ARTICLE IN PRESS

Global Environmental Change 18 (2008) 564–574

Contents lists available at ScienceDirect

Global Environmental Change

0959-37

doi:10.1

� Corr

E-m1 Te

journal homepage: www.elsevier.com/locate/gloenvcha

Vulnerability and adaptation to climate variability and water stress inUttarakhand State, India

Ulka Kelkar a,�, Kapil Kumar Narula b,1, Ved Prakash Sharma b,1, Usha Chandna b,1

a TERI (The Energy and Resources Institute), 4th Main, 2nd Cross, Domlur 2nd Stage, Bangalore 560071, Indiab TERI (The Energy and Resources Institute), Darbari Seth Block, India Habitat Centre, Lodhi Road, New Delhi 110003, India

a r t i c l e i n f o

Article history:

Received 29 May 2007

Received in revised form

1 September 2008

Accepted 3 September 2008

Keywords:

Climate change

Water stress

Agriculture

80/$ - see front matter & 2008 Elsevier Ltd. A

016/j.gloenvcha.2008.09.003

esponding author. Tel.: +9180 2535 6590; fax

ail address: ulkak@teri.res.in (U. Kelkar).

l.: +91112468 2100; fax: +9111 2468 2144.

a b s t r a c t

This paper presents a participatory approach to investigate vulnerability and adaptive capacity to

climate variability and water stress in the Lakhwar watershed in Uttarakhand State, India. Highly water

stressed microwatersheds were identified by modelling surface runoff, soil moisture development,

lateral runoff, and groundwater recharge. The modelling results were shared with communities in two

villages, and timeline exercises were carried out to allow them to trace past developments that have

impacted their lives and livelihoods, and stimulate discussion about future changes and possible

adaptation interventions.

& 2008 Elsevier Ltd. All rights reserved.

1. Introduction

The Fourth Assessment Report of the IntergovernmentalPanel on Climate Change (IPCC) points out that freshwateravailability in Asia is projected to decrease due to climatechange. By the middle of the 21st century, annual average riverrunoff and water availability are projected to decrease by 10–30%relative to 1900–70 over some dry regions at mid-latitudes and inthe dry tropics. Further, reduced water availability is projected forregions supplied by melt water from glaciers and snow (IPCC,2007).

India receives an average annual precipitation of about 4000billion cubic metres (BCM). Of this, utilizable surface water andgroundwater resources are estimated to be only 690 and 432 BCM,respectively (CWC, 2005). As a result of wide fluctuations in theavailability of water, spatially and temporally, water shortage isvirtually an annual feature in several parts of the country.Moreover, acceleration in the rate of consumption due to anincreasing population and changing lifestyles is a cause forconcern for effective sustainable management and utilization ofthis resource. Per capita annual water availability in India hassteadily declined from 1820 m3 in 2001 to 1703.6 m3 in 2005,coming close to the water stress threshold of 1700 m3 (CWC,2005). Agriculture accounts for more than 80% of water with-

ll rights reserved.

: +9180 2535 6589.

drawals in the country, using both surface and groundwaterresources. Water demand for irrigation is projected to rise from541 BCM in 2000 to 910 BCM in 2025, and to 1072 BCM in 2050(CWC, 2005).

Specifically for the Lakhwar sub-basin, part of the UpperYamuna sub-basin in Uttarakhand state, Narula and Bhadwal(2003) found that about 1500 km2 of the 4000 km2 sub-basinreceives an annual runoff of less than 1250 mm, and as a result ishighly sensitive to increased water stress due to climate change.The study projected a decrease of 20–30% in total flows due toclimate change alone by 2041–60 relative to 1961–90, withmonsoon rainfall likely to become less intense and more sporadic.The potential impacts of such changes could include thefollowing.

�

Reduced groundwater availability—Under the changed climatescenario, since the total water availability will go down byabout 30% relative to the reference period of 1961–90, thegroundwater recharge will also be reduced, translating into afall in groundwater tables. This will also lead to an increase inthe extraction costs because higher capacity equipment has tobe installed to maintain yields. � Reduced surface water availability—Decrease in total runoff

levels would reduce the availability of drinking water forhumans and livestock. This would specifically have seasonalpatterns worth studying. In conjunction with reduced ground-water availability, this could lead to increased conflicts overwater, and would certainly increase the burden of watercollection on women.

ARTICLE IN PRESS

U. Kelkar et al. / Global Environmental Change 18 (2008) 564–574 565

�

Declining crop yields—Since the changes in the rainfall patternwill have a direct impact on the groundwater rechargevolumes, intensity of irrigation in the region will be affected.Crops like sugarcane, rice, and wheat that are water-intensiveand are grown extensively in the region will be most severelyaffected. Restrictions in growth of sugarcane and rice can occurdue to changes in mean temperatures and rainfall patterns(Venkataraman and Krishnan, 1992). The start of the monsoonseason is an important factor in rice production. In some cases,the sowing and harvesting time of these crops may also have tobe shifted as temperatures and humidity levels vary, and anychange in these conditions may affect the growth of thesecrops and reduce yields. � Reduced water quality—Areas located in the north and north-

east parts of the sub-basin are more vulnerable to climatechange impacts since total water availability is alreadyless compared to other regions of the sub-basin. Such areasare also susceptible to pollution as economic activities growand water flows decline, which would have significant healthimpacts.

In light of the predicted impacts of climate change outlinedabove, this study aims to assess the vulnerability and adaptivecapacity of households engaged in agriculture in Lakhwar sub-basin to climate variability and water stress. Recognizing thatvulnerability needs to be understood at the local level, and thathouseholds engaged in agriculture employ a variety of measuresin response to changing stresses and incentives, the studycomplements watershed modelling with a participatory approachto gain insights through mutual learning and exchange with theaffected communities. It attempts to address the followingspecific research questions:

1.

What is the capacity of households in the study region to copewith current climatic variability and water stress?

2.

Are their responses only temporary coping measures, or wouldthese responses help households adapt in the long run?

3.

2 North India has two agricultural seasons. Rabi is the winter season and the

rabi crop is harvested in April. Kharif coincides with the monsoon or the rainy

season. Wheat is an important rabi crop while paddy is an important kharif crop.3 Polyandry is the practice of having more than one husband at one time.

Majumdar (1962) termed the practice, observed in this region, of fraternal

polyandry with multiple wives as ‘‘polygynandry’’.

What are the possible scenarios of interventions (at differentlevels—policy, institutional, technological, community, indivi-dual) that can help build adaptive capacity?

2. Case study area

The Yamuna river flows through the Punjab-Kumaon Hima-layas from Shimla in the north-west to Mussoorie in the south-east. It rises from the Yamunotri springs in the Himalayas. Afterflowing in the south-westerly direction for about 120 km, theYamuna is joined by its principal tributary, the Tons, nearDakpathar. The Lakhwar sub-basin constitutes the Himayalanreach of the Upper Yamuna basin and stretches to Dehradundistrict in Uttarakhand State. This sub-basin constitutes about4000 km2, out of the total catchment area of 12,000 km2 (Fig. 1). Inthe upper reach, the Yamuna has a steep bed slope of 1/16 in thefirst 30 km and flattens out with a bed slope of 1/500 at Tajewala.Since in the upper catchment, the river courses are well definedand confined between high banks, they present no flood problem.

Rainfall spells in the Yamuna basin are generally associatedwith monsoon or late-monsoon depressions either from the Bay ofBengal or the Arabian Sea. Normally, monsoon sets in over theupper catchment by about the end of the third week of June andwithdraws by about the middle or third week of September. Late-monsoon depressions may form till the end of September or evenup to the middle of October. The Yamuna catchment receivesaround 120 cm of rainfall throughout the year, of which 75% isreceived during the monsoon months (CWC, 1998).

Dehradun is the state capital of the north Indian state ofUttarakhand (earlier called Uttaranchal). Located at an altitude of640 m above sea level, the district is spread over 3088 km2,nestled among mountain ranges. About 51% of the district iscovered by forests (FSI, 2008) with the mountain slopesdominated by sal (Shorea robusta). Agriculture2 is practiced inthe valley, along with agroforestry, orchard cultivation, and teagardening.

Dehradun district has two sub-divisions—Dehradun andChakrata. The lower part of Chakrata is called Jaunsar and theupper part (at higher elevation) is called Bawar. This is difficultterrain, poorly connected by transportation facilities, and gen-erally neglected when Uttarakhand was part of the large UttarPradesh state. The inhabitants of Jaunsar claim descent from thePandavas of the Mahabharata, and traditionally practicedpolyandry.3

Cultivation takes place mainly in narrow patches of terracedfields cut into the hill slopes. About half of all landholdings areless than 0.5 ha in size, and 70% are less than a hectare. Tenantfarming and sharecropping are not common. As pointed out bySati (2005), geography and culture (i.e. the way of life in the hills)have created a relatively equitable, if impoverished, land distribu-tion in Uttarakhand.

The total water requirement for Uttarakhand state (includinghuman consumption, animal consumption, agriculture, andindustry) is estimated as only 3% of the annual precipitationreceived. However, rainfall is available for only 100 days, andflows out swiftly from the steep slopes constituting the major partof the state. The cultivable command area (excluding forestland,populated land, and non-irrigable land) of the state is 1.14 millionha, of which 0.55 million ha remains unirrigated and lies primarilyin the hilly areas of the state. In the hill districts, the irrigated areais merely 14%, compared with 46% in the foothills and plains(Government of Uttaranchal, 2004). The State’s draft water policy(Government of Uttaranchal, 2004, p. 1) notes that ‘‘It is a paradoxthat the local people of the state and their lands are facingshortage of water for domestic consumptive uses in the remainingperiod of the year.’’

Uttarakhand (and its neighbouring state Himachal Pradesh)have had a tradition of water harvesting (Table 1). Water forhousehold use was obtained from springs, mountain streams, orman-made rainwater harvesting structures. Open ponds and tanksprovided water for animals, irrigation, and washing. For humanconsumption, water was tapped from underground seepages (inbaoris/naulas) or springs (dharas). Terraced fields were irrigatedby diverting water (using boulders and branches) from nearbymountain streams through small gravity flow channels known asguhls. Typically a farmer floods his field and then removes a stoneplug at the outside edge of the field so that water can flow to thenext terrace below. Some channels also provide hydropower forgharats (water mills). All these structures were usually commonproperty resources, which were largely owned, used, and main-tained by local communities. However, an increasing number ofguhls have been taken over by state government agencies, andfallen into disrepair due to lack of a sense of ownership (PSI,2003). At the national level, there has been a significant shift inthinking with water being increasingly viewed as a commodity,rather than a natural resource. This is reflected in the National

ARTICLE IN PRESS

Digitized streamsSubbasinsStreams

# Linking stream added Outlet#

####

##

###

#

##

# #

# #

#

#

###

2

8

6

3

7

1

5

4

9

Koti

Damta

Simla

Solan

Dadahu

Lakhwar

Yamunotri

Dehradun

Chakrata

Mussoorie

YashwantNagar

UTTARKASHI

SIRMAUR DEHRADUN

SOLAN

TEHRI GARHWAL

SHIMLA

0

####

##

###

#

##

# #

# #

#

#

###

####

##

##

#

##

# #

# #

#

#

###

TajewalaYAMUNANAGAR

Outlets

20 40 Kilometres

Fig. 1. Location of Lakhwar watershed within Upper Yamuna basin.

U. Kelkar et al. / Global Environmental Change 18 (2008) 564–574566

Water Policy (2002), which formalizes the concept of the rights ofwater users, and allows trading in entitlements.

The mid-Himalayan ranges are dry, with small streams fed bywinter snow. Agriculture in Jaunsar-Bawar is characterized bynatural irrigation i.e. no lift irrigation, tubewells/borewells, butonly guhls. Sati (2005) cites a 1996 study which found that inabout half of Uttarakhand’s villages, springs had either ceased toyield water, or did so only during the rainy season, calling this the‘‘too little—too much syndrome’’. There has been a decrease inspring discharge ranging from 25% to 75%, and this has resulted ina considerable decrease in water flow; estimated to be around30–40% in the last decade or two.

Against this background, the draft water policy for the statenotes that production of food grains has increased from around0.5 million tonnes in the 1950s to about 1.79 million tonnes in

1999/2000, but stipulates that this will have to be raised toaround 2.5 million tonnes by the year 2025.

3. Framework

This study adopts a place-specific approach to understandvulnerability and adaptive capacity through mutual learning andexchange with the affected communities. While choosing aspecific watershed as the case study area, the stresses acting onthe exposure unit were studied as part of a larger context, withexternal forces acting as both constraints to, and opportunities for,coping responses.

The working definition of vulnerability adopted for this studyis that defined by Turner et al. (2003, p. 8074) as ‘‘the degree to

ARTICLE IN PRESS

Table 1Traditional water harvesting structures in central western Himalayas

Structure Use State

Chaal/khal/

chuptyaula

Animal consumption Uttarakhand

Naula/baori Domestic water use Uttarakhand

Dhara Drinking water, occasionally irrigation from

large dharas

Uttarakhand

Guhl Irrigation and operating gharats (watermills) Uttarakhand

Hauzi Irrigation Uttarakhand

Gharat Milling Uttarakhand

Chappri/talaai/

talaab

Water for livestock and irrigation Himachal

Pradesh

Baori/khatri Domestic water use Himachal

Pradesh

Naun Bathing and washing clothes Himachal

Pradesh

Chharedu/panihar/

nahun

Bathing and drinking water Himachal

Pradesh

Kuhl Irrigation and operating gharats Himachal

Pradesh

Gharat Milling Himachal

Pradesh

Source: PSI (2003).

U. Kelkar et al. / Global Environmental Change 18 (2008) 564–574 567

which a system, subsystem, or system component is likely toexperience harm due to exposure to a hazard, either a perturba-tion or stress’’. Various factors shape the differences in vulner-ability of individuals or groups: entitlements, personalheterogeneity, variations in social obligations, environmentallocation, livelihood diversification strategies, support networks,empowerment or power relations, access to knowledge, informa-tion, and technology (Noronha, 2003). A combination of factorsmay increase vulnerability or enhance resilience to stresses(i.e. the capacity to cope or respond to stress in different ways).Within the context of climate studies, the most vulnerable areconsidered to be those who are most exposed to perturbations,who possess a limited capacity for adaptation, and who are leastresilient to recovery (Bohle et al., 1994).

4. Methods

Schroter et al. (2005) propose a set of five criteria for globalchange vulnerability assessments to inform adaptation decisionmaking by stakeholders: they should have a knowledge base fromvarious disciplines and stakeholder participation, be place based,consider multiple interacting stresses, examine differential adap-tive capacity, and be prospective as well as historical. On the basisof these criteria, they suggest eight methodological steps. Fig. 2outlines the various steps taken and methods used in this studycorrespond to the Schroter et al. (2005) framework.

The study combined watershed modelling with a participatoryapproach to investigate vulnerability and adaptive capacity toclimate variability and water stress in the Lakhwar watershed.Water balance modelling for Lakhwar watershed was carried outusing SWAT (Neitsch et al., 2002) and MODFLOW (McDonald andHarbaugh, 1988) models. Areas that would be most affected due tochanges in flows and water stresses were identified, and twovillages were selected for community-level interactions. Thispaper does not go into the details of the watershed modellingcomponent, which is summarized below, but focuses more on theparticipatory assessment.

4.1. Water balance modelling for Lakhwar watershed

Two widely used models viz. SWAT and MODFLOW were usedto evaluate water resources for the Lakhwar watershed. While

SWAT models the land and water phase of the hydrological cyclethat includes surface runoff, soil moisture development andlateral runoff, and groundwater recharge or baseflows, MODFLOWmodels the groundwater movement.

The input database for the model analysis was built fromclimatic, pedological, geological, topographical, hydrological, anddemographic data listed in Table 2. The modelling, analysis andcartographic representation were carried out in the Arc/Infoand ARCVIEW geographic information system (GIS). The modelresults were validated by comparing the calculated runoff valueswith actual flows for the sub-basins measured at runoff gaugesmaintained by the Central Water Commission, Ministry of WaterResources, Government of India.

The hydrological model results for the Lakhwar watershedshow that the average daily surface runoff (water yield) is veryhigh. The peak flows are in the range of 600 to 41000 cubicmetres per second (cumecs). The average daily flows are in therange of 80–120 cumecs. Surface runoff was found to be the mostimportant component of the overall water balance for thiswatershed (Table 3). This implies that useful interventions forthis watershed could include:

�

capturing surface runoff through catchment level rainwaterharvesting and local check dams, � enhancing gravity schemes for water supply to villages, and � desilting and management of local ponds at the village level.

4.2. Site selection

Based on the water balance modelling, the five microwater-sheds in the Lakhwar watershed were ranked in terms of relativewater stress. The southernmost part of the watershed wasidentified as the most water stressed with total annual wateravailability estimated to be 995 mm, followed by the microwa-tershed immediately to its north where total annual wateravailability was estimated to be 1030 mm.

The modelling results were shared with a local NGO partner,the Society for Motivational Training and Action (SMTA), toground-truth the analysis and to identify one village each in thetwo most water stressed areas. We wanted to select villages thatwere typical in terms of the economic importance of agricultureand the type of cultivation practiced in the region. Whilepopulation density in the hills is often low, we attempted toselect villages with at least 30–40 households to get a reasonablesample size. Lakhwar village, identified in the most water stressedmicrowatershed, is characterized by purely rainfed farming, andChhotau village in the second most water stressed microwa-tershed has a mix of irrigated and rainfed farming.

4.3. Participatory methods

Participatory learning and action research has evolved as anapproach that utilizes the knowledge of poor, rural, naturalresource users-recognizing them as ‘‘crucial agents of change anddevelopment, and not merely targets of technological advice’’(Tyler, 2006, p. 20). This ‘‘lay knowledge’’ as opposed to ‘‘expertknowledge’’ (Kasemir et al., 2003) is being increasingly consideredrelevant to environmental decision making. The two publicationscited above demonstrate with respect to community-basednatural resource management and global climate change respec-tively, that understanding the complexity of social and ecologicalsystems and knowing current perceptions of problems and valuesis essential to effect sustained changes. Further, this approachattempts to conduct field-based social science research in a less

ARTICLE IN PRESS

1. Define study area withstakeholders

2. Get to know place over time

3. Hypothesize who isvulnerable to what

4. Develop a causal model ofvulnerability

5. Find indicators for theelements of vulnerability

6. Operationalise model ofvulnerability

7. Project future vulnerability

8. Communicate vulnerabilitycreatively

Interviews with district officials and NGOs working in theregion; collection and analysis of secondary data andliterature

Water balance modelling carried out to identify highlywater-stressed areasVulnerability profiling of households to understandpatterns of livelihood strategies, socioeconomic andenvironmental stresses, and risk perceptionsFocus on smallholder farmers and water stress in contextof economic pressures

Analysis of questionnaire data combined with results fromwater balance modelling to assess water stress andvulnerability of the community

Timeline exercises with vulnerable community to discusspast developments and identify future scenarios of waterstress; sharing of modelling results and brainstorming onadaptation options

Step in Schroter et al (2005)methodological framework Methods used in study

Time spent in field with community members and NGOworkers; oral history of village elders

Semi-structured interviews to elicit information on impacts,coping capacity, and factors affecting vulnerability; timebudgeting exercises with women

Fig. 2. Outline of study methodology.

Table 2Input data sets of the water-balance models for the river basin

Subject area Data basis Source

Basic data Boundaries of the river basin,

administrative boundaries,

flowing waters

Survey of India (SoI),

Central Pollution

Control Board

Climatic data Daily precipitation, daily

temperatures (mean, minimum

and maximum), mean annual

precipitation in the hydrological

winter months, mean annual

precipitation in the hydrological

summer months, mean annual

potential evaporation

Indian Meteorological

Department

Soil-physical data Soil characteristics (% silt,

sand and clay), effective root

penetration depth, useful

field capacity, capillary elevation,

influence on groundwater

or perched aquifer

National Bureau of

Soil Survey and Land

Use Planning

Land use data Ground cover SoI, Satellite

imageries, State

Agricultural Board

Hydrogeological data Groundwater-bearing

lithologic units

Geological Survey of

India (GSI)

Topographical data Mean slope; mean slope

exposure

SoI

Gauge data Mean flows and mean

minimum flows (monthly, annual

for 30 years)

Central Water

Commission, Ministry

of Water Resources,

Government of India

U. Kelkar et al. / Global Environmental Change 18 (2008) 564–574568

extractive manner, i.e. instead of taking information fromrespondents, researchers should facilitate the identification,understanding, and addressing of issues by the local community.

The authors drew inspiration from the methods described inKasemir et al. (2003) where climate change integrated assessment

model outputs were shared with citizen focus groups. Evans et al.(2006) also describe how natural resource dependent commu-nities can discuss today’s concerns as a starting point for arrivingat a vision for the future. ‘‘The way that the public understandsand defines the issues then becomes a complementary input tothe scientific assessment and ultimately the policy makingprocess’’ (Kasemir et al., 2003, p. 7).

Group discussions were held in each village during May–July2005 to elicit community perceptions about climatic changes andlearn about factors impacting agricultural livelihoods over time.Stratified random sampling was undertaken (details in Section5.1) and semi-structured interviews were conducted with differ-ent types of households to elicit information on their agriculturalpractices and non-agricultural responses in the face of waterscarcity. Key person interviews were carried out with villageelders and leaders. Time budgeting exercises were carried outwith women. The community in each of the two villages wasbrought together for group discussions and timeline exercises,separately with men and women. Water availability mapsdeveloped in the watershed modelling exercise were shared withthe community to stimulate brainstorming on a possible pool ofoptions that would enhance the adaptive capacity of vulnerablehouseholds in the long run. A timeline was developed by thecommunity of key developments in the village, including thechanging water stress situation, extending into future scenariosfor water stress and possible interventions.

5. A tale of two villages: preliminary observationsfrom case studies

Amartya Sen’s entitlements and capabilities approach givesemphasis to differential access, or ‘‘entitlement’’ to resources, as adeterminant of household/individual vulnerability (Sen, 1981,1999). To explore factors influencing vulnerability of agricultural

ARTICLE IN PRESS

Table 3Water balance of Lakhwar watershed

Period of simulation 5 years As % of precipitation

Average precipitation (mm) 1848

Actual evapotranspiration loss (mm) 663 36%

Recharge (mm) 144 8%

Surface runoff (mm) 1041 56%

Source: TERI analysis (2005).

U. Kelkar et al. / Global Environmental Change 18 (2008) 564–574 569

households over time, and coping measures employed by differenttypes of households, case studies were carried out in Lakhwar andChhotau villages. This section reports insights gained throughgroup discussions and individual interviews in the two villages.

5.1. Lakhwar and Chhotau: socio-economic profiles

Lakhwar is part of a dense cluster of villages comprisingLakhwar, Dhanpau, Bisnau, Mistau, Jakhnau, and Sawda, about60 km from Dehradun. The total cultivated and uncultivated areaof Lakhwar village specifically is 72 ha. The patwari (local revenueofficial) had a record of 47 households (5 with medium land-holdings, 15 with small landholdings, and 27 with marginallandholdings), of which we surveyed 30 (5 with medium landholdings, 9 with small land holdings, and 16 with marginal landholdings). Many of the farmers lost irrigated fields during theconstruction of the Lakhwar dam,4 and received compensationfrom the government.

Chhotau village is about 100 km from Dehradun and 10 kmfrom Chakrata town, and is at a height of about 1500 m abovemean sea level. The total number of households is 38, of which weinterviewed 32. Chhotau can be accessed only by a 1.5 km steepdownhill walk. It lacks basic amenities such as electricity,drinking water, toilets, health centre, road connectivity, tele-phones, and education beyond the primary school level. Housesare made of mud, stone, and wood, and indoor air quality is poordue to the use of traditional cookstoves. Only one house in thevillage has a toilet. The nearest primary health care centre is inChakrata and the nearest hospital is in Dehradun. After the fifthgrade, students have to walk 6–7 km to the school in Puroli orNagau. There is no collective or cooperative society5 in the village,which could help reduce vulnerability, for instance by enablingfarmers to effectively market their produce or women to engage inoff-farm income earning activities.

Perhaps the defining trend for this region is the massivemigration to cities. The entire Jaunsar-Bawar area was designatedas a tribal area in 1968, making its inhabitants eligible for theGovernment’s promotional schemes for tribals. However, Jaunsarisociety is caste stratified, with Brahmin and Rajput familiesholding most of the land (Fig. 3). It is the upper caste Brahmin andRajput ‘‘tribals’’ who have gained the most from these policies,while the lower caste Harijans have been unable to get eithereducation or employment benefits and continue to live inmarginalized and even ghettoized conditions.

As mentioned before, these communities have traditionallyhad a polyandrous patriarchal system, but the practice has

4 Work on Lakhwar dam on the Yamuna river started in 1976, but was halted

due to the high cost of production.5 A cooperative society is a collectively owned enterprise, for instance farmers

organizing themselves to market their produce collectively. Agricultural credit

cooperatives have existed in India for more than a hundred years. Successful

examples in sugar production in Maharashtra state and dairy marketing in Gujarat

state have shown how cooperatives can be vehicles of self-empowerment and rural

development.

declined over time. The general trend towards break-up of thejoint family system is observed here also, and has impacted uponlandholdings, particularly the irrigated patches of fields. Theaverage household size was found to be 7.5 in Lakhwar and 8.3 inChhotau.

Basu (1993) estimated that Jaunsar-Bawar has a literacy rate of46% for men and 15% for women. While many householdsinterviewed in Lakhwar and Chhotau had members who hadbachelor’s or even master’s degrees, these were all from uppercaste families. No Harijan had a college education; only 1 personof this caste in each village had actually completed high school,while the majority dropped out at the primary and middle schoollevels.

5.2. Climate variability and water stress: community perceptions

Almost all the households interviewed in the two villages feltthat rainfall has declined in quantity and that they could no longerrely on the timely onset of the monsoon (Table 4).

Respondents noted a decrease in scattered light rainfall usefulfor percolation, and an increase in intense rainfall which destroyscrops and runs off. Some of the indicators mentioned by Lakhwarcommunity members were:

�

Earlier one could not see the stars throughout the month ofshravana (July/August). � A few days ago temperature reached 38 1C which was

unthinkable earlier. Fans were never needed till 2–3 years ago.

� Mosquitoes have become a problem in the last 10 years,

although this is not just because of temperature rise but alsodue to poor drainage and sanitation.

� It used to rain a lot when I came to the village as a young bride.

These days the sunshine has more heat.

� Around 1984, rainfall would come in the second week of May,

but now it tends to be delayed by a month.

� Maize should have been planted by the beginning of June but

the rains have been coming late every year.

� We remember lush yields when we would get tired harvesting

the crop. Now it does not rain on time—what will grow in thefields?

� Groundwater has been declining. There is acute lack of water in

streams in the lean season.

In Chhotau the more obvious indicator was the decline insnowfall:

�

When we were children the December snow would stay on theroofs till March, forming layers of frost and ice. � During the annual fair in April in Chakrata Bazaar, the army

would have to push the snow to the sides of the road. Now youdon’t find snow even on Deoban (at a higher elevation).

� For many years there has been no snowfall in December. There

is late snow in January–February when temperatures arealready rising—it melts away and does not feed into thestreams.

� As children, as soon as we saw rain, we would bring down the

livestock from the hilly slopes, to save them from beingdrowned. Now the streams are all dry.

� 35–40 years ago it used to snow for stretches of 2–4 days when

livestock were kept in the house and men would stay indoorsand play cards. The snow was like manure—knee deep snowthat covered the ground and retained soil moisture under-neath.

� It should rain in vaisakha–jyaishtha (i.e. April/May) but the

rains are delayed by 2 months. This year the gagli and ginger

ARTICLE IN PRESS

Harijan, 5.3

Rajput, 68.7Average landholding size 2.5 ha Average landholding size 0.8 ha

Vaishya, 0.8

Joshi, 19.6

Harijan, 8.8 Hindu, 1.0

Fig. 3. Distribution of landholdings (ha) by caste in (a) Lakhwar and (b) Chhotau villages.

Table 4Community perceptions about rainfall variability

% of respondents Lakhwar Chhotau

Rainfall has declined over the last 10–15 years (%) 90 87.5

Onset of rainfall has become late (%) 65 25

Onset of rainfall has become erratic (%) 32 69

U. Kelkar et al. / Global Environmental Change 18 (2008) 564–574570

crop have all dried up because of poor monsoon rain, otherwisethey would have been waist high by the beginning of July.

There was concern that rainfall is lost to surface runoff, streamsand springs are drying up, and soil moisture has declined.

Many people linked these climatic changes to deforestationand forest fires. A farmer remembered collecting wood from thickforests as a young boy in the 1960s. Now there are eucalyptusplantations that soak up water excessively. Deforestation inUttarakhand has its roots in the expansion of the British empirein India—the growing demand for sleepers by the railways, timberand fuel for new emerging cantonments and hill stations—andreached a peak during the Second World War (Dangwal, 2005).The traditional forests in the hills were replaced by commercialmonoculture plantations of pine trees for railway sleepers andturpentine, which had a negative impact on the soil texture(Furtado, pers comm). However, at present the tendency is to cuttrees indiscriminately for timber and firewood, and to start forestfires to encourage regrowth for cattle fodder. There are no jointforest management programmes or awareness initiatives in thisregion. Without collective ownership, there is no incentive toprotect common lands.

5.3. Changes in cropping patterns and agricultural practices

The traditional cropping pattern in the region was called‘‘barahnaj’’, literally ‘‘twelve seeds’’6 that were grown together tooptimize productivity and soil fertility, ensure food security, meetdiverse household needs, and minimize expenses on agriculturalinputs like seeds and manure (Sati, 2005). However, farmers haveresponded to migratory stresses and price incentives by changingthe cropping pattern towards maize and cash crops that requireless effort and yield higher returns. But they continue to dependheavily on rainfall, and hence the challenges they face are two-

6 Mandua (finger millets), ramdana/chua (amaranthus), rajma (common

kidney beans), ogal (buckwheat), urad (green gram), moong (black gram), naurangi

(mix of pulses), gahath (horsegram), bhat (soybean), lobiya (French beans), kheera

(cucumber), bhang (cannabis) and other crops.

fold. On the one hand there are climatic and ecological stresses,while on the other, newly emerging economic and societalstructures—and these two kinds of stresses must be studied inconjunction to get the true picture of vulnerability (O’Brien andLeichenko, 2004).

1.

slop

inta

Economic pressures: The fragmentation of land holdings in thehills prevents economies of scale and creates challenges forirrigation and pest management. With poor road connectivityand frequent landslides during the rains, the lack of transpor-tation, storage, and marketing facilities is obviously a prohi-bitive barrier. Farmers do not consider it worthwhile totransport their small quantities of produce to distant marketswhere the price may be higher. As more and more land isbrought under vegetable cultivation, oversupply combinedwith the absence of cold storage facilities and the high price ofseeds means that farmers become price takers. A Lakhwarvillage elder remembered that in the 1960s, Lakhwar had aself-sufficient agriculture-based economy, with only oneperson in the village working in the service sector. But nowaspirations for city living standards and the lack of othereconomic opportunities in the villages have meant that theyounger generation no longer considers agriculture as a viablelivelihood.

2.

Climatic and ecological stresses: Although Chhotau has someirrigated fields,7 the river dries up in the summer. Crops likepotatoes, peas, gagli, and ginger have all been damaged in thelast 5–10 years due to late rains. The availability of fodder goesdown in the summer, livestock ownership is also reduced, andwith it milk production and manure availability declines. It isno longer possible to cultivate millets like mundhwa and todiya,and pulses like urad that grew on residual soil moisture afterthe maize crop had been harvested.8 The winter rains are alsodelayed, and the wheat crop has been very poor for the last10–15 years. About 15–20% of the agricultural land is leftbarren.

5.4. Women’s role in agriculture

Women bear the major burden of performing agriculturaloperations and gathering supplies for household needs. Timeaccounting exercises showed that they spend 14 h a day workingin the fields, and gathering fodder and firewood, preparingmanure, collecting drinking water, ropemaking, haymaking, etc.

7 About 10–15% of the cultivated land is river irrigated, while fields on higher

es are rainfed.8 Chopra and Pasi (2002) have highlighted the impact of this trend on protein

ke and food security.

ARTICLE IN PRESS

U. Kelkar et al. / Global Environmental Change 18 (2008) 564–574 571

Agrawal (2002) points out that men tend to perform agriculturaloperations where they have a dominant role involving animals ortools, whereas women’s jobs depend on manual labour, and hencehave a lower status. Often women are the only family membersremaining in the village to tend to the fields while the men lookfor city jobs. The result is an acute shortage of manpower, andrising cost of agricultural labour. Several girls are enrolled in cityschools and colleges but go to the city only to appear forexaminations, while working on the fields the rest of the year.But there is also some awareness of the successful examples ofnearby villages: in Dhanpau village, women made and soldorganic manure, while in Luhan village, farmers formed acooperative society and were able to command good prices forchillies and gagli.

5.5. Coping measures

Households engaged in agriculture can employ a range ofstrategies in responding to water scarcity (Narain, 2003):

�

Le

F

by improving their access to available water (e.g. makeshiftstorages, digging deeper tubewells, exchanging irrigationtimeshares, buying groundwater, and engaging in water theft),

� by reducing their demand for water (e.g. switching to less

water consumptive crops, adopting more efficient irrigationpractices, and altering dates for agricultural operations),

� by coping with the adverse impacts of periodic drought

(e.g. credit, sale of valuables and livestock, use of stored seedsand foodgrains) and

� by diversifying their sources of livelihood (e.g. alternative

employment opportunities, migration).

Fig. 4 shows coping options employed in Lakhwar village. Manyhouseholds report that in poor rainfall years they shift to lesswater-intensive crops or look for other sources of employment.Other options include obtaining loans or selling assets duringlow income periods. However, closer analysis reveals that it isthe higher-caste Rajput families who report that they takeloans, in addition to ‘‘other’’ options such as ‘‘son sends cash’’,‘‘help from brother’’, or ‘‘buying food from market’’. The lowercaste farmers have no option but to find other labour relatedjobs. The average annual income reported by Harijan house-holds was about USD 428, while that by Rajput householdswas about USD 1621 with agricultural income supplementedby income from service, pension, and money sent by familymembers.

In Chhotau the picture is more ‘‘equitable, if impoverished’’with Joshi (Brahmin) families reporting an average annual incomeof USD 408 and Harijan families reporting USD 380. Whenthe river starts drying up in the summer, farmers take turns to

Plant lesswater-

intensivecrops, 10

Find otherjobs, 17

ave village,0

Take loans,10

Sellvaluables, 0

Irrigate fields, 2

Sell land, 1

Sell livestock,1

ig. 4. Coping measures reported by farming households in Lakhwar village.

flood their fields and stay up all night till the fields slowlyfill up. People keep fewer livestock, and the quantity of milk isreduced.

Traditional money lenders charge interest of 5 Rupees permonth per 100 Rupees, i.e. 60% per annum. Commercial bankscharge 10–12 or 18% but there is cumbersome paperwork involvedand distrust due to poor understanding of terms and conditions.Reciprocal exchange of cash, produce, and most importantly,labour is a common coping strategy in the hills, particularly in theface of manpower shortages (Agrawal, 2002).

5.6. Scenarios and interventions

In group discussions carried out separately with men andwomen in the two villages, the results of the watershed modellingexercise were shared through a series of coloured charts (Fig. 5).The participants expressed keen interest in the exercise andrelated it to the water stress they were currently facing. They wererequested to recall and prepare a timeline of key developmentsthat have taken place in the village over the last fifty years, andhave impacted their lives and livelihoods (Fig. 6). They were askedto reflect on the watershed modelling results and brainstormabout future scenarios. This helped stimulate discussion aboutpossible adaptation interventions, which are summarized inTable 5.

6. Discussion

The survey in Lakhwar and Chhotau villages revealed thatcurrent coping capacity of people in the region to climatevariability and water stress is quite low. Households areconsiderably dependent on low-value rainfed agriculture. Institu-tional capacity is also poor, particularly in terms of connectivityand the availability of formal credit, which constrains their abilityto use their agricultural skills and assets more effectively. Theyalso have limited human resources in terms of formal education orvocational skills, which limits their options in seeking off-farmemployment opportunities.

The types of responses to poor rainfall reported by householdsare only temporary coping measures, some of which, like sellingassets or taking loans from traditional moneylenders, mayactually increase their vulnerability over time by worseningimpoverishment or indebtedness. As they move towards non-agricultural jobs in the city, they appear to be making a morelasting adaptation to climate stresses, but in so doing many arebecoming more vulnerable due to dislocation and disruption oftheir familiar way of life.

In general, the richer and more detailed scenario discussions inLakhwar reflect the higher education levels and relatively morecomfortable economic status of this community. Both villages,however, were similar in their belief that agriculture in its presentform is simply not a viable livelihood for future generations. Therewas concern about unemployment (‘‘crime will rise and we willget the same atmosphere as in the plains’’), and the discussionsfocused less on interventions related to water resources and moreon alternative livelihood opportunities. There was a sense that onecannot go back to the old way of life due to changing economicstructures, tastes, and aspirations.

Many of the desired interventions are highly ambitious andrequire not just technical inputs but demand surveys and areliable raw material sourcing and marketing chain. However,water harvesting interventions are clearly feasible, and are beingpromoted by the Government of Uttarakhand, albeit in a top-down manner without always understanding the ground situa-tion. One of the success stories reported by the Watershed

ARTICLE IN PRESS

Fig. 6. Timeline exercise in Lakhwar village.

Fig. 5. Sharing of watershed modelling results with women in Lakhwar village.

U. Kelkar et al. / Global Environmental Change 18 (2008) 564–574572

Management Directorate of Uttarakhand is water harvesting invillage Kui in Nir microwatershed. The village, with 31 house-holds, had only a pipeline with scanty and irregular flow, 1 kmbelow the habitation. Twenty-eight roof harvesting tankswere constructed, with the villagers being responsible for thepurchase of collection pipes. There was a positive impact onagriculture, hygiene, and women’s daily burden. While earlier thevillagers grew few vegetables (potato, garlic, coriander) in therainy season in their homestead to meet their household needs,they could now grow onion, green pea, and carrot, and increasethe production of potato, garlic, and coriander, making a profit ofRs 587–1030 per household. Hygiene too improved with avail-ability of water allowing regular bathing, washing of clothes andof animals. Time spent by women on collecting water alsoreduced considerably (Watershed Management Directorate Uttar-anchal, 2004).

7. Conclusions

The study was formulated as a pilot case for the application ofa participatory approach, whereby insights can be gained intovulnerability and adaptive capacity through mutual learning andexchange with the affected communities. The interactions withcommunities acutely highlighted the mismatch between top-down policy recommendations and ground-level needs andaspirations. It is difficult to reconcile a situation where there issevere lack of water and near abandonment of farming as alivelihood, with the National Water Policy (2002) which laysemphasis on the sale of water, and the right of the government orgram sabhas to sell excess water.

The sharing of modelling results with the community, however,can benefit through the presentation of more dynamic information.The replication and refinement of the study methodology can help

ARTICLE IN PRESS

Table 5Adaptation interventions identified by village communities

Adaptation interventions identified by

agricultural households in Lakhwar

Adaptation interventions identified by

agricultural households in Chhotau

Coping with current water levelsIndividual Community/individual

Grow pulses for self-consumption and

enhancement of soil nitrogen

Cultivation of medicinal/aromatic

plants, horticulture on barren patches

Resume cultivation of mundhwa which

can be used in baby food and wine,

and jhangura which is used in

pillows

Community/individual

Make and sell organic manure like

Dhanpau women’s society

Grow fruit-bearing trees (e.g. reetha

which does not need much water but

has commercial value for soaps and

shampoos) on barren land

Institutional

Teach agriculture, horticulture, and dairy

farming in high school

Extension workers should visit to

provide expert advice e.g. farmers

could grow 2-month hybrid maize

instead of 3-month maize to cope

with the late onset of the monsoon

Enhancement of water supplyPolicy/institutional Technological

Reforestation of the hills following the

example of Mussoorie. Incentives to

village panchayats in the form of

recognition or rewards for

reforestation or preventing forest

fires.

Rainwater harvesting storage tank

Rainwater harvesting to serve needs in

pre-monsoon months

(This can be supplemented by

constructing irrigation channels, check

dams, and percolation ponds)

Completion of the Lakhwar dam will not

bring irrigation to the village but will

raise the water table, enhance soil

moisture, and rejuvenate forests

Consolidation of land is essential for

commercial plantations and fruit

orchards that are key to the

prosperity of Himachal Pradesh

Alternatives to agricultureInstitutional Policy

Primary education standards need to be

drastically improved; the key issue is

regular attendance of teachers who

are unwilling to live in remote areas

Emphasis on ‘‘education and

employment’’

Vocational training Institutional

Policy Vocational training—sewing,

needlework

Promotion of tourism Establishment of value addition

enterprises (like packaging,

juicemaking, processing of medicinal

plants, matchstick production, etc.)

U. Kelkar et al. / Global Environmental Change 18 (2008) 564–574 573

develop a programme of participatory research on adaptationresponses to water stress that are evolved by the affectedcommunities themselves. Such a programme can help policy-makers more effectively target resources to minimize the adverseeffects of current and future water scarcity.

Acknowledgements

This research was supported by the Advanced Institute onVulnerability to Global Environmental Change, a programme

funded by the David and Lucille Packard Foundation andcoordinated by START in partnership with IIASA. The authorswould like to thank the Institute’s faculty, mentors, and super-visors, especially Ms. Barbara Huddleston, retd FAO; Ms. PreetyBhandari, TERI; Dr. Frank Biermann, Potsdam Institute; and Dr.Joanne Linerooth-Bayer, IIASA, for their valuable comments atdifferent stages of the study. We would like to express ourappreciation to SMTA staff and advisors, and particularly to itsDirector Mr. Ruben Furtado for facilitating our interactions withthe communities of Jaunsar-Bawar. Our deepest gratitude is to thepeople of Lakhwar and Chhotau villages for sharing their time andthoughts with us.

References

Agrawal, R., 2002. Small farms, women, and traditional knowledge: experiencesfrom Kumaon hills. In: 17th Symposium of the International Farming SystemsAssociation, 17–20 November 2002, Lake Buena Vista, FL, USA. /http://conference.ifas.ufl.edu/ifsa/papers/c/c1.docS.

Basu, S.K., 1993. Health status of tribal women in India. Status paper for theNational Workshop on ‘‘Status of Tribal Women in India,’’ organised by Councilfor Social Development, New Delhi and sponsored by National Commission forWomen, New Delhi, 22–23 December 1993.

Bohle, H.G., Downing, T.E., Watts, M.J., 1994. Climate change and socialvulnerability. Global Environmental Change 4 (1), 37–48.

Chopra, R., Pasi, S., 2002. Where are the Empty Thalis in Uttarakhand?: DistrictLevel Food Insecurity Analysis of Uttarakhand. People’s Science Institute,Dehradun.

CWC (Central Water Commission), 2005. Water Data Book. CWC, New Delhi/http://cwc.gov.in/main/downloads/Water_Data_Complete_Book_2005.pdfS.

Dangwal, D.D., 2005. Commercialisation of forests, timber extraction anddeforestation in Uttaranchal, 1815–1947. Conservation and Society 3 (1),110–133.

Evans, K., Velarde, S.J., Prieto, R.P., Rao, S.N., Sertzen, S., Davila, K., Cronkleton, P., deJong, W., 2006. In: Bennett, E., Zurek, M. (Eds.), Field Guide to the Future: FourWays for Communities to Think Ahead. Center for International ForestryResearch (CIFOR), ASB, World Agroforestry Centre, Nairobi, 87pp.

FSI, 2008. State of Forest Report 2005. Forest Survey of India, Ministry ofEnvironment and Forests, Dehradun.

Government of Uttaranchal, 2004. Draft water policy for Uttaranchal, 15 June2004.

IPCC, 2007. Climate change 2007: climate change impacts, adaptation andvulnerability. Working Group II contribution to the Intergovernmental Panelon Climate Change Fourth Assessment Report. Summary for policymakers,23pp.

Kasemir, B., Jager, J., Jaeger, C.C., Gardner, M.T., 2003. Public Participation inSustainability Science: A Handbook. Cambridge University Press, Cambridge,281pp.

Majumdar, D.N., 1962. Himalayan Polyandry: Structure, Functioning and CultureChange: A Field-Study of Jaunsar-Bawar. Asia Publishing House, New York, p.389.

McDonald, M.G., Harbaugh, A.W., 1988. A modular three-dimensional finite-difference ground-water flow model. In: Techniques of Water-ResourcesInvestigations. Department of the Interior, US Geological Survey, Reston, VA,USA, 576pp.

Narain, V., 2003. Institutions, Technology, and Water Control: Water UsersAssociations and Irrigation Management Reform in Two Large-Scale Systemsin India. Orient Longman, Hyderabad, 244pp.

Narula, K., Bhadwal, S., 2003. Impact of climate change on hydrology for betterdecision-making at a river basin level in India: a case study. In: TERI. 2003.Environmental Threats, Vulnerability, and Adaptation: Case Studies from India.The Energy and Resources Institute, New Delhi, pp. 55–78.

Neitsch, S.L., Arnold, J.G., Kiniry, J.R., Williams, J.R., King, K.W., 2002. Soil and WaterAssessment Tool Theoretical Documentation-Version 2000. Grassland, Soil &Water Research Laboratory, and Blackland Research and Extension Center,Temple, Texas, United States.

Noronha, L., 2003. An approach to assessing vulnerability and coping strategies.In: TERI. 2003. Environmental Threats, Vulnerability, and Adaptation:Case Studies from India. The Energy and Resources Institute, New Delhi,pp. 31–40.

O’Brien, K., Leichenko, R., 2004. Double exposure: assessing the impacts of climatechange within the context of economic globalization. Global EnvironmentalChange 10 (3), 221–232.

PSI, 2003. Survival Lessons: Himlayan Jal Sanskriti. People’s Science Institute,Dehradun, 66pp.

Sati, V.P., 2005. Systems of agriculture farming in the Uttaranchal Himalaya, India.Journal of Mountain Science 2 (1), 76–85.

Schroter, D., Polsky, C., Patt, A., 2005. Assessing vulnerabilities to the effects ofglobal change: an eight step approach. Mitigation and Adaptation Strategiesfor Global Change 10 (4), 573–595.

ARTICLE IN PRESS

U. Kelkar et al. / Global Environmental Change 18 (2008) 564–574574

Sen, A., 1981. Poverty and Famines: An Essay on Entitlement and Deprivation.Oxford University Press, Oxford.

Sen, A., 1999. Development as Freedom. Anchor Books, New York.Turner, B.L., Kasperson, R.E., Matson, P.A., McCarthy, J.J., Corell, R.W., Christensen, L.,

Eckley, N., Kasperson, J.X., Luers, A., Martello, M.L., Polsky, C., Pulsipher, A.,Schiller, A., 2003. A framework for vulnerability analysis in sustainabilityscience. Proceedings of the National Academy of Sciences 100 (4).

Tyler, S. (Ed.), 2006. Communities, Livelihoods, and Natural Resources: ActionResearch and Policy Change in Asia. International Development ResearchCentre, Ottawa, 420pp.

Venkataraman, S., Krishnan, A., 1992. Crops and Weather. Indian Council ofAgricultural Research, Pusa, 585pp.

Watershed Management Directorate Uttaranchal, 2004. Success stories. /http://gov.ua.nic.in/wmd/index.htmS.