouman et al 2005, Yang et al 2005).In this system, fields remain unsaturated throughout crop growth and water is applied to bring the soil water content to field capacity. Rice could be grown aerobically under irrigated conditions just like wheat or maize (Bouman 2001). The distinguishing feature of aerobic production systems is that crops are direct-seeded in freely drained unbunded or bunded fields in nonpuddled soil where no standing water layer is maintained in the field, and roots grow mainly in an aerobic environment (Atlin et al 2006). Aerobic rice is "improved upland rice" in terms of yield potential and "improved lowland rice" in terms of tolerance of water deficit (Bouman et al 2006).

63

Aerobic rice is a new type of rice that is aerobic-soil-adapted and input‐responsive. It grows well in nonpuddled and nonsaturated soil with water content of 70% to 100% of water-holding capacity throughout the growing season. Aerobic rice technology can be adopted in favorable upland to improve productivity. Because of the crop’s better input responsiveness and lodging tolerance, farmers can harvest a higher yield in normal rainfall years and even in a deficient rainfall year if some supplemental irrigation facilities can be created. Aerobic rice can also replace transplanted water deficit-prone shallow lowland rice wherever the availability of water in the early part of the season is insufficient for transplanting but rice can be grown as a direct-seeded crop. Aerobic rice is targeted to more favorable environments where the land is relatively flat, soil fertility is good with better water-holding capacity, and the soil can be frequently brought to near field capacity by rainfall or supplemental irrigation. Differences between upland rice and aerobic rice Upland rice is traditionally grown as a rainfed crop in unbunded, unflooded fields, undulated or terraced, where soil conditions in the root zone remain aerobic throughout most of the growing season (Prasad et al 1999).Upland rice is also grown on marginal lands suitable for other crops such as minor millets, oilseeds, and orchards, and in the relatively favorable semi-wet conditions merging into the shallow rainfed wetlands. Upland rice is defined as “rice grown in rainfed, naturally well-drained soil without surface-water accumulation, normally without phreatic water supply, and normally not bunded.” Upland rice varieties are low-yielding but tolerate water deficit and low fertility, thus giving low but stable yield under the adverse environmental conditions of upland. However, high inputs of fertilizer and supplemental irrigation to upland rice will lead to lodging and thus reduce yield. The most common components are (1) aerobic soil, (2) the absence of surface-water accumulation, (3) dependence on rain, (4) plain to sloping topography, and (5) direct seeding. However, semi-aerobic situations in which short-duration rice is direct seeded under dry conditions but where water accumulates for short periods of time are also considered under upland rice. All upland rice is aerobic but all aerobic rice is not grown in upland. Both aerobic and upland rice are adapted to aerobic soil conditions, but aerobic rice varieties are more input-responsive and higher yielding than traditional upland rice and can also be grown in shallow lowland without puddling and water accumulation. The traditional direct-seeding system in rainfed lowland of eastern India Before the introduction of semidwarf varieties, direct seeding was the usual practice of crop establishment for rainfed shallow lowland in the states of Bihar, Jharkhand, and Chhattisgarh. This method of rice cultivation is known as “beushening” or “biasi” and is common throughout the rainfed regions of eastern India (Singh et al 2000). In this system, rice is broadcast seeded using a higher seed rate before the onset of monsoon and the young crop at 25 to 35 days after germination is cross-plowed with a light country plow when about 15 cm of rainwater accumulate in the crop field. This is followed by a light hand weeding if the weed pressure is more. In some cases, seedling redistribution is also done after this operation to fill in blank patches and thin out the crowded patches to have a more uniform population. The beushening system of rice cultivation serves the purpose of weeding, thinning, and soil disturbance in the

64

root zone, which stimulates better crop growth. This beushening system requires less labor and spreads labor demand over a longer period of time (Singh et al1994). Farmers in beushening rice cultivation use traditional tall cultivars, which do not respond to higher doses of fertilizer application. As a result, productivity in the traditional beushening system of rice cultivation is often quite low, but, in a year with low rainfall at the beginning of the season, direct seeding followed by beushening is the only alternative as transplanting is not possible without adequate water for puddling. Moreover, bullock-driven direct seeding or mechanized tractor-cum-seed driller direct seeding provides a long-term solution to the acute labor shortage in agriculture. Where can aerobic rice be grown? Globally, water is becoming an increasingly scarce and precious commodity and eastern India is no exception. Water is the most limited and essential natural resource used in agriculture. It is thus essential to improve water-use efficiency in agriculture. Because of increasing water scarcity, there is a need to develop alternative systems that require less water (Bouman et al 2002). It has been estimated that, on average 3,000 to 5,000 liters of water are required to produce 1 kg of rice under transplanting (Bhuiyan et al 1995). Total water inputs in lowland rice in Asia reportedly vary from 400 mm in heavy clay soil to more than 2,000mm in coarse-textured soil, with 1,300 to 1,500 mm as common average values (Bouman and Tuong 2001, Tuong et al 2005). Alternative systems of rice cultivation need to be adopted to maintain rice production under this situation and one such strategy is the cultivation of rice aerobically. Aerobic rice could be successfully cultivated with 600 to 700 mm of total water in summer and entirely using rainfall in the wet season with good rainfall distribution (Shailaja Hittalmani 2007). In eastern India, aerobic rice can be grown on water-deficit-prone bunded lands, usually upper fields or fields with light-textured soil. These fields will have a rare accumulation of standing water for a long period; crops thus remain aerobic for most of the growing season even though they can experience frequent moisture stress, particularly during and after the reproductive phase. These lands are widely distributed across eastern India, notably in Jharkhand, Chhattisgarh, Odisha, and Bihar. These lands require varieties with durations ranging from 95 to110 days. All of these areas require short-duration, photoperiod-insensitive varieties combining tolerance of water deficit with high yield potential. Bunded upland or flat upland rice-growing areas where annual rainfall is high and rainfall with or without supplemental irrigation is sufficient to bring the soil water content close to field capacity, and where farmers can apply a recommended dose of fertilizer and adopt best management practices, are also suitable for aerobic rice. Rice can also be grown aerobically in water-short rainfed shallow lowland production systems where the soil is relatively coarse-textured, well‐drained, and ponding of water occurs for a short period in the season. Many times, because of delayed onset of monsoon, farmers are forced to transplant over aged seedlings and suffer a tremendous yield loss. The available water is insufficient for crop establishment through transplanting but sufficient for aerobic rice cultivation. Shifting from conventional flooded systems to direct seeding in nonpuddled, nonflooded field can reduce the water requirements for rice production by more than 50% by reducing percolation, seepage, and evaporation losses (Castañeda et al 2002). Atlin et al (2006) obtained a yield with tropical aerobic varieties comparable with that obtained in a favorable rainfed lowland environment in South and Southeast Asia, indicating

65

that aerobic rice-based production systems are likely to be attractive to farmers in water-deficit-prone rainfed lowland as well as upland. In water-deficit-prone rainfed lowland rice-producing areas such as the Bastar Plateau in Chhattisgarh, Chotanagpur Plateau in Jharkhand, and western Odisha, average yield is less than 2 t ha−1 and water-deficit risk is very high. In a year with low rainfall, failure to accumulate standing water in bunded fields can delay the transplanting process or may severely disrupt water-dependent weed control operations, such as beushening (Tomar 2002).The dry direct-seeded aerobic rice system, which does not require standing water to accumulate either for establishment or weed management, would be the most reliable rice cultivation option for this ecosystem. Moreover, aerobic rice varieties, because of their input responsiveness, semi tall plant type with better lodging resistance, and high yield potential, can be successfully cultivated in shallow lowland irrigated areas where farmers do not have sufficient water to keep rice fields flooded for a substantial period of time and currently cultivate varieties with 105 to 120 days’ duration through transplanting. The increasing scarcity and rising cost of water threaten the sustainability of irrigated lowland rice. It is expected that, by 2025, 13 million hectares of Asia's wet-season rice and 2.0 million hectares of its irrigated dry-season rice will experience physical water scarcity due to the overuse of water, exacerbated by unpredictable rainfall caused by a changing climate (Tuong and Bouman 2003). In the next 25 years, some 15 to 20 million ha of irrigated rice are projected to suffer from water scarcity, particularly in wet-season irrigated rice in parts of India, China, and Pakistan (Carriger and Vallee 2007). Aerobic rice can replace such irrigated rice where water scarcity has made rice cultivation uneconomical and uncertain. The appropriate plant height varies from region to region based on management practices—less plant height for the high-input irrigated ecosystem (95 to 110 cm) and slightly higher plant height (110 to 120 cm) for the rainfed ecosystem where occurrence of water deficit is highly likely to reduce plant height in addition to low-input management practices followed by farmers. Lodging resistance and better ability of plant roots to uptake nutrients under aerobic conditions are also key traits for the successful development of aerobic rice varieties for the shallow lowland ecosystem. Advantages of aerobic rice The following are the advantages of aerobic rice over transplanted rice: i. The advantages of the aerobic rice system over the transplanting system of rice cultivation is the water savings. Growing aerobic rice can reduce the amount of water required to around 40%, as the crop can be grown without standing water that is needed for lowland transplanted rice. The driving force behind the aerobic rice system is water economy. Castañeda et al (2002) reported a water savings of 73% in land preparation and 56%during crop growth. ii. The other benefit of using aerobic rice varieties is that no churning is done to puddle soil, which, although it is done for retention of water in the fields, it actually disturbs the soil structure and the concentration of micronutrients. Cultivation of aerobic rice will prevent the loss of micronutrients from the soil. iii. An important characteristic of aerobic rice technology that meets the requirements of farmers is that management is relatively simple, saving on labor inputs and reducing drudgery. Operations such as puddling, leveling, raising a nursery, and transplanting are not required. Direct seeding is faster and easier. Rice

66

cultivation has become challenging because of the scarcity of labor due to outmigration. People from villages are moving to cities and prefer to work in industry rather than agriculture. Ding et al (2008) found that saving labor was the most important attribute for the adoption of aerobic rice systems for farmers. iv. Other advantages of using aerobic rice technology are the low input requirement, early establishment, earlier crop maturity by 7 to 10 days, and low incidence of pests and diseases. Early maturity reduces the total duration of the crop, thus increasing the chances of establishing an early second crop using available moisture in the soil whenever possible. v. Aerobic rice reduces the generation of methane gas, a by-product of flooding. vi. In aerobic rice, intercropping or mixed cropping with a leguminous crop is possible, which can help to sustain soil fertility. Aerobic rice is responsive to high inputs, can be rainfed or irrigated, and can tolerate occasional flooding, unlike some other upland crops such as maize or soybean. vii. Weed infestation is a major constraint to the adoption of aerobic rice. However, with the availability of good pre- and postemergence herbicide in combination with appropriate varieties, weed problems can be managed effectively.

Aerobic rice trials: experiences from the upland shuttle network In India, research on aerobic rice is limited. The lack of high-yielding varieties that can provide higher yield under aerobic conditions is a major drawback to aerobic rice cultivation. The popular lowland varieties of medium and short duration (<120 days’ maturity period), currently used for aerobic cultivation, have high yield potential under transplanted conditions but are poorly adapted to the aerobic system. To identify suitable varieties for aerobic cultivation, a set of varieties was evaluated under bunded or favorable upland at three sites in eastern India (Hazaribag, Faizabad, and Raipur) and one site in western India (Nawagam) under the ICAR-IRRI Upland Rice Shuttle Breeding Network (URSBN) during the 2004WS(wet season). The trial was conducted under rainfed conditions without any supplementary irrigation but a higher fertility rate (NPK at 80:40:40) than normal was used for upland. The genotypes evaluated included four IRRI-developed first-generation aerobic rice varieties (IR74371-3-1-1, IR74371-46-1-1, IR74371-54-1-1, and IR74371-70-1-1), two from the Philippines’ national aerobic program (PR26406-B-B-2 and PR27699-B-D808-4-4), and 18 upland breeding lines (of both very early and early duration) developed at the collaborating centers in India. At all the sites, trials were established by dry direct seeding and weeds were controlled though a combination of early postemergence herbicide applications and manual weeding. The mean grain yield of all the genotypes over the sites was 2.84 t ha−1 (Table 1). The new aerobic lines developed at IRRI such as IR74371-3-1-1 (3.72 t ha−1), IR74371-70-1-1 (3.69 t ha−1), IR74371-54-1-1 (3.51 t ha−1), and IR74371-46-1-1 (3.36 t ha−1) yielded above 3.0 t ha−1 across sites under aerobic conditions. The aerobic rice line from PhilRice (Philippines), PR27699-B-D808-4-4 (3.46 t ha−1), also performed very well in this situation. These lines have given 15% to 27% yield superiority over popular upland variety Annada (2.93 tha−1) and 45% to 61% over water-deficit-tolerant variety Vandana (2.31 t ha−1).These genotypes flowered at around 80 days and they have intermediate stature (approx. 90 cm). Most of the upland breeding lines developed in the national systems were poor yielders (yielding < 3.0 t ha−1) except for NDR 118 (3.22 t ha−1) and NDR 1052-14 (3.0 t ha−1).

67

The same trial was also conducted in the 2005WS involving more sites (Semiliguda, Jagdalpur, Ambikapur, Ranchi, and Almora) in the network. Some of the poor-performing entries in 2004 were excluded and some new entries from IRRI and national systems were included. A total of 25 entries were evaluated during the 2005WS. At all the sites, a trial was conducted under rainfed conditions without any supplementary irrigation. During the season, some of the sites received less rainfall, distribution was not uniform, and the trial was affected by a severe water deficit. For the combined analysis of data, the sites were divided into two groups depending upon the trial condition, stressed and nonstressed. The genotype × site interaction was large, even after partitioning the trials for moisture availability. The means for grain yield, plant height, and crop duration under stressed and nonstressed environments are presented in Table 2. There were significant differences among varieties at both the stressed and nonstressed sites. In non-stressed environment, mean yields of more than 3.5 t ha−1 were achieved by B6144F-MR-6 (3.7 t ha−1), IR74371-70-1-1 (3.48 t ha−1), and PSBRc-9 (3.66 t ha−1). These genotypes outyielded local check variety Annada (3.19 t ha−1) by 15%. Under severe stress conditions, except for IR74371-70-1-1 (1.6 t ha−1), mostly upland varieties such as Annada (1.59 t ha−1), UPLRi 7 (1.31 t ha−1), NDR1055-6 (1.44 t ha-1), and NDR118 (1.3 t ha−1) yielded more than 1.3 t ha−1. From the yield obtained under stress and nonstressed conditions for the evaluated entries, it is clear that IR74371-70-1-1 yielded superior under both conditions and it has combined high yield potential as well as good yield under aerobic water-deficit conditions very well. During 2006, some new aerobic lines developed at IRRI and also by the NARES were included for evaluation and the same set of varieties was evaluated for two consecutive years. For the analysis of data, sites were again grouped into stressed and nonstressed based on moisture availability during the growth period. The mean grain yield of the cultivars for the experiments conducted during the 2006-07WS under stressed and nonstressed environments appears in Table 3. Across the years, new aerobic line IR74371-70-1-1 and improved upland variety from Indonesia B6144F-MR-6 yielded more than 3.0 t ha−1 in nonstressed environments and these were the highest-yielding entries in this trial. The interesting fact about these entries is that, even under stressed environments, they are not inferior to the best upland varieties in terms of grain yield. The new aerobic lines developed at UAS, Bangalore (ARB 2 and ARB 3), did not perform well in this particular trial, buta sister selection of these lines, ARB 6, has yielded 4 to 5 t ha−1 in nonstress and around 2.0 t ha−1 under water-deficit stress in the trials conducted under the drought network (Shashidhar 2007). IR74371-70-1-1 yielded 5.1 t ha−1 under nonstress and 2.8 t ha−1 in stress in 2007.

A promising aerobic line released as a variety The most promising line designated as IR74371-70-1-1-CRR-1 was nominated for testing under the All India Coordinated Rice Improvement Program (AICRIP). It consistently outyielded national, regional, and local check varieties in the national coordinated trials during the 2005-07 wet seasons. In AICRIP trials, it had a yield advantage of 29.2% and 19.1% over the national and regional check varieties under rainfed water-deficit conditions and a yield advantage of 22.8% and 31.4% over national and regional check varieties under nonstressed conditions (Table 4). In the IRRI-India water-deficit breeding network on-station breeding trials conducted at

68

eight locations for 3 years (2005 to 2007), it provided a yield advantage of 0.5 t ha−1 under moderate stress and 1.0 t ha−1 under severe water-deficit conditions over IR64 and IR36, the two prominent varieties grown in these regions, and it yielded on a par with them under irrigated control conditions. Based on its performance in multi-location trials, IR74371-1-1-CRR-1 (IET 19576) was released in 2009 and notified as Sahbhagi dhan by CVRC in 2010 for general cultivation in water-deficit areas of Odisha and Jharkhand. Sahbhagi dhan flowers in 80 to 85 days, has intermediate stature and non lodging characteristic with 5 to 8 effective tillers per plant, and it possesses long slender grains. This variety is resistant to leaf blast and moderately resistant to brown spot and sheath blight. In quality characteristics, it has high head rice recovery (64.7%), high alkali value (7.0), intermediate amylose content (24.7%), and medium gel consistency (57 mm). This variety was also tested in farmers’ fields in PVS trials and farmers liked it. The new variety is a good alternative to high-yielding varieties such as Annada, Narendra 97, Sadabahar, and IR36 because of its ability to tolerate water deficit, its high yielding ability, and its good grain quality. Conclusions Aerobic rice technologies have been developed for water-short irrigated environments where water availability at the farm level is too low or too expensive for economic rice production, such as the northwestern part of India. But, for eastern India, aerobic rice with tolerance of water deficit is seen as a potentially viable alternative to poor-yielding upland rice or rainfed shallow lowland rice. In India, the average yield of upland rice is around 1.0 t ha−1, whereas in Brazil the average yield is about 5.0 t ha−1. The high-input system in Brazil has developed as a function of the availability of improved responsive cultivars combined with management practices that reduce the risk of production in aerobic soil, including a shift to less water-deficit-prone areas (Guimaraes and Stone 2002). In the rainfed shallow lowland, adoption of aerobic rice would reduce the risk of crop establishment in the event of a delayed onset of monsoon. The low yield potential of current aerobic varieties compared with irrigated or lowland varieties is a concern for adoption by farmers, but the development of new aerobic rice varieties and optimal management techniques will ensure that on a par yield can be achieved in farmers’ fields. The main hurdle to the adoption of aerobic rice systems is the inability to control weeds. However, the availability of new pre- and postemergence herbicides would facilitate efficient and economic management of weeds. The main enablers for the adoption of aerobic rice by farmers are reduced labor and water requirements and the fact that aerobic rice is relatively water deficit tolerant, which brings more stability to rice production in the rainfed production system. References Atlin GN, Lafitte HR, Tao D, Laza M, Amante M,Courtois B. 2006. Developing rice

cultivars for high fertilizer upland systems in Asian tropics. Field Crops Res. 97:43-52.

Bhuiyan SI, Sattar MA, Khan MAK. 1995. Improving water use efficiency in rice irrigation through wet seeding. Irrig. Sci. 16:1-8.

Bouman BAM. 2001. Water efficient management strategies in rice production. Int. Rice Res. Notes 16(20):17-22.

Bouman BAM, Tuong TP. 2001. Field water management to save water and increase its productivity in irrigated rice. Agric. Water Manage. 49:11-30.

69

Bouman BAM, Xiaoguang Y, Huaqi Y, Zhiming W, Junfang Z, Changui W, Bin C. 2002. Aerobic rice (Han Dao): a new way of growing rice in water-short areas. In: Proceedings of the 12th International Soil Conservation Conference, Beijing, 26-31 May 2002.Beijing: Tsinghua University Press. p 175-181.

Bouman BAM, Peng S, Castañeda AR, Visperas RM. 2005. Yield and water use of irrigated tropical aerobic rice systems. Agric. Water Manage. 74:87-105.

Bouman BAM, Yang X, Wang H, Wang Z, Zhao J, Chen B. 2006. Performance of aerobic rice varieties under irrigated conditions in North China. Field Crops Res. 97:53-65.

Carriger S, Vallee D. 2007. More crop per drop. Rice Today 6(2):10-13. Castañeda AR, Bouman BAM, Peng S, Visperas RM. 2002. The potential of aerobic

rice to reduce water use in water-scarce irrigated lowland in the tropics. In: Bouman BAM, Hengsdijk H, Hardy B, Bindraban PS, Tuong TP, Ladha JK, editors. Water-wise rice production. Proceedings of the International Workshop on Water-Wise Rice Production. 8-11 April. Los Baños (Philippines): International Rice Research Institute. 356 p.

FAO (Food and Agriculture Organization). 2006. Prospects for food, nutrition, agriculture, and major commodity groups. World agriculture: toward 2030-2050, Interim report. Global Perspective Studies Unit. Rome: FAO. www.faostat.fao.org.

FAO (Food and Agriculture Organization). 2009. FAOSTAT database. Rome: FAO.www.faostat.fao.org.

Guimaraes EP, Stone F. 2002. Evolution and characterization of high yielding upland rice ecosystem in Brazil. In: Atlin GN, Lafittee HR, George T, editors. Upland rice research in partnership. II Proceedings of the Upland Rice Research Consortium Meeting, 4-8 September 2000. Los Baños (Philippines): International Rice Research Institute. 12 p.

Pandey S, Bhandari H, Hardy B, editors. 2007. Economic costs of drought and rice farmers’ coping mechanisms: a cross-country comparative analysis from Asia. Los Baños (Philippines): International Rice Research Institute. 203 p.

Prasad K, Mishra GN, Sinha PK, Shukla VD, Singh RK, Variar M, Maiti D, Singh CV, Mandal NP, Rajamani S, Dani RC. 1999. The CRURRS upland rice research: achievements and directions. Central Rainfed Upland Rice Research Station, Hazaribag, India.

Serageldin I. 2004. Speculations on the future of water and food security. International Food Policy Research Institute.

Shailaja Hittalmani. 2007. Aerobic rice cultivation. Brochure, MAS Lab, University of Agricultural Sciences, GKVK, Bangalore.

Shashidhar HE. 2007. Aerobic rice: an efficient water management strategy for rice production. In: Aswathanarayana U, editor. Food and water security in developing countries. UK: Taylor & Francis. p 131-138.

Singh VP, Hossain M. 2000. Rice area in different ecosystems in eastern India. In: Singh VP, Singh RK, editors. Rainfed rice: a source book of best practices and strategies in eastern India. Los Baños (Philippines): International Rice Research Institute. p 7-8.

70

Singh VP, Singh RK.2000. Rainfed rice: a source book of best practices and strategies in eastern India.Los Baños (Philippines): International Rice Research Institute. 292 p.

Singh RK, Singh VP, Singh CV.1994. Agronomic assessment of beushening in rainfed lowland rice cultivation in Bihar, India. Agric. Ecosyst. Environ. 51:271-280.

Singh VP, Singh RK, Singh CV. 2000. Beushening system in Bihar, India. In: Singh VP, Singh RK, editors. Rainfed rice: a source book of best practices and strategies in eastern India. Los Baños (Philippines): International Rice Research Institute. p110-114.

Tomar VS. 2002. The beushening system of rice crop establishment in eastern India. In: Pandey S, Mortimer M, Wade L, Tuong TP, Lopez K, Hardy B, editors. Direct seeding in Asian rice systems: strategic issues and opportunities. Proceedings of a Workshop, 25-28 January 2000. Bangkok, Thailand. Los Baños (Philippines): International Rice Research Institute.

Tuong TP, Bouman BAM. 2003. Rice production in water-scarce environments. In: Kijne JW, Barker R, Molden D, editors. Water productivity in agriculture: limits and opportunities for improvements. UK: CAB International. p 53-67.

Tuong TP, Bouman BAM, Mortimer M. 2005. More rice, less water: integrated approaches for increasing water productivity in irrigated rice-based systems in Asia. Plant Prod. Sci. 8:231-241.

Wade LJ, George T, Ladha JK, Singh U, Bhuiyan SI, Pandey S. 1998. Opportunities to manipulate nutrient-by-water interactions in rainfed lowland rice systems. Field Crops Res. 56:93-112.

Widawsky D, O'Toole JC.1990. Prioritizing the rice biotechnology research agenda for eastern India. New York (USA): Rockefeller Foundation.

Yang X, Bouman BAM, Wang H, Wang Z, Zhao J, Chen B. 2005. Performance of temperate aerobic rice under different water regimes in North China. Agric.Water Manage.74:107-112.

Notes Authors’ addresses: N.P. Mandal and M. Variar, Central Rainfed Upland Rice

Research Station (CRRI), ICAR, PB-48, Hazaribag 825 301, India; A. Kumar, International Rice Research Institute (IRRI), Los Baños, Philippines.

71

Table 1. Grain yield (t ha-1), duration, and plant height (cm) across sites*of upland and recently developed aerobic rice varieties in anaerobic trial, URSBN, 2004.

Variety Grain yield (t ha−1)

Days to flower

Plant height (cm)

Annada 2.93 82 69.1 Danteswari 2.84 82 78.5 GR 3 2.93 88 71.3 IR74371-3-1-1 3.72 84 93.5 IR74371-46-1-1 3.36 84 92.2 IR74371-54-1-1 3.51 83 90.9 IR74371-70-1-1 3.69 82 91.0 NDR1045-2 2.81 81 81.1 NDR1052-14 3.01 83 74.0 NDR1055-6 2.98 82 77.9 NDR118 3.22 86 88.4 NDR80 2.39 87 83.1 NDR97 2.54 80 73.0 Poornima 2.71 81 74.0 PR26406-B-B-2 2.70 87 103.4 PR27699-B-D808-4-4

3.46 83 105.9

R1027-2282-2-1 2.95 89 87.3 R979-2282-2-1 2.26 89 79.6 RR165-1160 2.38 72 76.5 RR166-645 2.51 69 91.1 RR361-2 1.73 65 64.2 Sadabahar 2.83 78 90.6 Tripuradhan 2.47 79 85.8 Vandana 2.31 70 91.8 Mean 2.84 LSD0 .05 0.77

*Faizabad, Hazaribag, Raipur, and Nawagam.

72

Table 2. Performance of upland and recently developed aerobic rice varieties in water-stressed and nonstressed environments, aerobic trial, URSBN, 2005.

Days to flower

(days) Plant height

(cm) Grain yield (t ha−1)

Variety NS1 S2 NS1 S2 NS1 S2

Annada 75 76 69 59 3.19 1.59 B6144F-MR-6 80 83 98 90 3.70 1.19 Danteswari 81 83 81 66 2.60 0.77 GR 3 80 87 70 66 2.27 1.08 IR71525-19-1-1 78 85 88 78 2.97 1.06 IR71700-247-1-1-2 80 84 78 69 2.71 0.87 IR74371-3-1-1 74 83 85 79 3.01 1.18 IR74371-46-1-1 77 80 86 79 3.01 0.94 IR74371-54-1-1 74 84 88 75 3.21 1.06 IR74371-70-1-1 77 85 86 74 3.48 1.60 IR74422-934-02-RR511-2 80 83 90 86 2.80 0.82 IR75013-81-LBN-1-1 85 92 85 66 2.95 0.07 NDR1052-14 75 82 77 68 3.06 1.18 NDR1054-4 76 80 88 78 2.82 0.65 NDR1055-6 79 81 79 72 2.79 1.44 NDR1087-10 79 80 81 64 3.17 0.80 NDR118 74 76 80 74 2.84 1.30 NWGR 2011 78 79 83 57 2.72 0.69 PR27699-B-D808-4-4 80 82 93 86 3.37 1.12 PSBRc 9 82 91 93 83 3.66 0.63 R1027-2282-2-1 82 76 85 79 3.05 0.56 RR388-1-9 74 76 98 85 2.28 0.73 RR51-1 78 75 71 57 2.28 1.12 UPLRi 7 80 80 90 91 3.32 1.31 VL 9807 79 80 107 109 2.48 1.19 Mean 78 82 85 76 2.90 1.00 LSD.05 3.5 4.0 5.9 7.4 0.37 0.31

1Hazaribag, Ranchi, Almora.2Faizabad, Semiliguda, Jagdalpur, Ambikapur, Nawagam, NS = nonstress, S = stress.

73

Table 3. Grain yield (t ha−1) of some upland and recently developed aerobic rice varieties in India and at IRRI under water-stressed and nonstressed environments, aerobic trial, URSBN, 2006-07.

2Hazaribag, Semiliguda, Ranchi, Nawagam.1Faizabad, Jagdalpur,

Lembuchera (2006). 2Hazaribag.1 Semiliguda, Ranchi, Santhapur, Nawagam (2007).

NS = nonstress, S = stress.

2006 2007

Entry NS2 S1 NS2 S1 Annada 2.95 1.34 4.00 2.70 B6144F-MR-6 3.45 1.99 4.90 2.70 UPLRi 7 2.78 2.21 5.20 2.70 PR27699-B-D808-4-4 2.91 1.53 4.20 2.60 PMK 3 2.81 1.90 4.50 2.70 IR71525-19-1-1 2.88 1.80 4.00 2.60 IR74371-3-1-1 2.85 1.80 3.90 2.60 IR74371-54-1-1 2.88 2.05 4.00 2.40 IR74371-70-1-1 3.17 2.08 5.10 2.80 IR72667-16-1-B-B-3 3.06 1.93 4.30 2.60 IR55419-04 2.60 1.69 4.80 2.70 NDR1052-14 3.01 1.84 4.00 2.30 NDR1055-6 3.14 1.62 4.00 2.60 NDR1087-10 2.90 1.98 3.70 2.60 NDR118 2.72 1.39 4.00 2.20 NWGR2007 2.16 2.17 1.90 2.00 NWGR2012 2.90 1.60 3.20 2.50 NWGR2018 1.88 0.53 3.40 1.70 VL30169 1.78 1.32 3.60 1.50 VL30183 1.85 1.20 2.80 1.70 RR511-3 2.65 1.32 4.80 2.30 RR433-2 2.49 1.80 4.40 1.80 ARB 2 2.45 1.42 3.70 1.70 ARB 3 2.69 1.14 3.60 2.20 Mean 2.71 1.65 4.00 2.34 LSD 0.05 0.40 0.33 0.43 0.34

74

Table 4. Grain yield (t ha-1) of IR74371-70-1-1-CRR-1 along with check varieties at water- deficit-affected and normal locations in national coordinated trials (IVT-E and AVT 2-E) conducted under AICRIP during the wet season of 2005-07.

Entry Days to flower

2005 2006 2007 Grain yield (t ha−1) Grain yield (t ha−1) Grain yield (t ha−1) Drought Normal Drought Normal Drought Normal

IR74371-70-1-1-CRR-1

78 2.10 3.87 2.15 4.97 2.80 2.90

Annada (NC) 75 1.62 3.15 2.17 5.10 2.55 2.46 Narendra97 (RC)

75 1.76 2.94 1.89 4.03 1.60 2.52

75

Paper 5 Varietal improvement and weed management for aerobic rice cultivation in the drought-prone eastern region of India

B.N. Singh, Krishna Prasad, A.K. Singh, and Pramod Kumar

Jharkhand is a drought-prone state in eastern India. Rice area, production, and productivity are strongly influenced by rainfall pattern and amount of rain during the wet season. Around 90% of the rice area is rainfed. The state has undulating terrain, where rainwater runoff is very high. Less rain in June-July reduces the rice area, whereas mid-season drought in August-September affects production and productivity. Research on aerobic rice cultivation by direct seeding of fertilizer-responsive high-yielding varieties with supplementary irrigation as a possible option began in 2009. Aerobic rice varietal trials, weed management under direct seeding, and demonstration of available technology were carried out at Birsa Agricultural University (BAU), Kanke, Ranchi, and at the drought-prone Zonal Research Station (ZRS), Chianki, Palamau, Jharkhand. Efforts were made to select better-yielding varieties and improved production technology for aerobic rice cultivation. A production technology package was demonstrated in the drought years of 2009 and 2010 on BAU rice research farms in Chianki and Ranchi. The varieties Naveen of medium duration and Rajshree of late duration were demonstrated for different toposequences (upper, middle, and lower). Dhaincha (Sesbania aculeata), sunnhemp (Crotalaria juncea), cowpea (Vignaunguiculata), or urd bean (Vigna mungo) was grown along with direct-seeded rice, uprooted, and used as green mulch, 25 days after seeding. Some 100:60:40 kg NPK ha−1was used for variety Naveen and 80:40:40 kg ha−1 for variety Rajshree. From seven varietal trials during the 2009and 2010 wet season in Kanke, Apo, IR74371-54-1-1 (Sahod Ulan1, Sookha dhan 2), IR83920-B-B-CRA-103-14-1-1-1, IR84898-B-16-5, CB 05-754, CR 2729-3-2, MAS 946, and MAS ARB 868 with medium maturity (120 days), and NDR 1141 with early maturity (106days),were found promising. Leaf and panicle blast and brown spot were the major diseases. In varietal trials, intermediate plant height (100 to 110 cm), non-lodging, and high yield (>4.0 t ha−1) were the major criteria for selection. Two to three irrigations were applied depending on rainfall pattern. For weed management, growing dhaincha or cowpea or urd bean with preemergence spray of pendimethalin at 0.75 kg a.i. and 2,4D at 0.8 kg a.i. ha−1 at 25 DAS was found effective, and produced higher yield using Naveen.

Abbreviations: AICRIP: All India Coordinated Rice Improvement Program; BAU: Birsa Agricultural University; AVT: advanced varietal trial; IET: initial varietal trial; ZRS: Zonal Research Station; WCE: weed control efficiency.

The newly created state of Jharkhand came into existence on 15 November 2000 as the 28th state carved out from erstwhile Bihar, India. Some 77% of its people are engaged in agriculture and 73% live in rural areas. Out of a total of 7.97 million hectares of geographic area, 1.81 million hectares are net sown cropped area, and cropping intensity is 114% (Mishra 2004). Only 10% of the area is irrigated, where vegetable crops are mainly grown. The average landholding is 1.58 ha and fertilizer consumption was 67 kg NPK ha−1 in 2006-07 compared with 113 kg nationally. Some 50% of the land holdings are in the submarginal category, having less than 0.4 ha, and another 23% (from 0.4 to 1 ha) are in the marginal category. Thus, around 73% of the farmers in Jharkhand State have less than 1 ha of land under cultivation (Table 1).

The major soil type of Jharkhand is red and lateritic light-textured soil in around 78% of the area. Another category of alluvial soils is found in 19% of the area (Table

76

2). Most of the upland and midland soils are in the first category while lowlands are in the second category. In undulating land topography, the upper part of the toposequence (Tanr 2 to Don 1) is light-textured soil and this leads to soil erosion and rainwater runoff. Tanr is the land located on the upper slopes and Don is the land on the lower slopes. These two classes have been further divided based on moisture status and soil organic matter into three subclasses: Tanr 1, 2, and 3 and Don 1, 2, and 3.For convenience of technology generation, these classes have been regrouped in uplands (Tanr 1 and 2), midlands (Tanr 3 and Don 3), and lowlands (Don 2 and Don1).Tanr 1 is bariland near the house or periphery of the village and it is quite fertile, where farmers prefer to grow vegetables or nursery raising for rice. Tanr 2 is upland, where rainfed upland crops such as direct-seeded rice, ragi, maize, urd bean, and pigeonpea are grown. Tanr 3 has runoff from Tanr lands and transplanted rice is grown. Don lands are lowlands, and bottom land in the toposequence is Don 1, followed by Don 2 and Don 3 toward upland. Moisture availability is for a longer period by runoff and seepage in Don 1. Acid upland soil is around 71%, where phosphorus deficiency is a major nutrient constraint. Some 22% of the soils have pH from 5.5 to 6.5, whereas 49% of the soils have pH from 5.5 to 4.5 (Table 2). In such soils, the application of nitrogen alone does not increase yield, until phosphorus is added to the soil. Application of single superphosphate will add calcium and sulfur in the soil, but it is not available. Only diammonium phosphate (DAP), urea, and muriate of potash are available. Application of lime increases the soil pH and nutrient availability, but transportation cost and access at the village level are constraints. Organic carbon of upland and midland soil is also low.

Rainfall pattern Jharkhand has subhumid climatic conditions. The eastern region has high humidity, while the western region has low humidity. The state receives rain from both southwest and northeast monsoon. In Ranchi, average rainfall is 1,289 mm, out of which 85% is received during four rainy-season months (June to September). Rainfall has declined since 2001 and the average rainfall is now 1,048 mm (Figure 1).The southwest monsoon contributes around 85% of the rainwater. The rainfall pattern from1951 to 2006 in Ranchi shows that 29 years had less than normal rainfall, while 6 years had rainfall less than 75% of the normal, which are considered “meteorological drought years.” In normal years, premonsoon rains are received in May, which helps in summer plowing. Monsoon usually breaks in mid-June, but late arrival and early cessation are not uncommon. The distribution of rainfall is also uneven and erratic. In 2009 and 2010, little rain occurred in June-July, and heavy rain fell in August in the first fortnight. In a saturated soil, more than 50 mm of rain on any day are required for puddling, and this did not happen in these both years. This led to a decline in rice area, production, and productivity (Figure 2). In the western palamau region also, rainfall is declining, and the area is slowly shifting toward semi-arid conditions (Figure 3). In a toposequence, water drains from Tanr to Don lands. A rainless period of a week in Tanr 2 having red and lateritic soil affects crop growth and water stress develops. Drought can be early season (June-July), mid-season (August-September), or late (October). Early-season drought reduces the area sown under rice, and mid- and late-season drought affects the production and productivity of the rice crop.

77

Rice cultivation Rice is the major crop of Jharkhand, grown on around 60% of the total cropped area, and 70% of the area is under food grains. The area fluctuates from year to year depending on rainfall and its distribution. It has varied from a maximum of 1.67 million ha in 2008-09 to a minimum of 0.72 million ha in 2010-11 (Figure 4). Production has also varied from 3.4 million tons in 2008-09 to1.04 million tons in 2010-11. Productivity has varied from 2.0 tons to 1.2 tons ha−1 accordingly. Rice area is maximum in years of good rainfall during June-July and declines in years of delayed rainfall. Likewise, production is maximum when rainfall is well distributed in a season for normal crop growth and minimum in years of drought in the vegetative or reproductive stage, or in both stages. Rice is grown in kharif (wet season), rabi (boro), and summer in the dry season. In the kharif season, it is now grown in uplands on around 0.3 million ha, under midlands on 0.7 million ha, and in lowlands on 0.5 million ha under rainfed conditions. Around 1,000,000 ha are under irrigated transplanted conditions during the wet season. Uplands are grown as rainfed under direct-seeded conditions, and in midlands and lowlands under transplanted conditions. Around 90% of the rice area is rainfed. Upland rice area has declined over the years because of weed problems and low productivity. In the dry season, area is limited to 10,000 ha in the eastern region, bordering west Bengal. Rice is grown in all three zones and in all 24 districts of the state. Some of the rice varieties under cultivation in Jharkhand are shown in Table 3. Possibility for aerobic rice in Jharkhand Drought is a recurring phenomenon in Jharkhand. Sometimes, rainfall is adequate throughout the country, but Jharkhand as a whole or in part is always affected by drought (Table 4). In one region or another, districts, or parts of a district, drought occurs almost every year. Based on the success of direct-seeded rice with sprinkler irrigation in Brazil, and with surface irrigation in China, research on direct-seeded rice with modern varieties, irrigation, and higher fertilizer inputs as compared with rainfed upland, low-yielding direct-seeded rice has been carried out on a small scale since 2006 in Ranchi. Systematic trials in collaboration with the All India Coordinated Rice Improvement Program (AICRIP) on aerobic rice began in India in 2009. In Brazil, higher yields with improved plant types under direct-seeded aerobic rice were obtained using input-responsive drought-tolerant and pest-resistant varieties with sprinkler irrigation or in regions with favorable rainfall distribution (Stone et al 1990, Guimaraes and Stone 2000, Pinheiro et al 2006). In northern China also in water-deficit environments, aerobic rice cultivation has been successful (Wang et al 2002, Yang et al 2005). In India also, research has been carried out at the University of Agricultural Sciences (UAS), Bangalore; Central Rice Research Institute (CRRI), Cuttack; and Indian Agricultural Research Institute (IARI), New Delhi, for the past 10 years. At UAS, drought screening was part of a strategy to develop aerobic rice cultivars, whereas, at other locations, only varietal evaluation was being carried out.

At BAU-Ranchi, weed management and varietal trials have been conducted. The rice crop is direct seeded in nonpuddled soil like wheat, bunds are made to harvest rainwater, and a higher dose of NPK as with high-yielding varieties (HYVs) is being recommended to obtain higher yield. Direct seeding in uplands is common in Jharkhand, but yield is low because of tall non-input-responsive varieties. Aerobic rice can be cultivated in the upper, middle, and lower part of the toposequence, and accordingly varieties of different maturity duration are needed. Supplementary

78

irrigation is provided through surface irrigation, when no rain and moisture stress are visible during any period of crop growth (Table 5).In direct-seeded rice, weeds are a major problem, so varieties of intermediate height with good seedling vigor are preferred for demonstration. Generally, rains begin in the third week of June and nursery sowing starts for rainfed transplanted crops. In the first fortnight of July, when soils are saturated and rainfall of about 50 mm occurs on a single day, water harvesting through bunds is done and the fields are puddled and transplanted. Growing rice in unpuddled aerobic soil, with the use of external inputs such as supplementary irrigation and fertilizer, and aiming for high yield under tropical conditions have also been proposed at IRRI (Bouman et al 2005). The main driving force behind aerobic rice is economical water use and timely sowing. As insufficient and delayed rain reduces rice area, and production in transplanted rice cultivation, a fundamental approach to reduce water inputs in rice is growing the crop like an irrigated upland crop, such as wheat or maize. Instead of trying to reduce water input in lowland paddy fields, the concept of having the field flooded or saturated is abandoned altogether (Bouman and Tuong 2001). In direct seeding, weeds are hardy and have a profuse root and shoot growth habit; they grow faster than rice, thereby checking the growth of rice plants by severe weed-crop competition. This can be managed by weedicides or with legume crops as live mulch. Materials and methods Birsa Agricultural University, Kanke, Ranchi, is situated at 23o17′ north latitude, 85o19′ east longitude, and an altitude of 625 m above mean sea level. This location has a subtropical sub-humid climate characterized by hot and dry summer, cold winter, and moderate annual rainfall. The experimental plot represents midland having red loam type of soil, which belongs to the “red yellow-light-gray” group representing the major soil order Alfisols of Jharkhand. The soils are well aggregated with high permeability and low water retention capacity due to the presence of hydrated oxides of iron and aluminum. Soil test values indicated that the soil of the experimental plots was sandy loam (sand 63%, silt 22%, and clay 15%) in texture, slightly acidic in nature, and moderately fertile, being low in organic carbon, low in available nitrogen, high in available phosphorus, and medium in available potassium.

Three types of experiments were carried out with the rice research unit, Kanke, Ranchi, at BAU during the 2009 and 2010 wet seasons (WS): one on the demonstration of an available technology package for farmers and extension personnel; the second on varietal trials to select higher-yielding, drought-tolerant, disease-resistant varieties; and the third on weed management in aerobic rice. A demonstration was also carried out at ZRS-Chianki, Palamau, when rainfall was less than in Ranchi (Figure 3). i. Demonstration of aerobic rice technology In the 2009WS, at BAU-Kanke and ZRS-Chianki, demonstrations with rice variety Naveen were carried out. Naveen (CR 749-20-2; IET 20-2) was released in Odisha State in 2005.It has intermediate height, 125 days’ duration, and good initial vigor, and it is suitable for medium land in the toposequence. In the 2010WS, demonstrations were further continued in Kanke, with Naveen and Rajshree. Rajshree (TCA 80-4) was released in Bihar in 1987. It has intermediate height, 145 days’ duration, good initial vigor, and nonlodging plant type, and it is suitable for shallow and medium-deep rainfed lowlands. In one of the demonstrations, sunnhemp (Crotalaria juncea) was grown with Naveen, a mixed crop with aerobic rice.

79

ii. Varietal evaluation for aerobic rice Varietal yield trials from AICRIP, Hyderabad, began in 2009 to select better-yielding varieties for midlands. Two yield trials, an initial evaluation trial (IET) and an advanced varietal trial (AVT), were conducted in the 2009WS, and five trials in the 2010WS in Kanke, Ranchi. The IET was grown in two replications and the AYT in three replications. The trials were direct seeded in the first week of July with a row-to-row distance of 20 cm and 80:60:40 kg NPK ha−1 was applied. A full dose of P and K was applied as basal, while N was applied as topdressing in two equal splits as urea at 25 and 50 days after seeding. Topdressing was done after hand weeding. Two irrigations were given as and when drought symptoms were visible in the field during the rainless period. iii. Weed management for aerobic rice As weeds are a major problem in direct-seeded rice, a legume crop such as dhaincha (Sesbania aculeata), cowpea (Vigna unguiculata), and urdbean (Vigna mungo) as green mulch with different nitrogen and weedicide treatments was used in Kanke during the2009 and 2010WS with the objectives to find the effects of nitrogen rates and weed management on growth, grain yield attributes, grain and straw yield, and weed control efficiency. During the 2009WS, three treatments comprising three nitrogen rates (N1=75, N2=100,and N3=125 kg ha−1) in the main plots and five treatments in the subplots (W1 = dhaincha in between rice rows + pendimethalin at 0.75 kg a.i. ha-1, W2 = rice + pendimethalin at 0.75 kg a.i., W3 = cowpea in between rice rows + pendimethalin at 0.75 kg a.i., W4 = weed-free check (two mechanical weedings at 20 and 40 DAS), and W5 = unweeded check). During the 2010WS, subplots with seven weed control methods were used:W1 = dhaincha in between rice rows + pendimethalin at 0.75 kg a.i. ha−1, W2 = rice + pendimethalin at 0.75 kg a.i., W3 = dhaincha in between rice rows + pendimethalin at 0.75 kg a.i. + 2,4D (0.8 kg a.i.) at 25 DAS, W4 = urdbean in between rice rows + pendimethalin at 0.75 kg a.i., W5 = urdbean in between rice rows + pendimethalin at 0.75 kg a.i. + 2,4D (0.8 kg a.i.) at 25 DAS, W6 = weed-free check, and W7 = unweeded check. In W1, W3, and W5 treatments, dhaincha, cowpea, or urd bean was uprooted and left in between rows as live mulch. The experiment was laid out in a split-plot design with three replications in both years. The experimental soil was slightly acidic (pH 6.2), sandy loam in texture, low in organic carbon (0.46%) and available nitrogen (228 kg ha−1), high in available phosphorus (35.3 kg ha−1), and medium in available potassium (157.1 kg ha−1). During the 2009WS, 40 kg P and 20 kg K and in the 2010WS 60 kg P and 40 kg K were applied as basal in all the treatments. N was applied as DAP (18% N and 46% P2O5) as basal, and urea (46% N) as topdressing. P and K were applied through DAP and muriate of potash (60% K2O). Plot size was 20 m2 each. The trials were seeded on 23 June in 2009 and 3 July in 2010.

Results and discussion (i) Demonstration of aerobic rice technology As per demand from the state Department of Agriculture, government of Jharkhand, aerobic rice technologies were demonstrated at BAU, Kanke, Ranchi, during the 2009 and 2010WS. In Ranchi, demonstration was carried out with an animal-drawn seeder in moist soil at row-to-row spacing of 25 cm. Variety Naveen was demonstrated in Ranchi and at ZRS-Chianki. For field days, a package of technology for aerobic rice cultivation was prepared with the available information (Table 5).In the 2009WS, grain yield of 3.5 t ha−1 at BAU-Kanke and grain yield of 2.9 t ha−1at

80

ZRS-Chianki were obtained. The low yield was mainly due to late seeding of aerobic demonstrations. In the 2010WS at BAU- Kanke, grain yield of 4.1 t ha−1 by Naveen and 4.3 t ha−1 by Rajshree was obtained. The demonstration response was very good as per observations of farmers during field days. (ii) Varietal evaluation for aerobic rice yield trials

2009 yield trials. Two trials, an initial varietal trial−medium-early duration (IET-ME)and an advanced variety trial–medium-early duration (AYT-ME), were evaluated. The IET-ME had 10 entries and was evaluated in two replications. The trial was direct seeded on 9 July2009. IR64 was the national check and Birsa Vikas Dhan 203 was the local check (Table 6). Days to 50% flowering varied from 81 to 124 days, plant height from 80 to 117 cm, and grain yield from 0.60 to 4.65 t ha−1. Late-duration varieties were affected by drought and had 55% to 69% grain sterility. Varieties with 110 to 120 cm plant height under direct seeding will have better weed competitive ability, and varieties with 115 to 120 days’ duration will be desirable for midland. IR83920-B-B-CRA-103-14-1-1-1 yielded a maximum of 4.65 t ha−1 compared with 3.89 t ha−1 by IR64.

In AYT-ME of the 2009WS, 22 entries were evaluated in three replications. The trial was direct seeded on 8 July 2009. IR64 and Naveen were used as checks (Table 7). Grain yield varied from 2.6 to 5.87 t ha−1. The three top-yielding entries were MASARB 868 (5.87 t ha−1), CR 2624 (5.17 t ha−1), and RR1218-509-2-452-1(5.12 t ha−1). Naveen outyielded IR64. Its yield was 4.62 compared with 3.78 t ha−1 by IR64. The plant height of Naveen (112 cm) was also higher and more appropriate than that of IR64 (86 cm). 2010 yield trials. Five aerobic rice yield trials from AICRIP were conducted during the 2010WS: an initial varietal trial-early (IVT-E), an initial varietal trial-medium early (IVT-ME), an advanced yield trial 1-mid-early (AVT-1 ME), an advanced varietal trial 2-early (AVT-2E), and advanced varietal trial 2-medium-early (AVT-2 ME).

In IVT-E, 21 entries were evaluated in two replications. They were direct seeded on 8July2010. Both brown spot and leaf blast infestation were observed in this trial (Table 8). Two entries, CR 2729-3-2 (3.66 t ha−1) and NDR1141 (3.52 t ha−1), outyielded Rasi (3.26 t ha−1). NDR1141 was early maturing, and it has seed-to-seed duration of 106 days, as compared with other entries of 120 days ‘duration. It was also taller (109 cm) to better compete with weeds.

In IVT-ME, 22 entries were evaluated in two replications. Direct seeding was done on 7 July2012. IR64 and Birsa Dhan 202 were the checks (Table 9). Brown spot and blast infestation was severe in certain entries. The check Birsa Dhan 202 was susceptible to brown spot and leaf blast. Four entries were higher yielding than check IR64 (3.12 t ha−1): IR84887-B-154 (4.85t ha−1), MAS 946 (4.57 t ha−1), IR83380-B-B-124-1 (4.57 t ha−1), and IR70844-10-SRN-43-1-B-1-1 (4.52 t ha−1).The plant height of MAS946 was 87 cm and that of IR70844-10-SRN-43-1B-1-1 was 85 cm. All four selected entries are of medium duration (113 to 127 days’ seed-to-seed duration).

In AVT1-ME, seven entries were evaluated in three replications. The trial was direct seeded on 5 July2010. Leaf blast and brown spot infestation was high in this trial, and local check Naveen was observed to be susceptible (Table 10).Two entries, MAS946 (4.64 t ha−1) and IR83920-B-B-CRA-103-14-1-1-1 (4.44 t ha−1), were top yielders. They had moderately resistant and resistant reactions to brown spot and leaf blast, respectively. IR83920-B-B-CRA-103-14-1-1-1 was also a top-yielding entry

81

in the 2009WS. These three entries also belong to the medium maturity group, with plant height of 91 to 101 cm, respectively.

InAYT-2E, 10 entries were evaluated in three replications in an RCBD. They were direct seeded on 6 July 2010. Rasi was used as a national check and Birsa Vikas Dhan 203 as a local check (Table 11).Brown spot infestation was observed in this trial. Three entries were higher yielding than the two checks: CB 05-754 (6.49 t ha−1), CR 2601 (6.19 t ha−1), and IR74371-54-1-1 (6.0 t ha−1). These entries were also moderately resistant to brown spot. IR74371-54-1-1 was released in 2009 as Sahod Ulan 1 in the Philippines and as Sookha Dhan 2 in Nepal in 2011. Its height is 115 cm, which is more appropriate for rainfed conditions.

In AVT2-ME, eight entries were evaluated in three replications. The trial was direct seeded on 5 July 2010. Both brown spot and leaf blast were observed. Variety Naveen was observed to be moderately susceptible to both (Table 12). IR55423-01 (Apo) was maximum yielding (4.4 t ha−1), followed by check IR36 (4.0 t ha−1). IR55423-01 was released as Apo in the Philippines for drought-prone areas. It has plant height of 105 cm and is moderately resistant to brown spot and blast. Apo was also recommended for release in Odisha in 2012.

All seven yield trials during 2009 and 2010 showed wide variability for grain yield, its components, plant height, maturity, and resistance to natural infestation of leaf blast and brown spot. Many entries outyielded the local or national checks Naveen, Rasi, IR36, and IR64.Varieties such as IR36 and IR64 were primarily developed for the irrigated ecosystem, and their height decreases in the direct-seeded rainfed ecosystem because of the occurrence of drought. There is consistency in the yield superiority of certain entries in both years. Three entries have already been released in India, the Philippines, and Nepal. MAS946, which performed well in two trials during 2010, has been released as Sharada in Karnataka, India. IR55423-01 was released in the Philippines as Apo, and now it has been recommended in Odisha, India, for aerobic conditions. IR74371-54-1-1 was released in the Philippines as Sahod dhan 1 in 2009 and in Nepal as Sookha dhan 2 in 2011. This shows their wide adaptation for the aerobic system of cultivation. These varieties were developed through screening for drought tolerance. This shows that, for developing varieties for aerobic conditions, especially for rainfed environments, drought tolerance should be considered as an important trait. All other high-yielding entries should be evaluated in multi-location trials and in farmers’ fields through participatory varietal selection. Entries susceptible to blast and brown spot should be discarded.

In weed management for aerobic rice, results of experiments during 2009 and 2010WS revealed that grain and straw yield were significantly higher at 125 kg N than at 75 and 100 kg N (Tables 13, 14, and15). The increase in grain yield arose mainly by an increase in values for yield components, that is, effective panicles m−2, fertile grains per panicle, and 1,000-grain weight. Weed dry matter (g m−2) at harvest was significantly higher at 125 kg N than with others, but decreased significantly in treatments with dhaincha, cowpea, and urd bean grown in between rice rows. The use of pendimethalin at 0.75 kg a.i. in 2009 and 2010 and/or 2,4D at 0.8 kg a.i. ha−1 in addition to pendimethalin also reduced weed flora and weed biomass. The weed flora in the trial is given in Table 16. In the 2009WS, grain yield was maximum in W1 (4.4 t ha−1), that is, rice+ dhaincha (1:1 interrow) with pendimethalin, followed by W2 (4.27 t ha−1), rice + cowpea (1:1 interrow) with pendimethalin (Table 13). Weed dry matter was also low in rice + dhaincha or rice +

82

cowpea plots compared with other treatments. Pendimethalin alone with rice was not able to reduce weed biomass (Table14). In the 2010WS, WCE was a maximum of 61.4% in W3, the combination of dhaincha in between rice rows + pendimethalin at 0.75 kg a.i. + 2,4D at 0.8 kg a.i. ha−1), followed by W1 (dhaincha in between rows+ pendimethalin at 0.75 kg a.i. ha−1).The use of dhaincha was better than urd bean. This shows that dhaincha should be preferred, but, if it is not available, then urd bean, which is widely grown as a grain legume crop, can also be used effectively as a live mulch or browning through spraying of 2,4D. Dry matter of weeds at harvest and WCE were higher in interrow crops of dhaincha or urd bean than in mechanically weeded plots.

Conclusions From the different varietal trials on aerobic rice during 2009 and 2010, it was observed that certain released varieties such as Apo, Sahod dhan 1, and Sharada are found to be higher yielding than Naveen, selected on the basis of duration and plant type. Naveen was found to be susceptible to leaf blast and brown spot. These varieties can be tested in multi-location trials and in farmers’ fields for wide adoption. New higher-yielding varieties with tolerance of major diseases and pests must continue to evolve. In weed management trials, rice variety Naveen grown under aerobic conditions with 125 kg N ha−1 was productive and profitable. Planting dhaincha in between rice rows + pendimethalin at 0.75 kg a.i. ha−1 +2,4D at 0.8 kg a.i. ha−1 at 25 DAS (W3) suppressed maximum weed density, recorded lower weed dry matter accumulation, provided maximum weed control efficiency, and resulted in higher productivity of rice. In the absence of dhaincha seed availability, cowpea, urd bean, and sunnhemp can also be grown as an intercrop or mixed crop. Policy and future outlook The research on aerobic rice in India and the drought-prone states such as Jharkhand and others isquite limited. In the era of climate change, and shifts in rainfall pattern, it is essential that systematic research on aerobic rice varietal development and screening for drought tolerance and other location-specific biotic and abiotic stresses be carried out through multidisciplinary team efforts. Singh and Chinnusamy (2007) have also demonstrated the potential of this technology in farmers’ fields of western Uttar Pradesh and the amount of water saved through adoption of this technology. Globally, Prasad (2011) has made a review on problems, potential, and possibilities in the different countries of the world. In collaboration with IRRI, varietal screening and agronomic experiments have begun at CRRI, Cuttack, and variety Apo was released in 2012 in Odisha. Earlier work at UAS, Bangalore, also led to the release of Sharda for Karnataka State. Aerobic rice technology seems promising, especially for Jharkhand, where drought at the early stage reduces the area and production under rice. If rains are delayed or fail also, aerobic rice demonstration through Krishi Vigyan Kendra and on-farm demonstration should be carried out. The extension system of the Department of Agriculture should be involved in training and dissemination of technology on a large scale.

References Bouman BAM, Tuong TP. 2001. Field water management to save water and

increase its productivity in irrigated rice. Agric. Water Manage. 49:11-30.

83

Bouman BAM, Peng S, Castañeda AR, Visperas RM. 2005. Yield and water use of irrigated tropical aerobic rice systems. Agric. Water Manage. 74:87-105.

GuimaraesEP, Stone LF. 2000. Current status of high yielding aerobic rice in Brazil. High yielding aerobic rice,7-8 September 2000. Los Baños (Philippines): International Rice Research Institute.12p.

Mishra RK. 2004. Agricultural statistics of Jharkhand at a glance. Birsa Agricultural University, Kanke, Ranchi, Jharkhand. 111 p.

PinheiroBS, CastroEM, Guimarᾶes CM. 2006. Sustainability and profitability of aerobic rice production in Brazil. Field Crops Res. 97:34-42.

Prasad R. 2011. Aerobic rice systems. Adv. Agron. 111:207-247. Singh AK, ChinnusamyV. 2007. Aerobic rice: a success story. Ind. Farm. 57(8):7-10. Stone LF, Pinero B da S, Silvera PM. 1990. Sprinkler irrigated rice under Brazilian

condition. In: Proceedings of the 17thInternational Commission, Goiania, Brazil. Int. Rice Comm. Newsl. 39:36-40.

Wang H, Bouman BAM, Zhao D, Wang C, Moya PF. 2002. Aerobic rice in northern China: opportunities and challenges. In: Bouman BAM, Hengsdijk H, Hardy B, Bindraban PS, Tuong TP, Ladha JK, editors. Water-wise rice production. Proceedings of the International Workshop on Water-wise Rice Production, 8-11 April.Los Baños(Philippines): International Rice Research Institute. p 143-154.

Yang X, Bouman BAM, Wang H, Wang Z, Zhao J, Chen B. 2005. Performance of temperate aerobic rice under different water regimes in North China. Agric. Water Manage. 74:107-122.

Notes Authors’ addresses: B.N. Singh, Krishna Prasad, and A.K. Singh, Birsa Agricultural

University (BAU), Ranchi, Jharkhand; Pramod Kumar, Zonal Research Station (ZRS), Chianki Palamu, Jharkhand, India.

Figure 1. Rainfall pattern in Kanke, Ranchi, from 1956 to 2010. Figure 2. Rainfall pattern in Kanke, Ranchi, during 2010 compared with the normal rainfall pattern. Figure 3. Declining rainfall pattern in Palamau, during the last 33 years. Figure 4. Sowing of aerobic rice by animal-drawn seeder.

Figure 5. Interrow cultivation of dhaincha (Sesbania aculeata) in between rows of direct-seeded rice.

Figure 6. Growing of sunnhemp (Crotalaria juncea) in between rows of direct-seeded rice in demonstration plot, 2010WS.

Figure 7. Variety Naveen suitable for aerobic rice cultivation in Jharkhand. Figure 8. Demonstration of aerobic rice production technology at Zonal Research

Station, Chianki, Palamau, Jharkhand, during 2009WS. Figure 9. Area, production, and productivity of rice in Jharkhand, India.

84

Table 1. Typology of landholdings in Jharkhand, India.

Landholding (ha) Typology Percentage (%) 0 Landless 9.3

Up to 0.4 Submarginal 49.8 0.41−1.0 Marginal 22.8 1.01−2.0 Small 12.5 2.01−4.0 Medium 4.4

> 4.0 Large 1.2 Total 100.0

85

Table 2. Soil profile of Jharkhand.

Indicator Unit (%) Land typology Soil with phosphorus deficiency 66.0 Upper, middle, and lower

portion of toposequence Acid upland soil 71.0 Upper part of

toposequence pH (<5.5) 49.0 Mainly upper part of

toposequence, forest soils

Soil type: pH (5.5−6.0) 22.0 Upper and middle part of toposequence

Red and lateritic (Tanr 2 and 3 and Don 3)

78.0 Upper and middle part of toposequence

Alluvium (Don 1 and 2) 19.0 Lower part of toposequence

86

Table 3. Rice varieties cultivated in Jharkhand.

Rice ecosystem

Varieties under cultivation

Traditional varieties

Area (million ha)

Upland (direct seeded)

Vandana, Kalinga 3, Birsa Dhan 108, Birsa Dhan 109

Brown Gora, Brown Gora, landraces

0.30

Midland transplanted

IR36, IR64, Lalat, Naveen, Sahbhagi

Br 8, many landraces

0.40

Lowland transplanted

Swarna (MTU 7029), Samba Mahsuri (BPT 5204), Birsamati, Rajshree, Rajendra Mahsuri

Br 9, Br 10, many landraces

0.60

Gall midge- affected area

Rajendra Dhan 201, Lalat, Naveen

Traditional 0.20

Hybrid ProAgro 6444, PHB 72, KRH 2

− 0.15

Total 1.65

87

Table 4. Drought scenario in Jharkhand.*

Year Drought and rainfall pattern 2001 In Jharakand, 11 districts and 57 blocks were affected by drought. 2002 All of Jharkhand was affected by drought. 2003 Eleven districts and 67 blocks were affected by drought in Jharkhand.

2004 All of Jharkhand was affected by drought and nine districts were near famine. 2005 Drought affected all of Jharkhand.

2006 Severe drought in Palamau and Bokaro districts of Jharkhand. The rest had good rainfall.

2007 Normal rainfall all over Jharkhand.

2008 Palamau was the worst affected. The rest of Jharkhand had good rainfall and maximum production of rice in the state.

2009 All of Jharkhand was affected. Severe drought. Little rain in June-July. 2010 All of Jharkhand was affected. Severe drought. Little rain in June-July.

*Compiled from various sources and newspapers.

88

Table 5. Good agronomic practices for aerobic rice in Jharkhand.

Component Recommendation Technology upgrading Varieties Naveen and Sahbhagi, for midland and

sowing till mid-July, and Rajshree (140 days) for sowing by June and in lower toposequence

New varieties yielding more than 5.0 t ha−1 are Apo, Sahod Ulan 1, and Sharda

Seeding time From third week of June to end of July, depending on rainfall pattern

Apply preemergence herbicide at1−3 DAS and pendimethalin at 0.75 kg a.i. ha−1

Seed rate Paddy seed at 80 kg ha−1 for direct seeding in rows behind plow; dhaincha/sunnhemp/cowpea/urd bean at 40 kg ha−1, as broadcast before

Use tractor-driven seed- cum-fertilizer seed drill for sowing paddy

Seed treatment

Carbendazim at 2 g kg−1 seed

Organic fertilizer

5 t ha−1 FYM or compost or 2 tons vermicompost or 1 ton neem cake/ karanj cake/mahua cake

Inorganic fertilizer

100:60:40 kg NPK ha−1 for Naveen 80:40:20 kg NPK ha−1 for Rajshree

125:60:40 kg NPK ha−1

Basal fertilizer Midland: 20kg N(DAP):60:40 kg NPK ha−1

Lowland: 20 kg N(DAP):40:20 kg NPK ha−1

Irrigation 3 to 4, as and when there is no rain Mulching Uproot dhaincha/sunnhemp/cowpea/urd

bean after 20−25 days and use as mulch in rice fields

Spray 2-4, Dat 0.8 kg a.i. ha−1at 25 DAS

First topdressing, 25 DAS

Before uprooting green manure mulch crop, topdress 40 kg N as urea in Naveen, and 20 kg in Rajshree

Hand weeding 50 DAS before N topdressing Second

topdressing, 50 DAS

40 kg N as urea for Naveen 20 kg N as urea for Rajshree

Third topdressing, 75 DAS

20 kg N as urea for Rajshree

Disease control

If leaf or panicle blast is observed, spray carbendazim (0.1%) at 25 to 30 DAS and at flowering

Yield potential 4−5 t ha−1 >5 t ha−1

89

Table 6. AICRIP initial varietal trial: mid-early-duration aerobic IVT-ME during 2009WS.

Designation DTF1 Plant height (cm)

Fertile grains panicle-

1

Grain sterility (%)

Effective panicles m−2

Grain yield (t ha−1)

IR83920-B-B-CRA-103-14-1-1-1 87 98 168 5.9 257 4.65

Birsa Vikas Dhan203** 94 84 118 13.7 263 3.92

IR64* 82 88 110 10.3 300 3.89 R1570-2649-1-1546-1 87 97 175 11.5 270 3.86

R1124-69-1-45-1 92 95 130 11.3 202 3.79 CR 2697 89 117 142 9.5 207 3.66 PNR 625 81 92 104 11.7 240 3.26 PNR 621 82 98 166 14.2 190 2.79 NDR 8016 124 85 47 55.2 197 0.63 NDR 8014 122 80 21 69.0 217 0.60 Mean 94.2 93.4 118.1 21.2 234.3 3.11

Range 81−124

80−117 21−175 5.9−69.

0 190−300

0.60−4.65

CV (%) 19.9 LSD 0.05 1.42 *National check and **local check. 1DTF =days to 50% flowering.

90

Table 7. AICRIP aerobic advanced varietal trial−mid-early duration (AVT-ME) in 2009WS at BAU, Kanke, Ranchi.

Designation DTF Plant height (cm)

Fertile grains panicle−1

Grain sterility (%)

Effective panicles m−2

Grain yield (t ha−1)

MAS ARB 868 94 89 206 6.6 240 5.87 CR 2624 90 108 159 7.1 236 5.17 R1218-509-2-452-1 89 96 131 10.0 254 5.12 Naveen** 87 112 152 7.0 272 4.62 CR 2496-50 98 87 141 9.6 261 4.46 PAU 201 94 81 151 7.8 263 4.42 RP 4092-128-104-95-12

103 89 113 9.9 260 4.42

Kadamba 122 82 130 11.7 262 4.24 CR 2597 106 94 114 10.4 274 4.02 IR64* 88 86 113 9.5 250 3.78 CB 05758 81 97 160 6.3 253 3.76 CR 2632 78 117 129 7.7 236 3.76 MAS ARB 25 87 85 92 9.5 248 3.74 RP 4092-126-75-11-4 92 93 125 11.2 273 3.74 NDR 8778 82 83 97 8.5 251 3.66 R1835-RF-38 81 106 119 12.1 235 3.56 CR 2631 78 107 128 8.0 213 3.56 RPHR 523-1-3-5-4 87 89 114 11.8 268 3.34 R1836-RF-39 77 93 99 8.5 262 2.95 CR 2496-24-5 102 84 152 7.4 263 2.73 CB 06803 82 101 124 9.8 283 2.66 CR 2499-50 108 75 124 10.4 252 2.51 Mean 91.2 93.4 130.6 9.1 254.9 3.91 Range 77−1

22 75−117 92−206 6.3−12.1 213−283 2.51−5.87

LSD(0.05) 1.30 CV (%) 20.10 *National check and **local check.

91

Table 8. AICRIP initial variety trial-early aerobic (IVT-E) at BAU, Kanke, Ranchi, 2010WS.

Designation DTF Plant height (cm)

Effective panicles

m−2

Brown spot

Leaf blas

t

Grain yield

(t ha−1)

CR 2729-3-2 90 95 278 4 4 3.66 NDR 1141 76 109 219 4 3.5 3.52 Rasi* 73 91 257 2 2 3.26 Rewa 1103 78 115 228 2 2 3.06 Birsa Dhan 201** 87 79 193 4 3.5 2.93 IR77196-169-NDR B-1-9-2 94 93 204 3 2.5 2.79

R 1576-1680-1-536-1 96 77 268 1 1 2.73 CB 001523 91 69 175 5 5 2.66 R 1037-649-1-1 91 79 287 4 3.5 2.46 JGL 17004 73 98 257 3 3 1.99 MAS 26 92 74 179 3 2.5 1.99 RP 5128-9-11-2 90 80 160 2 2 1.93 Rewa 1101 92 80 291 3 3 1.86 NP 25 74 71 239 2 2 1.73 NP 3114-2 105 73 247 4 4 1.66 NDR 1143 91 71 247 2 1.5 1.46 NDR 1142 72 103 278 4 3.5 1.26 R 2085-RF-97 86 79 165 3 3 1.06 CR 2728-5-1 85 98 190 2 1.5 0.93 CB 05-753 93 89 191 3 3 0.86 Rewa 1102 79 92 194 3 2.5 0.60

Range 72−105

69−115

160−291 1−5 1−5 0.60−3.6

6 Mean 85.7 86.4 226 3.0 3.0 1.61 CV (%) 0.9 4 12.24 LSD 0.05 1.6 20 0.41

*National check and **local check.

92

Table 9. AICRIP initial variety trial−mid-early aerobic (IVT-ME) during 2010WS at BAU, Kanke, Ranchi.

SN Designation DTF Plant height (cm)

Effective panicles

m-2

Brown spot

Leaf blast

Grain yield

(t ha−1) 1 IR84887-B-154 83 106 366 3 3 4.85

2 MAS 946 93 87 287 4 4 4.57

3 IR83380-B-B-124-1 92 98 345 4 4 4.57

4 IR70844-10-SRN-43-1-B-1-1 97 85 357 2 2 4.52

5 IR84899-B-185 89 105 367 0 0 4.12 6 IR84898-B-165 92 116 292 4 4 4.06

7 IR83927-B-B-279 80 112 272 3 3 3.92

8 IR84899-B-182-CRA-12-1 79 108 249 1 1 3.92

9 R1570-418-1-149-1 99 89 287 4 4 3.56

10 RP 4092-111-35-4-5 97 66 309 4 4 3.52

11 IR84899-B-185 77 117 234 3 3 3.36 12 IR64* 93 77 296 4 4 3.12

13 IR82590-B-B-32-2-150 93 95 261 4 4 3.12

14 IR82616-B-B-64-3 92 86 309 4 4 2.73

15 KMP 175 79 108 276 4 4 2.46 16 IR84899-B-184 78 105 230 1 1 1.99

17 RP 4092-103-18-6 106 77 224 0 4 1.73

18 R2086-RF-98 76 107 232 3 3 1.53 19 RP 1529-7-3-1 114 82 259 1 1 1.46

20 RP 4092-115-78-42 96 89 290 5 5 1.26

21 R1532-1246-1-1111-1 107 74 309 7 7 1.0

22 Birsa Dhan 202** 112 120 314 6 6 0.86 Range 76−114 66−120 224−367 0−7 0−7 0.86−4.85 Mean 91.5 95.8 289.0 3.2 3.4 2.26 CV(%) 0.7 2.3 7.54 LSD 0.05 1.3 14.1 0.35 *National check and **local check.

93

Table10. AICRIP advanced variety trial1−mid-early-duration aerobic (AVT-1ME), 2010WS

at BAU, Kanke, Ranchi.

Designation DTF Plant height (cm)

Effective panicles

Brown spot

Leaf blast

Grain yield

m−2 (t ha−1) MAS 946 88 91 438 3 3 4.64

IR83920-B-B-CRA-103-14-1-1-1 84 101 225 1 1 4.44

IR64* 89 83 359 4 4 3.87

R1570-2649-1-1546-1 95 94 258 2 2 3.52

IR84880-B-CRA- 277-2-3-2-1 91 113 281 1 1 3.30

Naveen** 94 99 272 7 7 2.59

R1124-69-1-45-1 95 91 432 6 6 2.27 Mean 91 96 324 3.4 3.4 4.30 Range 84−95 83−113 225−438 1−7 1−7 2.27−4.64 CV (%) 0.8 2.4 5.37 LSD 0.05 1.3 14.2 0.41

*National check and **local check.

94

Table11. AICRIP advanced varietal trial 2−early aerobic (AVT-2E) at BAU Kanke, Ranchi, 2010WS.

Designation DTF Plant height (cm)

Effective panicles m−2

Brown spot (1−9 scale)

Grain yield (t ha−1)

CB 05-754 91 103 271 3 6.49 CR 2601 89 101 296 3 6.19 IR74371-54-1-1 86 115 237 3 6.00 CB.0015-24 89 92 304 5 5.99 IR55419-04 85 95 226 1 5.59 Rasi* 85 98 344 4 5.42 MAS-26 90 100 326 5 5.10 Birsa Vikas Dhan 203** 106 101 306 4 4.93 R1207-4-290-1 89 105 330 6 3.19 R1565-2449-3-1484-1 92 90 296 7 2.13 Mean 90.2 100.0 293.6 4.1 5.10

Range 85−106 90−115 226−344 1−7 2.13−6.49

CV (%)

6.3

5.79 LSD (0.05) 0.8

31.9

0.60

*National check and **local check.

95

Table12. AICRIP advanced variety trial 2−mid-early-duration aerobic (AVT-2ME), 2010WS at BAU, Kanke, Ranchi.

Designation DTF Plant height (cm)

Effective Panicles m−2

Brown spot

Leaf blast

Grain yield (t ha−1)

IR55423-01 ( Apo) 95 105 354 4 4 4.44 IR36* 95 80 343 4 4 4.01 MAS 946 90 94 385 3 3 3.90 R1218-509-2-452-1 99 91 269 6 6 3.22 CR2597 103 97 362 6 6 2.92 Naveen** 96 103 304 5 5 1.86 R1835-RF-38 92 113 233 7 7 1.33 RP 4092-128-104-95-12

110 92 432 4 4 0.63

Mean 98 97 335 4.9 4.9 3.40 Range 90−110 80−113 233−432 3−7 3−7 0.63−4.44 CV (%) 0.8 4.74 5.43 LSD 0.05 1.3 27.8 0.32 *National check and **local check.

96

Table 13. Nitrogen × weed control method (N× W) interaction for grain yield (t ha−1) of aerobic rice variety Naveen during 2009WS at Kanke, Ranchi.

Treatment N1 (75 kg N ha−1)

N2 (100 kg N ha−1)

N3 (125 kg N ha−1)

Mean

W1: rice + dhaincha (1:1) + pend.at 0.75 kg a.i. ha−1

3.81 4.23 4.40 * 4.15

W2: pend. at 0.75 kg a.i. ha−1 + 1 H.W. at 60 DAS

3.41 3.83 3.98 3.74

W3: rice+ cowpea (1:1) + pend. at 0.75 kg a.i. ha−1

3.82 4.16 4.27 * 4.08

W4: two mechanical weedings (20 and 40 DAS)

3.42 3.79 3.93 3.71

W5: unweeded check 2.84 3.18 3.31 3.11 Mean 3.46 3.84 3.98 N at same level of W S.Em.=0. 15C.D.(P=0.05)− NS

W at same level of N S.Em.=0. 17C.D.(P=0.05)− NS

97

Table 14. N × Winter action for weed dry matter (g m−2) at harvest in aerobic rice during 2009WS at Kanke, Ranchi.

Treatment N1(75 kg N ha−1)

N2 (100 kg N ha−1)

N3 (125 kg N ha−1) Mean

W1: rice + dhaincha (1:1) + pend. at 0.75 kg a.i. ha−1

21.6 * 23.1 26.5 23.7

W2: pend. at 0.75 kg a.i. ha−1 + 1 H.W. 60 DAS

56.2 72.2 79.7 69.4

W3: rice + cowpea (1:1) + pend.0.75 kg a.i. ha−1

26.8 25.2 28.7 26.9

W4: two mechanical weedings (20 and 40 DAS)

54.3 67.5 85.3 69

W5:unweeded check

103.2 108.4 111.2 107.6

Mean 52.42 59.27 66.28

N at same level of W

S.Em.=3.64C.D.(P=0.05)−11.19

98

Table 15. Effect of nitrogen doses and weed control methods on grain yield, yield attributes, and weed dry matter of aerobic rice during 2010WS at Kanke Ranchi.

Treatment Effective panicles m−2

Fertile grains panicle−1

1,000-grainweight (g)

Grain yield (t ha−1)

Straw yield (t ha−1)

Weed dry matter (gm−2) at harvest*

WCE (%)

Main plot: N level N1:75 kg N ha−1 234 86.1 23.17 3.53 5.00 26.26

(709.6) 49.2

N2:100 kg N ha−1 256 90.5 23.58 3.74 5.52 27.34

(769.6) 44.9

N3:125 kg N ha−1 266 95.4

23.72 3.90 5.78 28.86

(850.7) 39.1

CD (P=0.05) 9.6 6.8 0.39 0.11 0.20 0.97 − Subplot: weed control method

W1: rice + dhaincha (1:1) + pend. at 0.75 kg a.i. ha−1

284 96.6

24.0 4.30 5.99 24.14

(583.9) 58.2

W2: pend. at 0.75 kg a.i. ha−1 + 1 H.W. at 60 DAS

227 85.2

23.5

3.42 4.96 28.32 (803.7) 42.4

W3: W1+2,4D at 0.8 kg a.i. ha−1 at 25DAS

295 98.1

24.0 4.37 6.17 23.16(538.9) 61.4

W4: rice + urd bean (1:1) + pend. at 0.75 kg a.i. ha−1

253 90.6

23.1 3.73 5.34 26.82

(721.3) 48.3

W5: W4 + 2,4D at 0.8 kg a.i. ha−1 at 25DAS

256 91.9

23.5 3.78 5.41 26.22(689

.6) 50.6

W6: weed-free check, H.W. at 20 and 40 DAS

257 92.0

23.3 3.94 5.97 26.47 (702.9) 49.6

W7: unweeded check 192 80.5

23.2 2.53 4.20 37.31

(1,396.3) −

LSD (0.05) 25.5 4.6 0.79 0.30 0.37 1.35 CV (%) 9.5 5.2 4.2 8.4 7.0 5.16 *Data subjected to square root transformation. Numbers in parentheses are the original value.

99

Table 16. Major weeds present in the experimental field of rice.

Botanical name English name Family Grasses

Echinochloa colona Water grass Poaceae Eleusine indica Goose grass Poaceae Digitaria sanguinalis Large crab

grass Poaceae

Paspalum distichum Knot grass Poaceae Brachiaria milliformis Signal grass Poaceae

Broadleaf weeds Ludwigia parviflora Water purslane Onagraceae Sphelenthus acmella Toothache

plant Sphenocleaceae

Eclipta alba False daisy Asteraceae Commelina benghalensis Day flower Commelinaceae

Sedges Cyperus iria Flat sedge Cyperaceae Fimbristylis miliaceae Fimbristylis Cyperaceae Kyllinga bravifolia Kyllinga Cyperaceae Cyperus difformis Nut sedge Cyperaceae

100

Figure 1. Rainfall pattern in Kanke, Ranchi, from 1956 to 2010.

101

Figure 2. Rainfall pattern in Kanke, Ranchi, during 2010 compared with normal rainfall pattern.

102

Figure 3. Declining rainfall pattern in Palamau, during the last 33 years.

103

Figure 4. Sowing of aerobic rice by animal-drawn seeder.

104

Figure 5. Interrow cultivation of dhaincha (Sesbania aculeata) in between rows of direct-

seeded rice.

105

Figure 6. Growing of sunnhemp (Crotalaria juncea) in between rows of direct-seeded rice in demonstration plot, 2010WS.

106

Figure 7. Variety Naveen suitable for aerobic rice cultivation in Jharkhand.

107

Figure 8. Demonstration of aerobic rice production technology at Zonal Research Station, Chianki, Palamau, Jharkhand, during 2009WS.

108

Figure 9. Area, production, and productivity of rice in Jharkhand, India.

1.421.50

1.38 1.361.28

1.35

1.62 1.64 1.67

0.98

0.72

1.76

2.70

2.07

2.30

1.90

1.56

2.97

3.33 3.40

1.47

1.041.20

1.79

1.49

1.691.59

1.15

1.83

2.02 2.03

1.50 1.45

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

2000-2001 2001-2002 2002-2003 2003-2004 2004-2005 2005-2006 2006-2007 2007-2008 2008-2009 2009-2010 2010-2011

Year

Are

a, p

rod

uc

tio

n a

nd

pro

du

cti

vit

y o

f ri

ce

Area(m ha) Production (m tons)productivity (tons/ha)

109

Paper 6 Rice varietal improvement and management practices under aerobic and alternate wetting and drying conditions in Nepal: progress and challenges

R.B. Yadaw, R.K. Mahato, A. Kumar, B.P. Tripathi, and S.N. Sah

Rice yield in the rainfed areas of Nepal is not only low but also highly unstable. Aerobic rice cultivation under direct seeding, and alternate wetting and drying (AWD) in transplanted rice, with suitable drought-tolerant varieties, will not only save water but also bring more area under cultivation, which will lead to increased rice production and productivity. Experiments on selecting suitable rice varieties for aerobic rice and AWD were carried out from 2006 to 2009 at different sites in Nepal by IRRI, NRRP (Hardinath), and RARS (Tarahara). Relevant components in production technology, weed management, and cropping systems were also evaluated through on-station and on-farm research. Selection among fixed lines was carried out in observation yield trials (OYTs) and advanced yield trials (AYTs) under aerobic and AWD conditions to select better-yielding lines than the local checks. Under AWD, five promising lines were further tested in a coordinated varietal trial (CVT) and in participatory varietal selection (PVS). In the PVS trials conducted in farmers’ fields, IR80411-B-49-1 and B 6144-MR 6 were found better than the checks and they were released in 2009asTarahara 1 and Hardinath 2, respectively, for aerobic rice cultivation in Nepal. For the AWD system, IR05N-449 and IR78937-B-B-B-B-1 have been found promising. Rice-maize and rice-rice cropping systems had produced higher grain yield and returns than rice-wheat and rice-mungbean cropping systems. Using the AWD system, 8−13% irrigation water can be saved without any yield decline. In aerobic rice, weed management by butachlor at 1.5 kg a.i. ha−1 sprayed 2 weeks after seeding (WAS) followed by one hand weeding at 4WAS was found most effective.

Abbreviations.AWD =alternate wetting and drying; AYT =advanced yield trial; CVT =coordinated varietal trial; JRP=Jute Research Program; OYT =observational yield trial; NRRP = National Rice Research Program; PVS =participatory varietal selection: WAS =weeks after seeding.

Rice is the staple food in Nepal and a mainstay for the rural population for food security. Rice is grown on 1.49 million hectares, with a total production of 4.46 million tons of milled rice (MoAC2010). Out of the total rice area in Nepal, about 51% of the area is rainfed and 49% is irrigated. Rainfed rice is important for the livelihood of 233,000 resource-poor farmers living in the hills and Terai region of Nepal (CBS 2002). Population growth combined with changes in diet will lead to an increase in the rice requirement to 8.3 million tons by 2025. The purpose of this project was to increase food security and farm income in rainfed areas of the Terai using water-saving technologies. Two representative sites in the central Terai, Kishanpur Village of Dhanusha District and Bengadabar Village of Mahottari District, were selected for PVS and on-farm research (Shreshtha et al 2007). So far, rice varieties Radha4, Radha 11, and Radha12 have been recommended for the rainfed rice environment, but these varieties have shown a poor performance in farmers’ fields. Radha11 is also reported to

110

be susceptible to blast and bacterial leaf blight and it has poor cooking quality (Chaudhary et al 2004). These varieties, although they are suited to more favorable lowland, perform poorly under drought, as they were not bred/ selected for drought tolerance. The productivity of rainfed rice is on a declining trend due to water shortage resulting from climate change and the erratic nature of monsoon. Rice grown under traditional practices in the Asian tropics and subtropics requires between 700 and 1,500 mm of water for a cropping season depending on soil texture (Bhuiyan 1992). Farmers are also using irrigation either by lifting water from rivers/canals/ponds or pumping underground water by diesel-run pump sets when drought conditions prevail. A majority of the farmers cannot afford to irrigate using a pump set because of the increasing price of fuel. Farmers are also not able to provide irrigation to rice fields in the tail portion of canal areas because of the shortage of water under drought conditions, resulting in low production. Considering the abovementioned problems, research on water-saving technologies was carried out under the ADB-supported project “Developing and disseminating water-saving rice technologies in South Asia” with the goal of enhanced livelihood of rice farmers in the rainfed rice production domain through the development and dissemination of aerobic rice and AWD water-saving technologies.

Manual transplanting of rice is the main method for rice establishment under irrigated and rainfed lowland ecosystems in the country. Rice is grown in three seasons: spring (chaite dhan), kharif (wet), and boro (dry). The total annual rainfall in Nepal varies between 1,000 and 4,000 mm, with an average of 1,814 mm. More than 75% of the rainfall occurs from June to September. Irrigation is 22% year-round and the rest is mainly dependent on rainfall (APP1999).The average operational landholding of a farmer is reported to be about 0.14 ha per capita in the Terai and 0.05 ha per capita in the hill region. Most of the farmers follow the transplanting method, which is labor-intensive and causes drudgery to women. Rice growers are looking for an alternative to manual transplanting. In aerobic rice, direct seeding is practiced in place of transplanting. Direct seeding substantially reduces labor and also helps in establishing the rice crop on time in case of delayed rain, with a reduction in the drudgery involved in manual transplanting. Direct seeding reduces transplanting cost by 36% over manual transplanting (Tripathi et al 2004). Studies on direct seeding showed that, under good management and weed control, no significant differences in grain yield between direct seeding and manual transplanting were observed (Tripathi et al 2004).Thus, farmers can adopt aerobic rice cultivation and take economic advantage of saving labor at the expense of rice transplanting.

Drought area in Nepal From 1960 to 2005, drought occurred 12 times in one part or another of the country (Gumma et al 2011). Drought has a devastating effect on rice production in upland and rainfed lowland areas, which rely solely on rainfall for rice cultivation. In Nepal, 0.10 million hectares of upland and 0.27 million ha of rainfed lowland rice area are drought-prone (IRRI 2003). In 2006, 13% of the rice area (284,000 ha) could not be planted due to early drought and crop production declined up to 70% in some areas of Nepal (NPC 2008). The eastern region of Nepal, mainly Saptari, Siraha, and Dhanusha districts, was severely affected by drought in 2006 (Figure 1).

111

Water-saving technologies for rice cultivation Aerobic rice and alternate wetting and drying (AWD) are the two major water-saving rice technologies for which research was undertaken in Nepal from 2006 to 2009.In aerobic rice cultivation, rice is grown under direct seeding like wheat, maize, millet, etc. Weeds are a major problem in direct seeding. Weeds are managed either manually or through the use of weedicide. AWD is practiced in transplanted rice, in which the rice crop is not flooded for most of the duration, except during early establishment and flowering, but is kept under saturated conditions. Irrigated rice varieties do not perform well under such conditions and varieties with drought tolerance could be a better alternative. In our study, efforts were made to select suitable varieties under aerobic and AWD conditions. Experiments on weed management and cropping systems under aerobic conditions were also carried out in order to identify the most effective weed control practice and cropping system for farmers.

Materials and methods

Four types of experiments were conducted on-station and on-farm: 1. Varietal selection. Pedigree selections from segregating generations under AWD, aerobic rice OYT and AYT, PVS for aerobic rice and AWD, and CVT. 2. Integrated weed management 3. Cropping system 4. Quantifying water in the AWD system

Different varietal trials were conducted to identify superior genotypes with higher grain yield and drought tolerance suited to aerobic and AWD conditions during 2006-09. Fertilizer was applied as per recommendation at 60:30:30 kg NPK ha−1. The trials were carried out in an augmented and randomized complete block design (RCBD) with three replications. Observational yield trials (OYTs) and advanced yield trials (AYTs) were planted/seeded in 5m ×1m plot size. In participatory varietal selection (PVS), weed control, and AWD trials, plot size of 5m×2 m was used. The spacing was 20×20 cm for all trials under AWD. Under aerobic seeding, row-to-row distance was 20 cm.

The integrated weed management experiment was conducted in a split-plot design with three replications using three rice varieties, Loktantra, Mithila, and Radha11 in 2007, and B6144F-MR-6-0-0, IR80411-B-49-1, and Radha-11 in 2008 and 2009. The varieties were treated as subplots and seven weed control treatments as main plots in both years. The plots were seeded directly in aerobic dry conditions. Seeding was done continuously within rows; with row-to-row spacing of 20 cm. Fertilizer was applied at 60:30:30 kg NPK ha−1. Half of the nitrogen and a full dose of phosphorus and potash were applied as basal. The topdressing of N was applied at 25 DAS. The herbicide butachlor and 2,4D at 1.5 kg a.i.ha−1 were sprayed in 2007 and 2008. Nominee gold at 1.2 kg a.i. ha−1 was sprayed instead of 2,4D at 1.5 kg a.i. ha−1 as post emergence in 2009.

For quantifying water in AWD and to study the performance of rice varieties under AWD and conventional irrigation systems, experiments were conducted at NRRP-Hardinath during the kharif season (WS) in 2007, 2008, and 2009. The experiments used a split-plot design with two irrigation systems (I1=conventional system andI2= AWD

112

system) as a main-plot factor and four rice varieties (Sabitri, Mithila, Hardinath1, and Radha 4) as a sub plot factor in 2007 and 2008; whereas, in 2009, three rice varieties (IR80411-B-49-1, IR78397, and Sabitri) were used as a subplot factor. Continuous water of 5 cm was maintained in the plot throughout the crop growth period in the conventional system. In AWD, irrigation was provided when the field showed hair cracks.

In a crop rotation experiment, Hardinath1 and Radha4 were tested in rice-rice, rice-maize (cv. Devati), rice-mungbean (cv. PS-16), and rice-wheat (cv. Gautam) systems during 2007, 2008, and 2009 at NRRP- Hardinath. The plot size was 5m×5m.The trial was conducted in an RCBD with three replications.

Results and discussion

To improve stability in production and to reduce production cost, due consideration was given to the development and selection of drought-tolerant varieties with higher yield than released cultivars. The following research activities were undertaken.

Varietal selection for aerobic rice and AWD

Pedigree selection from segregating generations Selection from11 segregating F2 generations developed at IRRI were made under transplanted AWD conditions. A total of 136 plants (F3) in the 2007WS, 168 plants (F4) in the 2008WS, and 67 plants (F5) in the 2009WS were selected (Table 1). i. Aerobic rice observational yield trial A total of 126, 114, and 18 entries along with the check were evaluated in the 2006WS, 2007WS (first set), and 2007WS (second set) at NRRP- Hardinath to identify superior genotypes with higher grain yield and drought tolerance suited for direct-seeded aerobic conditions. On the basis of higher grain yield, earlier maturity, and more drought tolerance than the higher-yielding check, Radha 11, a total of 17, 17, and 5 entries were selected (Tables 2, 3, and 4). Lines IR05N455, IR80411-B-28-4, and IR80411-B-49-1 were more preferred (Table 3). Apo is already a released variety in the Philippines in the aerobic drought-prone ecosystem. In the 2007WS OYT at JRP-Itahari, 102 entries along with the check were evaluated to identify superior genotypes for higher grain yield under aerobic conditions. Eight entries were found superior to the check, Kanchhi Masuli (Table 5). The genotypes IR80411-B-49-1, IR80411-B-28-4, IR80411-B-160-4, and Kanchhi Masuli were also found resistant to nematode. ii. Advanced yield trial Selected entries from the OYT were evaluated in an AYT. Five AYTs were conducted at NRRP-Hardinath from the 2007 to 2009WS. AYT 2007. In an AYT of less than 120 days’ duration, 12 selected lines were compared with local checks in the 2007WS at Hardinath to find out suitable lines under rainfed aerobic conditions. The lines IR80411-B-49-1, IR79328-126-3-1-2, and IR78581-12-3-2-2 were found to outyield check variety Kachorwa 26 (Table 6).

113

AYT 2008. Another 12 selected lines were evaluated in an AYT (<120 days) in the 2008WS and 12 in the 2009WS at Hardinath. The four selected lines in the 2008WS, IR78937-B-3-B-B-1, IR80013-B-141-4-1, IR78937-B-20-B-B, and IR84899-B (Table 7), and IR80013-B-20-B-4, IR78913-B-22-B-B-B-4, IR78937-B-B-B-B-1,and IR84899-B in the 2009WS (Table8), produced higher grain yield than the check varieties. In another trial, 12 and 16 selected lines of more than 120 days’ duration were evaluated under aerobic rainfed conditions at Hardinath during the 2008WS and 2009WS, respectively. Five and three lines outyielded check variety Rambilash under aerobic conditions in the 2008WS and 2009WS, respectively (Tables 9 and 10). AWD yield trial. One AYT of 16 entries was evaluated in three replications at NRRP-Hardinath under AWD in transplanting conditions during the 2009WS. Four lines with higher grain yield than checks Radha4 and Hardinath1, which are the most popular rice varieties among the farmers, were found: IR80411-B-49-1, IR05N-449, IR55435-5, and NR-1887 (Table 11). Participatory varietal selection under aerobic rice A PVS mother trial on direct-seeded aerobic rainfed rice was conducted during the 2008 and 2009WS to identify superior lines at two on-farm project sites, Bengadabar and Kishanpur. In PVS during the 2008WS at Bengadabar, four entries, B6144F-MR-6, RP2439, IR55435-5, and Vandana, produced higher grain yield than check Radha 4. Farmers liked these varieties because of their drought tolerance, higher grain yield, and disease and pest resistance (Table 12).

In PVS at Kishanpur during the 2008WS, four entries, IR80411-B-49-1, Anjali, WAB 272, and B6144F-MR-6, produced higher grain yield than the checks (Table 13). In PVS at Kishanpur during the 2009WS, five lines, IR80411-B-49-1, IR78913-B-22-B-B-B, IR84894-B, IR80991-B-336-4, and IR80416-B-32-B, produced higher grain yield than check Radha 4 (Table 14). IR84898-B had the lowest yield because of grain sterility. Among the tested genotypes, IR80411-B-49-1 and IR78913-B-22-B-B-B were preferred by farmers due to higher grain yield with drought tolerance, disease and pest resistance, good milling recovery after harvest, and their suitability to rice-wheat and rice-chickpea/lentil cropping systems.

In PVS at Bengadabar in the 2009WS, four lines, IR78913-B-22-B-B-B, IR80411-B-49-1I, IR80991-B-336-4, and IR80416-B-32-B, produced higher grain yield than check Radha4 (Table 15). However, IR80411-B-49-1 matured in 125 days, which was late in comparison with the popular farmer-preferred varietyHardinath1. Among the tested lines, most of the farmers liked IR80411-B-49-1, IR78913-B-22-B-B-B, and IR78937-B-B-B-B-1.

Participatory varietal selection under AWD A PVS mother trial of more than 120 days’ duration along with checks was carried out at Bengadabar and Kishanpur during the 2009WS.At Bengadabar, out of 12 entries under evaluation, five lines, IR80411-B-28-4, IR05N-449, IR55435-5, NR-1824, and NR-1887, outyielded Radha4 (Table 16). A majority of the farmers preferred IR80411-B-28-4, IR05N449, and IR55435-5 because of their drought tolerance, higher grain yield, grain type, and resistance to diseases and pests.

114

In PVS at Kishanpur in the 2009WS, 12 entries were evaluated with checks Radha 4 and Radha 11. Five lines, IR80411-B-28-4, NR-1824, IR80991-B-336-4, NR-1887, and NR-1893, were more promising than the check (Table 17).

Coordinated varietal trial It is mandatory to promote the promising genotypes in CVTs for multi-location tests in Nepal before submission for variety release. The most promising genotypes from AYTs, OYTs, and PVS were tested in a CVT. The genotypesIR78913-B-22-B-B-B, IR84894-B, IR80991-B-336-4, IR80416-B-32-B, and IR05N-449 were found outstanding for grain yield, drought tolerance, disease and insect resistance, good milling recovery, and good grain appearance under rainfed lowland aerobic conditions. Therefore, these genotypes were promoted for multi-location testing through a CVT. Among them, IR05N-449 and IR78937-B-B-B-B-1 are in the pipeline to be released. Two lines, B 6144F-MR 6 as Hardinath 2 and IR80411-B-49-1 were released as Tarahara 1 in 2012 in the project in Nepal and are being disseminated. Integrated weed management under aerobic conditions Weeds are the major problem in rice. They cause a yield reduction because of competition with water, nutrients, and sunlight provided to the rice crop. In recent decades, however, because of increasing labor cost, transplanting is being replaced by direct seeding and many herbicides are being used for effective weed control in direct-seeded conditions. Smith (1983) reported more than 350 weed species occurring in rice fields, listing 73 as economically important, with 15 and 36 classified as causing major and moderate yield losses, respectively. In certain conditions, the competitive pressure exerted by these weeds has resulted in a total loss of rice yield (Ampong-Nyarko and De Datta 1991). The common weeds exhibited in rice fields are Echinochloacrus-galli, E.colona, Cyeprus spp., Fimbristylismiliacae, Leptochloachinenesis, Cynodondactylon,etc. Weeds need to be removed from the field in the early stage of crop growth. Because of high weed competition, farmers are inclined toward the use of herbicides. However, hand weeding is the main practice in rice fields in Nepal but it is labor-intensive and costly. Weeds such as E. crus-galli and E.colonamimic rice and they are very difficult to identify at the younger stage and are also resistant to herbicide. In our weed control experiments, the highest grain yield was obtained with a spray of 2,4-D at 3 weeks after sowing (WAS) in 2007 and 2008 (Table 18). In addition, the interaction effect (varieties × weed control methods) was found significant for grain yield. Similarly, rice variety Mithila had higher grain yield than the other test genotypes (Table 19).Grain yield differed significantly among the treatments in all the tested rice genotypes. The highest grain yield was observed in weed-free plots, followed by hand weeding at 2WAS and 4WAS in all three tested rice varieties. B6144F-MR6 had showed the highest weed competitiveness among them (Tables 18 and 19). The statistical analysis showed a significant difference in observed traits except for tillers m−2 among the tested varieties at Bengadabar in 2009 (Tables 20 and 21). Similarly, hand weeding at 2 WAS and 4 WAS was found to be an effective method for weed control. Overall, the use of herbicides butachlor or 2,4-D plus one hand weeding at 5WAS for weed control was found as effective as two hand weeding at 3 and 5 WAS. Farmers were impressed with the control of weeds by herbicide because of the shortage and drudgery of labor

115

during weeding. Farmers have now started to use herbicide in the main season for rice for weed control because of its effectiveness and profitability.

Rice-based cropping systems under aerobic conditions An on-station trial at NRRP- Hardinath was conducted in the 2007-09WS under this research activity. Farmers are growing different succeeding crops in rice-based systems. But, they are not aware of suitable crops that can give them a high profit and also help to maintain field soil fertility. Therefore, different rice-based crop rotations were designed to validate a more profitable crop rotation that also enhances soil fertility. In the 2007WS, the total yield of a rice-maize cropping system was higher than that of rice-wheat. The second crop of rice was damaged because of severe drought (Table 22). In such a case, a drought-tolerant aerobic rice variety will produce more grain than a drought-sensitive one. In 2008, the total yield of rice-rice and rice-maize patterns was higher (Table 23). In the first rice crop, the variety was Hardinath 1, and in the second rice crop it was Makwanpur 1. In 2009, rice-maize followed by rice-rice produced higher total grain yield than a rice-wheat or rice-mung bean rotation (Table 24). Wheat in the dry season will require more water than mung bean in the summer season. Among the rotations, rice-maize and rice-rice were found to be more profitable than rice followed by wheat and mung bean. Quantifying water in the AWD system In Nepal, average rice yield in the rainfed region is only 1.5−2.5 t ha−1. Widespread and persistent rural poverty is a major problem in the country, particularly where farmers grow rice without irrigation, referred to as “rainfed rice.” Rice is the dominant crop in these areas because it is the only crop that can be grown. Drought is now becoming a serious problem for millions of rainfed rice farmers because of climate change. A trial was conducted at NRRP-Hardinath during the summer (chaite dhan) season in 2007, 2008, and 2009 to study the performance of rice varieties under the AWD system over the conventional irrigation system. The statistical analysis revealed a non significant difference in grain yield between alternate wetting and drying and the conventional irrigation method in 2007 and 2009 but the difference was significant in 2008 (Tables 25 and 26). However, the mean grain yield was found to be higher in AWD than in the conventional system in all the years. Overall, AWD saved 8−13% of the water compared with conventional irrigation under conditions when water input from rainfall was more than 1,000 mm over all three years. In that situation, the total water requirement of the crop has to be met by supplemental irrigation; AWD will provide more water savings than that observed in the current experiments. Policy issues Appropriate policy that provides top priority to increase rice production should be put in place in Nepal to meet the increasing demand for rice. In spite of the progress made so far, many constraints need to be overcome in the adoption of proven technologies in water-saving projects in Nepal. Major challenges to the large-scale adoption of water-saving technologies in the country are farmers' lack of awareness about available water-saving technologies, the traditional habit of farmers to cultivate rice using their own

116

knowledge, poor linkages among stakeholders, and inadequate government support. Similarly, the support of research and extension agencies will ensure the effective scaling-up of water-saving technologies among farmers in the country. The production environments of rice are diverse and are prone to drought; therefore, concerted efforts are needed in all areas for further improvement and better management. These can be met by adequate funds and commitments by government for the scaling-up of water-saving technologies to secure food and improve the livelihood of resource-poor farmers in rainfed areas.

Conclusions Rainfed rice area always faces the risk of water shortage where irrigation is not available. Therefore, rice production and productivity have remained low and stagnant in Nepal over the years. As a result, Nepal has become a large rice-importing country to meet current demand. More than 90% of the Nepalese people eat rice in their diet. The price of rice is increasing year by year. This is a major problem for poor and small farmers who spend most of their share of income to buy rice for daily use. The country can reduce imports and ensure food security by increasing yield and intensifying production through the AWD irrigation system in irrigated and rainfed transplanted areas and aerobic rice in hilly upland and shallow rainfed lowland irrigated and rainfed areas. The water-saving rice varieties B6144F-MR-6 and IR80411-B-49-1 were found to be outstanding in aerobic conditions. The rice varietiesB6144F-MR-6 and IR80411-B-49-1 were released with the names Hardinath-2 and Tarahara-1, respectively, in 2010 for general cultivation by farmers in Nepal. Some of the rice varieties found most promising under AWD conditionsareIR78937-B-B-B-B-1 and IR05N-449. For weed control, the use of butachlor at 1.5 kga.i.ha−1 as post emergence herbicide + one hand weeding 5 weeks after seeding was found effective. These technologies could be one of the options for increasing rice production and productivity of farm households in aerobic conditions. In addition, these technologies are cost-effective and also labor-saving for rice farming communities. The study demonstrated the efficiency of AWD to save 8−13% of water for irrigation without any yield decline even under conditions of a high amount of water available to rice through rainfall. For the extrapolation and dissemination of technologies for the wider community, the government and donor agencies should focus on the effects of these emerging technologies in the context of the ongoing process of climate change. Therefore, for the government and other supporting institutions need to upscale these successful technologies in wider areas in the country. Acknowledgments The authors are grateful to the involved staff, extension personnel, and farmers for their kind support during experimentation. We express our special thanks to ADB, IRRI, and NARC for their financial support and technical assistance for the successful completion of this project in Nepal. References Ampong-Nyarko K, De Datta SK. 1991. A hand book for weed control in rice. Manila

(Philippines): International Rice Research Institute.113 p.

117

APP (Agriculture Perspective Plan). 1999. Ministry of Agriculture and Cooperatives, Singh Durbar, Kathmandu Nepal.

Bhuiyan SI.1992. Water management in relation to crop production: case study on rice. Outlook Agric. 21:293-299.

CBS (Central Bureau of Statistics). 2002. National Planning Commission, Government of Nepal, Kathmandu. CBS Statistical Pocket Book, Nepal.

Chaudhary B, Yadav M, Akhtar T, Yadaw RB, Gharti DB, Bhandari D. 2004. Evaluation of rice genotypes for resistance to bacterial leaf blight in Nepal. In: Rice research in Nepal. Proceeding of the 24th summer crop workshop, 28-30 June 2004. Gumma MK, Gauchan D, Nelson A, Pandey S, Rala A.2011. Temporal changes in rice-growing area and their impact on livelihood over a decade: a case study of Nepal. Agric. Ecosyst. Environ. 142:382-392.

IRRI (International Rice Research Institute). 2003. Annual report. Los Baños (Philippines): IRRI. p 125-131.

MOAC (Ministry of Agriculture and Cooperatives).2010. Statistical information on Nepalese agriculture. Nepal Government Agricultural Business Promotion and Statistics Division, Singh Durbar, Kathmandu Nepal.

NPC (National Planning Commission). 2008. Three-year interim plan (2008/2009), Government of Nepal, Kathmandu.

Shrestha H, Akhtar K, Gautam AK, Tripathi BP. 2007. Comparison of rice production systems between two on-farm research sites in Dhanusha. In: Proceedings of the 8th National Outreach Workshop, 19-20 June 2007. National Agricultural Research Institute, Khumaltar. p 528-531.

Smith JR. 1983. Weeds of major economic importance in rice and yield losses due to weed competition. In: Proceedings on weed control in rice, 31 August-4 September 1981. Los Baños (Philippines): International Rice Research Institute. p 19-36.

Tripathi J, Bhatta MR, Justice S, Shakya NK. 2004. Direct seeding: an emerging resource in conservation technology for rice cultivation in the rice-wheat system. Proceedings of the Khumaltar, Lalitpur, Nepal Conference, 28-30 June 2004. Khumaltar, Lalitpur, Nepal.

Notes

Authors’ addresses: R.B. Yadaw and S.N. Sah, National Rice Research Project (NRRP), Hardinath, Nepal; R.K. Mahato, Regional Agricultural Research Station (RARS), Tarahara, Nepal; A. Kumar, International Rice Research Institute (IRRI), Los Baños, Philippines; B.P. Tripathi, Nepal-IRRI Office, Kathmandu, Nepal.

Figure 1. Drought area during 2006: comparative rainfall (mm) and normalized difference vegetation indices (NDVI) for a normal and severe drought year (2006).

118

Table 1. List of segregating generations from 2007 to 2009 under transplanted AWD conditions.

Designation Selected plants (no.) 2007—F3 2008—F4 2009—F5 IR83922-B 11 13 5 IR84880-B 9 4 2 IR84881-B 18 14 6 IR84882-B 9 16 5 IR84887-B 13 14 10 IR84894-B 16 23 12 IR84895-B 8 23 9 IR84896-B 10 4 2 IR84898-B 10 11 4 IR84899-B 12 19 6 IR84890-B 20 27 6 Total 136 168 67

119

Table 2. Grain yield and ancillary characteristics of selected genotypes (drought tolerance score 1) in OYTs under direct-seeded aerobic rainfed conditions at NRRP-Hardinath in the 2006WS.

Designation Days to

50% heading

Grain yield (t ha-1)

IR72667-16-1-B-B-3 75 3.0 IR78168-51-1-3-1-6 77 3.0 IR75417-R-R-R-R-267-3 76 3.0 IR80431-B-44-4 72 3.0 IR80413-B-10-3 75 3.0 IR80412-B-31-1 75 3.0 IR79328-126-3-1-2 84 3.0 IR80463-B-39-1 74 3.0 IR80411-B-49-1 87 3.0 IR78581-12-3-2-2 85 3.0 IR80930-B-17-4-1 76 3.0 IR81019-B-147-4-1 76 3.0 IR81022-B-493-1-1 69 3.0 IR81047-B-106-2-1 72 3.0 IR81022-B-333-1-2 77 3.0 IR81022-B-347-3-2 69 3.0 IR81047-B-106-4-3 70 3.0 Radha 4 82 2.0 Radha 11 88 2.5 Mean 74.4 1.8

120

Table 3. Grain yield and ancillary characteristics of selected genotypes (drought tolerance score 1) in OYT-first set under aerobic rainfed conditions at NRRP-Hardinath in 2007.

Designation Days to 50%

heading

Plant height (cm)

Phenotypic acceptabilit

y

Grain yield (t ha−1)

BP223E-MR-5 82 101 5 3.0 IR78875-207-B-1-B 81 103 1 3.2 IR78936-B-9-B-B-B 85 103 3 3.0 IR78937-B-20-B-B-4 79 101 3 4.2 IR78937-B-3-B-B1 81 103 3 3.0 IR80973-B-186-U1-2 79 102 5 3.6 IR80991-B-330-U1-1 82 103 3 3.0 IR05N379 82 102 3 3.0 IR05N386 87 97 3 3.0 IR05N412 86 103 5 3.0 IR05N419 85 94 5 3.2 IR05N449 91 89 3 3.4 IR05N455 91 99 1 4.0 IR80411-B-28-4 96 101 1 3.2 IR80411-B-49-1 96 99 3 3.6 IR80416-B-32-3 81 99 5 3.2 IR84894-B 72 112 3 3.2 Radha 11 115 110 5 2.3

Mean 77 105 4.2 2.2

Phenotypic acceptability: 1 = good, 3 = fair, 5 = poor

121

Table 4. Grain yield and ancillary characteristics of selected genotypes (drought tolerance score 1) in OYTs under aerobic rainfed conditions at NRRP-Hardinath in 2007.

Designation Plant height

(cm) Phenotypic

acceptability Grain yield

(t ha−1) APO 90 1 4.5 IR71203-10-CPA-4 92 1 3.8 IR80312-6-B 94 3 3.0 IR80411-B-49-1 93 1 4.0 IR81022-B-35-1-2 101 1 3.5 Radha 11 105 5 1.5 Mean 96 2.6

Phenotypic acceptability: 1 = good, 3 = fair, 5 = poor

122

Table 5. Ancillary characteristics of selected genotypes in OYTs under aerobic condition in Jute Research Program, Itahar, in 2007.

Vegetative vigor: 1 = very good, 2-3 = good, 3-5 = fair, 5-7 = poor, and 7-9 = very poor

Designation Vegetative vigor

Nematode galls (no.)

Days to heading

Plant height (cm)

Grain yield (t ha−1)

IR78875-207-B-1-B 6 14 88 56 2.6 IR78877-208-B-2 7 10 86 56 2.6 IR78908-193-13-3-B 3 8 73 61 3.4 IR79913-B-362-B-3 3 12 82 60 2.7 IR81039-B-173-U-3-3 3 9 80 56 2.9 IR80411-B-160-4 3 1 97 97 3.6 IR80411-B-28-4 1 1 96 96 3.7 IR80411-B-49-1 1 1 96 96 3.9 Kanchi Masuli 2 1 98 56 2.5 Mean 83 61 1.5

123

Table 6. Grain yield and ancillary characteristics of different genotypes in AYTs under aerobic rainfed conditions at NRRP-Hardinath in the 2007WS.

Designation Days to 50% headin

g

Plant height (cm)

Effective panicles

m−2

Fertile grains

panicle−1

1,000-grain

weight (g)

Grain yield

(t ha−1)

IR80411-B-49-1 99 102 249 121 29.1 4.2 B6144F-MR-6 90 101 135 130 26.1 2.0 IR75417-R-R-267 87 105 165 95 20.4 2.3 IR79328-126-3-1-2 99 98 217 132 24.8 3.6 IR70213-10-CPA4-

2 90 101 166 182 22.2 2.5

IR72667-16-1-B-B-3

89 89 213 113 21.9 1.5

IR78878-53-2-2-2 87 90 128 125 22.6 2.3 IR80430-B-73-3 91 83 206 106 30.0 2.9 IR80431-B-24-4 89 101 203 103 25.7 1.9 IR78581-12-3-2-2 96 93 234 156 28.6 3.8 Kachorwa 25 90 113 209 203 21.2 2.6 Kachorwa 26 97 119 247 113 21.3 3.3

Mean 92.1 101 197 131 24.5 2.7 CV (%) 1.79 3.75 21 14.4 5.1 18.3 LSD0.05 1.3 3.1 33 15.0 1.0 0.4

124

Table 7. Grain yield and its ancillary characteristics of different genotypes tested under AYTs (<120 days) during 2008 aerobic rainfed conditions at NRRP-Hardinath, Nepal.

Designation Days

to 50% headin

g

Plant height (cm)

Effective panicles

m−2

Fertile grains

panicle−1

1,000- grain

weight (g)

Grain yield

(t ha−1)

IR78937-B-20-B-B 66 89.2 354 79 28.4 2.7 IR78937-B-B-B-B-1 88 88.6 335 102 21.8 3.1 IR81044-B-B-112-42-

4 75 93.4 344 79 24.1 1.3

IR810430-B-B-149 75 115 305 90 23.2 1.5 IR04A305 86 94.4 297 66 22.3 1.4 IR84894-B 85 97.2 273 103 20.6 1.5 IR84898-B 75 98.0 277 95 20.4 1.7 IR84899-B 85 93.8 323 101 21.7 2.6 IR80013-B-141-4-1 72 100.0 407 102 22.7 2.9 IR78913-B-22-B-B-B 74 109.6 340 125 26.2 2.1 Hardinath1 65 85.2 349 82 22.5 1.6 Radha 4 86 96.3 292 65 25.3 2.3 Mean 77.7 96.7 322 89.9 23.3 2.1 CV (%) 1.31 6.8 23 9.2 3.9 17.6 LSD0.05 1.57 11.0 125 13.5 0.3 0.3

125

Table 8. Ancillary characteristics and grain yield of rice varieties evaluated in AYTs (<120 days) under aerobic rainfed conditions at NRRP-Hardinath during the 2009WS.

Designation Days to

50% heading

Plant heigh

t (cm)

Effective

paniclesm−2

Fertile grains

panicle−

1

Grain sterility

(%)

Grain yield

(t ha−1)

IR81429-B-31 80 116 272 84 20 2.5 IR80937-B-186-4 78 108 293 87 28 3.5 IR84898-B 80 97 190 93 20 3.6 IR78913-B-22-B-B-B-

4 79 110 235 99 21 4.1

OM-2516 80 86 230 96 27 3.5 OM-5796 80 86 233 102 24 4.1 IR78937-B-B-B-B-1 78 94 270 97 18 4.1 IR84899-B 80 101 255 105 17 4.0 IR80013-B-20-B-4 82 101 257 111 28 4.3 IR78937-B-20-B-B-4 79 96 258 97 31 3.7 Hardinath1 75 94 263 104 15 3.9 Radha4 82 90 244 97 18 3.3

Mean 79 98 250 98 22 3.7 CV (%) 2.9 7.8 17.8 11.8 31.1 17.6 LSD0.05 − 13.01 45.13 − − 0.61

126

Table 9. Grain yield and ancillary characteristics of different genotypes tested under AYTs (>120 days) under aerobic rainfed conditions during the 2008WS at NRRP-Hardinath, Nepal.

Designation Days to 50%

heading

Plant height (cm)

Effective

paniclesm−2

Fertile grains

panicle−1

1,000-grain

weight (g)

Grain yield

(t ha−1)

IR80973-B-186-4 86 100 254 103 20.7 2.6 IR80991-B-336-4 88 87 229 93 19.6 2.8 IR80411-B-28-4 111 87 264 83 24.3 1.9 IR80411-B-49-1 113 88 256 57 24.6 2.8 IR80416-B-32-3 86 100 232 115 21.7 1.6 IR05N 379 95 88 246 76 21.4 2.4 IR05N 449 100 83 352 74 22.1 3.0 IR05N 455 95 90 289 89 23.7 2.4 IR80900-B-17 100 135 220 107 25.1 2.1 IR75416-R-R-R-R-

156 106 104 330 118 18.0 2.0

Radha4 114 92 304 111 19.3 1.3 Rambilash 116 93 273 90 18.6 1.0 Mean 100.6 96 227 93 21.6 2.2 CV (%) 1.05 6.6 5.6 10.7 5 19.5 LSD0.05 1.46 9.4 16.3 1.5 0.7

127

Table 10. Ancillary characteristics and grain yield of rice varieties evaluated in advanced yield trials>120 days under aerobic rainfed conditions at NRRP-Hardinath during the 2009WS.

Designation Days to 50%

heading

Plant height (cm)

Effective panicles

m−2

Fertile grains

panicle−1

Grain sterility

(%)

Grain yield

(t ha−1)

IR81063-B-951-4-3 104 98 254 90 27 3.2 IR81040B-78-4-2 103 107 250 90 25 3.4 IR82098-B-B-3-1 104 118 244 103 28 2.8 IR82319-B-B-10-3-

2 103 105 220 78 28 3.5

IR74371-54-1-1 101 100 205 97 41 3.0 IR82318-B-B-35-2 101 114 261 90 27 3.7 IR05N-455 101 92 266 88 24 3.0 IR78399-157-3-6 105 114 231 115 28 3.2 IR82098-B-B-18-2 102 115 226 96 31 3.1 IR80991-B-336-4 105 115 288 100 27 3.3 IR05N-449 105 107 244 96 22 3.2 IR80416-B-32-3 105 98 213 98 17 3.6 IR83143-B-39-B-B 105 96 228 83 21 3.5 IR06G112 100 102 246 64 31 3.1 Radha11 105 124 216 91 31 2.6 Rambilash 107 119 222 105 31 3.4 Mean 104 107 243 93 28 3.3 CV (%) 2.4 12.4 25.4 18.9 11.1 23.1 LSD0.05 − − − 9.3 9.2 0.24

128

Table 11. Ancillary characteristics and grain yield of rice varieties evaluated in AYTs>120 days under AWD in transplanted conditions at NRRP-Hardinath during the 2009WS.

Designation Days to 50% headin

g

Plant height (cm)

Effective panicles

m−2

Fertile grains

panicle−1

Grain sterility

(%)

Grain yield

(t ha−1)

IR05N-379 101 101 290 134 24 3.8 IR05N-449 109 101 262 130 24 4.1 IR05N-455 109 101 211 111 32 3.5 IR80900-B-17 99 152 265 106 27 3.2 IR80411-B-49-1 101 101 326 151 20 4.4 IR80991-B-336-

4 99 104 269 101 29 3.8

NR-1824 108 122 329 121 30 3.8 NR-1887 106 105 288 111 16 4.1 NR-1893 105 102 244 132 17 3.6 IR55435-5 99 104 235 131 31 4.0 Radha 4 96 102 255 138 23 3.9 Hardinath 1 109 122 272 108 31 3.4

Mean 103 110 271 123 28 3.8 CV (%) 1.7 4.2 16.9 8.7 12.9 8.5 LSD0.05 3.04 7.85 − 18.14 − 0.34

129

Table 12. Mean grain yield and ancillary characteristics in a PVS mother trial under aerobic conditions during the 2008WS at Bengadaber, Mahotari.

Designation Days to maturity

Plant height (cm)

Effective panicles m−2

Grain yield (t ha−1)

IR80411-B-49-1 95 81 208 1.5 B6144F-MR-6 115 86 404 3.5 IR79328-126-3-1-

2 132 93 300 1.2

IR78581-12-3-2 125 69 312 2.5 Kachorwa 26 125 90 316 2.0 NR1824-21-2-1-1-

1 124 96 300 2.0

BRRIdhan 26 99 62 316 1.5 RP2439 122 70 264 3.5 WAB 272 112 75 340 2.5 IR55435-5 107 68 316 3.0 Anjali 86 65 260 2.5 PR101 86 65 258 2.0 Vandana 86 74 244 3.0 Hardinath1 88 64 232 2.5 Radha4 115 72 280 3.5 Ghaiya 2 86 64 200 1.5

130

Table 13. Grain yield and ancillary characteristics in a PVS mother trial under aerobic conditions during the 2008WS at Kisanpur, Dhanusha.

Designation Days to maturity

Plant height (cm)

Effective panicles

m−2

Grain yield (t ha−1)

IR80411-B-49-1 91 120 228 3.1 B6144F-MR-6 117 121 326 2.6 IR79328-126-3-1-2 118 92 386 2.1 IR78581-12-3-2 117 85 512 2.1 Kachorwa 26 119 105 560 2.1 NR1824-21-2-1-1-1 114 110 421 2.3 BRRIdhan 26 107 87 508 2.3 RP2439 100 75 472 1.9 WAB272 116 97 294 2.7 IR55435-5 108 114 290 2.2 Anjali 93 93 290 2.6 PR101 92 83 330 2.1 Vandana 97 112 350 2.2 Hardinath1 100 102 356 2.9 Radha4 114 94 340 2.9 Ghaiya 2 99 85 415 2.0 Mean 106 98 380 2.4 CV (%) 0.9 0.9 7.6 2.4 LSD0.05 2.00 1.94 59.71 0.12

131

Table 14. Ancillary characteristics and grain yield of rice varieties evaluated in a participatory varietal selection trial <120 days under aerobic rainfed conditions at Kishnpur, Dhanusha, during 2009.

Designation Days to

50% heading

Plant height (cm)

Effective panicles

m−2

Fertile grains

panicle−

1

Grain sterility

(%)

Grain yield

(t ha−1)

IR80411-B-49-1 95 104 263 160 11 4.2 IR78937-B-B-B-B-1 92 104 255 143 27 3.8 IR80013-B-141-4-1 94 113 251 83 24 3.9 IR84899-B 91 106 282 96 20 3.4 IR78937-B-20-B-B-4 93 109 265 113 16 3.7 IR78913-B-22-B-B-B 93 114 259 118 28 4.1 IR84898-B 90 92 235 73 37 3.2 IR84894-B 99 116 261 160 27 4.1 IR80991-B-336-4 96 107 238 155 15 4.1 IR80416-B-32-B 98 99 265 148 18 4.1 Radha4 95 102 252 147 15 3.7 Hardinath1 75 98 263 127 12 3.8

Mean 95 105.3 257 126 21 3.8 CV (%) 2.3 11.6 6.2 21.16 13.6 7.5 LSD0.05 3.71 - 27.08 9.90 11.50 0.48

132

Table 15. Grain yield and other characteristics in PVS in aerobic rainfed conditions at Bengadabar, Dhanusha, 2009WS.

Designation Days

to 50% headin

g

Plant height (cm)

Effective panicles

m−2

Fertile grains

panicle−

1

Grain sterility

(%)

Grain yield

(t ha−1)

IR80411-B-49-1 95 98 228 147 25 3.9 IR78937-B-B-B-B-1 92 84 304 130 25 3.4 IR80013-B-141-4-1 94 103 233 93 34 3.1 IR84899-B 91 95 248 104 22 3.1 IR78937-B-20-B-B-

4 93 111 225 133 22 3.4

IR78913-B-22-B-B-B

93 104 280 147 28 4.1

IR84898-B 90 92 224 98 38 2.9 IR84894-B 99 94 242 150 19 3.5 IR80991-B-336-4 96 96 258 147 21 3.8 IR80416-B-32-B 98 92 239 144 23 3.7 Radha4 95 107 246 151 21 3.7 Hardinath1 75 98 245 128 22 3.3

Mean 95 98 248 131 25 3.5 CV (%) 2.3 12.3 7.6 9.8 21.7 5.5 LSD0.05 3.71 − 31.89 21.85 9.18 0.32

133

Table 16. Grain yield and ancillary characteristics in PVS under AWD at Bengadabar, Dhanusha, in the 2009WS.

Designation Days to

50% heading

Plant height (cm)

Effective panicles

m−2

Fertile grains

panicle−

1

Grain sterility

(%)

Grain yield

(t ha−1)

IR05N-379 101 98 253 120 28 3.6 IR05N-449 109 100 259 110 33 3.9 IR05N-455 109 101 241 112 24 3.5 IR80900-B-17 99 157 262 115 34 3.3 IR80411-B-28-4 101 100 262 126 26 4.2 IR80991-B-336-

4 99 103 244 103 39 3.5

NR-1824 107 125 260 112 33 3.8 NR-1887 106 104 248 108 31 3.9 NR-1893 105 99 247 113 33 3.7 IR55435-5 99 104 242 120 37 3.8 Radha4 96 102 244 130 19 3.8 Radha11 109 119 263 111 31 3.3

Mean 103 109 253 115 31 3.7 CV (%) 1.7 3.3 6.1 8.9 15.5 6.7 LSD0.05 3.04 2.96 26.15 17.34 5.66 0.41

134

Table 17. Grain yield and ancillary characteristics in PVS under AWD at Kishanpur, Dhanusha, in the 2009WS.

Designation Days to

50% heading

Plant height (cm)

Effective panicles

m−2

Fertile grains

panicle−

1

Grain sterility

(%)

Grain yield

(t ha−1)

IR05N-379 101 101 278 118 30 3.3 IR05N-449 109 101 264 106 48 3.2 IR05N-455 99 101 250 115 29 3.2 IR80900-B-17 99 135 242 99 43 2.9 IR80411-B-28-4 101 101 290 145 17 4.2 IR80991-B-336-

4 99 104 255 105 39 3.5

NR-1824 108 122 256 112 34 3.5 NR-1887 106 105 252 123 31 3.4 NR-1893 105 102 249 111 33 3.4 IR55435-5 99 104 264 110 37 3.3 Radha4 96 102 262 110 30 3.2 Radha11 109 122 271 111 38 3.1

Mean 103 108 261 114 34 3.3 CV (%) 1.7 7.3 4.8 10.2 18.5 9.5 LSD0.05 3.04 13.43 21.11 19.74 6.18 0.53

135

Table 18. Grain yield and weeds in different treatments at NRRP-Hardinath under aerobic rainfed conditions in the 2007WS.

Treatment

Grain yield (t ha−1 ) Weed (dry weight, g m−2)

Loktantra Mithila Radha-11

Mean Loktantra Mithila Radha-11

No weeding: control

1.2 1.1 1.2 1.2 58.0 75.0 61.0

One H.W. at 3 weeks

3.3 3.4 3.5 3.4 37.1 28.0 32.0

Two H.W. at 3 and 5 weeks

3.6 3.7 3.5 3.6 10.3 12.1 11.0

Butachlor spray

3.2 3.6 3.6 3.5 17.0 22.0 15

Butachlor + one H.W. at 5 weeks

3.8 3.7 3.7 3.7 5 7 6.1

2,4-D spray at 3 weeks

3.1 3.2 3.1 3.2 15 17 14

2,4-D spray at 3 weeks + one H.W. at 5 weeks

3.2 3.3 3.3 3.3 10 11.0 8.0

Mean 3.1 3.2 3.1 3.3 LSD0.05 0.16 CV (%) 5.6

136

Table 19. Grain yield and weed biomass in a weed management trial at NRRP-Hardinath under aerobic rainfed conditions during 2008.

Treatment Varieties (grain yield, t ha−1) Varieties (dry weed

biomass, g m−2)

B6144F-MR-6

IR80411-B-28-

4

Radha-11

Mean

B6144F-MR-

6

IR80411-B-28-

4

Radha-11

No weeding 0.7 0.9 1.0 0.85 237 157 125 1 H.W. at 3

weeks 1.5 1.2 1.3 1.34 115 110 195

2 H.W. at 3 and 5 weeks

1.3 1.5 1.8 1.54 68 30 28

Butachlor 2.0 1.4 2.3 1.98 51 53 37 Butachlor+1 H.W.

at 3 weeks 1.8 2.1 1.8 1.92 21 41 19

2,4-D 2.3 1.6 2.4 2.10 48 44 49 2,4-D+1 H.W. at

5weeks 2.8 2.2 3.0 2.64 33 24 28

Mean 1.81 1.56 1.94 1.75 CV (%) 4.9 LSD0.05 671.

0

137

Table 20. Grain yield (t ha−1) of rice varieties in a weed management study under aerobic rainfed conditions at NRRP-Hardinath in 2009.

Treatment B6144F-MR-

6 IR80411-B-28-4 Radha-

11 Mean

No weeding 1.6 1.2 1.2 1.3 H.W. at 3WAT 3.3 3.3 2.3 3.1 Herbicide at 2WAT 3.4 3.1 3.1 3.2 H.W. at 3WAT and 5WAT 3.4 3.5 3.2 3.4 Herbicide at 2WAT and

4WAT 3.3 3.0 3.0 3.1

Weed-free 3.5 3.9 3.3 3.6 Mean 3.1 2.9 2.7 2.9 LSD0.05 0.224 CV (%) 4.7

WAT =weeks after transplanting.

138

Table 21. Dry weed weight (g m−2) in weed management study under aerobic rainfed conditions with different varieties at NRRP-Hardinath during 2009.

Treatment B6144F-MR6 IR-80411-B-49-1 Radha-11 3WAT 5WAT At

harvest Total 3WAT 5WAT At

harvest Total 3WAT 5WAT At harvest Total

No weeding − − 650 650 − − 715 715 − − 705 705 H.W. at 3WAT 40 − 72 112 45 − 79 124 41 − 75 116 Herbicide at

2WAT − − 165 165 − − 185 185 − − 162 162

H.W. at 3WAT and 5WAT

42 12 8 62 39 13 7.8 59.8 39 18 11 58

Herbicide at 2WAT and 4WAT

− − 124 124 − − 127 127 − − 123 123

Weed-free − − − − − − − − − − − −

139

Table 22. Grain yield of a rice-based crop rotation under aerobic conditions at NRRP-Hardinath in the 2007WS.

Rice-based crop Yield (t ha−1) of

first crop (rice) Yield (t ha−1) of

second crop Rice equivalent

yield (t ha−1)

Total yield (t ha−1)

Chaite rice - kharif rice 3.4 0.7* 0.7* 4.1*

Kharif rice - rabi wheat 3.4 2.6 3.1 6.5 Chaite rice - kharif maize

3.3 5.1 4.1 7.4

Chaiterice - kharif mungbean

3.5 0.3 1.0 4.5

*Low yield because of severe drought. Depending on price fluctuation, rice equivalent yield has been taken as wheat × 1.2, maize × 0.80, and mung bean × 3.5.

140

Table 23. Grain yield of different crops grown under aerobic rice conditions during the 2008WS.

Rice-based crop Yield (t ha−1) of

first crop (rice) Yield (t ha−1) of

second crop Rice equivalent

grain yield (t ha−1)

Total grain yield (t ha−1)

Chaite rice - kharif rice 4.3 4.7 4.7 9.0 Kharif rice - rabi wheat 4.6 2.8 3.4 8.0 Chaite rice - kharif maize

4.2 6.1 4.9 9.1

Chaite rice - kharif mungbean

4.7 0.4 1.4 6.1

141

Table 24. Grain yield of different crops grown under aerobic rice conditions during the 2009WS.

First crop Grain yield

(t ha−1) Second crop Grain yield

(t ha−1) Rice equivalent

yield (t ha−1) Total yield

(t ha−1) Chaite rice 4.1 Kharif rice 4.3 4.3 8.4 Kharif rice 3.9 Wheat 3.2 3.9 7.8 Chaite rice 4.2 Kharif maize 6.5 5.2 9.4 Chaite rice 4.4 Kharif mungbean 0.8 2.8 7.2

142

Table 25. Performance of rice varieties in different irrigation systems under AWD in transplanting conditions at NRRP-Hardinath during the 2007-08WS.

2007 2008

Variety Grain yield (t ha−1) Mean Variety Grain yield (t ha−1) Mean

Irrigation system

Irrigation system

AWD Conventional

AWD Conventional

Sabitri 3.4 3.3 3.35 Sabitri 3.5 3.7 3.6 Mithila 2.8 2.7 2.75 Mithila 3.5 3.7 3.6 Hardinath1 3.0 3.1 3.15 Hardinath 1 4.1 4.1 4.1 Radha4 3.2 3.1 3.15 Radha-4 3.7 3.8 3.7 Mean 3.10 3.7 LSD0.05 − 0.1 CV (%) 0.9 1.4 AXB NS NS

143

Table 26. Performance of rice varieties in different irrigation systems under AWD in transplanting conditions at NRRP-Hardinath in 2009.

Variety Grain yield (t ha−1) Mean

(t ha−1) Irrigation system AWD Conventional

IR80411-B-49-1 4.3 4.2 4.3 IR78397-B-B-B-B

4.2 3.9 4.0

Sabitri 4.1 4.2 4.1 Mean 4.2 4.1 4.1 CV (%) 6.1

144

Table 27. Water savings under conventional irrigation and alternate wetting and drying systems in rice cultivation at NRRP-Hardinath during the 2007 WS.

CSW =continuous standing water.

Treatment Water applied (mm) Water saved compared with CSW

(%) Irrigation Rainfall Total CSW system 245 1,086.4 1,331.4

AWD system 134 1,086.4 1,220.4 8.34

145

Table 28. Water savings under conventional irrigation and alternate wetting and drying systems in rice cultivation at NRRP-Hardinath during 2008.

Treatment Water applied (mm) Water saved compared with CSW (%)

Irrigation Rainfall Total CSW system 400 1,162 1,562

AWD system 214 1,162 1,376 11.94

146

Table 29. Water savings under conventional irrigation and alternate wetting and drying systems in rice cultivation at NRRP-Hardinath during 2009.

Treatment Water applied (mm) Water saved compared with CSW (%)

Irrigation Rainfall Total CSW system 415 1,055.6 1,470.6

AWD System 225 1,055.6 1,280.6 12.92

147

Figure 1. Drought area during 2006: comparative rainfall (mm) and normalized difference vegetation indices (NDVI) for a normal and severe drought year (2006).

Source: Gumma et al (2011).

148

Paper 7 Aerobic rice perspectives in Bangladesh: progress and challenges

H.U. Ahmed, K.M. Iftekharuddaula, M. Maniruzzaman, M. Shahidul Islam, and A.K.M. Zakaria

In Bangladesh, aerobic rice cultivation is possible in all three rice-growing seasons: aus, aman, and boro. Aerobic rice can be a suitable alternative to increase rice production in the water-short, drought-prone northwest region of the country. In the Asian Development Bank-supported project “Developing and disseminating water-saving rice technology in South Asia,” aerobic rice lines and varieties possessing drought tolerance were provided from IRRI. These lines were evaluated in observational yield trials (OYTs) and advanced yield trials (AYTs) during the aman and boro season. PSBRc82 and IR74963-262-5-1-3-3 showed superior performance to the checks under direct-seeded continuous water supply, and aerobic rice conditions. Among released varieties, BRRIdhan29 outperformed the other checks under direct-seeded and transplanted conditions during the boro season. Rice cultivation under aerobic conditions in soil saturated to field capacity (FC) resulted in water savings of 29% over continuous flooded conditions without any significant yield decline. The average water productivity was 0.79 kg m−3 in FC treatments compared with 0.55 kg m−3 in continuous flooded conditions. Aerobic rice line IR74963-262-5-1-3-3 showed maximum yield under FC conditions. The use of herbicide (Refit 25 EC) reduced the growth of different weed species, especially the sedge group, and herbicide combined with two hand weedings was found effective in managing weeds in aerobic conditions. There is a need to evaluate and disseminate selected aerobic lines under on-station and on-farm testing during all three seasons.

Rice requires two to three times more water than other cereals because of its higher water requirement for cultivation (Barker et al 1998, Tuong et al 2005, Carriger and Vallée 2007). However, availability of water is decreasing worldwide because of competition from industrial and urban needs, which is one of the major drivers for the development of water-saving technologies for rice cultivation (Kumar and Ladha 2011). Aerobic rice is a relatively new production system in which rice is grown under nonflooded, nonpuddled, and nonsaturated soil conditions (Bouman 2001). Aerobic rice genotypes require less water for their cultivation and have higher water productivity (Atlin and Lafitte 2002, Castañeda et al 2002, Wang et al 2002). For the cultivation of aerobic rice, the field should be leveled and drained, and the soil needs to be kept under saturation without any standing water in the field. For rainfed conditions as well as for water-short irrigated conditions, aerobic rice genotypes may require drought tolerance (Atlin et al 2004). Farmers put in a lot of effort and use many inputs to produce rice and, with increasing fuel costs and depleting groundwater table, the money spent for irrigation water is increasing year after year(Kumar and Ladha 2011). Selected aerobic rice varieties will help to minimize production costs by a judicious use of irrigation water in rice production with little or no loss of productivity.

Bangladesh is predominantly a rice-growing country. About 10 million hectares of cultivable land are available for year-round rice cultivation. Three rice-growing seasons exist in Bangladesh, of which dry-season or boro rice (December-May) and wet-season or aman rice (June-November), both transplanted, are two prominent seasons covering more than 90% of the area and producing most of the rice in Bangladesh (www.knowledgebank-brri.org/riceinban.php). The third season, aus, is

149

divided into both transplanted aus (April-first week of August) and direct-seeded aus (March-second week of August), and it covers the remaining 10% of the rice area. Though rice production in the aus season is lower than in the other two seasons (www.knowledgebank-brri.org/riceinban.php), some areas have no alternative but to cultivate aus rice. The direct-seeded rice probably matches with the aerobic rice but the basic difference is that direct-seeded rice is grown in Bangladesh under totally rainfed conditions whereas aerobic rice needs irrigation. This aus season seeding starts at the end of March and the crop is harvested between early July and mid-August depending on the start of the monsoon. Until the mid-1990s, the direct-seeded rice area was about 1 million ha, but now it has declined to less than 0.5 million ha because of its low productivity. The country produces 30.19 million tons of rice (BBS 2009), but, to feed the increasing population, which is likely to surpass 200 million by2025, Bangladesh must produce more rice. In the aman season, flood and drought are the two abiotic stresses that cause huge losses to rice production almost every year. In low-lying areas, floods regularly occur at the initial stage. Similarly, in Rajshahi and adjoining areas, occurrence of drought at the seedling stage due to failure or delay of early monsoon rain or drought at the reproductive stage due to early withdrawal of monsoon by the third week of September causes a yield reduction to the rice crop. Cultivation of early-duration drought-tolerant varieties of 105 to 115 days’ duration under direct-seeded aerobic conditions may help farmers to harvest a good yield.

At the beginning of 2000, the International Rice Research Institute (IRRI) started a research program to develop aerobic rice varieties. So far, IRRI-developed aerobic rice lines with short duration have shown an advantage in avoiding late-season drought in mid-October when the transplanted aman rice is in the booting or flowering stage in the drought-prone northwestern region (Salam et al1997).

The water requirement is very high in transplanted rice for both for the aman and boro season. However, in the aman season, rice receives a large portion of water from rainfall. In the boro season, mostly underground water sources such as deep wells and open wells are used to irrigate the rice crop. Large and repeated use of underground water has brought different problems, including depletion of water-table depth, resulting in unavailability of underground water and salinity problems. To achieve high productivity to meet the increased food requirement, farmers are not willing to leave transplanted rice cultivation and look for other alternatives.

Because of light soil texture and the prolonged winter season, the soil of northwestern Bangladesh is suitable for potato and wheat cultivation during the dry season. One of the major cropping patterns of this region in T. aman is potato-fallow/late boro or early aus. Harvesting the potato delays the planting of boro rice in late February toward March. After the harvesting of potato, about 80% of the land is being used for early aus cultivation. Early-maturing aerobic rice varieties can be suitable for cultivation in this short season (Ahmed et al 2008).

In 2003-04, BRRI received 22 aerobic rice genotypes under IRRI-BRRI collaboration, and studies were undertaken to explore the suitability of these lines in northwestern Bangladesh. Our study was undertaken to find out suitable early-maturing lines that can give higher yield under aerobic conditions during T. aman and

150

the boro season also and to select some lines that might perform well after the potato harvest in the aus season (Ahmed et al 2007). Breeding aerobic rice varieties At BRRI, experiments were conducted from 2007 to 2009 under the Asian Development Bank (ADB)-supported project “Developing and disseminating water-saving rice technologies for South Asia.”BRRI received segregating materials as well as advanced lines for evaluation from IRRI, Philippines. The development of breeding material A crossing program was undertaken at BRRI during the2007 wet season (WS). F1’s from six crosses (Table 1) were grown during the 2008 dry season (DS).The six F2populations were grown on the BRRI farm, Gazipur, during the 2008WS (T. aman). Each progeny was transplanted in a 5.4-m-long single-row plot maintaining a spacing of 25 × 15 cm. A single seedling was used per hill. The seedlings were transplanted in a lowland field maintaining continuous water for maximum phenotypic expression. The experiment was seeded on 11 July 2007 and planted on 10 August 2007. Fertilizer doses were 80:60:40:100:10 kg N, P2O5, K2O, gypsum, and ZnSO4 ha−1. All the fertilizers were applied during final land preparation except for N, which was applied in two splits as topdress. Weeding and other cultural practices were done as and when necessary. A total of 228 progenies were selected based on plant type, grain type, grain color, tolerance of lodging, and high yield potential (Table 1). Observational yield trial, 2007WS (T. aman) A set of 26 entries along with four checks were direct-seeded during the 2007WS on 21 July 2007 to evaluate line performance. The plot size for each entry was 4 rows of 5-m length with 25-cm line-to-line and 20-cm hill-to-hill spacing. Fertilizer was applied as in experiment 1. The plot was maintained under aerobic management. Out of 26 lines, only 10 entries survived and grain yield was recorded. Among the 10 entries, OM2718 and AS996 showed a higher yield performance than the best check (Table 2). (i) OYT 2008DS (boro) A total of 133 lines consisting of two local checks were evaluated under irrigated and aerobic conditions in an augmented design. The experiment was seeded on 8 January 2008 for direct-seeded aerobic conditions and a nursery raised for transplanted conditions on the same date. The seedlings were transplanted on 13 February 2008. The seedlings of each genotype were transplanted at two to three seedlings/hill with a spacing of 20 20 cm. In the irrigated field, 3 to 5 cm of continuous water was maintained, whereas water was applied in the aerobic field 5 days after the disappearance of water. Fertilizer doses were 100:80:60:100:10 kg N, P2O5, K2O, gypsum, and ZnSO4 ha−1. All the fertilizers were applied during final land preparation except for nitrogen. Nitrogen fertilizer was applied in three splits. Weeding and other cultural practices were done as and when necessary. Only 17 entries were selected from the irrigated field based on plant height and grain yield (Table 3) and 14 entries were selected from the aerobic field (Table 4). Grain yield was high in the aerobic trials. Six of the lines were common in both the aerobic and transplanted trials. In

151

irrigated transplanted conditions, the yield of the selected entries was on a par with that of check variety BR 28. However, under aerobic conditions, many breeding lines such asBP223E-MR-5, IR71525-19-1-1, and IR79907-B-425-B-4 yielded more than 1.0 t ha−1 higher than check variety BRRI dhan 28.This suggests that high-yielding lines rather than transplanted rice can be selected under aerobic conditions. (ii) OYT (boro) 2009DS A total of 237 lines in two sets (105+132 entries) of IRRI origin, with local checks, were evaluated under dry direct-seeded aerobic conditions. The experiment was seeded on 1December 2008 maintaining a row-to-row distance of 25 cm. Fertilizer dose and weeding were the same as in experiment 3. Under aerobic conditions, 11 entries were found to yield more than 4.0 t ha−1 (Table 5), whereas check variety BRRI dhan 28 totally failed to establish and it produced no grain. The highest yield was obtained from IR79477-65-3-1-1 (4.89 t ha−1) and IR77298-12-7-15 (4.88 t ha−1; Table 5). Advanced yield trial (AYT) 2008DS (boro) Rice genotypes with short and medium growth duration having early vegetative vigor and yield potential under irrigated and aerobic conditions were evaluated in a replicated trial with 24 entries, including four local checks (Table 6). The experiment was established in four sets: (1) direct-seeded rice maintained by continuous water supply, (2) direct-seeded rice maintained aerobically, (3) transplanted rice maintained by continuous water supply, and (4) transplanted rice maintained aerobically.

The experiment was seeded on 16 December 2007. The seedlings for each genotype were transplanted on 28 January 2008 at two to three seedlings per hill with a spacing of 20 20 cm in an RCBD with three replications. Fertilizer doses, weeding, and other cultural practices were the same as in experiment 1. The amount of water applied at each irrigation and groundwater table (2-m PVC pipe) were recorded every day and weekly rainfall was recorded in the aerobic field (Figures.1, 2, and 3).

The grain yield data are shown in Table 6.The amount of water saved in the aerobic field in transplanting and the direct-seeded experiment was 32% and 35%, respectively. The results clearly indicated that PSBRc 82 performed better in all four conditions than BRRI dhan 28. IR74963-262-5-1-3-3 can be a potential line under aerobic conditions. PSBRc82 seems to be promising for both aerobic and irrigated transplanted conditions, equally in the dry and wet season in Bangladesh. The performance of BRRI dhan 29 was found to be good in all conditions. Water management strategy for aerobic rice and water savings Water-use efficiency and weed management of aerobic rice A study was conducted on the BRRI farm, Gazipur, during the boro season, 2007-08, under aerobic conditions to find out the water requirement of promising lines under different water regimes, to determine efficient weed management practices, and to identify water-use-efficient genotypes for better yield performance. These lines were IR72176-140-1-2-2-3, IR74963-262-5-1-3-3, IR72022-462-3-3-2, IR772176-140-1-2, and IR77073-B-35-1-1, andBRRIdhan36 as a check. The recommended fertilizer dose was 150:30:75:18:5 kg N, P, K, S, Zn ha−1. All fertilizers except for N were applied as basal before the last plowing. N fertilizer was applied as topdress with three equal

152

splits at 25, 40, and 55 days after sowing (DAS). Three water treatments were used: continuous standing water (1 to5 cm), water applied at field capacity (FC), and water applied at 70% of FC. Two weed management practices were used: preemergence herbicide (Refit 25 EC) within 3 DAS with two hand weedings (at 35 and 55 DAS), and three hand weedings at 30, 45, and 60 DAS. Perforated PVC pipe of 150-mm diameter was installed at a depth of 40 cm below ground surface in the field for monitoring water-table depth. Water depth, rainfall, and evaporation were measured. Moisture conditions were monitored daily before the irrigation treatment to determine the FC and 70% of FC. The design of the field experiment was a split-split plot with water treatment in the main plot, weed management in the subplot, and variety in the sub-subplot.

The amount of applied irrigation water was measured by the volumetric method following the water management treatments. Rainfall and percolation data during the crop growing period were recorded through a rain gauge and percolation gauge, respectively. The depth of flooded water was recorded daily. A polythene sheet was placed along the boundary line of each plot to protect from seepage and fertilizer loss. Data on yield and yield-contributing traits were recorded. The data were analyzed statistically using IRRISTAT software package and least significant difference (LSD) value was used to determine the significance levels among the treatment means. Water productivity was also calculated as the weight of grain per unit of total water used from transplanting to harvest for each treatment.

The initial soil properties of the experimental plot were analyzed. The field soil texture varied from silt clay to clay loam, with pH 6.3 to 6.8 and low organic carbon from 0.71 to 1.2. The chemical properties indicated that the total nitrogen content was 0.05 to 1.1%, phosphorus ranged from 5.4 to 16 ppm, zinc from 0.70 to 2.2 ppm and potassium from 0.21 to 0.30 meq 100 g−1 soil.

Rainfall, pan evaporation, and field water level fluctuation patterns from seed sowing to harvest are shown in Figures 1 to 3. The data indicated that rainfall during the crop growing period was 240 mm and well distributed. The water depth pattern varied for all three water regimes (Figure 1). Maintenance of continuous standing water (CSW) required maximum irrigation water, that is, 732 mm. But, in other water treatments, 29% and 37% less water was applied for FC and 70% for FC treatments, respectively (Table 7). However, the 70% FC treatment required the lowest quantity of water (459 mm) compared with CSW. In the nonflooded period, the field water table dropped about 20 cm below the ground surface for the FC treatment, whereas it was about 30 cm below the ground surface for the 70% FC treatment. The average soil moisture content was 29% in the topsoil for the FC treatment and it was 21% for the 70% FC treatment. The total water input (rainfall + irrigation) ranged from 699 to 972 mm, of which 240 mm was rainfall. There was no drainage outflow because of low rainfall during the crop growing season. There was no significant difference in evapo-transpiration, but variation occurred in seepage and percolation losses because of nonflooded conditions of the aerobic treatments.

There was no significant yield difference among the interaction effect of irrigation treatments and genotypes, but among the genotypes significant variation occurred (Table 8), though the quantity of water applied varied from 699 to 972 mm. The highest grain yield was recorded in the FC treatment and after that a decreasing

153

trend was found (Table 8). There was no significant yield difference between the genotype and weed management (Table 9), and there was no significant interaction effect of weed management and irrigation treatments on grain yield (Table 10), though the weed biomass was always higher in the three hand-weeding treatments (Table 11).

Water productivity varied among the irrigation treatments and the yield difference was due to genotypes. Water productivity ranged from 0.56 to 0.87 kg m−3, with an average of 0.65 kg m−3, and it showed an increasing trend in the aerobic treatments, whereas, in FC and 70% FC, it was 0.84 and 0.87 kg m−3, respectively (Figure 3). Water savings in the 2009DS (boro) Rainfall, pan evaporation, and field water level fluctuation pattern from seed sowing to harvest are shown in Figure 4. The data indicated that rainfall during the base period of rice was 101 mm and was distributed up to the ripening period, which became congenial to rice production in the boro season and reduced irrigation frequency. The water depth pattern varied in all three water regimes (Figure 4). Maintenance of continuous standing water required maximum irrigation water and that was 790 mm. But, in other water treatments, 29% and 37% less water was applied from FC and 70% FC treatments, respectively (Table12). However, the 70% FC treatment required the lowest quantity of water (495 mm) compared with CSW. The highest field water depth at all times was recorded in the CSW treatment and the lowest in 70% FC. The average soil moisture content was 29% in the topsoil for the FC treatment and was 21% for the 70% FC treatment (Figure 4).

The total water input (rainfall + irrigation) ranged from 596 to 891 mm. Out of this, 101 mm was rainfall. There was no drainage outflow because of low rainfall during the crop growing season. There was no significant difference in evapo-transpiration, but variation occurred in seepage and percolation losses because of nonflooded conditions of the aerobic treatments (data not shown).

There was a significant yield difference for the interaction effect of irrigation treatments, weed management, and genotypes (Table13). The highest grain yield was recorded in CSW and a competitive yield was also found in treatments in which irrigation was given 3 days after water disappearance (DAWD). In 2009, a significant yield loss was observed under five DAWD treatments. There was a significant yield difference between the genotypes within the water management treatments but there was no difference among weed management treatments. In aerobic conditions, a slight decreasing yield trend was observed.

Water productivity varied among the irrigation treatments and the yield difference due to genotypes. Water productivity ranged from 0.53 to 0.74 kg m−3, with an average of 0.64 kg m−3, and an increasing trend was observed in FC treatments, with the value being 0.73 and 0.74 kg m−3 in FC and 70% FC, respectively (Figure 5).

Total water applied (mm), amount of water saved, and water productivity were analyzed over two consecutive boro seasons, 2007-08 and 2008-09, in an assessing water-use-efficiency experiment conducted in direct-seeded conditions. On average, 540 mm of water was applied in the FC treatment whereas it was 494 mm in the 70% FC treatment and water saved was 29% and 37% in FC and 70% FC treatments,

154

respectively. The average water productivity was as high as 0.81 kg m−3 in the 70% FC irrigation treatment while it was 0.79 kg m−3 in the FC treatment (Table 8). Weed control and management Weed species and weed biomass as influenced by weed management and irrigation treatment under aerobic direct-seeded rice (DSR) were evaluated during boro 2007-08.The different treatments were I1= continuous flooding, I2 = irrigation at 3 DAWD, I3 = irrigation at 5 DAWD, W1 = Refit at 7 DAT + 1 hand weeding (at 35 DAS), W2 = 3 hand weedings at 25, 50, and 70 DAS.

The population of weed biomass as well as the presence of different weed species was much higher at 65 DAS than at 30 DAS (Table15). Cynodon dactylon, Echinochloacrus-galli, and Kakpaya (grass); Cyperus rotundusand Scirpus sp. (sedge); and Kesoti, Helencha, and Photka (broadleaf) weeds were present. The experiment indicated that herbicide (Refit 25 EC) reduced the growth of different weed species, especially the sedge group, and irrigation applied at 3 DAWD favored weed growth.

At 30 DAS, there were no weeds in W1 (herbicide-treated plots) but higher weed biomass (2.1 g m−2) was observed in W2I2. At 45 DAS, both W1 and W2 showed higher weed biomass, although the biomass was much higher in W2 than in W1. The highest weed biomass (99.6 g m−2) was observed in W2I1, followed by W2I3 (Table 16). At 65 DAS, both W1 and W2 showed higher weed biomass, although the biomass was much higher in W2 than in W1. The highest weed biomass (90.4 g m−2) was observed in I2W2, followed by I2W1 (75.2 g m−2).The differences among irrigation treatments were less. This indicates that irrigation at field capacity and at 70% field capacity did not enhance weed infestation significantly. The dominant weed species were Fimbristylis miliaceae, Cyperus difformis, Echinochloa colona, Scirpus maritimus, C.iria, Kesoti, and some unknown broadleaf weeds. Some upland weeds were found in aerobic boro plots because the plots were converted to resemble the upland ecosystem. Policy issues It was mentioned earlier that aerobic rice culture is a new issue in rice production systems in Bangladesh. To increase awareness among farmers, it needs intensive research and on-farm validation. At the same time, the government and policymakers are still not aware of this system. Government officials, even extension workers, still express their confusion when discussing aerobic rice. Though scientists are fully aware of the water scarcity problems in the coming days, research at BRRI needs to reach farmers and extension workers, and this requires adequate funding support to demonstrate the technology to farmers in their fields. BRRI is pursuing continuous research on breeding aerobic rice varieties suited to target environments in Bangladesh and evaluating advanced lines in both on-station and on-farm experiments. To strengthen this research, strong and continuous support from the government and donor agencies is needed. The support from ADB under the project “Developing and disseminating water-saving rice technologies for South Asia” has provided in-depth knowledge, training, and capacity to undertake research on water-

155

saving technologies to overcome the problem of water shortage for sustainable rice cultivation. Conclusions Conventional approaches of rice cultivation under puddled conditions lead to heavy losses of water due to seepage and percolation and require more water for puddling and nursery raising. Aerobic rice is one of the water-saving technologies for rice cultivation that involves the cultivation of rice under nonpuddled aerobic conditions. This minimizes the water losses associated with continuous flooding of the field. In Bangladesh, aerobic rice cultivation is possible in all three rice-growing seasons, aus, aman, and boro. Aerobic rice is a more suitable alternative to increase rice production and productivity in the water-short, drought-prone northwest region of the country. It is also seen that varieties that are suitable for cultivation under irrigated transplanted conditions normally are not adapted to aerobic conditions. It is important that varieties with high yield potential and drought tolerance be developed for aerobic cultivation. In our study, improved breeding lines from IRRI were evaluated in observational yield trials and advanced yield trials during aman and boro seasons. Some of the breeding lines, such as PSBRc82 and IR74963-262-5-1-3-3, showed superior performance over the check varieties. BRRI dhan 29 also performed well in all the conditions tested and it may be evaluated in on-farm testing in Bangladesh. Rice cultivation under aerobic conditions resulted in water savings of up to 29% at field capacity over continuous flooded conditions without any significant yield decline. In addition, the use of herbicide combined with two hand weedings was found to be more effective in managing weeds under aerobic conditions.

References Ahmed HU, Ali MA, Yasmeen R, Salam MA. 2007. Performance of short duration

aerobic rice genotypes at drought prone ecosystem. Int. J. Biol. Res. 2(4):54-58. Ahmed HU, Ali MA, Sarker ABS, Salam, MA. 2008. Performance of short duration

aerobic rice genotypes after potato harvest. Int. J. Biol. Res. 4(5):72-75. Atlin GN, Lafitte HR. 2002. Developing and testing rice varieties for water-saving

systems in the tropics. In: Bouman BAM, Hengsdijk H, Hardy B, Bindraban PS, Tuong TP, Ladha JK, editors. Water-wise production. Proceedings of the International Workshop on Water-wise Rice Production, 8-11 April 2002, Los Baños, Philippines. Los Baños (Philippines): International Rice Research Institute.

Atlin G, Laza M, Amante M, Lafitte HR.2004. Agronomic performance of tropical aerobic, irrigated and traditional upland rice varieties in three hydrological environments at IRRI. In: New directions for a diverse planet: Proceedings of the 4th International Crop Science Congress, Brisbane, Australia, 26 Sept. - 1 Oct. 2004.

Barker R, Dawe D, Tuong TP, Bhuiyan SI, Guerra LC. 1998. The outlook for water resources in the year 2020: challenges for research on water management in rice production. In: Assessment and Orientation towards the 21st Century, 7-9September 1998. Proceedings of the 19th Session of the International Rice Commission. Cairo (Egypt): FAO. P 96-109.

156

Bouman BAM.2001.Water-efficient management strategies in rice production. Int. Rice Res. Notes 16(2):17.

BBS (Bangladesh Bureau of Statistics). 2009. Carriger S, Vallée D. 2007. More crop per drop. Rice Today 6(2):10-13. Castañeda AR, Bouman BAM, Peng S, Visperas RM. 2002. The potential of aerobic

rice to reduce water use in water-scarce irrigated lowlands in the tropics. In: Bouman BAM, Hengsdijk H, Hardy B, Bindraban PS, Tuong TP, Ladha JK, editors. Water-wise rice production. Proceedings of the International Workshop on Water-Wise Rice Production, 8-11 April 2002, Los Baños, Philippines. Los Baños (Philippines): International Rice Research Institute.

Hand Book of Agricultural Statistics, December, 1999, Dhaka, Bangladesh. www.knowledgebank-brri.org/riceinban.php Kumar V, Ladha JK.2011. Direct seeding of rice: recent developments and future

research needs. Adv. Agron. Volume 111, DOI: 10.1016/B978-0-12-387689-8.00001-1.

Salam MA, Ahmed HU, Hossain MA. 1997. Advanced breeding lines of rice for rainfed lowland drought prone environment in Bangladesh. Bangladesh J. Plant Breed. Genet. 10 (1and 2):13-18.

Tuong TP, Bouman BAM, Mortimer M .2005. More rice, less water: integrated approaches for increasing water productivity in irrigated rice-based systems in Asia. Plant Prod. Sci. 8:231-241.

Wang H, Bouman BAM, Zhao D, Wang C,Moya PF. 2002. In: Bouman BAM, Hengsdijk H, Hardy B, Bindraban PS, Tuong TP, Ladha JK, editors. Water-wise rice production. Proceedings of the International Workshop on Water-Wise Rice Production, 8-11 April 2002, Los Baños, Philippines. Los Baños (Philippines): International Rice Research Institute.

Notes Authors’ addresses: H.U. Ahmed, K.M. Iftekharuddaula, M. Maniruzzaman, and M. Shahidul Islam, Bangladesh Rice Research Institute (BRRI); A.K.M. Zakaria, Rural Development Academy (RDA), Bogra, Bangladesh. Figure 1. Rainfall, evaporation, and fluctuation patterns of water table in different water regimes during 2007-08 (boro). Figure 2. Soil moisture content of topsoil (0−15 cm) in different water regimes during

2007-08 (boro). Figure 3. Water productivity as influenced by water management during 2007-08

(boro). CSW = continuous standing water, FC = field capacity, 70% FC = 70% of field capacity.

Figure 4. Soil moisture content of topsoil (0−15 cm) in different water regimes during 2007-08 (boro).

Figure 5. Water productivity as influenced by water management during 2008-09 boro. CSW = continuous standing water, FC = field capacity, 70% FC = 70% of field capacity.

157

Table 1. List of progenies selected from F2 generation during T. aman, 2008.

Cross Selected progenies (no.)

IR64/Vandana 48 BRRIdhan39/IR76569-122-1-1-

3 60

IR73001-13-2-2/Vandana 40 IR76557-80-2-2-3/HO13-5-3-B4 10

BR16/Saita 26 Vandana/WAB-36-17-12-HB 44 Total 228

158

Table 2. Performance of dry direct-seeded lines grown in the 2007 WS (T. aman).

Designation Duration (days)

Plant height (cm)

Grain yield (t ha−1)

OM576 122 92 3.9 OM1490 125 89 3.2 OM2718 118 85 4.5 OM4498 126 80 4.2 OMCS2000 126 79 3.9 AS996 116 96 4.7 BRRI dhan28* 132 105 3.3 BRRI dhan42* 122 100 3.1 BRRI dhan43* 122 100 3.0 BRRI dhan45* 128 102 2.4

*Check.

159

Table 3. Selected entries from OYT (transplanted) during the 2007-08 DS (boro). Designation Growth duration

(days) Grain yield

(t ha−1) BP223E-MR-5 145 2.9 IR71525-19-1-1 144 2.4 IR78875-176-B-2-B 139 3.6 IR78875-190-B-1-3 141 2.8 IR78875-207-B-1-B 141 2.6 IR79907-B-425-B-4 142 2.4 IR79913-B-283-B-3 138 2.6 IR79971-b-201-2-4 140 3.4 IR79971-B-338-2-2 144 3.8 IR80508-B-57-4-B 147 3.9 IR80973-B-186-U4-2 148 2.6 IR81040-B-78-U2-1 150 2.4 IR72010-39-CPA-7-1-1-4-

2 142 2.4

IR80463-B-39-1 140 3.2 IRPSBRC82 138 2.2 IR72667-16-1-B-B-3 150 3.2 IR77298-14-1-2 146 2.4 BRRI dhan 28* 144 2.6

*Check.

160

Table 4. List of selected entries from OYT (aerobic) during the 2008 DS (boro). Designation Growth

duration (days) Grain yield

(t ha−1) BP223E-MR-5 145 5.0 IR71525-19-1-1 144 4.6 IR78875-176-B-2-B 139 3.8 IR79907-B-425-B-4 142 4.6 IR80973-B-186-U4-2 148 3.6 IR82299-B-306-B 140 3.4 IR70210-39-CPA-7-1-1-4 142 4.2 IRO5A233 (IR79478-67-3-3-

2) 137 2.5

IR80416-B-32-3 142 3.5 IR80420-B-22-2 144 2.9 IR68098-B-78-2-2-B-1 145 3.4 IR75417-R-R-R-R-267-3 147 3.0 BRRI dhan 28* 145 3.3

*Check.

161

Table 5. Selected entries in dry direct-seeding aerobic conditions, OYT, 2008-09 (boro).

Designation Days to

maturity Plant height (cm)

Grain yield (t ha−1)

IR81027-B-23-1-3 148 104 4.07 IR81027-B-88-3-4 146 107 4.07 IR81047-B-57-1-4 146 102 4.07 IR78555-3-2-2-2 146 112 4.07 IR79477-65-3-1-1 154 109 4.89 IR80420-B-22-2 146 112 4.07 IR81896-B-B-31 153 107 4.08 IR81896-B-B-443 153 110 4.06 IR81896-B-B-370 162 104 4.07 IR77298-12-7-15 149 69 4.88 IR83140-B-32-B-B 148 88 4.08 BRRI dhan 28* 153 − −

*Check, no yield and growth.

162

Table 6. Yield data of all the tested entries in four situations during the 2007-08 DS (boro).

Designation Grain yield (t ha−1) Average

DS-Aero DS-

CSW TP-Aero TP-CSW

IR8855-17 2.97 4.07 3.83 3.77 3.66 IR64 3.03 4.13 3.23 3.90 3.58 IR82853-36 3.23 3.40 3.47 3.67 3.44 IR82914-11 3.07 3.20 3.40 3.47 3.28 IR82851-23 3.33 4.67 3.43 3.87 3.83 IR72176-140-2-2-3 3.13 4.43 3.70 3.97 3.81 IR74963-262-5-1-3-3 4.00 5.17 3.83 3.90 4.23 IR73707-45-3-2-3 4.30 4.73 3.40 4.13 4.14 IR69515-KKN-4-UBN-4-2-1-

1 4.07 4.17 3.87 4.00 4.03 IR70179-1-1-1-B 3.77 4.73 3.37 3.47 3.83 IR72022-462-3-3-2 2.83 4.23 3.73 3.93 3.68 IR72176-140-1-2 3.43 3.73 3.60 3.83 3.65 IR74963-262-5-1-3-3 4.00 4.43 3.97 4.00 4.10 IR75003-95-5-1-3 3.13 4.13 3.33 3.73 3.58 IR77073-B-35-1-1 3.40 4.53 2.80 3.20 3.48 IR77030-B-34-1-1 3.23 3.73 4.13 4.13 3.81 IR77073-B-4-2-2 2.63 4.00 3.30 3.43 3.34 IR77080-B-6-2-2 3.20 4.13 3.43 3.60 3.59 IR77031-B-42-3-1 3.87 4.53 3.83 4.30 4.13 IR77034-B-4-3-3 3.27 4.07 3.63 3.80 3.69 IR74371-46-1-1 2.13 3.57 1.77 2.60 2.52 IR74963-262-5-1-3-3 3.70 4.63 3.60 4.13 4.02 IR62142-114-32-2-2

(PSBRc80) 3.40 4.73 3.27 3.53 3.73 IR64683-87-2-2-3-3

(PSBRc82) 4.13 4.07 3.90 4.43 4.13 BR14* 3.13 3.87 2.83 3.57 3.35 BRRI dhan36* 3.07 5.03 3.90 3.90 3.98 BRRI dhan28* 3.93 3.60 3.55 4.00 3.75 BRRI dhan29* 4.70 4.80 4.73 4.93 4.79

LSD0.05 0.96 1.07 0.77 0.76 *Check.

163

Table 7. Field-level water applied in different irrigation treatments during 2007-08 (boro).

Treatment Water applied (mm) Water saved compared with CSW

(%) Irrigation Rainfall Total

Cont. standing water (CSW)

732 240 972 −

Field capacity 520 240 760 29 70% of field capacity

459 240 699 37

164

Table 8. Grain yield (t ha−1) variation in different irrigation regimes during 2007-08 (boro).

Variety Irrigation regimes CSW FC 70% FC Variety mean

IR72176-140-1-2-2-3

4.03 4.19 4.04 4.09

IR74963-262-5-1-3-3

4.59 4.75 4.45 4.60

IR72022-462-3-3-2 4.01 4.23 3.61 3.95 IR772176-140-1-2 4.08 4.27 4.19 4.18 IR77073-B-35-1-1 3.80 4.51 3.91 4.08 BRRI dhan36 4.24 4.25 3.89 4.13 Irrig. mean 4.13 4.37 4.02 4.17 LSD0.05 0.69

CSW = continuous supply of water; FC = field capacity.

165

Table 9. Interaction effect of variety and weed management on grain yield (t ha−1) during 2007-08 (boro).

Variety Weed management

Herbicide + 2 hand weedings

Three hand weedings

IR72176-140-1-2-2-3 4.31 3.86 IR74963-262-5-1-3-3 4.92 4.28 IR72022-462-3-3-2 4.07 3.83 IR772176-140-1-2 4.07 4.29 IR77073-B-35-1-1 4.13 4.02 BRRI dhan36 4.21 4.04 LSD0.05 0.56

166

Table 10. Interaction effect of irrigation and weed management on yield (t ha−1) during 2007-08 (boro)

Irrigation regime Weed management

Herbicide + 2 hand weedings

Three hand weedings

Cont. standing water 4.20 4.05 Field capacity 4.54 4.20 70% of field capacity 4.12 3.92 LSD0.05 0.40

167

Table 11. Interaction effect of irrigation on weed biomass during 2007-08 (boro).

Irrigation regime Weed biomass (g) Herbicide + 2 hand

weedings Three hand weedings

Cont. standing water 10.25 15.66 Field capacity 12.32 18.55 70% of field capacity 13.57 22.82 Irrig. mean 12.05 19.01

168

Table 12. Field-level water applied and water saved in different irrigation treatments during 2008-09 (boro).

Treatment Water applied (mm) Water saved

compared with CSW (%)

Irrigation Rainfall Total

Cont. standing water 790 101 891 −

Field capacity 560 101 661 29 70% of field capacity 495 101 596 37

169

Table 13. Interaction effect on grain yield (t ha−1) of different varieties and weed management on different irrigation regimes during 2008-09 (boro).

Variety Irrigation regimes

CSW FC 70% FC Variety mean Herbicide + hand weeding

IR72176-140-1-2-2-3

4.21 4.11 3.63 3.98

IR74963-262-5-1-3-3

4.62 4.49 4.04 4.38

IR72022-462-3-3-2 4.16 4.01 3.53 3.90 IR772176-140-1-2 4.13 4.10 3.46 3.90 IR77073-B-35-1-1 3.93 3.77 3.59 3.76 BRRIdhan36 4.25 4.19 3.67 4.04 Irrig. mean 4.22 4.11 3.65 3.99

Two hand weedings IR72176-140-1-2-2-3

4.08 3.96 3.64 3.89

IR74963-262-5-1-3-3

4.68 4.48 3.67 4.28

IR72022-462-3-3-2 3.99 3.89 3.47 3.78 IR772176-140-1-2 3.95 3.84 3.44 3.74 IR77073-B-35-1-1 3.72 3.63 3.53 3.62 BRRI dhan36 4.28 4.03 3.50 3.93 Irrig. mean 4.12 3.97 3.54 3.88 LSD0.05 0.11

170

Table 14. Total water applied, water saved, and water productivity in different irrigation regimes.

Treatment Boro 2007-08 Boro 2008-09 Average

Total water applied (mm) CSW 732 790 761 FC 520 560 540 70% FC 459 495 477

Water saved FC 29 29 29 70% FC 37 37 37

Water productivity (kg m−3) CSW 0.56 0.53 0.55 FC 0.84 0.73 0.79 70% FC 0.87 0.74 0.81

171

Table 15. Weed species and weed biomass as influenced by weed management and irrigation treatment in the direct-seeded rice ecosystem in 2007-08 (boro).

Rep./treat. Number of weed species present m−2 Biomass (g

m−2) Grass Sedge Broadleaf Mixture

30 DAS I1W1 4 4 4 72 1.20 I1W2 16 8 48 72 1.44 I2W1 4 4 12 52 1.12 I2W2 12 4 44 112 2.07 I3W1 − − 12 56 0.88 I3W2 16 16 24 64 1.25

65 DAS I1W1 4 36 40 4 11.70 I1W2 12 172 68 4 27.17 I2W1 12 116 48 4 9.24 I2W2 4 256 108 12 28.24 I3W1 24 40 36 − 7.93 I3W2 4 240 72 8 22.43

I1 =continuous flooding, I2 =irrigation at 3 days after water disappearance, I3 =irrigation at 5 DAWD, W1 = Refit at 7 DAT + 1 hand weeding (at 35 DAS), W2 = 3 hand weedings at 25, 50, and 70 DAS.

172

Table 16. Weed species and weed biomass as influenced by weed management and irrigation treatment under direct-seeded rice during 2008-09(boro).

Treatment Number of weed species present m−2 Biomass (g

m−2) Grass Sedge Broadleaf 30 DAS

I1W1 0 0 0 0 I1W2 47 120 107 1.8 I2W1 15 0 0 0 I2W2 64 79 188 2.1 I3W1 0 0 0 0 I3W2 35 71 67 0.84

45 DAS I1W1 80 360 81 44.8 I1W2 48 552 263 99.6 I2W1 93 200 71 50.0 I2W2 67 420 321 79.6 I3W1 100 153 93 70.4 I3W2 64 596 293 92.4

65 DAS I1W1 76 496 109 52.0 I1W2 72 388 68 54.8 I2W1 72 464 33 75.2 I2W2 76 555 89 90.4 I3W1 56 325 45 46.0 I3W2 64 455 64 71.2

I1 = CSW, I2 =FC, I3 = 70% FC, W1 = herbicide + two hand weedings, W2 = three hand weedings.

173

Table 17. Weed biomass as influenced by weeding practices under aerobic conditions during 2007-09DS (boro).

Weeding practice

Weed biomass (g m−²) Average (g m−²) 30 DAS 45 DAS 65 DAS

W1 3 165 202 123 W2 10 272 294 192

W1: herbicide + two hand weedings, W2: three hand weedings.

174

Figure 1. Rainfall, evaporation, and fluctuation patterns of water table in different water regimes during 2007-08 (boro).

-350

-300

-250

-200

-150

-100

-50

0

50

100

150

06.01

.08

16.01

.08

26.01

.08

05.02

.08

15.02

.08

25.02

.08

07.03

.08

17.03

.08

27.03

.08

06.04

.08

16.04

.08

26.04

.08

05.05

.08

15.01

.08

25.01

.08

Date

Wat

er l

evel

(m

m)

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

RF

& E

V (

mm

)

RF EV CSW FC 70%FC

175

Figure 2. Soil moisture content of topsoil (0−15 cm) in different water regimes

during 2007-08 (boro).

0

10

20

30

40

FC 70% FC

Water regimes

MC

(%)

176

Figure 3. Water productivity as influenced by water management during 2007-08 (boro). CSW = continuous standing water, FC = field capacity, 70% FC = 70% of field capacity.

177

Figure 4. Soil moisture content of topsoil (0−15 cm) in different water regimes during 2007-08 (boro).

0

10

20

30

40

FC 70% FC

Water regimes

MC

(%)

178

Figure 5. Water productivity as influenced by water management during 2008-09 boro.

CSW = continuous standing water, FC = field capacity,70% FC = 70% of field capacity.

179

Paper 8 The alternate wetting and drying system of rice cultivation in Bangladesh: progress and challenges H.U. Ahmed, K.M. Iftekharuddaula, M. Maniruzzaman, M. Shahidul Islam, and A.K.M. Zakaria

The alternate wetting and drying (AWD) system of rice cultivation is a water-saving technology in which irrigation water is applied 3 to 5 days after the disappearance of water in the field. The total water use in a rice crop is 20−25% less than in continuous flooding. AWD can be practiced in transplanted aus (April-August), transplanted aman (June-November), and boro rice (December-May) cultivation in Bangladesh. Most of the varieties developed for irrigated conditions do not perform well under AWD. Targeted breeding efforts are required to identify key traits and develop varieties with better adaptation to AWD conditions. In Bangladesh, efforts have been made to select higher-yielding lines in the AWD system from observational yield trials (OYTs), advanced yield trials (AYTs), and in segregating populations in experiments conducted under the Asian Development Bank (ADB)-supported project “Developing and disseminating water-saving rice technologies for South Asia.” Varietal performance varied under AWD and a continuous supply of water (CSW). Among the lines tested under CSW, at least three breeding lines, IR80508-B-57-4-B, IR79971-B-338-2-2, and IR78875-176-B-2-13, yielded at least 1.0 t ha−1 more than the popular variety BRRI dhan 28 and were similar in maturity duration to the check. In AWD conditions, three breeding lines (BP223E-MR-5, IR71525-19-1-1, and IR79907-B-425-B-4) yielded at least 1.0 t ha−1 more than the popular variety BRRI dhan 28. The superior performance of different breeding lines clearly indicates the suitability of different lines and possible role of different sets of attributes in higher yield under AWD conditions compared with CSW.

Acronyms and abbreviations: AWD = alternate wetting and drying; BRRI = Bangladesh Rice Research Institute; CSW = continuous supply of water; DAWD = days after water disappearance; OYT = observational yield trial; AYT = advanced yield trial; TREY = total rice equivalent yield.

Rice is the staple food of the people in Bangladesh and it grows extensively in every corner of the country. The country produces 30.19 million tons of rice (BBS2009), which is marginally sufficient to meet the nation’s demand. However, the increasing population together with simultaneously less cultivable land because of urbanization and increasing yield losses due to abiotic stresses such as drought, submergence, and salinity are highly likely to deteriorate rice production in the country. The increasing cost of cultivation for irrigation in the boro (dry) season and the declining water table in the northwestern region of the country are some of the concerns in rice cultivation. With such conditions, high-yielding rice varieties for the highly productive dry season cultivated with techniques such as alternate wetting and drying (AWD) provide a solution for increasing rice production with less water use.

Bangladesh’s rice production system is principally the lowland ecosystem. The rice varieties developed by the Bangladesh Rice Research Institute (BRRI) for the

180

three prevailing rice-growing seasons existing in Bangladesh are mostly suitable for lowland conditions. Only a few upland varieties have been developed. These varieties have low productivity and their cultivation is dependent on rainwater. Increasing competition for fresh water from industrial and municipal sectors has led to a decline in the share of agriculture in freshwater resources (Kumar and Ladha 2011). Bangladesh has a network of rivers that flow throughout the country but, because of competition from increasing urbanization and human consumption and injudicious use of water sources, depletion of the groundwater table and a lack of water during the dry season have created an alarming situation for sustainable rice production in many parts of the country. The water shortage during the dry season is increasing day by day. To reduce the increasing cost of irrigation and to adequately overcome the water shortage in the long run, scientists are evaluating alternative systems of rice cultivation that require less water for irrigation and also maintain high productivity. The practical implementation of water-saving options in rice cultivation is emerging as a major challenge in many rice-growing countries, including Bangladesh (Lampayan et al 2003). However, appropriate implementation of available options of water-saving technologies in rice cultivation requires the development of rice varieties suitable for these technologies. The proper application of safe AWD technology not only saves water but also maintains rice productivity at levels similar to those obtained under a continuous supply of water (CSW) (Bouman et al 2007, Tuong 2009). AWD also provides an opportunity for plants to maintain nutrient uptake efficiency similar to that in CSW. It has also been reported that the addition of organic manure under AWD treatments significantly increases root growth and nutrient uptake (Yang et al 2004).Apart from this, AWD leads to lower weed competition than in aerobic rice because of initial puddling during land preparation. This system periodically allows the field to have standing water and keep the soil saturated even when water disappears from the aboveground surface for a few days. When the rice field becomes near saturated or at field capacity (FC), further irrigation is provided. In this system, the rice field is puddled and the crop is established by transplanting. This makes the target ecosystem of AWD as alternate flooding and drying up to FC without allowing any symptoms of drought to appear (Atlin and Lafitte 2002). However, if an assured irrigation facility is not available for providing timely irrigation to the crop, the crop may need tolerance of mild drought stress. Several parameters such as number of days after loss of water over the field surface, moisture content in the soil, and water-table depth in the field have been suggested to be considered for re-irrigation of a field under AWD management. However, safe AWD, which includes keeping the field flooded for 2 weeks, 1 week after transplanting, and for 2 weeks during flowering, and irrigating the field when the water table depth reaches 15-cm depth, has been reported to maintain productivity similar to that in continuous flooded conditions with 15−20% water savings (Bouman et al2007, Tuong 2009).At BRRI, extensive research was undertaken to develop rice varieties suitable to AWD.A screening methodology has been developed and is being followed for identifying breeding lines and varieties adaptable to AWD. In experiments conducted at BRRI, it was seen that BRRI dhan28, a mega-variety for irrigated rice ecosystems, is suitable for adoption under the AWD system of rice cultivation. A great deal of genetic variation is also observed under water stress (AWD) in comparison to continuous irrigation in the crop’s growth stages,

181

suggesting breeding and selecting cultivars suitable for water-saving options (Bouman and Tuong 2000). In Bangladesh, farmers’ expenses on irrigation are increasing day by day because of the increase in the price of fuel. Genotypes suitable to AWD can reduce irrigation requirements and enable farmers to reduce their production costs and improve their livelihood.

Experiments in different countries reported that a water savings up to 30% compared with that in a continuous water supply in AWD can be achieved without any significant reduction in yield potential (Cabangon et al 2001, Belder et al 2002). The two major seasons are transplanted aman (wet season), which is totally rainfed, and boro rice (dry season), in which cultivation is fully dependent on irrigation water. The AWD system of rice cultivation can be successfully applied during the boro season and can help farmers in saving water, reducing expenses on irrigation, as well as maintaining better underground water resources. Existing cultivation practices Among the three seasons of rice cultivation in Bangladesh, the water requirement is very high in the dry season (boro season). This is because of the almost complete absence of rainfall during this season. However, clear weather and high sunshine hours during the cropping season result in high productivity compared with that of the other two cropping seasons. Also, the chances of an occurrence of natural calamities that can affect rice production are very low in this season. High productivity obtained by farmers in the boro season is the major reason for their following this transplanted system of rice cultivation. However, farmers use puddled fields to transplant rice and traditionally keep fields flooded with water up to 2 weeks before the harvesting stage. Every year, boro cultivation is done on nearly 4.0 to 4.5 million hectares of land. The irrigation water comes mostly from underground water resources with a few exceptions of surface-water use. Because of this huge amount of water used from underground sources, the water table goes down, leading to a shortage of water, increased cost of irrigation, and intrusion of salinity. This failure of the underground water-table layer occurs all over the country but it occurs most predominantly in the northwestern region. Because of frequent water-table depletion and low discharge of water, productivity in this region is declining. Alternate wetting and drying technology adoption can be a viable solution to prevent groundwater depletion and use less water per unit of cultivated land. An AWD-adapted rice variety with high yield potential may reduce production costs. The following research activities were undertaken on AWD to select high-yielding lines and develop appropriate cropping systems to improve water-use efficiency. Breeding for AWD-tolerant rice varieties Research on the development and testing of rice varieties suitable to AWD in Bangladesh was begun by BRRI in collaboration with IRRI, Philippines, in 2006. BRRI received segregating materials as well as advanced lines for OYTs and AYTs from IRRI for selection and evaluation under AWD conditions. Irrigation and water management experts of BRRI evaluated water use in the AWD system. At the same time, BRRI undertook its own breeding program for the development of AWD-adapted rice varieties. Different experiments conducted in different years are described below.

182

Fertilizer application and other agronomic practices were as per recommendations in Bangladesh. (i) Selection from segregating generations. During the 2008 dry season (boro), 11 F3 generation crosses were grown at BRRI, Gazipur, in AWD conditions, and 204 high-yielding individual plants were selected. These plants were further grown in progeny rows during the 2009 boro season and 271 progenies were selected. The selection of the progenies proceeded further and 57 homozygous progeny rows were selected in the 2012 boro season that will be tested in a primary OYT in the 2013 boro season. (ii) Observational yield trial In the OYT conducted in boro 2006-07, 112 lines, including 16 BRRI varieties, were evaluated under the AWD and continuous supply of water (CSW) systems. Grain yield was compared in the two systems. Among the breeding lines, IR72022-462-3-3-2 and IR72176-140-1-2 outyielded the checks and other entries under both AWD and CSW (Table 1).

In the boro 2007-08 season, in another OYT, 130 lines were evaluated under CSW and AWD conditions. Among those under CSW, three breeding lines, IR80508-B-57-4-B, IR79971-B-338-2-2, and IR78875-176-B-2-13, yielded at least 1.0 t ha−1 more than popular variety BRRI dhan 28 and were similar in maturity duration (Table 2). In AWD conditions, three breeding lines (BP223E-MR-5, IR71525-19-1-1, and IR79907-B-425-B-4) yielded at least 1.0 t ha−1 more than popular variety BRRI dhan 28 (Table 3). The superior performance of different lines under irrigated and AWD conditions clearly indicates the suitability of different lines for AWD conditions and the possible role of different sets of attributes for higher yield under AWD conditions.

During the 2009 DS (boro), 237 lines from IRRI were evaluated with two local checks under AWD conditions. Twelve entries with superior performance over BRRI dhan 28 were identified and promoted for further evaluation (Table 4). (iii) Advanced yield trial.

In the 2008WS (aman), 34 entries, including three checks, were evaluated under AWDand CWS. Grain yield under CSW was higher than under AWD, but a few entries yielded well under both conditions. BRRI dhan 44 yielded maximum in both systems (Table 5).In practice, it is quite difficult to follow actual AWD conditions in the field in the aman season because of the high amount of rain. In boro 2008-09, 22 entries with three local checks were evaluated in both AWD and CSW. Under AWD conditions, none of the test entries exceeded the grain yield of check variety BRRIdhan28 (5.3 t ha−1; Table 6). The lower yield performance in the test entries might be associated with the higher percentage of spikelet sterility. Under CSW also, none of the entries exceeded the yield performance of check variety BRRI dhan 28 (5.8 t ha−1: Table 6). It is noted that the yield performance of the genotypes was comparatively lower when they were grown under AWD conditions than the yield performance of the genotypes under CSW. Rainfall, evaporation, and the fluctuation pattern of the water table in different water regimes in 2008-09 (boro season) are displayed in Figure 2. (iv) Participatory varietal selection

To evaluate the promising water-saving rice genotypes under AWD conditions in advanced multi-location trials for testing their specific and general adaptability

183

leading to their possible release as varieties, three test entries with BRRI dhan 28, the local check, were evaluated in five farmers’ fields of Gazipur, Rajshahi, Rangpur, Bagura, and Thakurgaon districts. The experiment was established under safe AWD conditions in a farmer’s field (water applied at 3 DAWD). Entries IR74963-262-1-3-3 and PSBRc82 were the highest yielders among all the tested entries, although their yield was not significantly higher than that of BRRI dhan 28 (Table 7). Water management for AWD (i) Quantification of water and association of weeds in AWD and irrigated

rice An experiment was conducted for 2 years during the boro season in 2007-08 and 2008-09. The boro season of 2007-08. A study was conducted on the BRRI farm, Gazipur, during 2007-08 to find out the water requirement of different varieties under different water regimes, to determine efficient weed management practices and to identify water-use-efficient genotypes for higher yield performance. The widely and popularly grown varieties and elite lines of the boro season were selected based on their water-use efficiency through a screening experiment of a water-short regime in 2006-07. The four lines (IR82853-36, IR82914-11, IR69515-KKN-4-UBN-4-2-1-1, IR70179-1-1-1-B) and two checks (BRRI dhan 28 and BRRI dhan 29) were evaluated as per the recommended dose of fertilizer and other management practices. Three water treatments were used: continuous standing water, water applied at 3 DAWD, and water applied at 5 DAWD. Two weed management practices were used: herbicide (Refit 25 EC at 7 days after transplanting (DAT)) with one hand weeding at 30 DAT and two hand weedings at 20 and 40 DAT. A perforated PVC pipe of 150-mm diameter was installed at a depth of 40 cm below the ground surface in the experimental field for monitoring the water-table depth. Field water depth, rainfall, and evaporation were measured. The moisture condition was monitored before applying irrigation in the AWD treatments. The design of the field experiment was a split-split plot with water treatment in the main plot, weed management in the subplot, and variety in the sub-subplot.

The amount of applied irrigation water was measured by the volumetric method following the water management treatment. Rainfall and percolation data were recorded with a rough rain gauge and percolation gauge, respectively. The depth of flooded water was recorded daily. Water level below the ground surface was monitored with perforated PVC pipes 40 cm below field level. A polythene sheet was placed along the boundary line of each plot to protect from seepage and fertilizer loss. Yield and yield-contributing data were recorded following the standard method. The data were analyzed statistically by using IRRISTAT. Water productivity was also calculated as the weight of grain per unit of total water used from transplanting to harvest for each treatment. The initial soil properties were analyzed and the field soil texture varied from silt clay to clay loam with pH 6.3 to 6.8, a low organic carbon content that ranged from 0.71 to 1.2, total nitrogen content ranged from 0.05% to 1.1%, phosphorus ranged from 5.4 to 16 ppm, zinc from 0.70 to 2.2 ppm, and potassium ranged from 0.21 to

184

0.30 meq 100 g−1 soil. These results indicated that the soil of the study site was poor to fair for crop production.

Rainfall, pan evaporation, and field water-level fluctuation pattern from transplanting to harvesting are shown in Figure 3. The data indicated that rainfall during the cropping period of rice was 132 mm and was distributed throughout the whole period, which became congenial to rice production in the boro season and reduced the irrigation frequency. The water depth pattern varied for all three water regimes. Maintenance of continuous standing water requires maximum irrigation water and that was 753 mm. But, in the AWD treatment, the amount of applied irrigation water was 21% and 30% less in the 3 DAWD and 5 DAWD treatments, respectively (Table 1). However, the 5 DAWD irrigation treatments require the lowest quantity of water (530 mm) compared with CSW. In the nonflooded period, the field water table dropped about 15 to 18 cm below the ground surface for the treatment of 3 DAWD, whereas it was about 25 to 29 cm below the ground surface for the treatment of 5 DAWD. The highest field water depth at all times was recorded in the CSW treatment and the lowest in 5 DAWD. The average soil moisture content was about 30% in the topsoil for the treatment of 3 DAWD and was about 24% for the treatment of 5 DAWD and there was a fluctuating pattern of moisture content with the variation in soil depth (Figure 4).

The total water input (rainfall + irrigation) ranged from 662 to 885 mm, out of which 132 mm was rainfall. There was no drainage outflow because of low rainfall during the crop growing season. There was no significant difference in evapo-transpiration, but variation occurred in seepage and percolation losses due to nonflooded conditions of the AWD treatments.

There was no significant yield difference among the interaction effect of irrigation treatments and genotypes, but significant variation was observed among the genotypes (Table 9), though the quantity of water applied varied from 662 to 885 mm. The percent water saved at 3 DAWD and 5 DAWD is shown in Table 8. The highest grain yield was recorded in CSW and a decreasing trend was found for the AWD treatments (Table 9). Water productivity varied among the irrigation treatments and the yield difference was obtained due to genotypes. Water productivity ranged from 0.82 to 1.15 kg m−3, with an average of 1.00 kg m−3, and it showed an increasing trend in the AWD treatments (Figure 5).

Boro 2008-09 season. The experiment was repeated in the boro2008-09 season to validate the findings of the previous season’s experiment. Rainfall, pan evaporation, and field water-level fluctuation pattern from transplanting to harvest are shown in Figure 6. The data indicated that rainfall during the base period of rice was 101 mm and was distributed up to the ripening period, which became congenial to rice production in the boro season and this reduced the irrigation frequency. The water depth pattern varied for all three water regimes. Maintenance of continuous standing water required maximum irrigation water and it was 755 mm. But, in AWD treatments, the amount of applied irrigation water was 24% and 34% less in 3 DAWD and 5DAWD treatments, respectively (Table 10). However, the 5 DAWD treatments required the lowest quantity of water (500 mm) compared with CSW. In the nonflooded period, the field water table dropped to 15 to 18 cm below the ground surface for the 3 DAWD treatments, whereas it was 25 to 29 cm below the ground surface for the 5 DAWD

185

treatments. The highest field water depth at all times was recorded in the CSW treatment and the lowest in the 5 DAWD treatments. The average soil moisture content was about 30% in the topsoil for the 3 DAWD treatments and it was 24% for the 5 DAWD treatments, and fluctuation was observed in the pattern of moisture content with the variation in soil depth (Figure 7).The total water input (rainfall + irrigation) ranged from 601 to 856 mm, out of which 101 mm was rainfall. There was no drainage outflow because of low rainfall during the crop growing season. There was no significant difference in evapo-transpiration, but variation occurred in seepage and percolation losses because of nonflooded conditions of the AWD treatments (data not shown).

There was a significant yield difference among the interaction effect of irrigation treatments, weed management, and genotypes (Table11). The highest grain yield was recorded in CSW and a competitive yield was also found in the 3 DAWD treatments. But, a significant yield loss was observed in the 5 DAWD treatments. There was a significant yield difference between the genotypes within the water management treatments, but no difference was observed among the weed management treatments.

Water productivity varied among the irrigation treatments and a yield difference was recorded due to genotypic effects. The average water productivity ranged from 0.74 to 1.01 kg m−3, with an average of 1.00 kg m−3, and there was an increasing trend of water productivity in the AWD treatments. In the 3DAWD and 5DAWD treatments, water productivity was 0.98 and 1.04 kg m−3, respectively (Table 12). On the whole, total water applied (mm), amount of water saved, and water productivity were analyzed over the consecutive boro seasons of 2008 and 2009 in the quantification experiment conducted under transplanted conditions. On average, 560 mm of water was applied in the 3 DAWD treatments and 494 mm in the 5 DAWD treatments. The percentage of water saved was 22% and 31% in 3 DAWD and 5 DAWD treatments, respectively (Table 12). Weed management for AWD Weed species and weed biomass. During boro2007-08, weed species and weed biomass were studied under different weed management and irrigation treatments in transplanted rice. The population of different weed species and weed biomass were much higher at 20 and 40 DAT in hand weeding and 3 and 5 days’ AWD plots (Table 13). The highest weed biomass was observed in two hand weedings and 3 and 5 days’ AWD-treated plots at 20 and 40 DAT. Cynodon dactylon (grass), Scirpus sp., and Cyperus difformis (sedge) and Marselia sp. and Photka (broadleaf) weeds were present.

In boro 2008-09, no weeds were found in W1 (herbicide-treated plots) at 20 and 40 DAT. At 20 DAT, higher weed biomass (12.88 g m−2) was observed in I2W2 (3 DAWD with hand weeding), followed by I1W2, when the irrigation treatments were not imposed (Table 14). At 40 DAT, when irrigation treatments were imposed, higher weed biomass (9.68 g m−2) was found in I3W2 at 5 DAWD and hand weeding. But, the differences in weed biomass among the treatments were very little. They might be due to less weed infestation. The dominant weed species were Scirpus maritimus, followed by Monochoria vaginalis and Cynodon dactylon.

186

In summary, application of herbicide with one hand weeding treatment in the case of AWD transplanted rice was found to be better in controlling weeds than two hand weedings as reflected in less weed biomass recorded (Tables 14 and 15).

Crop rotation for improving land productivity A study was conducted on the BRRI farm, Gazipur, during T. aman, rabi/boro season, 2008-09, to quantify the amount of water required for rice-based cropping systems, to increase year-round water-use efficiency and land productivity and also to identify a suitable cropping pattern for water-short environments. The different rice-based cropping systems were P1 = rice (BR11, transplanted: TP) - rice (BRRI dhan 29, TP) - fallow, P2 = rice (BRRI dhan 33, TP) – mustard -rice (Vandana, DS), P3 = rice (BRRI dhan 33, TP) – potato - rice (Vandana, DS), and P4 = rice (BRRI dhan 39) – potato - mungbean. Recommended fertilizer doses were applied as per the crops according to BRRI recommendations. For irrigation, the best-performing AWD treatment, that is, 3 DAWD, was applied during boro and proper supplemental irrigation was applied during T. aman season. T. aman and boro rice were transplanted on 25 July 2008 and 12 January 2009 and harvested on 3 November 2008 and 16 May 2009, respectively. Potato, mustard, and mung bean were sown on 28 November 2008, 28 November 2008, and 20 March 2009, respectively. Vandana was direct-seeded after the potato and mustard harvest. BR11, BRRI dhan 33, and BRRI dhan 39 were used in the T. aman season, and BRRI dhan 29 and Vandana were used in the boro season. Cardinal, BARI Mustard 14, and BARI mung 6 were used as potato, mustard, and mung bean varieties, respectively.

The amount of applied irrigation water was measured by the volumetric method by following the water management treatments. Rainfall, evaporation, and percolation data were recorded during crop growing periods through a rain gauge, pan evaporation, and percolation gauge, respectively. The depth of flooded water was recorded daily. The water level below the ground surface was monitored with perforated PVC pipes 40 cm below the field level. A polythene sheet was placed along the boundary line of each plot to protect from seepage and fertilizer loss. Yield and yield-contributing data of rice were recorded following the standard method. The collected data were analyzed statistically by using IRRISTAT software package and least significant difference (LSD) value was used to determine the significance levels among the treatment means. The level of confidence was set at 95%. Water productivity was also calculated as the weight of grain per unit of total water used from transplanting to harvest for each treatment. The yield of different crop sequences with rice equivalent yield is shown in Table 16. The yield of rice varied due to rice variety, field duration, and crop growing season. The highest rice yield (6.26 t ha−1) was found in BRRI dhan 29 during the boro season but in T. aman the highest yield was observed in BR11 and both are high-yielding long-duration varieties. Among the rabi crops, potato was the highest yielder (28.4 t ha−1) and mustard was the lowest (0.42 t ha−1). There was a significant difference in yield among the crop sequences. During the year, the maximum yield equivalence in terms of rice yield was recorded in rice (BRRI dhan 39) – potato - mung bean (23.1 t ha−1), followed by rice (BRRI dhan 33) -potato - rice (Vandana) (22.5 t ha−1) (Table

187

16). But, the year-round yield of the rice (BR11) - rice (BRRI dhan 29) - fallow crop sequence was 10.6 t ha−1and the minimum yield was found in the rice (BRRI dhan 33) – mustard - rice (Vandana) crop sequence. The variation in yield equivalence was mostly governed by rabi crops because T. aman was grown at the same level of irrigation. The highest yield equivalence in potato was due to the much higher yield and price of potato than other rabi crops. Irrigation water use and water productivity The maximum irrigation water (1,684 mm) was consumed by the rice (BR11) - rice (BRRI dhan 29) - fallow crop sequence, followed by rice (BRRI dhan 33) - potato – rice (Vandana) (1,435 mm) (Table 18).The lowest water was consumed by the rice (BRRI dhan 39) - potato - mung bean crop sequence (1,215 mm) as two other crops were used rather than rice. Water productivity was the highest (1.90 kg m−3) in the rice (BRRI dhan 39) - potato - mung bean crop sequence, followed by rice (BRRI dhan 33) - potato-rice (Vandana) (1.59 kg m−3) (Table 17), because higher potato yield and lower consumption of water were the only contributors to increasing water productivity.

The economics of different crop sequences A significant variation in net return was observed among different crop sequences (Table 18). Maximum net return (Tk. 215,869; US$3,128) was recorded in T4: rice (BRRI dhan 39) - potato - mung bean crop sequence, followed by T3: rice (BRRI dhan 33) - mustard - rice (Vandana) (Tk. 212,108; $3,074) and T1: rice (BR11) - rice (BRRI dhan 29) - fallow (Tk. 105,186; $1,524). Though the net return of the potato-based crop sequence was higher, its benefit-cost ratio (BCR) was lower due to the high input price. The best BCR was found in T1: rice (BR11) - rice (BRRI dhan 29) - fallow (1.93) (Table 18) because of the low cost involved in rice cultivation. Among the four crop sequences, based on water productivity, maximum net return and maximum yield were found from the rice (BRRI dhan 39) – potato - mung bean crop sequence. This crop sequence is also good for better maintenance of soil health as a legume crop in a rotation, which will improve soil quality. Policy issues The AWD system of rice cultivation is a new rice production system. It is already proven that the AWD system can save water in rice production. The development and identification of suitable rice varieties for the AWD system is the prime requirement before practicing AWD. Farmers are still accustomed to keeping stagnant water in their rice fields. In the government-supported irrigation project area, the farmers are paying whole-season money for a unit piece of land and pump operators are committed to providing water as and when necessary. In that case, farmers will not allow their land to stay dry even for a day. If the government adopts payment on the basis of the amount of water used, and if the farmers are fully convinced that, because of the dryness of the rice field, for the time being this will not hamper productivity, and then this AWD system of rice production will work. For farmers who manage their own source of irrigation, the AWD system may immediately come into force if farmers are motivated.

188

Conclusions Rice cultivation in Bangladesh mostly depends on the lowland ecosystem. The conventional method of growing rice under flooded conditions in this ecosystem requires a high input of irrigation water. Irrigation of rice fields in Bangladesh mostly relies on groundwater resources. Trends in climate change, increased drought incidence, and increased use of groundwater by industry and the municipal sector have increased competition for freshwater resources. In such conditions, water-saving technologies such as AWD can play an important role in achieving yield equal to that of CSW, with reduced water requirement. Apart from this, techniques such as AWD reduce the cost incurred on fuel for irrigation, leading to increased net gain. The success of such water-saving technologies depends on the combination of varieties suitable to the target environment and the production technology package itself. In our study, it was observed that BRRI dhan 29 recommended for boro rice cultivation showed higher yield than other varieties under the AWD system. During the wet season, under T. aman, BRRI dhan 44 was found to be higher yielding than BR 11 and others. In the ADB-supported project, the identification of new breeding lines BP223E-MR-5, IR71525-19-1-1, and IR79907-B-425-B-4 with 1.0 t ha−1 more yield than the popular variety BRRI dhan 28 indicates that, by cultivating new varieties under the AWD system, rice productivity of the flooded system can be maintained with a 20−25% savings of irrigation water.

References Atlin GN, Lafitte HR. 2002. Developing and testing rice varieties for water-saving

systems in the tropics. In: Bouman BAM, Hengsdijk H, Hardy B, Bindraban PS, Tuong TP, Ladha JK, editors. ‘Water-wise rice production. Proceedings of the International Workshop on Water-wise Rice Production, 8-11 April 2002, Los Baños, Philippines. Los Baños (Philippines): International Rice Research Institute.

BBS. 2009. Bangladesh Bureau of Statistics. Belder P, Bouman BAM, Spiertz JHJ, Lu G, Quilang EJP 2002. Water use of

alternately submerged and non submerged irrigated lowland rice. In: Bouman BAM, Hengsdijk H, Hardy B, Bindraban PS, Tuong TP, Ladha JL, editors. Water-wise rice production. Proceedings of the International Workshop, 8-11 April 2002, IRRI, Los Baños, Philippines. Los Baños (Philippines): International Rice Research Institute. p 51-61.

Bouman BAM, Lampayan RM, Tuong TP. 2007. Water management in irrigated rice: coping with water scarcity. Los Baños (Philippines): International Rice Research Institute. 54 p.

Bouman BAM, Tuong TP. 2000. Field water management to save water and increase its productivity in irrigated lowland rice. Agric.Water Manage. 1615:1-20.

Cabangon RJ, Castillo EG, Bao LX, Lu G, Wang GH, Cui YL, Tuong TP, Bouman BAM, Li YH, Chen CD, Wang JZ. 2001. Impact of alternate wetting and drying irrigation on rice growth and resource-use efficiency. In: Barker R, Loeve R, Li

189

YH, Tuong TP, editors. Water-saving irrigation for rice. Proceedings of the International Workshop, 23-25 March 2001, Wuhan, China. Colombo (Sri Lanka): International Water Management Institute. p 55-79.

Kumar V, Ladha JK. 2011. Direct seeding of rice: recent developments and future research needs. Adv. Agron. 111, DOI: 10.1016/B978-0-12-387689-8.00001-1.

Lampayan RM, Bouman BAM, de Dios JL, Lactaoen AT, Espiritu AJ, Norte TM, Quilang EJP, Tabbal DF, Llorca LP, Soriano JB, Corpuz AA, Malasa RB, Vicmudo VR. 2003. Adaption of water saving technologies in rice production in the Philippines. In: International Workshop “Transitions in agriculture for enhancing water productivity,” Tamil Nadu, India, September 2003.

Tuong TP. 2009. Promoting AWD in Bangladesh. RIPPLE 4(3):1-2. Yang C, Yang L, Yang Y, Ouyang Z. 2004. Rice root growth and nutrient uptake as

influenced by organic manure in continuously and alternately flooded paddy soils. Agric. Water Manage. 70(1):67-81.

Notes Authors’ addresses: H.U. Ahmed, K.M. Iftekaruddaula, M. Maniruzzaman, and M. Shahidul Islam, Bangladesh Rice Research Institute, Gazipur, Bangladesh; A.K.M. Zakaria, Rural Development Academy, Bogra, Bangladesh.

Fig 1. Rainfall, evaporation, and fluctuation pattern of water table in different water regimes during 2008 T. aman in an advanced yield trial (AWD) field.

Fig 2. Rainfall, evaporation, and fluctuation pattern of water table in different water regimes during 2008-09 boro season.

Fig 3. Rainfall, evaporation, and fluctuation pattern of water table in different water regimes during 2007-08 boro season.

Fig 4. Soil moisture content at different soil depth during alternate wetting and drying treatments during 2007-08 boro season.

Fig 5. Water productivity as influenced by water management during 2007-08 boro season. CSW = continuous standing water, 3 DAWD = water applied at 3 days after water disappearance, and 5 DAWD = water applied at 5 days after water disappearance.

Fig 6. Rainfall, evaporation, and fluctuation pattern of water table in different water regimes during 2008-09 boro season.

Fig 7. Soil moisture content at different soil depth during alternate wetting and drying treatments during 2008-09 boro season.

190

Table 1. Selected genotypes from OYT, 2006-07 DS (boro season).

Designation Grain yield (t ha−1)

Growth duration (days)

AWD CSW AWD CSW IR82855-17 6.1 7.1 160 155 IR64 5.9 7.0 160 154 IR82853-36 5.9 7.1 151 153 IR82914-11 6.1 6.1 153 155 IR82851-23 6.1 5.5 153 152 IR72176-140-1-2-2-3 6.1 6.1 152 151 IR74963-262-5-1-3-3 6.1 6.1 151 151 IR73707-45-3-2-3 5.7 6.7 162 161 IR69515-KKN-4-UBN-4-2-1-1 6.3 7.1 153 153 IR70179-1-1-1-B 6.5 5.1 153 153 IR72022-462-3-3-2 6.7 7.1 160 157 IR72176-140-1-2 6.7 6.9 160 158 IR74963-262-5-1-3-3 6.1 6.9 159 155 IR75003-95-5-1-3 5.9 6.9 161 158 IR77073-B-35-1-1 6.3 7.1 160 157 IR77030-B-34-1-1 6.3 5.7 160 153 IR77030-B-4-2-2 6.3 4.6 153 153 IR77080-B-6-2-2 5.8 5.9 154 155 IR77031-B-42-3-1 5.7 6.9 152 153 IR77034-B-4-3-3 5.7 6.1 161 157 IR74371-46-1-1 5.7 5.9 151 153 IR74963-262-5-1-3-3 5.7 6.1 151 153 IR62142-114-32-2-2 6.3 7.1 164 164 IR64683-87-2-2-3-3 5.8 6.1 153 153 BR14* 5.9 6.3 158 156 BRRI dhan 36* 5.7 5.8 153 153

*Check.

191

Table 2. List of selected entries from OYT-irrigated (CSW), 2007-08 DS (boro season).

Designation Grain yield

(t ha−1) Growth duration

(days) BP223E-MR-5 2.9 145 IR71525-19-1-1 2.4 144 IR78875-176-B-2-B 3.6 139 IR78875-190-B-1-3 2.8 141 IR78875-207-B-1-B 2.6 141 IR79907-B-425-B-4 2.4 142 IR79913-B-283-B-3 2.6 138 IR79971-b-201-2-4 3.4 140 IR79971-B-338-2-2 3.8 144 IR80508-B-57-4-B 3.9 147 IR80973-B-186-U4-2 2.6 148 IR81040-B-78-U2-1 2.4 150 IR72010-39-CPA-7-1-1-4-2

2.4 142

IR80463-B-39-1 3.2 140 IRPSBRc82 2.2 138 IR72667-16-1-B-B-3 3.2 150 IR77298-14-1-2 2.4 146 BRRI dhan 28* 2.6 144

*Check.

192

Table 3. List of selected entries from OYT under AWD, 2007-08 DS (boro season).

*Check.

Designation Grain yield (t ha−1)

Growth duration (

days) BP223E-MR-5 5.0 145 IR71525-19-1-1 4.6 144 IR78875-176-B-2-B 3.8 139 IR79907-B-425-B-4 4.6 142 IR80973-B-186-U4-2 3.6 148 IR82299-B-306-B 3.4 140 IR70210-39-CPA-7-1-1-4 4.2 142 IRO5A233 (IR79478-67-3-3-2) 2.5 137 IR80416-B-32-3 3.5 142 IR80420-B-22-2 2.9 144 IR68098-B-78-2-2-B-1 3.4 145 IR75417-R-R-R-R-267-3 3.0 147 BRRI dhan 28* 3.3 145

193

Table 4. Selected entries under AWD, OYT, boro 2008-09.

*Check.

Designation Days to maturity

Plant height (cm)

Number of panicles

Grain yield (t ha−1)

IR81047-B-57-1-4 146 124.8 10.4 5.49 IR81022-B-310-1-1 144 131.8 9.8 5.51 IR81026-B-126-4-2 146 98.0 11.0 6.26 IR77080-B-34-3 154 101.6 12.2 6.13 IR80420-B-22-2 154 110.0 10.4 6.58 IR77298-12-7-13 155 91.0 11.0 6.48 IR83140-B-36-B-B 154 87.4 12.0 7.05 IR83140-B-32-B-B 155 93.0 10.6 6.07 IR83140-B-33-B-B 156 92.4 11.8 5.92 IR83141-B-18-B-B 154 92.2 11.6 6.33 IR83142-B-71-B-B 157 94.2 13.4 6.40 IR83143-B-32-B-B 155 97.2 10.8 6.32 BRRI dhan 28* 150 93.4 12.0 5.81

194

Table 5. Plant height (cm), growth duration, and grain yield of the genotypes for both CSW and AWD conditions in AYT, T. aman, 2008.

Designation Plant height Days to

maturity Grain yield (t ha−1)

CSW AWD CSW AWD CSW AWD BP223E-MR-5 102 104 117 114 4.1 3.8 IR71525-19-1-1 135 129 122 120 3.2 2.9 IR78875-176-B-2-B 113 119 122 121 3.7 3.9 IR78875-190-B-1-3 114 115 118 118 3.6 3.6 IR78875-207-B-1-B 112 112 126 123 3.9 3.8 IR79907-B-425-B-3 108 110 123 122 3.2 3.2 IR79913-B-283-B-3 119 115 126 124 3.1 2.9 IR79971-B-201-2-4 123 126 124 124 2.7 2.1 IR79971-B-338-2-2 138 138 116 117 3.0 2.3 IR80508-B-57-4-B 132 138 124 125 3.3 3.4 IR80973-B-186-U1-2 125 118 124 125 2.7 3.6 IR82299-B-306-B 101 105 118 117 2.4 3.3 IR70210-39-CPA-7-1-1-4 116 121 124 123 4.2 4.0 IR05A233(IR79478-28-3-2-2)

98 106 127 127 3.3 4.1

IR80416-B-32-3 104 107 123 124 4.1 5.2 IR80420-B-22-2 107 108 126 127 3.3 3.3 IR68098-B-78-2-2-B-1 104 106 126 126 4.3 3.9 IR70210-39-CPA-1-1-4-2 112 112 123 126 2.7 3.5 IR75417-R-R-R-R-267-3 152 140 124 124 3.1 3.4 IR80463-B-39-1 115 113 128 126 4.8 3.6 PSBRC 82 113 112 126 125 4.0 3.8 IR72667-16-1-B-B-3 102 103 127 125 3.0 3.2 IR77080-B-4-2-2 105 102 124 124 3.5 3.9 PSBRC82 109 107 127 127 4.5 5.0 IR74963-262-5-1-3-3 108 108 128 128 4.1 4.2 IR77031-B-42-3-1 99 101 126 126 1.9 2.0 IR73707-45-3-2-3 108 104 129 129 4.1 3.7 IR69515-KKM-4-UBN-4-2-1-1

106 104 123 125 3.1 2.4

BR11* 110 110 141 140 5.9 5.5 BRRI dhan 30* 110 111 138 137 6.4 6.1 BRRI dhan 44* 119 119 139 138 7.1 6.3 LSD 0.05 1.03 1.09

*Check.

195

Table 6. Performance of AYT under AWD and CSW conditions during 2008-09DS (boro).

Designation Days to

maturity Plant height (cm)

Number of panicles

Sterility (%)

Biomass (g)

Grain yield (t ha−1)

AWD CSW BP223E-MR-5 157 92.5 12.1 38.0 48.4 4.1 4.7

IR71525-19-1-1 157 115.5 9.7 36.1 57.5 3 3.9 IR78875-176-B-2-B 153 106.8 10.1 24.5 54.2 4.1 4.3 IR78875-190-B-1-3 155 110.3 10.4 20.3 46.4 3.9 4.7 IR78875-207-B-1-B 159 108.3 10.3 32.0 58.4 4.9 5.3 IR79907-B-425-B-3 147 99.9 10.9 25.8 44.8 4.0 5.0 IR79913-B-283-B-3 154 107.0 10.6 40.4 43.7 3.9 4.1 IR79971-B-201-2-4 154 113.7 9.5 40.4 58.0 3.6 4.0 IR79971-B-338-2-2 148 116.3 9.0 34.0 56.1 3.7 3.4 IR80508-B-57-4-B 154 117.2 9.4 26.9 38.7 3.6 4.3 IR80973-B-186-U1-2 158 113.9 9.7 49.3 51.5 3.3 4.2 IR82299-B-306-B 154 103.1 9.0 29.8 55.3 3.9 4.3 IR70210-39-CPA-7-1-1-4

152 111.0 10.9 31.1 54.3 4.0 4.6

IR05A233(IR79478-28-3-2-2)

158 90.2 11.3 37.0 43.3 3.6 4.4

IR80416-B-32-3 152 91.1 11.3 25.5 62.5 4.3 5.2 IR80420-B-22-2 155 97.1 11.6 36.8 49.8 4.4 5.6 IR68098-B-78-2-2-B-1 155 83.2 10.3 27.6 49.8 4.1 4.8 IR70210-39-CPA-1-1-4-2

155 105.0 9.4 37.4 64.3 3.8 4.3

IR75417-R-R-R-R-267-3

144 132.2 12.2 11.0 56.2 4.3 4.2

IR80463-B-39-1 155 95.5 10.9 38.0 46.5 4.8 4.6 IR72667-16-1-B-B-3 152 89.8 12.8 30.7 58.3 4.4 4.7 IR77080-B-4-2-2 153 86.0 11.9 23.6 61.8 3.6 4.8 BR14 (check1) 158 107.4 10.1 26.8 54.3 4.5 4.9 BRRI dhan 28* 151 96.9 12.5 23.2 45.5 5.3 5.8 BRRI dhan 29* 160 100.8 12.4 35.5 64.9 4.9 6.2

*Check.

Table 7. Yield performance of the genotypes, ALART, 2009 DS (boro).

196

Genotype Locations+

Grain yield (t ha−1) L1 L2 L3 L4 L5 Mean

IR74963-262-5-1-3-3 6.24 5.33 5.63 6.29 5.67 5.83 IR69515-KKN-4-UBN-4-2-1-1 6.38 5.05 5.32 6.30 5.73 5.76 PSBRc82 6.26 5.69 5.03 6.25 5.95 5.83 BRRI dhan 28* 6.33 5.05 5.21 6.70 5.83 5.82 Mean 6.30 5.28 5.30 6.39 5.80 5.81 LSD 0.05 0.72 0.32 *Check. +L1 = Gazipur, L2 = Rajshahi, L3 = Rangpur, L4 = Bogra, L5 = Thakurgaon.

197

Table 8. Field-level water applied in different irrigation treatments during boro, 2007-08.

Treatment Total water (mm) Water saved compared with CSW (%) Irrigation Rainfall Tota

l Cont. standing water (CSW)

753 132 885 −

Water at 3 days after water disappearance (DAWD)

595 132 727 21

Water at 5 DAWD 530 132 662 30

198

Table 9. Grain yield (t ha−1) variation in different irrigation regimes during boro, 2007-08.

Variety Irrigation regimes CSW 3 DAWD 5 DAWD Variety

mean IR82853-36 5.37 5.23 5.32 5.31 IR82914-11 5.94 5.92 5.92 5.93 IR69515-KKN-4-UBN-4-2-1-1 6.07 6.00 5.95 6.01 IR70179-1-1-1-B 5.37 5.24 5.19 5.27 BRRI dhan 28* 6.32 6.10 5.46 5.96 BRRI dhan 29* 8.02 7.73 7.70 7.82 Irrig. mean 6.18 6.04 5.92 6.05 LSD0.05 0.46

*Check.

199

Table 10. Field-level water applied and water saved in different irrigation treatments during boro, 2008-09.

Treatment Water applied (mm) Water saved

compared with CSW (%)

Irrigation Rainfall

Total

Cont. standing water (CSW) 755 101 856 − Water at 3 days after water disappearance (DAWD)

573 101 674 24

Water at 5 DAWD 500 101 601 34

200

Table 11. Interaction effect on grain yield (t ha−1) of different varieties and weed management in different irrigation regimes during boro, 2008-09.

Variety Irrigation regimes

CSW 3 DAWD 5 DAWD Variety mean

Herbicide + hand weeding IR82853-36 4.79 4.94 4.43 4.72 IR82914-11 5.51 5.66 5.04 5.41 IR69515-KKN-4-UBN-4-2-1-1 5.95 5.93 5.38 5.75

IR70179-1-1-1-B 5.43 5.20 4.71 5.11 BRRIdhan28 5.82 5.83 5.28 5.64 BRRIdhan29 6.26 6.20 5.26 5.91 Irrig. mean 5.63 5.26 5.02 5.43

Two hand weedings IR82853-36 4.87 4.98 4.30 4.72 IR82914-11 5.80 5.83 5.45 5.69 IR69515-KKN-4-UBN-4-2-1-1 5.85 5.86 5.53 5.74

IR70179-1-1-1-B 4.88 5.04 4.57 4.83 BRRI dhan 28 5.92 5.72 5.25 5.63 BRRI dhan 29 6.38 6.25 5.26 5.96 Irrig. mean 5.62 5.61 5.06 5.43 LSD0.05 0.23

201

Table 12. Total water applied, water saved, and water productivity in different irrigation regimes.

Treatment Boro 2006-07 Boro 2007-

08 Boro 2008-09 Average

Total water applied (mm) CSW 595 753 755 701 Water at 3 DAWD 513 595 573 560 Water at 5 DAWD 452 530 500 494

Water saved (mm) Water at 3 DAWD 21 21 24 22 Water at 5 DAWD 29 30 34 31

Water productivity (kg m−3) CSW 0.85 0.82 0.74 0.80 3 DAWD 0.91 1.04 0.98 0.98 5 DAWD 0.96 1.15 1.01 1.04

202

Table 13. Weed species and weed biomass as influenced by weed management and irrigation treatment under transplanted rice conditions, boro, 2007-08.

Replication/ treatment

Number of weed species present m−2 Biomass (g m−2) Grass Sedge Broad leaf

20 DAT I1W1 − − − − I1W2 8 64 20 1.96 I2W1 8 4 36 2.90 I2W2 16 − 72 7.88 I3W1 8 8 20 2.90 I3W2 20 8 28 4.10

40 DAT I1W1 16 4 16 3.60 I1W2 16 300 4 6.84 I2W1 16 8 40 5.59 I2W2 36 528 8 10.36 I3W1 12 8 24 3.92 I3W2 12 536 8 10.00

I1 = continuous flooding, I2 = irrigation at 3 days after water disappearance, I3 = irrigation at 5 DAWD, W1 = Refit at 7 DAT + 1 hand weeding, W2 = 2 hand weedings at 25 and 50 DAT.

203

Table 14.Weed species and weed biomass as influenced by weed management and irrigation treatment under transplanted rice condition during boro, 2008-09.

Treatment Number of weed species present

m−2 Biomass (g m−2)

Grass Sedge Broad leaf 20 DAT

I1W1 − − − − I1W2 − 791 − 9.8 I2W1 − − − − I2W2 28 313 12 12.9 I3W1 − − − − I3W2 12 − 4 5.6

40 DAT I1W1 − − − − I1W2 − 263 4 9.4 I2W1 − − − − I2W2 8 285 16 7.0 I3W1 − − − − I3W2 4 320 16 9.7

I1 = CSW, I2 = 3 DAWD, I3 = 5 DAWD, W1 = herbicide + one hand weeding, W2 = two hand weedings.

204

Table 15. Weed biomass as influenced by weeding practices under transplanted AWD conditions, boro, 2007-09.

W1 = herbicide + one hand weeding, W2 = two hand weedings.

Weeding practices

Weed biomass (g m−²) Average (g m−²) 20 DAT 40 DAT

W1 5.8 13.11 9.5 W2 42.24 53.3 47.8

205

Table 16. Crop yield and total rice equivalent yield (TREY) of different rice-based crop sequences, BRRI, Gazipur, 2008-09.

Crop sequence T. aman (t ha−1)

Rabi/ boro (t ha−1)

Kharif-I/aus (t ha−1)

TREY (t

ha−1) Rice (BR11)-rice (BRRI dhan 29) - fallow

4.38 6.26 − 10.64

Rice (BRRI dhan 33) – mustard - rice (Vandana)

3.35 0.42 3.48 9.12

Rice (BRRI dhan 33) - potato - rice (Vandana)

3.37 28.40 3.52 22.46

Rice (BRRI dhan 39) - potato - mungbean

3.40 28.40 1.04 23.14

206

Table 17. Total water used and water productivity of different rice-based crop sequences, BRRI Farm, Gazipur, 2008-09.

Crop sequence T.

aman

(mm)

Rabi/ boro (mm)

Summer/ aus (mm)

Total water use (mm, inclusive rainfall)

TREY* (t ha−1)

Water productivity (kg m−3)

T1: rice (BR11) - rice (BRRI dhan 29) - fallow

710 974 − 1,684 10.64 0.63

T2: rice (BRRI dhan 33) -mustard - rice (Vandana)

680 141 460 1,281 9.12 0.71

T3: rice (BRRI dhan 33) -potato - rice (Vandana)

680 295 460 1,435 22.46 1.59

T4: rice (BRRI dhan 39) -potato -mungbean

680 295 240 1,215 23.14 1.90

*TREY =total rice equivalent yield.

207

Table 18. Net return and benefit-cost ratio (BCR) of different rice-based crop sequences, BRRI Farm, Gazipur, 2008-09.

Crop sequence TREY

(t ha−1) Return

per year (Tk

ha−1)*

Cost per year (Tk ha−1)*

Net return

(Tk ha−1)

BCR

T1: rice (BR11) - rice (BRRI dhan 29) -fallow

10.64 159,600 ($2,313)

54,414 ($788)

105,186 ($1,524)

1.93

T2: rice (BRRI dhan 33) - mustard - rice (Vandana)

9.12 136,800 ($1,982)

62,124 ($900)

74,676 ( $1,082)

1.20

T3: rice (BRRI dhan 33) - potato - rice (Vandana)

22.46 336,900 ( $4,882)

124,792 ($1,808)

212,108 ($3,074)

1.70

T4: rice (BRRI dhan 39) - potato -mungbean

23.14 347,100 ($5,030)

131,231 ($1,902)

215,869 ($3,128)

1.64

*US$1= taka 69. TREY = total rice equivalent yield.

208

Figure 1. Rainfall, evaporation, and fluctuation pattern of water table in different water regimes during 2008 T. aman in an advanced yield trial (AWD) field.

-100.0

-50.0

0.0

50.0

100.0

150.0

200.0

15.08

.08

25.08

.07

04.09

.07

14.09

.07

24.09

.07

04.10

.07

14.01

0.07

24.10

.07

03.11

.07

13.11

.07

Date

CSW

& A

WD

(mm

)

0.0

100.0

200.0

300.0

400.0

500.0

600.0

RF &

EV

(mm

)

RF (mm) EV (mm) Field Water Level

209

Figure 2. Rainfall, evaporation, and fluctuation pattern of water table in different water regimes during 2008-09 boro season.

-350.0

-300.0

-250.0

-200.0

-150.0

-100.0

-50.0

0.0

50.0

100.0

150.0

17.01

.09

27.01

.09

06.02

.09

16.02

.09

26.02

.09

08.03

.09

18.03

.09

28.03

.09

07.04

.09

17.04

.09

27.04

.09

Date

CSW

& A

WD

(mm

)

0.0

10.0

20.0

30.0

40.0

50.0

60.0

RF &

EV

(mm

)

RF (mm) EV (mm) CSW AWD

210

Figure 3. Rainfall, evaporation, and fluctuation pattern of water table in different water regimes during 2007-08 boro season.

-350.0

-300.0

-250.0

-200.0

-150.0

-100.0

-50.0

0.0

50.0

100.0

150.0

22.01

.08

01.02

.08

11.02

.08

21.02

.08

03.03

.08

13.03

.08

23.03

.08

02.04

.08

12.04

.08

22.04

.08

02.05

.08

11.05

.08

Date

Wat

er le

vel (

mm

)

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

RF &

EV

(mm

)

RF EV CSW 3 DAWD 5 DAWD

211

Figure 4. Soil moisture content at different soil depth during alternate wetting and drying treatments during 2007-08 boro season.

0.00

10.00

20.00

30.00

40.00

0-15 15-30 30-45Soil depth (cm)

MC

(%)

3 DAWD 5 DAWD

212

Figure 5. Water productivity as influenced by water management during 2007-08 boro. CSW = continuous standing water, 3 DAWD = water applied at 3 days after water disappearance, and 5 DAWD = water applied at 5 days after water disappearance.

213

Figure 6. Rainfall, evaporation and fluctuation pattern of water table in different water regimes during 2008-09 boro season.

-350.0

-300.0

-250.0

-200.0

-150.0

-100.0

-50.0

0.0

50.0

100.0

150.0

11.01

.09

21.01

.09

31.01

.09

10.02

.09

20.02

.09

02.03

.09

12.03

.09

22.03

.09

01.04

.09

11.04

.09

21.04

.09

01.05

.09

Date

Wat

er le

vel (

mm

)

0.0

10.0

20.0

30.0

40.0

50.0

60.0

RF

& E

V (m

m)

RF EV CSW 3 DAWD 5 DAWD

214

Figure 7. Soil moisture content at different soil depth during alternate wetting and

drying treatments during 2008-09 boro season.

0.00

10.00

20.00

30.00

40.00

0-15 15-30 30-45Soil depth (cm)

MC

(%)

3 DAWD 5 DAWD

215

Paper 9 Aerobic and alternate wetting and drying rice production systems in Pakistan: progress and challenges

Riaz A. Mann, Muhammad Ijaz, and Shahbaz Hussain

When looking at the serious water and farm labor scarcity in Pakistan, it seems very difficult to grow rice with conventional methods (puddled, transplanted, and flooded). Hence, to overcome this problem, the Pakistan Agricultural Research Council (PARC), Islamabad, collaborated with the International Rice Research Institute (IRRI), Philippines, in an Asian Development Bank (ADB)-supported project titled “Development and dissemination of water-saving rice production technologies in South Asia” from 2006 to 2009. During this period, Pakistan received a total of 251 rice lines of different maturity groups and with suitability to aerobic or alternate wetting and drying (AWD) conditions from IRRI, out of which 59 lines were selected for further evaluation and use in the rice improvement program in Pakistan. Both water-saving rice production technologies, aerobic and alternate wetting and drying technologies, demonstrated a minimum water savings of 25% and 20%, respectively, over the conventional rice production method at the farmers' field level. Weed management in aerobic rice, a major concern, was tackled successfully with the use of herbicides such as ethoxysulfuron and sodium 2,6bis-benzoate, leading to control of most grasses/sedges and a significant increase in rice grain yield over that of untreated rice plots. Aerobic rice technology also improved the productivity of rice-based cropping patterns. In Sindh, rice-chickpea produced higher profit than rice-wheat and rice-mustard. However, in Sadhoke, Punjab, wheat yield increased by 13% over that of the crop planted after a conventional rice crop. Farmers' were educated about the technology through on-farm demonstrations of water- and labor-efficient innovative technologies and by organizing farmers' field days in Chiniot areas. Having a good experience with aerobic rice technology, PARC launched a project on the dissemination of this technology on a large scale. During 2009, aerobic rice technology was demonstrated at three sites (Sadhoke, Chunnian, and Chiniot) on 135 plots, one acre each. In 2010, the technology was disseminated in two districts, Hafizabad and M. Bahuddin, with very encouraging results. The collaborative project with IRRI immensely benefited our local scientists and research institutions through scientific interaction and training and benefited farmers through demonstrations of water- and labor-saving technologies.

Abbreviations: AWD = alternate wetting and drying; GoP = government of Pakistan; KSK = Kala Shah Kaku; OFWM = on-farm water management; PARC = Pakistan Agricultural Research Council; PVS = participatory varietal selection.

In Pakistan, rice is the second important staple food crop after wheat and a major export commodity after cotton. Rice accounts for 6.7% of value added to the agricultural sector, and it has a 1.6% share in total GDP. Rice contributes 11% to total cropped area in the country, and accounts for 17% of the total cereals produced annually. About 40% of rice production (2.6 million tons) is consumed

216

locally while 60% is exported mainly to the Middle East, Africa, Asia, and Europe. Pakistan is well known for producing aromatic rice (Basmati) and earns reasonable foreign exchange by exporting both aromatic and nonaromatic rice annually. By 2025, projected rice demand in Pakistan will be about 3.35 million tons. Research on rice in Pakistan commenced in 1920 at Larkana, Sindh, and in 1926 at the Rice Research Station, Kala Shah Kaku, in the Kallar tract of Punjab, which is the homeland of world-famous and fine-grain aromatic varieties, namely, basmati rice. In 1970, both the Rice Research Stations at Dokri (Larkana) and Kala Shah Kaku were upgraded to Rice Research Institutes, with a wider mandate. A coordinated research program on rice at PARC began in 1975. In addition to coordination and monitoring and evaluation of research and development activities, PARC conducts research on many aspects not being undertaken by other institutions. The research and development infrastructure in Pakistan has the potential to cater to the needs of rice growers, the rice industry, and foreign markets. At present, two provincial institutes are working solely on rice, while 11 other institutions (four national, five provincial, and two in the private sector) are also engaged in rice research and development in the country.

Rice is grown as a single crop in all provinces of Pakistan under four different agro-ecological zones, with a cropping period from May to November. Zone I consists of the northern mountainous areas of the country such as Swat Valley, where cold-tolerant rice varieties are cultivated (Figure 1). Zone II lies between the Ravi and Chenab rivers, called the Kallartract, where basmati rice is predominantly grown (Mann and Ashraf 2001). Zone III comprises the areas lying on the west bank of the Indus River, where high temperature-tolerant rice varieties such as IR6 are cultivated. Zone IV is the Indus delta, which consists of vast spill flats and basins. The climate is arid tropical marine with no marked seasons and is suited for growing coarse rice varieties. Punjab is the largest rice-producing province, with 68% of the area and a 59% share in production, followed by Sindh, with 24% of the area and 33% of production, Baluchistan, with 6% area and 6% production, and KPK, with 2% area and a 2% share in total rice production. During 1981-82, total rice area was 1.93 million ha, which increased to 2.67 million ha (38%) in 2009-10 (Table 1). Likewise, rice production during 1980-81 was 3.12 million tons, which rose to 6.68 million tons (a 114% increase) during 2009-10. This exponential increase in rice production (by 114%) and increase per hectare (by 54%) were possible due to the introduction of favorable and farmer-friendly government policies, institutional reforms, the development of high-yielding and fertilizer-responsive varieties, and dissemination of better production technologies to farmers. However, both area and production declined to 2.28 million ha and 4.72 million tons during 2011-12 because of heavy rainfall and devastating floods in different parts of the country. The economic conditions of rice-growing farmers are generally weak compared with those of farmers of cotton or sugar cane-growing belts. This is because rice farmers own small and fragmented landholdings, and earn comparatively low farm income due to the high cost of rice production (mainly on irrigation water) and low paddy prices.

217

Current rice cultivation practices Rice in Punjab Province is traditionally grown with manual transplanting of 30-to 35-day-old seedlings (raised somewhere else) in flooded and puddled fields. However, in Sindh Province, the fields are not puddled. The entire rice fields are transplanted within a period of 45 to 60 days. The manual transplanted rice crop has only 150,000 seedlings per hectare compared with our recommendation of 200,000 to 250,000 seedlings per hectare. Low plant density in rice fields has been the major constraint to low paddy yield in Pakistan (PARC 1983). Poor crop management practices also contribute to low crop productivity. The present rice cultivation system is not very productive, resource-efficient, and sustainable. Rice yield is about one-half of the demonstrated potential yields obtained at research stations. Continuous puddling over the decades has led to deterioration in the physical properties of soil through structural breakdown of soil aggregates and capillary pores and clay dispersion. Puddling forms a compacted layer that restricts percolation of water, causing temporary water-logging, which restricts root penetration and growth of succeeding crops after rice. Nitrogen-use efficiency in paddy (≤40%) and low plant population are significant factors that lead to poor paddy yield with traditional methods. Thus, complete reliance on labor and the water-intensive rice cultivation method are not going to be economical and sustainable (Mann et al 2011). In order to grow more quality food from scarce resources (labor, water, inputs) on marginal/degraded lands, the current rice crop production system must be improved to make it more viable and eco-friendly. A few years back, social scientists conducted a baseline survey in the Punjab target area (nonconventional rice belt). The results revealed that 41% of the farmers use tube-well water while 53% use water from canals and tube wells (PARC 2007). Wetland preparation (puddling) is rarely done in this area. Only 6% of the farmers do partial puddling with only two to three wet plowings while 94% of the farmers reported only dry plowing along with one planking in the flooded field. For the first 25 days, farmers keep the paddy field in submerged conditions. Later on, the interval between two irrigations increases, ranging from 4 to 9 days. Almost all the farmers apply herbicides to control weeds in the paddy crop while hand weeding is not practiced. The cost of irrigation water is quite high, ranging from Rs. 8,500 ha−1 (through electric tube wells) to Rs. 20,000 ha−1

(through diesel tube wells). The high cost of irrigation indicates the severity of the water shortage in the area. A majority of the farmers use phosphate fertilizers (as a basal) and 100 to 150 kg ha−1 of nitrogenous fertilizers as topdressing. The use of phosphate fertilizers is not very common in the traditional rice area. However, paddy yield is much higher than that of the conventional rice area because of better soil quality. Some 68% of the farmers obtained paddy yield surpassing 4.0 t ha−1 from basmati rice varieties or surpassing 6.0 t ha−1 from coarse varieties. Small farmers do manual harvesting and threshing while combine-harvesting is also used on relatively larger areas. Tractor cultivation is common and there are many service providers. From the survey, it is very clear that farmers follow the

218

recommendations of rice experts for growing a paddy crop to achieve the potential yield. Water scarcity Pakistan has a total cultivated area of about 22 million ha, out of which 14.7 million ha or 68% of the area is primarily irrigated by the Indus Basin Irrigation System (IBIS). This system provides 54 MAF (million acre feet) of water at the farm gate and about 48 MAF of water are supplemented through one million tube wells. There is a shortfall of 22.7 MAF of water in agriculture, which is projected to double in the 2020s. The country has recently experienced a serious water deficit as high as 40%. Water resources, on both the surface and underground, have been shrinking rapidly since 1990. The situation is quite alarming and it threatens the nation’s current irrigated agriculture and national food security. All the rice cultivated in Pakistan is irrigated and it consumes 30% of the available freshwater resources. The luxurious use of water in rice production systems is becoming unaffordable for all farmers due to diminishing water resources and high energy cost. Wetland preparation (puddling) consumes 15% to 20% more water, leading to low yield of the following rabi (wheat) crop. Moreover, 70% of the one million tube wells installed in the rice area pump out brackish water, which poses a serious threat to crop productivity and soil sustainability (OFWM 2007). The groundwater level has declined by 10 feet, with little recharge. Farm labor in many rice-growing districts of Punjab has shifted to other sectors because of industrialization in rural areas. The situation is alarming and threatens the nation’s food security. Because of the increasing water scarcity, it is crucial to identify alternative rice crop production methods that require less irrigation water than that used in transplanted rice. Therefore, the substitution of transplanting by a low-cost method of crop establishment such as aerobic rice could be advantageous (Johnkutty et al 2002). In Punjab, more than 70% of the rice area is covered by basmati varieties (fine, long-grain, and aromatic), whereas coarse varieties cover Sindh Province (GoP 2011). Lahore and Gujranwala, Punjab, are considered as the traditional basmati rice belt, where high-quality basmati rice has been produced for many decades. The soil of this belt is heavy textured, poorly drained, and saline-sodic. During the monsoon season, rainfall is higher than in the adjacent area to the south. A shortage of canal water prevails while a large number of tube wells have been installed to supplement irrigation water for crop production. Over the past 12 to 15 years, rice area has expanded to the adjacent area of Faisalabad, Sahiwal, and Multan divisions because of the shift from cotton or sugarcane crops to rice (RRI 1998). In these areas, scarcity of irrigation water is common because of deep and brackish groundwater and the inadequate supply of canal water. The area has fertile, clay-loamy, and well-drained soils. The major cropping patterns are rice - wheat, rice – potato - maize, cotton – wheat, and maize - wheat. In the nonconventional rice area, a majority of the farmers do not practice puddling because they cannot afford a luxurious use of water during land preparation for rice. Farmers grow mixed varieties of rice (basmati as well as

219

nonbasmati), considering the crops that follow rice. However, the basmati rice produced in this area does not have the same grain quality and cooking characteristics as the rice from the traditional belt. In view of this, in the ADB-supported project, the decision was made to focus on the new rice area comprising the divisions of Sargodha, Sahiwal, and Multan (about 519,000 ha or 31% of the total Punjab rice area) for the dissemination of new water-saving rice technologies. In Sindh Province, although monsoon rainfall is quite low, the adequate supply of canal water during the rice season does not create any problem of water shortage in many areas of upper Sindh. However, the area at the tail end of the canal areas and some areas in the lower Sindh (downstream of the Indus River) do have a serious problem of water shortage. Sindh farmers do not puddle the soil during land preparation of the paddy crop. In Thatta Division (Lower Sindh), farmers are already practicing direct seeding of rice in wet fields. So, the canaltail area in the districts of Larkana and Shahdad Kot (Upper Sindh) and Thatta and Badin (LowerSindh) could be potential target areas (195,000 ha or 33% of the total rice area) in Sindh Province.

Water-saving rice production options

Aerobic rice is a new way to cultivate rice that requires less water than irrigated lowland rice. It entails the direct seeding of rice in aerobic soil, aiming at high yield. A shift from continuously flooded to aerobic conditions may have profound effects on the entire cropping system. Aerobic rice has several advantages over conventional rice transplanting in flooded and puddled soils. With aerobic rice, 30−35% water savings can be achieved and several operations such as nursery raising, its care, nursery uprooting, and its transportation and manual transplanting can be eliminated (Table 2). The rice crop in aerobic conditions does not lodge and land preparation after the aerobic rice harvest becomes easier, leading to improved wheat crop establishment and productivity. The aerobic rice cultivation system is more environment-friendly; as greenhouse gases such as carbon dioxide, methane, etc., are not produced. The salient features of the aerobic rice cultivation methodology follow.

Rice can be direct-seeded in two ways. First, in well- and finely prepared land, seed can be drilled into soil at 7-inch-wide rows. In this method, irrigation can be applied immediately. In the absence of a seed drill, farmers can broadcast rice seed in the well-prepared field in optimum moisture conditions, and then planking can be done to cover the seed. Soaking of seed for 10 to 15 hours in the broadcast method is recommended for better germination. Proper rice varieties for aerobic cultivation are under evaluation and in the experimental stage (Mann et al 2007). Seed depth should be 2 to 3 cm for a good crop stand. Placing seed below 3 cm adversely affect seed emergence because of rapid drying of the soil surface in peak summer days. The best planting time of aerobic

220

rice is 15 to 20 days earlier than the traditional transplanting period, which will enable earlier planting of the following wheat crop, resulting in its increased yield as against its late sowing in the traditional rice - wheat system.

The aerobic rice crop does not require continuous flooding and can be safely irrigated. The first irrigation can be given just after sowing; thereafter, plants require irrigation water, most probably at 8 to 12 days after sowing. Subsequent irrigations can be given at 5- to 8-day intervals depending on the soil type to keep the field saturated. Water stress should be avoided at tillering, panicle initiation, and grain-filling stages, which are crucial for obtaining higher yield. Again, an unleveled field causes wastage of land and low irrigation efficiency, resulting in lower yield. There are many ways to level the field, but the best one is through laser leveling. Thus, a leveled field ensures irrigation water savings, uniform seed germination, increased fertilizer-use efficiency, and higher crop yield. Aerobic or direct-seeded rice is being disseminated to nonconventional rice belts such as Sahiwal, Multan, Jhang, Bahawalpur, Thatta, and Badin districts.

The alternate wetting and drying (AWD) technique deals with transplanted rice. Transplanted and puddled rice crops can be successfully grown under the AWD method rather than continuously flooding the rice field. Transplanted rice 12 to 15 days after transplanting can be irrigated at 6- to 9-day intervals to keep the field at optimum moisture. In this method, more than a 20% water savings along with a substantial yield increment was achieved in a basmati rice crop over a conventional flooded crop. Moreover, high water productivity in the basmati crop with AWD was obtained due to the efficient use of soil nutrients and plant vigor. This confirms that a rice crop with AWD can be successfully grown without any major yield penalty. This technique can be successfully practiced in the conventional basmati-growing belt.

Testing and evaluation of local rice varieties During the 2006 rice season, field experiments with local rice varieties/lines were conducted to test their suitability under the new water-saving technologies (AWD and aerobic conditions) at both stations. At Kala Shah Kaku, variety KSK-133 among the nonbasmati group produced the highest rice grain yield (5.02 t ha−1) under AWD, whereas line # 99421 produced the maximum yield of 4.82 t ha−1

under aerobic conditions (Table 3). Variety KSK-133 has already been approved by the department for general cultivation in Punjab, with potential paddy yield of 7.2 t ha−1 under normal field conditions. Seed of this variety was distributed to the farmers of water-deficit areas. Likewise, among the basmati group, two varieties, Basmati 2000 and Super Basmati, had the highest yield under AWD and aerobic conditions (Table 3). At Dokri, varieties such as DR-92 and Shahkar produced the maximum rice yield under water-stress conditions (Table 4). Shahkar has recently been

221

released to farmers for general cultivation in Sindh Province. Obviously, the trial has generated valuable information that varieties such as KSK-133, Super Basmati, Basmati 2000, and Shahkar have wide adaptability and can perform well under water-saving production techniques. Farmers of the target area also received seed of some of the latest and modern high-yielding varieties in a short time.

Testing and evaluation of IRRI material for aerobic and AWD conditions During the entire course of the project (RETA 6267), both the Rice Research Institutes at Kala Shah Kaku and Dokri received a total of 251 lines from IRRI-Philippines during 2006-09, under the categories of aerobic, rainfed, and irrigated lowland for testing and evaluation under local environments. Our scientists at KSK selected seven lines out of 36 entries under the category of aerobic lowland (Table 5). The results revealed that the aerobic entries with designations IR79597-56-1-2-1, IR80416-B-152-4, IR80416-B-32-3, and IR70210-39-CPA-7-1 produced higher grain yield than the local checks. The local check (KSK-133) produced a maximum yield of 5.6 t ha−1, while the IRRI entries produced yield in the range of 5.4 to 5.9 tha−1. A sufficient amount of seed was produced for such lines, which was used in a further breeding process. Likewise, seven lines were selected out of 41 (rainfed lowland) IRRI entries at KSK (Table 6). Entry IR82082-B-B-96-1had grain yield of 6.1 t ha−1, followed by IR82311-B-B-1-2 and IR82293-B-B-31-1, with grain yield of 6.0 t ha−1. Most of these entries have desired traits. Meanwhile, the local checks produced grain yield of 5.5 to 6.1 t ha−1. These lines have up to 110 days’ maturity period, with moderate plant height and high tillering capacity. Likewise, nine entries were planted at Thatta (Lower Sindh). Rice grain yield is presented in Table 7, which revealed that the IRRI lines took 110 to 116 days to mature, with plant height of 97 to 125 cm. The grain yield varied from 4.02 to 4.68 t ha−1 against 4.41 t ha−1 of the local check (DR-92). The grain yield of IRRI lines is highly encouraging and the lines have the potential to replace the year-old varieties. There is now a desire to advance to seed multiplication on a large scale. These lines were tested for another crop season to further evaluate their performance under local environments. The exchange of this valuable material from IRRI played an important role in improving the local breeding program for water-saving rice conditions.

Participatory varietal selection under aerobic and AWD conditions The most promising rice varieties were planted under the AWD system on Haji Sons Farm, near Chiniot, Punjab. The result revealed that maximum paddy yield of 8.4 t ha−1 was produced by IR82082-B-B-96-1, followed by KSK-282 with 8.2 t ha−1 (Table 8). Plant height ranged from 110 to 134 cm while productive tillers varied from 190 to 384 m−2. Likewise, in Sindh Province, eight IRRI lines along with two local checks were drilled, using a seed rate of 50 kg ha−1. A fertilizer

222

dose at 120-60-37 kg NPK ha−1 was applied with all phosphorus and potash as basal. The results (Table 9) revealed that grain yield ranged from 1.72 to 5.34 t ha−1. The entries tested under a mother trial were of medium stature and had medium maturity. The lines IR67017-8-2-1-141 and IR82068-99-1-3-3 took more days to mature than other entries. Entry IR79906-B-192-2-4 produced significantly higher yield (5.34 t ha−1) than check varieties DR-92 and Shahkar. The same entry produced comparable longer panicle. Farmers of the area visited the PVS fields at the flowering stage of the rice crop and showed great interest in the performance of rice varieties. The objective had been to evaluate the promising rice varieties at the farm level and convince neighboring farmers for their adoption in the next season. KSK-133 was demonstrated in the Chiniot area while Shahkar and DR-92 were demonstrated in the Larkana area of Sindh Province. Farmers' participatory research played a major role in disseminating the innovative technology and varieties in the shortest possible time.

Suitable crop rotations for aerobic and AWD technologies In Pakistan, wheat yield after the paddy harvest is always lower than that of wheat planted after nonpaddy crops. The reason is that puddling leads to deteriorated physical properties of soil through structural breakdown of soil aggregates and capillary pores and clay dispersion. Puddling forms a compacted layer (plow plate) that restricts percolation of water, causing temporary water logging, which restricts root penetration and growth of the succeeding wheat crop. Moreover, land preparation in the paddy field becomes difficult and it requires high cost because of several plowings, diskings, and harrowings to pulverize the soil. Contrary to this, land preparation in the aerobic rice field is very easy and the soil becomes soft with less cost and time. The results indicate that productive tillers, spike length, total grains/spike, and grain yield of a wheat crop following aerobic rice increased substantially over that of the crop when it was raised after flooded and puddled fields (Table 10). Hence, the present rice cultivation system needs to be replaced with aerobic technology for sustaining the productivity of the rice-wheat cropping system and improving farmers' livelihoods and environmental protection. At RRI-Dokri, an experiment continued for three rice-based cropping patterns, rice -wheat, rice - chickpea, and rice - mustard. The seeding rate for rice was 50 kg ha−1, with line-to-line distance of 20 cm. The results depicted that rice-chickpea was the most productive and profitable cropping pattern compared with the rice-wheat or rice - mustard rotation. Rice-chickpea gave an average net income of Rs. 57,434 ha−1, followed by rice-wheat with net income of Rs. 32,104 ha−1 (Table 11). It is evident that the yield of upland crops following puddled rice is badly affected because of anaerobic soil conditions, difficult land preparation, and poor crop establishment. On the other hand, the aerobic rice system does not pose a serious threat to the productivity and sustainability of wheat or pulse crops, following rice, due to favorable soil structure, good tillage, and good crop establishment. Hence, there is a great need to promote pulses in the rice-based

223

cropping system to reduce the import bill of pulses to meet the demand of the ever-increasing population in Pakistan. Suitable weed control and management strategies Weed infestation continues to be a serious problem in dry-seeded rice. Aerobic soil conditions and dry-tillage practices, besides alternate wetting and drying conditions, are conducive to the germination and growth of highly competitive weeds, which cause grain yield losses of 50% to 91% (Elliot et al 1984, Fujisaka et al 1993). Manual weed control is the most expensive method that none of the farmers could afford. Chemical weed management is considered as the cheapest technique as many new products are available in the market (Gupta et al 2003). Several preemergence herbicides, including butachlor, thiobencarb, pendimethalin, oxadiazon, oxyfluorfen, and nitrofen alone or supplemented with hand weeding, have been reported to provide a fair degree of weed control (Janiya and Moody 1988, Moorthy and Manna 1993, Pellerin and Webster 2004). An experiment was conducted on the farm of RRI-Kala Shah Kaku, using direct-seeding technique with seven treatments. Basmati rice variety Super Basmati was used at a seed rate of 30 kg ha−1 with direct seeding. The treatments included (1) postemergence herbicide, Nominee, at 250 mL ha−1 (ethoxysulfuron) at 2 to 3 weeks; (2) postemergence herbicide (Nominee) + one hand weeding in the fifth week; (3) preemergence herbicide, Sunstar, at 200 g ha−1 (sodium 2,6 bis-benzoate); (4) preemergence herbicide (Sunstar) applied after seeding + one hand weeding in the fifth week; (5) one hand weeding in the third week after seeding; (6) two hand weedings in the third and fifth week; and (7) no weeding. The trial was conducted with a spilt-plot design with three replications. The recommended dose of fertilizer (120-60-0 kg NPK ha−1) with all P as basal and N in three splits (at 25, 50, and 70 days after seeding) was applied. From the data during 2008, it is evident that one application of postemergence herbicide (Nominee) along with one hand weeding was quite effective for controlling weeds in dry direct-seeded aerobic rice, with grain yield of 4.0 t ha−1 (Table 12). The untreated plots produced grain yield of only 0.9 t ha−1. The single application of Nominee gave yield of 3.8 t ha−1. The number of productive tillers varied from 30 to 297 m−2. Preherbicide weeds ranged from 480 to 577 m−2, whereas the postherbicide weed count was 9 m−2. Sedges were predominantly found in the direct-seeded rice crop, followed by broad leaf weeds (Table 13). After the use of herbicides and manual weeding, weed density declined sharply. Nominee, a postemergence herbicide, effectively controlled grasses and sedges in the rice crop. The grain yield of the direct-seeded rice crop was adversely affected by the presence of weeds that are difficult to control through cultural and manual practices. Chemical weed control remains an economical and sustainable technique that can work only with the use and availability of an effective herbicide. New products of postemergence herbicides have been found effective to control grasses such as Echinochloacolona or E. crus-galli and sedges such as Cyperusrotundus and C. iria. With the introduction and use of

224

herbicides such as Sunstar and Nominee, it is hoped that the area under cultivation of direct-seeded aerobic rice in the country will increase.

Suitable water management strategies for aerobic and AWD conditions The water requirement of lowland rice varies from 1,650 to 3,000 mm (Tuong and Bouman 2003, Lampayan and Bouman 2005). The aerobic rice production system eliminates continuous seepage and percolation losses, greatly reduces evaporation as no standing water is present at any time during the cropping season, and effectively uses rainfall and thus helps in enhancing water productivity and concomitant loss of soil sediments, silt, and fertility from the soil. A comparison of the water requirement of lowland flooded rice and aerobic rice clearly showed that the aerobic rice system can save about 45% of water (Shashidhar 2007). In aerobic and AWD rice plots, water use was measured by the number of irrigations and volume of water used per field. Water use in the rice crop was measured by a “water meter,” while the groundwater table in the field was monitored by 2-m PVC pipe. The water contribution through rainfall was also taken into consideration. An experiment was conducted with four local basmati varieties (Bas-385, Super Basmati, Bas-2000, and Bas-515), transplanted in main plots with two irrigation regimes (AWD and flooded) as subplots. Necessary data regarding plant height, productive tillers, yield, and water applied, and water saved were recorded from this study. The data showed that the basmati rice crop with AWD technique used 975 mm of water compared with 1,320 mm of water for the conventional flooded rice crop, contributing about 35% water savings in the case of AWD (Table 14). However, paddy yield with the AWD technique was slightly lower. From a number of water-saving studies in the past, it is concluded that 30% to 35% water can be saved in basmati rice through the AWD technique. To avoid a yield reduction and attain similar quality analysis, the safe-AWD technique that consists of keeping 5 cm of water in the field upto 2 weeks after transplanting and for 2 weeks during flowering and irrigating the rice field following AWD when water-table depth in the field reaches 15-cm depth as can be measured in PVC pipes can be applied in basmati rice. The safe-AWD technique causes no yield penalty, but provides water savings of 20% to 23%. Up-scaling of water-saving technologies Keeping in view the significance of water scarcity in rice cultivation, PARC, through funding from the government of Pakistan, launched a project on up-scaling of water-saving rice production technologies during 2009-11. The demo plots were conducted at 25 locations of three main rice-growing areas, Sadhoke (Gujranwala District), Chunian (Kasur District), and Chiniot, of Punjab Province. In this regard, cooperation from the agricultural extension department was also sought. At each site, dry-seeded rice with AWD was compared with the conventional farmers’ practice (puddled, transplanted, and continuously flooded rice). A total of 135 plots (one acre each) were established. All the inputs

225

(fertilizer, seed, herbicides, and pesticides) were supplied to the farmers by the project team while farmers managed the rice plots with varieties Super Basmati and KSK-133, under the supervision of the project team. The results revealed that dry-seeded rice crops used an average of 1,410 mm of irrigation water against 1,850 mm with the conventional farmers’ practice (Table 15), a 25% irrigation water savings. The maximum water savings was recorded at Sadhoke site, on clay soil (with high water-holding capacity and low percolation rate), while the lowest savings was recorded at Chunian site, on light-textured soil. At all three sites, dry seeding resulted in more productive tillers, with site averages of 388 to 451 m−2 and an average increase of 48.6% over puddled transplanted rice (Table 16). These results are consistent with those of Wiangsamutt et al (2006) and Ullah et al (2007), who found more productive tillers in dry-seeded rice than in transplanted rice. The average yield of dry-seeded rice was 5.35 t ha−1 compared with 4.20 t ha−1with the conventional farmers’ practice, a 27.3% increase (Table 17). These results are also consistent with those of Wiangsamutt et al (2006), Ullah et al (2007), and Gangwar et al (2008), who recorded maximum rice yield from dry seeding. Partial budget analysis showed that dry seeding of rice generated Rs. 32,550 ha−1 (US$ 383 cha−1) more than the conventional rice production practice. Thus, dry seeding with AWD can help overcome the problems of low plant density leading to low yield, and scarcity of water and labor, in the current rice farming system, leading to enhanced land and water productivity and farmers’ income. On some farms, however, grain yield under dry direct seeding was slightly lower than in the conventional transplanted/puddled rice crop. Farmers were quite happy to save high-value tube-well water, with considerable grain yield. The farmers showed their interest in adopting the water-saving technology on a larger area in the next season. In Lower Sindh (Thatta and Badin districts), baby trials at 10 sites were also established with dry direct seeding. A seed rate of 50 kg ha−1 was used and sowing was done in June. Grain yield varied from 4.15 to 4.55 t ha−1, which was on a par with the control plots. However, farmers reported that they saved three to six irrigations compared to traditional planting. It was a new experience to work with the rice farming community in Lower Sindh, and farmers took a great interest in the aerobic rice cultivation system. However, they did not strictly follow the irrigation schedule of the project teams, thinking that it would drastically reduce grain yield. In the following year, PARC launched the project in two major districts, Hafizabad and Mandi Bahuddin, in collaboration with the Agricultural Extension Department. A total of 398 rice plots with variety Super Basmati were established in two districts. The field results showed a 60% increase in productive tillers, 26% water savings, and grain yield increment of 25% in aerobic rice over the conventional rice crop in Hafizabad (Table 18). In Mandi Bahuddin, productive tillers increased by 42%, grain yield by 18% and water savings by 15%. Aerobic rice technology does not include rice nursery raising, its uprooting, shifting and transplanting, puddling, and continuously flooding the field. Farmers are very

226

keen to adopt this technology in the next rice season, as it is cost-effective and highly productive. There was no observation of iron deficiency in both districts. However, zinc sulfate at 25 kg ha−1 was applied in all rice plots. Farmers’ education was given much emphasis and importance so that the rice farming community could adopt modern water-saving rice production technologies in both provinces. Farmers remain the most important segment of rice stakeholders and they are the best judges of any technology. Therefore, meetings with the farmers of target environments play an important role in gaining their confidence on new technologies or varieties. In this regard, a farmers’ gathering was held with the cooperation of agricultural extension staff during the rice season in Punjab and Sindh. Farmers’ field days were also organized at the key demonstration plots, in which a large number of farmers visited the demo plots and showed their great interest in adopting dry seeding or AWD on their own farms. Such gatherings have been proven to be very effective in educating and motivating farmers to adopt improved/innovative rice production practices. Leaflets and brochures on aerobic rice production techniques and AWD were printed and distributed among the farmers in local dialect. Farmers were found very keen to learn/adopt new technologies, which would lead to improved sustainability and productivity of rice-based cropping systems. All of this was very helpful to disseminate water-saving rice production technologies in the target areas.

Policy decisions It is very difficult to work with farmers rather than work for farmers. Farmers preferred to learn from the experience of their fellow farmers rather than listening to experts. Healthy and positive feedback from farmers is a good indication for the promotion of scientific findings in the country. Learning from the experience of the ADB-funded project (RETA 6267), the Pakistan Agricultural Research Council launched a project with a total cost of Rs. 95 million (funding by the Ministry of Food and Agriculture) in two provinces (Punjab and Sindh). The results were very encouraging and convincing to policymakers, ministry higher-ups, and other provincial key managers. This stimulated other researchers in universities and provincial institutions to develop research proposals/projects and local funding agencies in the public sector allocated funds on this important aspect. The Agricultural Extension Department, particularly of Punjab Province, also played a major role in up-scaling of water-saving rice production technologies in nonconventional basmati areas. At present, researchers of three institutions are working on this aspect with funding from local agencies.

Conclusions

In Pakistan, water scarcity for rice production is becoming a crucial issue because of the shortage of canal water and high pumping cost of underground

227

water. The water-saving rice production system aimed at food stability and security would help reduce the heavy reliance on water use. During the past five years, activities such as testing and evaluation of IRRI material for aerobic, rainfed lowland, and drought areas; improved rice establishment techniques; weed management in DSR; seed multiplication of promising varieties; baby trials; and farmers’ meetings have made significant progress. A lot of valuable information/data has been generated toward developing a complete package of water-saving rice production technologies in the country. From IRRI germplasm, more than 59 lines have been selected for further varietal testing in Pakistan. Research trials on the AWD system, crop establishment techniques, weed management in direct- seeded rice, rice-based cropping patterns, and PVS and plant sampling for soil-borne diseases have produced very encouraging results. Likewise, technology dissemination/demonstration activities in the target environment helped a lot to buildup new confidence of farmers in innovative water-saving rice production technologies. Extension department staff played a major role in farmers’ identification and site selection for the establishment of demonstration plots, and up-scaling of such technologies in the Chiniot area. The PARC also took initiative to upscale water-saving technologies in rice at five sites in Punjab and Sindh. At such sites, water savings from 25% to 32% was recorded with a 14% yield increment over the farmers’ practice. Aerobic rice technology contributed Rs. 33,000 ha−1 (US$402) over conventional rice cultivation techniques. Overall, the IRRI-executed project has been successful in the context of generating new knowledge for overcoming the problem of water scarcity in rice production systems and strengthening the research capability of two major provincial rice research institutions. If the varieties for water-short conditions and water-saving technologies are adopted on about 50% of the area in the target environment, comprising the divisions of Faisalabad, Sargodha, Multan, and Bahawalpur, total savings in terms of less water use in rice production systems could be as high as Rs. 17.32 billion ($211 million). The considerable increase in rice area due to new rice varieties and technologies would be another advantage. Hence, innovative varieties or crop establishment technologies would highly benefit the rice farmers of the nonconventional rice belt of Punjab and farmers of Lower Sindh, thus increasing their farm income and improving their prosperity and livelihoods.

References Elliot PC, Navarez DC, Estario BD, Moody K. 1984. Determining suitable weed

control practices for dry-seeded rice. Philipp. J. Weed Sci. 11:70-82. Fujisaka S, Moody K, Ingram K. 1993. A descriptive study of farming practices for

dry seeded rainfed lowland rice in India, Indonesia and Myanmar. Agric. Ecosyst. Environ. 45:115-128.

Gangwar KS, Tomar OK, Pandey DK. 2008. Productivity and economics of transplanted and direct-seeded rice (Oryza sativa)-based cropping systems in Indo-Gangetic plains. Indian J. Agric. Sci. 78(8):655-658.

228

GoP (Government of Pakistan). 2011. Agriculture Market News Information from APEDA. Gupta RK, Naresh RK, Hobbs PR, Jiaguo Z, Ladha JK. 2003. Sustainability of

post-green revolution agriculture: the rice–wheat cropping systems of the Indo-Gangetic Plains and China. Rice Wheat Consortium, New Delhi, India.

Janiya JD, Moody K. 1988. Effect of time of planting, crop establishment method, and weed control method on weed growth and rice yield. Philipp. J. Weed Sci. 15:6-17.

Johnkutty I, Mathew G, Mathew J. 2002. Comparison between transplanting and direct-seeding methods for crop establishment in rice. J. Trop. Agric. 40:65-66.

Lampayan RM, Bouman BAM. 2005. Management strategies for saving water and increasing its productivity in lowland rice-based ecosystems. In: Proceedings of the first Asia-Europe workshop on sustainable resource management and policy options for rice ecosystems (SUMAPOL), 11-14 May 2005, Hangzhou, Zhejiang Province, China. On CD, Altera, Wageningen, Netherlands.

Mann RA, Ashraf M. 2001. Improvement of basmati and its production practices in Pakistan. In: Chaudhary RC, Tran DV, Duffy R, editors. Specialty rices of the world: breeding, production and marketing. Rome: Food and Agricultural Organization of the United Nations. p 129-148.

Mann RA, Hussain S, Saleem M. 2011. Impact of dry seeding with alternate wetting and drying on rice productivity and profitability in Punjab, Pakistan. In: Resilient food system for a changing world. Proceedings of the 5th world congress on conservation agriculture, Brisbane, Australia. p 395-396.

Mann RA, Ahmad S, Hassan G, Baloch MS. 2007. Weed management in direct seeded rice crop. Pak. J. Weed Sci. Res. 13(3-4):219-226.

Moorthy BTS, Manna GB. 1993. Studies on weed control in direct seeded upland rainfed rice. Indian J. Agric. Res. 27:75-80. OFWM. 2007. Tube well. Water Analysis Report, On-Farm Water Management, Department of Agriculture, Government of Punjab, Lahore.

PARC. 1983. Rice maximization report. Rice Program, National Agricultural Research Centre, Pakistan Agricultural Research Council, Islamabad.

PARC. 2007. Annual report of Rice Program. National Agricultural Research Centre, Pakistan Agricultural Research Council, Islamabad.

Pellerin KJ, Webster EP. 2004. Imazethapyr at different rates and timings in drill and water seeded imidazolinone-tolerant rice. Weed Technol. 18:223-227.

RRI (Rice Research Institute). 1998. Annual report for 1996-97. Rice Research Institute, Kala Shah Kaku, Lahore.

Shashidhar HE. 2007. Aerobic rice: an efficient water management strategy for rice production. In: Aswathanarayana U, editor. Food and water security in developing countries. London (UK): Routledge, Taylor and Francis. p 131-139.

229

Tuong TP, Bouman BAM. 2003. Rice production in water-scarce environments. Paper presented at the Water Productivity Workshop, 12-14 Nov. 2001, Colombo, Sri Lanka.

Ullah MA, Zaidi SAR, Razzaq A, Bokh SDH. 2007. Effect of planting techniques (direct seeding vs. transplanting) on paddy yield in salt-affected soil. Int. J. Agric. Biol. 9(1):179-180.

Wiangsamutt B, Mendoza CT, Lafarge AT. 2006. Growth dynamics and yield of rice genotypes grown in transplanted and direct-seeded fields. J. Agric. Technol. 2(2):299-316.

Notes Authors’ addresses: Riaz A. Mann and Shahbaz Hussain, Pakistan Agricultural

Research Council, Islamabad; Muhammad Ijaz, Rice Research Institute, Kala Shah Kaku, Pakistan.

Figure 1. Rice-growing zones of Pakistan.

230

Table 1. Rice area, production, and yield during last 12 years.*

* API2012.

Years Area (000 ha)

Production (000 t)

Rice yield (kg ha−1)

1981-82 1,933 3,125 1,620 2000-01 2,377 4,803 2,021 2001-02 2,115 3,882 1,834 2002-03 2,201 4,228 2,013 2003-04 2,440 4,876 1,998 2004-05 2,501 4,948 1,978 2005-06 2,620 5,548 2,118 2006-07 2,575 5,693 2,211 2007-08 2,520 5,540 2,198 2008-09 2,963 6,952 2,347 2009-10 2,673 6,687 2,501 2010-11 2,379 5,393 2,267 2011-12 2,590 6,280 2,425

231

Table 2. Field operations with different rice production systems.

Rice culture

Dry land preparation

Nursery raising

Nursery uprooting

Puddling Nursery shifting

Manual transplant

Flooding Optimum moisture

Conventional √ √ √ √ √ √ √ x

AWD √ √ √ √ √ √ x √ Dry aerobic

√ x x x x X x

√

232

Table 3. Grain yield (t ha−1) of local rice varieties/lines, planted with three methods, RRI, KSK, Punjab, in 2006.

*Check.

Variety/line Grain yield (t ha−1) Transplanted and flooded

AWD Dry direct seeding

Nonbasmati group IR6* 4.53 4.36 3.86 IR9 5.04 4.38 4.71 KS 282 4.87 4.34 4.22 KSK-133 4.98 5.02 4.16 99421 4.96 4.12 4.82 IR54742-31-9-26-15-

2 4.44 4.11 4.25

PK-3300-12-2 4.24 4.34 4.21 PK-3699-43 4.32 4.18 4.07 PK-3355-5-1-4 4.60 3.80 4.60 99723 3.56 3.44 2.67 Basmati group Bas 370 4.27 3.64 3.85 Bas 198 4.00 3.82 3.90 Bas Pak 3.57 3.73 2.99 Bas 385 4.42 4.10 2.67 Super Basmati* 4.30 4.20 4.59 Basmati 2000 4.77 4.24 4.10 60001 4.58 4.07 3.12 49731 3.75 3.54 2.83 40265 3.67 3.87 3.57 52773-2 4.45 3.75 3.10 33608 4.31 4.39 2.44 97502 4.41 3.86 2.09 Shaheen Basmati 3.32 3.57 2.51 98316 3.67 3.64 3.32 99417 2.75 2.82 1.67 98506 3.57 3.53 2.19 PK-5261-1-2-1 4.28 3.99 2.36 PB-95 3.64 3.35 2.04 SSRI-8 3.15 3.46 1.52 SSRI-13 3.49 4.13 1.50

233

Table 4. Grain yield (t ha−1) of local coarse rice varieties/lines, planted with three methods, RRI, Dokri, Sindh, in 2006.

Variety/line Grain yield (t ha−1)

Transplanted and flooded

AWD Direct seeding

IR8 7.35 4.99 4.53 IR6* 6.23 4.10 3.06 DR 82 5.71 4.88 2.11 DR 83 5.53 4.57 4.16 DR 92 6.99 5.49 5.99 S. Hayat 5.60 2.60 2.48 Lateefy 3.55 2.28 2.38 DR 64 6.61 4.21 4.08 DR 65 4.16 2.32 2.42 DR 66 4.26 2.52 2.43 DR 58 7.48 4.38 3.10 DR 67 4.52 3.36 3.27 Shahkar* 8.36 5.74 5.22 LSD0.05 0.80 0.17 0.53 CV (%) 8.22 8.42 7.30

*Check.

234

Table 5. Selected lines from advanced yield trial-I (aerobic) at RRI farm, KalaShah Kaku, in 2008.

Variety/line Plant

height (cm) Effective

tillers plant−1

Panicle length

Paddy yield

(t ha−1) IR72 79.8 14.6 26.3 5.5 IR77080-B-34-1-1 108.9 11.8 25.6 5.7 IR79597-56-1-2-1 104.9 12.2 26.2 5.9 IR80416-B-152-4 111.5 12.9 26.2 5.8 IR80416-B-32-3 107.0 14.7 26.8 5.9 IR70210-39-CPA-7-

1 113.7 14.7 24.1 5.8

IR82870-11 82.1 14.1 24.3 5.4 KSK-133* 105.1 13.9 25.4 5.6 KS 282* 102.1 14.6 26.6 5.2 KSK-401* 102.5 16.1 30.9 5.6 IR6* 98.2 14.1 25.8 5.0 LSD0.05 1.52 1.23 0.96 0.16

CV (%) 7.59 6.53 5.37 6.56 *Check.

235

Table 6. Selected lines from advanced yield trial-II (rainfed lowland) out of 47 lines, at RRI farm, KSK, in 2008.

Variety/line Plant

height (cm)

Effective tillers plant−1

Panicle length (cm)

Grain yield (t ha−1)

IR82287-B-B-77-2 117.9 15.9 25.4 5.6 IR82082-B-B-96-1 132.1 16.1 25.9 6.1 IR82298-B-B-86-1 120.3 14.4 24.1 4.4 IR81388-B-B-66-3 119.5 14.2 22.6 5.1 IR81026-B-126-4-

2 106.8 11.4

26.4 5.6 IR82311-B-B-1-2 115.1 11.7 24.8 6.0 IR82293-B-B-31-1 136.4 11.8 27.9 6.0

KSK-133* 106.3 14.3 26.7 5.5 KS 282* 93.1 16.7 26.0 5.6

KSK-401* 87.8 17.0 26.4 6.1 IR6* 96.9 16.6 26.0 5.5 LSD0.05 3.35 2.34 NS 0.41 CV(%) 6.24 5.23 6.72 6.85

236

Table 7. AYT-aerobic, 110−120 days’ duration planted at RRI, Dokri, in 2008. Entry Days to maturity Plant height

(cm) Grain yield

(t ha−1) IR81413-B-B-75-2 111 107 4.53 IR81449-B-B-109-3 112 112 4.58 IR81449-B-B-116-4 113 92 4.48 IR81449-B-B-128-1 113 102 4.54 IR81449-B-B-51-4 112 101 4.64 IR82319-B-B-103-2 116 115 4.73 IR81454-B-B-57-1 107 105 4.51 IR82098-B-B-18-2 117 124 4.27 IR82320-B-B-29-2 108 103 4.45 DR 92* 113 102 4.55

LSD0.05 1.51 1.43 0.55 CV(%) 6.78 5.77 6.28 *Check.

237

Table 8. Participatory varietal selection: yield trial at Haji Sons Farm, Chiniot, Punjab.

Variety Plant height

(cm) Effective tillers m−2

Panicle length (cm)

Grain yield (t ha−1)

IR82082-B-B-96-1 128.2 254 25.0 8.4 IR82311-B-B-1-2 126.0 227 28.0 7.4 IR82293-B-B-31-1 101.4 248 25.0 7.5 IR79597-56-1-2-1 119.2 210 24.8 6.2 IR80416-B-152-4 126.0 190 25 7.2 IR80416-B-32-3 134.2 193 29 7.6 IR70210-39-CPA-7-1 117.6 384 30 8.2 Mean 121.7 243.7 26.7 7.5 LSD0.05 4.05 1.96 1.45 1.30 CV(%) 6.87 5.33 6.21 9.80

238

Table 9. Participatory varietal selection trial, near RRI-Dokri, Larkana, in 2008.

Entry Days to maturity

Plant height (cm)

Panicle length (cm)

Spikelet’s panicle-1

Grain yield (t ha−1)

IR67017-8-2-1-141 114a 83.7e 22.7bc 107.3cd 1.7f IR83885-1-1-4 108bc 93.7cd 19.3d 87.0f 2.8de IR82068-99-1-3-3 115a 103.7b 22.0c 94.0ef 3.5cd IR80312-6-B-3-2-B 108c 108.0b 23.3bc 102.0de 4.4bc DR 92 106c 101.7bc 26.7a 115.0bc 4.4abc Vandana 92e 105.0b 22.3bc 120.0ab 4.0bc IR789787-B-22-B-B 102d 88.7de 21.7cd 104.3d 2.2ef IR80013-B-141-4-1 107c 108.0b 25.0ab 121.0ab 5.0ab Shahkar* 114a 88.0de 26.7a 128.0a 4.9ab IR79906-B-192-2-4 111b 123.7a 26.7a 107.0cd 5.4a

Mean 107.7 100.4 23.6 108.6 3.8 *Check. values with similar letter(s) are not significantly different at p<0.05.

239

Table 10. Effect of rice planting methods on the following wheat crop, Sadhoke, in 2009.

Rice planting method

Plant height (cm)

Effective tillers m−2

Spike length (cm)

Grains spike−1

Grain yield (t ha−1)

Conventional (transplanted and flooded)

96a

226b

9.34a

44.56a

3.76b

Aerobic rice 92a 265a 9.56a 47.85a 4.25a % Increase −4.17 17.5 2.3 7.4 13.0*

* significant at P<0.05 values with similar letter(s) are not significantly different at p<0.05.

240

Table 11. Aerobic rice-based cropping pattern as tested at RRI, Dokri, 2007.

Cropping pattern

Grain yield (t ha−1)

Variable cost (Rs. ha−1)

Gross income (Rs. ha−1)

Net income (Rs. ha−1)

Rice - wheat Rice = 3.49 Wheat = 3.22

64,246

96,350

32,104

Rice - chickpea Rice = 4.15 Chickpea = 2.93

51,891

109,325

57,434

Rice - mustard Rice = 4.63 Mustard = 0.48

50,655

70,725

20,070

Note: Prices: rice = Rs. 526/40 kg; wheat = Rs. 625/40 kg; chickpea = Rs. 747/40 kg; mustard = Rs. 820/40 kg. US$1= Rs. 60.

241

Table 12. Weed management in DSR at RRI, KSK, in 2008.

Treatment Number of weeds m−2

(before herbicide use)

Number of weeds m−2

(after herbicide)

Plant height (cm)

Effective tillers−2

Panicle length (cm)

Grain yield (t ha−1)

Nominee 570.7 24.7 110.2 257.0 26.8 3.8 Nominee + hand weeding

549.3 9.0 103.7 296.7 28.2 4.0

Sunstar 528.7 137.7 99.5 89.7 22.5 1.7 Sunstar + hand weeding 547.7 67.0 101.8 224.3 23.6 3.3 One hand weeding 526.7 44.0 101.7 218.3 24.3 2.7 Two hand weedings 486.3 30.7 108.8 269.3 26.4 3.3 Control 577.0 645.3 95.0 29.7 19.4 0.9 Mean 540.9 136.9 103.0 197.1 24.4 2.8 LSD0.05 0.65 0.26 0.31 0.40 0.38 0.67 CV(%) 9.12 6.34 8.53 6.49 8.23 9.84

242

Table 13. Weed species in DSR at RRI, KSK, in 2008. Treatment No. of weeds (before herbicide

application) No. of weeds (after herbicide

application) Grasses Sedge

s Broad leaves

Grasses Sedges Broad leaves

Nominee 19.0 358.7

148.3 20.0 47.7 70.0

Nominee + hand weeding 22.7 382.3

146.3 6.7 25.3 35.0

Sunstar 26.7 388.7

159.3 1.3 8.3 15.0

Sunstar + hand weeding 19.7 403.3

119.3 0.3 5.3 3.3

One hand weeding 21.0 362.

7 144.3 6.3 25.0 12.7

Two hand weedings 19.0 368.

3 97.0 2.0 17.7 11.0

Control 19.0 454.

3 103.7 23.3 476.7 145.3

Mean 21.4 388.3 131.2 8.6 86.6 41.8 LSD0.05 0.94 0.35 0.26 0.18 1.34 0.23 CV(%) 5.67 6.33 5.11 7.15 5.09 6.31

243

Table 14. Grain yield and water savings in four basmati rice varieties with AWD

and conventional puddled rice at RRI, KSK, in 2008.

Variety Treatment Plant height (cm)

Tillers plant−1

Yield (t ha−1)

Water applied (mm)

Super Basmati AWD 114.0 13.3 4.45 975 Basmati 2000 AWD 133.7 13.3 4.51 975 Basmati 385 AWD 127.3 12.7 4.11 975 Basmati 515 AWD 126.0 14.0 4.42 975 Super Basmati Flooding 120.3 14.3 4.74 1,350 Basmati 2000 Flooding 137.1 15.1 4.88 1,350 Basmati 385 Flooding 129.5 14.6 4.61 1,350 Basmati 515 Flooding 130.2 14.5 4.64 1,350 Water savings in AWD 38.4% Cost of water savings Rs. 8,500 ha-1 Cost of yield difference Rs. 7,000 ha-1 Net monetary benefit Rs. 1,500 ha-1

or US$24.20

244

Table 15. Average irrigation water use and savings of dry-seeded rice (DSR) compared with farmers’ practice at three sites in 2009.

Project site

Irrigation water input (mm)

% water savings of DSR over farmers' practice

Farmers’ practice (puddled, transplanted, and flooded)

Dry-seeded rice

Sadhoke (n=52) 2,160a 1,450c 33 Chunian (n=43) 1,730b 1,300d 25 Chiniot (n=40) 1,680b 1,470c 12 Mean 1,850 1,410 25 SE 1,390.78 0.11 LSD0.05 84.54 75.5 CV(%) 5.21 6.02

245

Table 16. Average productive tillers of dry-seeded rice (DSR) compared with farmers’ practice at three sites in 2009.

Project site Productive tillers m−2 % increase of DSR over

farmers’ practice Farmers’ practice (puddled, transplanted,

and flooded)

Dry-seeded rice

Sadhoke (n=52) 287b 452a 57 Chunian (n=43) 235c 397a 69 Chiniot (n=40) 310b 388a 25 Mean 277 412 49 SE 1.112 1.334 LSD0.05 2.389 2.617 CV(%) 5.31 5.28

values with similar letter(s) are not significantly different at p<0.05.

246

Table 17. Average grain yield of dry-seeded rice (DSR) compared with farmers’ practice at three sites in 2009.

Project site Grain yield (t ha−1) % increase of DSR over farmers’ practice Farmers’ practice

(puddled, transplanted and flooded)

Dry-seeded rice

Sadhoke (n=52) 4.0b 6.0a 48 Chunian (n=43) 3.5b 4.6b 31 Chiniot (n=40) 5.1a 5.5a 8 Mean 4.2 5.4 27 SE 0.001 0.018 LSD0.05 0.7 0.3 CV(%) 3.39 6.52

values with similar letter(s) are not significantly different at p<0.05.

247

Table 18. Effect of aerobic and AWD technology on rice crop and yield in two districts in 2010.

Technology Hafizabad Mandi Bahuddin

Effective tillers m−2

Grain yield (t ha−1)

Water use (acre-inches)

Effective tillers m−2

Grain yield (t ha−1)

Water use (acre-inches)

Conventional practice 219.72 3.74 78 268.32 3.62 71 Dry direct seeding 352.76 4.68 57 380.39 4.29 60 % increase over conventional

60.5 25.1 26.4 41.8 18.5 15.5

% increase is a decrease for water use, which is an improvement.

248

Figure 1. Rice-growing zones of Pakistan.

ZoneZone--IIIIII

ZoneZone--IVIV

ZoneZone--IIII

ZoneZone--IIRICE IN PAKISTAN

249

Paper 10 Suitable management strategies for the aerobic and alternate wetting and drying systems of rice cultivation in Odisha A. Ghosh, S.K. Pradhan, O.N. Singh, and P. Samal

Indian water resources are becoming depleted faster and the country needs to promote water-saving rice production technologies to ensure rice food security and cope with the burgeoning population as the major contribution of around 70% to total rice production in India comes from the irrigated ecosystem. In an Asian Development Bank-supported project at the Central Rice Research Institute (CRRI) from 2007 to 2009, two promising water-saving rice production technologies, aerobic rice and the alternate wetting and drying system (AWD), were evaluated. Under aerobic conditions, studies on rice genotypes vis-à-vis irrigation management estimated an overall savings of around 25% of irrigation water across the year, accounting for higher water-use efficiency (2.75 to 5.0 kg grain ha−1 mm−1 water applied) than that (2.50 to 4.25 kg grain ha−1 mm−1) under conventional management. In the AWD system, overall savings of irrigation water was around 33%, accounting for higher water-use efficiency (3.4 to 3.5 kg grain ha−1 mm−1 water applied) than that (2.6 to 2.7 kg grain ha−1 mm−1) under conventional management. However, mean grain yield in aerobic rice (2.4 to 4.2 t ha−1) and in AWD (4.9 to 5.0 t ha−1) was lower by 20% and 12%, respectively, compared with that in conventional management. Taking into account the severe weed pressure in aerobic cultivation, studies showed greater weed competitiveness in rice genotypes CR Dhan 200, IR74371-3-1-1, and Anjali, producing 10% more grain yield than other genotypes. Results also indicated higher benefit with the application of preemergence herbicide followed by one manual weeding at the 3-week stage. In view of the yield penalty in the aerobic rice-based cropping system, non-rice crops such as groundnut or mung bean grown sequentially in the dry season enhanced the grain yield (4.3 to 4.6 t ha−1) of aerobic rice. Therefore, three years’ results of the project have shown potential for adopting aerobic rice and AWD water-saving technologies in the country.

Occupying the largest rice area, India contributes 26% to global rice production, following only China (33%). In India, average rice productivity is around 2.2 t ha−1. Rice contributes 43% of total food grain production and 53% of total cereal production in the country. The yield gap between potential and actual productivity varies widely across ecosystems over states. This is estimated to be 20% to 30% in irrigated rice, 30% to 50% in shallow lowlands, and beyond 50% in uplands and other unfavorable ecosystems. India requires 3% annual growth in rice productivity to achieve a production target of 122 million tons of rice by 2030. As a result, the country has to add at least 2.5 million tons of milled rice annually to sustain its current level of self-sufficiency. With the strong impact of climate change, the future increasing demand for rice production has to be met with less water, labor, and land (Table 1). Achieving this targeted rice production with reduced available resources and increased input costs poses a challenge to rice farmers, researchers, and planners. Thus, efficient crop, soil, and water

250

management strategies are needed to reduce the unit cost of production, which will help maintain higher profits for farmers and keep prices affordable to consumers.

Four major rice ecosystems prevail in India: irrigated, rainfed upland, rainfed lowland, and flood-prone deepwater and coastal environment. In India, irrigated rice constitutes the majority of rice area (20.5 million ha) that accounts for 46% of the total rice area, with a 70% contribution to total rice production in the country (Figure 1). Irrigated rice is grown in the states of Punjab, Haryana, Uttar Pradesh, Jammu and Kashmir, Andhra Pradesh, Tamil Nadu, Sikkim, Karnataka, Himachal Pradesh, and Gujarat. Rainfed upland rice (6 million ha) occupies 14% of the total area in the country. These areas lies in the eastern zone comprising Assam, Jharkhand, Chhattisgarh, Madhya Pradesh, western Odisha, eastern Uttar Pradesh, West Bengal, and the Northeastern Hill region. The rainfed lowland rice ecosystem (13 million ha) occupies 29% of the total area under rice. The flood-prone deepwater and coastal rice ecosystem (5 million ha) occupies around 11% of the area (Pandey and Ghosh 2010). Rainfed rice cultivation is highly unpredictable as this ecosystem represents a complex, diverse, and risk-prone growing environment. After 1960, the Green Revolution boosted rice production, but also further increased its dependence on input-oriented management practices in which the role of water appears to be the most important for the successful implementation of these practices. Irrigated rice areas in India, based on water availability and irrigation sources, can be classified as follows: i. River-irrigated Traditional rice cultivation is based on river water, which can be divided into rivers with perennial water flow and seasonal water flow. Success of the rice crop largely depends on the quantum and period of water available in the rivers or reservoirs on account of seasonal rainfall. Sometimes, to save standing crops, intermittent irrigation is practiced to alleviate midseason drought. This practice is mostly followed in Karnataka and Tamil Nadu. ii. Tank-irrigated In these areas, rice cultivation is very common if sufficient water is available in the tank, built across slopes to harvest rainfall. Started as a “dry crop” with direct seeding, rice grows as a rainfed crop. This typical “semi-dry” rice cultivation is popular in flood-prone areas of Assam, Bihar, and Uttar Pradesh, coastal regions in West Bengal, Odisha, and Andhra Pradesh. iii. Groundwater-irrigated Here, farmers depend solely on available groundwater for rice cultivation. These are typical bore-well-/dug-well-based irrigated rice ecosystems, mostly prevalent under boro rice in West Bengal. Rice in Punjab and Haryana is also largely grown exploiting groundwater resources.

Apart from this ecosystem-based classification, the crop establishment method prevailing in rice-growing areas could be another criterion, which is largely governed by three prime resource factors, land topography, labor, and water availability. Accordingly, rice cultivation could be classified into

251

i. Dry-seeded rice (DSR) This system is followed primarily under rainfed upland and lowland conditions. In rainfed rice regions in India, DSR in different forms is practiced during the wet season, when dry seed is broadcast, drilled, or dibble-sown in dry soil, preferably before the onset of premonsoon rains, and seed germination and emergence are facilitated by premonsoon showers. Subsequent crop growth continues with the seasonal rains on account of southwest monsoon. ii. Wet-seeded rice (WSR) WSR is the practice of sowing sprouted seeds under puddled conditions as a direct-sown crop. The WSR system is adopted as direct wet-seeded rice in irrigated conditions or even under shallow favorable lowland conditions in which water is available for field preparation and initial crop establishment. iii. Transplanted rice (TPR) This practice of rice cultivation is widely practiced in India, mostly in the irrigated ecosystem. TPR is also adopted in shallow favorable lowlands, where field inundation starts in the last week of June to first week of July and the field remains flooded for the major part of the crop’s growing period. Water shortage as a problem The quantum of water used for irrigation in the past century was 428 km3, which comprised 300 km3 of surface water and 128 km3 of groundwater (Kumar et al 2005). Estimates indicate that the water requirement for irrigation by 2025 will be around 561 km3 for the low-demand scenario and 611 km3 for the high-demand scenario (Kumar et al 2005). On the other hand, demand for water in various nonagricultural sectors in the country reduces its availability for agriculture; as a result, the agricultural sector is likely to suffer from a 10% to 15% reduction in irrigation water by 2025 (Ghosh et al 2009). The worst scenario would pertain to the rice system as it alone consumes more than 50% of the total amount of water available for the agricultural sector. Thus, the looming water crisis necessitates a strategic development of water-saving rice production systems, lest the country become a net importer of rice by 2020 (Narasimhan 2008).

The International Water Management Institute, in its survey report, projected the average consumption of water for agriculture during 2015-50 to be around 80% and emphasized that the requirement of growing water-intensive crops such as wheat, rice, and sugarcane would be around 275 trillion liters of water by 2025 (Anonymous 2009). Compared with that of China, India’s agricultural sector needs water consumption more than the current reserves, projected to be 56% compounded annual growth rate (CAGR). On the flip side of these data, it is noticed that Indian water footprints will be depleting sooner.

Factors such as rising labor costs and inadequate labor during the peak period, dwindling water resources, etc., enforce the paradigm shift to alternate crop establishment methods in place of transplanted rice presently practiced on more than 50% of the rice area in India. Thus, the need for less labor, water, and land-demanding rice cultivation practices could transform inundated transplanted rice cultivation into dry direct or wet direct-seeded rice. Direct seeding, unlike

252

transplanting, can reduce the labor requirement by up to 50% (Santhi et al 1998). Farmers may, however, end up using most of the labor saved from transplanting for weed management operations. As a result, even though the total labor requirement shows no substantial reduction, the demand for labor spreads over a longer period of time than with transplanting. This enables farmers to make full use of labor, thus efficiently avoiding bottlenecks in overall cultural practices (Pandey and Velasco 1999). Apart from saving labor and water, nonflooded rice cultivation may bring other advantages such as reducing methane emissions, maintaining soil structure beneficial to nonrice crops in the rotation, and promptly switching over to a subsequent crop season by reducing the turnaround time for land preparation. The practice of both puddling and nonpuddling operations was reported to be equally effective for crop growth, producing comparable rice yield (Bajpai and Tripathi 2000). In addition, in a rice-wheat system prevalent in the Gangetic alluvium region in Uttar Pradesh and West Bengal, zero tillage under nonpuddling conditions was reported to significantly support higher wheat yields than that followed by puddled rice cultivation. Simultaneous changes that have stimulated the switch include the development of modern rice varieties with early vegetative vigor (EVV) and high tillering ability that increase the crop’s ability to compete with weeds more effectively under direct-seeded conditions (De Datta and Nantasomsaran 1991). Available Water-Saving Technology Water-saving rice production technologies such as aerobic rice systems, alternate wetting and drying (AWD), saturated soil culture (SSC), bed planting, the system of rice intensification (SRI), etc., have been developed to minimize the irrigation water requirement of crops so that water-use efficiency can be enhanced without compromising productivity much. However, the prospects for their adoption are not very impressive, probably because of the shortfall of the technology per se, lack of awareness of farmers, inadequate dissemination mechanisms, or poor extension machinery. Aerobic rice It is evident that rice grown in a submergence regime has the lowest water-use efficiency. Understanding of this mechanism of saving water has explored the idea of growing rice under aerobic conditions, that is, with oxygen in the soil throughout the growing season compared with traditional irrigated rice, which is “anaerobic,” requiring more water (Bouman et al 2007). In this aerobic system, high-input responsive varieties having drought tolerance of upland types with high-yielding characteristics of lowland varieties are direct-sown in a well-drained, nonpuddled, and nonsaturated soil (Bouman et al 2005). This aerobic rice can be grown in either rainfed or irrigated conditions, where the amount of irrigation water should preferably match with the evapotranspiration of the crop. The fundamental approach for reducing the water requirement in aerobic rice is to balance water outflows (evapotranspiration, seepage, percolation, and runoff) with water inflows (irrigation and rainfall). This new way of growing rice started as

253

early as the mid-1980s in China. Now, it can be considered a mature technology in temperate regions in northern China and Brazil; but, in the Asian tropics, it is still very much in the research and development phase (Bouman et al 2007, Xue et al 2008).

Alternate wetting and drying system (AWD) Another water-saving technology is growing rice in the AWD system of irrigation management. Here, a predetermined wetting and drying phase is imposed in a cyclic manner during crop growth, thus maintaining adequate soil moisture regime in transplanted rice (Ghosh et al 2010). It is location-specific, governed by the nature of soil, meteorological parameters, and groundwater contribution. In the Philippines, a 10% to 40% yield decline was reported when moisture tension was allowed to reach 0.1 to 0.3 bar at 10- to 20-cm soil depth. Peng and Bouman (2007), while reviewing studies on this issue, reported higher yield even beyond 0.1 bar tension. Therefore, the schedule of irrigation in the AWD system needs to be standardized, taking into account the fluctuation of the groundwater table. It implies a trade-off between crop and water productivity in optimizing water-use efficiency as well as sustainable grain yield.

Saturated soil culture and bed planting This practice implies maintaining the soil water regime as close to saturation as possible. The hydraulic head of the ponded water decreases seepage and percolation losses. In this system, shallow irrigation is provided, obtaining 1-cm ponded water depth a day or so after the disappearance of standing water. Thus, frequent shallow irrigation along with good water control at the field level is the prerequisite of this practice. This system is most efficient in saving water by 5% to 50%, with an average of 23%; however, yield is likely to decline by 0 to 12%, with an average of 6%. Several other studies reported its merit in economizing the irrigation water requirement efficiently (Bouman et al 2006). In southern New South Wales, Australia, this practice could reduce both irrigation water and yield by around 10%.I In Australia, planting on raised beds was explored to facilitate SSC practice. Beds with a dimension of 120 cm separated by furrows of 30-cm width and 15-cm depth remained saturated with a continuous flow of water within the furrows; thus, 34% water savings compared with that of flooded rice was possible, which was at the cost of a 16% to 34% yield decline. The prospects of raised-bed planting curtailing the considerable amount of water are yet to be confirmed because the water use for land preparation/puddling and initial crop establishment seems to be high in this practice. System of rice intensification (SRI) This is another improved management practice in which, despite applying irrigation in the AWD system, yield does not decline; most often, it may increase. However, SRI is a labor-intensive technology with different modifications

254

suggested by different researchers to manage the rice crop. This technique is also under large-scale adoption as a promising resource conservation technology but, when up-scaling this technology, more emphasis needs to be given to other management parameters rather than water savings per se.

Aerobic rice is a very important technology for saving water; however, reports of yield decline and decreased availability of nutrients to plants under aerobic conditions seem to be the major concern when up-scaling this technology widely. In rice belts in parts of northern and eastern India, the success of this technology in farmers’ fields was validated in the Asian Development Bank (ADB)-supported project “Developing and disseminating water-saving technologies in South Asia.” Farmers have shown keen interest in adopting this technology. AWD is a well-validated water-saving technology for transplanted rice. However, introducing the concept of safe AWD where there are no threats of yield decline is very important. Materials and methods Field studies were conducted at the Central Rice Research Institute, India, during 2007, 2008, and 2009, successively, to examine the sustainability of water-saving rice technologies under the ADB project. The experimental site was located at a geographic position of 20o 30′N and 86oE, and at an altitude of 22 m above mean sea level, where the groundwater table remained within 100 cm throughout the cropping season. The soil was an Aeric Haplaquept with 8−10% sand, 32−35% silt, 60−65% clay,0.83% organic C, 0.09% total Kjeldahl N, 12 mg kg−1 soil available phosphorus, and 65 mg kg−1 soil available potash with pH of 6.8. The texture of the soil was clay loam with bulk density of 1.3−1.6 g cc−1 and field capacity was 2.8 mm cm−1 depth of soil, estimating 45 mm of water-holding capacity at root-zone (30-cm) depth.

Mean meteorological parameters averaged over the study period, 2006 to 2009, recorded a range of 1,572 to 1,950 mm of annual rainfall, of which only 29 to 105 mm of rainfall were received during the dry season (actual cropping season), which was 1.5−5.75% of the total annual rainfall. Maximum and minimum temperature varied within 27.8−37.2 oC and 14.6−26.3 oC, and evaporation was 3.0−7.4 mm day−1. The groundwater table (GWT) during the cropping season remained within 30 mm to 100 cm (Figure 2). It was 30−35 cm at the beginning of the season, with a gradual depletion toward the advancement of the season as rainfall during the dry season is of rare occurrence, and happens sporadically, once in a blue moon. In the first 30 days after sowing, GWT remained at 55 cm, and then fall to around 80 cm in the next 30 days; thereafter, it declined sharply. Four types of experiments were carried out, as mentioned below. Experiment 1: Aerobic rice system In the field experiments conducted under an aerobic soil moisture regime, the water requirement of promising medium-duration (100 to 120 days), intermediate-height rice genotypes CR Dhan 200, Annada, Shatabdi, Anjali,

255

IET18665,RR 3-88-5,IR78877-181-B-1-2, IR78875-53-2-2-2, IR55419-04, IR74371-3-1-1, and IR74963-262-5-1-3-3was estimated during 2007, 2008, and 2009 successively. Subsequently, percent water savings in aerobic rice was determined in comparison with that in saturated soil conditions under conventional irrigated conditions. Dibble-seeding was carried out in the first week of January in dry soil at a spacing of 20 × 15 cm. The sowing was done in 5 × 4-m plots using a seed rate of 80 kg ha−1. Fertilizers were applied at the recommended dose of 80 kg N, 17.5 kg P, and 33.2 kg K ha−1. The entire dose of P and K and 50% of N were applied at the time of sowing. The remaining N was applied in two equal splits at 45 and 60 days after sowing. The experiment was laid out in a split-plot design with irrigation in the main plot and genotypes in the subplots with five replicates. The subsurface water level was monitored using piezometers installed down to a depth of 125 cm at the center of each plot, which recorded daily fluctuations of the subsurface water level and helped in interpreting variability in soil water status, particularly in the root zone. To maintain aerobic soil conditions, irrigation was applied when the surface soil showed hairy cracks (equivalent to 40 kPa soil moisture potential), while saturated soil conditions were maintained with adequate irrigation. Water was applied in every treatment plot using a polythene pipe attached to the pump delivering water from a nearby bore well. The amount of water applied in each treatment plot was measured with a water flow meter connected to the tail end of the pipe measuring water flow in liter units (1 ha mm ~ 10,000 L). This instrument quantified the water flow rate, which was 3.0 ha mm h−1. The total quantity of water applied per irrigation was also recorded in this water meter. To restrict the passage of water from the adjacent plots, a set of double drains of 50-cm depth was dug in between all treatment plots. Those drains were connected to a 75-cm-deep drain around the periphery of the site, wherein seepage water was accumulated and then discharged outside the experimental sites. In addition, thin-gauge metal sheets were inserted down to 40 cm deep in the bunds of all plots. Results and discussion The water requirement, averaged over the year, was higher (1,240 ha mm), ranging from 1,079 to 1,400 ha mm, maintaining soil saturation in the conventional system compared with that of 855 ha mm in aerobic conditions, with a range of 830−880 ha mm across years (Table 2). To apply this much water, 15 irrigations were required in the conventional system compared with 10 irrigations in the aerobic system. This resulted in a 25.15% overall savings of irrigation water, with a range of 23.7−26.6%, accounting for higher mean water-use efficiency (3.84 kg grain ha−1mm−1 water applied), ranging from 2.73 to 4.95 kg grain ha−1 mm−1 water applied in aerobic conditions compared with 3.37 kg grain ha−1 mm−1 water applied in soil saturation conditions.

Under aerobic conditions, plant height ranged from 87 to 94 cm, with an average of 90 cm; whereas, under saturated conditions, average plant height was 95 cm, with a range of 92−98 cm. Panicle number also varied according to soil conditions, with a higher panicle number (255) under saturated conditions, ranging from 225 to 275 cm, with maximum panicles in IR74371-3-1-1. In aerobic

256

conditions, panicle number declined to 240, ranging from 215 to 270, with maximum panicles in CR Dhan 200. However, panicle length and weight and 1,000-grain weight did not vary significantly across soil conditions, while it varied among cultivars, with higher values recorded in CR Dhan 200, followed by IR74371-3-1-1.Grain yield declined in all genotypes grown under aerobic conditions compared with yield in the conventional system. Across years, crops grown under soil saturation produced significantly higher grain yield of 3.54−4.71 t ha−1; whereas, under aerobic conditions, a mean yield decline of 14% was observed (Table 2). Evaluation of genotypes across years showed RR 3-88-5 and IET18665 in 2007, CR Dhan 200 (Pyari) and IR78877-181-B-1-2 in 2008, and CR Dhan 200 and IR74371-3-1-1 in 2009 as promising genotypes under aerobic conditions (Figure 3). This phenomenon implied a significant correlation of deficit moisture stress with a decline in grain yield and yield parameters (Choudhury and Singh 2007). Yield decline at increasing water stress could be attributed to a complex phenomenon in which anatomical, morpho-physiological, and biophysical factors are responsible either in combination or alone (Bouman et al 2005). Contrary to that, those detrimental effects were not observed in saturated conditions.

Averaged over the total irrigation application during the growing period, less requirement of irrigation water was recorded at the initial stage and also toward crop maturity. This resulted in variability in water application efficiency, and this variability was also governed by other climatic parameters such as rainfall. The practical implications of the study are to encourage farmers to grow an appropriate variety of aerobic rice when they encounter water scarcity. This approach will not only help them harvest more crop per drop of water, but will also enable them to bring more area under irrigated rice cultivation, thus ensuring higher total rice production. However, keeping in view the prospects of aerobic rice in water-scarce areas, a future study should emphasize how to sustain higher aerobic rice productivity, identifying and alleviating the causative factors responsible for its yield decline (Bouman et al 2005). Other research should examine nutrient management, especially Zn and Fe availability, and weed management to make the technology more viable and acceptable to farmers. Experiment 2: Rice in the AWD system The quantum of water savings while scheduling irrigation under the AWD system was compared with the conventional system in field experiments conducted during 2007, 2008, and the 2009DS, successively. The experiment studied the yield performance of medium-maturity-duration (120 to 130 days) genotypesIR64, Naveen, and Lalat in the first year; IR64, Lalat, and Pusa 44 in the second year; and IR64, Naveen, Lalat, and Chandan in the third year under two conditions of irrigation management, the AWD system of irrigation and the conventional method of maintaining 5 cm of standing water throughout the growth period.

The nursery was raised in mid-December and transplanting was done with 25-day-old seedlings at 15 cm plant-to-plant spacing during the first week of

257

January every year in plots measuring 20 m2. The recommended dose of N, P, and K fertilizer was applied at 120 kg N, 26.2 kg P, and 33.3 kg K ha−1, of which 50% N, 100% P, and 100% K were applied during planting and the remaining N was applied in two equal splits, at tillering and panicle initiation stage. The experiment was laid out in a split-plot design with irrigation level in the main plot and genotypes in subplots with five replications. The subsurface water level was monitored using piezometers installed down to a depth of 125 cm at the center of each plot, which recorded daily fluctuations of the subsurface water level and helped in interpreting variability in soil water status, particularly in the root zone. In the AWD system, irrigation water was applied when the subsurface water level went below 30-cm depth, which was monitored with the piezometer installed down to root-zone depth of 30 cm. On the other hand, 5 cm of ponded water in the conventionally irrigated plots were maintained with adequate irrigation application. Water was applied in every treatment plot using a polythene pipe attached to the pump delivering water from a nearby bore well. The amount of water applied in each treatment plot was measured with a water flow meter connected to the tail end of the pipe measuring water flow in liter units (1 ha mm ~ 10,000 lit). This instrument quantified water flow rate, which was 3.0 ha mm h−1. The total quantity of water applied per irrigation was also recorded with the water meter. To hydrologically separate the plots for restricting the passage of water from the adjacent plots, a set of double drains of 50-cm depth was dug in between all treatment plots. Those drains were connected to a 75-cm-deep drain around the periphery of the site, wherein seepage water was accumulated and then discharged outside the experimental sites. In addition, thin-gauge metal sheets were inserted down to 40 cm deep in the bunds of all plots. Results and discussion Averaged over years, the water requirement was higher (2,126 ha mm), ranging from 2,047 to 2,175 ha mm, maintaining conventional submergence with 5 cm of standing water throughout the growth period compared with 1,432 ha mm in the AWD system, with a range of 1,356−1,475 ha mm across years (Table 3). To apply this much water, 26 irrigations were required in the conventional system compared with 18 irrigations in the AWD system. This resulted in an overall 32.8% savings of irrigation water, with a range of 32−33.9% across years in the AWD system, accounting for higher mean water-use efficiency (3.43 kg grain ha−1 mm−1 water applied), ranging from 3.35 to 3.55 kg grain ha−1 mm−1water applied compared with 2.64 kg grain ha−1 mm−1 water applied in the conventional system of irrigation management.

In AWD conditions, plant height ranged from 86 to 98 cm, with an average of 95 cm; whereas, under normal irrigated conditions, mean plant height was 98 cm and ranged from 95 to 100 cm. Panicle number also varied under the two conditions: panicle number was higher (410) under normal irrigated conditions, with a range of 375 to 430. In AWD conditions, panicle number declined to 350, with a range of 340 to 410. However, panicle length and weight and 1,000-grain weight did not vary significantly across soil conditions, while it varied among

258

cultivars, with higher values for Naveen, followed by Lalat. Grain yield declined in all genotypes grown under the AWD system compared with the conventional system. This seems to be due to the deficit soil moisture beyond a certain range that impaired crop growth. On the other hand, the higher water-use efficiency in the AWD system was attributed to the fact that the magnitude of yield decline in this moisture regime was comparatively lower than the extent of water saved.

Across years, crops grown under the conventional system produced significantly higher grain yield of 5.55−5.64 t ha−1; whereas, under AWD conditions, a 12.5% mean yield decline compared with that of the conventional saturated system was observed (Table 3). Evaluation of genotypes across years showed significantly higher grain yield in Lalat in 2007, Pusa 44 in 2008, and Naveen in 2009 under the AWD irrigation management (Figure 4). It was also noticed that the grain yield of Pusa 44 under the AWD system was comparable with that in conventional irrigation, which was attributed to its longer duration. The field study revealed that rice does not require continuous submergence throughout its growth period, which corroborated an earlier report (Belder et al 2005). Concurrently, our results indicated prudent use of water in rice cultivation, which is of paramount importance for expanding rice cultivation. Results showed that a system of AWD cycle of irrigation management affected the water-use pattern, which ensured a varying amount of water savings depending on the soil moisture tension at which it was applied. Nonetheless, the frequency of irrigation application was also determined, usually at 5- to 6-day intervals at the beginning of the growing season, which decreased to 3- to 4-day intervals upon advancement of soil dryness at a later stage (Ghosh et al 2010).

Thus, the study could advocate applying irrigation water following the AWD system to save water in water-scarce regions. However, future studies may converge in reducing the magnitude of yield decline by suitably adjusting the depth of the subsurface water level at which irrigation could be applied without compromising grain yield, that is, “safe AWD.” In that situation, water savings may not be to the extent as recorded in our study, but farmers would be satisfied achieving optimum water savings without any yield loss. Experiment 3: Sustainable crop rotation Aerobic rice is an up-and-coming vibrant technology combating the water crisis with 30−40% water savings. Although growing rice under aerobic conditions could enable farmers to sustain rice cultivation with less water, in reality, this system seems to be still immature, in a nascent stage of development. Aerobic rice, when grown continuously, causes a yield penalty in the short run to the extent of 8−70%, and even a total yield collapse in the long run (Xiao et al 2010).The actual reason is not yet clear; it is likely to be related to altered plant growth due to root-shoot signaling, impaired ability to acquire nutrients, or a disruption of soil-water-plant equilibrium, in a deviation from normal. This is believed to be governed by potentially interwoven effects of both biotic and abiotic factors that could be called “auto-toxicity.” Nonetheless, prolonged aerobic soil steadily decreases soil health over the years, resulting in a

259

permanent “soil sickness,” leading to yield collapse, the major crux of the problem. This means striving for the development of a sound alleviating mechanism. Thus, some refinement and fine tuning are essential in the existing management of aerobic rice to ensure its wider dissemination for greater adoption by the rice farming community that faces a water crisis.

Long-term rice-based cropping system research began in 2007 with the objective of determining the extent of aerobic rice yield decline over the years, alleviating soil sickness, and developing a suitable crop rotation for sustainable aerobic rice yield. Three crop rotations, rice (wet season; WS) - groundnut (dry season; DS), rice (WS) - mungbean (DS), and rice (WS) - maize (DS), were studied during 2008; whereas, another crop, cowpea (DS), was included in the rotation in 2009. The performance of rice in those rotations was compared with that of the aerobic rice (WS) - aerobic rice (DS) rotation. Since 2009-10, the program has been carried out more intensively, introducing other rotations such as aerobic rice (WS) - nonrice/aerobic rice (DS) in a1- or 2-year cropping sequence. In the first year of the study, a uniformity trial was carried out by sowing medium-maturity-duration rice variety CR Dhan 200 during the first week of June in 2007 as the first crop in the rotation. Results and discussion A root study was carried out to determine root health as well as the formation of nodules in the root as an indication of nematode infestation (Table 4). Development of the rhizosphere and rhizoplate was determined by measuring root volume and root fresh and dry biomass. The root volume of rice was higher when grown in sequence with nonrice crops (34 to 40 cc plant−1) compared with that (33 to 35 cc plant−1) in the rice-rice rotation. Fresh and dry biomass of rice root were also higher (21−28 and 7−12 g plant−1, respectively) compared with those (19−20 and 6−7 g plant−1) in the rice - rice rotation. Gall formation in the rice root was higher (25−30 galls plant−1) than in the rice with mung bean and cowpea rotation (20−22), except in the initial year, 2007. Aerobic rice, during the wet season in 2007, produced 4.17 t ha−1 grain yield, which remained more or less consistent across years until the wet season of 2009 (Table 5). This implied no yield decline up to the third year of growing aerobic rice consistently. On the other hand, there was no significant yield decline also in aerobic rice grown during the dry season in 2008 and 2009, with 4.4 and 4.35 t ha−1, respectively. Moreover, analyzing grain yield over the years vis-à-vis across seasons showed no significant decline, with 4.17 t ha−1 in the 2007WS and 2009WS. On the other hand, nonrice crops grown sequentially in the DS enhanced grain yield in rice grown in rotation during the WS, which was more pronounced when growing groundnut and mung bean in sequence with rice. Thus, rice in the groundnut-rice rotation produced significantly higher grain yield in both 2008 (4.59 t ha−1) and 2009 (4.63 t ha−1). Similarly, the grain yield of rice in the rice - mung bean rotation was 4.34 and 4.31 t ha−1, respectively, in 2008 and 2009.

Probably, this implies that these rotations could emerge as potential alleviating mechanisms for sustainable aerobic rice production. However, future

260

studies should be carried out examining the management of soil, tillage, and crop residue, developing an improved irrigation schedule based on critical crop growth stages, etc. Experiment 4: Weed management strategies in aerobic rice Aerobic rice varieties have been developed with moderately high production potential under limited water conditions. Most often, notwithstanding their high yield potential, actual yields happen to be much lower because of improper management of weed pressure, one of the major impediments in aerobic rice cultivation (Singh et al 2008). Unlike other rice ecosystems, in an aerobic environment, a transitional soil water regime of dry and wet spells alternately creates a congenial micro-environment prompting the emergence and growth of highly competitive complex weed flora in different flushes. Weed pressure becomes severe because of the lack of a “head start” of rice seedlings over weeds (Singh et al 2008). This may call for a comprehensive measure taking into account the compatibility of both the weed-suppressive ability (WSA) of rice genotypes and the efficacy of chemical and/or manual methods of weed control.

In field studies conducted in 2007, 2008, and 2009WSsuccessively, the weed competitiveness (WC) of five medium-duration (100 to 110 days), photosensitive, medium- stature (80 to 100 cm) aerobic rice genotypes, Anjali, CR Dhan 200,Satabdi, IR74371-3-1-1, and RR 3-88-5, was studied with different weed management practices constituting manual and/or chemical methods. The stages for applying different weed management were decided with the principle of keeping weed pressure as minimum as possible during the first 30 to 35 days, the most critical stage of crop growth for medium-duration rice genotypes. Results and discussion A mixed weed flora comprising 70−75% grassy weeds at the beginning of the season and sedges and broadleaf weeds in the later stages of crop growth was observed in the experiments. The predominant grassy weeds were Echinochloacolona, E.crus-galli, Dactyloctenium aegyptium, and Eleusine indica; sedges were Cyperusiria, C.rotundus, Fimbristylismiliaceae, Amaranthus spinosus, Alternanthera sessilis;and broadleaf weeds were Ecliptaalba, E. prostrata, Commelinabenghalensis,Portulacaoleracea, and Euphorbia hirta.

Overall results showed greater WSA of CR Dhan 200 and IR74371-3-1-1, eliminating weed competition more substantially than other genotypes. Compared with the weedy check across years, they ensured low weed density (55−70%) with less fresh weed (14−15%) and dry weed (34−36%) biomass (Table 6). Lower weed pressure was attributed to the WSA of the genotypes on account of faster initial crop growth ensuring higher early vegetative vigor (EVV) that enabled the genotypes to suppress the first flush of weeds at the initial stage, while the second flush of weeds could not compete with crop growth and remained smothered under the crop canopy, supplemented with effective management (Singh et al 2008).

261

Weed control efficiency (WCE) and weed index The WCE of different treatments varied significantly on account of the WC of genotypes (26−36%) as well as efficacy of weed management (20−50%) across years. Higher WCE was achieved in integrated weed management (IWM) (48.7%), followed by manual weeding twice in the second and fourth week of crop growth (40.6%), or the treatment comprising preemergence herbicide application along with manual weeding in the third week of crop growth (37.0%). The lowest weed index (11.5%) occurred with IWM, indicating a relatively higher yield advantage, followed by manual weeding twice in the second and fourth week of crop growth (14.2%) and preemergence herbicide application along with manual weeding in the third week of crop growth (18.8%). Grain yield and yield parameters Results showed higher plant height of 120 cm with IWM, comparable with the treatment of preemergence herbicide along with manual weeding once in the third week of crop growth. Similarly, higher panicle numbers (264) and weight (3.20 g) were also recorded in the IWM treatment. Accordingly, the higher performance of genotype CR Dhan 2000 was noticed, with significantly higher grain yield, closely followed by genotypes IR74371-3-1-1 and Anjali, with 8−10% more yield advantage than other genotypes (Table 6). The WSA of CR Dhan 200 was attributed to its better EVV, implying faster initial growth that smothered the weed population (Zhao et al 2007).

Regarding weed management, manual weeding twice in the second and fourth week of crop growth produced significantly higher grain yield, comparable with that of crops treated with IWM and preemergence herbicide coupled with manual weeding once in the third week, accounting for an 85−90% yield advantage compared to the weedy check. When comparing their relative efficacy with that of weedy and weed-free conditions, crops treated with IWM or manual weeding alone recorded 13−16% less yield than crops in weed-free conditions. Crops treated with preemergence herbicide coupled with manual weeding showed a higher yield decline (40%). On the other hand, comparing their relative yield advantages with those of the weedy check showed 8−10% higher benefit with IWM or manual weeding alone over preemergence herbicide coupled with manual weeding once.

When deriving a relationship between dry weed biomass and grain yield, a negative correlation was noticed across all genotypes over the years (Figure 5). This suggests that the high yielding ability of the genotypes was associated with lower weed biomass, indicating that high yielding ability and strong WSA are compatible. Thus, this implied a need to minimize weed density for achieving optimal aerobic rice grain yield. Additionally, it indicated a31.29, 27.39, and 28.28 g m−2 decrease in biomass, essential for achieving every unit increase in grain yield in 2007, 2008, and 2009, respectively. Production economics Net return ($307−308 ha−1) was maximum in both manual weeding twice and application of preemergence herbicide along with manual weeding, even though the latter weed management treatment could bring more benefit as adjudged by

262

the higher benefit-cost ratio (1.64), attributed to lesser cost incurred for weed management. IWM, accounting for a $299.10 ha−1net return, closely followed, with a B:C ratio of 1.53 (Table 7). The study suggested application of preemergence herbicide along with manual weeding once in the third week of crop growth to be a promising agro-technique for weed suppression with optimum grain yield, cost effectiveness, and higher economic return in aerobic rice cultivation (Singh et al 2008). Conclusions The aerobic rice and alternate wetting and drying systems are promising water-saving rice production technologies that minimize the irrigation requirement of the crop so that water-use efficiency can be enhanced without compromising actual productivity. Our field studies reported around a 25% overall savings of irrigation water in aerobic rice, with grain yield of 2.4 to 4.2 t ha−1, accounting for higher water-use efficiency of 3.84 kg grain ha−1 mm−1 water applied in aerobic conditions as compared with 3.37 kg grain ha−1 mm−1 water applied in saturated soil conditions. The application of preemergence herbicide along with manual weeding once in the third week was found to suppress weed populations effectively under aerobic conditions. Rice under the AWD system (grain yield of 4.82 to 5.0 t ha−1) resulted in an overall 33% savings of irrigation water, accounting for higher water-use efficiency of 3.43 kg grain ha−1 mm−1 water applied as compared with 2.64 kg grain ha−1 mm−1 water applied in the conventional system of irrigation management. The study estimated 12−14% lower yield in the AWD system and 25−30% lower yield in aerobic rice, respectively, compared with that in the conventional irrigation system. Growing groundnut or green gram after an interval of 1 or 2 years was found to be a promising crop rotation practice for sustainable aerobic rice production. However, more studies are recommended to be carried out to better examine the management of soil, tillage, and crop residue, and to develop an improved irrigation schedule based on critical crop growth stages. References Anonymous. 2009. Water: the India story. Grail Research

LLC.www.ce.utexas.edu/prof/mckinney/ce397/Topics/Rice/Rice(2010).pdf. Bajpai RKK, Tripathi RP. 2000. Evaluation of non-puddling under shallow water

tables and alternative tillage methods on soil and crop parameters in a rice-wheat system in Uttar Pradesh. SoilTill. Res. 55:99-106.

Belder P, Bouman BAM, Spiertz JHJ, Peng S, Castañeda AR, Visperas RM. 2005. Crop performance, nitrogen and water use in flooded and aerobic rice. Plant Soil 273:161-182.

Bouman BAM, Humphreys E, Tuong TP, Barker R. 2006. Rice and water. Adv. Agron. 92:187-237.

Bouman BAM, Feng L, Tuong TP, Lu G, Wang H, FengY. 2007. Exploring options to grow rice using less water in northern China using a modeling

263

approach. II. Quantifying yield, water balance components, and water productivity. Agric.Water Manage. 88:23-33.

Bouman BAM, Peng S, Castañeda AR, Visperas RM. 2005.Yield and water use of irrigated tropical rice systems. Agric. Water Manage. 74:87-105.

Choudhury BU, Singh AK. 2007. Performance of rice (Oryza sativa) planted on raised bed under different soil moisture tensions. Indian J. Agron. 52(4):305-310.

De Datta SK, Nantasomsaran P.1991. Status and prospects of direct seeded flooded rice in tropical Asia. In: Direct seeded flooded rice in tropics. International Rice Research Conference, 27-31 August 1990, Seoul, Korea.Los Baños (Philippines): International Rice Research Institute. p 1-16.

Ghosh A, Rao KS, Singh ON, Dash RN, Samal P, Pandey MP, Kumar A, Zhao D. 2009. Conserving water in rice cultivation: its potential and prospects under looming crisis of water resources. In: 4th World Congress on Conservation Agriculture, New Delhi, 4-7 February. p 161-162.

Ghosh A, Singh ON, Rao KS. 2010. Improving irrigation management in dry season rice cultivation for optimum crop and water productivity at non traditional rice ecologies. Irrig. Drain. DOI-10.1002/ird.572.

Kumar R, Singh RD, Sharma KD. 2005. Water resources of India. Curr. Sci. 89(5):794-811.

Kumar P, Shinoj P, Raju SS, Kumar A, Msangi S.2010. Factor demand, output supply elasticity and supply projections for major crops of India. Agric. Econ. Res. Rev. 23(1):9.

Narasimhan TN. 2008. A note on India’s water budget and evapotranspiration. J. Earth Syst. Sci. 117(3):237-240.

Pandey S,Velasco L.1999. Economics of direct seeding in Asia: patterns of adaptation and research priorities. Int. Rice Res. Notes 24:6-11.

Pandey MP, Ghosh A. 2010. Rice production strategies for sustainable agriculture. In: Sengar RS, Sharma AK, editors. Stable food production and sustainable agriculture:a challenge ahead in 21stcentury. Studium Press (India) Pvt. Ltd. p 81-99.

Peng S, Bouman B. 2007. Prospects for genetic improvement to increase lowland rice yields with less water and nitrogen. In: Spiertz JHJ, Strunk PC, Van Laar HH, editors. Scale and complexity in plant systems research: gene-plant-crop relations. Springer. p 251-266.

Santhi P, Ponnuswamy K, Kempuchetty N.1998.A labor-saving technique in direct-sown transplanted rice. Int. Rice Res. Notes 23:25-36.

Singh S, Ladha JK, Gupta RK, Bhushan L, Rao AN. 2008. Weed management in

aerobic rice systems under varying establishment methods. Crop Prot. 27:660-671.

Xiao DQ, Zhang HY, Spiertz JHJ, Yu J, Xie GH, Bouman BAM. 2010. Crop response of aerobic rice and winter wheat to nitrogen, phosphorus and

264

potassium in a double cropping system. Nutr. Cycl. Agroecosyst. 86:301-315.

Xue C, Yang X, Bouman BAM, Deng W, Zhang Q, Yan W, Zhang T, Rouzi A,Wang H. 2008. Optimizing yield, water requirement, and water productivity of aerobic rice for the North China. Irrig. Sci. 26:459-474.

Zhao DL, Bastiaans L, Atlin GN, Spiertz JHJ. 2007. Interaction of genotype x management on vegetative growth and weed suppression of aerobic rice. Field Crops Res. 100:327-340.

Notes Authors’ address: Central Rice Research Insitute, Cuttack, Odisha, India. Figure 1. Percent area under rice in different ecosystems in India. Figure 2. Groundwater table during the crop-growing dry season (mean of 3 years). Figure3. Grain yield (t ha−1) of promising aerobic rice genotypes under variable

soil moisture conditions during 2007, 2008, and 2009 DS. Figure 4. Grain yield (t ha−1) of promising rice genotypes under AWD and

conventional irrigation management during 2007, 2008, and 2009 DS. Figure 5. Relationship between grain yield (t ha−1) of aerobic rice and dry weed

biomass (g m−2) across genotypes during 2007, 2008, and 2009.

265

Table 1. Demand-supply scenarios of rice production in India. Year Population

(million) Demand projection for rice (million t)

Supply projection for rice (million t) at production growth rate

Demand supply gap (million t) at production growth rate

1.34% 1.14% 1.34% 1.14% 2015 1,288 106.8 101.5 100.5 5.3 6.3 2020 1,370 112.8 108.5 106.4 4.3 6.4 2025 1,445 117.3 116.0 112.5 1.3 4.8 2030 1,523 121.6 123.9 119.1 −2.3 2.5 Source: Kumar et al (2010).

266

Table 2. Irrigation requirement of aerobic rice under variable soil moisture conditions during 2007, 2008, and 2009. Soil condition

Grain yield (t ha−1) Irrigation water requirement (ha mm)

Water saving (%) WUE, kg grain ha−1 mm−1 water applied

2007 2008 2009 2007 2008 2009 2007 2008 2009 2007 2008 2009 Aerobic soil conditions

2.40 (32%)*

3.96 (13.2%)

4.23 (10.31%)

880 830 855 26.6 23.7 24 2.73 4.77 4.95

Saturated soil conditions

3.54 4.56 4.71 1,400 1,079 1,125 2.53 4.22 4.04

* Yield decline.

267

Table 3. Irrigation requirement of rice in the AWD system under different levels of irrigation during 2007, 2008, and 2009 DS.

*Yield decline.

Soil condition

Grain yield (t ha−1) Irrigation water requirement (ha

mm)

Water savings (%) WUE (kg grain ha−1 mm−1 water applied)

2007 2008 2009 2007 2008 2009 2007 2008 2009 2007 2008 2009 AWD system

5.00 (11.35%)*

4.82 (13.15%)*

4.91 (13%)*

1,475 1,356 1,465 32.5 33.9 32 3.39 3.55 3.35

5 cm ponded water

5.64 5.55 5.64 2,175 2,047 2,155 − − − 2.59 2.71 2.62

268

Table 4. Root parameters of aerobic rice in dry season during 2007, 2008, and 2009.

Treatment/ crop rotation

Galls plant−1 Rice root volume (cc plant−1)

Rice root fresh biomass (g

plant−1)

Rice root dry biomass (g

plant−1) Year 200

7 2008 2

009

2007

2008 2009 2007

2008

2009

2007

2008

2009

Rice - rice 16 25 30

30.0 33.7 35.0 16.5 19.0 20.6 7.0 6.7 7.3

Rice - mungbean

− 21 22

− 37.5 38.7 − 23.0 23.8 − 9.0 8.5

Rice - maize − 19 25

− 36.2 35.5 − 22.2 20.8 − 8.0 7.7

Rice - groundnut

− 20 26

− 37.5 40.6 − 24.5 28.0 − 10.0 12.6

Rice - cowpea − 22 20

− 34.2 38.3 − 21.2 22.5 − 7.2 8.1

LSD (5%) − 2 2 − 3.67 3.7 − 4.02 4.2 − 1.2 1.2

269

Table 5. Grain yield of rice and nonrice crops in aerobic rice-based cropping system during 2007, 2008, and 2009.

Treatment/crop rotation

2007 2008 2009 WS* DS WS DS WS

Rice yield (t ha−1)

Rice/nonrice yield ( t ha−1)

Rice yield (t ha−1)

Rice/nonrice yield (t ha−1)

Rice yield (t ha−1)

Rice - rice 4.17 4.40 3.88 4.35 4.00

Rice - mungbean

− 0.60 4.34 0.65 4.31

Rice - maize − 5.64 3.97 4.53 3.88 Rice - groundnut

− 2.00 4.59 2.02 4.63

Rice - cowpea − − − 1.17 4.10 CD at 5% − − 0.65 − 0.40 * Uniformity trial with rice, variety Apo (CR Dhan 200).

270

Table 6. Influence of rice genotype and weed management on weed growth and grain yield of aerobic rice during 2007, 2008, and 2009.

Treatment Weed density (no.m−²)

Fresh weed biomass (g m−²)

Dry weed biomass (gm−²)

WCE (%) Grain yield (t ha−¹)

2007 2008 2009 2007 2008 2009 2007 2008 2009 2007 2008 2009 2007 2008 2009

Rice genotype CR Dhan 200 25.8 26.67 47.5 217.7 225.3 161.2 90.0 94.3 96.13 36.0 35.2 60.4 3.70 3.62 4.00

Shatabdi 32.6 − − 228.6 − − 100.0 − − 28.9 − − 3.36 − − Anjali 30.0 − − 225.0 − − 95.0 − − 32.5 − − 3.50 − − IR74371-3-1-1 28.0 24.22 − 220.5 230.0 − 92.0 96.6 − 34.6 33.6 − 3.65 3.63 −

RR3-88-5 34.3 − − − − − 102.8 − − 26.9 − − 3.35 − − Weed management

No weeding/ weedy check

70.5 85 111 260.5 270.2 230.2 140.7 145.5

141.1 − − − 2.00 1.75 1.58

Weed-free − − 22 − − 103.4 − − 40.41 − − 64.0 − − 4.37 Manual weeding once in third week

25.8 − − 212.8 − − 92.8 − − 34.0 − − 3.40 − −

Manual weeding twice in second and fourth weeks

20.5 25.7 32 205.4 210.0 158.9 83.6 85.4 105.8 40.6 41.3 58.0 4.50 3.67 3.75

Preemergence herbicide, butachlor at 2.5 L ha−1

35.5 28 − 235.7 220.0 − 105.5 90.6 − 25 37.7 − 3.25 3.31 −

271

Table 7. Production economics on weed management in aerobic rice during 2009. Treatment Cost of weed

management Cost of production*

Net return

B:C ratio

$ ha−1 T1: weedy check, 0.0 400 −50.70 0.88 T2: weed-free 370 770 200.10 1.26 T3:IWM 160 560 299.10 1.53 T4: weedicide + HW 85 485 205.40 1.42 T5: manual weeding

125 525 307.50 1.56

T4: Preemergence herbicide, butachlor, at 2.5L ha−1 + manual weeding once in third week; T5: manual weeding twice in second and fourth weeks. * Cost of crop management other than weed management: $400ha−1. MSP of rice: $222 t−1 (Rs. 10,000 t−1 at ~$45).

272

Figure 1. Percent area under rice in different ecosystems in India.

Irrigated

46%

Upland 14%

Lowland 29%

Deep water

9%

Coastal

2%

273

Figure 2. Groundwater table during the crop-growing dry season (mean of 3 years).

274

CD 5%

Period of study Factor 2007 2008 2009 Irrigation 0.15 0.22 0.06 Genotypes 0.27 0.48 0.10 Interaction 0.32 0.89 0.14

Figure 3. Grain yield (t ha−1) of promising aerobic rice genotypes under variable soil moisture conditions during2007, 2008, and 2009 DS.

275

CD 5%

Period of study Factor 2007 2008 2009 Irrigation 0.12 0.71 0.12 Genotypes 0.16 0.63 0.16 Interaction 0.23 0.90 0.23

Figure 4. Grain yield (t ha−1) of promising rice genotypes under AWD and conventional irrigation management during 2007, 2008, and 2009 DS.

11.5

22.5

33.5

44.5

55.5

66.5

IR 6

4

Nav

een

Lal

at

IR 6

4

Nav

een

Lal

at

IR 6

4

Lal

at

Pusa

44

Cha

ndan

IR 6

4

Nav

een

Lal

at

2007 2008 2009

Genotypes

Gra

in y

ield

t ha

-1Conventional irrigationAWD system

276

Figure 5. Relationship between grain yield (t ha−1) of aerobic rice and dry weed biomass (g m−2) across genotypes during 2007, 2008, and 2009.

y = -31.29x + 195.04

R2 = 0.581,(2007)

y = -27.394x + 151.36

R2 = 0.3435, (2008)

y = -28.282x + 194.9

R2 = 0.5945, (2009)

2030405060708090

100110120130140150160170

1 1.5 2 2.5 3 3.5 4 4.5 5Grain yield (t ha

-1)

Dry

bio

mass

, g

m-2

200720082009200720082009

277

Paper 11 Weed management strategies in dry-seeded rice systems

B.S. Chauhan

Rice is grown traditionally by direct seeding or transplanting seedlings into puddled soil. However, farmers are concerned about shifts in rainfall pattern, water and labor shortages, and high production costs.Farmers in many countries have responded to some of these concerns by shifting from transplanting to dry-seeded rice (DSR). Weeds, however, are the major constraint to the successful production of DSR. Herbicides are being used to manage weeds in DSR, but they do not provide complete and season-long weed control. Also, there are concerns about the evolution of herbicide resistance in weeds and the availability of broad-spectrum herbicides. Therefore, there is a need to integrate different weed control strategies, including the use of clean seed and equipment, the stale seedbed practice, thorough land preparation, weed-competitive varieties, high seeding rate (75 to 125 kg ha−1), narrow row spacing, crop residue as mulches, herbicides, and intercultural operations to achieve effective and sustainable weed control in DSR. Thorough land preparation to kill existing weeds, followed by the use of a weed-competitive cultivar, sown at narrow row spacing (18 to 20 cm), followed by the application of a preemergence herbicide such as oxadiazon and postemergence such as fenoxapron or bispyribac-sodiummay help to control weeds more effectively than using a single method of weed control. Further research on DSR needs to focus on the integration of appropriate agronomic and varietal aspects, such as crop plant density and spacing, varieties with early vigor, herbicide application timing, and crop rotation in order to improve the effectiveness of weed management approaches.

Rice is a principal source of food for more than half of the world population. It is grown on approximately 156 million hectares, of which 133 million ha are in Asia (FAI 2009, Prasad 2011). About 55% of the total rice area is irrigated and concerns are increasing about the supply of irrigation water in these areas because of competition with urban areas (Prasad 2011). In the future, rice farmers are likely to have limited availability of irrigation water. By 2025, most of the dry-season areas in South and Southeast Asia may suffer from “economic water scarcity” (Tuong and Bouman 2003). Rice is commonly grown by transplanting seedlings into puddled soil (intensive soil tillage under wet conditions). Puddling consumes a significant amount of water in transplanted rice. Scientists are thereforestriving to develop water-saving technologies such as alternate wetting and drying, continuous soil saturation, or irrigation at an interval (Prasad 2011). In addition to concerns about the water supply are the increased scarcity of labor because of migration of rural labor to cities and labor’s preference for nonagricultural jobs.

Because of increased production costs and/or decreased availability of labor or water, farmers in many countries have been or are in the process ofchanging their rice establishment method from manual transplanting to dry-seeded rice (DSR). In a DSR crop, dry paddy seeds are sown into a prepared seedbed after tillage or under zero-till conditions and depending on water availability; soils are kept aerobic, continuously saturated, or flooded. A DSR crop grown under nonpuddled soil conditions and without standing water, intended to save water, is referred to as aerobic rice.

278

Weeds, however, are major constraints to rice production in DSR systems. The risk of crop yield losses due to weeds is greater in dry-seeded rice than in transplanted rice for two reasons: an absence of the seedling size differential between the crop and weed, and an absence of standing water in the early stages of crop establishment (Chauhan and Johnson 2010e). The weed species commonly associated with DSR are shown in Table 1. The weeds in different DSR systems can cause rice grain yield losses from 50% to 90% (Chauhan and Johnson 2011c, Chauhan et al 2011b). Manual weeding and/or herbicides are used to control weeds in DSR. However, manual weeding is becoming less common due to the nonavailability of labor at critical times and the high wages of labor. In addition, manual weeding can be done only when weeds have reached a sufficient height to be pulled. By that time, a yield loss may already have occurred. Herbicides are easy to use; however, there are concerns about the evolution of resistance in weeds, shifts in weed populations, less availability of new and effective herbicide products, and concerns about the environment (Chauhan and Johnson 2010e). Therefore, there is a need to integrate different cultural practices with herbicide use to achieve effective and sustainable weed control in DSR.

Weed management strategies To manage weeds in DSR, various management strategies, including the use of clean seed and machinery, the stale seedbed technique, good land preparation, weed-competitive varieties, high seeding rate and narrow row spacing, crop residue as mulches, appropriate water management, and herbicides, are needed (Chauhan and Johnson 2010b,e, Mahajan and Chauhan 2011b). (i) Clean seed and machinery The use of clean paddy seed for sowing and machines for land preparation is the first and the most important weed management strategy in any crop, including DSR. Rice seed contaminated with weed seeds may introduce a new and problematic weed species to a field or may add weed seeds to the existing weed seed bank. In direct-seeded rice, weedy rice (Oryza sativa L.) is becoming an increasing weed problem in many Asian countries, mainly due to seed contamination and spread by harvesting machines. The use of weed-free rice seeds for sowing can greatly reduce weed infestation in DSR. In addition, bunds and irrigation canals should be kept free from weeds to reduce the spread of weed seeds through irrigation water. (ii) Stale seedbed practice In the stale seedbed practice, weeds are allowed to germinate by giving light irrigation or after light showers, and thereafter weed seedlings are killed by using a nonselective herbicide (glyphosate or paraquat) or shallow cultivation. The cultivation used to kill emerged seedlings, however, stimulates further weed seedling emergence. The use of the stale seedbed practice helps to reduce the weed seed bank and weed population in the crop. This practice has been found effective in reducing problems of Cyperus rotundus L. and weedy rice (Delouche et al 2007, Mahajan and Chauhan 2011b). (iii) Tillage system and land preparation Any kind of tillage system influences the vertical weed seed distribution in the soil profile. There is a need to understand the effect of different tillage systems on weed

279

distribution because this distribution has the potential to influence weed seedling emergence. In a recent study in the aerobic rice system, very limited soild is turbance in zero-till resulted in more than 75% of the weed seeds being retained in the top 2-cm soil layer, whereas the consequence of the high soildisturbance in conventional-till resulted in 62% of the weed seeds being buried to a depth of 2 to 5 cm (Figure 1; Chauhan and Johnson 2009). Weed seeds were not found in the 5- to 10-cm soil layer in zero-till, whereas the conventional-till system buried 25% of the seeds to this layer. These results suggest that low soil-disturbance systems will leave most of the weed seeds on or close to the soil surface. Germination conditions are usually more suitable for the seeds present near the soil surface than when buried deep in the soil (Chauhan and Johnson 2010e). On the other hand, weed seeds present on the soil surface are more prone to weed seed predation (Chauhan et al 2010).

Small-seeded species often require light for germination and their seedlings mainly emerge from the surface layer because seeds may not have enough reserves to support their seedlings to emerge from deep depths (Chauhan and Johnson 2009). Emergence from different soil depths has been reported to be proportional to seed size (Chauhan and Johnson 2010e). A relationship between seed weight of different weed species and burial depth required to completely inhibit seedling emergence is shown in Figure 2. Therefore, zero-till and reduced-till under aerobic rice systems may favor seedling emergence of weeds with a small seed size and those that require light for germination (Figure 3). When dry rice seeds are sown after tillage, a thorough land preparation is required for making a seedbed, weed control, and residue incorporation. Tillage operations such as plowing, disking, and harrowing may help in reducing weed populations by providing a weed-free environment at the sowing time of dry-seeded rice. The final operation of cultivation is usually followed by land leveling. The land is traditionally leveled by using a wooden plank; however, traditionally leveled fields have frequent dikes and ditches, which often result in poor crop establishment and weed control because of the uneven depth of seeding and water distribution (Jat et al 2006). The use of laser leveling has been found to lead to better crop establishment of dry-seeded rice, water savings, and improved weed control (Jat et al 2009). Therefore, if possible, a dry-seeded rice field should be leveled using a laser leveler to improve weed control efficiency. (iv) Weed-competitive varieties The choice of varieties in DSR plays an important role in rice-weed competition due to their diverse morphological traits. Tall and traditional varieties with droopy leaves are more competitive against weeds than short and improved varieties but they are often lower in yield potential and there is a trade-off between yield potential and competitive ability (Chauhan 2012). In aerobic rice systems, vegetative vigor scored at 2 weeks after sowing and weed-free yield accounted for 87% of the variation in yield between varieties under weed competition (Zhao 2006). These traits could be an efficient means of indirect selection for improving rice yield under weedy conditions. Therefore, a variety with early vigor, high early growth, and early canopy cover ability could be used under DSR to compete better with weeds. In addition, early-maturinghybrids and varieties (e.g., NK6320, Bio404, and Pant Dhan16) in Punjab, India,have also been found to have high

280

weed-competitive ability due to improved vigor and a tendency of early canopy cover (Mahajan et al 2011). Most importantly, there is a need to develop varieties for DSR under the conditions in which they will be used. The varieties currently used under dry-seeded conditions were bred for transplanted conditions.

In a recent study on the relative importance of shoot and root competition in DSR, shoot competition reduced rice grain yield by 76% to 84% with Echinochloa colona (L.) Link (Table 2). A similar trend was observed for rice panicles. In contrast, root competition with E. colona plants reduced rice grain yield by 44% to 55%. The authors concluded that shoot competition for light was the primary mechanism determining competitive outcomes between DSR and E. colona. Rice grain yield was highly correlated with aboveground and belowground rice dry weights (Figure 4). There was a linear relation between rice grain yield and root dry weight, with 91% of the variation in grain yield explained by the association. The study illustrated that shoot traits can be selected as criteria for increasing rice yield in competition with weeds. The authors also suggested that root competition, as it accounted for up to 55% reductions in rice grain yield when rice was grown with weed plants, should not be overlooked when examining rice-weed interactions (Chauhan and Johnson 2010c). (v) Seeding rate In northwestern India, a seeding rate of 15 to 25 kg seed ha−1 is being used for DSR where farmers have access to good-quality seed and a well-leveled seedbed, and losses due to weeds, bird, rats, insects, and nematodes are avoided (Chauhan et al 2011b). Sowing at such low seeding rates (estimated seeding rates shown in Figure 5) becomes possible with the availability of appropriate seeding machines with precise seed-metering devices. However, in most areas in Southeast Asia, farmers still use a high seeding rate for DSR. In addition to compensating for poor seed quality and crop establishment, high seeding rates in DSR are used to suppress weeds at the early stages, in situations where partial or poor weed control is expected.

A recent study (Chauhan et al 2011b) showed that maximum grain yield of a DSR crop in the presence of weeds was achieved at 95 to 125 kg seed ha−1; however, seeding rates of 15 to 125 kg ha−1 had little effect on the grain yield of rice under weed-free conditions (Figure 6). The results also indicated a significant decrease in weed biomass as the seeding rate increased from 15 to 125 kg ha−1 under weedy conditions (Figure 7). In an earlier study, a seeding rate of 300 viable seeds m−2 (approximately 80 kg ha−1) was needed to avoid a large yield loss due to weeds (Zhao 2006). In another study, Ludwigia hyssopifolia (G. Don) Exell shoot biomass decreased by 96% when grown under rice interference compared with its growth without interference, which indicates that crop interference could be a powerful tool to reduce competition from this weed (Chauhan et al 2011a). Similar results were reported for Cyperus iria L., Echinochloa crus-galli (L.) Beauv., and Leptochloa chinensis (L.) Nees (Chauhan and Johnson 2010d, 2011b). These studies suggest that high seeding rates may help to suppress weed growth and reduce grain yield losses from weed competition if no or partial weed control is expected (Chauhan et al 2011b). High seeding rates are, however, not advised when farmers use hybrids or expensive rice seed. Farmers use 15 to 25 kg seed ha−1 for hybrids. (i) Row spacing

281

Farmers in many areas broadcast rice seed at very high seeding rates. Although high seeding rates help to suppress weeds initially, it is difficult to conduct mechanical or manual weeding or intercultural operations in a broadcast crop. Some weeds, such as E. colona, E. crus-galli, and weedy rice, are hard to distinguish from rice plants at the seedling stage and they thereby escape manual weeding operations. Therefore, a crop grown in rows will allow farmers to practice interrow cultivation. The use of narrow row spacing improves the crop’s competitive ability against weeds by developing faster canopy cover compared with wider row spacing.

A DSR crop grown in wide row spacing may have higher weed biomass than a crop grown in narrow row spacing. In a recent study in aerobic rice systems, for example, weeds growing in 30-cm rice rows had 34% greater biomass than weeds grown in 15-cm rice rows (Chauhan and Johnson 2011c). In addition, the critical periods for weed control were also less for crops grown in 15-cm rows than in 30-cm rows (Figure 8). In a similar study, E. colona and E. crus-galli, problematic weeds of rice, emerging until 30 days after rice emergence had greater biomass and seed production under 30-cm rows than under 20-cm rows (Chauhan and Johnson 2010a). Therefore, a DSR crop should be grown under narrow row spacing (15 to 20 cm) to obtain a fast canopy cover and less penetration of light and ultimately less weed growth. Recently, it was reported that varieties for aerobic rice cultivation differ in weed-suppressing ability when sown in paired rows (15-30-15-cm row spacing) and the weed competitiveness of some rice varieties can be enhanced by using the paired row system (Mahajan and Chauhan 2011a). (ii) Crop residue Crop residues present on the soil surface can influence germination, emergence, and growth of weeds by influencing the conditions surrounding weed seeds. The practice of zero-till systems allows the retention of previous crop residues in the field, which can suppress seedling emergence of many weed species. The response of weed emergence to residue, however, varies among the species (Figure 9). The seedling emergence of some weed species can be suppressed by using a low amount (ca. 2 t ha−1) of residue, whereas a large amount of crop residue is needed to considerably reduce emergence and growth of other weed species. Because of concern about the depletion of soil organic matter and environmental pollution due to the burning of crop residues, retention of crop residues has been proposed in some intensive systems (Chauhan and Johnson 2010e). Where residue is used, it will be important to balance the quantities of residue required to suppress weeds with those that will not have negative effects, such as reducing crop emergence, complicating tillage and crop management, and reducing the effectiveness of soil-active herbicides (Chauhan and Johnson 2011a). Machines capable of seeding in a residue amount of up to 6 t ha−1 are available; however, there is still room for further improving the efficiencies of these machines, especially for small-scale farmers. (iii) Water management Water has been a very effective tool for weed management in any system of rice. However, the timing, duration, and depth of flooding should be optimum to effectively suppress weed growth in water-short environments. Exploiting the knowledge and understanding the differences between rice and weeds in their tolerance of early

282

flooding could enable more effective weed management through more precise timing of flooding (Chauhan and Johnson 2010e). Recent studies showed that intermittent flooding or alternate wetting and drying (2 days out of 7 days for 28 days) can also help to suppress seedling emergence of Cyperus difformis L., C. iria, L. chinensis, and L. hyssopifolia (Figure 10); however, the flooding needs to be integrated with other control measures to manage weeds completely and more effectively. In water-short environments, flooding after hand weeding or herbicide application could largely prevent the subsequent growth of weeds and reduce the need for further weeding. In situations where farmers have a limited water supply, early rather than later flooding would make the best use of water to control weeds effectively. (iv) Herbicides Herbicides are an important tool for weed management in DSR and their use is likely to increase in thefuture because of increasing labor costs. The use of herbicides has become even more important where there is a morphological similarity between weeds (e.g., E. colona and E. crus-galli) and rice, and simultaneous emergence of weeds and rice under dry-seeded conditions. However, herbicide should be used in conjunction with the cultural practices mentioned above. Factors such as availability of herbicides, incidence of weed flora, and timing and method of application should be considered before developing any herbicide program. Some preemergence herbicides may have a phytotoxic effect on rice germination if heavy rain occurs immediately after herbicide application. In situations where heavy rain is expected after spray, it is advisable to wait for the application of early postemergence herbicides rather than spraying preemergence herbicides.

Various preemergence (pendimethalin, oxadiazon, oxadiargyl, etc.) and postemergence (azimsulfuron, bispyribac-sodium, fenoxaprop, penoxsulam, metsulfuron, ethoxysulfuron, 2,4-D, their commercial mixtures, etc.) herbicides are recommended and used under DSR but often these herbicides have a narrow weed control spectrum, especially in upland and aerobic rice systems. Bispyribac-sodium and penoxsulam, for example, are poor in controlling L. chinensis, Dactyloctenium aegyptium (L.) Willd., and C. rotundus. Under assured irrigated conditions, frequent irrigations after postemergence herbicide application may control weeds effectively. However, in aerobic rice systems, there may be a need for one handweeding or mechanical weeding (by rotary or conoweeder) in addition to the application of postemergence herbicides. After the weeding, the crop should have a dense canopy to avoid any future growth of subsequent weeds. There is a need to have an effective herbicide (with both pre- and postemergence components) that can be applied at 4−5 weeks after rice sowing (Chauhan 2012). The postemergence component would kill the emerged seedlings and the preemergence component would not allow new weeds to emerge for another 2−3 weeks and, by that time, the crop would have a closed canopy to suppress any further weed emergence.

Herbicide must be handled safely so as to reduce excessive waste, environmental concerns, and injury to adjacent crops and beneficial organisms. Several strategies can reduce the development of herbicide resistance in weeds: (1) rotation of herbicides with different modes of action, (2) use of herbicide mixtures and recommended rates of herbicides, (3) pulling outweeds escaped from herbicide

283

application, (4) cleaning of machinery to prevent spread of resistant weeds from one area to another, (5) adoption of integrated weed control practices, and (6) adoption of crop rotations and diversification. (v) Integrated weed management. Reliance on a single herbicide could result in the development of resistance in weeds and a shift in weed populations. No single method can effectively control weeds; therefore, there is a need to integrate different weed control strategies to manage weeds more effectively (Figure 11; Chauhan 2012). For example, thorough land preparation to kill existing weeds followed by the use of high seeding rates (75 to 125 kg ha−1) for sowing at narrow row spacings (18 to 20 cm) followed by the application of pre- and/or postemergence herbicides may help to control weeds more effectively than using a single method. Similarly, under zero-till or reduced-till systems, allowing weed seeds present on the soil surface to germinate and predation followed by killing emerged weed seedlings by the use of a nonselective herbicide followed by sowing the crop under residue followed by an application of a postemergence herbicide may help to manage weeds very effectively. Planting legumes, such as mungbean and cowpea, in between two cereal crops may also help to reduce buildup of weed populations in DSR. Rotation of tillage systems, crop establishment methods, or crops may help to deflect the “trajectories” of likely weed population shifts. For example, the repeated use of a zero-till system in rice may lead to greater densities of E. colona and Digitaria ciliaris (Retz.) Koel, which, in turn, could be discouraged by shifting to other tillage or crop establishment methods (Chauhan and Johnson 2010e). Theoretically, an integrated weed management approach is easy to implement; however, in practice, more serious consideration is needed in establishing how different weed management strategies can be integrated. Future research in DSR needs to focus on the integration of appropriate agronomic and varietal aspects, such as crop plant density and spacing, nutrition, varieties with early vigor, herbicide application timing, and crop rotation, in order to improve the effectiveness of weed management approaches. Insect pests Buhler et al (2000) compared some of the characteristics of weeds (relative to insect pest biology) that may affect the development of integrated weed management approaches: weeds as a producer vs. insects as a consumer, a weed population usually stable once established vs. a rapid increase or decrease in insect population due to natural processes, weed generation time of at least 1 year vs. insect generation time of 1 year or less, weed population seldom synchronous vs. insect population often synchronous, weeds are immobile within a generation vs. insects are mobile during at least one growth stage, high reproduction of weeds from a single generation vs. low reproduction of insects from multiple generations, and highly variable yield loss per individual weed plant vs. predictable yield loss per individual insect. The authors also suggested that these differences would not apply to all weeds or insects, but the information would indicate the implications for integrated management in weeds for a differential response of weeds and insects to resource management and cultural practices.

284

Conclusions DSR is a water- and labor-saving technology; however, weeds are the major constraint to the successful production of a DSR crop. A single strategy may not provide effective control of weeds. Therefore, various weed management strategies, including the use of clean seed and machinery, the stale seedbed practice, good land preparation, weed-competitive varieties, high seeding rate and narrow row spacing, crop residue, water management, and herbicides, need to be integrated to achieve effective and sustainable weed control in DSR. References Buhler DD, Liebman M, Obrycki JJ. 2000. Theoretical and practical challenges to an IPM

approach to weed management. Weed Sci. 48:274-280. Chauhan BS. 2012. Weed ecology and weed management strategies for dry-seeded

rice in Asia. Weed Technol. 26:1-13. Chauhan BS, Johnson DE. 2009. Influence of tillage systems on weed seedling

emergence pattern in rainfed rice. Soil Till. Res. 106:15-21. Chauhan BS, Johnson DE. 2010a. Implications of narrow crop row spacing and delayed

Echinochloa colona and Echinochloa crus-galli emergence for weed growth and crop yield loss in aerobic rice. Field Crops Res. 117:177-182.

Chauhan BS, Johnson DE. 2010b. Opportunities to improve cultural approaches to manage weeds in direct-seeded rice. In: Zydenbos SM, editor. 17th Australasian Weeds Conference, Christchurch, New Zealand. New Zealand Plant Protection Society. p 40-43.

Chauhan BS, Johnson DE. 2010c. Relative importance of shoot and root competition in dry-seeded rice growing with junglerice (Echinochloa colona) and ludwigia (Ludwigia hyssopifolia). Weed Sci. 58:295-299.

Chauhan BS, Johnson DE. 2010d. Response of rice flatsedge (Cyperus iria) and barnyardgrass (Echinochloa crus-galli) to rice interference. Weed Sci. 58:204-208.

Chauhan BS, Johnson DE. 2010e. The role of seed ecology in improving weed management strategies in the tropics. Adv. Agron. 105:221-262.

Chauhan BS, Johnson DE. 2011a. Ecological studies on Echinochloa crus-galli and the implications for weed management in direct-seeded rice. Crop Prot. 30:1385-1391.

Chauhan BS, Johnson DE. 2011b. Phenotypic plasticity of Chinese sprangletop (Leptochloa chinensis) in competition with seeded rice. Weed Technol. 25:652-658.

Chauhan BS, Johnson DE. 2011c. Row spacing and weed control timing affect yield of aerobic rice. Field Crops Res. 121:226-231.

Chauhan BS, Migo T, Westerman PR, Johnson DE. 2010. Post-dispersal predation of weed seeds in rice fields. Weed Res. 50:553-560.

Chauhan BS, Pame ARP, Johnson DE. 2011a. Compensatory growth of ludwigia (Ludwigia hyssopifolia) in response to interference of direct-seeded rice. Weed Sci. 59:177-181.

285

Chauhan BS, Singh VP, Kumar A, Johnson DE. 2011b. Relations of rice seeding rates to crop and weed growth in aerobic rice. Field Crops Res. 121:105-115.

Delouche JC, Burgos NR, Gealy DR, de San Martin GZ, Labrada R, Larinde M, Rosell C. 2007. Weedy rices: origin, biology, ecology and control. FAO Plant Production and Protection Paper 188. FAO, Rome.

FAI. 2009. Fertilizer statistics 2009-10. The Fertilizer Association of India, New Delhi. Jat ML, Chandana P, Sharma SK, Gill MA, Gupta RK. 2006. Laser land leveling: a

precursor technology for resource conservation. Rice-Wheat Consortium Technical Bulletin Series 7, New Delhi, India. 48 p.

Jat ML, Gathala MK, Ladha JK, Saharawat YS, Jat AS, Vipin, Kumar AS, Sharma SK, Kumar V, Gupta RK. 2009. Evaluation of precision land leveling and double zero-till systems in the rice-wheat rotation: water use, productivity, profitability and soil physical properties. Soil Till. Res. 105:112-121.

Mahajan G, Chauhan BS. 2011a. Effects of planting pattern and cultivar on weed and crop growth in aerobic rice system. Weed Technol. 25:521-525.

Mahajan G, Chauhan BS. 2011b. Weed management in direct drilled rice. Indian Farming, April 2011. p 6-9.

Mahajan G, Ramesha MS, Kaur R. 2011. Screening for weed competitiveness in rice: way to sustainable rice production in the face of global climate change. In: Proceedings of International Conference on Preparing Agriculture for Climate Change, 6-8 Feb. 2011. PAU, Ludhiana, India. 115 p.

Prasad R. 2011. Aerobic rice systems. Adv. Agron. 111:207-247. Tuong TP, Bouman BAM. 2003. Rice production in water-scarce environments. In: Kijne

JW, Barker R, Molden D, editors. Water productivity in agriculture: limits and opportunities for improvements. UK: CAB International. p 53-67.

Zhao D. 2006. Weed competitiveness and yielding ability of aerobic rice genotypes. PhD thesis. Wageningen University, The Netherlands. 142 p.

Notes Author’s address: International Rice Research Institute, Los Baños, Philippines. Some of the contents of this chapter are taken from B.S. Chauhan, “Weed ecology and weed management strategies for dry-seeded rice in Asia,”Weed Technology 26(2012):1-13.

286

Figure 1. Effect of tillage systems on vertical seed distribution (Chauhan and Johnson 2009).

Figure 2. Relationship between burial depths required to completely inhibit seedling emergence of different weed species and their seed weight (Chauhan and Johnson 2010e).

Figure 3. Effect of tillage systems on seedling emergence of different weed species under aerobic rice systems (Chauhan and Johnson 2009).

Figure 4. Relationship between grain yield and leaf dry weight (DW), straw DW, and root DW of rice (Chauhan and Johnson 2010c).

Figure 5. Estimated seeding rates when two seeds are planted per hole at 20-cm plant-to-plant spacing and 20-cm row spacing.

Figure 6. The relationship of rice grain yield to rice seeding rate under weedy (solid line for hybrid and short broken line for inbred) and weed-free (long broken line for hybrid and dotted line for inbred) conditions in aerobic rice systems in 2008 and 2009 in the Philippines (Chauhan et al 2011b).

Figure 7. The relationship between weed biomass and rice seeding rates in aerobic rice systems (adapted from Chauhan et al 2011b).

Figure 8. Effect of row spacing and different periods of weed competition on rice grain yields in aerobic rice systems. The critical weed-free periods to achieve 95% of maximum yield are shown between the broken vertical lines.

Figure 9. Effect of residue amount (Mg ha−1) on percent seedling emergence (± SEm) of different weed species (Chauhan and Johnson 2010).

Figure 10. Effect of flooding depth and duration (intermittent: 2 days out of 7 days for 28 days, and continuous: for 28 days) on seedling emergence of four weed species (Chauhan and Johnson 2010e).

Figure 11. Different weed management strategies for aerobic and dry-seeded rice systems (Chauhan 2012).

287

Table 1. Weed species associated with dry-seeded rice in Asia. Weed species Family Group Ageratum conyzoides L. Asteraceae Broadleaf Alternanthera sessilis (L.) R. Br. ex DC. Amaranthaceae Broadleaf Amaranthus spinosus L. Amaranthaceae Broadleaf Amaranthus viridis L. Amaranthaceae Broadleaf Caesulia axillaris Roxb. Asteraceae Broadleaf Celosia argentea L. Amaranthaceae Broadleaf Commelina benghalensis L. Commelinaceae Broadleaf Commelina diffusa Burm. f. Commelinaceae Broadleaf Cynodon dactylon (L.) Pers. Poaceae Grass Cyperus difformis L. Cyperaceae Sedge Cyperus esculentus L. Cyperaceae Sedge Cyperus iria L. Cyperaceae Sedge Cyperus rotundus L. Cyperaceae Sedge Dactyloctenium aegyptium (L.) Willd. Poaceae Grass Digitaria ciliaris (Retz.) Koel Poaceae Grass D. sanguinalis (L.) Scop. Poaceae Grass Echinochloa colona (L.) Link Poaceae Grass Echinochloa crus-galli (L.) P. Beauv. Poaceae Grass Echinochloa glabrescens Munro ex Hook. f. Poaceae Grass Eclipta prostrata (L.) L. Asteraceae Broadleaf Eleusine indica (L.) Gaertn. Poaceae Grass Eragrostis japonica (Thunb.) Trin. Poaceae Grass Fimbristylis dichotoma(L.) Vahl Cyperaceae Sedge Fimbristylis miliacea (L.) Vahl Cyperaceae Sedge Imperata cylindrica (L.) Raeuschel Poaceae Grass Ipomoea triloba L. Convolvulaceae Broadleaf Ischaemum rugosum Salisb. Poaceae Grass Leptochloa chinensis (L.) Nees Poaceae Grass Ludwigia hyssopifolia (G. Don) Exell Onagraceae Broadleaf Ludwigia octovalvis (Jacq.) P. H. Raven Onagraceae Broadleaf Ludwigia prostrata Roxb. Onagraceae Broadleaf Murdannia nudiflora (L.) Brenan Commelinaceae Broadleaf Oryza sativa L. (weedy rice) Poaceae Grass Portulaca oleracea L. Portulacaceae Broadleaf Rottboellia cochinchinensis (Lour.) Clayton Poaceae Grass Scirpus juncoides Roxb. Cyperaceae Sedge Trianthema portulacastrum L. Aizoaceae Broadleaf

288

Table 2. The reduction in growth of rice due to full, shoot, and root competition with Echinochloa colona (Chauhan and Johnson 2010c).

E. colona plants (no.)

Type of competition

Percent reduction in growth Panicles Grain yield

4 Full 75 86 4 Shoot 72 76 4 Root 11 44 8 Full 84 93 8 Shoot 81 84 8 Root 13 55

289

Figure 1. Effect of tillage systems on vertical seed distribution (Chauhan and Johnson 2009).

Tillage systemsZero-till Conventional-till

Seeds recovered (%)

0

20

40

60

80

100

0-2 cm 2-5 cm 5-10 cm

Fig. 1. Effect of tillage systems on vertical seed distribution (Chauhan and Johnson 2009).

290

Figure 2. Relationship between burial depths required to completely inhibit seedling emergence of different weed species and their seed weight (Chauhan and Johnson 2010e).

D = 1.94 + 4.24x - 0.54x2

R2 = 0.63

Seed weight (mg)0 1 2 3 4 5 6

Depth (cm)(required to completely inhibit emergence)

0

2

4

6

8

10

12

291

Figure 3. Effect of tillage systems on seedling emergence of different weed species under aerobic rice systems(Chauhan and Johnson 2009).

Weed species

D. cilia

ris

E. colo

na

E. pros

trata

E. indic

a

Seedling emergence (%)

0

5

10

15

20

25Zero-tillConventional-till

292

Figure 4. Relationship between grain yield and leaf dry weight (DW), straw DW, and root DW of rice (Chauhan and Johnson 2010c).

y = 0.30 x + 0.39adj. r2 = 0.92

Leaf

DW

(g p

lant

-1)

0

3

6

9

12

15

18

y = 0.89 x + 1.11adj. r2 = 0.94

Stra

w D

W (g

pla

nt-1

)

0

10

20

30

40

50

y = 0.06 x + 0.15adj. r2 = 0.91

Grain yield (g plant-1)

0 10 20 30 40 50 60

Roo

t DW

(g p

lant

-1)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

293

Figure 5. Estimated seeding rates when two seeds are planted per hole at 20-cm plant-to-plant spacing and 20-cm row spacing.

--------------------- 1 m ---------------------

------

------

-- 1

m --

------

------

20 cm

20 c

m

Example, 2 seeds hole-1 = 50 seeds m-2 = 10.0-12.5 kg ha-1

294

Figure 6. The relationship of rice grain yield to rice seeding rate under weedy (solid line for hybrid and short broken line for inbred) and weed-free (long broken line for hybrid and dotted line for inbred) conditions in aerobic rice systems in 2008 and 2009 in the Philippines (Chauhan et al 2011b).

20080

1000

2000

3000

4000

5000Weedy-Hybrid;y = 505 + 48.9x - 0.28x2; R2 = 0.62Weed free-Hybrid;y = 3212 + 9.5x - 0.08x2; R2 = 0.04Weedy-Inbred;y = 315 + 30.6x - 0.11x2; R2 = 0.80Weed free-Inbred;y = 2319 + 33.6x - 0.23x2; R2 = 0.36

2009

Seed rate (kg ha-1)0 25 50 75 100 125

Grain yield (kg ha-1)

0

1000

2000

3000

4000

5000

Weedy-Hybrid;y = -54 + 47.2x - 0.26x2; R2 = 0.72Weed free-Hybrid;y = 4082 + 11.0x - 0.12x2; R2 = 0.21Weedy-Inbred;y = -134 + 39.6x - 0.21x2; R2 = 0.75Weed free-Inbred;y = 3618 + 9.5x - 0.10x2; R2 = 0.08

295

Figure 7. The relationship between weed biomass and rice seeding rates in aerobic rice

systems (adapted from Chauhan et al 2011b).

Seeding rate (kg ha-1

)

0 25 50 75 100 125

Weed b

iom

ass (

g m

-2)

0

100

200

300

400

500y = 370 - 2.41x; r = 0.75

296

Figure 8. Effect of row spacing and different periods of weed competition on rice grain yields in aerobic rice systems. The critical weed-free periods to achieve 95% of maximum yield are shown between the broken vertical lines.

15-cmGrain yield (% of weed free yield)

0

20

40

60

80

100

Weedy until daysWeed free until days

30-cm

Days after sowing0 20 40 60 80 100 120

0

20

40

60

80

100

297

Figure 9. Effect of residue amount (Mg ha−1) on percent seedling emergence (± SEm) of different weed species (Chauhan and Johnson 2010).

Chromolaena odorata

0

20

40

60

80

100

Digitaria ciliaris

Seed

ling

emer

genc

e (%

)

0

20

40

60

80

100

Digitaria longiflora

Residue amount (Mg ha-1)0 1 2 4 6

0

20

40

60

80

100

Echinochloa colona

0

20

40

60

80

100

Eleusine indicaSeedling em

ergence (%)

0

20

40

60

80

100

Melochia concatenata

0 1 2 4 60

20

40

60

80

100

298

Figure 10. Effect of flooding depth and duration (intermittent: 2 days out of 7 days for 28 days, and continuous: for 28 days) on seedling emergence of four weed species (Chauhan and Johnson 2010e).

Cyperus difformis

0

20

40

60

80

100

Cyperus iria

0 2 4 6 8 10Seed

ling

emer

genc

e (%

of 0

cm

floo

ding

dep

th)

0

20

40

60

80

100

Leptochloa chinensis

0

20

40

60

80

100

Ludwigia hyssopifolia

Depth of flooding (cm)0 2 4 6 8 10

Seedling emergence (%

of 0 cm flooding depth)

0

20

40

60

80

100

Intermittent floodingContinuous flooding

299

Figure 11. Different weed management strategies for aerobic and dry-seeded rice

systems (Chauhan 2012).

300

Paper 12 Adoption and dissemination of alternate wetting and drying technology in pump irrigation system areas in Bangladesh

Florencia G. Palis, Rubenito M. Lampayan, Ekkehard Kürschner, and Bas Bouman

The water-saving technology alternate wetting and drying (AWD) was introduced in Bangladesh in 2004 to overcome the problem of water scarcity, especially during the dry season. Various organizations, both government and nongovernment, including the private sector, have conducted independent on-farm validation and AWD dissemination since then. AWD adoption, however, is still in its early stage. The perceived benefits with AWD reported by farmers and pump owners are that AWD saves water, reduces irrigation frequency and irrigation costs, has higher yield, and saves fuel and electricity consumption. The increase in yield was attributed by farmers to the increase in the number of tillers and panicles, bigger and heavier grains with good shape, and that plants are sturdier and thereby lodging decreases. Some scientists, however, attributed the increase in yield to the increase in the number of tillers, thus increasing rice production per unit area. The key factors that were identified to influence AWD adoption by farmers included unreliable water and energy supply, farmer organization and collective action, type of irrigation system, incentives for pump operators for deep tube wells (DTW) and pump owners for shallow tube wells (STW), and the payment arrangement for irrigation services.

Capitalizing on the momentum from the initial dissemination efforts by various disseminating organizations through field demonstrations and training during the boro season, the following strategies are recommended: strengthening partnerships among stakeholders, strengthening farmer organization, training and mentoring field staff/stakeholders to influence policy change at the local institutional level, and developing incentive mechanisms for farmers and pump owners.

Abbreviations and acronyms: AWD = alternate wetting and drying; BADC = Bangladesh Agricultural Development Corporation; BRRI = Bangladesh Rice Research Institute; NARES = national agricultural research and extension systems; RDA = Rural Development Academy; DTW = deep tube well; STW = shallow tube well.

Rice is the staple food of 150 million peoplein Bangladesh.It is grown on 8.3 million hectares, about 77% of the country’s total cultivated area (BBS 2008). Around 60% of national rice production comes from boro rice (dry season), which is grown on more than 4.8 million hectares of land and is mostly irrigated by groundwater pumps. Thus, the stability of boro rice production is most important for national food security (Fujita 2004), as the production of aman rice is quite variable depending on weather conditions. Water resources are becoming scarcer worldwide, and Bangladesh is no exception. Although the country is well known for its abundant rainfall during the monsoon season, water shortage is always the main problem during the boro season (dry season). Water is scarcest in the northwest and southwest regions during the boro season because of the low amount of annual rainfall (Chowdhury 2012).

In Bangladesh, 80% of groundwater is used for crop production where boro rice alone uses 73% of total irrigation water. This results in a declining water table during the

301

boro season, requiring deep setting of pumps to continually draw groundwater and making irrigation cost expensive. Correspondingly, the demand for energy is increasing, as electricity or diesel fuel is needed to operate pump systems, thus affecting the chronically deficient energy situation in the country. The irrigation cost for rice represents 30−40% of the total cost for rice production, which is much higher than in other countries of South and Southeast Asia (Sattar et al 2009, Hossain 2009).

Groundwater arsenic contamination has also been reported in many of the eastern and western districts (Chowdhury et al 2000). Arsenic was claimed to be present in the food chain, which threatensthe use of groundwater for irrigation. The judicious use of water in intensive irrigated areas is therefore crucial to maintain sustainable and cost-efficient rice production, a better environment, and human health (Chowdhury et al 2000).

To address water scarcity, water-use efficiency, and associated irrigation problems in Bangladesh, alternate wetting and drying (AWD) as an efficient practice for water management in rice cultivation was introduced by the International Rice Research Institute (IRRI) tonational agricultural research and extension system (NARES) partnersin the country, starting with the key partner—the Bangladesh Rice Research Institute (BRRI). Jointly, AWD was introduced to farmers through a simple tool—a perforated water tube, now widely known as a pani pipe (pani means water in Bangla). The pani pipe is inserted into the soil to observe the perched water table in the field. The water inside the panipipe reflects the level of water in the field. When the water drops below 15 cm from the surface, as can be seen in the pani pipe, this signals farmers to apply irrigation water to the rice plant. The tube can be used in any field or soil type, except in soil that is too sandy to be good for growing paddy rice. In clay soil with low percolation rate, the period of nonflooding may last up to 10 days, whereas, in more loamy and permeable soils, it may last only 2 to 3 days. This chapter discusses the adoption and dissemination of AWD in Bangladesh.

Evolution of AWD dissemination in Bangladesh AWD was introduced in Bangladesh in 2004 with IRRI playing a central role in promoting the technology (Kürschner et al 2010). IRRI, in collaboration with BRRI and the Rural Development Academy (RDA), initially tested AWD technology on-station under the ADB-funded research on the development and adaptation of rice germination under drought situations (Sattar et al 2009, Miah et al 2009 a, b). The IRRI-Bangladesh office observed that AWD technology is simple for farmers to follow and staff encouraged BRRI scientists to validate it in a real production system through farmer participatory research in farmers’ fields.

In the 2006-07 boro season, Gazipur and Rangpur BRRI research stationsconducted farmer participatory research in farmers’ fields, which involved 30 farmers, to validatereal AWD performance. At this time, AWD was already a component of the Irrigated Rice Research Consortium activities in Bangladesh, and of the ADB-funded project on “Water-saving technology for South Asia.”

Also, in 2006, a workshop was conducted jointly by IRRI and BRRI to showcase findings from the AWD trials to facilitate wide-scale dissemination of the technology.

302

Workshop participants included those from the Bangladesh Agricultural Development Corporation (BADC), RDA, Barindra Multipurpose Development Authority (BMDA), the Department of Agricultural Extension (DAE), and NGOs such as the Rangpur Dinajpur Rural Services (RDRS), Practical Action, and PETRACHEM. Preliminary trials in farmers’ fields confirmed the potential benefits of the technology, especially for boro rice. The summarized benefits are as follows: irrigation frequency decreases by 3 to 7 in a season, which varied with soil type; fuel consumption declines by 30 liters of diesel per hectare; power-operated tube wells decrease cost by about BDT 1,000 (US$12) per hectare; 20% water is saved per hectare; and average rice yield increases by half a ton per ha (Miah et al 2009 a, Sattar et al 2009, Alam et al 2009). Further validation trials on AWD were carried out at various sites to spread the AWD practice. In May 2007, a crop-cut ceremony at the boro rice harvest on a BADC farmwas held and led by the secretary of agriculture. It was also attended by key scientists from BRRI and IRRI, key government agencies, NGOs, the private sector, and farmers. This event was followed by a training course on AWD in August 2007 given by several IRRI water scientists to DAE key extension staff nationwide, including other government and nongovernment organizations, members of the private sector, pump owners, farmer leaders, and farmers. In 2008, a separate training of trainers (ToT) on AWD was given to senior officers of Syngenta and to newly recruited BRRI staff. With the training activities and the results of many validation trials that re-confirmed the initial findings, a number of government, nongovernment, and private-sector organizations in Bangladesh have participated in piloting and testing AWD in the country. Although some organizations participated only for a short duration (i.e., BADC and RDA), several key organizations incorporated AWD into their extension strategies and programs (Table 1).

One important milestone, which is considered as the tipping point for AWD dissemination in Bangladesh, was the conduct of the first national workshop for AWD in 2009. In this workshop, the secretary of agriculture directed the DAE to include AWD in DAE extension programs throughout the country. This event, which was facilitated by IRRI, also provided an opportunity to review dissemination processes, share results and experiences, and assess the extent of the spread of the technology (Sattar et al 2009). In December 2011, an international workshop on AWD was held in Bangladesh to further promotethe technology on a larger scalethrough exchange of information and experiences among countries about the technology. The countries represented were the Philippines, Vietnam, India, Nepal, and Bangladesh. AWD dissemination strategies The following four dissemination strategies were adopted: (i) Training of trainers (ToT) Field and extension staff deployed in the field played a vital role in AWD dissemination. They are responsible for taking knowledge about new technologies to farmers, who often rely on them for advice whenever they encounter problems in their agricultural activities. Therefore, extension staff needs to have adequate knowledge on the technology, good rapport with farmers, and the ability to effectively communicate the

303

technology tofarmers. The extension staff well trained on AWD is better prepared and more confident in explaining the technology and training farmers on it.

ToT was conducted strategically by using existing extension networks and building on the mandates of the participants’ organizations. Target participants for the ToT have their own extension programs. This was evident in the ToT conducted in 2007 and 2008 that resulted in widespread piloting, testing, and dissemination of AWD (Table 1).

BRRI played an important role in providing guidance for AWD dissemination. However, its mandate and capacity are not oriented toward disseminating the technology on a larger scale. The DAE, with a national mandate to disseminate agricultural technologies to farmers in Bangladesh, is a major player and has incorporated AWD in most of its rice- and irrigation management-related extension activities. This was evident inits two specific projects: the World Bank-funded National Agricultural Technology Project (NATP) and the Agricultural Engineering Technologies Extension Project (AETEP). NATP is a 15-year national project with a large extension component. It started in July 2008 and aimed to improve agricultural productivity and farm income with a focus on small and marginal farmers (www.natpdae.gov.bd). Originally, the AETEP was run by another government agency but was continued by DAE from 2007 to 2009. The BMDA is mandated for infrastructure development, particularly irrigation in the Barind region, where water is a major problem.

BMDA has mobilized its own resources in disseminating AWD through the water management system of its DTWs (Hossain et al 2009). The RDRS is a major NGO in the northern part of Bangladesh engaged in rural development and empowering the rural poor. It disseminated AWD through its ongoing projects with farmers and network of local NGO partners in the region. Syngenta is a private multinational company and a major supplier of agrochemicals in Bangladesh. It uses itsown funds to provide PVC tubes for AWD to its customers through retail sellers, contract farmers, and during farmers’ meetings. For Syngenta, disseminating AWD is a win-win strategy: the company helps farmers improve their production and income, and at the same time strengthen relationships between Syngenta and farmers—their clients (Kabir 2009). (ii) Demonstration trials and field days Farmer participatory research was carried out through demonstration plots or trials in farmers’ fields. Participatory research focuses on a process of sequential reflection and actions for which local knowledge is appreciated, and used for research planning. Farmer field days were also organized in the demonstration plots to showcase the technology to a larger audience to promote AWD technology faster.

Demonstration plots or trials were coursed through many ways: individual pump owners, farmer field schools, and by group or block. Through demonstration plots, farmers were able to have practical experience with the technology, while other farmers had the opportunity to observe the performance of the standing crops. For instance, in the beginning, farmers were using the perforated perched tube or pani pipe to determine the appropriate time to irrigate their respective rice fields. However, after experiencing the AWD technology, farmers were already using available materials to substitute for the pani pipes, such as Coca Cola bottles. At the harvest time of key demonstration

304

trials/sites, farmer field days were organized and attended by farmers within and among neighboring villages as well as by private and public organizations, and mass media.

Farmer participatory training provided to farmers about AWD played a key role in the successful dissemination of AWD (Kürschner et al 2010). The active participation of farmers in validation and demonstration trials induces farmer-to-farmer learning and sharing. From the 121 farmers surveyed, around 21% of the farmers who did not adopt said that they did not receive adequate training or did not understand AWD. The timing in the conduct of AWD training, however, is important and the training should be given in the early part of the dry season. (iii) Workshops Workshops can be effective in up-scaling any technology if they are coupled with strategic action. The first national workshop for AWD conducted in 2009 was launched by the secretary of agriculture and participated in by key government organizations whose mandate was on water management, nongovernment organizations, and the private sector. It was during this workshop when the secretary of agriculture directed the DAE to include AWD in DAE extension programs throughout the country. (iv) Mass media and others Likewise, opinion leaders (i.e., experienced farmers, seed or agrochemical dealers, and local government representatives) and mass media campaigns also provided strong support to the diffusion process of AWD in the country (Kürschner et al 2010). TV programs and English and Bangla newspapers also engaged in raising awareness about AWD (Miah 2008). This was also evident at a press conference on AWD at the BRAC center with television and media personnel during the national workshop on AWD in 2009 (Sattar et al 2009). Adoption of AWD Indeed, dissemination of AWD in Bangladesh has been progressing strongly. However, regardless of so many dissemination activities, unless farmers perceive that the technology is effective, more profitable, simple to implement, and culturally acceptable, adoption will not be realized (Palis et al 2004). Adoption is defined as the decision of an end user to continue full use of an innovation, in this case AWD technology.

The adoption of AWD by rice farmers in Bangladesh is still in its early stage. From the 272 farmers and pump owners surveyed by Kürschner et al (2010) in Rajshahi and Rangpur divisions of northwest Bangladesh, 121 farmers and pump owners (44%) adopted AWD. However, about 70% started adopting AWD only during the 2010 boro season, 26% after the 2009 boro season, and just about 4% of the farmers adopted the technology after 2008. Most of the respondents had received AWD training and orientation only in 2010.

However, these farmers and pump owners have already been unknowingly practicing some form of AWD to cope with water scarcity and as a means to save fuel cost, making “unintended” AWD a common practice (Kürschner et al 2010). That is, farmers who do not receive irrigation water at the agreed time are forced to let the water level drop below the soil surface. Farmers don’t have a choice but to wait for irrigation water to come. About 24% of the total farmers surveyed reported practicing “unintended”

305

AWD, which can be further classifiedinto “severe” (longer drying period) and “mild” (water level is just at saturation or a few cm below the soil surface before re-irrigation). Farmers’perceptions and factors constraining AWD adoption

A number of perceived advantages and risks with AWD were reported by farmers and pump owners (Kürschner et al 2010, Palis et al 2010, 2011). These are that AWD saves water, saves fuel and electricity for irrigation, thereby incurring less irrigation cost, and increases yield.

Figure 1 shows the positive perceptions of using AWD by farmers and pump owners. More than 70% perceived that AWD saves water, more than 30% perceived that AWD incurs less irrigation cost, 24% to 37% perceived that AWD increases yield, 14% to 21% perceived that AWD saves fuel, and 5% to10% perceived that AWD saves electricity (Kürschner et al 2010). Empirical evidence from on-farm field experiments (Sattar et al 2009) and farmers’ reports from the survey (Kürschner et al 2010) indicated that AWD saves water from 15% to 30%, reduces irrigation frequency by 28% or from 3 to 7 irrigations in a season depending upon the soil type, reduces irrigation cost by 20% due to a reduction in irrigation frequency equivalentto 30 liters of diesel ha-1, reduces the electric bill for power-operated tube wells by BDT 1,000 or $12 ha−1, and increases yield by 0.5 ton ha−1 or 10%. The increase in yield was attributed by farmers to the following: an increase in number of tillers and panicles, bigger and heavier grains with good shape, and AWD plants are sturdier, thereby reducing lodging. Sattar et al (2009) surmised that, because of proper aeration due to AWD, the number of tillers per hill increased by 2 to 5 and production per unit area increased, resulting in an average increase in rice yield by 0.5 ton ha−1. The reported increase in yield due to AWD, however, needs further experimental validation in farmers’ fields.

Regardless of the positive perceived advantages reported by farmers and pump owners, AWD adoption can still be considered low (Kürschner et al 2010, Palis et al 2010, 2011). The key factors that constrain AWD adoption were investigated. These factors included unreliable water and energy supply, farmer organization and collective action, type of irrigation system, incentives for pump operators for DTWs and pump owners for STWs, and the payment arrangement for irrigation services (Kürschner et al 2010, Palis et al 2010, 2011).

The influence of unreliable water and energy supply largely depends on the local situation. Power and water scarcity is considered as a driving force in increasing awareness of the value of water savings and water productivity (more crop per drop). The extent of power or energy scarcity is vital because, if it is severe, resulting in a prolonged lack of water, there is an increasing risk for yield loss. As AWD prolongs the irrigation interval, the risk of crop losses will be higher when the power supply fails at the time whencrops need water. Because of this perceived risk, farmers tend to use more water than what is actually needed during their irrigation schedule when electricity is available. Also, when water is available in the canal, farmers usually take advantage to irrigate their crops even if the field is still not due for irrigation. Hence, to fully practice AWD in Bangladesh, water and energy are jointly indispensable. That is, the more reliable the water and energy supply is, the more likely farmers will adopt AWD.

306

Structural issues within the irrigation system may also influence AWD adoption. Bangladesh has five types of irrigation system: the traditional or local method, the canal irrigation project of the government, low-lift pumps, and shallow and deep tube wells (Chowdhury 2012). The low-lift pump is used to pump water from surface-water sources while shallow or deep tube wells are used for groundwater. These minor irrigation devices or equipment are operated byeither diesel or electricity.

Each irrigation system can vary in terms of the organization of water users, regulations, irrigation schedules, and payment arrangements. Most deep tube wells are government owned and maintained by the public authority; others are run by a cooperative or joint ownership. DTWs irrigate a relatively larger area (up to 25 ha) than STWs (1−2 ha), and involve 100−200 farmers, requiring more coordination than STW farmer users (Kürschner et al 2010). Because of the high investment in DTWs aside from the high transaction costs in system management, and the scarcity of surface water in rivers and canals, farmers prefer using shallow tube wells. Shallow tube wells and power pumps accounted for 71% of the total irrigated area in 2000 (Hossain et al 2007).

In northwest Bangladesh, the study area of Kürschner et al (2010), farmers are using two major groundwater irrigation systems: deep tubewells (DTW) and shallow tube wells (STW). The DTWs are either state-owned or privately owned. The state-owned DTWs are managed by the BMDA. The privately owned DTWs are managed by 15 to 45 farmers. In both cases, an operator is appointed to operate the pump, and the irrigation schedule is arranged prior to the start of theboro season. Distribution of water in DTW systems is normally based on a rotational schedule, in which farmers’ fields are divided into irrigation blocks. Most irrigation schedules are fixed and usually decided by pump operators, and are adjusted frequently according to power availability.

In the Rajshahi area, 57% of the farmers do not have control or influence on the timing of irrigation. With farmers having less influence on their own actual irrigation schedules, implementation of AWD may be limited, especially if only a few and not all farmers within a block intend to adopt the technology. Hence, for successful AWD implementation, a coordinated approach between farmers and pump owners/operators for water scheduling is needed in which both will mutually benefit. And, this requires a collective voice and collective action from farmers to have strong bargaining power with pump owners. At present, however, farmers are not formally organized into groups,especially those using STWs. Considering this situation (community organization to bring about collective action), the role of pump owners and operators is essential in AWD implementation.

Since the early 1980s, the market for irrigation technologies has been largely liberalized and privatized, resulting in the expansion of STW irrigation. Since then, a competitive market for irrigation water has evolved. The water payment varies from one area to another along with the type of well within a particular area (Biswas and Mandal 1993). Farmers are paying an equivalent of 25% to 30% of their rice harvest for irrigation (Sattar et al 2009). In recent years, though, the market has moved toward fees on a per hour basis of tube well operation (Chowdhury 2012).

The payment arrangement for irrigation services is crucial for influencing farmer AWD adoption. It establishes motivationfor farmers and pump owners to adopt AWD. It

307

determines whether economic benefits are transferred to farmers, remain with pump owners alone, or both are mutually benefited.

In the study area of Kürschner et al (2010), two main payment forms were normally practiced: seasonal fixed and consumption-based rates. A majority of the payment arrangements, however, are based on seasonal fixed rates, which included (1) a seasonal payment arrangement per STW irrigation block where fuel is provided by the pump owner; (2) fixed rates per season in DTWs with rates previously decided in meetings; and (3) a seasonal fixed rate set by the pump owner and where farmers buy their own fuel. On the other hand, consumption-based rates included (1) rentals of STWs to farmers on an hourly basis and (2) the use of prepaid cards based on hourly consumption (as implemented by BMDA), withsome extra charges by pump operators. Payments in kind were uncommon and found only a few times in Kushtia District.

Unless the payment arrangement would benefit farmers significantly, AWD is lesslikely to be adopted on a larger scale by farmers. The consumption-based STW system can directly provide benefits to farmers through a reduction in fuel costs and pump lending fee (Table 2; Kürschner et al 2010). This may lead, though, to areduction in income for pump owners but this can be compensated when the pumps are rented out to more farmers. Similarly, the prepaid card system in DTWs of BMDA has great potential to directly provide benefits to farmers. The prepaid cards, however, were introduced to only a limited number of DTWs at the time of interviews.

In terms of fixed-rate payment schemes, the scheme guarantees pump owners a secure and regular income (Table 2; Kürschner et al 2010). By default, only pump owners would benefit from this setup. However, in the fixed-rate system in which farmers provide their own fuel, both parties will benefit from the technology. Farmers can reduce the fuel cost, and, since pumps will be used less intensively, maintenance cost by pump owners may also decrease.

Likewise, the previous studies showed that the probability of farmers to adopt AWD is higher if the irrigation payment arrangement is through cash and if farmers have a larger irrigated area. Farmers use less water in the cash payment system because they have to operate STWs with their own diesel and a large farmer with a large irrigated area has more incentive to apply AWD because he has to irrigate all his land with his own diesel. For the STW owners, though the profit is less, the transaction cost is largely reduced. Evidence of socioeconomic impacts

Considering that AWD adoption is still at the initial stage, only some short-term socioeconomic impacts were being assessed. Partial economic analysis showed that AWD is truly a beneficial technology for farmers (Kürschner et al 2010). Farmers were able to decrease the total number of irrigations by 28%, which translated to irrigation costs decreasing by about 20%. However, monetary profit from reduced irrigation frequency was observed only when payment was based on a consumption basis. Also, weeds can be one of the major problems with AWD, especially if they are not wellcontrolled during the early stages of rice crop growth (by mechanical weeding, herbicides, or flood water). This results in an increased labor cost for hand weeding. In

308

terms of yield, farmers reported a 0.4−0.5 t ha-1 yield increase. Farmers claimed that AWD provided more tillers and panicles, and plants were healthier in general.

The potential social impact of AWD is in the reduction in potential conflicts over water issues in larger irrigation systems. However, many structural problems (power shortages, groundwater shortages, and social power asymmetries) still need to be addressed. Strong collaboration and collective actions among various stakeholders are imperative to manage and implement AWD successfully in a larger scheme, and to ensure benefits in terms of water savings and economic returns (Kürschner et al 2010).

Conclusions

The adoption of AWD in Bangladesh is still at the early stage. However, severalorganizations in Bangladesh have committed to disseminating the technology within their geographic mandate. Field validations were already successfully done by various organizations, and data and farmers’ feedback from these validations were used to support their dissemination efforts. Considering that AWD is knowledge-intensive, it seemed that most organizations were limited by the capability of well-trained extension staff. This was partly attributed to inadequate training received from their organizations. As a consequence, quality training for farmers is lacking and transfer of knowledge is insufficient. There is substantial demand for AWD by farmers, but actual adoption depends on many factors, either internal or external, that influence irrigation systems. Frequent power shortage was one of the major stumbling blocks to adoption, as this obstructs farmers from practicing in-time irrigation. Irrigation payment schemes and the mode of irrigation delivery to farmers also greatly influenced adoption. Fixed-rate arrangements discouraged farmers from adopting AWD since irrigation charges are infixed amounts, which have already been decided prior to the start of the season. In addition, most pump owners are still not passing on the economic benefits of AWD to the farmers who adopted the technology.

Social changes at the level of farmers could not be observed yet, as wide-scale adoption has not yet occurred. Economic benefits, however, were already perceived by farmers. Among the advantages that AWD brought to farmers were the reduction in the number of irrigations by 28%, the drop in irrigation costs (fuel/electricity) by 20%, and the increase in yield by 0.4−0.5 t ha−1. However, given the high demand and potential that the technology offers, the assessment of the dissemination and adoption of AWD in Bangladesh has not been fully realized. Using the potential in the near future is a challenging task, as further spread of the technology depends on actions taken and efforts to be made at the organizational level to improve and institutionalize the dissemination process in the country.

The development of AWD in Bangladesh demonstrated the importance of external and local catalysts, the role of local champions, and the engagement of various stakeholders. IRRI was the external catalyst, while BRRI was the local catalyst. Along the way, various local champions were identified and some key stakeholders became partners in disseminating the technology altogether. Both IRRI and BRRI played considerable roles in promoting AWD in Bangladesh nationally by hosting workshops to

309

inform policymakers, and conduct training to build the capacity of the staff of various public and private stakeholders.

Capitalizing on the momentum of the initial dissemination efforts by various disseminating organizations through field demonstrations and training during the boro season, the following strategies will be followed: (i) Strengthening partnerships among stakeholders This involves organizing different stakeholders, facilitating the exchange of ideas and experiences, and recording the commitment of stakeholders. Regional meetings with stakeholders before and after the boro season will be conducted. An annual review and planning workshop with policymakers and stakeholders for participatory decision-making will also be organized. Currently, initial steps have been taken to incorporate AWD into policies and structures at the national level, but whether these efforts will change the way irrigation is being organized remains to be seen. A strong "national alliance of stakeholders" for scaling up of the technology will be formed for sustainability. (ii) Strengthening farmer organization Unless farmers are well organized, theywill continue to negotiate with pump owners individually. A strong farmer organization or group (irrigation block) could generate bargaining power with pump owners through a collective voice to bring about a win-win situation. Further, a cohesive farmer group could diffuse conflict, that is, water distribution and timing, and implement an efficient irrigation management system. (iii) Training and mentoring field staff/stakeholders in the selected location, and

influencing policy change at the local institutional level Field staff of the disseminating organizations, who trained the farmers, played an important role in spreading AWD locally. However, field staff seemed to be limited in their capacity and resources to intensify the training activities for a greater number of farmers. There is a need to enhance the ability and capacity of the field staff of the disseminating organizations to strengthen the quality of training given to farmers. Field staff also needs to be trained and mentored on the aspects of monitoring and evaluation so that data will be available to influence policymakers (i.e., the system of using electricity) within the local administration and the power development board. For long-term adoption, a certain level of water and energy security is required, especially becausethe chronically unreliable electricity supply affects adoption negatively as AWD requires farmers to receive water in time when crops need it. (iv) Developing incentive mechanisms: exploring the potential of a “metering

system” in deep tubewells The payment system for pump irrigation services plays a crucial role in the wide-scale adoption of safe AWD as it determines whether economic incentives through reduced pumping costs are transferred to farmers or not. The BMDA has pioneered an example of a consumption-based rate system by providing a group of farmers with prepaid cards as the metering device to operate the pumps in Rajshahi Division. More individual AWD farmers will benefit if this system is successfully implemented on a larger scale. However, changing payment schemes for irrigation towardthe more consumption-based form requires a participative negotiation process between farmers and the people incharge of irrigation management. Adaptive research is needed to assess the potential of the metering device and its compatibility in local-level settings.

310

References Alam MS, Islam MS, Salam MA, Islam MA. 2009. Economics of alternate wetting and

drying method of irrigation: evidences from farm level study. In: The Agriculturist. 7(1 and 2):82-89.

BBS (Bangladesh Bureau of Statistics). 2008. Yearbook of agricultural statistics of Bangladesh, 2006.Ministry of Planning, Government of the People's Republic of Bangladesh. Dhaka, Bangladesh.

Biswas MR, Mandal MAS. 1993. Irrigation management for crop diversification inBangladesh. Dhaka (Bangladesh): University Press Limited.

Chowdhury NT. 2012. Irrigation institutions of Bangladesh: some lessons, problems, perspectives and challenges of agricultural water management. In: Kumar M, editor. In: Tech. Available from www.intechopen.com/books/problems-perspectives-and-challenges-of-agricultural-water-management/irrigation-institutions-of-Bangladesh-some-lessons.

Chowdhury UK, Biswas BK, Chowdhury TR,Samanta G, Mandal BK, Basu GC, Chanda CR, Lodh D, Saha KC, Mukherjee SK, Roy S, Kabir S, Quamruzzaman Q, Chakraborti D. 2000. Ground water arsenic contamination in Bangladesh and West Bengal, India. Environ. Health Persp. 108(5), May 2000.

Fujita K. 2004. Transformation of groundwater market in Bengal: implications to efficiency and income distribution. Centre for Southeast Asian Studies, Kyoto University, Japan. Available at www.sasnet.lu.se/EASASpapers/8KoichiFujita.pdf.Accessed 1 Feb. 2012.

Hossain I, Rahman S, Mannan A. 2009. AWD technology at BarindMultipurpose Development Authority. In: Sattar MA, editor. National Workshop Proceedings on AWD Technology for Rice Production in Bangladesh, 15 July 2009. Gazipur (Bangladesh): Bangladesh Rice Research Institute.

Hossain M, Lewis D, Bose M, Chowdhury A. 2007. Rice research, technological progress and poverty: the Bangladesh case in agricultural research, livelihoods, and poverty. Studies of economic and social impacts in six countries. Baltimore, Md. (USA): The Johns Hopkins University Press.

Hossain M. 2009. The impact of shallow tube wells and boro rice on food security in Bangladesh. IFPRI Discussion Paper 917. Washington, D.C: International Food Policy Research Institute.

Kabir L. 2009. Benefit-cost analysis of demonstration on AWD irrigation method. In: Sattar MA, editor. National Workshop Proceedings on AWD Technology for Rice Production in Bangladesh, 15 July 2009. Dhaka (Bangladesh): International Rice Research Institute.

Kürschner E, Henschel C, Hildebrandt T, Jülich E, Leineweber M, Paul C. 2010. Water saving in rice production: dissemination, adoption and short term impacts of alternate wetting and drying (AWD) in Bangladesh. SLE Postgraduate Studies on International Cooperation,study commissioned by the Advisory Service on Agricultural Research for Development of German Technical Cooperation (GTZ-

311

BEAF) in collaboration with the International Rice Research Institute (IRRI). Humboldt University, Berlin, Germany: SLE Publication Series S241.

Miah H, Sattar MA,Tuong TP. 2009b. Role of alternate wetting and drying technology in resource conservation for rice cultivation in Bangladesh. Paper presented at the 4th World Congress on Conservation Agriculture. New Delhi.

Miah H, Sattar MA, Tuong TP. 2009a. Water saving irrigation practices in rice-based canal system. Unpublished paper. International Rice Research Institute. Dhaka, Bangladesh.

Miah H. 2008. Practicing AWD technology for saving irrigation water in winter rice (boro rice) in Bangladesh. Paper presented at the annual review and planning of the ADB-funded project “Development and Dissemination of Water-saving Rice Technologies in South Asia,” 19-21 April 2008, at the Central Rice Research Institute, Cuttack. India.

Palis FG, Cenas PAA, Bouman BAM, Hossain M, Lampayan RM, Lactaoen AT, Norte TM, Vicmundo VR, Castillo GT. 2004. Farmer adoption of controlled irrigation in rice: a case study in Canarem, Victoria, Tarlac. Philipp. J. Crop Sci. 29(3):3-12.

Palis FG, Singleton GR, Brown PR, Huan NH, Nga NTD. 2010. Socio-cultural factors influencing the adoption of ecologically based rodent pest management. In: Singleton GR, Belmain SR, Brown PR, Hardy B, editors. Rodent outbreaks: ecology and impacts. Los Baños (Philippines): International Rice Research Institute. p 153-170.

Palis FG, Singleton GR, Brown PR, Huan NH, Umali C, Nga NTD. 2011. Can humans outsmart rodents? Learning to work collectively and strategically. Wildlife Res. 38:568-578.

Sattar MA, Rashid MA, Hassan MN, Molla HR, Khan AK, Parveen S, Roy D, Mahamud H. 2009. AWD technology for water saving in boro rice production for selected locations. In: National Workshop Proceedings on AWD technology for rice production in Bangladesh. In: Sattar MA, Maniruzzaman M, Kashem MA, editors. Collaboration with Bangladesh Rice Research Institute, International Rice Research Institute, and Krishi Gobeshona Foundation, 15 July 2009. Gazipur (Bangladesh): Bangladesh Rice Research Institute.

Notes Authors’ addresses: F.G. Palis, R.M. Lampayan, and B. Bouman, International Rice Research Institute, Los Baños, Philippines; E. Kürschner, team leader, SLE, Humboldt University, Berlin, Germany.

Figure 1. Percent of farmers and pump owners reporting advantages of AWD. Kürschner et al 2010.

312

Table 1. Key organizations contributing to AWD dissemination.

BRRI DAE BMDA RDRS Syngenta Sector Government

organization Government organization

Government organization

NGO Private

Core function Rice research

Agricultural extension

Infrastructural projects and development

Rural development to reduce poverty

Supplier of agricultural products

Target group Mostly small-scale farmers and some pump owners

Small- and medium-scale farmers/pump owners

Innovative pump operators, farmer leaders, members of irrigation association

Innovative medium- scale farmers (70% women)

Innovative farmers with at least 2−3 acres of land; Syngenta clients

Motivation Research agenda, validation and training

Improving irrigation management

Improving deep-well irrigation system management

For participatory development projects

Means to strengthen customer relation-ships

Geographic focus

Nationwide —9 research stations; many are in northwest Bangladesh

Nationwide Barind Tract (i.e., Rajshahi and Rangpur divisions)

Rangpur Division

Nationwide

Scope of dissemination

900 farmers in farmers’ fields at 9 research stations. Since 2007-09

528 demo plots (NATP); 26,880 trained pump owners (AETEP). Since 2007-08

86 demo plots, 25,000 trained farmers and pump owners. Since 2007-08

325 demo plots through 13 partner NGOs. Since 2009-10

50,000 panipipes distributed to farmers in 6 districts. Since 2008-09

Source: Adapted from Kürschner et al (2010), Sattar et al (2009), Kabir (2009).

313

Table 2. Economic benefits for farmers and pump owners when adopting AWD by irrigation payment schemes.

Payment scheme Farmers’ benefit Pump owners’ benefit STW pump rented, hourly basis

Save fuel cost and pump lending fee

Less income (unless renting out the pump to more farmers)

DTW prepaid, hourly basis

Save irrigation cost Pump operator receives less income

Fixed seasonal rate for STW and DTW

No direct benefits/savings

Save fuel/electricity. Increases net return (selling less water at the same price)

Fixed rate plus own fuel for STW

Save fuel cost Unchanged income, but reduced use of pump can prolong pump’s life

Source: Kürschner et al (2010).

314

Figure 1. Percent of farmers and pump owners reporting advantages of AWD. Kürschner et al (2010).

Paper 13

0

10

20

30

40

50

60

70

80

90

100

Saves water Lessirrigation

costs

More yield Saves fuel Saveselectricity

Measurewater level

others

Farmers

Pump owners

315

Farmers’ participatory research and adoption of aerobic rice in the Philippines

Rubenito M. Lampayan, Florencia G. Palis, Junel B. Soriano, and B.A.M. Bouman

Aerobic rice is a technology that uses less water than irrigated rice. As the global problem of declining water supply continues, IRRI took the lead in research on and development of aerobic rice in the Philippines in 2001. Farmer participatory research and development and adoption surveys were some of the initial activities undertaken in the Philippines. Preliminary testing of three promising aerobic rice varieties in Tarlac and Nueva Ecija showed that, in farmers’ field conditions, Apo has the highest yield (4.0 to 5.5 t ha−1) among the three varieties. However, further testing of Apo in Bulacan revealed a minimum yield of 2.0 t ha−1. One breeding line, IR74371-54-1-1, was highly preferred by farmers and it was released in the Philippines in 2009 as Sahod Ulan 1, its local name, withNSIC Rc 192 (IRRI 148) as its designation. This line yielded 5.26 t ha−1 in Bulacan, and is now adopted by farmers in Bulacan and La Union provinces. A case study conducted in Bulacan revealed that, initially, farmers’ interest in adopting was highly driven by the perceived advantage in crop establishment and less water usage of aerobic rice. When gaining more experience and knowledge, farmers consider specific criteria such as marketability and follow a certain decision tree in deciding whether to adopt or not adopt any new technology. Results of socioeconomic assessment of 80 farmers consisting of adopters and nonadopters in Bulacan revealed that the yield of aerobic rice may be lower than that of lowland rice, but it has comparable profitability. Hence, aerobic rice production can be a good alternative for farmers in rainfed and water-short areas but not a substitute for flooded rice. Location-specific varieties that take into consideration water availability and existing high-/low-input management practices are needed for aerobic rice technology development and adoption.

Around 60% of the 3.4 million hectares of rice are under irrigation in the Philippines, with the majority of the production coming from Central Luzon,the country’s rice bowl (IRRI 2002). About 130,000 ha of rice land in Central Luzon is irrigated by the reservoir-backed gravity irrigation system known as the Upper Pampanga River Integrated Irrigation System (UPRIIS). Besides UPRIIS, shallow tube wells (STWs) and deep-well pumps (DWPs) are commonly found in this area. However, in the last decade, the growing scarcity of water globally has become a serious problem (Bouman et al 2007). Given that the availability of water for irrigation is declining, the sustainability ofrice production in the irrigated lowland has now been threatened. As water scarcity becomes a serious problem, researchers have been looking for ways to reduce water use in rice production but at the same time increase crop efficiency.

Aerobic rice is a production system in which modern input-responsive rice varieties are grown under nonflooded, nonpuddled, and nonsaturated soil conditions (Bouman and Tuong 2001). Since aerobic rice needs less water than conventional flooded lowland rice, the system of aerobic rice is an attractive option for water-short irrigated or rainfed lowland environments (Bouman et al 2005, Atlin et al 2006). To achieve high yield under aerobic conditions, new varieties with combined drought-resistance characteristics of upland varieties and high-yielding characteristics of lowland varieties are required. Evidence of a feasibility study from northern China showed that

316

breeders’ first-generation (temperate) aerobic rice varieties used only 50% of the water requirement in lowland rice. This aerobic rice technology is considered sufficiently mature for dissemination in the Yellow River basin. However, more research and development (R&D) are still needed for most of tropical Asia to create sustainable and high-yielding aerobic rice systems (Bouman 2008). Likewise, according to Bouman (2008), aerobic rice should be recognized as a special crop type besides “lowland rice and upland rice” to facilitate the collection of statistics on the adoption and diffusion of aerobic rice, and to encourage its wide dissemination by extension agencies.

The research on and development of aerobic rice in the Philippines started in 2001. IRRI led strategic research to develop sustainable aerobic rice systems for water-scarce irrigated and rainfed environments through the Swiss Development Cooperation-funded Irrigated Rice Research Consortium (IRRC) Phases 2 to 4 and the Challenge Program on Water and Food (CPWF) on developing a System of Temperate and Tropical Aerobic Rice in Asia (STAR). From 2001 to 2009, strategic research and development activities were undertaken to (1) identify and develop aerobic rice varieties with high yield potential; (2) understand crop management strategies, including water, nutrient, and weed management; (3) identify key sustainability issues; and (4) develop practical technologies for crop establishment. Pot and field experiments and on-farm farmer participatory trials were employed to test together with farmers whether the technology worked. These on-farm farmerparticipatory trials were conducted in various locations in Central Luzon. Both quantitative (farmer surveys) and qualitative methods (focus group discussions and key informant interviews) were conducted to capture farmers’ feedback, experiences, and perceptions about aerobic rice, as well as to understand the biophysical and socioeconomic factors that influence farmer adoption in the Philippines.

Field experiments were performed both at the IRRI experiment station and in Central Luzon. The findings were reported in graduate theses (Grassi 2006, Vermeulen 2007, Soriano 2008, Acosta 2009, Gunarathna 2010), in conference proceedings (Lampayan et al 2005, Soriano et al 2006, Ganotisi et al 2009), and published in refereed journals (Bouman et al 2005, Nie et al 2008, Kreye et al 2009a,b, Lampayan et al 2010). Table 1 shows the range of activities conducted from 2002 to 2008 in the Philippines. This chapter focuses on on-farm farmer participatory research and development and farmer adoption surveys. Participatory research on and development of aerobic rice

Farmer participatory research and development have been carried out by the water-saving workgroup of the IRRC since 2002. Preliminary testing of three promising aerobic rice varieties was conducted during the 2002 wet season (WS), 2003WS, and 2003 dry season (DS) in several farmers’ fields in deep-well pump irrigation systems in Tarlac (six farmers’ fields) and shallow tube-well systems in Nueva Ecija (three farmers’ fields). These varieties were PSB Rc 9 (Apo), UPLRi-5, and Magat. Apo and UPLRi-5 were inbred varieties, while Magat was a hybrid. Farmers’ field sizes planted with aerobic rice in Tarlac and Nueva Ecija ranged from 700 to 2,200 m2. The main objectives of the trials

317

were to explore the performance of these varieties under aerobic field conditions. Farmers and project researchers worked closely together, especially during the dry season’s land preparation, crop establishment, and crop management. Seeds were sown (direct seeding) in rows 25 cm apart, about 2 cm deep, in relatively dry soil at 80 to 100 kg ha−1 seeding rate. The establishment of furrows for row seeding was facilitated using a traditional implement known as a “lithao,” after which the seeds in the furrows were covered. After seeding, a light irrigation of 2 cm was applied to the field to facilitate seed germination. For fertilizer application, a total of 160 kg N ha−1 was applied in three splits (50% at basal, 25% at maximum tillering, and 25% at panicle initiation) to ensure that N fertilizer would not be a limiting factor in aerobic rice crop performancein farmers’ field conditions. Previous reports indicated that irrigated lowland rice farmers in Central Luzon used high fertilizer N rates up to 180 kg ha−1 during the dry season to achieve high yield (PhilRice 2000). Relatively higher basal N application (50% of the total N) was applied to ensure good N availability in the soil for early crop use, although this had resulted in inefficient use of N fertilizer as the rice plant doesn’t require high doses of N at the early stage of crop growth. Basal application of phosphorus (30 kg P ha−1), potassium (30 kg K ha−1), zinc sulfate (20 kg ZnSO4 ha−1), and iron sulfate (20 kg FeSO4 ha−1) were also given to the rice plants. Weed management was done using early postemergence herbicides (Nominee), and two to three hand weedings. Mechanical weeding was also carried out using thelithao (Figure 1). Crop-cut samples from the aerobic rice fields were collected and measured for yield comparison between varieties.

In the 2002WS, under farmers’ field conditions, variety Apo had the statistically highest yield (5.5 t ha−1), followed by Magat (5.0 t ha−1) and UPLRI-5 (4.7 t ha−1) in Tarlac. In spite of its relatively higher seed cost, Magat was known to be a drought-tolerant hybrid and farmers were interested in trying it in their own fields. In Nueva Ecija, UPLRI-5 had higher yield (4.5 t ha−1) than Apo (4.1 t ha−1) because of lodging of Apo during the grain-filling stage (Lampayan et al 2003). The high fertilizer N dose of 160 kg, which produced taller plants and higher biomass of Apo, was responsible for lodging (Lampayan et al 2003). Generally, aerobic rice yields were responsiveto applied nitrogen, but the risk of lodging also increased, especially atN rates beyond 90 kg ha−1 (Lampayan et al 2010). Magat was not planted in Nueva Ecija during the 2002WS. The total amount of rainfall from seeding to harvesting at both sites was about 1,500 mm, with more than 50% of the total rainfall occurring from the last week of June to the middle of July. With sufficient rainfall, no irrigation was applied during this season.

In the 2003DS, testing of promising aerobic rice varieties was continued but under supplementary irrigation. Two establishment techniques were also evaluated with 20 farmer-participants in Tarlac Province. Each farmer-participant tested one of the three promising varieties, sowing them under the aerobic rice technique using either “low-tech” or “high-tech” implements or both. The “low-tech” referred to the use of a lithao in creating furrows for seeding. Seeds were then hand sown, and basal fertilizer was manually broadcast. The “high-tech” referred to the use of a mechanical seeder that was pulled by a big tractor for direct seeding and direct placement of basal fertilizer N (Figure 2). A similar fertilizer application rate was recommended in both establishment methods. Again, a high dose of N (total 200 kg ha−1) was applied in four splits (80 kg ha−1 basal; 40 kg ha−1 at 20 to 30 DAS; 40 kg ha−1 at maximum tillering; 40 kg ha−1 at panicle

318

initiation). Basal application of P (30 kg ha−1), K (30 kg ha−1), and ZnSO4 (20 kg ha−1) was also provided. Most of the farmer-participants tried both “low-tech” and “high-tech” by either contributing two plots or splitting one big plot into two subplots. Plot sizes used for each of the establishment techniques ranged from 920 to 1,178 m2. Both “low-tech” and “high-tech” fields were laser-leveled before seeding. To facilitate aerobic conditions, only light irrigation was applied each week according to deep-well pumps’ rotational delivery schedule. Farmers were asked to record the date and duration of each irrigation. However, the actual amount of irrigation was not measured in the trials. Yield data were collected (Figure 3) from farmers’ interviews and from crop-cut samples obtained from two 10-m2 sampling spots in each field. To compare yields from aerobic rice (“low-tech” and “high-tech”) with those of farmers’ practices, neighboring farmers (noncooperators) were interviewed and rice yield was obtained from them. Results showed that “low-tech” fields (from farmers’ interviews) had an average yield of 2.4 t ha−1, ranging from 0.5 to 5.7 t ha−1. The crop-cut calculations had higher yield estimates, ranging from 1.7 to 5.9 t ha−1, with an average yield of 3.5 t ha−1. Under “high-tech,” average yield from farmers’ interviewswas about 2.3 t ha−1, and about 4.1 t ha−1 using crop-cut yield data (Figure 3). Paired comparisons of the two yield estimates (crop-cut data vs. farmers’ interview data) indicated nonsignificant differences using a t-test. The difference in yield between “low-tech” and “high-tech” was also nonsignificant using a t-test. In terms of yield performance by variety, the average yield for Apo and UPLRi-5 was lower (2.6 t ha−1 and 2.1 t ha−1) than from the crop-cut estimate (4.0 t ha−1 and 3.4 t ha−1), respectively. Reported yield data from neighboring farmers’ fields (data not shown) ranged from 2.2 to 3.5 t ha−1 (2.6 t ha−1 on average), which was comparable with that of aerobic rice farmers’ reported yield. For both “low-tech” and “high-tech,” light irrigations (about 3- to 5-cm depth per irrigation) were applied 14 times. The same frequency of irrigation was also recorded in neighboring farmers’ fields, but higher depths of irrigation water (8 to 10 cm) were applied in their fields at each irrigation application (data not shown). This means that aerobic rice farmers saved about 50% of irrigation compared with nonaerobic rice farmers in the area.

While aerobic rice trials were going on, visits to the research-extension sites and briefings to demonstrate aerobic rice were conducted. These activities also involved nonparticipating farmers in the area. Although farmers were enthusiastic in adopting aerobic rice during the project implementation, some salient issues have emerged, such as water stress, weed pressure, possible yield reduction due to incidence of nematodes under aerobic conditions, and nutrient management (Lampayan et al 2003). There is a need to investigate these issues before a wide-scale adoption can be achieved. Dissemination and promotion From Tarlac and Nueva Ecija, aerobic rice technology was further validated and disseminated in Bulacan, another province of Central Luzon, which has predominantly rainfed conditions. IRRI and the Bulacan Agricultural State College (BASC) started working with seven farmer-participants in San Ildefonso, Bulacan, in the 2004 WS, and 13 farmers in 2005 until the 2008WS. Variety Apo was also demonstrated to some farmers in five neighboring towns (Norzagaray, San Miguel, Angat, San Rafael, and

319

Doña Remedios Trinidad). The average yield in the 2007WS ranged from 3.0 to 6.3 t ha−1 (adjusted to 14% MC), while it was 2.0 to 5.0 t ha−1 in the 2008DS (Soriano et al 2008). Similar ranges of yield were reported in the 2005WS (2.4 to 6.8 t ha−1), 2006DS (2.4to 6.3 t ha−1), and 2006WS (2.1 to 6.9 t ha−1). The relatively low yield in the 2008DS was due to rodents and insect problems (i.e., stemborers) since most neighboring farmers’ fields were under fallow conditions.

Through the support of IRRI, BASC established partnerships primarily with state universities and colleges (SUCs) and local government units (LGUs) in the promotion and development of aerobic rice within and outside Bulacan Province. Demonstration farms were established in nine provinces in the country in collaboration with SUCs and LGUs in the respective provinces, and Department of Agriculture Regional Field Units (DA-RFUs). Seminars and training activities were also conducted before the establishment of demonstration farms.Farmers’ field days was organized at harvest, which were attended by local agricultural technicians, farmers, and staff of partner SUCs. As shown in Table 2, from 2006 to 2010, 271 demonstration farmswere established, 240 technicians/researchers and 2,227 farmers were trained, and about 1,250 farmers estimated to be adopters of aerobic rice in the country were identified (Table 2; Soriano 2011). Farmers’ testimonies revealed some advantages of aerobic rice over other varieties, thus convincing them to adopt it. Among the noted advantages were low input requirement, manageability for growing (less labor requirement), competitive yield, higher income, early establishment (improves the farming system and productivity), high resistance to pests and diseases, and weedcompetitiveness. One of the strategies undertaken to promote and disseminate aerobic rice technology was the production and distribution of a technology guide (Figure 4) to farmers and technicians. Local and national media (articles in daily newspapers and TV programs) were also tapped in the promotion (Soriano 2010).

Farmer participatory varietal selection (PVS) trials were also conducted from 2004 to 2006 to evaluate cultivars for aerobic conditions. IRRI provided about nine “potential” cultivars and these were sown under the aerobic rice system and evaluated on BASC farms. Selection of potential varieties was carried out by a group of farmers using their own set of criteria at the crop ripening stage. After field evaluation, farmers gathered together and discussed the results obtained. For example, in San Ildefonso, Bulacan, during the 2004DS, Apo was most preferred by farmers among the varieties tested (Table 3). Although Apo’s computed yield was not the highest, most farmers still preferred this variety because of its good stand, long panicles that resulted in more grains, and its resistance to drought.Apo (IRRI 132) was released in the Philippines in 2001 as NSIC Rc 9 for the rainfed ecosystem. Aside from Apo, farmers preferred the early-maturing lineIR74371-54-1-1. This breeding line was officially released in 2009 as NSIC Rc192 (IRRI 148), with local name Sahod Ulan1, and it is currently widely planted by farmers aside from Apo in La Union and Bulacan provinces. Adoption studies for aerobic rice Two case studies were conducted between 2004 and 2007 in Bulacan.The first case study documented farmers’ perceptions and the adoption process (Flor et al 2007) while

320

the second case study assessed the socioeconomic benefits of aerobic rice (Quicho et al 2008). (i) Farmers’ perceptions and adoption study This case study followed the cooperators and other farmers in one village (Pala-pala) in San Ildefonso, Bulacan, who tried the technology from 2004 to the 2006WS. The study was carried out between 2005 and 2007 through focus group discussions, participants’ observations, and key informant interviews with farmers. A knowledge, attitudes, and practices (KAP) survey was also implemented in 2005 for 29 farmers who planted aerobic rice (particularly Apo) in the village.

To see how aerobic rice was accepted by farmers through the years, trends in the number of farmers who planted it and the area in which it was planted were observed. It started with only five cooperators in the 2004WS (other cooperators were not in the same village). Then, it spontaneously increased to 29 farmers in the 2005WS, but decreased to 10 farmers in the 2006WS (Flor et al 2007). This number, however, may not be absolute as there may have been more farmers planting aerobic rice in the 2006WS but they were not included in the study since they were recent adopters. In terms of size of the area planted with aerobic rice, the average area cultivated by each farmer increased from 0.32 ha to 2.6 ha in the 2005 WS after getting good results. However, cultivated area declined as the number of farmers planting aerobic rice dwindled in the 2006WS.

After the first season of farmer participatory trials of aerobic rice, many farmers within the village became interested in it. A number of positive and negative attributes of aerobic rice were perceived by farmers in the village (Table 4). What farmers liked most about aerobic rice were that it did not need a lot of water or irrigation and it was drought tolerant. This was very important because Pala-pala Village is a rainfed area and only a very few farmers have supplementary irrigation sources. Some farmers who were forced to irrigate two to three times through pumps even said that their irrigation frequency was cut back to about half the number of irrigations they used to apply before adopting aerobic rice. Farmers also noticed that aerobic rice was responsive to fertilizer inputs, requiring less fertilizer. With approximately 50 to 60 kg ha−1 of applied N, yields of more than 4.0 t ha−1 were achieved. During crop growth, aerobic rice was said to have good vigor and the ability to grow tall, making the crop competitive with weeds. Resistance to insect attacks was also a positive attribute of aerobic rice that refrained some farmers from using insecticides for the entire season. Other positive attributes perceived by farmers were shorter crop duration, good yield under drought conditions, good milling recovery, and less labor cost. Shorter crop duration was valuable to farmers who wanted to plant more than one crop a year. Farmers also said that although aerobic rice did not get as much water, it still produced yield comparable with that of their other varieties. Furthermore, they liked aerobic rice because of its higher grain weight and the grains have good milling recovery. Overall, some farmers said that using aerobic rice incurred lesser cost because of the lower labor requirement in crop establishment and irrigation. On the other hand, farmers also saw negative characteristics of aerobic rice technology. Although aerobic rice has a good trait of being responsive to fertilizer, recommended fertilizer application may be too high and may cause plants to lodge, thus affecting yield. In addition, the panicles of Apo were exposed such that, at ripening stage, birds could

321

easily make them their target. Moreover, Apo had short and bold grains and traders may consider this “ordinary rice” when milled and that fetches a lower price in the market.

Several conflicting opinions and perceptions also emergedon weed problems and eating quality. Some farmers had problems with weeds because of the direct seeding method of establishment. These farmers had to invest in herbicides and labor for weeding; however, the farmers also said that the plants could compete with weeds even if no herbicides were applied. In terms of eating quality, some said that Apo was not good, but others said that, when cooked, the rice was soft and had a good taste and smell. These conflicting perceptions show how the different experiences affect the way that farmers respond to aerobic rice initially. However, this was also recognized by the farmers themselves. They knew that there was still more to know about aerobic rice technology and the main issue of water availability still existed and affected them enough so that they wanted to gain more experience with aerobic rice to see if this could be a technology to address their water concerns in rice production.

After three trial seasons, most of the farmers had tried aerobic rice and planted it in the following cropping season in a different area in their rice field. Farmers believed that planting it in the area might make the plants vulnerable to diseases. This group of farmers had similar perceptions as those of farmers interviewed in 2005 as presented in Table 4. However, easiness in crop establishment through dry direct seeding was an additional characteristic that encouraged farmers to continue planting aerobic rice. After more experience with the technology and specifically in planting with Apo, farmers decided to adopt or not adopt aerobic rice based on specific criteria, including extension pathways, economic advantages such as yield and marketability, fitness of the technology with the farmers’ cropping pattern, and other farmer-specific factors (Flor et al 2007). The extension efforts by BASC and observing the aerobic rice grown by fellow farmers who tried the technology were very relevant to farmers’ decision to try aerobic rice. Higher yield and marketability were very important considerationsfor most farmers, but the idea of subsistence value of aerobic rice was also equally important. Although farmers planted aerobic rice to sell, they were also using it for household consumption. (ii) Socioeconomic assessment Another study was carried out from 2004 to 2006 to assess the economic advantage of aerobic rice compared with conventional transplanted lowland rice cultivation in Bulacan. A total of 80 farmers from the towns of San Ildefonso, San Miguel, and San Rafael were interviewed using an input-output survey questionnaire at the end of the 2005WS and 2006WS. These farmers planted aerobic rice alone, lowland rice alone or both. There was no difference in the average yield of aerobic and lowland rice in the 2005WS. However, in 2006, a few farmers planted high-yielding varieties and in their fieldslowland rice yielded higher than aerobic rice. The average area planted under lowland rice was slightly higher than that of aerobic rice. The seed rates used by farmers for both aerobic and lowland transplanted conditions were relatively high. Some 106 kg ha−1 in the 2005WS and 125 kg ha−1 in the 2006WS were used under aerobic conditions, whereas, in lowland transplanted conditions, 99 kg ha−1 in the 2005WS and 98 kg ha−1 in the 2006WS were used. In the Philippines, high seeding rates of more than 150 kg seeds ha−1 are not uncommon for direct-seeded rice (PhilRice 2004) in rainfed lowland rice-growing areas to guard against weeds, poor germination, and damage to seeds by biotic

322

and abiotic stresses. For lowland transplanted rice, the reported average seeding rate in the country is 94 kg seeds ha−1 (PhilRice2004), although much lower seeding rates of 40 to 60 kg ha−1 using good seeds are now widely promoted. Fertilizer used was almost equal in 2005 for both aerobic and lowland rice (Table 5). But, in 2006, the amount of phosphorus used in lowland rice was higher than in aerobic rice. The increase in the usage of phosphorus may be attributed to the planting of high-yielding varieties thatrequired high fertilizer use. With regard to herbicide, aerobic rice production used a higher amount than lowland rice. This is quite predictable since, according to farmers, controlling weeds is one of the major problems in aerobic rice. On the other hand, the use of molluscicide washigher in lowland rice production. In terms of irrigation, as expected, significantly lower irrigation frequency was required in aerobic rice than in lowland rice, which was 0.4 and 2.0 to 2.7, respectively. The amount of labor used was almost equal in all major farm activities except for crop establishment (Figure 5). Since aerobic rice uses direct seeding, the labor requirement was significantly lower than in lowland rice, which employs transplanting. By employing direct seeding in aerobic rice production, as much as 17 person-days ha−1 of labor were saved. Furthermore, less total hired labor was needed.

The major cost component in both aerobic and lowland rice was input cost, which accounted for more than 50% of the total cost (Table 6). The high material input cost in aerobic rice was primarily due to herbicide cost, amounting to almost US$40 ha−1 as compared with $14 ha−1 for lowland rice in 2005, and $29 ha−1 as compared with $10 ha−1 for lowland rice in 2006. In terms of irrigation cost, aerobic rice production in this case saved as much as $17 and $15 ha−1 for 2005 and 2006, respectively. On average, hired labor cost was higher in lowland rice by $80 ha−1 in both years. As stated earlier, the average yield of lowland rice and aerobic rice was the same in 2005. However, in 2006, the use of high-yielding varieties by some farmers boosted the yield higher by about 1,000 kg compared with that of aerobic rice. The average price of aerobic rice and lowland rice sold was $0.214 and $0.211 kg−1, respectively. This contradicted farmers’ perception that aerobic rice is usually priced lower than inbred lowland rice. The average gross return for two years of aerobic rice and lowland rice was $317.86 ha−1 and $329.34 ha−1, respectively. These values indicated that economic performance and profitability of aerobic rice were the same as those of lowland rice despite differences in yield. No data were available yet after 2006 with new aerobic rice varieties from the project.

Conclusions Aerobic rice production can be economically viable or feasible in areas where there is inadequate water supply or in rainfed areas where farmers mostly depend on rainfall for rice cultivation. However, aerobic rice technology cannot be used as a substitute for flooded rice. In the trials conducted, aerobic rice demonstrated a good yield even though it was sometimes lower than the yield of irrigated lowland rice. The higher material inputs used in aerobic rice, however, are outweighed by the lower labor inputs and lower water requirement. Looking into economic performance, aerobic rice can be as profitable as irrigated lowland rice.

323

Farmers are enthusiastic about trying to adopt the technology as demonstrated by the increasing number of new adopters. However, upon having more knowledge about and experience with the technology, farmers' decision whether to continue to adopt or not adopt passes through a weighing of criteria that encompasses cultural, economic, and technical considerations. With the new variety (Sahod Ulan 1) available now, there is a need to do training and demonstration with extension officers of the Department of Agriculture, PhilRice, and IRRI for large-scale adoptionand dissemination in rainfed areas in the wet season, and in irrigated areas in the dry season.

References Acosta JE. 2009. Analysis and simulation of water and crop management effects on the

field performance of aerobic rice (Oryza sativa L.) in Los Baños, Laguna, Philippines. PhD thesis. University of the Philippines at Los Baños, Philippines.

Atlin GN, Lafitte HR, Tao D, Laza M, Courtois B. 2006. Developing rice cultivars for high fertility upland systems in the Asian tropics. Field Crops Res. 97:43-52.

Bouman BAM, Tuong TP. 2001. Field water management to save water and increase its productivity in irrigated lowland rice. Agric. Water Manage. 49:11-30.

Bouman BAM, Lampayan RM, Tuong TP. 2007. Water management in irrigated rice: coping with water scarcity. Los Baños (Philippines): International Rice Research Institute. 54 p.

Bouman BAM. 2008. Developing a System of Temperate and Tropical Aerobic Rice in Asia (STAR). CPWF Project Report. International Rice Research Institute, Los Baños, Philippines.

Bouman BAM, Peng S, Castañeda AR, Visperas RM. 2005. Yield and water use of tropical aerobic rice systems. Agric. Water Manage. 74:87-105.

Flor RJB, Palis FG, Cenas PAA, Lampayan RM, Bouman BAM, Singleton GR. 2007. Perceptions and decision-making: a case study on the adoption of aerobic rice. Paper presented at the Federation for Crop Science Societies of the Philippines 19th Scientific Conference, Tagaytay City, Philippines, 13-15 June 2007.

Ganotisi ND, Batuac RA, Castro RC, Lampayan RM. 2009. Performance of AR as affected by crop establishment and seeding rate under Ilocos conditions. Proceedings of the 20th Federation of Crop Science Society of the Philippines Scientific Conference. Siliman University, Dumaguete City, Philippines18-23 May 2009.

Grassi C. 2006. The aerobic rice project: evaluation of the water use efficiency in rice-growing. MSc thesis, Universit adegli Studi di Firenze, Florence, Italy.

Gunarathna MHJP. 2010. Improving crop water productivity of aerobic rice production system using sprinkler. MSc thesis. Central Luzon State University, Philippines.

IRRI (International Rice Research Institute). 2002. Rice almanac. Los Baños (Philippines): IRRI.

Kreye C, Bouman BAM, Reversat G, Fernandez L, Vera Cruz C, Elazegui F, Faronilo JE, Llorca LP. 2009a. Biotic and abiotic causes of yield failure in tropical aerobic rice. Field Crops Res. 112:97-106.

324

Kreye C, Bouman BAM, Castañeda AR, Lampayan RM, Faronilo JE, Lactaoen AT, Fernandez L. 2009b. Possible causes of yield failure in tropical aerobic rice. Field Crops Res. 111:197-206.

Lampayan RM, Bouman BAM, de Dios JL, Espiritu AJ, Soriano JB, Lactaoen AT, Faronilo JE, Thant KM. 2010. Yield of aerobic rice in rainfed lowlands of the Philippines as affected by nitrogen management and row spacing. Field Crops Res. 116:165-174.

Lampayan RM, Bouman BAM, de Dios JL, Lactaoen AT, Quilang EJP, Tabbal DF, Llorca LP, Norte TM, Soriano JB, Corpuz AA, Espiritu AJ, Malasa RB, Vicmudo VR. 2003. Technology transfer for water savings (TTWS) in Central Luzon, Philippines: results and implications. In: Proceedings of the 53rd Philippine Agricultural Engineering Annual National Convention, 21-25 April 2003, Davao, Philippines. p 239-251.

Lampayan RM, Bouman BAM, Faronillo J, Castañeda A, de Dios JL, Espiritu AJ, Lactaoen AT, Norte TM, Soriano JB, Silverio LB, Valdez JA, Villanueva B, Janiya J, Johnson D, Yang X, Wang H. 2005. Management strategies of aerobic rice for water savings and water productivity increase in tropical and temperate environments. In: Proceedings of the PSAE Convention, General Santos City, April 2005.

Nie L, Peng S, Bouman BAM, Huang J, Cui K, Visperas RM, Xiang J. 2008. Alleviating soil sickness caused by aerobic monocropping: responses of aerobic rice to nutrient supply. Field Crops Res. 107:129-136.

PhilRice. 2000. Philippine rice R&D highlights 2000. Philippine Rice Research Institute, Maligaya Science City of Muñoz, Nueva Ecija, Philippines. 69 p.

PhilRice. 2004. Philippine rice statistics 1970-2002, volume 2. Philippine Rice Research Institute (PhilRice), Maligaya Science City of Muñoz, Nueva Ecija, Philippines.

Quicho ED, Flor RJB, Lampayan RM, Bouman BAM, Palis FG. 2008. Introduction of aerobic rice in Bulacan: learnings from participatory development. Paper presented at the 38th Annual Scientific Conference of the Crop Science Societies of the Philippines, 12-16 May 2008, Iloilo City, Philippines.

Soriano JB, Valdez JA, Silverio LB, Pastrana MI, Cabiles DMS, Mendoza JP, Bouman BAM, Lampayan RM, Villanueva BV, Peralta W. 2006. Aerobic rice technology in Bulacan: a research and development (R&D) project. CLARRDEC Sci. Technol. J. 1(1):1-136.

Soriano JB, Valdez JA, Vigilia NB, Cruz GA. 2008. Water-saving technologies for rice production: increasing productivity and income in rainfed areas. Paper presented at the CLARRDEC Regional Symposium on R&D Highlights, Central Luzon State University, Science City of Muñoz, Nueva Ecija, Philippines, 16 August 2008.

Soriano JB. 2008. Water regime and nutrient requirements for aerobic rice production system. PhD thesis, Central Luzon State University, Philippines.

Soriano JB. 2010. Water-saving technologies for rice production: increasing rainwater productivity and income in water-scarce areas of the Philippines. Progress report submitted to IRRC-Water-Saving Workgroup, IRRI, Philippines, November 2010.

325

Soriano JB. 2011. Water-saving technologies for rice production: increasing rainwater productivity and income in water-scarce areas of the Philippines. Progress report submitted to IRRC-Water-Saving Workgroup, IRRI, Philippines, November 2011.

Vermeulen H. 2007. A pot experiment on ‘soil sickness’ in aerobic rice: a biotic problem? MSc thesis, Wageningen University, Wageningen, The Netherlands.

Notes Authors’ address: International Rice Research Institute, Los Baños, Philippines.

Figure 1. A lithao, used for mechanical weeding. Figure 2. Crop establishment in aerobic rice production using traditional technology

(lithao and hand sowing) and an automatic seeder. Figure 3. Rice yield in the two levels of technology across varieties based on two

sources of information, 2003DS, Tarlac, Philippines. Figure 4. Aerobic rice technology guide distributed to farmers. Figure 5. Labor use by activity, Bulacan, 2005-06WS.

Table 1. Activities conducted in connection with the development of aerobic rice in the Philippines through STAR and IRRC projects.

326

Activities Objectives Location Years conducted

Continuous aerobic rice trials

To understand long-term sustainability issues of aerobic rice, crop performance, and water productivity

IRRI 2001-07

Breeding and screening of “aerobic rice,” participatory varietal selection trials (PVS)

To identify varieties suitable for aerobic conditions

IRRI, Tarlac, Nueva Ecija, Bulacan

2001-07

Field experiments: water × variety, water × N (time, split), N × row spacing, N × variety

To establish interactions of different parameters; to study N dynamics in the soil; to derive optimum combination of N, water application, and variety for high yield and water productivity

Bulacan, Tarlac, Nueva Ecija, IRRI

2003-08

Crop rotation experiment To understand aerobic rice performance under different crop rotation schemes

Tarlac, IRRI 2003-05

Weed experiments To identify effective weed management options under aerobic rice systems.

Tarlac 2004-06

Root health/micronutrient experiments

To understand soil health problems and yield failure of aerobic rice under tropical conditions

IRRI 2005-07

Farmer-participatory R&D To test and evaluate on-farm performance of promising aerobic rice varieties in a farmer participatory R&D mode

Tarlac, Bulacan, Nueva Ecija

2002-09

Table 2. Estimated number of trained agricultural workers and farmers,

established demonstration farms, and farmer adopters in the Philippines as of September 2010 (adapted from Soriano et al 2010).

327

Province Year starte

d

No. of demo farmsa

No. of trained

technicians/ researchers

No. of trained farmers

No. of farmer adopte

rsb

Yield range (t ha−1)c

Bulacan 2006 80 35 550 350 3.5−6.0 (4.0) Bataan 2008 55 20 155 55 3.0−4.5 (3.7) Palawan 2007 25 20 632 460 3.8−5.0 (4.3) La Union 2008 45 35 430 206 3.5−5.5 (4.2) Aurora 2008 25 15 200 35 3.0−4.0 (3.4) Pampanga 2008 15 5 45 15 4.0−6.0 (4.7) Isabela 2008 20 35 125 65 3.5−6.5 (4.0) Tarlac 2010 5 5 55 10 3.0−5.5 (3.9) Occidental Mindoro

2010 1 10 35 − 4.5−5.0 (4.7)

Others 2010 60 n.a.d Total 271 240 2,227 1,196

aDemonstration trials conducted in wet season only. bAdopters were farmers who planted aerobic rice in the wet season. cValues in parentheses are average values. dn.a. = not applicable.

Table 3. Results of farmers’ field evaluation of nine aerobic rice cultivars, dry season,

2003-04, in Bulacan.

Line/variety Days to Yield Preferenc Remarks/observatio

328

maturity (t ha−1)

e rank ns from farmers

IR74371-46-1-1 102 5.87 (0.17)a

4 More tillers, good crop stand, long panicles

IR74371-54-1-1

100 5.26 (0.21)

2 Early maturing, good grains, strong tillers, long panicles

IR77080-B-6-2-2 105 5.04 (0.30)

− n.a.b

UPLRi- 7 107 4.44 (0.20)

5 Good grain complexion, big grain size

Apo 110 4.41 (0.26)

1 Good crop stand, long panicles, resistant to drought, more grains per panicle

IR77078-B-17-3-2 110 3.90 (0.36)

− n.a.

PSB Rc 80 110 3.41 (0.24)

− n.a.

IR55419-04 110 3.22 (0.25)

3 More tillers, long panicles

IR64 110 1.10 (0.28)

− None

aValues in parentheses are standard deviation. bn.a. = not applicable.

Table 4. Positive and negative attributes of aerobic rice as perceived by farmers in Pala-

pala, San Ildefonso, Bulacan. Positive attributes Negative attributes

329

Lesser need for irrigation High fertilizer requirement and may result in lodging

Lesser need for fertilizer Exposed panicles Flexibility: good in drought or flooded conditions

Grain is circular and small

Competitive with weeds Low market value Resistance to insect attacks Relatively shorter duration No dormancy period Better weight of grain and good milling recovery

Comparable yields with conventional system

Less overall costs (especially on labor) Source: Flor et al 2007.

330

Table 5. Grain yield and input use by type of rice planted, Bulacan, 2005WS and 2006WS.

Item 2005WS 2006WS

Aerobic Lowland TPR

Aerobic Lowland TPR1

Sample size 41 23 25 17 Grain yield (t ha−1) 3.7 4.2 3.8 4.8***

Area planted (ha) 1.3 1.7 1.3 1.8 Material inputs Seed (kg ha−1) 106 99 125 98 Fertilizer (kg ha−1) N 59 63 47 69 P 33 31 21 41*** K 14 14 12 25 Pesticide (kg a.i. ha−1) Herbicide 0.9*** 0.6 0.9*** 0.5 Insecticide 0.2 0.1 0.1 0.3 Molluscicide 0.0 0.1*** 0.0 0.3*** Irrigation frequency 0.4 2.7*** 0.4 2.0*** Labor use (8-h person-days ha−1) Total labor 34 54 36 53 Hired 25 47*** 31 47*** Family 9 7 5 6 1 TPR =transplanted rice. *** Significant at 1%.

331

Table 6. Comparative economic performance by type of rice planted, Bulacan, 2005WS and 2006WS.

Item

2005WS 2006WS Aerobic Lowland

TPR Aerobic Lowland

TPR1 Production value (US$)1 848.03 916.06 757.65 974.75 Material input costs 297.53 311.68 252.26 344.27 Power cost 103.56 101.43 110.35 127.47 Seed 24.05 37.55*** 23.85 35.48 Fertilizer 96.04 100.95 68.68 112.59*** Other nutrients 7.31 5.68 3.69 17.21 Herbicide 39.61 14.38*** 29.48 9.53*** Insecticide 9.33 14.63 5.80 10.03 Molluscicide 0.00 5.75*** 0.77 3.65 Irrigation 2.88 19.43*** 2.79 17.91** Miscellaneous 14.76 11.68 6.84 10.44 Labor cost 222.97 285.48 197.21 290.70 Hired 171.29 252.24*** 169.56 248.35*** Imputed family labor 51.68 33.24 27.65 42.35** Total cost 520.50 597.16 449.47 634.97 Gross returns 327.53 318.90 308.18 339.78 1 Used PhP 44.16 = US$1. *** Significant at 1%.1 TPR = transplanted rice.

332

Figure 1. A lithao, used for mechanical weeding.

333

Figure 2. Crop establishment in aerobic rice production using traditional technology (lithao and hand sowing) and automatic seeder.

334

Figure 3. Rice yield in the two levels of technology across varieties based on two sources of information, 2003DS, Tarlac, Philippines.

2358 2315

0

1000

2000

3000

4000

5000

Low High

Mea

n yi

eld

(kg/

ha)

Technology level

Interview yield

3508 4058

0

1000

2000

3000

4000

5000

Low High

Mea

n yi

eld

(kg/

ha)

Technology level

Crop-cut yield

335

Figure 4. Aerobic rice technology guide distributed to farmers.

IntroductionAerobic rice entails the growing of rice in aerobic (nonpuddled, nonflooded) soil, with the use of supplementary irrigation and fertilizers to achieve high yields. Aerobic rice varieties combine traits of traditional upland rice (drought resistance) andmodern lowland rice (lodging resistance, high yield potential). The target environment for aerobic rice are areas where wateris not sufficiently available to grow flooded lowland rice, but sufficiently available to grow upland crops such as maize. Current aerobic rice systems have a 20-30% lower yield potential than flooded rice, but need about 50% less water. Aerobic rice should not be grown on the same field in two consecutive dry seasons, but be rotated with upland crops.

• High Tech – use of mechanical seeder pulled by tractor for direct seeding at a seeding rate of 80-100 kg ha-1

Seeding and establishment

Fertilizer management• Apply N in dry season at 120-140 kg ha-1, and in wetseason at 100-120 kg ha-1, in the following splits: 30%

10-14 DAE; 30% at early tillering; 40% at PI• P, K: local recommendations (balanced fertilization)• Zinc and iron sulfate: 15-20 kg ha-1

Integrated weed management

• Use pre-emergence herbicides• Mechanical weeding with “lithao”, “sagad”, or plough withboard removed

• Spot hand weeding

IRRI PhilRice NIA BASC (2005)

Aerobic Rice: A water-saving technology

• Traditional – use of “lithao” to open furrows; hand sowing of seed at a seeding rate of 80-100 kg ha-1,and close slits with feet

Water management

If not enough rainfall, do the following:• Flush irrigate to promote germination• During crop growth, flush irrigate before crop wilts• Keep soil wet around flowering by 2 flush irrigations

• Dry-till the land after harvest of previous crop

ManagementLand preparation

• Ensure that fields are well-levelled and harrowed

Harrowing

Varieties Apo, UPLRI-5, UPLRI-7, Magat, NSIC 192Yield potential of about 5-6 t ha-1

Apo

336

Figure 5. Labor use by activity, Bulacan, 2005-06WS.

337

Paper 14 Developing and disseminating alternate wetting and drying water-saving technology in the Philippines Rubenito M. Lampayan, B.A.M. Bouman, Florencia G. Palis, and Rica Joy Flor

Most rice farmers in Asia practice continuous flooding and maintain more than 5-cm water depth in lowland irrigated rice fields throughout the cropping seasonto obtain maximum yield. This practice uses more than the actual amount of water needed to produce riceand much of it is actually lost through seepage, evaporation, and percolation. To save water, farmers primarily have to reduce these losses while maintaining the transpiration requirements of plants to ensure a good yield. Alternate wetting and drying (AWD) is atechnology that does not require rice fields to be continuously flooded. In the Philippines, AWD was initially implemented through a project called “Technology Transfer for Water Savings (TTWS)” in 2001-04, which aimed to develop and implement a framework for the transfer, adaptation, adoption, and dissemination ofAWD among farmers in the Philippines. The dissemination of safe AWD practices included novel tools such as the “field water tube” and easy-to-understand information materials such as flyers, brochures, and posters to guide the implementation of AWD in farmers’ fields. The first 2 years of the project were designed as a participatory learning phase with selected farmers who were using irrigation water from pump systems in Central Luzon, Philippines. After the trials, the farmers from these systems reported savings of 16% to 30% of irrigation water and an increased net profit of more than PHP 2,000. Later, more farmers in these systems adopted the technology, and it spilled over to gravity irrigation systems in the country. Our expanding impact-pathway networks (both scaling out and scaling up) have been driven by widespread training activities, while building on necessary outscaling mechanisms, including documenting evidence from local success stories, increasing buy-in from local R&D partners, and policy support.

Rice is the staple food of half of the global population. A harvested area of 75 million hectares of irrigated rice fields, which annually provide 560 million tons of rough rice (Maclean et al 2002), is the world’s single biggest user of diverted fresh water resources. Rice occupies 30% of the world’s irrigated cropland but, because it is flooded, receives about 40% of the irrigation water (Dawe 2005). The bad news is that worsening water scarcity threatens the sustainability of the world’s premier staple food. An inadequate supply of water during crop establishment or during the vegetative or reproductive stage of the crop generally leads to a significant reduction in rice yield (Wopereis et al 1996, Bouman and Tuong 2001), thus making water a critical component.

Irrigated lowland rice is grown under flooded conditions. On average, an irrigated rice field in Asia receives some 1,500 millimeters (mm) of water per growing season (Bouman et al 2007). During field turnaround and the crop-growth period, water outflows are runoff over the bund, transpiration, evaporation, seepage, and percolation. Only transpiration is a productive water flow as it contributes to crop growth. Runoff, evaporation, seepage, and percolation are unproductive water flows and therefore losses at the field level, though some of this water may be recaptured at additional cost and reused further downstream. As the combined unproductive water flows can amount to 60% to 80% of all water inputs to rice fields, scientists have been looking for ways to

338

reduce unproductive outflows while maintaining transpiration and, hence, crop growth and yield (Bhuiyan et al 1995, Bouman and Tuong 2001, Li 2001, Tabbal et al 2002). One of these technologies is alternate wetting and drying (AWD). AWD is a water-saving technology that lowland (paddy) rice farmers can apply to reduce their water use in irrigated fields. In AWD, irrigation water isapplied to flood the field for a certain number of days after the disappearance of ponded water. Hence, the field is alternately flooded and nonflooded.

In the Philippines, particularly in Central Luzon, water availability for rice production has declined in the last decades due to increasing population (Pingali et al 1997) and water quality degradation (Castañeda and Bhuiyan 1993). Many irrigation systems were also destroyed and clogged by the earthquakes of 1990 and Mt. Pinatubo in 1991 (NIA 1996), and cropping intensity dropped from 169% in the 1990s to 137% in the early 2000s. This prompted the country’s National Irrigation Administration (NIA) to enhance irrigation water availability by infrastructure development and the promotion of sound field-water management (NIA 1996). This is particularly true in Tarlac Province, one of the seven provinces of Central Luzon. The system damaged by the eruption of Mt. Pinatubo in 1991 causing great economic disturbance for thousands of Tarlac farmers prompted NIA to reactivate 72 deep-well pump (DWP) irrigation systems in Tarlac Province through the Tarlac Groundwater Irrigation Systems Reactivation Project (TGISRP) starting in 1997. These DWPs are managed by farmers’ groups known as irrigators’ service cooperatives (ISC). However, with DWPs, farmers facea high cost of pumping; thus, rice farming is less profitable. In one of the deep-well systems under the TGISRP (Figure 1), a project aimed at facilitating adaptation and eventual adoption of water-saving technologies, particularly AWD and aerobic rice, was launched in 2001 (Lampayan et al 2004). The Technology Transfer for Water Savings (TTWS) project was implemented through an interagency collaboration among the International Rice Research Institute (IRRI), NIA in Tarlac Province, and the Philippine Rice Research Institute (PhilRice) (Bouman et al 2002). The project employed a farmer participatory approach in the introduction of two water-saving technologies, AWD and aerobic rice, for which farmers and researchers worked together in technology validation. The pilot site for AWD through the TTWS project was in one deep-well system (P-38) within the village of Canarem in Victoria, Tarlac. Following the participatory introduction of safe AWD in P-38, this chapter discusses the results of the implementation and adoption, and the impacts of safe AWD at the study site and its diffusion to other large irrigation systems in the Philippines.

Developing a tool for safe AWD implementation In AWD, the field is not continuously flooded. Instead, the soil is allowed to dry out for some days after the disappearance of ponded water before it is flooded again. During these periods, seepage and percolation losses decline to nearly zero because of the almost absence of hydraulic head (Bouman et al 2007). This practice has long been investigated and has found practical implementation in many countries, including China and Japan. However, implementation has been very site specific in terms of the timing, duration, and frequency of nonflooded periods; promoted to farmers using difficult and

339

complex recommendations; and difficult to scale up for use by large numbers of farmers. The number of days that the soil is left dry can vary from 1 to 10.

The development of a practical way to implement AWD started in 2001 through the TTWS project of the Irrigated Rice Research Consortium’s Water-Saving Workgroup (Lampayan et al 2004). The workgroup developed a novel method of AWD called “safe AWD” that reduces water needs by 15% to 30% while maintaining yield and sometimes even increasing it (Figure 2). As safe AWD is easy to implement, extension services and farmers can easily adopt it. The only time that the field is necessarily kept flooded is at flowering to avoid spikelet sterility and resulting yield loss. A perforated tube about 30 cm long (that can be made of plastic pipe or bamboo or any indigenous material) is embedded in the paddy field to reveal the ponded water level even when it drops below the soil surface. Irrigation is applied to flood the field again with a depth of around 5 cm when the ponded water level has dropped to about 15 cm below the surface. The tube can be used in any field or soil type, except in soil that is too sandy to be good for growing paddy rice. In clay soil with low percolation rates, the period of nonflooding can last up to 10 days, while in more loamy and permeable soil it may last only 2 to 3 days. With the use of the tube, threshold irrigation criteria were defined using field experiments at various locations, extrapolation using crop growth and water balance simulation models, and on-farm participatory adaptive research and technology evaluation (Bouman et al 2007). The threshold of 15 cm is called “safe AWD” as this will not cause any yield decline since the roots of the rice plant will still be able to take up water from the almost nearly saturated soil and the perched water in the root zone. At 15-cm depth, soil-water potential is still above −10 kPa (Figure 3), reflecting that water in the soil is very much available for plant use. Under safe AWD, no special N management regime is needed and local recommendations as for flooded rice can be used (Belder et al 2004, Cabangon et al 2011). Farmer participatory trials in P-38, Canarem, Tarlac, Philippines From among the 63 farmers drawing water from P-38, 11 farmer-cooperators volunteered to demonstrate AWD technology on their respective farms. Each farmer-cooperator contributed two neighboring plots with a size of 500 to 1,000 m2 each, one representing the current farmers’ practice (FP) and one the AWD technology (Lampayan et al 2004). In the deep-well system of Canarem (P-38), the frequency of irrigation water application was determined by the rotation delivery system, and farmers followed certain irrigation delivery schedules (usually once a week). To realize nonflooded conditions at the end of a rotation period, the standard amount of 5- to 7-cm water application at each application in the farmers’ practice (flooded fields) was reduced to 3−4 cm in the controlled irrigation field. Production inputs were the same in both plots. The fertilizer rate was 90-40-40 kg NPK ha-−1, with 40 kg NPK ha−1 basally applied a day before transplanting. PSBRc 98 (a 115-day-duration variety) was used and grown by farmers in nurseries for 21 to 25 days with a seeding rate of 60 kg ha−1. Irrigation water input in both plots was measured using trapezoidal weirs, and field water tubeswere installed to monitor daily ponded and perched water depth in the plots. Grain yield in AWD and FP

340

was measured and an input-output survey was carried out using a semi-structured questionnaire at the end of the harvest. Results showed that, after the participatory trials in the 2002-03DS, farmers gave positive feedback about the effectiveness of AWD, which resulted in their positive acceptance of the technology (Palis et al 2004). This positive feedback was supported by the data collected from the farmer-cooperators’ fields. Farmers generally perceived that yield in the FP plots were similar to yield in the AWD plots, even when less water was used. From the field data, average yield did not vary significantly between AWD and FP plots during the 2002 and 2003 dry seasons (Figure 4A). In the 2002DS, yield also did not vary across toposequences and ranged from 5.3 to 5.5 t ha−1. Yield was higher in 2003 than in 2002, which may have been caused by a variety of reasons beyond the usual effects of differences in weather. In 2003, farmers used more fertilizer (Table1) and were better instructed in overall good crop management practices through cooperation in the project. Figure 4Bshows the total water use in FP and AWD plots during the 2002 and 2003DS. The average difference in total water use between FP and AWD plots was 16% in 2002 and 24% in 2003. The largest water savings were 24% in 2002 and 33% in 2003. From Figure 4C, it can be seen that the alternate wetting and drying period of the paddy field was taking place in both FP and AWD plots, although AWD plots maintained only 2-cm-depth standing water after each irrigation period, in contrast to the usual farmers’ practice of 5 to 10 cm. This suggested that, with the weekly irrigation delivery interval in P-38, farmers were already adopting AWD water management by default. The deeper perched water depths (>30 cm) in Figure 4C were the response to the decision of P-38 management to experiment with the threshold level by increasing the irrigation interval from 1 week to about 2 weeks in the latter part of the crop growth period (Palis et al 2004), and this had resulted in a savings of about two irrigations for farmers. The higher savings in 2003 reflected the effects of the learning process: at this time, farmers were already confident about the performance of AWD and were willing to take more risk in saving more water. All farmer-cooperators acknowledged that AWD saves time, labor, and expenses. It reduced cost by using 20% to 25% less fuel and oil (Table 1). It reduced labor as farmers spent fewer hours in irrigation (depending on the distance). Farmers viewed AWD as not only a water-saving technology but also a fuel-saving technology.

AWD diffusion and impact in TGISRP The AWD demonstration trials in selected farmers’ fields in P-38 also served as a “lighthouse” for other farmers in P-38 and for the entire TGISRP. With the experience of AWD farmer-cooperators in P-38, the officers of the ISC implemented AWD in the whole P-38 service area, after consultations with the members of the ISC. Farmers’ field dayswereregularly conducted at the site, bringing in farmers from other systems to see the demonstration trials and to interact with their fellow farmers. Series of training activities and briefings were also conducted for interested farmers in selected DPWs from 2002 to 2004, as NIA introduced the technology to the other ISCs within the TGISRP. To assess the level of diffusion, adoption, and impact of AWD in the whole P-38, key informant interviews and focus group discussions were conducted intermittently

341

from 2003 through early 2007. A knowledge, attitudes, and practices (KAP) survey was implemented in 2005 to 2006 among the 194 rice farmers of 25 DWP ISCs. An economic input-output survey was also done with a smaller sample of 146 farmers within the KAP sample. Comparison was made between AWD farmer-adopters and non-AWD farmer-adopters.

Diffusion and impacts in P-38 From Table 2, it is clear that there was a reduction of 49% in hours spent on irrigation in 2004 compared with 2003 in P-38. Although the irrigation hours had already decreased by the 2003DS, further reduction was evident as the farmers became more confident of using safe AWD. This had translated into a difference in average costs for fuel in irrigation of about PHP 3,000 (about US$55) between 2003 and 2004. Farmers who continuously used AWD had consistently higher net returns. This further encouraged them to use AWD for their entire area as opposed to having plots of AWD only. In 2005, in a KAP survey among P-38 farmers, some 64.7% said that they had heard about AWD before (Table 3). This made it appear that the AWD technology was heard of and practiced only by some farmers in P-38. However, the 1 to 2 weeks’ rotation or water delivery interval of the pump suggests that there is enforced AWD because of the schedule. Through the interviews and participant observations in farmers’ fields, it was verified that, at irrigation in specific crop stages when AWD was to be applied, farmers in P-38 irrigated only after at least 8 days since their last irrigation application because of the pump schedule. At this time, the soil was already dry with some cracks visible at the paddy surface. These data were supported by the cooperative’s record on the use of deep-well and associated fuel consumption. Based on the record, the average irrigation frequency for P-38 farmers was 10 for the whole season, with an average interval of 9 days. Furthermore, their irrigation time on average was 50.4 hours per hectare for the whole season. Most of them did not have shallow tube wells to support deep-well irrigation. This made their practice actually an AWD. The figures obtained in Table 3 are characteristic of the process of AWD adoption in P-38. The farmers of P-38 actually implemented AWD, albeit not consciously, because it was enforced by the schedule. AWD had become part of the way the P-38 ISC functions. Some farmers copied the technique of irrigation from farm neighbors. They did this with the idea of saving cost since the ISC policy was for them to bring their own fuel for irrigation. Farmer learning of techniques and practices in production came about through observation from neighboring farmers. The observation regarding irrigation practices was made in the field, and could be spontaneously copied and reproduced. After the implementation and consequent adoption of AWD in P-38, an increase in yield between the 2003 and 2005DS was seen (Table 4). One factor that may have led to the 10% increase in yield and gross returns was the better management practice as the farmers had become more informed about crop management. There was also a definite reduction in the way that the farmers used water in P-38 after the system-wide implementation of AWD (Table 4). Farmers had seen that reducing the amount of water did not have negative effects on yield.

342

Another effect of the adoption of AWD is the reduction in the number of hours devoted to irrigation, a 48% reduction. For some farmers, the reduction in the hours spent on irrigation translates into more time doing other things on the farm or even time for off-farm work. Another change among the farmers is that, despite the long intervals of 1 to 2 weeks, farmers were no longer worried about the schedule for irrigation. In fact, the farmers were even satisfied with their schedule (Table 5). This translated into lesser tension among them in access to irrigation. Competition for water was also minimized because the farmers did not use the pump for long periods at one time. The schedule of water distribution was easily implemented, while improving pump management and decreasing conflict among members within the irrigators’ association. Likewise, water saved because of AWD, which was 20% to 30%, allowed for expanding the irrigation service area. From 52 ha in 2000, it became 60 ha after 2002. Diffusion and impacts of AWD in other pumps within TGISRP Table 6 shows the comparison of average yield, cost, and returns of AWD farmer-adopters and non-AWD farmer-adopters from the other 32 DWPs in Tarlac during the 2005DS. Among the respondents interviewed, only 18% (29 out of 161) said that they had practiced AWD, and the rest (82%) did not practice it. It had to be noted, however, that this non-AWD variable is the response of farmers to the question of whether they had heard about the technology, and whether they had practiced AWD before, and not on actual measurement of their water use. Considering the scheduling of the deep-well pump systems, those using the deepwell generally use less water, somewhat in the AWD sense, although the farmers may not know it. The average yield per hectare of AWD farmer-adopters was lower by 201 kg of rice than the average yield of the non-AWD adopters. Statistically, however, the difference was not significant. Looking at the cost variables, the only production expense that differed significantly was the irrigation cost or the cost of fuel consumption of the pump. The AWD farmer-adopters saved about PHP 5,000, or about a 48% reduction in irrigation cost. Thus, in return, the AWD adopters had significantly higher average net returns (by 25%) than the non-AWD adopters. Farmers reported that, from 17 times to irrigate the farm, upon practice of AWD, they irrigated the rice farm only 10 times on average. These results suggest that the practice of AWD has been diffused to other rice farmers within the deep-well system aside from P-38 users.

Introduction of AWD in canal irrigation systems in the Philippines In the course of project implementation in Tarlac, a number of information, training, and extension materials (bulletins, flyers, leaflets, flipcharts, videos, e-learning modules, TV segments, press releases) were developed to support the training and promotion activities for AWD (Lampayan 2010). These materials werealso already made available in IRRI’s Cereal Knowledge Bank. Seminars, workshops, and training activities were conducted in many regions in the country in response to the large number of requests from various national agricultural research and extension (NARES) partners in the country. With success in pump systems in Tarlac, AWD was introduced in several national irrigation systems in the Philippines, including the Bohol Integrated Irrigation

343

System (BIIS) in 2005 and the Upper Pampanga River Integrated Irrigation System (UPRIIS) in 2007.

BIIS is composed of three national irrigation systems covering a total area of 10,260 ha: the Capayas Irrigation System (CIS) in Ubay (1,160 ha), the Bohol Irrigation System 1 (BIS 1; 4,960 ha), and the Bohol Irrigation System 2 (BIS 2; 4,140 ha). The three irrigation systems are served by reservoir-type dams: the Malinao Dam of BIS 1 in Pilar, the Bayongan Dam of BIS 2 in San Miguel, and the Capayas Dam in Ubay. Among the challenges faced by NIA in the BIIS is how to improve the performance of irrigation systems as well as how to increase water productivity. Since the start of operations in 1998, BIS 1 has performed poorly because of inefficient water use.

The dam has been beset by problems—declining available water, a synchronous farming activities resulting in wasteful use of water, and poorly maintained irrigation facilities. All of these, in turn, affected farm productivity and contributed to low farmer income. For this, NIA created an action plan for BIIS in 2005 focusing on improving water distribution equity and efficiency; improving operations; strengthening coordination among NIA, irrigators’ associations, and local government units; rehabilitating and upgrading irrigation facilities; and establishing demonstration farms on water-saving technologies. The project opened with the training of 31 NIA technical staff in 2005 and training of 200 irrigators’ association (IA) leaders in 2006. From 2006 to 2007, 21 AWD demonstration farms were established in cooperators’ fields. Following this, a total of 19 workshops on water savings brought knowledge to some 3,000 farmers from 19 IAs. NIA introduced water savings to 34 IAs or 2,815 farmers prior to implementation in the BIS 1 service area. To document the initial adoption and impact of AWD, IRRC conducted in 2007 a baseline survey with 225 farmers, focus group discussions, interviews with 15 key informants, and a mini-survey with 70 AWD farmers in BIS after one season with AWD (Flor 2009). The results of the initial adoption study showed that AWD facilitated an increase in irrigated area in BIS1 by 20% (from 2,324 ha in 2005 to 2,819 ha in 2006). Farmers claimed that, with AWD, they realized that they could still grow good rice crops even with less available water (Flor 2009). In 2008, NIA had fully implemented AWD in the whole system by revising its irrigation delivery policy to “enforce” adoption of the technology (Lampayan 2010).

In UPRIIS, AWD was demonstrated with the farmers’ Pook Malaya Irrigators’ Association in Sto. Domingo, Nueva Ecija (Sibayan et al 2010). The IA has 291 farmer-members, with 257 hectares that are divided into nine turnout service areas (TSAs). In 2007, AWD was implemented in one of these turnout service areas, with 39 farmers involved in managing their water on their respective farms. On-site briefings were conducted before the start of the safe AWD implementation in the farmers’ fields. During implementation, field-water monitoring was carried out taking readings from the installed field-water tubes. Field days and focus group discussions were also conducted to allow neighboring farmers to see and compare AWD and non-AWD fields in the TSA. In the following year (2008), AWD implementation was expanded to more TSAs in Pook Malaya IA. For two years of technology demonstration, farmers reported that there was no yield penalty at AWD sites despite the reduction of 15% to 30% in irrigation water. The technology also improved relationships among water users and facilitated good harvesting operations because of better soil condition that promotes mechanization

344

(Sibayan et al 2010). Through the strong partnerships with farmers and other institutions (IRRI, PhilRice, and NIA), a number of lessons were learned on AWD implementation in Pook Malaya IA. First, farmers will not adopt a technology if they don’t believe in it and seeing and experiencing by themselves or learning by doing are still the best way of convincing farmers. Second, collective effort is important and clustering is an effective tool, and capitalizing on the cohesiveness of the IA is a facilitative option. Third, incentives can be introduced but they are not a must in order to disseminate any technology. And, last, partners and farmers need to be trained to enable them to transfer the knowledge well to others. The success of AWD demonstrations in Pook Malaya IA encouraged UPRIIS management to scale out AWD to the whole UPRIIS system, whose total irrigated area is 109,551 hectares of rice land.

The successful adoption of AWD in TGISRP, UPRIIS, and BIIS became the basis in the formulation of Administrative Order (AO) 25 on “Guidelines for the adoption of water-saving technologies (WST) in irrigated rice production systems in the Philippines.” In March 2008, the Water-Saving Workgroup of IRRC in the Philippines met at IRRI to form a team to come up with policy support to bring about nationwide dissemination and adoption of AWD. In April 2008, the office of the Department of Agriculture issued a Special Order (SO no. 266) creating the technical working group (TWG) tasked to formulate implementing guidelines in the adoption of water-saving technologies for rice in the Philippines (DA 2008). After several meetings among TWG members, a draft of the guidelines was presented for consultations throughout the country. Four island-wide consultation meetings were conducted from June to August 2009, which were participated in by NIA personnel at the regional and system levels, officials of confederated IAs, regional and provincial agriculture officials, representatives from the state colleges and universities, and NGOs. In November 2009, DA Administrative Order 25 was issued by the DA secretary, which provides “Guidelines for the adoption of water-saving technologies (WST) in irrigated rice production systems in the Philippines” (DA 2009). DA-AO 25 is the policy support for the nationwide adoption of AWD. After the issuance of the order, a series of technical briefings was carried out in different regional irrigation offices in the country for effective implementation of the technology (Sibayan 2011). In May 2011, a conservative estimate of 81,687 farmers were reported to be adopting AWD in the Philippines (Table 7), and this figure was expected to increase by threefold as NIA has an ongoing program to increase, rehabilitate, and come up with irrigated areas to support the country’s Staple Food Self-Sufficiency Program. NIA already targeted 59 irrigations systems throughout the country as models for irrigation management transfer (IMT) programs. Of these, initial AWD activities have been carried out in 22 systems.

With the success of AWD in the Philippines and through large IRRC networks, participatory research to validate AWD and develop training and extension materials in local languages has also been carried out by a number of NARES partners in South and Southeast Asia since 2005. Large-scale dissemination of AWD is also happening in Vietnam and Bangladesh, while field demonstrations and validations have been conducted in other countries such as Myanmar, Indonesia, Laos, and India (IRRC 2011). Diffusion pathways for AWD dissemination

345

When the IRRI-based Water-Saving Workgroup of the Irrigated Rice Research Consortium started the TTWS project in 2001, only a few NARES partners in the Philippines were involved in AWD activities. With a modest operating budget, the network of partners expanded rapidly after the success of AWD adoption by farmers in P-38, and in other deep-well systems in Tarlac Province. Originally, two Philippine national partners were included as official project partners: the Philippine Rice Research Institute and the National Irrigation Administration in Tarlac. Through the success and visibility of their activities (they organized and attended many national workshops and training activities), more and more R&D partners became interested in AWD and other water-saving technologies. The most prominent were Bulacan Agricultural State College (BASC), Bureau of Soils and Water Management, Central Luzon State University, Bohol Agricultural Promotion Center (BAPC), Agricultural Training Institute, national and regional offices of the NIA in the country, local government units, and other state colleges and universities (SCUs). They also became members of the Water-Saving Workgroup and fully participated in project planning and workshops. These partners picked up different components of the impact-pathway circle (Figure 5) according to their own mandates and interests, ranging from research to teaching, training, and extension. PhilRice and NIA jointly initiated a range of adaptive research and dissemination activities on AWD for large gravity irrigation systems in the Philippines. Each year, training was provided and farmers’ field schools were organized in many parts of the country and support was mobilized through local government units and village leaders. Within a few years, a rather complete set of activities along the impact-pathway circle had been developed at the local scale (encompassing a few villages, but with ambition to scale up and out nationally). Some local partners were successful by obtaining national R&D grants to finance their activities, and it was their participation in high-profile programs such as the IRRC that contributed to that success.

All project partners benefited a lot from working in partnership, and the value of interaction at joint planning meetings, in training activities, and in workshops cannot be overstated. Though we did not systematically document it, we believe that significant partner changes along the impactpathway have occurred over time. Three key lessons we learned follow: (1) let partners take ownership and give them freedom to modify and adapt concepts and practices; (2) be flexible in partnership arrangements; follow new initiatives developed by new partners; include new and exciting partners in the project; and (3) create opportunities for new partnerships through training.

Conclusions Alternate wetting and drying irrigation is a mature technology ready for dissemination in irrigated rice areas where water availability is limited. In P-38, AWD was successfully introduced to farmer-cooperators and it saved 16% to 24% of irrigation water without a yield reduction. The adoption of AWD in a deep-well system requires close cooperation among members and good management of the system. The technology was successfully adopted in the whole P-38 deep-well system, diffused to other deep-well pumps in Tarlac, and implemented in a number of small and large canal irrigation systems in the country. Two years after the introduction, the economic, social, and

346

cultural impacts are evident. The introduction of AWD in P-38 led to institutional impacts among those who were partners with IRRI in the introduction (NIA and PhilRice) as well as with other stakeholders or new partners. The success story of AWD adoption in the P-38 deep-well irrigation system in Tarlac and the training activities conducted by Water-Saving Workgroup partners paved the way for the spontaneous rapid outscaling of the water-saving technologies. The potential adoption domain for safe AWD is far beyond the Philippines and includes the irrigated rice areas in Asia, especially during the dry season, when farmers incur pumping cost or when water scarcity occurs. The success of safe AWD adoption in pump-irrigated areas has paved the way to more widespread validation and adoption of safe AWD in the Philippines and other Asian countries such as Bangladesh and Vietnam. However, the speed and extent of adoption will depend very much on the long-term commitments of NARES partners and their supporting governments.

References Belder P, Bouman BAM, Spiertz JHJ, Cabangon R, Guoan L, Quilang EJP, Li Yuanhua,

Tuong TP. 2004. Effect of water and nitrogen management on water use and yield of irrigated rice. Agric. Water Manage. 65:193-210.

Bhuiyan SI, Sattar MA, Tabbal DF. 1995. Wet seeded rice: water use efficiency, productivity, and constraints to wider adoption. In: Moody K, editor. Constraints, opportunities, and innovations for wet seeded rice. Los Baños (Philippines): International Rice Research Institute. p 143-155.

Bouman BAM, Tuong TP. 2001. Field water management to save water and increase its productivity in irrigated rice. Agric. Water Manage. 49(1):11-30.

Bouman BAM, Tabbal DF, Lampayan RM, Cuyno RV, Quiamco MB, Vicmudo VR, Norte TM, Lactaoen AT, Quilang EJP, de Dios JL. 2002. Knowledge transfer for water-saving technologies in rice production in the Philippines. Paper presented at the 12th Philippine Agricultural Engineering Week (52nd Annual National Convention) of the Philippine Society of Agricultural Engineers, Puerto Princesa City, Palawan, Philippines, 22-26 April 2002.

Bouman BAM, Tuong TP, Lampayan RM. 2007. Water management in irrigated rice: coping with water scarcity. Los Baños (Philippines): International Rice Research Institute. 54 p.

Cabangon RJ, Corcuera F, Angeles O, Lampayan RM, Bouman BAM, Tuong TP. 2008. Field water tube: a simple tool for managing water under alternate wetting and drying irrigation. Crop Science Society of the Philippines 19th Scientific Conference, May 2008.

Cabangon RJ, Castillo EG, Tuong TP. 2011. Chlorophyll meter-based nitrogen management of rice grown under alternate wetting and drying irrigation. Field Crops Res. 121:136-146.

Castañeda AR, Bhuiyan SI. 1993. Sediment pollution in a gravity irrigation system and its effects on rice production. Agric. Ecosyst. Environ. 45:195-202.

Dawe D. 2005. Increasing water productivity in rice-based systems in Asia: past trends, current problems, and future prospects. Plant Prod. Sci. 8:221-230.

347

DA (Department of Agriculture). 2008. DA Special Order No. 266: Creation of the technical working group to formulate implementing guidelines in the adoption of water-saving technologies for rice in the Philippines, April 2008.

DA (Department of Agriculture). 2009. The Department of Agriculture Administrative Order No. 25 series of 2009: Guidelines for the adoption of water-saving technologies (WST) in irrigated rice production systems in the Philippines, November 2009.

Flor RJ. 2009. Adoption and impacts of AWD in Bohol Integrated Irrigation System (BIIS). IRRC Country Outreach Program (ICOP), International Rice Research Institute, Los Baños, Philippines.

IRRC (Irrigated Rice Research Consortium). 2011. IRRC (Phase IV) 2010 Annual report. Report submitted to Swiss Agency for Development and Cooperation (SDC).

Lampayan RM, Bouman BAM, de Dios JL, Lactaoen AT, Espiritu AJ, Norte TM, Quilang EJP, Tabbal DF, Llorca LP, Soriano JB, Corpus AA, Malasa RB, Vicmudo VR. 2004. Adoption of water-saving technologies in rice production in the Philippines. Food and Fertilizer Technology Center Extension Bulletin 548. Republic of China on Taiwan: FFTC. 15 p.

Lampayan RM. 2010. Research to impact: the story of alternate wetting and drying water saving technology. IRRI Seminar, September 2010.

Lampayan RM. 2011. IRRC Water-Saving Workgroup (Phase 4): development and dissemination of technologies to help farmers cope with water scarcity. Report presented during the IRRC Review Meeting, 3 September 2011, IRRI, Philippines.

Li YH. 2001. Research and practice of water saving irrigation for rice in China. In: Proceedings of the International Workshop on water saving irrigation for paddy rice, 23-25 March, Wuhan, China. Wuhan University, Wuhan, China. p 135-144.

Maclean JL, Dawe D, Hardy B, Hettel GP, editors. 2002. Rice almanac. Los Baños (Philippines): International Rice Research Institute. 253 p.

NIA (National Irrigation Administration). 1996. Annual report 1996. NIA, Manila, Philippines. 46 p.

Palis FG, Cenas PAA, Bouman BAM, Hossain M, Lampayan RM, Lactaoen AT. 2004. Farmer adoption of controlled irrigation in rice: a case study in Canarem, Victoria, Tarlac. Philipp. J. Crop Sci. 29:3-12.

Pingali PL, Hossain M, Gerpacio RV. 1997. Asian rice bowls: the returning crisis? Los Baños (Philippines): International Rice Research Institute, and Oxon (UK): CAB International. 341 p.

Sibayan EB, de Dios JL, Lampayan RM. 2010. Outscaling AWD in a public-managed reservoir-type irrigation system: a case study in the Philippines. In: Palis FG, Singleton GR, Casimero MC, Hardy B, editors. Research to impact: case studies for natural resource management for irrigated rice in Asia. Los Baños (Philippines): International Rice Research Institute. 370 p.

Sibayan EB. 2011. National dissemination of AWD in the Philippines: experiences and the role of policy support. International workshop on “Alternate wetting and drying for resource conservation and reduction of environmental pollution” held at BRAC Center, Dhaka, Bangladesh, on 13-14 December 2011.

348

Tabbal DF, Bouman BAM, Bhuiyan SI, Sibayan EB, Sattar MA. 2002. On-farm strategies for reducing water input in irrigated rice: case studies in the Philippines. Agric. Water Manage. 56:93-112.

Wopereis MCS, Kropff MJ, Maligaya AR, Tuong TP. 1996. Drought-stress responses of two lowland rice cultivars to soil water status. Field Crops Res. 46:21-39.

Notes Authors’ address: International Rice Research Institute, Los Baños, Philippines.

Figure 1. Locations of P-38 and other deep-well pump (DWP) systems in Tarlac, Philippines.

Figure 2. AWD bulletin to guide NARES partners and farmers in their AWD dissemination activities.

Figure 3. Relationship between soil water potential and fieldwater depth (adopted from Cabangon et al 2008).

Figure 4. Grain yield (A), total water used (B), and example field water status (C) in AWD and FP plots in P-38 during 2002 and 2003DS (adopted from Lampayan et al 2004).

Figure 5. Research-to-impact pathway networks for AWD in the Philippines.

349

Table 1. Average yield, cost, and returns of rice grown under alternate wetting and drying (AWD) and farmers’ practice (FP) in P-38, Canarem, Victoria, Tarlac, for the 2002DS and 2003DS, respectively.

2002DS (n=11) 2003DS (n=12) Item FP AWD FP AWD Yield (kg ha−1) 5,400 5,400 5,230 6,080 Standard error of mean yield (kg ha−1) 150 190 380 410 Gross return (PHP ha−1) 51,300 51,300 62,348 60,785 Material Cost (PHP ha−1) 15,701 13,329 13,370 12,085 Seeda 1,950 1,950 600 600 Fertilizer 2,400 2,400 5,362 5,362 Pesticide 650 650 252 252 Fuel and oil 9,841 7,469 6,769 5,484 Others 860 860 387 387 Labor (PHP ha−1) 8,567 8,521 15,174 14,935 Land preparation 2,500 2,500 3,089 3,085 Crop establishment 2,700 2,700 3,658 3,658 Crop care 867 821 1,731 1,496 Postharvest labor 2,500 2,500 6,696 6,696 Total production cost (PHP ha−1) 24,267 21,849 25,845 27,021 Net profit (PHP ha−1) 27,032 29,451 33,803 33,765 Net profit (US$ ha−1)b 530 577 624 623

a In the 2003 DS, seeds were provided to farmers by PhilRice at subsidized cost. bAverage exchange rates: US$1 = PHP 51.002 (2002) and PHP 54.206 (2003). Source: Lampayan et al 2004.

350

Table 2. Average irrigation hours and cost in P-38ISC in 2003 and 2004DS. Item 2003DS 2004DS Differencea Number of farmers interviewed 54 42 12 Average size of farm (ha) 0.9 0.9 0 Fuel consumption (L h−1) 8.0 8.0 8 Total hours’ irrigation 1,143.8 768.4 375.4* Fuel cost (PHP)a 9,150 6,147 3,003* a Average exchange rates: US$1 = PHP 54.206 (2003) and PHP 56.041 (2004). * Significant at 1% level using t-test.

351

Table 3. AWD knowledge and practice according to P-38 ISC farmers in 2005 KAP survey. Item Frequency (n=34) Percentage (%) Heard about it 22 64.7 Trained 15 44.1 Practiced AWD before 12 35.3 Continued 9 26.5

352

Table 4. Comparison in average gross returns, total irrigation, and fuel consumption for all farmers (n=63) in P-38 between 2003DS and 2005DS.

Items 2003DS 2005DS % Change Yield (kg ha−1) 4,551 5,072 +11.5 Gross returns (PHPa ha−1) 45,510 50,720 Irrigation Total number of irrigations 10 10 0 Mean hours per irrigation 7.6 4.0 −48.0 Fuel usage Total fuel consumption (Lha−1) 323.7 193.7 −40.0 Total fuel cost (PHP ha−1):

nominal 6,147 7,338 +19.4

Total fuel cost (PHPha−1): real (base year = 2003) 6,147 3,676 −40.0

aAverage exchange rates: US$1 = PHP 54.206 (2003) and PHP 55.098 (2005).

353

Table 5. Farmers’ knowledge and perceptions relating to irrigation (2005 KAP survey in P-38). Items Frequency (n=35) Percent Less water can hurt the crop - Agree 7 20.0 - Disagree 28 80.0 The more water, the better for the plants

- Agree 8 22.9 - Disagree 19 54.3 - Neither 7 20.0 - Depends on the crop stage 1 2.9 My irrigation schedule is the best for my farm

- Agree 33 94.3 - Disagree 2 5.7

354

Table 6. Average yield, cost, and returns of AWD adopters (n=30) and non-AWD

adopters (n=116) from 32 deep-well systems, 2005DS.

Item

AWD adopters (n=30)

Non-AWD adopters (n=116)

Difference (AWD-non-AWD)

Yield (kg ha−1) 4,756.3 4,957.4 −201.1 Standard error of mean yield (kg ha−1) 450.0 320.0 n.a.b Gross returns, PHP ha−1 49,647 51,963 −2,316 Materials cost,PHP ha−1 17,385 23,042 −5,657 Seed 1,712 1,939 −227 Fertilizer 5,717 6,415 −698 Pesticide 701 776 −75 Fuel and oil (pump irrigation) 5,555 10,428 −4,873* Others (food, tractor and land rent) 3,701 3,484 216 Laborcost, PHP ha−1 10,367 12325 −1,958

Land preparation 505 845 −340 Crop establishment 2,307 2,546 −239 Crop care 34 55 −21 Postharvest 6,752 7,838 −1,086 Others 770 1,042 −272 Total production cost, PHP ha−1 27,752 35,367 −7,615 Net returns, PHP ha−1 Net returns, US$a ha−1

21,895 397

16,596 301

5,299 96

aAverage exchange rate in 2005: US$1 = PHP 55.099. * Means significant at 1% level of probability using t-test.bn.a.= not applicable.

355

Table 7. Estimated number of farmers adopting AWD in the Philippines by region as of May 2011.

AWD adoption* Region Area (ha) Number of

farmers Cordillera Autonomous Region (CAR)

10,910 10,888

Region 1 4,099 10,102 Region 2 3,312 4,941 Region 3 26,652 20,938 Region 6 195 147 Region 7 8,232 7,577 Region 11 27,853 17,294 Region 12 11,760 9,800 Total 93,014 81,687 * Estimates were taken from national irrigation systems with ongoing wide-scale

implementation of AWD, and with ongoing demonstration trials and training activities. Source: Lampayan 2011.

356

Figure 1. Locations of P-38 and other deep-well pump systems (DPWs) in Tarlac, Philippines.

NU EVA EC IJA

PAN GA SINA N

Baka y Rive rTo C amiling

To C amilingTar lac R

i ve r

To C a milin g

Bu lsa River

O’ D

o nn e

l Riv

er

TA RL AC

NIA -PM O fo r T GIS RP

LA PA Z

To Zarago sa

Ba mb an R iver

CO NC EP CIO N

Bam ban

To M a n ila

PA M PA NG A

Ch i

co R

i ver

Tal av era

Riv er

TA RLA C

SA N M A NUE L

To U rd an eta/B ag uio

M ONC A DA

PA NI Q UI

TG-05

TG-02

TG-5 7

TG-29

TG-08

P-38 TG-0 9TG-12P -3 7

TG-11

TG-01

TG-14

TG-2 4

TG -1 0TG-1 3

RA M O S

PUR A

GE RO NA

VIC TO RIA

To G uimb a

Bal oy R

iver

To Guim ba

0 5 1 0 1 5 2 0 K m

N

R oads

R iv ers

Provincia l boundary

D eepw el l sy s te m

Le ge nd

NU EVA EC IJA

PAN GA SINA N

Baka y Rive rTo C amiling

To C amilingTar lac R

i ve r

To C a milin g

Bu lsa River

O’ D

o nn e

l Riv

er

TA RL AC

NIA -PM O fo r T GIS RP

LA PA Z

To Zarago sa

Ba mb an R iver

CO NC EP CIO N

Bam ban

To M a n ila

PA M PA NG A

Ch i

co R

i ver

Tal av era

Riv er

TA RLA C

SA N M A NUE L

To U rd an eta/B ag uio

M ONC A DA

PA NI Q UI

TG-05

TG-02

TG-5 7

TG-29

TG-08

P-38 TG-0 9TG-12P -3 7

TG-11

TG-01

TG-14

TG-2 4

TG -1 0TG-1 3

RA M O S

PUR A

GE RO NA

VIC TO RIA

To G uimb a

Bal oy R

iver

To Guim ba

0 5 1 0 1 5 2 0 K m

N

R oads

R iv ers

Provincia l boundary

D eepw el l sy s te m

Le ge nd

357

Figure 2. AWD bulletin to guide NARES partners and farmers in their AWD dissemination activities.

358

Figure 3. Relationship between soil water potential and field water depth (adopted from Cabangon et al 2008).

H u i b e i I R R I

Field water depth (cm)

Soil water potential (kPa) at 15 cm soil depth

Decreasing soil moisture Moist soil Dry soil

AWD with yield penalty

Safe AWD

Saturated soil -60

-45 -30

-30

-20

-25

-10

-20

-15

-10

-5

0

0 -50

-40

-40

-35

359

Figure 4. Grain yield (A), total water used (B), and example field water status (C) in AWD and FP plots in P-38 during 2002 and 2003DS (adopted from Lampayan et al 2004).

0200400600800

100012001400

Upper

Mid

dle

Low

er

Avera

ge

Upper

Mid

dle

Low

er

Avera

ge

AWDFP

2002 dry season 2003 dry season

P-38, Canarem

Wate

r use (

mm

)

0123456789

Upper

Mid

dle

Low

er

Ave

Upper

Mid

dle

Low

er

Ave

AWDFP

2002 dry season 2003 dry season

P-38, Canarem

Gra

in y

ield

(t/

ha)

(a) (b)

-100-90-80-70-60-50-40-30-20-10

01020

12-J

an

19-J

an

26-J

an

2-F

eb

9-F

eb

16-F

eb

23-F

eb

2-M

ar

9-M

ar

16-M

ar

23-M

ar

30-M

ar

6-A

pr

FPAWD

Perc

hed w

ate

r depth

(cm

)

(c)

360

Figure 5. Research-to-impact pathway networks for AWD in the Philippines.

Basic researchApplied research

Participatory R&D

Knowledge sharing

Dissemination

Farmer uptake

Key institutions

ImpactTechnologies

Adaptation

NARES uptake

“Poverty”

alleviation

Dissemination in other countries (Bangladesh, Vietnam, etc)

Workshops, meetings

IRRI, NIA, PhilRice

Included in training programs

Policy supportAdmin Order passed by Dep Agric Secretary

TTWS (2002-2004)Techno-demo

Adoption study

(done in the 80s and 90s)

Private sector (NGOS)Local authorityExtensionNGOs

Training

Field Water tube

Development of extension materials

Palay Check

361

Paper 15 The rice root-knot nematode Meloidogyne graminicola: a new challenge for water-saving rice production systems in Asia Dirk De Waele, Zin Thu Zar Maung, Pa Pa Win, Pyone Pyone Ki, Yi Yi Myint, Luzviminda Fernandez, Teodora Cabasan, Judith Galeng, Bas Bouman, Casiana Vera-Cruz, and Arvind Kumar

The rice root-knot nematode Meloidogyne graminicola is considered one of the most important pathogens of rice, especially in South and Southeast Asia, in a range of rice production systems. Recently, this nematode species has also been identified as a major causal agent of yield decline and even yield failure in tropical aerobic rice. Meloidogyne graminicola has a short life cycle (only 3 weeks in many regions). In a susceptible rice cultivar, which has for instance a crop cycle of 120 days under optimal conditions, it can produce up to six generations during a single crop cycle, resulting in a high nematode population density and severe damage to the rice plants. Water regime is an extremely important environmental factor that influences the incidence, development, and population dynamics of this nematode species, and the damage and yield loss it can cause to rice. Options to manage M. graminicola are limited. Crop rotation, flooding, and the use of nematicides are practices that are often used to manage plant-parasitic nematodes in infested fields. The feasibility of controlling M. graminicola by crop rotation has been reported to be limited because of the wide host range of this nematode species. Flooding effectively inactivates or kills most soil nematodes but is not a nematode management option in areas where water scarcity is experienced. Chemical soil sterilization is effective but expensive and environmentally harmful. Moreover, the application of many nematicides has been or is being banned. Resistance to M. graminicola has been identified in Oryza longistaminata and in African rice (O. glaberrima) but the introgression of resistance into Asian rice (O. sativa) has not been very successful so far as the interspecific progenies do not express the same degree of resistance observed in O. glaberrima. The emergence of M. graminicola on rice is a prime example of how the looming water shortage for irrigated lowland rice production in Asia drives change (a combination of agricultural, environmental, socioeconomic, and policy change) that in turn affects the pest status of a plant-parasitic nematode. If effective management strategies (based on an integrated combination of adapted water regimes and crop rotation sequences unfavorable for the build up of high M. graminicola populations, and the use of M. graminicola-resistant or -tolerant rice cultivars) are not developed, the rice root-knot nematode may become a threat to the sustainability of rice cultivation in water-saving production systems in Asia.

Nematodes (phylum Nematoda) are unsegmented, thread-like (round) worms. Their natural habitat includes terrestrial and aquatic ecosystems ranging from mountains to oceans. However, for these organisms, it is crucial that their body always be surrounded by a thin film of water for their mobility and survival. Nematodes are highly adaptive and diverse. They are the most numerous metazoans on Earth: four out of every five organisms are nematodes (Bird and Kaloshian 2003). Recently, a nematode species (Halicephalobus mephisto) has been found in Earth’s terrestrial deep subsurface, expanding the known metazoan biosphere (Borgonie et al 2011). Nematodes are either free-living or parasites of plants and animals. Some plant-parasitic nematode species can damage plants aboveground but the majority of the plant-parasitic nematode species are root pathogens. Most of the plant-parasitic nematodes are smaller than 1 mm. All plant-parasitic nematodes have four juvenile stages. The second-stage juveniles (J2s) usually develop inside the eggs, hatch, and infect host plants.

362

Based on their life cycle and feeding habit, plant-parasitic nematodes can be classified as (1) ectoparasites that live completely outside the plant and feed on the outer cell layers of the roots with the aid of a stylet, (2) semi-endoparasites that enter the roots partially and feed on the inner cell layers of the roots while a part of the nematode’s body remains outside in the soil, and (3) endoparasites that penetrate the root completely and feed on the cortical and/or vascular cells of the roots. Endoparasitic nematodes can be further subdivided into two types: (1) migratory endoparasitic nematodes, of which the juveniles and adults remain mobile and continuously feed on the cortical cells of the roots as they migrate inside the roots; and (2) sedentary endoparasitic nematodes, of which the females become sessile after inducing specialized feeding cells on which they exclusively feed. This last category includes the cyst and root-knot nematodes, which can have an enormous impact on agriculture worldwide. Rice nematodes Rice is being cultivated in a variety of agroecosystems and, as a result, a high diversity of plant-parasitic nematodes infects rice. More than 200 species of plant-parasitic nematodes have been found associated with rice (Gerber et al 1987, Bridge et al 2005) but, as is the case in most agricultural crops, only a limited number of these species are of economic importance and can cause damage and yield loss (Prot and Rahman 1984, Bridge et al 2005). The most important rice nematodes are root-knot nematode species (Meloidogyne spp.), which cause root gall formation; the foliar nematodes Aphelenchoides besseyi and Ditylenchus angustus, the causal agents of white tip and ufra diseases, respectively; and some species belonging to the migratory endoparasitic genera Hirschmanniella and Pratylenchus. All of these nematodes can cause mechanical damage and/or malfunctions of the physiological processes involved in plant growth, development, and reproduction, resulting in poor growth and yield loss of rice (Bridge et al 2005).

In general, economically important nematode species differ in lowland and upland rice, and in irrigated and rainfed rice, but some species can cause damage in all rice agroecosystems while other species are more restricted (Prot and Rahman 1984, Bridge et al 2005). In deepwater rice, D. angustus and Meloidogyne graminicola (Golden and Birchfield 1965) are the predominant plant-parasitic nematode species in addition to A. besseyi, while in irrigated rice Hirschmanniella spp., especially H. oryzae, are omnipresent. In upland rice, Meloidogyne spp. and Pratylenchus spp. have the greatest potential to cause damage (De Waele and Elsen 2007).

The population densities of plant-parasitic nematodes, including M. graminicola, fluctuate throughout the year. The most important factors responsible for these fluctuations are various soil factors such as soil moisture, soil temperature, and soil pH; climatic factors such as rainfall and air temperature; host-plant growth stage; and crop cycle duration (Yeates 1973, Wallace 1973, Robbins and Baker 1974, Yeates and Risk 1976, Trudgill and Phillips 1997, Kyi et al 2001, Siddiqui 2007, Giat et al 2008). A good knowledge of the population dynamics of a pathogenic nematode during and in between the crop cycle(s) may assist in making management decisions and in developing new or more effective management practices (Trudgill and Phillips 1997, Giat et al 2008).

Soil sickness of aerobic rice and Meloidogyne graminicola

363

Continuous cropping of aerobic rice in the same field can result in yield decline (Peng et al 2006), rapid yield losses (George et al 2002, Pinheiro et al 2006), and yield failure (Kreye et al 2009c). Many factors, such as nutrient deficiencies, soil alkalinity, and the buildup of root-infecting pathogens (fungi and nematodes), are apparently involved in the soil sickness of aerobic rice, but the complex interactions between these harmful abiotic and biotic factors make it difficult to identify the primary agent(s) involved (Kreye et al 2009 a, b). However, there are more and more indications that M. graminicola is most likely one of the most important factors affecting growth and yield adversely in tropical aerobic rice. One of these indications is the remarkable increase in frequency of occurrence and population densities of M. graminicola, which the shift from paddy rice to rice cultivation with less water has caused. The effect of change on Meloidogyne graminicola The case of the emergence of M. graminicola on rice in South and Southeast Asia is a prime example of how the looming water shortage for irrigated lowland rice production in Asia drives change (a combination of agricultural, environmental, socioeconomic, and policy change) that in turn affects the pest status of a plant-parasitic nematode. Interestingly, several authors (such as Prot 1994 and Reversat et al 2003) have predicted the emergence of M. graminicola as a new major pest of rice in Asia. Because of a combination of socioeconomic and environmental (climate) changes, South and Southeast Asia are increasingly experiencing water shortages. Water shortage is not only increasing the cost of rice production in Asia but is also severely limiting the yield of rice, thus threatening food security. As a consequence, rice production, especially in the lowland, has raised international concern since the traditional paddy production system consumes a high amount of water and in many areas of Asia the water requirement is too high to allow for this type of rice production to be continued in the future. It takes 3,000 to 5,000 liters of water to produce 1 kg of rice, which is two to three times more than to produce 1 kg of other cereals such as wheat or maize. This is due to the high unproductive water loss (>80% of water applied) by evaporation, surface runoff, etc. Furthermore, concurrent water demand by the urban population and industry forces legal restrictions on the use of water for agricultural purposes; thus, the arable land allowed for the cultivation of lowland rice with its inherently high water demand is decreasing. In Asia, out of a total area of 79 million hectares of irrigated paddy rice, 17 million ha may experience physical water scarcity and 22 million ha economic water scarcity by 2025 (Tuong and Bouman 2003). This is why water-saving rice production systems, such as direct wet or dry seeding, intermittent irrigation, alternate wetting and drying, the cultivation of aerobic rice varieties, etc., are being developed and increasingly used. However, more and more observations indicate that the large-scale introduction of these systems is favoring the development of high M. graminicola population densities, thus drastically increasing their economic significance. Meloidogyne graminicola Meloidogyne graminicola was first described in 1965 from grasses and oats in Louisiana in the United States (Golden and Birchfield 1965). It has since been found on rice, mainly in South and Southeast Asia (India, Nepal, Pakistan, Bangladesh, Sri Lanka, Myanmar, Malaysia, Singapore, Thailand, Vietnam, Laos, Indonesia, the Philippines, and Taiwan), and also on wheat grown in rice-wheat cropping area (Sharma 2001). Recently, Jain et al (2012) documented the geographic distribution of

364

M. graminicola on rice in India and reported that during the past decade this nematode species had spread to numerous new rice-producing areas in the country. Meloidogyne graminicola is equally prevalent on lowland (irrigated), upland (rainfed), and deepwater rice. It causes swellings and galls throughout the root system. Infected root tips become swollen and hooked, a symptom that is characteristic for this nematode species (Figure 1).

In upland conditions and shallow intermittently flooded land, all Meloidogyne species can cause severe growth reduction, chlorosis, wilting, reduced tillering, unfilled spikelets, and poor yield (Babatola1984), but M. graminicola is considered by far the most damaging Meloidogyne species on rice (Figure 2). Meloidogyne graminicola appears to be exceptionally well adapted to flooded conditions, thus enabling it to continue multiplying in host tissues even when the roots are deeply covered in water. Second-stage juveniles (J2s) invade rice roots in upland conditions just behind the root tip (Rao and Israel 1973). They cannot invade rice under flooded conditions but quickly invade when infested soils are drained (Manser 1968). Females developed within the roots and eggs are laid mainly in the cortex (Roy 1976) as in most other Meloidogyne species. However, the J2s of M. graminicola can remain in the maternal gall or migrate intercellularly through the aerenchymatous tissues of the cortex to new feeding sites within the same root (Bridge and Page 1982). This behaviour is part of its adaptation to flooded conditions. The life cycle duration of M. graminicola varies considerably in different environments, ranging from a very short life cycle of only 15 days at 27 to 37°C (Jaiswal and Singh 2010) to a rather long life cycle of up to 51 days in some regions in India (Rao andIsrael 1973). A study at IRRI showed that the life cycle duration of an M. graminicola population from the Philippines was 20 days at 26 to 29°C under nonflooded conditions and 19 days at both 26 to 29°C and 32 to 36°C under flooded conditions, which is similar to the life cycle duration of an M. graminicola population from Bangladesh at 22 to 29°C (Bridge and Page 1982). Although higher temperatures can affect the rate of nematode development, the study shows that temperature did not substantially reduce or increase the rate of development and reproduction of M. graminicola. This observation may explain the wide distribution of M. graminicola in a wide range of rice-based agroecosystems in Asia. It also indicates that higher temperature due to climate change will have no effect on the incidence of this nematode species. The short life cycle duration also demonstrates the reproductive and damage potential of M. graminicola: in a susceptible rice cultivar, such as UPLRi-5, which has a crop cycle of 120 days under optimal conditions, it can produce up to six generations during a single crop cycle, resulting in a high nematode population density and, as a result, severe damage to the rice plants. Although numbers of M. graminicola decline rapidly after 4 months, some egg masses and J2s can remain viable for at least 5 to 14 months in waterlogged soils (Roy 1982, Bridge and Page 1982). Meloidogyne graminicola has a wide host range, which includes many of the common weeds of rice fields (MacGowan and Langdon 1989), and it can also be damaging to agricultural crops that are grown in rotation with rice such as onion (Gergon et al 2001, 2002). The effect of water regime on Meloidogyne graminicola

365

The water regime is a very important environmental factor that influences the incidence, development, and population dynamics of M. graminicola, and the damage and yield loss it can cause to rice. The yield response of a rice variety to this nematode is affected by whether this crop is cultivated in flooded or upland (i.e., free draining, without surface-water accumulation) conditions (Tandingan et al 1996). In upland conditions and shallow intermittently flooded land, M. graminicola is considered to be the most damaging Meloidogyne species on rice (De Waele and Elsen 2007). In M. graminicola-infested lowland rainfed rice in Bangladesh, nematicide application resulted in a yield increase of 16% to 20% or about 1 tha−1 (Padgham et al 2004). In M. graminicola-infested upland rice fields in Thailand, nematicide application resulted in a yield increase of 12% to 33% (Arayarungsarit 1987), and in Indonesia a 28% to 87% yield increase (Netscher and Erlan 1993). Greenhouse studies have also shown the importance of the interaction between M. graminicola and the water regime. Soriano et al (2000) reported that yield losses ranged from 11% to 73% in simulated intermittently flooded conditions. In simulated upland conditions, yield losses ranged from 20% to 80% (Plowright and Bridge 1990, Prot and Matias 1995, Tandingan et al 1996). The effects of the water regime and nematodes on rice plants are illustrated by three studies recently carried out in Myanmar. During 2007, Maung et al (2010) carried out a monsoon rice nematode survey (May-October).A total of 539 soil samples and 539 root samples were collected from 11 monsoon rice cultivars in 12 regions of Myanmar. All regions surveyed and 90% of the 539 monsoon rice field samples were infested with the rice root nematode H. oryzae but, remarkably, M. graminicola was not found. During 2009, Win et al (2011) carried out two rice nematode surveys. The first survey was conducted during the dry summer season (mid-January to mid-May) in ten regions representing the summer-irrigated lowland rice ecosystem in the lower Ayeyarwady delta area of Myanmar. The second survey was conducted during the rainy (monsoon) season (mid-May to mid-October) in three regions representing the rainfed upland rice ecosystem in the northern hilly area of Myanmar. In total, 552 soil samples and 552 root samples were collected from 15 locally cultivated rice cultivars in lowland and upland rice ecosystems. Seventy-eight percent of the 450 lowland rice fields and 9% of the 102 upland rice fields sampled were infested with M. graminicola. The number of M. graminicola and the root galling index (on a scale of 0 to 5) averaged 867 J2/3 g roots and 4.1, respectively, in lowland rice and 11 J2/3 g roots and 1.2, respectively, in upland rice. In lowland rice, the frequency of occurrence of M. graminicola was higher in delayed irrigation than in early irrigation (87% vs. 54%, respectively). The root galling index averaged 4.5 in delayed irrigation and 1.2 in early irrigation. The rice root nematode H. oryzae was also detected during the lowland rice survey but not during the upland rice survey. It was more frequently observed in lowland rice fields under early irrigation than in delayed irrigation (22.5% vs. 7.7%, respectively). The authors concluded that M. graminicola was unquestionably a major pest of the summer-irrigated lowland rice agroecosystem in Myanmar and that the combination of direct wet seeding, delayed irrigation, rice monoculture, high cropping intensity, and the use of high-yielding rice varieties may be responsible for the prevalence of M. graminicola. From December 2009 until December 2010, the population dynamics of M. graminicola and H. oryzae in a double rice cropping sequence (rainfed monsoon rice-irrigated summer rice) was monitored in lowland rice cultivated in the Ayeyarwady river delta in Myanmar, in a naturally infested field, on rice cultivars Yatanartoe (an

366

irrigated cultivar) and Taungpyang (a rainfed lowland cultivar). During the summer-irrigated rice-growing season, the population density of M. graminicola J2s in roots showed two distinct peaks: at the maximum tillering stage of the rice plants in January and at the heading stage in March 2010. With the onset of the monsoon rain, the J2 population density in the roots of ratoon rice plants gradually decreased in May. During the rainfed monsoon rice-growing season, very low population densities of M. graminicola J2s were detected in the roots of the rice plants while the root population density of H. oryzae juveniles and adults showed two distinct peaks: at the maximum tillering stage of the rice plants in August and at the heading stage in October 2010. With the onset of the dry season, the population density of H. oryzae in the roots reached the lowest density at harvest in November. Root galling caused by M. graminicola followed the same trend as the J2 population densities throughout the summer-irrigated rice-growing season. In contrast, no root galls were observed during the rainfed monsoon growing season. This population dynamics study clearly shows that flooding before and just after planting of rice plants can prevent penetration of rice roots by M. graminicola J2s. Host response of rice genotypes to M. graminicola under aerobic conditions Following the terminology of Bos and Parlevliet (1995), resistance/susceptibility on the one hand and tolerance/sensitivity on the other hand are defined as independent, relative qualities of a host plant based on comparison between genotypes. A host plant may either suppress (resistance) or allow (susceptibility) nematode development and reproduction; it may suffer either little injury (tolerance), even when heavily infected with nematodes, or much injury (sensitivity), even when relatively lightly infected with nematodes. Resistance/susceptibility can be determined by measuring the nematode population densities in and on the roots, whereas tolerance/sensitivity can be determined by measuring the effect of the nematode population on plant growth, yield-contributing traits, and/or yield (Cook and Evans 1987). In a study at IRRI (De Waele et al, submitted), the host response of 19 aerobic, 7 upland, and 4 lowland rice genotypes that are either being used in IRRI’s aerobic rice breeding program or are already cultivated by farmers in Asia was evaluated under aerobic soil conditions in an outdoor raised-bed experiment (Table 1). This study showed a large variation in susceptibility and sensitivity to M. graminicola infection among the genotypes examined. Resistance comparable with that of the resistant reference genotypes included in the study (CG14, TOG5674, TOG7235) was not found, but, in terms of susceptibility, the upland genotype Morobereken and the aerobic genotypes IR78910-23-1-3-4 and WAB638-1 may be less susceptible than the susceptible reference genotypes and other genotypes included in the study (Table 2). The resistant or less susceptible rice genotypes not only alleviate the problem caused by M. graminicola but may also prevent yield reduction caused by M. graminicola. In this respect, excluding the reference genotypes, six rice genotypes with low yield reduction under M. graminicola infection that can already be used by farmers were identified in the study (in decreasing order of yield): IR78877-208-B-1-2, WAB450-24-2-3-P-38-1-HB, CT6510-24-1-2, UPLRi-7, IRAT216, and Azucena. Of these, the high-yielding genotypeIR78877-208-B-1-2 is the most interesting and it should be grown in areas where M. graminicola is a problem (Figure 3).

367

Also interesting are the genotypes that combine high yield with less susceptibility and less sensitivity, such as IR78910-23-1-3-4. Except for the above-mentioned tolerant genotypes, the results of the study also allow the identification of rice genotypes that are either highly susceptible or hypersensitive to M. graminicola infection. Excluding the reference genotypes, cultivation of highly susceptible genotypes (for instance, upland rice genotype Aus196), hypersensitive upland rice genotype Dinorado as well as aerobic genotypes B6144F-MR-6, Way Rarem, and WAB638-1 should be avoided infields infested with M. graminicola. Use of these genotypes will result in increased populations of M. graminicola and will have adverse effects on crop production. On average, M. graminicola caused an almost 30% reduction in yield. Excluding the two susceptible and three resistant reference genotypes included in the experiment, the most-affected trait was dry shoot biomass (23.6% reduction), followed by root length, which was more affected than fresh root weight (19.8% vs. 8%) and grain filling (17.3%) while plant height and the number of spikelets panicle−1 were less affected (10.2% and 8.1%, respectively). Neither tillering nor the number of panicles plant−1 were affected (Figure 4). Management of Meloidogyne graminicola Options to manage the population density of M. graminicola are now limited. Crop rotation, flooding, and the use of nematicides are practices that are often used to manage plant-parasitic nematodes in infested fields. Rotating upland rice with mung bean (Ventura et al 1981), mustard, sesame, millet, or guzitil (Guizotia abbysinica) can effectively decrease the population of M. graminicola and reduce yield losses (Rahman 1990) but the feasibility of controlling M. graminicola by crop rotation has been reported to be limited because of the wide host range of this nematode species. Flooding effectively inactivates or kills most soil nematodes but is of course not a nematode management option in areas where water scarcity is experienced. Chemical soil sterilization is effective but expensive and environmentally harmful. Moreover, options have been few for chemical control of nematodes since the application of DBCP (1, 2-di-bromo-3-chloropropane) and EDB (ethylene di-bromide) was banned (Boerma and Hussey 1992). In this context, the search for rice genotypes that are either resistant to or tolerant of M. graminicola may offer an alternative to manage this nematode species in tropical aerobic rice. Resistance to M. graminicola has been identified in Oryza longistaminata and, although M. graminicola has not been found in African rice (O. glaberrima) so far (Plowright et al 1999, Soriano et al 1999), the introgression of resistance into Asian rice (O. sativa) has not been very successful as the interspecific progenies do not express the same degree of resistance observed in O. glaberrima (Plowright et al 1999). Differences in host response to M. graminicola infection were also observed among Asian rice (O. sativa) genotypes (Jena and Rao 1977a, b, Sharma-Poudyal et al 2004, Bridge et al 2005, Prasad et al 2006). Conclusions It is obvious from all the evidence collected so far that there is a relationship between the incidence of M. graminicola and the use of less water for rice cultivation, which leads to yield loss. New water-saving rice production systems, such as direct wet or dry seeding, intermittent irrigation, alternate wetting and drying, and the cultivation of aerobic rice varieties, are being developed and increasingly applied on a large scale

368

in Asia. The effect of the implementation of these systems on the incidence of M. graminicola is largely unknown but, if effective management strategies (based on an integrated combination of adapted water regimes and crop rotation sequences unfavorable for the build up of M. graminicola population densities, and the useof M. graminicola-resistant or -tolerant rice cultivars) are not developed, M. graminicola may become a threat to the sustainability of rice cultivation in water-saving production systems in Asia. References Arayarungsarit L. 1987. Yield ability of rice varieties in fields infested with root-knot

nematode. Int. Rice Res. Newsl. 4:14. Babatola JO. 1984. Rice nematode problems in Nigeria: their occurrence, distribution

and pathogenesis. Trop. Pest Manage. 30:256-265. Bird D, Kaloshian I. 2003. Are roots special? Nematodes have their say. Physiol. Mol.

Plant Pathol. 62:115-123. Boerma HR, Hussey RS. 1992. Breeding plants for resistance to nematodes. J.

Nematol. 24:242-252. Borgonie G, Garcia-Moyano A, Litthauer D, Bert W, Bester A, Van Heerdeb E, Moller

C, Erasmus M, Onstott TC. 2011. Nematoda from the terrestrial deep subsurface of South Africa. Nature 474:79-82.

Bos L, Parlevliet JE. 1995. Concepts and terminology on plant/pest relationships: towards consensus in plant pathology and crop protection. Annu. Rev. Phytopathol. 33:69-102.

Bridge J, Page SLJ. 1982. The rice root-knot nematode, Meloidogyne graminicola, on deep water rice (Oryza sativa subsp. indica). Rev. Nématol. 5:225-232.

Bridge J, Plowright RA, Peng D. 2005. Nematode parasites of rice. In: Luc M, Sikora RA, Bridge J, editors. Plant parasitic nematodes in subtropical and tropical agriculture. Wallingford (UK): CAB International. p 87-130.

Cook R, Evans K. 1987. Resistance and tolerance. In: Brown RH, Kerry BR, editors. Principles and practice of nematode control in crops. Sydney (Australia): Academic Press. p 179-231.

De Waele D, Das K, Zhao D, Tiwari RKS, Shrivastava DK, Vera-Cruz C, Kumar A. Submitted. Host response of rice genotypes to the rice root-knot nematode (Meloidogyne graminicola) under aerobic soil conditions. J. Phytopathol.

De Waele D, Elsen A. 2007. Challenges in tropical plant nematology. Ann. Rev. Phytopathol. 45:457-485.

George T, Magbanua R, Garrity DP, Tubaña BS, Quiton J. 2002. Rapid yield loss of rice cropped successively in aerobic soil. Agron. J. 94:981-989.

Gerber K, Smart GCJ, Esser RP. 1987. A comprehensive catalogue of plant parasitic nematodes associated with aquatic and wetland plants. Gainesville, Florida, Agricultural Experimental Station, Institute of Food and Wetland Sciences, University of Florida.

Gergon EB, Miller SA, Davide RG. 2001. Occurrence and pathogenicity of the rice root-knot nematode (Meloidogyne graminicola) and varietal reaction of onion (Allium cepa). Philipp. Agric. Sci. 84:43-50.

Gergon EB, Miller SA, Halbrendt JM, Davide RG. 2002. Effect of the rice root-knot nematode on growth and yield of yellow granex onion. Plant Dis. 86:1339-1344.

Giat E, Kaspi R, Anderson CA, Westerdhal BB. 2008. Seasonal population dynamic of the plant-parasitic nematode, Anguina pacificae, on golf course putting greens in California. J. Nematol. 40:252-257.

369

Golden AM, Birchfield W. 1965. Meloidogyne graminicola (Heteroderidae), a new species of root-knot nematode from grass. Proc. Helminthol. Soc. Wash. 32:228-231.

Jain RK, Khan MR, Kumar V. 2012. Rice root-knot nematode (Meloidogyne graminicola) infestation of rice. Arch. Phytopathol. Plant Prot. 45:635-645.

Jaiswal PK, Singh KP. 2010 A technique for studying the life cycle of Meloidogyne graminicola in rice roots. Int. Rice Res. Notes35:1-3.

Jena RN, Rao YS. 1977a. Nature of resistance in rice (Oryza sativa L.) to the root-knot nematode (Meloidogyne graminicola Golden and Birchfield). I. Mechanisms of resistance. Proc. Indian Acad. Sci. 8:31-38.

Jena RN, Rao YS. 1977b. Nature of resistance in rice (Oryza sativa L.) to the root-knot nematode (Meloidogyne graminicola Golden and Birchfield). II. Histopathology of nematode infection in rice varieties. Proc. Indian Acad. Sci. 8:87-91.

Kreye C, Bouman BAM, Castañeda AR, Lampayan RM, Faronilo JE, Lactaoen AT, Fernandez L. 2009a. Possible causes of yield failure in tropical aerobic rice. Field Crops Res. 111:197-206.

Kreye C, Bouman BAM, Faronilo JE, Llorcaz L.2009b. Causes for soil sickness affecting early plant growth in aerobic rice. Field Crops Res. 114:182-187.

Kreye C, Bouman BAM, Reversat G, Fernandez L, Vera Cruz C , Elazegui F, Faronilo JE, Llorca L. 2009c. Biotic and abiotic causes of yield failure in tropical aerobic rice. Field Crops Res. 112:97-106.

Kyi PP, Long PG, Mercer CF. 2001. The population dynamics of Heterodera trifolii in a manawatu pasture. N. Zeal. Plant Prot. 54:136-140.

MacGowan JB, Langdon KR. 1989. Hosts of the rice root-knot nematode, Meloidogyne graminicola. Gainesville, Fla. (USA), Nematology Circular No. 172.

Manser PD. 1968. Meloidogyne graminicola, a cause of root-knot of rice. Plant Prot. Bull. 16:11.

Maung ZTZ, Kyi PP, Myint YY, Lwin T, De Waele D. 2010. Occurrence of the rice root nematode Hirschmanniella oryzae on monsoon rice in Myanmar. Trop. Plant Pathol. 35:3-10.

Netscher C, Erlan X. 1993. A root-knot nematode, Meloidogyne graminicola, parasitic on rice in Indonesia. Afro-Asian J. Nematol. 3:90-95.

Padgham JL, Duxbury JM, Mazid AM, Abawi GS, Hossain M. 2004. Yield losses caused by Meloidogyne graminicola on lowland rainfed rice in Bangladesh. J. Nematol. 36:42-48.

Peng S, Bouman BAM, Visperas RM, Castañeda A, Nie L, Park HK. 2006. Comparison between aerobic and flooded rice in the tropics: agronomic performance in an eight season experiment. Field Crops Res. 96:252-259.

Pinheiro BDS, Castro EDMD, Guimaraes CM. 2006. Sustainability and profitability of aerobic rice production in Brazil. Field Crops Res. 97:34-42.

Plowright RA, Bridge J. 1990. Effect of Meloidogyne graminicola (Nematoda) on the establishment, growth and yield of rice cv. IR 36. Nematology 36:81-89.

Plowright RA, Coyne DL, Nash P, Jones MP. 1999. Resistance to the rice nematodes Heterodera sacchari, Meloidogyne graminicola and M. incognita in Oryza glaberrima and O. glaberrima x O. sativa interspecific hybrids. Nematology 1:745-751.

Prasad R, Vijayakumar CHM, Sankar M, Varaprasad KS, Srinivasa M, Kondala Rao Y. 2006. Root-knot nematode resistance in advanced back cross populations of rice developed for water stress conditions. Nematol. Mediterr. 34:3-8.

370

Prot JC, Matias DM. 1995. Effects of water regime on the distribution of Meloidogyne graminicola and other root-parasitic nematodes in a rice field toposequence and pathogenicity of M. graminicola on rice cultivar UPLRi-5. Nematologica 41:219-228.

Prot JC, Rahman ML. 1984. Nematode ecology, economic importance, and management in rice ecosystems in South and Southeast Asia. In: Teng PS, Heong KL, Moody K, editors. Rice pest science and management. Los Baños (Philippines): International Rice Research Institute. p 129-144.

Prot JC. 1994. Effects of economic and policy changes on the status of nematode rice pests in Vietnam and the Philippines. Fund. Appl. Nematol. 17:195-198.

RahmanML.1990. Effect of different cropping sequences on the root-knot nematode, Meloidogyne graminicola, and yield of deepwater rice. Nematol. Mediterr.18:213-217.

Rao YS, Israel P. 1973. Life history and bionomics of Meloidogyne graminicola, the rice root-knot nematode. Indian Phytopathol. 26:333-340.

Reversat G, Bouman B, Fernandez L, Lampayan R, Castañeda A. 2003. How water shortage in irrigated Asian rice production affects the potential threat of plant parasitic nematodes. AFPP-Seventh international conference on plant diseases, Tours, France, 3-5 December 2003. 8 p.

Robbins RT, Baker KR. 1974. The effects of soil type, particle size, temperature, and moisture on reproduction of Belonolaimus longicaudatus. J. Nematol. 6:1-6.

Roy AK. 1976. Pathological effects of Meloidogyne graminicola on rice and histopathological studies on rice and maize. Indian Phytopathol. 29:359-362.

Roy AK. 1982. Survival of Meloidogyne graminicola eggs under different moisture conditions in vitro. Nematol. Mediterr. 10:221-222.

Sharma SB. 2001. Plant parasitic nematodes in rice-wheat based cropping system in South Asia: present status and future research thrust. J. Crop Prod. 4:227-247.

Sharma-Poudyal D, Pokharel RR, Shrestha SM, Khatri-Chhetri GB. 2004. Evaluation of common Nepalese rice cultivars against rice root-knot nematode. Nepal Agric. Res. J. 5:33-36.

Siddiqui MA. 2007. Seasonal fluctuation in nematode population associated with mango, Mangifere indica L. Arch. Phytopathol. Plant Prot. 40:389-394.

Soriano IR, Prot JC, Matias DM. 2000.Expression of tolerance for Meloidogyne graminicola in rice cultivars as affected by soil type and flooding. J. Nematol. 32:309-317.

Soriano IR, Schmit V, Brar DS, Prot JC, Reversat G. 1999. Resistance to rice root-knot nematode Meloidogyne graminicola identified in Oryza longistaminata and Oryza glaberrima. Nematology 1:395-398.

Tandingan I, Prot JC, Davide RG. 1996. Influence of water management on tolerance of rice cultivars for Meloidogyne graminicola. Fund. Appl. Nematol. 19:189-192.

Trudgill DL, Phillips MS. 1997. Nematode population dynamics, threshold levels and estimation of crop losses. FAO Plant Prod. Prot. Paper 144:45-58.

Tuong TP, Bouman BAM. 2003. Rice production in water–scarce environments. In: Kijne JW, Barker R, Molden D, editors. Water productivity in agriculture: limits and opportunities for improvement. Wallingford(UK): CAB International. p 53-67.

Ventura W, Watanabe I, Castillo MB, Dela Cruz A. 1981. Involvement of nematodes in the soil sickness of a dryland rice-based cropping system. Soil Sci. Plant Nutr. 27:305-315.

Wallace HR. 1973. Nematode ecology and plant disease. London, Edward Arnold; New York, Crane, Russak. 228 p.

371

Win PP, Kyi PP, De Waele D. 2011. Effect of agro-ecosystem on the occurrence of the rice root-knot nematode Meloidogyne graminicola on rice in Myanmar. Aust. Plant Pathol. 40:187-196.

Yeates GW, Risk WH. 1976. Annual cycle of root nematodes on white clover in pasture. III. Heterodera trifolii in a yellow-brown earth. New Zeal. J. Agric. Res. 19:393-396.

Yeates GW. 1973. Annual cycle of root nematodes on white clover in pasture. I. Heterodera trifolii in a yellow-grey earth. New Zeal. J. Agric. Res. 16:569-574.

Notes Authors’ addresses: Dirk De Waele, University of Leuven (KU Leuven), Willem de Crovlaan 42, 3001 Heverlee, Belgium; North-West University, Private Bag X6001, 2520 Potchefstroom, South Africa; and International Rice Research Institute (IRRI), DAPO Box 7777, Metro Manila, Philippines; Zin Thu Zar Maung, Pa Pa Win, Pyone Pyone Ki, and Yi Yi Myint, Plant Protection Division, Department of Agriculture, Ministry of Agriculture and Irrigation, Bayint Naung Road, West Gyogone, P.O. Box 1011, Insein, Yangon, Myanmar; Luzviminda Fernandez, IRRI; Teodora Cabasan, IRRI and University of Southern Mindanao, 9407 Kabacan, Cotabato, Philippines; Judith Galeng, Bas Bouman, Casiana Vera-Cruz, and Arvind Kumar, IRRI. Figure 1. Rice roots infected with Meloidogyne graminicola showing the typical galling

of the root tips (swollen and hooked). Figure 2. Direct-seeded summer-irrigated rice field in the lower Ayeyarwady

delta area of Myanmar infested with Meloidogyne graminicola. Figure 3. Effect of Meloidogyne graminicola infection on the yield of 30 rice

genotypes at maturity grown in uninfested and infested (82 J2 g−1soil) soil under aerobic soil conditions in outdoor raised beds.

Figure 4. Effect of Meloidogyne graminicola infection on the vegetative growth of 30 rice genotypes grown in uninfested (left) and infested (right; 82 J2 g−1 soil) soil under aerobic soil conditions in outdoor raised beds.

372

Table 1. Rice genotypes included in the outdoor raised-bed Meloidogyne graminicola host response study at IRRI. Accession Ecotype Species Apo Aerobic O. sativa Aus196 Upland O. sativa Aus257 Upland O. sativa Azucena Upland O. sativa B6144F-MR-6 Aerobic O. sativa CG14R Aerobic O. glaberrima CT6510-24-1-2 Aerobic O. sativa Dinorado Upland O. sativa IR60080-46A Aerobic O. sativa IR64S Lowland O. sativa IR71525-19-1-1 Aerobic O. sativa IR72 Lowland O. sativa IR78877-163-B-1-1 Aerobic O. sativa IR78877-208-B-1-2 Aerobic O. sativa IR78878-53-2-2-2 Aerobic O. sativa IR78910-23-1-3-4 Aerobic O. sativa IR80508-B-194-3-B Aerobic O. sativa IR80508-B-57-3-B Aerobic O. sativa IRAT216 Aerobic O. sativa Morobereken Upland O. sativa Palawan Upland O. sativa TOG5674R Lowland O. glaberrima TOG7235R Lowland O. glaberrima UPLRi-5S Aerobic O. sativa UPLRi-7 Aerobic O. sativa Vandana Upland O. sativa WAB450-24-2-3-P-38-1-HB Aerobic O. sativa WAB638-1 Aerobic O. sativa WAB880 SG 42 Aerobic O. sativa Way Rarem Aerobic O. sativa SSusceptible reference genotype RResistant reference genotype

373

Table 2. Reproduction of Meloidogyne graminicola on 30 rice genotypes grown under aerobic soil conditions in outdoor raised beds and its effect on root galling and yield reduction at maturity

Root weight (g plant-1)

Yield reduction plant-1 (%)

Gall rating (0-5 scale)3

Nematode population

density J2 plant-1

J2 g-1 roots

Accession U1 I2 i I I x 1,000 x 1,000 Apo 14.5 10.7 28.3 3.1 40.7 4.6 Aus196 19.9 12.1* 24.3 3.4 77.1 6.5 Aus257 13.7 12.3 24.0 3.4 47.8 4.4 Azucena 14.2 14.8 26.1 2.2 75.5 6.9 B6144F-MR-6 12.5 11.8 67.8** 3.8 44.4 3.9 CG14 13.6 14.4 33.1* 0.6 17.9 1.6 CT6510-24-1-2 13.2 13.4 5.5 2.7 35.0 2.7 Dinorado 16.9 15.4 87.8** 3.1 45.3 4.0 IR60080-46A 15.5 10.4 41.6 3.0 15.0 2.1 IR64 17.8 11.0* 60.6 3.7 69.3 5.7 IR71525-19-1-1 15.7 12.6 45.2 4.0 32.0 2.8 IR72 14.2 13.9 26.3 1.9 26.8 2.3 IR78877-163-B-1-1 10.0 8.2 22.8 3.8 27.8 6.6 IR78877-208-B-1-2 15.6 14.6 -0.6 2.6 27.3 2.2 IR78878-53-2-2-2 8.7 7.3 26.2 3.6 22.7 3.0 IR78910-23-1-3-4 13.6 14.2 15.7 2.1 22.8 1.6 IR80508-B-194-3-B 13.2 12.6 40.0** 4.7 30.3 2.5 IR8058-B-57-3-B 13.0 13.8 16.0 2.8 54.4 5.8 IRAT216 12.1 15.3 5.0 3.1 34.1 2.1 Morobereken 21.2 19.2 34.3 2.5 13.6 0.5 Palawan 16.5 15.0 51.8 2.9 37.3 4.9 TOG5674 20.6 9.4* -69.5 0.7 1.0 0.3 TOG7235 21.0 20.8 -4.7 0.7 1.7 0.4 UPLRi-5 14.9 12.4 66.2** 3.5 70.4 8.3 UPLRi-7 10.8 11.2 13.7 2.4 35.7 3.1 Vandana 8.6 8.2 25.6 4.5 37.7 4.9 WAB450-24-2-3-P-38-1-HB 12.1 9.9 6.5 3.5 24.8 3.1 WAB638-1 15.2 13.8 74.4 2.3 15.6 1.6 WAB880 SG 42 12.8 11.2 29.3 3.3 41.8 4.6 Way Rarem 19.7 21.0 57.1** 2.5 48.6 3.1 Trial mean 14.7 13.0 29.4 2.9 35.8 3.5 LSD 0.05 6.7 6.6 1.3 29.8 3.2

374

Figure 1. Rice roots infected with Meloidogyne graminicola showing the typical galling of the root tips (swollen and hooked).

375

Figure 2. Direct-seeded summer-irrigated rice field in the lower Ayeyarwady delta area of Myanmar infested with Meloidogyne graminicola.

376

Figure 3. Effect of Meloidogyne graminicola infection on the yield of 30 rice genotypes at maturity grown in uninfested and infested (82 J2 g-1 soil) soil under aerobic soil conditions in outdoor raised beds.

377

Figure 4. Effect of Meloidogyne graminicola infection on the vegetative growth of 30 rice genotypes grown in uninfested (left) and infested (right; 82 J2 g-1 soil) soil under aerobic soil conditions in outdoor raised beds.

378

Paper 16 Traits for dry direct-seeded rice Nitika Sandhu and Arvind Kumar

Rice is one of the most important food crops in the world. Increasing water scarcity and rising labor costs drive the search for alternative methods of rice cultivation. Dry direct-seeded rice (DDSR) has received much attention over the past few years because of its low-input use. It involves sowing seed into a well-prepared dry soil. The development of early-maturing varieties, improved nutrient management techniques, and increased availability of chemical weed control methods have encouraged many farmers in the Philippines, Malaysia, Thailand, and India to switch from transplanting to the DDSR method of rice cultivation. This shift should substantially reduce crop water requirements, soil organic-matter turnover, nutrient relations, carbon sequestering, weed biota, greenhouse gas emissions, and, most important, the labor requirement. The development of rice varieties for dry direct-seeded conditions can be accelerated by selecting suitable traits. The traits likely to be important for direct-seeded conditions are anaerobic germination, early uniform seedling emergence, early vegetative vigor, seedling-stage cold tolerance (for dry-season rice), root plasticity, proper nutrient uptake under dry conditions, lodging resistance, disease and insect resistance (blast, bacterial leaf blight, brown spot, sheath blight, false smut, brown planthopper, and gall midge),nematode resistance, drought tolerance for rainfed and water-short areas, heat tolerance at the reproductive stage for dry-season rice, preharvest seed dormancy for cyclone-hit areas, no dormancy for rice-rice/rice-wheat areas, and preferred grain quality. This chapter discusses the important traits for improving the adaptation and productivity of rice under direct-seeded conditions and the identification of QTLs with large and consistent effects for traits considered to be beneficial for direct-seeded rice, and suggests likely future patterns of changes in rice cultivation.

Food and water are two of the most important necessities for survival, but, with an increasing demand for food and a looming water crisis, a shortage of both may be on the horizon unless innovative technologies are developed (Parthasarathi et al 2012). In view of the emerging water crisis, it is necessary to develop varieties that have the ability to survive and provide high yield under water-deficit conditions. The increasing scarcity of water has threatened traditional rice cultivation practices globally (Tuong and Bouman 2003). The prediction shows that, among the different constraints, abiotic stresses are more prominent and yield limiting than biotic stresses, and, among the abiotic constraints, water scarcity, drought, and unfavorable temperature are likely to be more devastating to rice yield (Tuong and Bouman 2003). The rice crop is also occasionally subjected to drought due to poor quality of water and inadequate water supply at critical stages (Bouman 2002). India is the largest user of groundwater in the world—more than a quarter of the global total and about 60% of the aquifers in India will be in a critical condition in another 15 years if the trend of indiscriminate exploitation of groundwater continues. Therefore, the sustainability of rice ecosystems and the ability to increase production in pace with population growth with reduced water and labor use are major concerns.

With predictions suggesting that many Asian countries will have severe water problems by 2025, direct-seeded rice gives hope to farmers who do not have access to enough water to grow flooded lowland rice (Parthasarathi et al 2012). The Water-Saving Work Group of the Irrigated Rice Research Consortium (IRRC) is further developing this new technology and making it available to farmers in Asia. Dry direct-seeded rice (DDSR) is a feasible alternative to conventional puddled transplanted rice with good potential to save water, reduce the labor requirement, mitigate

379

greenhouse gas (GHG) emissions, and adapt to climatic risks, and its yield can be comparable with that of transplanted rice if the crop is properly managed (Kumar and Ladha 2011). Still, weed infestations can cause large yield losses in DDSR. Increasing incidences of blast disease, crop lodging, impaired kernel quality, and stagnant yield across the years are additional major challenges. Although the development of suitable varieties and agronomic packages for promoting DDSR is underway (Pathak et al 2011), so far no variety for the rainfed/irrigated ecosystem has been developed that possesses traits specifically needed to produce high yield under dry direct-seeded conditions.

An equally or more important aspect of sustainable rice cultivation is the availability of labor to grow rice under puddled transplanted conditions as the rural population moves away from farming to work in urban areas in industry and other nonfarming activities. Transplanting is a resource- and cost-intensive method since the preparation of seedbeds, the raising of seedlings, and transplanting are labor- and time-intensive operations. Rice cultivation needs to adapt to the mechanization process that requires significantly less human labor, operates at a faster pace, and is more profitable by engaging fewer people per unit of land for cultivation, thereby becoming more profitable to the rural population. DDSR as a rice cultivation system involves a mechanized process from sowing to harvesting that can be carried out quickly with a significantly lower labor requirement. Dry seeding can save as much as 50% of the total labor used on a typical farm relative to transplanting (Pandey and Velasco 2002). However, the current challenge is to decipher the complexities of DDSR adaptation and exploit all available genetic resources to produce rice varieties combining DDSR adaptation with high yield potential, good grain quality, and resistance to biotic stresses (Bouman et al 2007). This involves the development of high-throughput, high-precision phenotyping systems to allow genes for yield components under dry direct-seeded conditions to be efficiently mapped and their effects assessed for a range of traits, and then move the most promising genes into widely grown rice mega-varieties, while scaling up gene detection and delivery for use in marker-aided breeding.

Yield reduction under continuous monocropping has been reported in other upland crops such as mungbean, cowpea, and maize (corn) (Ventura and Watanabe 1978). Yield decline/failure of monocropped direct-seeded rice is a constraint to the widespread adoption of DDSR technology and is generally believed to be caused by soil sickness that may include biotic factors such as nematodes (Nishizawa et al 1971), soil-borne pathogens (Ventura et al1981), abiotic factors such as changes in soil nutrient availability (Lin et al 2002), soil properties, climate, rice cultivars, and management practices. Despite the many constraints such as continuous cropping obstacles and weed infestation, DDSR is considered a useful strategy for maintaining the sustainability of rice production under future water shortage caused by global climate change as efficient weed control strategies have evolved over recent years.

As more rice is needed to be produced with less and less water to feed the ever-growing population, which needs the introduction of judicious water management practices and suitable water-saving technologies in rice cultivation, more research emphasis should be given to developing water-saving devices through novel irrigation systems and developing rice genotypes that are high-yielding even under water-limited conditions.

380

The importance of rice in food security Rice is the staple food of about 60% of the world’s population and demand is expected to continue to grow as population increases. Globally, rice is grown over an area of about 164 million hectares, with an annual production of 723 million tons (FAO 2012). More than 90% of the world’s rice is grown and consumed in Asia, where 60% of Earth’s population lives. China and India, which account for more than one-third of the global population, supply more than half of the world's rice.

Food security worldwide is challenged by increasing food demand and threatened by declining water availability. The situation is further aggravated by drought, global warming, methane emissions, adverse climatic changes, over-pumping of groundwater causing aquifer resources to decline, and the increasing “cost” of water (Tuong and Bouman 2003).The sustainability of irrigated rice culture is threatened by the increasing water scarcity in agriculture, caused by competition for water from urbanization and industrialization.

The need for resource conservation Water deficit is a key constraint that affects rice production in different countries. The affinity of the rice crop toward water is universally known. Rice ecosystems are generally classified into four types: irrigated, rainfed lowland, deepwater, and rainfed upland (Poehlman and Sleper 1995). Comparisons among irrigated lowland, direct-seeded, rainfed lowland, and upland rice are given in Table 1. The conventional rice production ecosystem (puddled transplanted) requires an average of 2,500 liters of water to produce 1 kg of rice (Bouman et al 2007), which is two to three times more than for other cereals (Barker et al 1996, Tuong et al 2005). It is also reported that about 50% of the diverted fresh water in Asia is used to irrigate rice fields (Barker et al 1996). Seasonal water inputs for puddled transplanted rice vary from 660 to 5,280 mm depending on the growing season, climatic conditions, soil type, and hydrological conditions (Tuong and Bouman 2003). Although evapo-transpirational losses of water in rice are similar to those in wheat (Kumar and Ladha 2011), it is also reported that the higher water requirements of rice are due to water required for puddling, and high seepage and percolation losses associated with continuous flooding (Hafeez et al 2007).

Worldwide, freshwater availability for irrigation is decreasing because of increasing competition from urban and industrial development, degrading irrigation infrastructure, and degrading water quality (Molden 2007). Rice is a heavy consumer of water, and is less efficient in the way it uses much more water than either wheat or maize (Shen et al 2001). Rice cultivation involves intensive land preparation, which entails repeated plowing, leveling, puddling, and management of bunds throughout the season. Overexploitation of groundwater in the last decades has caused serious problems in many parts of India, including Haryana and Punjab. Groundwater tables (Figure 1) have dropped on average by 0.5−1.0 m y−1 in various states of India (Bouman and Tuong 2001). The challenges faced by paddy farmers under the present scenario are 1. Nonavailability of adequate labor at the right time, at affordable prices, and in

required numbers. 2. Lack of adequate water throughout the season, over-pumping of groundwater

causing aquifer resources to decline, and the high “cost” of water when available. 3. Continuous cropping and yield decline in irrigated habitats. 4. Lack of remunerative prices for produce.

381

5. Rapid fall of water table, leading to an increase in energy requirement for pumping groundwater and increased tube-well infrastructure cost.

6. Contamination of surface water and groundwater due to leached nitrogenous fertilizers and pesticides.

7. Contamination of atmosphere due to methane emissions from paddy plants (ebblusion), paddy fields, and nitrous oxide leached out of fields.

8. Salinization of prime (those nearest the reservoirs) paddy lands. 9. Over-irrigation, over-fertilization, and over-protection due to intensive cropping

and depletion of silicon, an essential element for paddy, in paddy soils. 10. Longer turnaround time after PTR (puddled transplanted rice) delays wheat

planting, resulting in yield losses of 35 to 60 kg d−1 ha−1of wheat, particularly in eastern Indo-Gangetic plain.

11. PTR cultivation system deteriorates soil physical properties by destroying soil aggregates, forming hard-pans at shallow depth, filling of macropores with finer soil particles, and lowering permeability in the subsurface layer, all of which are believed to have a negative effect on the following wheat crop. By 2025, it is expected that 2 million ha of Asia’s irrigated dry-season rice and 13

million ha of its irrigated wet-season rice will experience “physical water scarcity,” and most of the approximately 22 million ha of irrigated dry-season rice in South and Southeast Asia will suffer “economic water scarcity” (Tuong and Bouman 2002). Severe water shortages have also led to a recent rapid increase in the area cultivated in upland rice in northern China (Wang et al 2002). In principle, water is always scarce in the dry season when the lack of rainfall makes cropping impossible without irrigation.

There is a need to develop such a production system, which is expected to cause profound changes in water conservation, labor requirement, soil organic matter turnover, nutrient dynamics, carbon sequestration, soil productivity, weed ecology, and greenhouse gas emissions. The challenge is to develop effective integrated natural resource management interventions that allow profitable rice cultivation with increased soil aeration while maintaining the productivity, environmental protection, and sustainability of rice-based ecosystems.

Direct-seeded rice One technology that enables rice to be grown on dry land without flooding, and helps farmers to cope with water scarcity, is the direct-seeded rice (DSR) system. There are three principal methods of establishing DSR (Table 2): dry seeding (sowing dry seeds into dry soil), wet seeding (sowing pregerminated seeds on wet puddled soils), and water seeding (sowing seeds into standing water). Wet-DSR is primarily done under labor-shortage conditions, and is currently practiced in Malaysia, Thailand, Vietnam, the Philippines, and Sri Lanka. But, with the elevating shortages of water, the incentive to develop and adopt DDSR has increased in Asian countries in both irrigated and rainfed ecosystems. Water seeding is widely practiced in the United States, primarily to manage weeds such as weedy rice, which are normally difficult to control. Prior to the 1950s, direct seeding was most common, but was gradually replaced by puddled transplanting (Rao et al 2007). Direct seeding of rice has the potential to provide several benefits to farmers and the environment over conventional practices of puddling and transplanting. Direct seeding helps reduce water consumption by about 30% (0.9 million L acre−1) as it eliminates raising of seedlings in a nursery, puddling, and transplanting under puddled soil, and maintains 4−5 inches of water at the base of the transplanted seedlings. The farmer saves

382

about INR 1,400 acre−1 in cultivation cost. Dry direct seeding, on the other hand, avoids nursery raising, seedling uprooting, puddling, and transplanting, and thus reduces the labor requirement. DDSR helps in making full use of family labor and having less dependence on hired labor. The development of such DSR systems should start with crop management practices that involve soil management, land practices, time of sowing, seed rate, seed establishment, varieties, nutrients, and water and weed management. Soil management practices involve soil texture, soil properties, soil structure, soil compaction, water-holding capacity, porosity, infiltration rate, soil organic matter, total and available nutrients, soil alkalinity and acidity, presence of toxic substances, and microbial populations. For proper water management, we need to increase irrigation water productivity; decrease energy requirements, decrease deep drainage losses, and implement a proper irrigation schedule. Differences between conventional irrigated rice and eco-friendly direct-seeded rice are depicted in Table 3.

All kinds of management are highly location-specific; no general recommendation is possible for all situations. These management practices are formidable challenges to ensure the productivity, profitability, and sustainability of DDSR.

In Southeast Asia, DDSR is more often adopted in the dry season than in the wet season probably because of better water control; but, dry-season rice accounts for less than one-quarter of rice production in this region. At present, 23% of rice is direct-seeded globally (Rao et al 2007). Mitigating the impact of the current scenario: potentially useful traits Suitable traits, QTLs, and genes that can be considered for developing high-yielding direct-seeded adapted rice varieties are summarized in Table 4. Although DDSR fields may experience soil moisture regimes that range from aerobic to flooded, adequate understanding of the adaptive traits and pathways will help in designing effective breeding strategies to develop rice varieties suitable for direct-seeding systems.

Traits likely to be important for developing varieties for dry direct-seeded conditions are the following:

1. Anaerobic germination 2. Early uniform germination 3. Early vigor 4. Root plasticity 5. High nutrient (N, P, Fe, Zn) uptake under variable conditions ranging from

flooded to aerobic 6. Cold tolerance at the seedling stage 7. Drought tolerance at the reproductive stage 8. Heat tolerance at the reproductive stage 9. Tolerance of/resistance to nematodes 10. Lodging resistance 11. Uniform maturity 12. High grain yield 13. No seed dormancy for rice-rice areas 14. Seed dormancy for rice-other crop areas prone to cyclone at maturity 15. Tolerance of blast 16. Resistance to bacterial blight 17. Resistance to gall midge

383

18. Resistance to brown planthopper 19. Resistance to brown spot 20. High grain quality

(i) Anaerobic germination Soil flooding at the early stage is one of the most important abiotic constraints to rice yield, with complete submergence of plants being particularly serious for rice farmers in the humid and semi-humid tropics of Asia (Jackson and Ram 2003). Early flooding due to uneven field and rainfall slows down seed germination, reduces seed germination and hinders crop establishment, and causes a high percentage of seedling mortality. Tolerance of anaerobic conditions at the early stage is a prerequisite for effective direct-seeded rice in both irrigated and rainfed areas. Unraveling the mechanisms involved in adaptation to flooding will help design management options that will allow tolerant rice genotypes to adequately establish in flooded soils while simultaneously suppressing weeds. This will consequently result in enormous savings in production costs as opposed to when rice is transplanted. It can also reduce the cost of manual or mechanical weeding and the use of hazardous chemicals for weed control. Recently, a small number of rice genotypes with greater tolerance of anaerobic germination and early seedling growth were identified following large-scale screening of 8,000 genebank accessions and breeding lines at IRRI (Ismail et al 2009, Angaji et al 2010). On the basis of consistent performance, a highly tolerant genotype, Khao Hlan On (Oryza sativa L. subsp. japonica), was selected as a potential donor parent to incorporate tolerance of flooding during germination into breeding lines (Table 4). Septiningsih et al (2013) identified six significant QTLs on chromosomes 2, 5, 6, and 7 using a BC2F3 population derived from a cross between IR42, a susceptible variety, and Ma-Zhan Red, a tolerant landrace from China. The largest QTL on chromosome 7, having an LOD score of 14.5 and an R2 of 31.7%, was confirmed. This kind of QTL may provide promising targets for use in marker-assisted selection to develop resilient varieties with improved tolerance of anaerobic conditions during germination for direct-seeded systems. (ii) Early uniform germination After shifting from transplanting to direct seeding, uniform crop establishment becomes a critical factor that affects subsequent growth, development, and yield in dry direct-seeded rice. Seed vigor is a complex and important agricultural trait that defines the seed properties, which determine the potential for rapid, uniform emergence, and for the development of normal seedlings under a wide range of field conditions. The rate, uniformity, and percentage of seedling emergence are critical determinants of crop establishment and yield. Germination and seedling growth are an extremely important trait for rice production in temperate rice-growing areas, at high altitudes in both tropical and subtropical areas, and in areas with a cold irrigation water supply. Rapidly germinating seedlings could emerge and produce a deep root system before the upper soil layers dry out. In the DDSR system, the toxicity of ammonia volatilization may be one of the main causes responsible for poor and non uniform seed germination and poor early seedling growth (Haden et al 2011, Bremner 1995). In order to improve initial crop establishment and the weed competitiveness of direct-seeded rice, varieties with higher germination and faster seedling emergence with more vigorous seedlings under dry direct-seeded conditions must be selected to minimize the risks encountered in direct seeding (Azhiri-Sigari et al 2005). Seed priming, optimum depth of seeding (3−4 cm), proper

384

soil moisture, precise land leveling, and proper seed rate (20−25 kg ha−1) can be an effective technique for rapid uniform germination of the rice crop. Recently, a major and consistent QTL for early uniform emergence on chromosome 11 contributed by Moroberekan was identified at IRRI (Table 4). (iii) Early vigor Early vigor (biomass accumulation) is a useful but complex trait in rice (Oryza sativa L.). It can be defined as a high relative growth rate (RGR) during exponential growth before canopy closure (Dingkuhn et al 1999, Poorter and De Jong 1999, Shipley 2006). Early vegetative vigor depends on the assimilate source (light capture and photosynthetic rate) as well as the sink constituted by structural growth (leaf appearance rate, potential size, and tiller outgrowth).It may also accelerate the depletion of soil-water reserves, making less water available for later crop stages (Zhang et al 2005). Varieties with early seedling vigor help in reducing the competitive effects of weeds and increase yield stability in dry direct-seeded environments (Okami et al 2011). Differences in the early vegetative vigor affecting crop establishment and yield under direct-seeded conditions have earlier been reported in Asia (Garrity et al 1992, Caton et al 2003), Latin America, and Africa (Fischer et al 1997, Johnson et al 1998, Dingkuhn et al 1999). Usually, early vegetative vigor is associated with an increase in plant height, which leads to lodging. Recently, Sandhu et al (2014) identified a QTL, qEVV9.1, for early vigor under direct-seeded conditions, and this QTL was found to be co-located with the grain yield QTL, but, interestingly, not associated with plant height (Table 4). (iv) Root architecture and plasticity Rice is characterized by a shallow root system compared with other cereal crops, having limited water extraction below 60 cm (Fukai and Inthapan 1988). The form of the rice root system also varies with the cultivation method (Yoshida and Hasegawa 1982). In upland conditions with direct sowing, the root system generally develops deeper than in transplanted planting in lowland conditions (Gowda et al 2011). Rice plants with a deep root system are therefore beneficial in avoiding water stress by absorbing water from deep soil layers (Clark et al 2002). Root characteristics such as thickness, depth of rooting, root number, root volume, and dry root biomass have been associated with adaptation to direct seeding in rice (Ekanayake et al 1985, Yoshida and Hasegawa 1982). A deeper root system has been shown to allow direct-seeded rice to extract more water from the soil, resulting in higher yield potential, and varieties with a high deep-root weight to shoot weight ratio exhibit enhanced water uptake in dry direct-seeded rice (Fukai and Cooper 1995). Direct-seeded rice has a very well developed, thick, dense, and long root system, which allows it to survive under water-scarce conditions. A study at CCSHAU, Hisar, comprising basmati (Taraori Basmati, Pusa1121, and Pusa Basmati 1460), indica (PAU201 and HKR47), and aerobic rice genotypes (MAS25, MAS26, MAS109, MAS-ARB25, and MAS-ARB868) revealed the same fact that root length and biomass were significantly influenced by the rice varieties (Sandhu et al 2011, 2012) (Figure 2). Aerobic rice genotypes performed better under water-limited conditions, which could be attributed to their relatively longer root length and/or higher biomass. Regression analysis showed that grain yield was positively correlated with root length (r2 = 0.7324) (Figure 3) and biomass (r2 = 0.8401) (Figure 4), suggesting that root length and biomass may be the limiting factor for grain yield under water-saving conditions (Sandhu et al 2011). Martin et al (2007) reported that rice varieties ADT

385

39 and PMK 3 that were suitable for cultivation under direct-seeded conditions had a longer and deeper root system than other rice varieties.

Root traits are typically complex and controlled by many genes, each with a small genetic effect; such traits are typically controlled by quantitative trait loci (Sharma et al 2011). Identifying genetic variation and QTLs for root traits can contribute to our understanding of their role in plant performance under direct-seeded conditions. Direct-seeded conditions decrease the number of nodal roots per internode and stimulate the elongation of nodal roots (Kondo et al 1999). Other traits that may be important for direct-seeded rice are root hair growth and lateral root growth, which are beneficial for nutrient uptake. The root angle also plays an important role in rice in avoiding water stress. The Deeper Rooting 1 (DRO1) gene makes the roots of rice plants grow downward instead of outward, thus enabling them to reach water held deeper in the soil (Uga et al 2013).

Root plasticity may have important adaptive effects, minimizing the deleterious effects of the environment and maximizing adaptation and productivity in specific environments for future sustainable agriculture. Plasticity is especially important in the root system since roots adopt a particular architecture under one condition, but can change architecture in response to a change in conditions, that is, a shallow root system under lowland conditions but a deep root system under upland conditions. Roots show adaptive morphogenesis, hydrotropism (Liao et al 2004), and accelerated elongation in response to moisture gradient and direct-seeded cultivation conditions (Takahashi 1997, Takahashi et al 2003, Rajaniemi 2007). Therefore, a better understanding of root-related traits and how root trait QTLs interacts to affect soil resource acquisition across a range of environments will be an important tool in the breeding of direct-seeded rice varieties. (v) Nutrient uptake

Switching over from anaerobic to aerobic conditions has several implications. Land preparation and water management are the principal factors governing the nutrient dynamics in both DDSR and TPR systems. Reduced nutrient uptake, especially of nitrogen and phosphorus under direct-seeded cultivation conditions vis-à-vis flooded conditions, has been the most important factor for lower yield in direct-seeded systems than in flooded systems of rice cultivation (Kumar and Ladha 2011). Increased oxygen in the rhizosphere for long periods favors oxidation of NH4 to NO3 and thereby more losses of N via leaching. One method of maintaining soil N as NH4is to add nitrification inhibitors along with fertilizer, which also increase nutrient-use efficiency (NUE) and crop yield. For example, dicyandiamide is a commercially available nitrification inhibitor, which is used with solid chemical fertilizer in rice fields.Micronutrient deficiencies are of concern in DSR—imbalances of such nutrients (e.g., Zn, Fe, Mn, and S) result from improper and imbalanced N fertilizer application (Gao et al 2006, Saleque and Kirk 1995). In DSR, Fe oxidation by root-released oxygen reduces rhizosphere soil pH and limits the release of Zn from highly insoluble fractions for availability to the rice plant (Kirk and Bajita 1995). A shift from TPR to DDSR may also affect Zn bioavailability in rice. Nutrient-efficient plants produce higher yield per unit of nutrient applied or absorbed than other plants grown under similar agroecological conditions (Fageria et al2008). Although it is often emphasized that enhancing nutrient-use efficiency is critical across the globe both to sustain yield in low-input agriculture and to reduce fertilizer inputs in high-input agriculture, it is not clear whether traits that enhance efficiency in low-input systems also confer efficiency in high-input systems (Rose et al 2012).

386

Agricultural production in the 21st century is predicted to be more limited because of lower availability and increased cost of water and nutrient resources (Lal 2007). This emphasizes the urgency of improving the rice root system so that plants are able to capture nutrients more efficiently. Strategies of efficient nutrient acquisition include root morphology for exploring nutrients in soil through the growth of axial roots with shallow angles and more dispersed lateral roots (Lynch 2011) and the development of longer lateral roots with more root hairs (Rose et al 2012).Phosphorus uptake is closely related to root hair length, and plants grown under phosphate-limiting conditions form longer root hairs (Bates and Lynch 1996, Zhang et al 2003). Increased lateral root growth increased phosphorus uptake in common bean (Lynch and Van Beem 1993) and maize (Zhu and Lynch 2004), nitrogen uptake (Sen et al 2013) and yield under varying water and nutrient regimes (Hochholdinger and Tuberosa 2009) in maize, and water uptake under drought in rice (Suralta et al 2008, 2010). The increased lateral root elongation to target nutrients (Zhang and Forde 1998, Hodge 2004) is more evident in the case of upland dry direct-seeded cultivars (Yang et al 2003). Sandhu et al (2014) reported a significant and positive correlation between some root traits (root hair density, and lateral root and nodal root number) and yield under dry direct-seeded conditions, indicating the role of improved water and nutrient uptake at the seedling stage in improving rice yield for direct-seeded conditions. A number of QTLs associated with nutrient uptake have been detected in rice (Ni et al 1998, Wissuwa et al 1998, Ming et al 2000), maize (Zhu et al 2005), wheat (Su et al 2006,2009), common bean (Liao et al 2004, Yan et al 2004), and soybean (Li et al 2005, Liang et al 2010). The characterization of root-related traits and QTLs associated with these root traits (Sandhu et al 2013, 2014) that increase nutrient uptake under direct-seeded conditions can help to develop direct-seeded rice varieties with high yield potential. Coexisting chromosomal regions/loci governing different traits provide a unique opportunity for breeders to introgress such regions together as a unit into high-yielding lowland varieties through marker-assisted backcrossing and to develop cultivars possessing increased adaptation to dry direct-seeded conditions. (vi) Cold tolerance Low temperature is a common production constraint in dry direct-seeded rice cultivation in temperate zones, high-elevation environments, and, even in some subtropical rice-growing areas such as northern Laos, the temperature can often drop below 15oC at the time of sowing (Sihathep et al 2001). Low temperature during December-January is a severe constraint to dry-season rice in South Asia. Under low temperature, delayed germination and seedling growth may result in nonuniform seedling growth and weak seedlings, which may affect final grain yield (Cruz et al 2006). Low temperature weakens photosynthetic ability by inducing leaf discoloration, reduces plant height, produces degenerated spikes, delays days to heading, reduces spikelet fertility, and causes irregular grain maturity and poor grain quality. In parts of South and Southeast Asia, an estimated 7 million hectares cannot be planted with modern varieties because of low-temperature stress. Indica rice, associated with tropical environments, is more sensitive to low temperature. The more tolerant japonica rice is divided into tropical and temperate groups. Andaya and Mackill (2003) identified a major QTL on chromosome 12, designated as qCTS12a, which was closely associated with cold tolerance during the vegetative stage, and accounted for 41% of the phenotypic variation. Suh et al (2010) identified three main-

387

effect QTLs (qPSST-3, qPSST-7, and qPSST-9) on chromosomes 3, 7, and 9 controlling cold tolerance at the reproductive stage. (vii) Drought tolerance Water stress is the biggest challenge for rice productivity in the rainfed rice ecosystem. Rainfed rice occupies about 38% of the total cropped area and contributes 21% to total rice production. In the context of current and predicted water scarcity, increasing irrigation is generally not a viable option for alleviating drought problems in rainfed rice-growing systems. It is therefore critical that genetic management strategies be undertaken for cultivating rice with less water and focusing on maximum extraction of available soil moisture and its efficient use in crop establishment and growth to maximize biomass and yield.

Drought resistance is a trait controlled by many genes having different effects, and it is affected by drought timing and severity. Another way to explain the complexity of drought is that drought resistance involves an interaction between the genes involved in yield potential per se (which are numerous) and the genes for drought resistance. Breeding drought-resistant direct-seeded rice is therefore an increasingly important goal. Large-effect QTLs for grain yield under drought have been recently identified (Bernier et al 2007, Venuprasad et al 2009) and their successful introgression has established a yield advantage under drought (Swamy et al 2013). (viii) Heat tolerance Increases in greenhouse gas emissions and relative humidity influence the degree to which high temperature can affect rice productivity. The global mean temperature could increase by 2.0−4.5 oC by the end of this century. Spatial analysis using cropping pattern data from the Rice almanac (Maclean et al 2002) showed susceptible stages of rice (i.e., flowering and early grain filling) coinciding with high-temperature conditions in Bangladesh, eastern India, southern Myanmar, and northern Thailand (Wassmann et al 2009). With the gradual shift from intensive irrigated systems, in which standing water creates a cooler microclimate, future water-saving technologies (e.g., alternate wetting and drying, aerobic rice, and direct-seeded rice) could be more vulnerable to the adverse effects of high temperatures. Also, dry- season rice in South Asia is highly vulnerable to high temperature during the reproductive stage. There is thus an urgent need to address high-temperature-induced yield losses in rice under current climatic conditions and more so in the face of a changing climate scenario. Regions such as Pakistan and northern and southern India have hot, extended summers suitable for screening rice germplasm to identify prospective heat-tolerance donors that can be used by breeders to develop high-temperature-tolerant direct-seeded rice varieties. Using a diverse set of genotypes, aus-type variety N22 was identified as an ideal donor of the high-temperature tolerance gene at the flowering stage (Yoshida et al 1981, Prasad et al 2006, Jagadish et al 2008). (ix) Nematode tolerance/resistance Nematode infestation is the major bottleneck in DDSR, especially in dry field conditions. Nematodes are prevalent in Asia and Africa, causing up to 70% yield loss to upland and lowland rice. In India, they are reported to cause 17−30% yield loss due to poorly filled kernels (Jain et al 2007).There is a need to incorporate tolerance of nematodes in rice varieties. Nematicides pose a great threat to target organisms,

388

to our aquatic ecosystems, to our drinking-water supplies, and to the ozone layer. Pratylenchus spp. and Meloidogyne spp. have the highest potential of causing economic damage in Southeast Asian upland rice ecosystems. Rice root-knot nematode is widely distributed in the countries of South and Southeast Asia, Myanmar, Bangladesh, Laos, Thailand, Vietnam, India, China, and the Philippines, and in the United States. A lack of resistance to nematodes has been a major factor hindering the genetic improvement of cultivated rice. Oryza glaberrima exhibited a hypersensitive resistant reaction by forming necrotic tissues in invaded roots and consequently suppressing nematode development. The use of resistant cultivars is a low-cost and sustainable option for the control of nematodes in the long term, which does not impose unwanted changes in traditional agronomic practices. Recently, at IRRI, QTLs linked to resistance and tolerance genes for M. graminicola in crosses of O. glaberrima and O. sativa have been identified on chromosomes 3, 5, and 10 (Table 4). (x) Lodging resistance Lodging is a common problem resulting in reduced yield, quality of production, and pace as well efficiency of mechanical harvesting (Kono 1995).The success of the Green Revolution convinced many breeders of the suitability of semidwarf varieties for increased crop productivity, and many improved semidwarf varieties have been developed and cultivated worldwide. However, despite the short stature conferred by the semidwarf1 (sd1) gene, lodging remains a problem in many improved rice varieties. A major factor contributing to lodging is the phenotypic characteristics of rice plants, particularly when they are tall with long internodes. Tall plants with shallow roots in combination with weather conditions after heading, such as strong winds and heavy rainfall, can cause severe lodging (Watanabe 1997). However, to develop new varieties with increased plant biomass and harvest index to further increase grain yield, it is also necessary to develop new strategies to improve lodging resistance in rice. Intermediate plant height, large stem diameter, and thick stem wall with high lignin content are characters for lodging resistance, and lower positioning of panicles in the plant’s canopy is associated with increased tolerance of lodging. In fact, the use of semidwarf genes such as sd1 and Rht1 successfully eliminated the lodging problem in rice and wheat, respectively, and also increased crop productivity. However, in japonica varieties that normally produce weak and fine culms, sd1 is ineffective in enhancing culm strength. Hence, the lodging resistance of these varieties has been improved by decreasing the length and weight of the above-ground parts of the plant. Instead of relying solely on the sd1 gene, major genes and QTLs controlling culm strength can be used in developing improved rice varieties. The use of lodging-resistance genes in addition to the dwarfing gene is a promising new approach in improving lodging resistance and further increasing productivity in rice. (xi) Uniform maturity Maturity of rice occurs at different times irrespective of planting method and variety. Uniform maturity is helpful for machine harvesting. If the crop does not mature uniformly, then premature cutting of the rice keeps the grain from reaching maturity, and can cause serious losses in product quality. Early mature panicles in cultivars with nonuniform maturity over a longer period of time are likely to undergo shattering. In the process to develop suitable varieties for DDSR, selection for uniform maturity is considered an important trait.

389

(xii) High grain yield The yield of direct-seeded rice under farmers’ field conditions tends to be lower than that of transplanted rice because of poor and uneven establishment; nematode problems; growth of weeds that compete with rice for nutrients, moisture, and sunlight; and biotic stresses (Pathak et al 2011). Incorporation of traits needed for direct-seeded rice may lead to increased yield and adaptability of rice to direct-seeded conditions. DSR can have yield comparable with that of puddled transplanted conditions if the crop is properly managed (Kumar and Ladha 2011). The identification of QTLs conferring improved adaptation to direct seeding may facilitate the development of high-yielding direct-seeded rice varieties. Several QTLs related to grain yield under dry direct-seeded conditions, qGY1.1, qGY2.1, qGY2.2, qGY6.1, qGY8.1, and qGY10.1, have recently been identified and can be incorporated into future QTL introgression work to improve rice yield under dry direct-seeded drought conditions (Sandhu et al 2013, 2014). (xiii) Seed dormancy

Seed dormancy is an important agronomic trait affecting grain yield, quality, and processing performance. Seed dormancy has both advantages and disadvantages for plants, especially crops, in terms of cultivation and use, because weak dormancy leads to uniform germination, whereas deep dormancy prevents preharvest sprouting but inhibits germination.

Seed dormancy is needed for rice-other crop (such as wheat) areas that are prone to cyclones at the maturity stage of rice. This system is found in the fertile, hot semiarid to hot subhumid regions of the Indus and Gangetic alluvial plains of Bangladesh, India, Nepal, and Pakistan. Dormancy was regarded as the primary inner factor that led to rice resistance to preharvesting sprouting.In rice-rice as well as in rice-wheat areas, no seed dormancy will be an important trait to breed for. In areas prone to cyclones around maturity in South Asia, seed dormancy for 7−10 days will prevent the lodged mature crop seed to germinate in the panicle itself. Based on the prevailing situation, no seed dormancy/seed dormancy should be included as a trait in the breeding program for DDSR.

Indica rice variety Kasalath exhibits desirable seed dormancy, stronger than that of Nipponbare, a japonica rice variety. Lin et al (1998) reported five putative QTLs affecting seed dormancy on chromosomes 3, 5, 7 (two regions), and 8. Nipponbare alleles increased the germination rate whereas Kasalath alleles showed desirable seed dormancy. The donor having this kind of characteristic can be used in a marker-assisted breeding program to develop direct-seeded rice varieties adapted to particular environments.

(xiv) Blast tolerance A breakdown of blast resistance is the major cause of yield instability in several rice-growing areas. Efforts are underway to develop rice varieties with durable blast resistance. More than 40 major genes as well as QTLs for blast resistance have been identified. Monogenic resistance to blast is less stable but varieties with pyramided monogenes or QTLs are durably resistant. Rice research should focus on identifying genes with more durable resistance, tagging these genes with molecular markers, and pyramiding these genes or QTLs through molecular marker-aided selection. Pi-9, one of the identified blast-resistance genes, exhibited excellent resistance to rice blast races from 43 different countries (Liu et al 2002), which was considered as the widest spectrum resistance resource favored by breeders in the

390

cloned rice blast-resistance genes. The Pita locus in rice has also been effectively used to manage rice blast worldwide, and also in India. The genes Pita and Pita2occupy a near-centromeric region on chromosome 12 and are linked closely at 0.4 cM with Pita2reportedly having broader resistance (Bryan et al 2000, Fjellstrom et al 2004, Shikari et al 2013). Three indica landrace cultivars, Tadukan from the Philippines, Tetep from Vietnam, and TeQing from China, are known as sources of Pita resistance worldwide (Moldenhauer et al 1990). (xv) Resistance to bacterial blight Bacterial blight (BB), caused by Xanthomonas oryzae pv. oryzae (Xoo), is one of the most destructive diseases active in the major rice-growing countries of Asia, as rice is grown under irrigated and high-fertilizer-input conditions that are conducive to disease development. In severe epidemics, yield losses ranging from 20% to 40% have been reported. So far, 14 dominant (Xa1, Xa2, Xa3, Xa4, Xa7, Xa10, Xa11, Xa12, Xa14, Xa16, Xa17, Xa18, Xa21, and Xa22(t)) and six recessive (xa5, xa8, xa13, xa15, xa19, and xa20) resistance genes for BB have been identified. Long-term cultivation of rice varieties carrying a single resistance gene has resulted in a significant shift in pathogen-race frequency and consequent breakdown of resistance. One tangible solution to resistance breakdown is the pyramiding of multiple resistance genes in the background of modern high-yielding varieties. The Xa4 gene was durable as a minor gene for partial resistance (Koch and Parlevliet 1991). The Xa21 gene was identified in the wild species O. longistaminata (Khush et al 1990), and expressed resistance to all Philippine and Indian races of Xanthomonas oryzae. There was a large residual effect in genes Xa21 and Xa4, xa5 had a smaller effect and xa13 had no residual effect (Li et al 2001). Loan et al (2006) reported high resistance in combinations of three dominant resistance genes (Xa4+Xa7+Xa21) in IRBB 62, followed by Xa4+xa5+xa13+Xa21 in IRBB 60. (xvi) Resistance to gall midge Rice gall infestation is a serious rice disease caused by a dipteran insect pest known as gall midge (Orseolia oryzae). The disease is prevalent in India, China, Southeast Asia, and Africa, with severe losses reported in regions of Bangladesh, China, India, Indonesia, Myanmar, Sri Lanka, and Vietnam, causing significant yield loss mainly during the kharif season. After hatching, the larvae drill into rice plants along the leaf sheath and eat the growing point of rice. The leaf sheath develops into onion shoot-like galls with an apical shortened lamina. Breeding rice varieties with pest resistance genes such as W1263 (with Gm1), Phalguna (Gm2), RP2068 (gm3), Abhaya (Gm4), and ARC5984 (Gm5), and with other resistance sources with unknown genes such as Bhumansan, NHTA8, Banglei, Aganni, and ARC6605, has been a viable, ecologically acceptable approach for the management of gall midge. (xvii) Resistance to brown planthopper The brown planthopper (BPH), Nilaparvata lugens, causes serious yield reduction by directly sucking the plant sap and acting as a vector of various diseases such as rice grassy stunt and ragged stunt. It is a major insect pest of rice throughout the Asian rice-growing countries. So far, 16 major effective BPH resistance genes have been identified in indica cultivars and four wild relatives. Reissig et al (1985) observed that the dense canopy of direct-sown rice provides an ideal condition (moister microenvironment) for the multiplication of brown planthopper. Future research should investigate whether reduced seeding rate or increased spacing would reduce

391

insect pressure in direct-seeded rice. Alternatively, resistant germplasm can be sought since the use of insecticides is being discouraged. A Sri Lankan indica rice cultivar, Rathu Heenati, was found to be resistant to all four biotypes of BPH. Linkage mapping analysis confirmed that resistance to BPH in Rathu Heenati is likely to be controlled by a major resistance gene, BPH17, located on chromosome 4, and two minor genes, Qbph3 and Qbph10 (Sun et al 2005). Pyramiding of BPH17 and other BPH resistance genes through molecular breeding can be proposed to enhance resistance to BPH. (xviii) Resistance to brown spot Rice brown spot (BS), a chronic disease, affects millions of hectares of rice every growing season, more often in South and Southeast Asian countries. BS is conventionally perceived as a secondary problem found in rice crops that experience physiological stresses, for example, drought and poor soil fertility, especially the lack of nitrogen, rather than a true infectious disease. Three QTLs were detected in cultivar Tadukan (qBS2, qBS9, qBS11) on chromosomes 2, 9, and 11, respectively (Sato et al2008), with qBS11being considered as having a major effect. (xix) Grain quality Good grain quality is one factor that influences acceptance by consumers. Rice quality is a combination of physical and chemical characteristics that are required for a specific use by a specific user. With improved economic conditions and a better standard of living, people are looking for better-quality rice. Improvement in grain quality coupled with productivity will enhance export potential and help to sustain marketability in trade and commerce. The production of high-quality grain has been one of the major concerns in DSR. The challenges for dry direct-seeded rice quality are grain appearance, chalkiness, head rice recovery, amylose content, gelatinization temperature, and grain dimensions. As the availability of nutrients will be less under dry direct-seeded conditions, the effects on grain quality of rice under such conditions need to be assessed. A reduction in grain weight under direct-seeded cultivation conditions may be due to a decrease in starch accumulation. This may change grain composition and quality. Grain quality is affected by water management also. Under water-deficit conditions, starch levels have been observed to decrease in anthers of plants during flowering (Lalonde et al 1997), which may reduce pollen viability (Garrity et al 1986, Farooq et al 2011) and hence panicle fertility. The number of sterile spikelets increased as well as abortive, opaque, and chalky kernels in DDSR compared with TPR (Farooq et al 2007, 2009). High-yielding breeding lines and varieties were not accepted by the farmers in many countries because of poor grain quality. Thus, a new strategy was followed to introduce grain quality characteristics (such as medium-long slender grains) from improved (indica) germplasm.

Research over the last few years has led to the identification of donors, QTLs, and genes for many of the traits important to the development of rice varieties for dry direct-seeded conditions. On the other hand, recent advances in molecular technology provide a unique opportunity to combine these traits/QTLs/genes efficiently to develop new varieties for DDSR. At the same time, it is equally important to continue strategic basic research to identify new donors, QTLs, and genes to make a stronger base to enhance yield under DDSR. Conclusions

392

Over the centuries, lowland rice has proven to be a remarkably sustainable system for rice production mostly because of its luxurious water availability. But, the present-day scenario threatens the sustainability of lowland rice production and necessitates the adoption of direct-seeded cultivation. Most of the technologies reduce inputs only at the expense of yield. Under different rice production zones across the continents, the need exists to develop a site-specific package of production technologies for different rice production systems. Incorporation of all traits beneficial for direct-seeded rice adaptation may allow the development of suitable varieties for DDSR that may further allow adoption of this technology more quickly and easily. Effective management strategies, well-developed biotechnological and genetic approaches, and a better understanding of traits will help to solve the problems in DDSR.Currently, IRRI and other national research institutes have begun work on direct-seeded rice systems and it is expected that, in the next few years, with the development of new varieties more suitable to DDSR, the technology will be accepted by farmers on a large scale. References Andaya VC, Mackill DJ. 2003. Mapping of QTLs associated with cold tolerance

during the vegetative stage in rice. J. Exp. Bot. 54(392):2579-2585 Angaji SA, Septiningsih EM, Mackill DJ, Ismail AM. 2010. Identification of QTLs

associated with tolerance of anaerobic conditions during germination in rice (Oryza sativa L.). Euphytica. 172:159-168.

Azhiri-Sigari T, Gines H, Sebastian LS, Wade L. 2005. Seedling vigor of rice cultivars in response to seeding depth and soil moisture. Philipp. J. Crop Sci. 30(1):53-58.

Barker R, Dawe D, Tuong TP, Bhuiyan SI, Guerra LC. 1996. The outlook for water resources in the year 2020: challenges for research on water management in rice production. Assessment and orientation towards the 21st century.Proceedings of 19th Session of the International Rice Commission, 7-9 September 1996. Cairo (Egypt): FAO. p 96-109.

Bates TR, Lynch JP. 1996. Stimulation of root hair elongation in Arabidopsis thaliana by low phosphorus availability. Plant Cell Environ.19:529-538.

Bernier J, Kumar A, Venuprasad R, Spaner D, Atlin GN. 2007. A large-effect QTL for grain yield under reproductive-stage drought stress in upland rice. Crop Sci. 47:507-516.

Bouman BAM. 2002. Aerobic rice (Han Dao): a new way of growing rice in water short areas. Mol. Gen. Genet. 4:53-61.

Bouman BAM, Lampayan RM, Tuong TP. 2007. Water management in irrigated rice: coping with water scarcity. Los Baños (Philippines): International Rice Research Institute. 54 p.

Bouman BAM, Tuong TP. 2001. Field water management to save water and increase its productivity in irrigated rice. Agric. Water Manage. 49(1):11-30.

Bremner JM. 1995. Recent research on problems in the use of urea as a nitrogen fertilizer. Fert. Res. 42(1-3):321-329.

Bryan GT, Wu K, Farrall L, Jia Y, Hershey HP, McAdams SA, Faulk KN, Donaldson GK, Tarchini R, Valent B. 2000. A single amino acid difference distinguishes resistant and susceptible alleles of the rice blast resistance gene Pi-ta. Plant Cell. 12:2033-2045.

393

Caton BP, Cope AE, Mortimer AM. 2003. Growth traits of diverse rice cultivars under severe competition: implications for screening for competitiveness. Field Crops Res. 83:157-172.

Chu Z, Fu B, Yang H, Xu C, Li Z, Sanchez A, Park YJ, Bennetzen JL, Zhang Q, Wang S. 2006. Targeting xa13, a recessive gene for bacterial blight resistance in rice. Theor. Appl. Genet. 112:455-461.

Clark LJ, Cope RE, Whalley WR, Barraclough PB, Wade LJ. 2002. Root penetration of strong soil in rainfed lowland rice: comparison of laboratory screens with field performance. Field Crops Res.76:189-198.

Cruz RPD, Milach SCK, Federizzi LC. 2006. Inheritance of rice cold tolerance at the germination stage. Genet. Mol. Biol. 29:314-320.

Dingkuhn M, Johnson DE, Sow A, Audebert AY. 1999. Relationships between upland rice canopy characteristics and weed competitiveness. Field Crops Res. 61(1):79-95.

Ekanayake IJ, O’Toole JC, Garrity DP, Masajo TM 1985. Inheritance of root characters and their relations to drought resistance in rice. Crop Sci. 25:927-933.

Fageria NK, Baligar VC, Li YC. 2008. The role of nutrient efficient plants in improving crop yields in the twenty first century. J. Plant Nutr.31:1121-1157.

FAO. 2012. Available at www.faostat.fao.org. Farooq M, Siddique KHM, Rehman H, Aziz T, Lee DJ, Wahid A. 2011. Rice direct

seeding: experiences, challenges and opportunities. Soil Till. Res. 111:87-98. Farooq M, Basra, SMA, Ahmad N. 2007. Improving the performance of transplanted

rice by seed priming. Plant Growth Regul. 51:129-137. Farooq M,Wahid A, Lee DJ, Ito O, Siddique KHM. 2009. Advances in drought

resistance of rice. Crit. Rev. Plant Sci. 28:199-217. Fischer AJ, Ramirez HV, Lozano J.1997. Suppression of jungle rice

(Echinochloacolona (L.) Link) by irrigated rice cultivars in Latin America. Agron. J.89:516-552.

Fjellstrom R, Concetta A, Conaway-Bormans A, McClung M, Marco A. Marchetti, Shank RA, William DP. 2004. Development of DNA markers suitable for marker assisted selection of three Pi genes conferring resistance to multiple Pyricularia grisea pathotypes. Crop Sci. 44:1790-1798.

Fukai S, Inthapan P. 1988. Growth and yield of rice cultivars under sprinkler irrigation in south-eastern Queensland. Aust. J. Exp. Agric. 28:249-252.

Fukai S, Cooper M. 1995. Development of drought-resistant cultivars using physio-morphological traits in rice. Field Crops Res. 40:67-86.

Gao XP, Zou CQ, Fan XY, Zhang FS, Hoffland E. 2006. From flooded to aerobic conditions in rice cultivation: consequences for zinc uptake. Plant Soil 280:41-47.

Garrity DP, Movilon M, Moody K. 1992. Differential weed-suppressing ability in upland rice cultivars. Agron. J.84:586-591.

Garrity DP, Oldeman LR, Morris RA. 1986. Rainfed lowland rice ecosystem: characterization and distribution. Progress in rainfed lowland rice. Los Baños(Philippines): International Rice Research Institute. p 3-23.

Gowda VRP, Henry A, Yamauchi A, Shashidhar HE, Serraj R. 2011. Root biology and genetic improvement for drought avoidance in rice. Field Crops Res. 122:1-13.

394

Haden, VR, Xiang J, Peng S. 2011. Relative effects of ammonia and nitrite on the germination and early growth of aerobic rice. J. Plant Nutr. Soil Sci. 174(2):292-300.

Hafeez MM, Bouman BAM, Van de Giesen N, Vlek P. 2007. Scale effects on water use and water productivity in a rice-based irrigation system (UPRIIS) in the Philippines. Agric. Water Manage. 92:81-89.

Hochholdinger F, Tuberosa R. 2009. Genetic and genomic dissection of maize root development and architecture. Curr. Opin. Plant Biol. 12:172-177.

Hodge A. 2004. The plastic plant: root responses to heterogeneous supply of nutrients. New Phytol. 162:9-24.

Ismail AM, Ella ES, Vergara GV, Mackill DJ. 2009. Mechanisms associated with tolerance to flooding during germination and early seedling growth in rice (Oryza sativa). Ann. Bot. 103:197-209.

Jackson M, Ram PC. 2003. Physiological and molecular basis of susceptibility and tolerance of rice plants to complete submergence. Ann. Bot. 91:227-241.

Jagadish SVK, Craufurd PQ, Wheeler TR. 2008. Phenotyping parents of mapping populations of rice (Oryza sativa L.) for heat tolerance during anthesis. Crop Sci. 48:1140-1146.

Jain RK, Mathur KN, Singh RV. 2007. Estimation of losses due to plant parasitic nematodes on different crops in India. Ind. J. Nematol. 37:219-220.

Jairin JK, Phengrat S, Teangdeerith A, Vanavichit, Toojinda T. 2007. Mapping of a broad-spectrum brown planthopper resistance gene, Bph3, on rice chromosome 6. Mol. Breed. 19:35-44.

Johnson DE, Dingkuhn M, Jones MP, Mahamane MC. 1998. The influence of rice plant type on the effect of weed competition on Oryza sativa and Oryza glaberrima. Weed Res.38:207-216.

Joshi E, Kumar D, LalB, Nepalia V, Gautam P, Vyas AK. 2013.Management of direct seeded rice for enhanced resource-use efficiency. Plant Know. J. 2(3):119-134.

Khush GS, Bacalangco E, Ogawa T. 1990. A new gene for resistance to bacterial blight from O. longistaminata. Rice Genet. Newsl. 7:121-122.

Kirk GJD, Bajita JB. 1995. Root induced iron oxidation, pH changes and zinc solubilization in the rhizosphere of lowland rice. New Phytol. 131:129-137.

Koch M, Parlevliet JE. 1991. Residual effects of the Xa-4 resistance gene in three rice cultivars when exposed to a virulent isolate of Xanthomonas campestris pv. oryzae. Euphytica 55:187-193.

Koide Y, Ebron LA, Kato H, Tsunematsu H, Yanoria MJT, Kobayashi N, Yokoo M, Maruyama K, Imbe T, Fukuta Y. 2011.A set of near-isogenic lines for blast resistance genes with an indica-type rainfed lowland elite rice (Oryza sativa L.) genetic background. Field Crops Res. 123(1):19-27.

Kondo M, Murty MVR, Aragones DV, Okada K, Winn T, Kwak KS. 1999. Characteristics of root system and water uptake in upland rice. In: Ito O, O'Toole J, Hardy B, editors. Genetic improvement of rice for water-limited environments. Manila (Philippines): International Rice Research Institute. p 117-134.

Kono M. 1995. Physiological aspects of lodging. In: MatsuoT, Kumazawa K, Ishii R, Ishihara K, Hirata H, editors. Science of the rice.II. Physiology.p 971-982.

Kumar V, Ladha JK. 2011. Direct seeding of rice: recent developments and future research needs. Adv. Agron. 111:297-413.

395

Lal R. 2007. Anthropogenic influences on world soils and implications for global food security. Adv. Agron.93:69-93.

Lalonde S, Beebe D, Saini HS. 1997. Early signs of disruption of wheat anther development associated with the induction of male sterility by meiotic-stage water deficit. Sex. Plant Reprod. 10:40-48.

Li Z, Sanchez A, Angeles E, Singh S, Domingo J, Huang N, Khush GS. 2001. Are the dominant and recessive plant disease resistance genes similar? A case study of rice R genes and Xanthomonas oryzae pv. oryzae races. Genetics 159:757-765.

Li ZC, Mu P, Li CP, Zhang HL, Li ZK, Gao YM, Wang XQ. 2005. QTL mapping of root traits in a doubled haploid population from a cross between upland and lowland japonica rice in three environments. Theor. Appl. Genet. 110:1244-1252.

Liang Q, Cheng XH, Mei MT, Yan XL, Liao H. 2010. QTL analysis of root traits as related to phosphorus efficiency in soybean. Ann. Bot.106:111-117.

Liao H, Yan XL, Rubio G, Beebe SE, Blair MW. 2004. Genetic mapping of basal root gravitropism and phosphorus acquisition efficiency in common bean. Funct. Plant Biol. 31:959-970.

Lin S, Dittert K, Sattelmacher B. 2002. The ground cover rice production system GCRPS: a successful new approach to save water and increase nitrogen fertilizer efficiency? In: Bouman BAM, Hengsdijk H, Hardy B, Bindraban PS, Tuong TP, Ladha JK, editors. Water-wise rice production. Proceedings of the International Workshop on Water-wise Rice Production, 8-11 April 2002, Los Baños, Philippines. Los Baños (Philippines): International Rice Research Institute. p 187-196.

Lin SY, Sasaki T, Yano M. 1998. Mapping quantitative trait loci controlling seed dormancy and heading date in rice, Oryza sativa L., using backcross inbred lines. Theor. Appl. Genet. 96:997-1003.

Liu G, Lu G, Zeng L, Wang GL. 2002. Two broad-spectrum blast resistance genes, Pi9(t) and Pi2(t), are physically linked on rice chromosome 6. Mol. Genet. Genom. 267(4):472-480.

Loan LC, Ngan VT, Van PD. 2006. Preliminary evaluation on resistance genes against rice bacterial leaf blight in Can Tho province - Vietnam. Omon Rice 14:44-47.

Lynch J, Van Beem JJ. 1993. Growth and architecture of seedling roots of common bean genotypes. Crop Sci. 33:1253-1257.

Lynch JP. 2011. Root phenes for enhanced soil exploration and phosphorus acquisition: tools for future crops. Plant Physiol.156:1041-1049.

Maclean JL, Dawe DC, Hardy B, Hettel GP, editors. 2002. Rice almanac: source book for the most important activity on Earth. Los Baños (Philippines): International Rice Research Institute, West Africa Rice Development Association, International Center for Tropical Agriculture, and Food and Agriculture Organization.

Martin G, James Padmanathan PK, Subramanian E. 2007. Identification on suitable rice variety adaptability to aerobic irrigation. J. Agric. Biol. Sci. 2(2):1-3.

Marzougui S, Sugimoto K, Yamanouchi U, Shimono M, Hoshino T, Hori K, Kobayashi M, Ishiyama K, Yano M. 2012. Mapping and characterization of seed dormancy QTLs using chromosome segment substitution lines in rice. Theor. Appl. Genet. 124:893-902.

396

Ming F, Zheng XW, Mi GH, He P, Zhu LH. 2000. Identification of quantitative trait loci affecting tolerance to low phosphorus in rice (Oryza sativa L.). Chin. Sci. Bull. 45:520-524.

Mishra KK, Vikram P, Yadaw RB, Swamy BPM, Dixit S, Sta Cruz MT, Maturan P, Kumar A. 2013. qDTY12.1: a locus with a consistent effect on grain yield under drought in rice. BMC Genet. 14:12-22.

Molden D. 2007. Water for food, water for life: a comprehensive assessment of water management in agriculture. London: International Water Management Institute.

Moldenhauer KAK, Lee FN, Norman RJ, Helms RS, Well RH, Dilday RH, Rohnian PC, Marchetti MA 1990. Registration of ‘Katy’ rice. Crop Sci. 30:747-748.

Nair S, Kumar A, Srivastava MN, Mohan M, 1996. PCR-based DNA markers linked to a gall midge resistance gene, Gm4t has potential for marker-aided selection in rice. Theor. Appl. Genet. 92:660-665.

Ni JJ, Wu P, Senadhira D, Huang N. 1998. Mapping QTLs for phosphorus deficiency tolerance in rice (Oryza sativa L.). Theor. Appl. Genet.97:1361-1369.

Nie L, Peng S, Chen M, Shah F, Huang J, Cui K, Xiang J. 2012. Aerobic rice for water-saving agriculture: a review.Agron. Sustain. Dev. 32:411-418.

Nishizawa T, Ohshima Y, Kurihara H. 1971. Survey of the nematode population in the experimental fields of successive or rotative plantation. Plant Prot. Soc. 18:121-122.

Okami M, Kato Y, Yamagishi J. 2011. Role of early vegetative vigor in adaptation of rice to water-saving aerobic culture: effects of nitrogen utilization and leaf growth. Field Crops Res. 124(1):124-131.

Pandey, S, Velasco. 2002. Economics of direct seeding in Asia: patterns of adoption and research priorities. In: Pandey S, Mortimer M, Wade L, Tuong TP, Lopez K ,Hardy B, editors. Direct seeding: research issues and opportunities. Proceedings of the international workshop on direct seeding in Asian rice systems: strategic research issues and opportunities, 25-28 January 2000, Bangkok, Thailand. Los Baños (Philippines): International Rice Research Institute. p 3-14.

Parthasarathi T, Vanitha K, Lakshamanakumar P, Kalaiyarasi D. 2012. Aerobic rice:mitigating water stress for the future climate change. Int. J. Agron. Plant Prod. 3(7):241-254.

Pathak H, Tewari AN, Sankhyan S, Dubey DS, Mina U, Singh V, Jain N, Bhatia A. 2011. Direct-seeded rice: potential, performance and problems. A review. Curr. Adv. Agric. Sci. 3(2):77-88.

Perumalsamy S, Bharani M, Sudha M, Nagarajan P. 2010. Functional marker-assisted selection for bacterial leaf blight resistance genes in rice (Oryza sativa L.). Plant Breed. 129:400-406.

Poehlman JM,Sleper DA 1995. Breeding field crops. Iowa State University Press, Ames, IA. p 278-299.

Poorter H, De Jong R. 1999. A comparison of specific leaf area, chemical composition and leaf construction costs of field plants from 15 habitats differing in productivity. New Phytol. 143(1):163-176.

Prasad PVV, Boote KJ, Allen LH, Sheehy JE, Thomas JMG. 2006. Species, ecotype and cultivar differences in spikelet fertility and harvest index of rice in response to high temperature stress. Field Crops Res. 95:398-411.

Qu S, Liu G, Zhou B, Bellizzi M, Zeng L, Dai L, Han B, Wang GL. 2006. The broad-spectrum blast resistance gene Pi9 encodes a nucleotide-binding site–

397

leucine-rich repeat protein and is a member of a multigene family in rice. Genetics 172:1901-1914.

Rajaniemi TK. 2007. Root foraging traits and competitive ability in heterogeneous soils. Oecologia. 153:145-152.

Rao AN, Johnson DE, Sivaprasad B, Ladha JK, Mortimer AM. 2007. Weed management in direct-seeded rice. Adv. Agron. 93:153-255.

Reissig WH, Heinrichs EA, Litsinger JA, Moody K, Fielder L, Mew TW, Barrion AT. 1985. Illustrated guide to integrated pest management in rice in tropical Asia. Los Baños (Philippines): International Rice Research Institute. 411p.

Rose TJ, Impa SM, Rose MT, Pariasca-Tanaka J, Mori A, Heuer S, Johnson-Beebout SE, Wissuwa M. 2012. Enhancing phosphorus and zinc acquisition efficiency in rice: a critical review of root traits and their potential utility in rice breeding. Ann. Bot. 112(2):331-345.

Saleque MA, Kirk GJD. 1995. Root induced solubilization of phosphate in the rhizosphere of lowland rice. New Phytol. 129:325-336.

Sama VSAK, Himabindu K, Naik SB, Sundaram RM, Viraktamath BC, Bentur JS. 2012. Mapping and marker-assisted breeding of a gene allelic to the major Asian rice gall midge resistance gene Gm8. Euphytica 187:393-400.

Sandhu N, Jain S, Battan KR, Jain RK. 2011. Aerobic rice genotypes displayed greater adaptation to water-limited cultivation and tolerance to polyethyleneglycol-6000 induced stress. Physiol. Mol. Biol. Plants 18(1):33-43.

Sandhu N, Jain S, Chowdhury VK, Jain RK. 2012. Microsatellite diversity among aerobic and lowland indica rice genotypes with differential water requirements. Ind. J. Genet. 72(4):463-467.

Sandhu N, Jain S, Kumar A, Mehla BS, Jain R. 2013. Genetic variation, linkage mapping of QTL and correlation studies for yield, root, and agronomic traits for aerobic adaptation. BMC Genet. 14:104-119.

Sandhu N, Torres R, Sta Cruz MT, Maturan PC, Jain R, Kumar A, Henry A. 2014. Traits and QTLs for development of dry direct-seeded rainfed rice varieties. J. Exp. Bot. (communicated) (MS ID#: JEXBOT/2014/120410).

Sato H, Ando I, Hirabayashi H, Takeuchi Y, Arase S, Kihara J. 2008. QTL analysis of brown spot resistance in rice (Oryza sativa L.). Breed. Sci. 58:93-96.

Sen S, Smith ME, Setter TL. 2013. Analysis of maize root traits in response to low nitrogen. Asian J. Plant Sci. Res. 3(3):121-125.

Septiningsih EM, Ignacio JC, Sendon PM, Sanchez DL, Ismail AM, Mackill DJ. 2013. QTL mapping and confirmation for tolerance of anaerobic conditions during germination derived from the rice landrace Ma-Zhan Red. Theor. Appl. Genet. 126(5):1357-1366.

Sharma S, Xu S, Ehdaie B, Hoops A, Close TJ, Lukaszewski AJ, Waines JG. 2011. Dissection of QTL effects for root traits using a chromosome arm-specific mapping population in bread wheat. Theor. Appl. Genet. 122:759-769.

Shen L, Courtois B, McNally KL, Robin S, Li Z. 2001. Evaluation of near-isogenic lines of rice introgressed with QTLs for root depth through marker-aided selection. Theor. Appl. Genet.103:75-83.

Shikari AB, Khanna A, Krishnan SG, Singh UD, Rathour R, Tonapi V, Sharma TR, Nagarajan M, Prabhu KV, Singh AK. 2013. Molecular analysis and phenotypic validation of blast resistance genes Pita and Pita2 in landraces of rice (Oryza sativa L.). Indian J. Genet. 73(2):131-141.

398

Shipley B. 2006. Net assimilation rate, specific leaf area and leaf mass ratio: which is most closely correlated with relative growth rate? A meta-analysis. Funct. Ecol. 20(4):565-574.

SihathepV, Sipaseuth, Phothisane C, Thammavong A, Sengkeo, Phamixay S, Senthonghae M, Chanphengsay M, Linquist B, Fukai S. 2001. Response of dry-season irrigated rice to sowing time at four sites in Laos. ACIAR Proceedings. 101:138-146.

Song WY, Pi LY, Wang GL, Gardner J, Hoisten T, Ronald PC. 1997. Evolution of the rice Xa2I disease resistance gene family. Plant Cell 9:1279-1287.

Su JY, Qi Z, Li HW, Li B, Jing RL. 2009. Detection of QTLs for phosphorus use efficiency in relation to agronomic performance of wheat grown under phosphorus sufficient and limited conditions. Plant Sci. 176:824-836.

Su JY, Xiao YM, Li M, Liu QY, Li B. 2006. Mapping QTLs for phosphorus-deficiency tolerance at wheat seedling stage. Plant Soil 281:25-36.

Sugimoto K, Hoshino T, Marzougi S, Takeuchib Y, Yano M. 2013. Sdr7, a quantitative trait locus for seed dormancy in rice, encodes an ortholog of the Arabidopsis protein delay of germination. 1. Plant and Animal Genome XXI, 12-16 January 2013, San Diego, California.

Sugimoto K, Takeuchib Y, Ebanaa K, Miyaoa A, Hirochikaa H, Haraa N, Ishiyamac K, Kobayashic M, Band Y, Hattorid T, Yano M. 2010. Molecular cloning of Sdr4, a regulator involved inseed dormancy and domestication of rice. Proc. Natl. Acad. Sci. USA 107(13):5792-5797.

Suh JP, Jeung JU, Lee JI, Choi YH, Yea JD, Virk PS, Mackill DJ, Jena KK. 2010. Identification and analysis of QTLs controlling cold tolerance at the reproductive stage and validation of effective QTLs in cold-tolerant genotypes of rice (Oryza sativa L.). Theor. Appl. Genet. 120:985-995.

Sun L, Su C, Wang C, Zhai H, Wan J. 2005. Mapping of major resistance gene to the BPH in rice cultivar Rathu Heenati. Breed. Sci. 55:391-396.

Suralta RR, Inukai Y, Yamauchi A. 2008. Genotypic variations in responses of lateral root development to transient moisture stresses in rice cultivars. Plant Prod. Sci. 11:324-335.

Suralta RR, Inukai Y, Yamauchi A. 2010. Dry matter production in relation to root plastic development, oxygen transport and water uptake of rice under transient soil moisture stresses. Plant Soil. 332:87-104.

Swamy BPM, Ahmed HU, Henry A, Mauleon R, Dixit S, Kumar A. 2013. Genetic, physiological, and gene expression analyses reveal that multiple QTL enhance yield of rice mega-variety IR64 under drought. PLoS ONE 8(5):e62795.

Swamy P, Ajay N, Panchbhai, Dodiya P, Naik V, Panchbhai SD, Usha B. Zehr, Kasi Azhakanandam, Bharat RC. 2004. Evaluation of bacterial blight resistance in rice lines carrying multiple resistance genes and Xa21 transgenic lines. Australian Agronomy Conference. 12th AAC, 4th ICSC, Australia.

Takahashi A, Kobayashi A, Kakimoto Y, Fujii N, Takahashi H. 2003. Novel hydrotropism mutants of Arabidopsis thaliana and their altered waving response and phototropism. Biol. Sci. Space 17(3):209-210.

Takahashi H. 1997. Hydrotropism: the current state of our knowledge. J. Plant Res. 110(1098):163-169.

Tuong TP, Bouman BAM. 2003. Rice production in water-scarce environments. In: Kijne JW, Barker R, Molden D, editors. Water productivity in agriculture: limits and opportunities for improvement. Wallingford (UK): CABI. p 53-67.

399

Tuong TP, Bouman BAM. 2002. Rice production in water-scarce environments. Proceedings of the Water Productivity Workshop, Colombo, Sri Lanka. Colombo (Sri Lanka): International Water Management Institute.

Tuong TP, Bouman BAM, Mortimer M. 2005. More rice, less water: integrated approaches for increasing water productivity in irrigated rice-based systems in Asia. Plant Prod. Sci. 8:231-241.

Ullah I, Jamil S, Iqbal MZ, Shaheen HL, Hasni SM, Jabeen S, Mehmood A, Akhter M. 2012. Detection of bacterial blight resistance genes in basmati rice landraces. Genet. Mol. Res. 11(3):1960-1966.

Uga Y, Sugimoto K, Ogawa S, Rane J, Ishitani M, Hara N, Kitomi Y, Inukai Y, Ono K, Kanno N, Inoue H, Takehisa H, Motoyama R, Nagamura Y, Wu J, Matsumoto T, Takai T, Okuno K, Yano M. 2013. Control of root system architecture by DEEPER ROOTING 1 increases rice yield under drought conditions. Nat. Genet. 45:1097-1102.

Ventura W, Watanabe I. 1978. Growth inhibition due to continuous cropping of dryland rice and other crops. Soil Sci. Plant Nutr. 24:375-389.

Ventura W, Watanabe I, Castillo MB, dela Cruz A. 1981. Involvement of nematodes in the soil sickness of a dryland rice-based cropping system. Soil Sci. Plant Nutr. 27:305-315.

Venuprasad R, Bool ME, Dalid CO, Bernier J, Kumar A, Atlin GN. 2009. Genetic loci responding to two cycles of divergent selection for grain yield under drought stress in a rice breeding population. Euphytica 167:261-269.

Vikram P, Swamy BPM, Dixit S, Ahmed HU, Sta Cruz MT, Singh AK, Kumar A. 2011. qDTY1.1, a major QTL for rice grain yield under reproductive-stage drought stress with a consistent effect in multiple elite genetic backgrounds. BMC Genet. 12: 89-98.

Wang H, Bouman BAM, Zhao D, Wang C, Moya PF. 2002. Aerobic rice in northern China: opportunities and challenges. In: Bouman BAM, Hengsdijk H, Hardy B, Bindraban PS, Tuong TP, Ladha JK, editors. Water-wise rice production. Proceedings of the international workshop on water-wise rice production, 8-11 April 2002, Los Baños, Philippines. Los Baños (Philippines): International Rice Research Institute. p 143-154.

Wang L, Cheng J, Lai Y, Du W, Huang X, Wang Z, Zhang H. 2014. Identification of QTLs with additive, epistatic and QTL × development interaction effects for seed dormancy in rice. Planta 239:411-420.

Wassmann R, Jagadish SVK, Sumfleth K, Pathak H, Howell G, Ismail A, Serraj R, Redoña E, Singh RK, Heuer S. 2009. Regional vulnerability of rice production in Asia to climate change impacts and scope for adaptation. Adv. Agron. 102:91-133.

Watanabe T. 1997. Lodging resistance. In: Matsuo T, Futsuhara Y, Kikuchi F, Yamaguchi H, editors. Science of the rice plant.III. Genetics. Food andAgriculture Policy Research Center, Tokyo, Japan. p 567-577.

Wissuwa M, Yano M, Ae N. 1998. Mapping of QTLs for phosphorus-deficiency tolerance in rice (Oryza sativa L.). Theor. Appl. Genet. 97:777-783.

Yan XL, Liao H, Beebe SE, Blair MW, Lynch JP. 2004. QTL mapping of root hair and acid exudation traits and their relationship to phosphorus uptake in common bean. Plant Soil 265:17-29.

Yang L, Zheng B, Mao C, Yi K, Liu F, Wu Y, Wu P. 2003. cDNA-AFLP analysis of inducible gene expression in rice seminal root tips under a water deficit. Gene 314:141-148.

400

Yoshida S, Hasegawa S. 1982. The rice root system: its development and function. Drought resistance in crops with emphasis on rice. Manila (Philippines): International Rice Research Institute. p 97-114.Yoshida S, Satake T, Mackill D. 1981. High temperature stress. IRRI Research Paper Series 67:1-15.

Zhang H, Forde BG. 1998. An Arabidopsis MADS box gene that controls nutrient-induced changes in root architecture. Science 279:407-409.

Zhang Y, Lynch JP, Brown KM. 2003. Ethylene and phosphorus availability have interacting yet distinct effects on root hair development. J. Exp. Bot. 54:2351-2361.

Zhang Z, Qu X, Wan S, Chen L, Shu Y. 2005. Comparison of QTL controlling seedling vigour under different temperature conditions using recombinant inbred lines in rice. Ann. Bot. 95:423-429.

Zhu J, Lynch JP. 2004. The contribution of lateral rooting to phosphorus acquisition efficiency in maize (Zea mays) seedlings. Funct. Plant Biol. 31:949-958.

Zhu JM, Kaeppler SM, Lynch JP. 2005. Mapping of QTL controlling root hair length in maize (Zea mays L.) under phosphorus deficiency. Plant Soil. 270:299-310.

Notes Authors’ address: International Rice Research Institute, Los Baños, Philippines.

401

Table 1. Comparisons among irrigated lowland, direct-seeded, rainfed lowland, and upland rice.

Irrigated

lowland rice Direct-seeded

rice Rainfed lowland rice

Upland rice

Water availability High Moderate Moderate Low Fertilizer input level High High Moderate Low Puddling Yes No Yes No Crop establishment T/WDS/DDS DDS T/WDS/DDS DDS Drought tolerance Low Low to

moderate Moderate High

Yield High Moderate Moderate Low T = transplanted, WDS = wet direst seeding, DDS = dry direct seeding. Source: Modified from Nie et al (2012).

402

Table 2. Classification of the direct-seeded rice (DSR) system.

Source: Modified from Joshi et al (2013).

System of direct seeding

Seed bed condition

Tillage Sowing method

Seed environment

Suitable ecology

Direct seeding in dry bed

Dry seeds are sown in dry and mostly aerobic soil.

Dry tillage, reduced tillage, zero tillage

Broadcasting, drilling, or sowing in rows at depth of 2−3 cm

Aerobic

Mainly in rainfed area, some in irrigated areas with precise water control

Direct seeding in wet bed

Pregerminated seeds are sown in puddled soil; may be aerobic or anaerobic.

Dry and wet tillage

Broadcasting, line sowing, drilling

Aerobic, anaerobic

Mostly in favorable rainfed lowlands and irrigated areas with good drainage facilities

Direct seeding in standing water

Dry or pregerminated seeds are sown mostly in anaerobic conditions in standing water.

Dry and wet tillage

Broadcasting on standing water of 5−10 cm

Anaerobic

In areas with red rice or weedy rice problem and in irrigated lowland areas with good land leveling

403

Table 3. Difference between direct-seeded and irrigated rice. Eco-friendly direct-seeded rice Conventional irrigated rice

1. Precise land leveling Land should be normally leveled 2. Direct seeding Nursery raising is needed 3. Puddling,raising nursery,and

transplanting, not required Puddling, leveling, raising nursery,and transplantingare required

4. Matures 7−10 days early, harvest early, so that farmers can start sowing a subsequent wheat crop, leading to higher yield of the crop.

Matures late

5. 40–50% water savings; no constantmaintenance of water level

Constant maintenance of water level

6. Weeding can be mechanized (bullock pair)

Not possible (rotary weeder)

7. No need for trimming bunds andplugging holes

Requires constant attention by way of plugging holes and trimming bunds

8. Intercropping of any other arable crop is possible

Not possible

9. Crop rotation can be practiced Not common 10. Aerobic conditions in soil Anaerobic conditions prevail 11. Soil structure is maintained Destroyed. Subsurface hard pan

is made by repeated plowing. 12. Faster organic matter

decomposition Slower organic matter decomposition

13. Oxygenated rhizosphere is found

Not found

14. Higher water-use efficiency Low WUE 15. Efficient use of rain water Less efficient use of rain water 16. No occurrence of

methanogenesis Methanogenesis occurs

17. Better mineral nutrient dissolubility

Less dissolubility

18. Adsorption of NO2−/ NH4

+ to soil particles

Not so

19. Production of toxins such as ethanol and lactate absent

Toxins are produced

20. Reduced humidity in microclimate; healthy crop

High humidity

21. Incidence of diseases and pests is significantly low

High incidence of diseases and pests

22. Cost of cultivation is significantly low (less labor)

High cost of cultivation

Source: Modified from Parthasarathi et al (2012).

404

Table 4. Traits, donors, QTLs, and genes for direct-seeded rice.

Trait Donor Population QTLs/genes Interval markers References

1 Anaerobic germination

IR93312-30-101-20-13-66-6

Donor:Khao Hlan On Recipient: IR64

qAG9.1

RM8303−RM5526

Angaji et al (2010)

2 Blast resistance

WHD127 Two BAC clones, 75-1- 127BAC12 and 75-1-127BAC3, were fully sequence

Pi9 NBS2-Pi9 candidate gene, forward primer:195R-1, reverse primer: 195F-1; NBS3-Pi9 candidate gene, forward primer: NBS3-F, reverse primer: NBS3-R

Qu et al (2006)

3 Blast resistance

WHD-1S-75-1-127

Recipient: IR49830-7-1-2-2 BC6F5 population

Pi9 RM19814−RM3 Koide et al (2011)

4 Blast resistance

Tadukan Germplasm screening, used as resistant check

Pita2 RM7102−RM155 Fjellstrom et al (2004), Shikari et al (2013)

5 BPH Rathu Hennati

Recipient: KDML105 (BC3F2)

BPH3 RM589−RM588 (short arm of chromosome 6)

Jairin et al (2007)

6 BPH Rathu Hennati

Recipient: 02428 (F2 population)

BPH17 RM8213−RM5953 (short arm of chromosome 4)

Sun et al (2005)

7 Gall midge Abhaya F1, F2, and F3 populations from cross of BG380-2 and Gurmatia with Abhaya

Gm 4 Chromosome 4: E20−E20570(susceptible) E20−E20583(resistant)

Nair et al (1996)

8 Gall midge Aganni F2 plants and F10 GM 8 Chromosome8, Sama et al (2012)

405

RILs of the cross TN1/Aganni

flankingSSR, RM22685, RM22709, RM22674

9 Bacterial blight

IRBB60 F2 populations ofADT43 × IRBB60 and ASD16 × IRBB60

xa4+xa5+xa13+Xa21 Xa4: STS, MP1, and MP2, xa5 gene: FM-F, FM-R;RM122 (F and R); RG556; xa13 gene: (CAPS)RG136, RP7, and ST12; Xa-21: pTA248

Ullah et al (2012). Perumalsamy et al (2010), Chu et al (2006),Swamy et al (2004), Song et al (1997)

10 Drought tolerance

IR74371-46-1-1

IR74371-46-1-1/2* Sabitri

qDTY12.1

, qDTY2.3

, qDTY3.2

qDTY12.1

RM28166−RM28199, qDTY

2.3 RM3212−RM250, qDTY

3.2 RM22-RM545

Mishra et al (2013)

11 Drought tolerance

IR6322-34-223

Donors: N22, Apo N22/Swarna, Apo/2*Swarna

qDTY3.1

, qDTY1.1

, qDTY2.1

qDTY3.1 :RM416; qDTY

1.1: RM11943−RM12091; qDTY

2.1:RM324

Vikram et al (2011)

12 Grain yield DS

IR94225-B-82-B

Aus276 × IR64 qGYDS1.1

, qGYDS6.1 (Donor:

Aus276), qGYDS8.1 (Donor:

IR64), qGYDS

9.1,qGYDS

10.1(Donor: Aus276), qNR

4.1 (Donor:

IR64)

Flanking markers:id6010515−id6015531, id8000536−id8000845, id10005369-id10006378, id4001205−id4002844; peak markers: id6015531, id8003773/ud8001270, id10006378, id4001205

Nitika et al, unpublished

13 Early vigor, nodal root

IR94226-B-177-B

Aus276 × MTU1010

qNR5.1 (Donor: Aus276),

qEVV9.1 (Donor: Aus276),

qRHD1.1(Donor:Aus276)

Flanking markers: id5000759−id5001182, ud9000737−id9002704, id1005271−id1006691;peak markers: id5001182,

Nitika et al, unpublished

406

ud9000737,id1005271

14 Nematode tolerance

IR97153-B-114 or IR97153-B-123

IR7877-208-B-1-2 × Dinorado

qGYNT3.1

,qGYNT5.1

Linked markers: id3000090, id5004150 Marker region: qGYNT3.1-1.3; qGYNT5.1–32

Judith et al, unpublished

15 Nematode tolerance

IR97152-B-219/IR97152-B-280

IR7877-208-B-1-2 × Dinorado

qGYNT10.1

Linked marker: id10004327 Marker region: qGYNT

10.1-64.6

Judith et al, unpublished

16 Lodging resistance

IR91648-B-289-B

Donor: Swarna qLDG3.1

id3000090−id300057 Dixit et al, unpublished

17 Lodging resistance

IR91648-B-289-B

Donor Moroberekan

qLDG4.1

id4004461−id4009390 Dixit et al, unpublished

18 Early uniform emergence

IR91648-B-32-B

Donor: Moroberekan Recepient: Swarna

qEUE11.1

id1100085−id11001535 Dixit et al, unpublished

19 Seedling stage cold

IR66160-121-4-4-2

F7 and F8, Recipient: cold- sensitive japonica Geumobyeo

qPSST-3

RM569−RM231

Suh et al (2010)

20 Seedling stage cold

IR66160-121-4-4-2

F7 and F8, Recipient: cold- sensitive japonica Geumobyeo

qPSST-7

RM3767−RM1377

Suh et al (2010)

21 Seedling stage cold

IR66160-121-4-4-2

F7 and F8, Recipient: cold-

qPSST-9

RM24427−RM24545

Suh et al (2010)

407

sensitive japonica Geumobyeo

22 No seed germination dormancy

Nipponbare 98 BC1 F5 lines (backcross inbred lines) derived from a backcross of Nipponbare (japonica)/Kasalath (indica)//Nipponbare

RFLP markers:C1488 (chromosome3), R830 (chromosome5), R1440(chromosome7), R1245(chromosome7)

RM14784(chromosome3), RM2010/RM17800, RM17799 (chromosome 5),RM21556,RM21557,RM21558,RM1135 (chromosome7), RM21915,RM21917,RM21918(chromosome7) Source: Supplementary Table 18,information on all Class I SSRs(http://archive.gramene.org/markers/microsat/)

Lin et al (1998)

23 Seed dormancy

Kasalath Seed dormancy 4 (Sdr4) Japonica cultivars have only the Nipponbare allele (Sdr4-n), which gives reduced dormancy, whereas both the Kasalath allele (Srd4-k) and Sdr4-n are widely distributed in the indica group

Sdr4 was mapped in the interval between markers SNP1 and SNP8, recombination happened between RM1365 and SNP5 on chromosome7; 87535F and 88688R

Sugimoto et al (2010)

24 Seed dormancy

Progeny of SasanishikiHabataki

Seed dormancy 7 (Sdr7) 46-kb genomic region on chromosome 1, sequence comparison of the candidate region of each parent identified a homolog of ArabidopsisDelay of Germination 1 (DOG1) as a probable candidate

Sugimoto et al (2013)

408

25 Seed dormancy

Donor:Nonabokra chromosomesegment substitution lines (CSSLs SL-501 and SL519) derived from a cross between Nona Bokra (strong dormancy) and Koshihikari (weak dormancy)

Sdr6, Sdr9, Sdr10 (chromosome1)RM7278−RM3425, RM6902; (chromosome 9)RM7488−RM7311,RM5963; (chromosome 10) RM1161−RM3498,RM3207

Marzougui et al (2012)

26 Seed dormancy

150 recombinant inbred lines (RILs) (F2:9)

Derived from Jiucaiqing (japonica) × IR26 (indica)

4 weeks after heading qSD1.1,qSD2.2,qSD4.1,qSD4.2 5 weeks after heading qSD5.1 6 weeks after heading qSD2.1, qSD3.1,qSD7.1

RM128–RM6950,RM5404–RM208,RM273–RM3839, RM6246–RM280,RM3777–RM249,RM5699–RM13297, RM282–RM6080,RM8261–RM5426

Wang et al (2014)

27 Brown spot resistance

IR86126-45-B-B,IR86126-49-B-B, IR86126-104-B-B

(from IR36 × Dinorado)

bs1andbs2 (chromosome 12) RM277, RM1261 (SSR, co-dominant), SNP block markers for brown spot resistance on chromosome12 also identified in MAGIC population, lines and markers available with Chitra Raghavan and Hei Leung

Source: Dr. C.M.Veracruz, IRRI

409

Arsenic!

1-1.5 m/y

0.7 m/y

Groundwater depletion

Figure 1. Groundwater depletion under different rice culture systems.

410

Figure 2. Comparison of root length and biomass of indica/basmati and aerobic rice genotypes under direct-seeded aerobic cultivation conditions.

411

Figure 3. Correlation between root length and yield under dry direct-seeded aerobic conditions.

412

Figure 4. Correlation between root biomass and yield under dry direct-seeded aerobic conditions.

413

Paper 17 Alleviating soil sickness in tropical aerobic rice: a role for abiotic and biotic interactions

C.G.B. Banaay, C. Kreye, M. Hofte, A. Kumar, D. De Waele, V.C. Cuevas, C.M. Vera Cruz, and B.A.M. Bouman

Aerobic rice production systems are used to help ease water constraints to productivity. In aerobic systems, improved adapted rice varieties are grown in nonpuddled aerobic soil without standing water. Successive cultivations of this crop, however, lead to symptoms of soil sickness. Although breeding for increased yield potential of aerobic rice along with tolerance of drought and occasional flooding and resistance to diseases is continuous, yield decline caused by soil sickness due to successive aerobic rice monocropping needs to be immediately addressed to ensure sustainability of the system. The factors involved in soil sickness due to continuous aerobic rice monocropping include, but are not limited to, nutrient deficiencies, soil alkalinity, and buildup of root-infecting nematodes and fungi. However, the interactions between these factors make it difficult to exactly identify the primary agents involved in order to come up with effective control strategies. Several studies aiming to alleviate soil sickness in aerobic rice production, that target the single factors mentioned, only partly alleviate symptoms. Nonetheless, the reduction in symptoms was significant. Studies point to the importance of site-specific nutrient management as well as ensuring the supply of good-quality irrigation water. Cultural management involving crop rotations and fallow periods was also helpful in diminishing the harmful effects of continuous cultivation. Control of soil-borne root diseases through host-plant resistance, nonhost crop rotations, and organic matter amendment also plays a role in alleviating soil sickness. Because of the complex nature of soil sickness, no single solution can address all the factors involved and therefore an integration of several strategies is required. There is also a need to elucidate early events in the development of soil sickness in the hope of identifying the primary predisposing or triggering factor leading to its development.

Key words: aerobic rice, soil sickness, yield decline, nutrient management, crop rotation, soil-borne pathogens In 2009, a record one billion people globally suffered from unremitting hunger, of which an estimated two-thirds were in Asia (Asia Society and IRRI Task Force 2010). As the region continues to urbanize at unprecedented rates and as Asia’s population increases unrelentingly, food insecurity in the region could worsen unless immediate action is taken. Asia must increase food production despite water, land, and labor scarcity, while overcoming the new challenges posed by climate change. Rice plays a major role in ensuring food security as it is a staple food for most of the people in Asia and a source of income for millions of rice farmers. Rice production, however, is currently subject to ever-increasing environmental constraints that threaten its sustainability. Reduced availability of agricultural water in nearly all rice-producing countries (Arnell 1999, Vorosmarty et al 2000) is just one of the problems. The increasing water scarcity with decreasing quality, decreasing resources, and increasing competition from other uses calls for the development of new rice production systems and approaches that require less water than traditional flooded rice (Bouman et al 2007). In order to reduce water inputs and increase water-use efficiency, alternative agronomic practices are being proposed. These are saturated soil culture, alternate wetting and drying, and aerobic rice production systems (Bouman 2001). In the first two systems, floodwater depth is

414

decreased instead of keeping the rice field continuously flooded at 5–10 cm of water in order to reduce unproductive water outflow (i.e., seepage and percolation), thereby increasing water productivity (Tuong and Bouman 2003). In the third system, seepage and percolation are eliminated and evaporation is reduced as rice is grown under completely aerobic conditions like an irrigated upland crop such as maize and wheat. Among the three water strategies, the aerobic rice system potentially provides the greatest water savings, but with an accompanying yield reduction (Bouman et al 2007).

To achieve high yield under aerobic soil, new varieties that are high-yielding, nutrient-responsive, drought-tolerant, and weed-competitive are required (Atlin et al 2006, Zhao et al 2010). In Brazil and China, locally adapted aerobic rice varieties yielding 4.5–7 t ha−1 have been developed and production systems have become sustainable and profitable (Pinheiro et al 2006, Wang et al 2002, Yang et al 2005). However, Pinheiro et al (2006) point out that a yield decline occurs in continuous aerobic rice production in Brazil, and is therefore not sustainable as a monocrop, but this is circumvented by crop rotation. Likewise, Bouman et al (2007) have reported yield declines in successive aerobic rice cropping in the Philippines that have been partially alleviated by fallowing for one year. These progressive yield declines in Brazil and the Philippines were attributed to an auto-toxicity effect and to nematode buildup, respectively, which are reminiscent of the soil sickness syndrome observed in other crops (Grodzinsky 2006). This soil sickness problem has curtailed the adoption of tropical aerobic rice production systems and it needs to be addressed immediately. Although Prasad (2011) has given a general overview of the whole aerobic rice system, this review focuses on the problem of soil sickness.

Several studies have been done in the Philippines to determine the causes and alleviate the effects of soil sickness in continuous aerobic rice production. This paper aims to review the current research on the causes of yield decline in continuous aerobic rice monocropping in order to provide a coherent perspective of what has been done and what else can be done regarding this complex problem. The present discussion is likewise concerned with the interaction of abiotic and biotic factors leading to soil sickness as it occurs in tropical aerobic rice cultivation and shows how primary factors can predispose the host plant to detrimental effects of secondary factors that influence plant productivity. This review shows the complexity of soil sickness as various factors interact through time. Aerobic rice production: water savings, yield gap, and yield decline Aerobic rice is a water-saving technology being developed for water-scarce environments (i.e., water-short irrigated areas, rainfed lowlands, and favorable uplands) with access to irrigation. In this system, nutrient-responsive improved rice varieties are grown in nonpuddled aerobic soil without standing water, in which irrigation is applied to bring the water content in the root zone to field capacity after it has reached a certain lower threshold, usually halfway between field capacity and the wilting point (Bouman 2001). In this way, water evaporation is lessened since there is no standing water, and both seepage and percolation are largely diminished. The amount of deep percolation water, however, depends on soil properties and irrigation efficiency. Field application efficiencies of irrigation in this system range from 60−70% in flash/furrow irrigation to 80–90% in sprinkler/drip irrigation (Bouman et al 2005). Water savings can be as much

415

as 51% compared with flooded fields, but with a consequent yield reduction of 22–32% compared with that under flooded conditions (Bouman et al 2005) mainly due to water-stress effects under conditions when N is not limiting (Belder et al 2005). This difference between the yield of aerobic rice and flooded rice is called the yield gap. Breeding efforts are now directed toward increasing the yield potential of aerobic rice cultivars with stable performance across environments. Zhao et al (2010), describing a two-screening-environment methodology used in the aerobic rice breeding program of the International Rice Research Institute (IRRI), report on the development of several second-generation varieties that outperform the first-generation reference aerobic rice varieties Apo and UPLRi-7 in both nonstressed and stressed conditions. The new elite aerobic rice varieties identified in this study are valuable for water-short irrigated and drought-prone rainfed areas in tropical regions (Zhao et al 2010).

The aerobic rice production system is located in an area of “transition” from the lowland irrigated production system with its submerged/saturated and reduced soil conditions to the upland rice production system with its adapted usually low-yielding cultivars growing in thoroughly aerobic soils that may experience prolonged drought spells. In the aerobic rice system, soils are aerobic but the crop usually receives regular supplementary irrigation. Tropical aerobic rice does not usually experience more severe soil water tensions than the measurement range of a field tensiometer of up to 80 kPa. By comparison, the general wilting point for plants is designated to be pF 4.2 (1,585 kPa). Recommended irrigation thresholds for wheat and maize are often lower than that for rice. As outlined above, yields decline from flooded systems toward traditional upland systems.

Rice, being a semiaquatic plant, is very well adapted to growing in the favorable system of the irrigated lowlands for which high-yielding rice varieties have been developed. Depending on the cultivar, varieties may be extremely sensitive to even slight water deficits (Bouman et al 2007). Nevertheless, rice varieties adapted to “mild” water stress with “high-yielding” (up to 6 t ha−1 under nonstressed conditions compared with up to 10 t ha−1 of high-yielding lowland cultivars) characteristics (such as Apo) could be developed. In fact, further improved varieties are available now (Zhao et al 2010). Traditional upland varieties are often low yielding because they tend to lodge when fertilized and are often grown in unfavorable soils that include water stress as well as low fertility.

But, when the system is changed from predominantly flooded/irrigated to aerobic, the plant experiences not only a certain degree of “water stress” but also a change in soil chemistry and, with it, availability of nutrients and living conditions for soil microorganisms and pests and diseases (Buresh and Haefele 2010). The flooded irrigated lowland system is often very sustainable and resilient and can in extreme cases sustain continuously three rice crops per year (given adequate fertilizer application). For aerobic systems, this is not necessarily the case and often typical upland crops such as cereals are integrated into crop rotations to overcome negative effects of monocropping. In temperate aerobic rice, successive crops are also avoided.

At the IRRI farm, a “long-term” field experiment comparing aerobic with flooded rice was started to further identify the yield attributes responsible for the yield gap between aerobic rice and flooded rice, and to test for potential adverse effects of the new cropping system (Peng et al 2006). The soil at this site was a typic Tropaqualf, with

416

59% clay, 32% silt, and 9% sand, a total C content of 19.8 g kg−1, and pH of 6.7. Puddled plots were kept flooded with initially 2-cm water depth gradually increasing to 10 cm at full crop development. Irrigation practice in the aerobic plots was through flood irrigation when the soil at 15-cm depth reached −30 kPa except during flowering stage, when the threshold for irrigation declined to −10 kPa (field capacity). Thus, severe water stress in the aerobic treatment was avoided. Initial results were analyzed after growing eight seasons of aerobic and flooded rice. In this study, the yield gap was attributed more to differences in biomass production and to sink size (spikelets m−2). Xie et al (2008) likewise identified sink size as a contributing factor to limitations in aerobic rice yield and recommended future research on the effects of water regimes on tiller dynamics since these have been determined to affect spikelet number per panicle.

In addition to a yield gap, a yield decline through seasons of continuous aerobic cropping was also observed in the study by Peng et al (2006). The yield difference between first cropping (6.32 t ha−1) and seventh cropping (3.77 t ha−1) equated to a yield reduction of 40%. According to their study, the yield decline was attributed more to changes in biomass production than to harvest index since, unlike biomass production, there was no increasing trend in the differences in harvest index between aerobic and flooded rice through the seasons, and there was no significant difference between the harvest index of first- and seventh-season aerobic rice. This phenomenon of yield decline has been reported previously (Belder et al 2005, Bouman et al 2005) and the development of crop management strategies that can reverse it as well as the development of new varieties with a minimum yield gap were recommended (Peng et al 2006). The emerging problem of soil sickness in continuous aerobic rice production A decline in the quality (including biomass, yield, and general plant vigor) of a crop after successive plantings on the same land has been observed since ancient times (Grodzinsky 2006). Farmers long ago knew about the negative effect of successive monoculturing on crop productivity; hence, the development of different cropping systems. Studies on the specific causes of these negative effects started under field conditions in the beginning of the 19th century. But, it was only in the 1960s when the whole syndrome of decreasing crop quality, progressive yield decline, and outbreaks of soil-borne diseases caused by successive replanting of the same crop on the same land was called “soil sickness.”

Soil sickness, however, is a general term for the whole phenomenon and it does not identify the specific causal agents because of the multifaceted nature of the problem. To add complexity to the whole issue, each case of soil sickness is specific to the site and to the crop. As such, it is hard to find a consensus as to the nature and origin of the problem. Nevertheless, several interwoven factors have been proposed to explain, at least in part, the observed continuous decline in crop productivity. These factors comprise a buildup of soil-borne pathogens (such as nematodes and fungi) and other pests, nutrient deficiencies/imbalances, adverse change in soil structure and physico-chemical properties, and allelopathy (Grodzinsky 2006, Nishio and Kusano 1975, Ventura and Watanabe 1978, Ventura et al 1981).

417

In a series of field experiments from 1974 to 1977, Ventura and Watanabe (1978) examined soil sickness in dryland rice-based cropping systems at the International Rice Research Institute (IRRI). Their results indicated that, despite sufficient nutrient and pest management, the inhibitory effect of continuous cropping occurred as early as the second or third successive cropping and the harmful effects persisted in the soil once established. Converting to flooded soil for the wet season, fallowing for 5 months, drying the soil for at least 30 days, and chemical soil sterilization, however, all partially reduced soil sickness. The inhibitory agents also seem to remain in root residues and may have served as a source of infection for the succeeding susceptible crops.

In follow-up experiments conducted in 1978 and 1979 (Ventura et al 1981), interaction of root-infecting fungi and nematodes was suspected. The study also showed that interactions among biotic and abiotic factors were involved in soil sickness. It was clear from these early studies that the development of soil sickness in dryland rice-based systems was rapid, persistent, crop-specific, involved several biotic and abiotic factors, and, perhaps most importantly, was reversible or at least reducible. Unfortunately, these studies were terminated in the 1980s and few follow-ups were made until about 10 years ago when interest in aerobic rice production was renewed in response to the looming water crisis. Starting in 2000, IRRI has dedicated significant resources to the development of a new generation of aerobic rice varieties that have higher yield than traditional varieties, are drought-tolerant, possess resistance to/tolerance of diseases, and are responsive to high inputs. The problem of soil sickness, however, persisted in the succeeding field experiments and research once again focused on this particular constraint in aerobic rice cropping.

In 1998-2000, experiments on continuous aerobic rice monocropping were conducted in irrigated fields in Siniloan, Laguna, Luzon, Philippines (George et al 2002). This report also included similar studies in favorable rainfed upland fields in Claveria, Misamis Oriental, Mindanao, Philippines, which were conducted in earlier years. In both cases, rapid yield loss of rice grown in successively cropped aerobic soils was observed despite sufficient management of soil water and nutrients to maintain relatively favorable conditions. Yield declined by as much as 73% while crop biomass exhibited decreases of 16% to 79%. Lower grain yield in succeeding years was associated with reduced plant height and tiller production and hence decreased biomass production. It was clear from the experiments that the large yield losses could not be accounted for by a soil fertility (i.e., soil nutrient content) decline. Furthermore, the observed decline in crop performance after a high first-season yield was observed in all rice varieties tested but not in maize. Interestingly, the yield decline of successive cropping was partially reversed in this study when a one-season break was introduced through cultivation of a maize-cowpea intercrop. These observations clearly indicated that soil sickness was specific to successive aerobic rice monocropping and that nonrice crop rotation mitigates the rapid yield decline.

The synchronous interaction between host, pathogen, and environment governs the development of disease and is commonly known as the disease triangle. Whether or not a disease condition develops depends on the influence of environmental factors on the inherent response of the host plant to the presence of the pathogen or its metabolites. If one of these factors is absent, the disease does not develop; hence, the three are equally important. In this context, the environment is perceived not only as

418

providing the conditions conducive to the growth of the pathogen and its interaction with the host during infection but also as an agent acting prior to infection to affect the response of the host plant and make it more prone to disease—a concept called predisposition. This concept goes both ways since the environment may induce changes in the plant toward either greater or less susceptibility; therefore, an environmental factor or set of factors that predisposes the host plant to disease can be manipulated in order to prevent the disease.

Unfavorable environmental factors are known to cause stress to the plant and reduce plant vigor or vitality and therefore also its overall performance. This stressful environment also influences disease development as it affects host susceptibility, pathogen growth and aggressiveness, as well as the host-pathogen interaction (Schoeneweiss 1975). In the case of soil sickness, no single pathogen or environmental factor can be presently identified to cause the sickness, rather a multitude of factors have been implicated to contribute to the whole disease syndrome. However, the development of soil sickness in aerobic rice cultivation occurs under similar circumstances (i.e., successive monoculture of plants in aerobic soil), which indicates action of similar agents at least at the start, prior to establishment of the soil sickness syndrome. However, there is a wide variety of symptoms and range of expressions of this sickness, which clearly indicates various intermediate effects after the onset of the primary effect(s). Herein lies the differential influences of crop, soil type, agricultural management regime, and various biotic factors, which aggravate/intensify the negative effects of the primary factor(s).

In the case of aerobic rice, the primary predisposing environmental factor, which can be referred to as a triggering factor, seems to be related to water stress coupled with continuous monocropping since this is the common thread in all soil sickness syndromes observed so far. This hypothesis is proposed from field observations of rice productivity under different water regimes. Long-term experiments conducted at IRRI show the sustainability of continuous paddy rice cultivation yet soil sickness develops in as few as 1−3 years for continuous aerobic rice cultivation (George et al 2002, Nie et al 2007b, Peng et al 2006) and 1 year for upland rice (Ventura and Watanabe 1978). Other water-saving technologies, such as saturated soil culture (SSC) and alternate wetting and drying (AWD), have not been observed to lead to the development of soil sickness (Wassman et al 2009). In SSC, soil is kept saturated and plants do not suffer drought stress. In AWD, controlled regulation of water levels (irrigation at a threshold soil water potential at 15-cm depth of −20 kPa) does not allow plants to experience prolonged drought stress, especially at critical stages, thus making this a widely accepted technology in China (Bouman et al 2007). It was recently proposed for aerobic rice cultivation that more frequent irrigation be applied or re-watering be done at much milder deficits. This was initially suggested in order to prevent a reduction in yield potential but perhaps this may also prevent triggering the development of soil sickness.

A possible role for allelochemicals in the early phase of soil sickness development in continuous aerobic rice cultivation The rapid onset of soil sickness symptoms (i.e., yield decline in the second cropping) could possibly point to the chemical nature of the primary factor or to biotic factors that quickly build up to exert enough damage such as root-knot nematodes that have fast

419

generation times and cause damage as early as the first cropping season. The role of biotic factors will be discussed further in the last section.

Since nutrients have been adequately managed in earlier experiments in which soil sickness was observed, it seems that the primary factor originates from the plant or from microbial activities in the rhizosphere as a response to the stress imposed by suboptimal water conditions and that nutrient deficiency, if present, is only an intermediate effect. Nutrient deficiencies have been previously identified as secondary factors that intensify the negative effect of primary factors contributing to soil sickness (Grodzinsky 2006).

Allelochemicals have been proposed as one of the factors that act initially during the development of soil sickness owing to its early detection (Dzubenko 2006). Rice root exudates as well as root residue decomposition products have been suggested to cause auto-toxicity problems in continuous monocropping of upland rice and in one instance in double-cropped rice systems in Taiwan, Brazil, and the Philippines (Chou 1998, Fageria and Baligar 2003, Jensen et al 2001, Olofsdotter 2001). This indicates that auto-toxicity effects occur as early as the second cropping season or within a year of continuous planting. This is earlier than the onset of nutrient deficiencies and disease-causing organisms, both of which occur after several more seasons of continuous planting. In wheat, water deficit has been shown to induce the production and accumulation of allelochemicals (Zuo et al 2010), which makes it a possible primary factor in the development of soil sickness because it is a direct response of the plant to water stress. Similarly, rice allelochemicals are released in response to environmental stimulus and may inhibit plant nutrient uptake, leading to problems of nutrient deficiencies later on (Fageria and Baligar 2003). Auto-intoxication in rice has been reported as early as the second cropping season, leading to reduced yield, N unavailability, and reduced concentration of cationic micronutrients (Chou 1998). Some studies also implicate allelochemicals as agents causing increased susceptibility to pathogens by weakening plant immunity (Morozov 2006). The effect of water stress on rice allelochemical production in the context of soil sickness development should be further investigated. Allelochemical auto-toxicity may not only cause nutrient deficiencies but also predispose plants to disease susceptibility, both of which, in turn, could further aggravate the toxic effects of allelochemicals with continuous cropping, thus creating a vicious circle of intensifying sickness through time.

Information about rice allelochemicals and their interaction with other abiotic and biotic factors is fragmentary and more research is needed before these can be clearly identified as one of the primary factors contributing to the development of soil sickness. Allelochemicals will not be dealt with in detail in this review due to a lack of information relating them to soil sickness in tropical aerobic rice cultivation. However, this is an important area of research that could help elucidate the link between water stress, successive monoculture, and soil sickness.

At present, very little is known about the primary factors and the early events leading to soil sickness in continuous aerobic rice production. However, an ongoing study at IRRI is exploring the effect of different water regimes on aerobic rice performance together with monitoring of abiotic and biotic constraints to productivity. This study may help to further elucidate the interactions between level of water stress and the development of soil sickness. Likewise, another project aims to investigate

420

aerobic rice root residues in different water regimes and cropping schemes. Studies have already shown that water stress predisposes plants to different types of diseases, including those infecting roots (Schoeneweiss 1975). Predisposition of plants to diseases as affected by water stress imposes an elastic strain and this can be alleviated; hence, subsequent recovery of resistance after removal of the stress is often observed (Schoeneweiss 1975). Experiments on the effect of different water regimes, including various levels of exposure to water stress, on plant disease susceptibility should be investigated.

Despite the lack of information on the initial events leading to soil sickness, many studies have been conducted that point to the effect of secondary abiotic and biotic agents and these will be reviewed in the succeeding sections. Abiotic factors contributing to soil sickness in continuous aerobic rice cultivation Alleviating soil sickness through nutrient management Efforts to further understand and alleviate soil sickness caused by aerobic rice monocropping became the theme of several studies at IRRI conducted from 2000 to 2007. A series of field and pot experiments, centered on the “long-term” field experiment from which the yield decline had been reported by Peng et al (2006), was conducted during those years to determine several management options to lessen or reverse the effect of soil sickness. These studies also tried to understand whether the main direct factor of the soil sickness problem was of biotic (pests and diseases) or abiotic (nutrition) origin. Nie et al (2007b, 2008, 2009a) identified nutrition, especially N, as an important constraining factor. They set up several pot tests and field trials in which a long-term field experiment started in 2001. Nie et al (2007b) reported that oven-heating of sick soil up to at least 90 C for 12 hours and to 120 C for as short as 3 hours was enough to bring about significant growth enhancement. The heating could have killed the biotic factors contributing to aerobic soil sickness such as nematodes and fungi (although this was not tested in their study) but it may also have facilitated the release of nutrients from the soil through enhancement of mineralization or transformation of nutrients to available forms. Alternatively, heating may have also transformed a phytotoxic compound to a nontoxic one. Although biotic factors cannot be discounted altogether, the authors suggested that abiotic factors were more likely to have caused soil sickness associated with continuous aerobic rice monocropping. The linear responses in leaf area and total biomass to linear mixtures of heat-treated and untreated aerobic sick soil also implied that the cause of soil sickness was probably chemical in nature although the exact identity of the chemical agent was unknown. It was also determined that growing aerobic rice for two consecutive seasons already caused soil sickness and that rotation with flooded rice for three seasons recovered only 50% of the yield loss, which was consistent with previous field experiments done by Ventura and Watanabe (1978). However, this also seems congruent with the effect of allelochemicals, hence further strengthening the need for more studies in this area. To tease out the nutrient issues, Nie et al (2008) observed that, of the three macronutrients applied, N improved plant growth while P and K had no effect. Although a previous study (Nie et al 2007a) showed that solophos (P fertilizer) improved plant growth in aerobic sick soil, the mechanism of growth promotion could not be explained clearly

421

because it was inconsistent with the soil analysis and the history of P application in the field. Other results of the study likewise suggested that P deficiency was unlikely to have caused soil sickness. Nie et al (2008) were able to confirm that P deficiency was not a problem and only increasing the rate of N (as urea) application improved the growth of continuously cultivated aerobic rice. Their results suggest that N deficiency due to poor soil N availability or reduced plant N uptake might be contributory to the yield decline in continuous cropping of aerobic rice. Shifts in water management from flooded to aerobic conditions are known to influence the availability and form of N present in the soil (Savant and De Datta 1982, Nie et al 2009a). N fertilizers, upon application to the soil, undergo numerous reactions, transformations, and various N loss mechanisms such as NH3 volatilization, nitrification and denitrification, leaching, chemical and microbial immobilization, and surface runoff (Sharif Zia et al 2001, Robertson and Groffman 2007); thus, they may be unavailable for plant uptake and supplies may become increasingly low with continuous aerobic cropping.

Determining the response of aerobic rice to various nitrogen sources was the next logical course of action since N application was found to partly alleviate soil sickness in continuous aerobic rice cultivation. Two N forms are usually used as fertilizers in rice cultivation: ammonium-N and nitrate-N. Nie et al (2009a) were able to show in their field experiment that growth of aerobic rice was generally better with ammonium-N than with nitrate-N. Qian et al (2004) conducted pot experiments to study the influence of NO3

− to NH4+ ratio on the yield and uptake of N by rice grown under

upland conditions. Their study showed that, although the rice crop did not show a significant preference for either NH4

+ or NO3− alone, in general, ammonium-N was a

better single-N source than nitrate in bringing about better vegetative growth of rice grown in aerobic soil conditions. One possible reason is that the nitrate form of N is more subject to leaching owing to its negative charge that is repelled by cation exchange centers in the soil (Wolf 1999). Also, it is the ammonium-N form that is directly assimilated into plant cells through the GS-GOGAT reactions.

Both nitrate- and ammonium-N sources supplied in equal amounts, however, significantly enhanced the growth and N-use efficiency of upland rice. One possible reason is the effect of NO3

− on increasing abundance of GS-GOGAT enzymes (specifically chloroplastic GS2 and Fd-GOGAT) in rice cells (Hayakawa et al 1990), which may serve to enhance NH4

+ assimilation derived from multiple sources (Bernard and Habash 2009, Cren and Hirel 1999). The advantages of a dual- or mixed-N source in rice growth are corroborated by other studies (Duan et al 2007, Kronzucker et al 1999, 2000, Lin et al 2005). It can be inferred from Nie et al (2009a) experiments that, under field conditions in which, presumably, part of the ammonium-N fertilizer applied is transformed to nitrate due to the aerobic conditions, the dual-/mixed-N source condition is actually in effect, albeit inadvertently. In addition, this study also showed that ammonium sulfate was more effective in improving vegetative plant growth, N nutrition, and grain yield than urea at high N rates even though they are both ammonium-N fertilizers. The reason why ammonium sulfate was more effective than urea in alleviating soil sickness was not clear in this study but may be related to the greater soil-acidifying effect of ammonium sulfate or to urea-induced ammonia toxicity (Haden et al 2011). Alternatively, ammonium sulfate could have contributed to the production of sulfur-containing compounds in plants that have a significant role in abiotic stress and

422

pathogen defense responses (e.g., glutathione S-transferases) that may be ultimately translated into improved vegetative plant growth. For a more extensive review of the advantages of ammonium sulfate over other N and S fertilizers, readers are referred to a recent paper by Chien et al (2011).

A follow-up study was conducted to determine the effect of soil acidification on soil nutrient availability, plant nutrition, and growth of continuously cropped aerobic rice, especially since it was discovered that alkaline irrigation water (pH 8.1) could have helped promote soil sickness in the experimental field being studied (Xiang et al 2009). Pot experiments were carried out using soil from a field where aerobic rice had been grown for 13 consecutive seasons (K6/K7). The soil was acidified through the addition of varying amounts of sulfuric acid to achieve a range of soil pH. Aerobic rice was then sown and grown with different rates of N application. Plant growth and N uptake improved significantly with soil acidification regardless of N rates or N sources. This improved growth was associated with an improvement in plant N nutrition as reflected by a 5.5-fold increase in soil ammonium and a 1.5-fold increase in soil nitrate. An increase in available Zn, exchangeable Mg, and exchangeable Ca was also observed in acidified soil. Increased plant uptake of P, K, S, Zn, Al, and Cu was also observed as well as a reduction in in planta concentrations of Na. The application of N, however, had greater positive effects on plant growth and N uptake than soil acidification alone. The growth response to soil acidification declined as the rate of N application increased. Plant growth was also found to be consistently better with ammonium sulfate than with urea, thus confirming earlier reports by Nie et al (2009a). Xiang et al (2009) suggested that the yield decline of continuous aerobic rice is probably associated with a reduction in soil N availability and plant N uptake as a result of a gradual increase in soil pH since acid treatments significantly improved soil and plant N status even without additional N applied. They hypothesized that three processes might explain the observed trends. Acidification could have temporarily stimulated N mineralization, it could have reduced N immobilization, or it could have hydrolyzed organic N forms already present in the soil.

Aside from studies on the role of nitrogen, other unpublished studies conducted at IRRI and at other IRRI experiment stations in other Philippine provinces suggest the involvement of micronutrient deficiencies in soil sickness. However, results were inconclusive and more studies need to be conducted to verify their involvement. In one investigation, however, iron deficiency has been definitely identified as one of the serious nutritional problems of aerobic rice and foliar and soil applications ameliorated the Fe-deficiency syndrome (Pal et al 2008)

All the experiments described earlier agree that proper application of the correct N fertilizer together with the provision of optimum levels of essential micronutrients and close monitoring of soil pH may help alleviate soil sickness caused by continuous aerobic rice monocropping. Further studies on possible management strategies are warranted to further examine nutrient issues in relation to aerobic rice. Nevertheless, other causes of yield decline such as allelopathy or the negative interaction of the crop with soil life in general may not be completely ruled out and this needs to be studied. Alleviating soil sickness through crop rotation

423

Crop rotation is an agronomic practice that involves growing different crops (preferably from different families) on the same land in recurrent succession and definite sequence (Sumner 1982). This results in a temporal diversity that interrupts pest life cycles by eliminating the host in successive cropping (Brown and Marten 1986). Beneficial effects derived from crop rotation are attributed to control of diseases, insects, and weeds; improvement of overall soil quality, including soil fertility; and improvement of nutrient- and water-use efficiency as different plants have different ways and rates of withdrawing nutrients and water from the soil (Fageria and Baligar 2003, Sumner 1982). Earlier studies have indicated the effectiveness of crop rotation in reducing damage caused by soil sickness in continuous monocropping of upland rice (Fageria and Baligar 2003, George et al 2002, Ventura and Watanabe 1978) and this system may prove to alleviate soil sickness caused by aerobic rice monocropping since upland and aerobic rice are planted in similar environments characterized by water stress, albeit at a milder water-deficit condition for aerobic rice due to the definite irrigation schemes prescribed for the technology.

To alleviate soil sickness caused by successive aerobic rice cultivation, it was suggested that continuous monocropping be avoided altogether and crop rotation be encouraged (Pinheiro et al 2006). Nie et al (2009b) examined the response of aerobic rice to fallow, flooding, and crop rotation in the same “long-term” aerobic rice field where the studies on nutrient management were also conducted. The field experiments showed that the yield of aerobic rice after two-season fallow and after three-season flooded rice was significantly higher than that of continuous aerobic rice. Yield was also greater after two seasons of upland crops than after two seasons of fallow. The experiments suggest that crop rotation and fallow could be used to reduce the yield decline caused by continuous aerobic rice cultivation. This supports the findings of George et al (2002) on the beneficial effect of crop rotation in partially alleviating yield decline in continuously cropped aerobic rice. Ventura and Watanabe (1978) also observed partial alleviation of soil sickness in crop rotation with sorghum and with leguminous crops such as mungbean and cowpea, as well as after a fallow period of at least 5 months. The beneficial effects of crop rotation on growth and yield of aerobic rice could be attributed to improving soil fertility, increasing water-use efficiency, and controlling diseases, insects, and pests. In particular, root-knot nematodes (RKN such as Meloidogyne graminicola) were found in this study (Nie et al 2009b) to be less in both continuously flooded rice and three seasons of flooded rice after continuous aerobic rice than in continuously aerobic fields. Soriano and Reversat (2003) also recommended that, to ensure higher upland rice yield, M. graminicola populations should be maintained at low density by nonhost (such as the legumes peanut and cowpea) crop rotations, ideally for two seasons before re-planting rice. Leguminous plants have been successfully used in intercropping and crop rotation for pest management (Brown and Marten 1986) and have been recommended for crop rotation with cereals such as rice (Fageria and Baligar 2003). In addition, appropriate crop rotation schemes have been recommended to reduce or eliminate allelochemical phytotoxicity (Fageria and Baligar 2003). Aside from preventing buildup of plant parasitic nematodes, some Rhizobium spp. in leguminous crops are known to be biocontrol agents of root diseases such as Pythium damping-off aside from contributing to soil fertility by virtue of their N2-fixing ability (Huang and Chou 2005).

424

Biotic factors contributing to soil sickness in continuous aerobic rice cultivation Aside from abiotic factors, previous studies on continuous upland rice cultivation indicated the involvement of biotic factors such as root-knot nematodes (Soriano and Reversat 2003) and root-infecting plant pathogenic fungi (Nishio and Kusano 1975) in yield decline. Initial studies on the possible biotic causes of yield decline in continuously cropped aerobic rice in IRRI field stations have been conducted since 2005. Several soil-borne root-infecting pathogens were observed and isolated. Four fungal genera, Pythium, Rhizoctonia-like, Fusarium, and Sclerotium-like, were isolated from discolored roots of affected plants (Kreye et al 2009b). Among these, Pythium isolates were the most predominant. Pythium species are oomycetes that commonly infect the roots of seedlings, and are said to adapt well to variable environmental conditions because of their highly flexibile life cycles (Jeger and Pautasso 2008). Their oospores have been reported to either germinate directly or produce cysts through sporangia and zoospores. Succeeding pathogenicity tests on aerobic rice variety Apo, however, revealed that only a few of these isolates are actually pathogenic. In particular, Pythium arrhenomanes was found to be highly pathogenic, causing preemergence damping-off and stunting of seedlings that were able to emerge (Banaay 2011, Van Buyten et al 2013). A few of the Pythium graminicola isolates are only slightly pathogenic, whereas Pythium inflatum isolates are slightly pathogenic to nonpathogenic, with some apparently exhibiting growth-promoting activity.

Although highly pathogenic species of Pythium have been isolated in soil-sick fields planted to aerobic rice, it seems unlikely that they are primary factors causing soil sickness. Their occurrence is low in the field and population dynamics studies in aerobic rice roots show their decline after the tillering stage (Pinili et al 2009). Also, initially infected seedlings often survive and recover at maturity. The presence of other antagonistic microorganisms in the soil may also limit their buildup to potentially damaging levels. However, continuous cropping of a susceptible plant may cause a differential increase in pathogenic species after several seasons. There have been indications that a resurgence of pathogenic Pythium spp. implicated in the soil sickness in aerobic rice cultivation was observed after two years of continuous planting (Pinili et al 2009). This particular field plot was planted with aerobic rice for a few years when pathogenic Pythium spp. were initially detected. A strong typhoon inundated the field with new soil from somewhere else and the Pythium population composition changed from predominantly pathogenic to predominantly nonpathogenic. However, after two years of re-establishment of aerobic rice monocropping in the area, pathogenic Pythium spp. were once again observed. This delay in the buildup of Pythium may mean that it is only an intermediate factor in soil sickness development since it comes in later compared with other factors.

In contrast, root-knot nematode buildup occurs quickly and studies indicate yield losses as early as the first year of cropping (Soriano and Reversat 2003) for susceptible varieties. This may indicate that RKN is a potential primary factor contributing to the establishment of soil sickness whose damage may be further aggravated by the effect of intermediate factors such as Pythium spp. and the resulting nutrient deficiencies and root function impairment caused by the two. To prevent buildup of RKN and mitigate

425

downstream effects of its interaction with confounding factors, crop rotations with nonhost plants and planting of resistant varieties are recommended.

Studies on the population dynamics of both Pythium spp. and Meloidogyne graminicola in an IRRI experimental field plot show that Pythium spp. are more frequently isolated during the earlier stages of growth, with a peak during maximum tillering stage, while M. graminicola populations continuously increase up to maturity, reaching 100% frequency of occurrence. Between Pythium and M. graminicola, the latter appears more widespread and able to maintain high populations until harvest in the field investigated (Pinili et al 2009). The population dynamics and distribution map presented in the initial survey conducted by Pinili et al (2009) indicated an active role played by the two organisms. The occurrence of M. graminicola in 100% of the root samples of plants at the reproductive growth stage is of great concern since this parasite has the potential to cause significant yield losses in upland and rainfed lowland rice production (Bridge et al 2005, Prot and Matias 1995, Soriano and Reversat 2003). This parasite will likely emerge as an important contributor to yield decline as the trend toward intensification of production and the increasing tendency for aerobic rice production due to scarcity of irrigation water could support the proliferation of nematode populations (Prot and Rahman 1994).

To better understand the relative contribution of Pythium and M. graminicola to the observed disease symptoms in the field, a follow-up study was done under controlled conditions. The response of several selected varieties of upland rice to single and simultaneous inoculation of a virulent strain of Pythium arrhenomanes and M. graminicola was evaluated under the controlled conditions of a phytotron (Banaay et al 2010). During the seedling stage, stunting was more severe and mean root length was shorter in Pythium-inoculated plants than in RKN-inoculated plants. In plants inoculated with both pathogens, growth trends either follow those of plants treated with Pythium alone or display more severe disease symptoms than singly inoculated plants. As the plants mature, however, RKN becomes more damaging while the effect of Pythium diminishes. Root galling rates increase while root discolorations in Pythium-inoculated plants decrease. These trends are consistent with the known lifestyles of both Pythium and RKN. Whereas Pythium causes seedling disease infecting soft root tissues, RKN is more prominent in mature plants when roots have become more extensive. This study also opens the possibility of either a disease complex or an antagonistic relationship developing when both organisms are infecting the same plant. Even though the effect of Pythium may be overcome at maturity, its effect may be retained in relation to its ability to either enhance or reduce the damaging effects of RKN on the plant at later stages of growth. Whether a synergistic or antagonistic relationship occurs between Pythium and RKN infecting the same plant depends on the inherent susceptibility, tolerance, or resistance of such a plant to the pathogens in question.

Banaay et al (2010) reported that the observed tolerance of rice variety Apo of M. graminicola can break down when populations of RKN in the soil are very high. The resistance to or tolerance of these disease-causing organisms, be they major or minor contributors to yield decline, is important for the sustainability of aerobic rice production. Seedling vigor is also a very important trait to ensure good establishment in direct-seeded aerobic fields. It may also be worth investigating whether tolerance of/resistance to these organisms will break down while facing nutrient deficiencies or soil pH

426

fluctuations, which is the reality under field conditions. There is some indication, albeit unconfirmed at present, that this is so. Currently, several traditional varieties and IRRI rice lines have been screened for tolerance of/resistance to either Pythium sp., M. graminicola, or both. Confirmation of responses and testing of other varieties are ongoing, especially against M. graminicola.

The root damage observed in aerobic rice soil sickness is most likely of complex etiology. Multiple infections are more likely to occur than single infections because these root-infecting organisms occupy similar niches. The wounds caused by one infection may be entry points for other infections. The injuries caused by so-called minor pathogens such as Pythium spp. may be heightened in mixed infections such as in the case of disease complexes with RKN. The effect of other abiotic agents such as water stress, phytotoxins, nutrient deficiencies, and other suboptimal edaphic conditions is also likely to intrude upon the disease etiology; hence, attempts should be made toward defining these predisposing factors. A scheme designed to address only one factor becomes ineffective and therefore control measures should be integrated.

Organic matter (OM) amendment is one control measure that can be considered when designing integrated management schemes to address the soil sickness problem. Soil OM can address phytotoxicity since allelochemicals may be adsorbed in humic fractions and organic acids may neutralize the toxins (Morozov, 2006). Nutrient deficiencies are likewise prevented since soil OM is a repository for plant nutrients. Injurious biotic factors are also minimized with increased soil OM. Banaay et al (2013) observed that OM amendment decreases Pythium spp. occurrence and % root galling caused by RKN in aerobic rice variety Apo. The effect of OM amendment in this study was related to the increase in both bacterial and fungal diversity. This could mean a general suppression exerted by the whole microbial community. A specific suppression, however, cannot be discounted since the activity of biocontrol organisms has been substantiated in many studies investigating suppressive soils. Although soil sickness in continuous aerobic rice cultivation is a major problem, it is worth mentioning at this point that the problem of emerging root diseases such as those caused by root-knot nematodes is also occurring in rainfed upland areas as well as in irrigated lowland rice fields where irrigation is becoming a problem. For instance, under irrigated lowland and rainfed upland rice ecosystems in Myanmar, 78% of 450 fields and 9% of 102 fields, respectively, were infested with M. graminicola, with the frequency of occurrence higher in delayed irrigation than in early irrigation (Win et al 2011). This survey confirms the role of Meloidogyne spp. and other plant parasitic nematodes such as Hirschmanniella spp. as important emerging pathogens of rice in Asia. Further, infested nurseries seem to play a major role in infecting otherwise nematode-free fields. There have been earlier reports of nematode infestation in seedbeds across India, Nepal, and Bangladesh. The problem becomes more severe when seedlings are transplanted into fields that cannot be continuously flooded due to a lack of water as the condition further promotes nematode growth. Over the last few years, this pattern of nematode infestation of seedbeds being transferred to rice fields seems to be increasing. The development of aerobic rice and the resolution of the problem of soil sickness therefore become increasingly important as lowland rice areas continuously experience suboptimal water control.

427

Yield failure in aerobic rice fields involving abiotic and biotic factors The observed soil sickness in various aerobic rice fields may be unique to the particular site and cultivar used in the studies, as is the case in a sandy-soil aerobic rice field in Tarlac, Philippines. Aside from experiments done at the IRRI experiment station, aerobic rice studies were also simultaneously conducted in a farmer’s field in Dapdap, Tarlac, from 2004 to 2007. Rice crops at this particular study site showed unexpected sudden yield failure, which is very different from the gradual yield decline generally characterizing soil sickness in continuously cropped aerobic rice in clay loam soils elsewhere. The first two years (2004-05) of field experiments consisted of water × N trials that gave very low yield despite large amounts of N fertilizer and adequate irrigation applied. Initial investigations were conducted to determine the possible causes of yield failure. Kreye et al (2009a) suggested, based on their analysis of field observations and using simulation modeling, that RKN and micronutrients were possible major constraints. Because this field experiment was not initially intended to identify the causes of yield failure, a follow-up experiment was conducted precisely to address this concern. In the following two dry seasons (2006-07), experiments were conducted in the same field to identify the major causes of such yield failure (Kreye et al 2009b). It was concluded that the interaction of RKN and micronutrient deficiencies with increasing soil pH led to yield failure. To further investigate the causes of yield failure, three pot experiments using sandy soil from the study site in Tarlac were conducted under controlled conditions at IRRI in 2007 (Kreye at al 2009c). The results of these experiments suggested that symptoms of soil sickness were partly caused by the use of shallow-tube-well irrigation water, which in turn increased soil pH and probably induced micronutrient deficiencies that impaired plant growth and induced chlorosis. The application of ammonium sulfate reduced soil pH, thus increasing the availability of micronutrients, leading to improved plant growth. Ammonium sulfate was better in reducing soil pH and consequently improving plant uptake of micronutrients and plant growth in general than urea application as observed in previous studies on alleviating soil sickness at IRRI experiment stations. Apparently, urea lacked an acidifying effect in the sandy soil type found in Tarlac, whereas ammonium sulfate was able to acidify the soil alkalinized by poor irrigation water. It was further suggested that, during the seedling stage and in areas affected more severely by poor irrigation water due to impeded drainage, RKN and other soil microorganisms affecting root health were most likely secondary factors that may have aggravated the abiotic component at later stages of growth. The ameliorating effect of ammonium sulfate associated with soil acidification supports the findings of Xiang et al (2009) as previously described.

Although not actively investigated in the three studies mentioned, the sandy-soil type in Tarlac may also play a major role in the development of soil sickness, especially in relation to the low activity of antagonistic soil microflora, the high prevalence of root-infecting nematodes, low water-holding capacity, low cation exchange capacity, and high leaching rates, which are characteristic of sandy soil types (Wolf 1999). All these effects may act in concert to intensify the effects of single factors, leading to synergistic interactions that aggravate the soil sickness syndrome as a whole.

In summary, all three papers by Kreye et al (2009a,b,c) suggested that, at the sandy-soil site in Tarlac, plant health was compromised due to a combination of factors including, but not limited to, increased soil pH and nutrient deficiencies possibly

428

predisposing the plants to RKN infection that in turn aggravated the effect of the abiotic stressors. This case of soil sickness illustrates how severe the effect of multiple factors acting together can be. The complexity of the soil environment together with the multiple causes and possible interactions of confounding factors make soil sickness a difficult malady to solve. No single solution can possibly affect all factors but perhaps the elimination of the primary factor(s) may offer a better prospect of control. Therefore, the need to elucidate the early events leading to soil sickness is of high importance. Conclusions Soil sickness is not unique to continuous aerobic rice monocropping. It has been observed time and again in various crops that have been grown on the same land for several consecutive seasons. Although the soil sickness syndrome in different crops exhibits similar characteristics, the causes of soil sickness are site- and crop-specific, as is the case for aerobic rice. Studies focusing on the identification of the primary predisposing factors are important as these factors set the stage for the development of soil sickness. At present, however, available information points to the roles played by soil nutrients, specific soil characteristics (i.e., soil pH, soil type), irrigation water quality, root-knot nematodes, and root-infecting pathogens such as Pythium spp., as well as the interactions between these factors in the establishment of soil sickness.

Soil nitrogen, which is subject to many transformations in and losses from the soil, seems to be an important growth-limiting factor together with micronutrients. Its availability for plant uptake, and the plant’s ability to assimilate it, is affected by allelochemicals, root-infecting organisms, soil pH, and soil type. Its addition to field crops is therefore a good option for alleviating soil sickness symptoms as investigated in several studies already described. However, it is also most prone to over-application that may lead to other problems such as pollution, toxicity, and increased susceptibility of the plant to other pests and diseases.

The preponderance of soil-borne, root-infecting pathogens found in sick soil points to the loss of the natural biocontrol mechanism of the soil microbial community. The affected plants may likewise have been predisposed to the effects of biotic agents as their response to pathogens is modulated by the environment. In addition to host-plant resistance, which continues to be the backbone of disease management in the tropics, alternative practices that increase soil fertility and biodiversity and increase plant vigor such as crop rotations with legumes and application of organic amendments may help overcome problems of soil sickness due to nutrient deficiencies, pathogen buildup, reduced plant immunity, as well as auto-toxicity to allelochemicals. A more direct approach to preventing the development of soil sickness is simply to avoid successive aerobic rice cropping in order to circumvent buildup of any predisposing factors, be they chemical or biological (Pinheiro et al 2006). Crop rotations, fallowing, and planting of cover crops (green manure) are examples of beneficial cropping schemes. These ecologically based agronomic practices help retain the capacity of soil to function as a vital living system to sustain biological productivity, and promote environmental quality and plant health (Doran and Zeiss 2000). The soil must be kept robust enough to subsidize and support ecosystem functions. In the end, all these illustrate the importance of adequate management of soil health in agricultural sustainability, for indeed healthy soils beget healthy crops and continue to do so long into the future.

429

Acknowledgments This work was funded by the International Rice Research Institute (IRRI)–Asian Development Bank (ADB) project titled “Developing and Disseminating Water-Saving Rice Technologies in South Asia.” References Arnell NW. 1999. Climate change and global water resources. Global Environ. Change

9:S31-S41. Asia Society and International Rice Research Institute Task Force. 2010. Never an

empty bowl: sustaining food security in Asia. Asia Society. 71 p. Atlin GN, Lafitte HR, Tao D, Laza M, Amante M, Courtois B. 2006. Developing rice

cultivars for high-fertility upland systems in the Asian tropics. Field Crops Res. 97:43-52.

Banaay CGB. 2011. Rhizosphere microbial community dynamics and the potential impact of organic matter amendment and Trichoderma ghanense on the control of diseases associated with soil sickness in a continuous aerobic rice monocropping system. PhD thesis. University of the Philippines Los Baños, Laguna, Philippines.

Banaay CGB, Fernandez L, Pinili MS, Das K, De Waele D, Kumar A, Dule Z, Vera Cruz CM 2010. Evaluation of response of selected upland rice varieties to simultaneous inoculation with Pythium arrhenomanes and Meloidogyne graminicola. Poster paper presented at the 6th International Rice Genetics Symposium, 16-19 November 2009. Manila, Philippines.

Banaay CGB, Vera Cruz CM, Cuevas VC. 2013. Effect of organic matter amendment on the rhizosphere microbial community and root-infecting pathogens of aerobic rice variety Apo. Phil Science Lett. 6:107-118.

Belder P, Bouman BAM, Spiertz JHJ, Peng, S, Castañeda AR, Visperas RM. 2005. Crop performance, nitrogen, and water use in flooded and aerobic rice. Plant Soil 273:167-182.

Bernard SM, Habash DZ. 2009. The importance of cytosolic glutamine sythetase in nitrogen assimilation and recycling. Lawrence Berkeley National Laboratory. p 14. http://escholarship.org/uc/item/4h95w3sd.

Bouman BAM. 2001. Water-efficient management strategies in rice production. Int. Rice Res. Notes. 26(2):17-22.

Bouman BAM, Peng S, Castañeda AR, Visperas RM. 2005. Yield and water use of irrigated tropical aerobic rice systems. Agric. Water Manage. 74:87-105.

Bouman BAM, Humphreys E, Tuong TP, Barker R. 2007. Rice and water. Adv. Agron. 92:187-237.

Bridge J, Plowright RA, Peng D. 2005. Nematode parasites of rice. In: Luc M, Sikora RA, Bridge J, editors. Plant parasitic nematodes in subtropical and tropical agriculture. Second ed. Oxfordshire (UK): CAB International. p 87-130.

Brown BJ, Marten GG. 1986. The ecology of traditional pest management in Southeast Asia. In: Marten GG, editor. Traditional agriculture in Southeast Asia: a human ecology perspective. Boulder, Col. (USA): Westview Press. p 241-272.

Buresh RJ, Haefele SM. 2010. Changes in paddy soils under transition to water-saving and diversified cropping systems. In: Gilkes RJ, Prakongkep N, editors. Soil

430

solutions for a changing world. Proceedings of 19th World Congress of Soil Science: Working Group 3.5, Brisbane, Australia, 1-6 August 2010. p 9-12.

Chien SH, Gearhart MM, Villagarcia S. 2011. Comparison of ammonium sulfate with other nitrogen and sulfur fertilizers in increasing crop production and minimizing environmental impact: a review. Soil Sci. 176:327-335.

Chou CH. 1998. Adaptive autointoxication mechanisms in rice. In: Olofsdotter MO, editor. Allelopathy in rice. Los Baños (Philippines): International Rice Research Institute. p 99-115.

Cren M, Hirel B. 1999. Glutamine synthetase in higher plants: regulation of gene and protein expression from the organ to the cell. Plant Cell Physiol. 40:1187-1193.

Doran JW, Zeiss MR. 2000. Soil health and sustainability: managing the biotic component of soil quality. Appl. Soil Ecol. 15:3-11.

Duan Y, Yin X, Zhang Y, Shen Q. 2007. Mechanisms of enhanced rice growth and nitrogen uptake by nitrate. Pedosphere 17(6):697-705.

Dzubenko N. 2006. Soil sickness: field crops. In: Grodzinsky AM, editor. Allelopathy in soil sickness. Jodhpur (India): Scientific Publishers. p 11-49.

Fageria NK, Baligar VC. 2003. Upland rice and allelopathy. Commun. Soil Sci. Plant Anal. 34:1311-1329.

George T, Magbanua R, Garrity DP, Tubaña BS, Quiton J. 2002. Rapid yield loss of rice cropped successively in aerobic soil. Agron. J. 94:981-989.

Grodzinsky AM. 2006. Soil sickness: history. In: Grodzinsky AM, editor. Allelopathy in soil sickness. Jodhpur (India): Scientific Publishers. p 1-10.

Haden VR, Xiang J, Peng S, Ketterings QM, Hobbs P, Duxbury JM. 2011. Ammonia toxicity in aerobic rice: use of soil properties to predict ammonia volatilization following urea application and the adverse effects on germination. Eur. J. Soil Sci. doi: 10.1111/j.1365-2389.2010.01346.x.

Hayakawa T, Kamachi K, Oikawa M, Ojima K, Yamaya Y. 1990. Response of glutamine synthetase and glutamate synthase isoforms to nitrogen sources in rice cultivars. Plant Cell Physiol. 31:1071-1077.

Huang HC, Chou CH. 2005. Impact of plant disease biocontrol and allelopathy on biodiversity and agricultural sustainability. Plant Pathol. Bull. 14:1-12.

Jeger MJ, Pautasso M. 2008. Comparative epidemiology of zoosporic plant pathogens. Eur. J. Plant Pathol. 122:111-126.

Jensen LB, Courtois B, Shen L, Li Z, Olofsdotter M, Mauleon RP. 2001. Locating genes controlling allelopathic effects against barnyard grass in upland rice. Agron. J. 93:21-26.

Kreye C, Bouman BAM, Castañeda AR, Lampayan RM, Faronilo JE, Lactaoen AT, Fernandez L. 2009a. Possible causes of yield failure in tropical aerobic rice. Field Crops Res. 111:197-206.

Kreye C, Bouman BAM, Reversat G, Fernandez L, Vera Cruz CM, Elazegui F, Faronilo JE, Llorca L 2009b. Biotic and abiotic causes of yield failure in tropical aerobic rice. Field Crops Res. 112:97-106.

Kreye C, Bouman BAM, Faronilo JE, Llorca L. 2009c. Causes for soil sickness affecting early plant growth in aerobic rice. Field Crops Res. 114:182-187.

Kronzucker HJ, Siddiqui MY, Glass ADM, Kirk GJD. 1999. Nitrate-ammonium synergism in rice: a subcellular flux analysis. Plant Physiol. 119:1041-1045.

431

Kronzucker HJ, Glass ADM, Siddiqui MY, Kirk GJD. 2000. Comparative kinetic analysis of ammonium and nitrate acquisition by tropical lowland rice: implications for rice cultivation in yield potential. New Phytol. 145(3):471-476.

Lin S, Li J, Sattelmacher B, Brück H. 2005. Response of lowland and aerobic rice to ammonium and nitrate supply during early growth stages. J. Plant Nutr. 28:1495-1510.

Morozov PA. 2006. Soil sickness: fruit trees. In: Grodzinsky AM, editor. Allelopathy in soil sickness. Jodhpur (India): Scientific Publishers. p 50-83.

Nie L, Peng S, Bouman BAM, Huang J, Cui K, Visperas RM, Park HK. 2007a. Solophos fertilizer improved rice plant growth in aerobic soil. J. Integrat. Field Sci. 4:65-70.

Nie L, Peng S, Bouman BAM, Huang J, Cui K, Visperas RM, Park HK. 2007b. Alleviating soil sickness caused by aerobic monocropping: responses of aerobic rice to soil oven-heating. Plant Soil 300:185-195.

Nie L, Peng S, Bouman BAM, Huang J, Cui K, Visperas RM, Xiang J. 2008. Alleviating soil sickness caused by aerobic monocropping: responses of aerobic rice to nutrient supply. Field Crops Res. 107:129-136.

Nie L, Peng S, Bouman BAM, Huang J, Cui K, Visperas RM, Xiang J. 2009a. Alleviating soil sickness caused by aerobic monocropping: responses of aerobic rice to various nitrogen sources. Soil Sci. Plant Nutr. 55:150-159.

Nie L, Xiang J, Peng S, Bouman BAM, Huang J, Cui K, Visperas RM. 2009b. Alleviating soil sickness caused by aerobic monocropping: responses of aerobic rice to fallow, flooding and crop rotation. J. Food Agric. Environ. 7:723-727.

Nishio M, Kusano S. 1975. Effect of root residues on the growth of upland rice. Soil Sci. Plant Nutr. 21:391-395.

Olofsdotter M. 2001. Rice: a step toward use of allelopathy. Agron. J. 93:3-8. Pal S, Datta PS, Rattan RK, Singh AK. 2008. Diagnosis and amelioration of iron

deficiency under aerobic rice. J. Plant Nutr. 31:919-940. Peng S, Bouman BAM, Visperas RM, Castañeda AR, Nie L, Park HK. 2006.

Comparison between aerobic and flooded rice in the tropics: agronomic performance in an eight-season experiment. Field Crops Res. 96:252-259.

Pinheiro BS, De Castro EM, Guimaraes CM. 2006. Sustainability and profitability of aerobic rice production in Brazil. Field Crops Res. 97:34-42.

Pinili MS, Banaay CGB, Elazegui F, Steelandt A, Das K, Kreye C, Kumar A, Bouman B, Hofte M, Vera Cruz CM. 2009. Population dynamics of Pythium sp. and Meloidogyne graminicola from an aerobic rice field planted to variety Apo. Poster paper presented at the 20th FCSSP Scientific Conference “Meeting the Challenges of Water and Energy Crisis,” 18-23 May 2009. Siliman University, Dumaguete City, Philippines.

Prasad R. 2011. Aerobic rice systems. Adv. Agron. 111:207-247. Prot JC, Matias DM. 1995. Effects of water regime on the distribution of Meloidogyne

graminicola and other root-parasitic nematodes in a rice field toposequence and pathogenicity of M. graminicola on rice cultivar UPLRi5. Nematology 41:219-228.

Prot JC, Rahman ML. 1994. Nematode ecology, economic importance and management in rice ecosystems in South and South-East Asia. In: Teng PS, Heong KL, Moody K, editors. Rice pest science and management. Manila (Philippines): International Rice Research Institute. p 129-144.

432

Qian X, Shen Q, Xu G, Wang J, Zhou M. 2004. Nitrogen form effects on yield and nitrogen uptake of rice crop grown in aerobic soil. J. Plant Nutr. 27(6):1061-1076.

Robertson GP, Groffman PM. 2007. Nitrogen transformation. In: Paul EA, editor. Soil microbiology, biochemistry, and ecology. New York (USA): Springer. p 341-364.

Savant NK, De Datta SK. 1982. Nitrogen transformation in wetland rice soils. Adv. Agron. 35:241-302.

Schoeneweiss DF. 1975. Predisposition, stress, and plant disease. Ann. Rev. Phytopathol. 13:193-211.

Sharif Zia M, Mahmood IA, Aslam M, Yasin M, Khan MA. 2001. Nitrogen dynamics under aerobic and anaerobic soil conditions. Int. J. Agric. Biol. 3:458-460.

Soriano I, Reversat G. 2003. Management of Meloidogyne graminicola and yield of upland rice in South-Luzon, Philippines. Nematology 5(6):879-884.

Sumner DR. 1982. Crop rotation and plant productivity. In: Rechcigl MJ, editor. CRC handbook of agricultural productivity. Vol. I. Plant productivity. Boca Raton, Fla. (USA): CRC Press, Inc. p 273-313.

Tuong TP, Bouman BAM. 2003. Rice production in water-scarce environments. In: Kijne JW, Barker R, Molden D, editors. Water productivity in agriculture: limits and opportunities for improvement. Wallingford (UK): CAB International. p 53-67.

Van Buyten E, Banaay CGB, Vera Cruz CM, Höfte M. 2013. Identity and variability of the Pythium species associated with yield decline in aerobic rice cultivation in the Philippines. Plant Pathol. 62:139-153.

Ventura W, Watanabe I. 1978. Growth inhibition due to continuous cropping of dryland rice and other crops. Soil Sci. Plant Nutr. 24:375-389.

Ventura W, Watanabe I, Castillo MB, De la Cruz A. 1981. Involvement of nematodes in the soil sickness of a dryland rice-based cropping system. Soil Sci. Plant Nutr. 27:305-315.

Vorosmarty CJ, Green P, Salisbury J, Lammers RB. 2000. Global water resources: vulnerability from climate change and population growth. Science 289:284-288.

Wang H, Bouman BAM, Zhao D, Wang C, Moya PF. 2002. Aerobic rice in Northern China—opportunities and challenges. In: Bouman BAM, Hengsdijk H, Hardy B, Bindraban PS, Tuong TP, Ladha JK, editors. Water-wise rice production. Proceedings of the International Workshop on Water-wise Rice Production, 8-11 April. Los Baños (Philippines): International Rice Research Institute. p 143-154.

Wassman R, Jagadish SVK, Heuer S, Ismail A, Redoña E, Serraj R, Singh RK, Howell G, Pathak H, Sumfleth K. 2009. Climate change affecting rice production: the physiological and agronomic basis for possible adaptation strategies. Adv. Agron. 101:59-122.

Win PP, Kyi PP, De Waele D. 2011. Effect of agro-ecosystem on the occurrence of the rice root-knot nematode Meloidogyne graminicola on rice in Myanmar. Austr. Plant Pathol. 40(2):187-196.

Wolf B. 1999. The fertile triangle: the interrelationship of air, water, and nutrients in maximizing soil productivity. Binghampton, N.Y. (USA): The Haworth Press, Inc. 463 p.

Xiang J, Haden VR, Peng S, Bouman BAM, Visperas RM, Nie L, Huang J,Cui K. 2009. Improvement in nitrogen availability, nitrogen uptake and growth of aerobic rice following soil acidification. Soil Sci. Plant Nutr. 55:705-714.

433

Xie G, Yu J, Wang H, Bouman BAM. 2008. Progress and yield bottleneck of aerobic rice in the North China Plain: a case study of varieties Handao 297 and Handao 502. Agric. Sci. China 7(6):641-646.

Yang X, Bouman BAM, Wang H, Wang Z, Zhao J, Chen B. 2005. Performance of temperate aerobic rice under different water regimes in North China. Agric. Water Manage. 74:107-122.

Zhao DL, Atlin GN, Amante M, Sta Cruz MT, Kumar A. 2010. Developing aerobic rice cultivars for water-short irrigated and drought-prone rainfed areas in the tropics. Crop Sci. 50:2268-2276.

Zuo S, Zhi J, Shao H, Zhao G. 2010. Allelopathy regulates wheat genotypes at the enhancement stage by soil water and prohydrojasmon (PDJ). Afr. J. Biotechnol. 9:5430-5440.

Notes Authors’ addresses: C.G.B. Banaay, Institute of Biological Sciences, College of Arts and Sciences, University of the Philippines, and International Rice Research Institute (IRRI), Los Baños, Laguna, Philippines; C. Kreye, University of Bonn, Karlrobert-Kreiten-Strasse 13, 53115 Bonn, Germany; M. Hofte, Laboratory of Phytopathology, Department of Crop Protection, Faculty of Bioscience Engineering, Ghent University, Coupure Links 653, Gent, Belgium; A. Kumar, IRRI; D. De Waele, IRRI, and Laboratory of Tropical Crop Improvement, Faculty of Bioscience Engineering, University of Leuven (K.U. Leuven), Belgium; V.C. Cuevas, Institute of Biological Sciences, College of Arts and Sciences, University of the Philippines Los Baños, Laguna, Philippines; C.M. Vera Cruz, IRRI; and B.A.M. Bouman, IRRI.