47-62 ballesteros:Layout 1 - Agrometeorologiaagrometeorologia.it/documenti/Rivista2015_3/... ·...

achieve a rational use of irrigation water, which re-quires phenological studies to set the timing of thedifferent crop stages.When estimating the crop water requirement, theFood and Agriculture Organization of the UnitedNations (FAO) advises defining the crop coefficient(Kc) values for each of the phenological stages de-scribing the Kc curve (Allen et al., 1998). Manystudies, such as those of Wright (1982), Allen et al.,(1998) and Kato and Kamichika (2006), have deter-mined the growth intervals based on the differentslopes of the green canopy cover (GCC) curvethroughout the progression of the crop cycle, con-sidering GCC as the fraction of the soil surface thatis covered by the crop canopy (Steduto et al., 2009).A common practice is to use the same Kc curve fordifferent years without taking into account the in-fluences of the environment and other factors oncrop development for a particular year (Martínez-Cob, 2008). Plant development is controlled by sev-eral environmental factors, such as the photoperiod,soil moisture and solar radiation, but it is primarilyaffected by temperature (Viator et al., 2005). Be-cause all biological processes respond to tempera-

Calibration and validation of thermal requirementmodels for characterizing phenological stagesRocío Ballesterosa*, Miguel Ángel Morenoa, José Fernando Ortegaa

Ital

ian

Jour

nal o

f Agr

omet

eoro

logy

- 3/

2015

Riv

ista

Ita

liana

di A

grom

eteo

rolo

gia

- 3/2

015

47

1. INTRODUCTIONIn the last few decades water scarcity has become amajor concern in food security, as agriculture con-sumes 70% of the total freshwater in the world, withlarge differences between countries and evenamong different regions (FAO, 2013).In Spain, agriculture uses 75% of the freshwater re-sources, reaching 90% in areas with limited re-sources (MARM, 2010). Although recent irrigationmodernization has decreased water demands, fittingcrop water supplies to their actual requirements isone of the main approaches to conserving water inagriculture (Rodríguez-Díaz et al., 2011). In thiscontext, the activity that was developed by the Irri-gation Advisory Services (IASs) involves not onlyimportant water conservation, but also an improve-ment in the irrigation water management of an area(Ortega et al., 2005). One of the main tasks of theIAS is supplying irrigation scheduling to farmers to

Abstract: When estimating crop water requirements, the lengths of crop growth and development stages must be welldefined. The aim of this study is to analyse different thermal time calculation methods using maximum and minimumdaily temperatures together with the threshold temperatures that provide the best estimates of thermal requirementsat each phenological stage for different crops. Additionally, the green canopy cover (GCC), which was calculated withimages that were obtained with an unmanned aerial vehicle (UAV), was compared to the crop coefficent (Kc) patternthat was obtained by the growing degree-days (GDD) methods. The proposed methodology is applied in HydrogeologicUnit 08.29 (Spain). The GDD and threshold temperatures were computed for: spring wheat, spring barley, maizeFAO-700, onion and grapes using improved traditional calculation methods. The results show how the selection ofdifferent threshold temperatures at different stages and GDD calculation methods leads to an improvement upontraditional GDD calculation.Keywords: growing degree-days, Thermal Time, Threshold Temperatures, phenology, green canopy cover, crop model.

Riassunto: Quando si stima fabbisogno idrico delle colture, la durata della crescita delle colture e delle fasi di sviluppodevono essere ben definite. Lo scopo di questo studio è quello di analizzare diversi metodi di calcolo del tempo termicomediante l’utilizzo di massime e minime giornaliere insieme con le temperature di soglia che forniscono le miglioriprevisioni del fabbisogno termico in ogni fase fenologica per diverse colture. Inoltre, la canopy (GCC), che è statacalcolata con le immagini ottenute con un drone (UAV), è stato confrontata con i risultati ottenuti attraversol’applicazione del modello Kc, basato sul metodo dei gradi giorno (GDD). Le temperature e le soglie di GDD sonostate calcolate per: grano primaverile, orzo primaverile, mais FAO-700, cipolla e uva, migliorando i risultati ottenuticon i metodi di calcolo tradizionali. I risultati mostrano come la selezione di differenti temperature soglia in diversefasi e metodi di calcolo GDD porta ad un miglioramento rispetto ai sistemi tradizionali.Parole chiave: gradi giorno, Tempo termico, Soglie termiche, fenologia, canopy, resa.

* Corresponding author’s e-mail: [email protected] a Regional Centre of Water Researc (CREA). UCLM. Ctra. delas Peñas, km 3,2. 02071, Albacete. Received 20 January 2015, accepted 18 May 2015

47-62 ballesteros:Layout 1 19-01-2016 12:36 Pagina 47

Ital

ian

Jour

nal o

f Agr

omet

eoro

logy

- 3/

2015

Riv

ista

Ita

liana

di A

grom

eteo

rolo

gia

- 3/2

015

48

time calculation methods using TMAX and TMIN, aswell as to establish the threshold temperatures thatprovide the most accurate estimates for each phe-nological stage in different crops to contribute to ir-rigation scheduling and crop modelling.

2. METHODOLOGY

2.1. Area description and climatic dataThe proposed methodology is implemented in aspecific area named Hydrogeologic Unit (H.U.)08.29. This unit is located in the southeast of Spain,on the eastern side of the La Mancha plains, with atotal surface area of 8500 km2 (IGME, 1980) (Fig.1) and with relatively uniform agronomic features. Nine agro-climatic stations from the Agro-climateInformation System for Irrigation (IASI) werewithin H.U. boundaries. These stations provideddaily and half-hourly data for the considered period(2001-2012). However, most of the agro-climaticstations that are used by the IAS only provide dailyregisters as open-access data. Therefore, the se-lected GDD method should only use TMAX and TMIN

in order to be used by as many users as possible.The agro-climatic stations are designed to automat-ically register all of the required values to calculatethe reference evapotranspiration (ETo) (Allen et al.,1998), such as air humidity and temperature, solarradiation, precipitation and wind direction andspeed. The air temperature and other meteorological vari-ables were recorded from four agro-meteorological

ture, using thermal units for describing Kc curves isa more accurate approach than considering a con-stant pattern. Authors, such as Wang and Engel(1998) and Weikai and Hunt (1999), have reporteddifferent equations to estimate the rates of growthand development with a small number of parame-ters for different crops. The cited researchersadopted cardinal temperatures (maximum, mini-mum and optimum) as parameters for their equa-tions. The estimation of degree days by calculatingthe area under the curve of hourly temperatures isconsidered the reference standard method becausebiological development rates are linearly related totemperature, but temperature has a diurnal trend(Wheeler et al., 2000). However, thermal data fromthe agro-meteorological stations that are used byIAS are typically reduced to maximum daily tem-perature (TMAX) and minimum daily temperature(TMIN). Thus, the GDD calculation method needsto be adapted not only to crop phenology but also tothe availability of meteorological data. Much efforthas been made to improve the procedures for cal-culating the integration of the temperature curveusing TMAX and TMIN to accurately simulate the be-haviour of the accumulated temperatures under thehourly data curve (McMaster and Wilhem, 1997).The concept of GDD involves not only a propercombination of temperatures and time but also thethreshold temperatures that, if exceeded, involve lit-tle or no growth. All biological responses to tem-perature can be summarized in terms of the base orminimum (TLMIN), below which plants will not grow,and the maximum (TLMAX), temperatures abovewhich growth ceases completely. Most crop models simulate the effect of tempera-ture using different methods to estimate the GDD.Some of these methods, such as that implementedin AquaCrop, require a lower and upper tempera-ture for GDD calculation. These temperatures canbe considered by users as conservative parameters,meaning that they do not change with time, man-agement practice, or geographic location (Raes etal., 2009). It is important to realize that thesethreshold temperatures that are provided by theAquaCrop database and that are considered con-servative parameters may need to be adjusted to im-prove model results or even model structure (Hsiaoet al., 2009). Cropsyst offers two resolution modesfor thermal time calculation, but threshold temper-atures should also be specified by users. Therefore,it is necessary to know the right values of thresholdtemperatures to optimize the thermal time calcula-tions (Stockle and Nelson, 1994). The aim of thisstudy is to calibrate and validate different thermal

Fig. 1 - Location of 08.29 Hydrological Unit in Castilla-LaMancha (Spain), the studied commercial plots and the agro-meteorological stations.Fig. 1 - Posizione di 08,29 Unità idrologica in Castilla-LaMancha (Spagna), le aree commerciali studiate e le stazioniagro-meteorologiche.


Ital

ian

Jour

nal o

f Agr

omet

eoro

logy

- 3/

2015

Riv

ista

Ita

liana

di A

grom

eteo

rolo

gia

- 3/2

015

49

set systems, except for grape cv. ‘Cencibel’, whichwas irrigated using drip irrigation systems, andgrape cv. ‘Airén’, which was a rain-fed crop.All of the crops were established in commercialplots in which conventional cultivation techniqueswere applied by farmers (de Juan Valero et al.,2003). The crops were also grown in plots of thesame standard soil and irrigated with the same pro-cedures. Ten years (2001-2010) of weekly field observationswere performed by IAS (Ortega et al., 2005; Mon-toro et al., 2011). Changes in the GCC of crops, theheight and the leaf area, which are related to dif-ferences in evapotranspiration during the four grow-ing stages, were registered. The progression andvariation in GCC were measured by naridal imagesfrom White et al., (2000), using a digital camera be-cause it is considered the easiest and the most reli-able technique (Gitelson et al., 2002). The startingdate of each stage was defined in terms of percent-age of the population reaching that stage, and thechanges should be applied to at least 50% of theplants (Zadoks et al., 1974). At the same time de-tailed phenological changes were also registered fol-lowing the Bundessortenamt, Chemische IndustrieBBCH Scale (Meier, 2001). According to FAOmethodology (Allen et al., 1998) the observed grow-ing cycles were divided into the four FAO stages.The GDD and threshold temperatures were inde-pendently calibrated for every crop stage. The samecultivars and varieties were used throughout the dif-ferent seasons at all sites. For woody crops the sameplants were observed each year (Tab. 1).

2.3. Validation data Two additional seasons of field observations wereperformed to analyse the sensitivity of the estima-tion into the H.U. boundaries: 2010-2011 and 2011-2012 (Fig. 1). These measurements were recordedfollowing the same criteria and methodology asthose that were used for the calibration. The samecrops and varieties were considered during the val-idation process.In addition, during these two validation seasonssome field sampling was included to record theGCC pattern throughout the phenological cycle.Maize and onion were the selected crops for thesefield experiments due to their high water require-ments under semiarid conditions. Both of thesecrops were irrigated using solid set systems. High-resolution georeferenced images were taken duringthese sampling events using an unmanned aerial ve-hicle (UAV) to characterize GCC according to themethodology that was proposed by Córcoles et al.,

stations that were located in El Picazo (39.46N lat,2.09W long, and 720 m.a.s.l.) Tarazona de La Man-cha (39.25N lat, 1.91W long, and 722 m.a.s.l.), Al-bacete (38.95N lat, 1.90W long, 677 m.a.s.l.) andPozocañada (38.80N lat, 1.75W long, and 872m.a.s.l.), which are located close to the study plots(Fig. 1). Agro-climatic stations show precipitationvalues from 424 to 377 mm year-1. The precipitationvalues are not higher than 82 mm from June to Sep-tember (summer time). The annual ETo valuesrange from 1200 mm year-1 in the North of 08.29.H.U. to 1300 mm year-1 in the Southeast, while inthe summer range from 550 to 575 mm. Incorrect and questionable climatic data were iden-tified by applying a climatic series analysis and vali-dation such that faulty records could be eliminatedto avoid artificial variability. In this study, seven lev-els of quality tests as proposed by Estévez et al.,(2011) were checked. These levels were grouped ac-cording to three rules that were proposed by Meekand Hatfield (1994): 1) the range tests, which usedynamic or fixed ranges for every climatic register;2) the step tests, which use dynamic or fixed rangesfor successive observations; and 3) the persistenceor time consistency tests, which use ranges to verifyif two successive records are identical or have lowvariability. In addition to these rules, other impor-tant control procedures include internal consistencytests that verify the physical and climatologic con-sistencies of every record or between different vari-ables. Thus, the validation data structure (level 0),range test, step test, internal consistency test, per-sistent test, spatial consistency test and visual vali-dation test were applied for meteorological dataquality control. The meteorological data that did notpass level 0 or the fixed range test were automati-cally flagged as erroneous. The data that were de-tected by the other tests were flagged as suspect,and were evaluated using manual inspection (Level6). The validation procedures were applied in theframework of ISO/TC146/SC5 (ISO, 1994;AENOR, 2004).

2.2. Calibration dataThe studied crops were selected according to theirdistribution in the H.U., their economic and socialimportance and their water requirements. Consid-ering these aspects, the GDD were computed for:spring wheat (Triticum aestivum L. cv. ‘Estero’),spring barley (Hordeum vulgare L. cv. ‘Pewter’),maize FAO-700 cycle (Zea mays L. cv. ‘Dracma’),onion (Allium cepa L. cv. ‘Cíclope’), and grapes(Vitis vinifera L. cv. ‘Cencibel’ and cv. ‘Airén’).These crops were irrigated using permanent solid


peratures together with the GDD calculationmethod that best simulates plant growth were ob-tained simultaneously for each phenological stage.For example, for barley, 209 paired threshold tem-

50

Ital

ian

Jour

nal o

f Agr

omet

eoro

logy

- 3/

2015

Riv

ista

Ita

liana

di A

grom

eteo

rolo

gia

- 3/2

015

(2013) (Fig. 2). The sampling dates were selectedaccording to the phenological and GCC changes.

2.4. Temperature thresholds Maximum and minimum threshold temperature(TLMAX and TLMIN) selection is critical in GDD esti-mation. The identification of threshold tempera-tures will provide a starting point for accurate GDDcalculation and for identifying the effects on cropyield and yield components (Porter and Gawith,1999).Before selecting the threshold ranges some previousstudies were revised to establish a range of maximumand minimum temperatures to be considered as limittemperatures for each crop (Tab. 2). Different rangesof threshold temperatures were examined for everyspecies, with an interval of 1 ºC, meaning that everymaximum temperature was paired with every mini-mum temperature (Tab. 2). The GDD was calculatedusing all of the proposed methods and consideringall of the paired maximum and minimum thresholds.Thus, the maximum and minimum threshold tem-

Crops n aMean date Earliest date (year) Latest date (year) Initial (FAO-Stage I)

Spring wheat 7 4 Jan. 27 Dec. (2002) 8 Jan. (2005) Spring barley 9 18 Jan. 13 Jan. (2002) 26 Jan. (2003) Maize 8 3 May. 24 Apr. (2009) 22 May. (2002) Onion 9 16 Apr. 10 Apr. (2003) 24 Apr. (2005) Grape cv. ‘Airén’ 9 7 Apr. 21 Mar. (2009) 19 Apr.(2003) Grape cv. ‘Cencibel’ 6 14 Apr. 7 Apr. (2009) 20 Apr. (2010) Crop development (FAO-Stage II) Spring wheat 7 12 Mar. 5 Mar. (2002,) 18 Mar. (2003) Spring barley 9 5 Mar. 25 Feb. (2009) 18 Mar. (2006) Maize 8 28 May 22 May (2009) 6 June (2002) Onion 9 29 May 24 May (2001) 5 June (2002) Grape cv. ‘Airén’ 9 28 May 17 May (2006) 6 June (2004) Grape cv. ‘Cencibel’ 6 24 May 15 May (2009) 30 May (2005) Mid-season (FAO-Stage III) Spring wheat 7 7 May 27 Apr. (2009) 15 May (2003, 2004) Spring barley 8 28 Apr. 20 Apr. (2009) 18 Mar. (2006) Maize 8 17 July 12 July (2001) 18 July (2003) Onion 9 24 June 17 June (2009) 1July (2004) Grape cv. ‘Airén’ 9 14 June 4 June (2009) 25 June (2010) Grape cv. ‘Cencibel’ 6 17 June 10 June (2006) 24 June (2010) Late-season (FAO-Stage IV) Spring wheat 7 26 May 21 May (2009) 5 June (2003) Spring barley 9 23 May 14 May (2009) 1June (2003) Maize 7 10 Aug. 3 Aug. (2006) 22 Aug.(2002) Onion 9 4 Aug. 29 July (2009) 15 Aug. (2001) Grape cv. ‘Airén’ 9 30 July 22 July (2009) 6 Aug. (2004) Grape cv. ‘Cencibel’ 6 30 July 24 July (2006) 7 Aug. (2010) Full Senescence/Beginning of dormancy

Spring wheat 7 19 June 11 June (2009) 26 June (2003) Spring barley 9 14 June 4 June (2009) 23 June (2006) Maize 7 14 Sep 6 Sep. (2006) 19 Sep. (2007) Onion 9 22 Aug. 13 Aug. (2003) 30 Aug. (2006) Grape cv. ‘Airén’ 8 1Sep. 28 Aug. (2005) 8 Sep. (2009) Grape cv. ‘Cencibel’ 6 6 Sep. 22 Aug. (2006) 18 Sep. (2010) a Mean phenological starting date considering the studied period (2001-2010).

Tab. 1 - Mean and extremestarting dates, and extremedates (earliest and latest) of the six species that werestudied from 2001-2010and the number of recorded years (n).Tab. 1 - Date mediee estreme (prima e ultima)delle sei specie che sonostate studiate 2001-2010 e il numero di anni registrati (n).

Fig. 2 - Quadracopter aircraft.Fig. 2 - Quadricoptero.


peratures were evaluated (from 0-10 as TLMIN andfrom 20 to 38 as TLMAX) for every GDD calculationmethod. Thus, 1254 GDD final values were evalu-ated for each day for a particular FAO-stage.

2.5. Model SelectionFor each pair of maximum and minimum thresholdtemperatures, the GDD was calculated using: thehalf-hourly temperature (THH) as the referencemethod; this method accounts for the daily variationin temperature and, therefore, reproduces mostclosely the temperature regime that was experi-enced by the crop (Wang, 1960; Cesaraccio et al.,2001) and TMAX and TMIN using different methodsover 10 years (2001-2010). All of these temperatureswill be from agro-climatic station records from IASI. 1) The GDD from half-hourly data (GDDHH) wascomputed as follows:

where GDDHH is the computed GDD using half-hourly temperature data throughout one phenolog-ical stage, DDHH is the sum of accumulatedhalf-hourly degree days, m is the number of half-hourly data for each day, and n is the length of thephenological stage (days). 2) The GDD from TMAX and TMIN were calculatedusing different methods: a) averaging (GDDAVER-

AGE) (Arnold, 1960), b) single triangle (GDDST)(Lindsey and Newman, 1956), c) double triangle(GDDDT) (Sevacherian et al., 1977), d) sine-wave(GDDSS) (Baskerbill and Emin, 1969), e) double

=

==n

1i i

m

1jHH

HH m

DD

GDD

i

sine-wave (GDDDS) (Allen, 1976) and f) cosine(GDDCOS) (Wann et al., 1984). All of these meth-ods were compared with GDDHH, as it was con-sidered the reference method, to determine whichmethod is more accurate. Statistical hypothesistesting using Student’s t-test (t) and Chi-squared(X2) test was performed to determine the methodand threshold temperatures that best simulate cropgrowth. After accepting the null hypothesis, thecalculation model and the pair of threshold tem-peratures with the best statistical values are se-lected (Fig. 3). The selected model was verifiedand temperatures were validated to avoid erro-neous conclusions. For this, a residual analysis wasperformed using linear regression (Fig. 3). The validation of the selected threshold tempera-tures and GDD calculation methods was performedusing phenological and climatic data from two grow-ing seasons (2010-2012). With the aim of perform-ing spatial in addition to temporal validation, sixcommercial plots were selected in the 08.29 H.U.(Fig. 1). The number of GDD for each crop andphenological stage were calculated using the differ-ent methods that were selected and threshold tem-peratures to be compared with GDD obtained forthe 2001-2010 period using confidence intervals toindicate the reliability of the estimation.

3. RESULTS

3.1. Selected methods and threshold temperaturesTab. 3 shows in bold the different methods thatwere studied and the pairs of threshold tempera-tures that offered the best results with the t and

51

Ital

ian

Jour

nal o

f Agr

omet

eoro

logy

- 3/

2015

Riv

ista

Ita

liana

di A

grom

eteo

rolo

gia

- 3/2

015

Crops Range of studied TLMIN (ºC)

Range of studied TLMAX (ºC)

References

Spring wheat 0-10 20-38 (Dofing and Karlsson, 1993; Saiyed et al., 2009; Luo, 2011) Spring barley 0-10 20-38 (Kirby et al., 1982;

Juskiw et al., 2001) Maize 6-12 26-40 (Tollenaar et al., 1979;

Derieux and Bonhomme, 1982; Ritchie and Nesmith, 1991; Brown and Bootsma, 1993; Birch et al., 1998; Hsiao et al., 2009)

Onion 0-10 20-38 (Lancaster et al., 1996; Tei et al., 1996) Grape 5-15 25-38 (Winkler, 1962;

Gutierrez et al., 1985; Carbonneau, et al., 1992; Amorós et al., 2010)

Tab. 2 - Range of selectedthreshold temperatures and studied references.Tab. 2 - Gamma di temperature di soglia selezionate e riferimenti studiati.


52

Ital

ian

Jour

nal o

f Agr

omet

eoro

logy

- 3/

2015

Riv

ista

Ita

liana

di A

grom

eteo

rolo

gia

- 3/2

015

and 29 ºC for barley at the reproductive stage and36 ºC at maturation. The minimum threshold tem-perature ranged from 0 to 5 ºC for wheat and from0 to 3 ºC for barley.

3.1.2. Summer crops with high water requirementsThe double triangle and sine methods obtainedbest fittings for the different phenological stagesin maize. Attending to statistical results GDDDS

was selected as best method for the entire grow-ing cycle (t=0.639; X2=0.972). The maximumthreshold temperatures varied from 26 ºC for theearliest stage to 40 ºC during maturation, whilethe TLMIN was approximately 6 ºC for the entirecycle, except for FAO Stage-III when the TLMIN

increased to 12 ºC. The double triangle method was considered thebest fitting for the entire cycle for onion crop(t=0.674; X2=0.958). The minimum thresholdtemperature was approximately 5 ºC during theentire cycle, except at the reproductive stage,when this temperature increased to 9 ºC. Onionshowed a high tolerance to heat, with a TLMAX thatwas higher than 35 ºC during most of the pheno-logical cycle.

3.1.3. GrapeThe triangle methods, double for cv. ‘Airén’(t=0.977; X2=0.721) and simple for cv. ‘Cencibel’(t=0.999; X2=0.887) were considered as best fit-

X2 statistical tests. The methods and thresholdtemperatures for each growth stage were selectedaccording to the highest statistical significancefrom both tests. Selecting a different method foreach phenological stage did not statistically im-proves the option of selecting the same methodfor all of the phenological stages. However, theproper selection of TLMAX and TLMIN was crucial toobtaining accurate results (Tab. 3). The best sta-tistical values for the t test ranged from 0.639 formaize to 0.999 for grape cv. ‘Cencibel’. For the X2

test these values ranged from 0.624 for barley to0.972 for maize crop.

3.1.1. Spring cropsThe double triangle, which uses the minimum tem-perature from consecutive days and the maximumtemperature from the current day, was selected forboth cereals with spring vegetative growth as themost suitable. This method presents good statisticalresults for the entire phenological cycle (t=0.855;X2=0.693 for wheat and t=0.911; X2=0.624 for bar-ley). A single triangle could also be selected as bestfitting for some stages especially for spring wheat.However, despite GDDST being a very similarmethod to GDDDT, the application of GDDST forthe entire cycle did not improve compared to theresults that were obtained by GDDDT. Very similar maximum thresholds were obtained forboth crops: 20 ºC for the establishment stage, 28 ºCfor the vegetative growth stage, and 30 ºC for wheat;

Fig. 3 - Workflow of the growing degree-days (GDD) method and threshold temperature selection.Fig. 3 - Diagramma di flusso del metodo dei gradi-giorno (GDD).


tings for grapes. The minimum threshold tem-perature remained constant, from 5 to 7 ºC, fromthe bud break to flowering. The maximum thresh-olds followed a similar tendency for both varieties,although cv. ‘Airén’ was more sensitive than cv.‘Cencibel’.

3.2. Relationships between crop growthstages and growing degree-daysAfter calculating the method and threshold tem-peratures for every stage that best fit the observeddata, the mean GDD is presented in Fig. 4 for thevalidation period (2001-2010).

3.2.1. Spring cropsThe average GDD value for wheat was 1926 forthe calibration period (2001-2010). The longestphenological stage was vegetative growth (FAOStage-II) with 693 GDD, followed by maturity(FAO Stage-IV), with 504 GDD. Barley required1496 GDD to complete its phenological cycle:750 GDD for the two first stages and 746 fromflowering to maturity. The results that were obtained when the se-lected GDD method and threshold tempera-tures were applied for the two validation seasonspresent good fit regarding the confidence inter-

53

Ital

ian

Jour

nal o

f Agr

omet

eoro

logy

- 3/

2015

Riv

ista

Ita

liana

di A

grom

eteo

rolo

gia

- 3/2

015

Method FAO Stage-I FAO Stage-II FAO Stage-III FAO Stage-IV All stages

TLMIN TLMAX t X2 TLMIN TLMAX t X2 TLMIN TLMAX t X2 TLMIN TLMAX t X2 t X2

Spring wheat

GDDAVERAGE 2 21 0.000 0.950 0 28 0.909 0.901 0 36 0.537 0.939 5 30 0.188 0.980 0.683 0.010

GDDST 0 25 0.673 0.701 0 28 0.999 0.918 5 31 0.523 0.991 0 31 0.494 0.949 0.852 0.675

GDDDT 0 20 0.645 0.620 0 28 0.946 0.771 5 30 0.588 0.948 0 36 0.497 1.000 0.855 0.693 GDDSS 0 23 0.408 0.706 0 23 0.737 0.681 0 36 0.537 0.939 0 37 0.438 0.890 0.915 0.421 GDDDS 0 24 0.384 0.907 0 20 0.993 0.918 0 36 0.537 0.939 4 37 0.497 0.999 0.934 0.364 GDDCos 3 25 0.163 0.941 0 28 0.737 0.681 0 36 0.834 0.693 0 37 0.758 0.534 0.792 0.461

Spring barley

GDDAVERAGE 2 20 0.000 0.924 0 23 0.995 0.050 0 30 0.599 0.957 0 36 0.527 0.928 0.837 0.036

GDDST 5 25 0.967 0.197 1 28 0.995 0.974 5 31 0.681 0.542 3 31 0.711 0.610 0.738 0.754

GDDDT 0 20 0.521 0.658 1 28 0.932 0.944 3 29 0.838 0.432 0 36 0.578 0.950 0.911 0.624 GDDSS 0 20 0.256 0.483 0 20 0.951 0.838 10 35 0.710 0.777 0 36 0.527 0.928 0.716 0.492 GDDDS 0 20 0.243 0.993 0 30 0.841 0.893 10 36 0.793 0.687 3 35 0.577 0.952 0.789 0.717

GDDCos 0 20 0.248 0.902 0 31 0.613 0.479 0 36 0.830 0.573 0 36 0.771 0.703 0.758 0.324

Maize

GDDAVERAGE 6 26 0.000 0.000 6 33 0.043 0.721 12 38 0.964 0.944 6 35 0.903 0.903 0.000 0.000

GDDST 6 26 0.640 0.964 6 28 0.981 0.436 11 38 0.795 0.574 6 39 0.947 0.673 0.824 0.454

GDDDT 7 26 0.736 0.988 7 31 0.525 0.997 11 38 0.777 0.834 8 36 0.995 0.997 0.663 0.884 GDDSS 6 26 0.588 0.475 6 39 0.279 0.607 12 38 0.702 0.524 11 32 0.985 0.963 0.361 0.894

GDDDS 6 26 0.695 0.537 7 31 0.525 0.997 12 38 0.802 0.928 6 40 0.997 0.956 0.639 0.972 GDDCos 6 40 0.805 0.311 12 39 0.575 0.065 12 28 0.715 0.802 6 40 0.657 0.633 0.516 0.287

Onion

GDDAVERAGE 0 36 0.774 0.483 8 37 0.621 0.979 10 37 0.319 0.990 10 35 0.493 0.977 0.447 0.423

GDDST 5 23 0.796 0.910 5 37 0.507 0.437 9 35 0.632 0.972 6 38 0.613 0.306 0.615 0.877 GDDDT 5 24 0.838 0.921 5 37 0.563 0.406 9 35 0.620 0.995 6 38 0.553 0.546 0.674 0.958 GDDSS 0 26 0.651 0.520 10 38 0.617 0.929 10 35 0.331 0.987 10 37 0.612 0.355 0.391 0.719

GDDDS 0 26 0.742 0.625 10 38 0.692 0.985 8 29 0.315 0.986 10 36 0.543 0.634 0.577 0.642

GDDCos 0 36 0.970 0.266 7 35 0.853 0.998 9 25 0.106 0.998 8 38 0.512 1.000 0.675 0.218

Grape cv. ‘Airén’

GDDAVERAGE 5 28 0.994 0.027 5 36 0.691 0.599 5 39 0.959 0.819 8 39 0.685 0.675 0.994 0.316

GDDST 5 25 0.633 0.980 5 30 0.645 0.874 14 34 0.976 0.997 9 33 0.904 0.936 0.898 0.787

GDDDT 5 27 0.661 0.988 5 31 0.647 0.993 15 33 0.885 0.890 10 32 0.998 0.989 0.977 0.721 GDDSS 10 36 0.977 0.208 5 25 0.939 0.519 5 39 0.959 0.819 8 39 0.685 0.675 0.945 0.700

GDDDS 5 36 0.749 0.622 5 36 0.694 0.858 5 39 0.989 0.508 8 39 0.627 0.465 0.855 0.799

GDDCos 5 36 0.972 0.316 5 37 0.616 0.463 5 39 0.989 0.508 8 39 0.627 0.465 0.970 0.442

Grape cv. ‘Cencibel’

GDDAVERAGE 5 26 0.996 0.034 10 39 0.928 0.974 10 39 0.928 0.974 5 35 0.791 0.903 0.975 0.545

GDDST 5 29 0.844 0.994 7 39 0.892 0.889 13 39 0.885 0.916 11 35 0.595 0.728 0.999 0.887 GDDDT 5 29 0.900 0.928 7 34 0.959 0.937 15 34 0.978 0.974 11 35 0.664 0.930 0.936 0.902

GDDSS 5 37 0.946 0.787 12 39 0.923 0.993 12 39 0.923 0.993 13 35 0.531 0.923 0.884 0.716

GDDDS 5 37 0.997 0.856 5 39 0.997 0.759 5 39 0.950 0.759 9 35 0.693 0.973 0.944 0.849

GDDCos 5 37 0.848 0.336 5 32 0.971 0.465 5 32 0.971 0.465 10 35 0.911 0.793 0.909 0.609

Tab. 3 - Statistical values of Chi-squared (X2) and Student’s t-test (t) for the six methods that were evaluated for every phe-nological stage and for the entire cycle of the studied crops.Tab. 3 - I valori statistici di Chi-squared (X2) e test t (t) per i sei metodi che sono stati valutati per ogni fase fenologica e perl’intero ciclo delle colture studiate. Temperature, metodo e soglia selezionata in grassetto.


vals (Tab. 4). If the required GDD is repre-sented in days, the difference between predictedand observed GDD that was necessary to com-plete each phenological stage can be appreci-ated. For wheat and barley the error is no morethan four days, except for the establishmentstage, in which higher error can be noticed.

3.2.2. Summer crops with high water requirementsThe average value in maize GDD for the consid-ered period (2001-2010) was 1986: 938 GDD forvegetative growth and 1048 GDD from floweringto phenological maturity (Fig. 4). The length ofthe cycle from emergence to physiological matu-rity varied between 1979 and 2203 GDD. In-creases in the total cycle length were related toincreases in the duration of the vegetative growth(FAO Stage-II) and physiological maturity (FAOStage-IV) periods (Fig. 5c). The highest variationin confidence intervals was obtained at the devel-opment and maturity stages. The validation sea-sons showed good results. However, duringflowering the required GDD was below this in-terval for the 2011-2012 season (Fig. 5c). The onion crop requires 1977 GDD to complete thegrowing cycle (Fig. 4) and 972 GDD to start bul-bing.

3.2.3. GrapeResults for the cv. ‘Airén’ (2419 GDD) showedhigher heat requirements than did those for cv.‘Cencibel’ (1964 GDD) to complete the annualgrowth cycle (Fig. 4). The main differences wereobserved during the flowering stage where cv.‘Airén’ requires double the amount of GDD thandoes cv. ‘Cencibel’. However, the confidence inter-val for both varieties shows low variability for everyphenological stage (Fig. 5e, f). The validation resultsshow a good fit, with all of the obtained GDD closeto median values.

3.3. Green canopy cover estimationsAfter obtaining orthoimages from UAVs and per-forming the different sampling events throughoutthe growing cycle for maize and onion, the GCCand GDD were then used to characterize the grow-ing cycle (Fig. 6). For maize crop, the results showed GCC valuesranging from 7.29% to 95.36% and the total amountof GDD was 2059 from emergence to maturity forthe maize growing season (Fig. 6a). The maximumvalues of GCC were reached at FAO-Stage III 1208GDD. The end of FAO-Stage IV occurred at 2059GDD, at which GCC continued decreasing to 0until 2199 GDD (Fig. 6a).Fig. 6b shows the progression of GCC during the

54

Ital

ian

Jour

nal o

f Agr

omet

eoro

logy

- 3/

2015

Riv

ista

Ita

liana

di A

grom

eteo

rolo

gia

- 3/2

015

Fig. 4 - Mean patternfor phenological events(2001-2010). Each colourrepresents a different phenological stage whose size is proportional to the phenological length.The phenological stagelength in days is in parentheses.Fig. 4 - Andamendo mediodelle fasi fenologiche (2001-2010). Ogni colorerappresenta una diversafase fenologica la cui dimensione è proporzionale alla sua durata (i valori in giorni sono tra parentesi).


55

Ital

ian

Jour

nal o

f Agr

omet

eoro

logy

- 3/

2015

Riv

ista

Ita

liana

di A

grom

eteo

rolo

gia

- 3/2

015

entire crop cycle for onion crop. The GCC valuesranged from a minimum of 2.13% to a maximum of51.00%. Through FAO Stage-I, GCC values werelower than 10.00%. Green canopy cover valuesgreater than 15% could not be found until 833GDD at the end of the crop development stage. Ac-cording to the described results for maize crop, themaximum values for GCC were reached at midsea-son (FAO Stage-III), at 1570 GDD, although theGCC was near the maximum at 1300 GDD.

4. DISCUSSION

4.1. Threshold temperatures and growing de-gree estimations4.1.1. Spring cropsThe average model was used in many studies asGDD calculation method for spring cereals andwith a constant TLMIN as the only threshold con-

sidered for the entire growth cycle (Bauer et al.,1992). Although GDDAVERAGE could be consideredas an acceptable method with X2 values approxi-mately 0.9 in some phenological stages, thismethod offered poor values at FAO Stages I andIV including weak statistical results for the entirecycle (Tab. 3). This method could be applied asan alternative for simplifying the calculationprocess, but the accuracy would be reduced over-all if the same threshold temperatures were usedthroughout growth.The major difference in GDD between wheat andbarley is the lower heat unit requirements toreach maturity (Tab. 4). However, barley andwheat cultivars required approximately the samenumber of GDD from anthesis to maturating.Wheat is traditionally considered to reach an op-timum growth rate with a temperature from 17 to

Site (Year)

FAO Stage-I FAO Stage-II FAO Stage-III FAO Stage-IV Complete Phenological cycle

Length

Error (Days)

Length Error (Days)

Length Error (Days)

Length Error (Days)

Length

GDDOBS Days GDDOBS Days GDDOBS Days GDDOBS Days GDDobs Days

Spring wheat

Pozocañada (2010/2011)

449.50 70 +1 734.56 59 +2 302.38 12 0 468.78 25 +2 1955.22 166


457.84 74 +2 650.20 59 -2 322.92 10 +1 472.09 28 +2 1903.05 171

Spring barley

Tarazona de La Mancha (2010/2011)

368.99 51 +4 470.58 53 +2 339.46 24 -1 461.66 28 +2 1640.68 156


380.52 50 +6 427.67 52 -4 322.68 21 -2 487.89 31 +3 1618.76 154

Maize

Tarazona de la Mancha (2010/2011)

388.87 31 +1 555.04 34 +4 442.17 35 +1 697.88 40 +3 2083.96 140

Tarazona deLa Mancha (2011/2012)

381.76 35 +1 649.81 37 +9 363.31 42 -5 664.31 44 +1 2059.20 158

Onion


493.808 64 -4 468.880 29 +1 648.81 49 0 278.48 16 -3 1889.98 158


495.765 69 -4 404.632 32 -2 735.16 52 +4 329.00 19 -1 1964.56 172

Grape cv. Airén


782.90 65 -2 276.25 17 -1 884.92 48 +1 414.75 28 -5 2358.82 158


748.67 66 -2 318.54 19 +1 911.18 52 +2 379.65 26 -6 2358.04 163

Grape cv. Cencibel


530.558 44 +2 115.735 17 -1 855.80 34 +3 561.401 52 +1 2063.49 147


521.859 39 +2 120.888 15 -2 834.41 36 +2 521.59 51 -1 1998.74 141

Tab. 4 - Observed Growing degree-days (GDDOBS) and observed days for each phenological stage in the validation period(2010-2012) and error between the expected length and observed length where positive values indicate an overestimationbetween the expected and observed length and negative value indicates an underestimation between the expected and ob-served stages.Tab. 4 - Gradi giorno osservati (GDDOBS) e giorni osservati per ogni fase fenologica del periodo di validazione (2010-2012)e l’errore tra la durata prevista e osservati in cui i valori positivi indicano una sovrastima tra la lunghezza atteso e osservatae i valori negativi indica una sottostima.


23 ºC over the course of an entire growing season,with a TLMIN of 0 ºC and a TLMAX of 37 ºC, beyondwhich growth stops (Porter and Gawith, 1999).The threshold temperatures that were obtainedin this study across different growing stages wereconsistent with the research of Porter and Gawith(1999) who reported: 10.6 ºC for terminalspikelet, 21 ºC for anthesis and 20.7 ºC for grainfilling as optimum temperatures. Although lowerthermal units were needed for barley, similarthreshold temperatures were obtained. In addi-tion, a similar behaviour related to the optimumrange was observed: from 15 ºC during vegetati-ve stage to 17-18 ºC for anthesis (López-Belli-do, 1991). The maximum threshold temperatures increasedwith crop growth and development progress (Tab.

3): the TLMAX increased from 30 ºC at the flower-ing stage to 36 ºC at the end of the cycle, (Luo,2011). There was a low tolerance to cold temper-atures at the flowering stage because of their sen-sitivity to late frost events during the reproductiveperiod (Rawson et al., 2000; Luo 2011); thereforeso accurately estimating the TLMIN for FAO Stage-III would be an essential step for cereals.

4.1.2. Summer crops with high water requirementsMaize requires a long development season andwarm weather, its growth is not possible whereminimum mean temperatures in the summer areunder 19 ºC or night minimum temperatures arelower than 13 ºC (López-Bellido, 1991). Although38 ºC is considered by some authors as a lethal

56

Ital

ian

Jour

nal o

f Agr

omet

eoro

logy

- 3/

2015

Riv

ista

Ita

liana

di A

grom

eteo

rolo

gia

- 3/2

015

Fig. 5 - Box plot results for the growing degree-days (GDD) and selected threshold temperatures (2001-2010) and the re-sults for the two validation years.Fig. 5 - Box plot per i valori dei gradi giorno (GDD) e delle soglie selezionate (2001-2010) e i risultati dei due anni di vali-dazione.


temperature for leaf photosynthesis (Crafts-Brandner and Salvucci, 2002), the TLMAX that wasobtained at this research at reproductive stage ex-ceeded this threshold (40 ºC). The minimumthreshold temperature increase at FAO Stage-IIIis highly influenced by the sensitivity to cold dur-ing the reproductive stage as for spring cereals.To reduce this variability it is essential to accu-rately estimate the threshold temperatures fromsilking to phenological maturity.As for spring crops, the GDDAVERAGE is the refer-ence method in different environments for onioncrops (Lancaster et al., 1996; Tei et al., 1996). Ad-ditionally, a low temperature is often the onlythreshold that is considered for the whole growthcycle. However, selecting the GDDAVERAGE is notthe best method to estimate the GDD. Lancaster

et al., (1996) reported 600 GDD as the minimumvalue for onion cv. ‘Early Longkeeper’ to reach thebulbing stage. The results that were obtained inthis study for cv. ‘Cíclope’ showed higher values(Fig. 4) to reach this stage (972 GDD). Accordingto this author the reason for this crop behaviour isthe minimum day length that is required to startbulbing (13.5 h). In this study onion varieties weresowed early (around the 15th of February), involv-ing a higher number of GDD to reach bulbing.Therefore, the accurate determination of TLMIN isa key factor regarding to the bulbing stage.

4.1.3. GrapeSome authors reported optimum temperatureranges from 12 to 33 ºC for Italy, Spain and Por-tugal locations (Poni et al., 2012). Although tem-

57

Ital

ian

Jour

nal o

f Agr

omet

eoro

logy

- 3/

2015

Riv

ista

Ita

liana

di A

grom

eteo

rolo

gia

- 3/2

015

Fig. 6 - GCC patterns and crop coefficient (Kc)curve throughout thephenological cycle for a) maize and b) onion.Kcini: initial kc value of Kc;Kcmid: midseason kc value;and Kcend: final kc value.According to Allen et al.(1998), the recommendedvalues for Kcini, Kcmid

and Kcend are 0.3; 1.20 and 0. for maize and 0.7;1.05 and 0.75 for onion, 35 respectively.Fig. 6 - Andamenti di GCCe dei coefficienti colturali(Kc) per tutto il ciclo fenologico per a) granturcoe b) cipolla. Kcini: Valore iniziale di Kc;Kcmid: Valore a mezza stagione; e Kcend: Valore Kcfinale. Secondo Allen et al.(1998), i valori consigliatiper Kcini, Kcmid e Kcendsono 0,3; 1.20 e 0. per il granturco e 0.7; 1.05 e 0.75 per la cipolla


peratures below 3.5 ºC are considered on budbreak on 34 Portuguese cultivars (Lopes et al.,2008), they seem to be a too low for this area.Minimum threshold temperatures from 5 to 8 ºCfor 12 different cultivars for leaf appearance, re-ported by Moncur et al., (1989), could be morerepresentative of thermal requirements. On theother hand, the TLMAX that was selected for matu-rity was 35 ºC, which would avoid the negative im-pacts in grape production at temperatures higherthan 35 ºC in either the growing or ripening sea-son (Gladstones, 1992).

4.2. Thermal time durations

4.2.1. Spring cropsGenerally, the early development of barley couldbe more attributable to a high rate of vegetativegrowth prior to anthesis, due to the difference indevelopmental patterns between the two cropsrelative to the number of leaves, spikelet devel-opment and kernel weight (Wong and Baker,1986; Bauer et al., 1992). The onset of floweringin spring wheat usually begins after the head ex-tension stage (spikes are visible above the flagleaf collar and peduncles continue to elongate)according to Haun (1973) for em. ‘Thull’. In con-trast, the onset of flowering in spring barley be-gins at approximately the mid boot stage and iscompleted by the time the head has emergedfrom the flag leaf collar. Thus, the plant devel-opment stage at which flowering begins con-tributes to the main differences between springwheat and spring barley in the time interval fromemergence to maturity (Bauer et al., 1992). Thehighest errors in days found at establishment(Tab. 4) could be due to this stage being the mostsensitive to the accuracy of the method andthreshold temperatures (Rawson and Macpher-son, 2000).Although, Andarzian et al., (2011) reported 1756GDD for more than 10 spring wheat genotypesto the end of maturity for wheat crop, this value isquite lower than the values that were obtained inthis study. These differences can be due to se-lecting threshold temperatures (0 and 26 ºC) asconservative variables, which are considered con-stant throughout the growing cycle and used asthe GDD calculation method GDDAVERAGE.

4.2.2. Summer crops with high water requirementsSome authors obtained similar results for thesame area and cultivar: 2210, 2144 and 2061 as

GDD values for the whole growth cycle for threeconsecutive years of maize field experiments(2001-2003) (de Juan Valero et al., 2005). Thenumber of GDD at every growth stage offerssome differences with respect to the values thatwere obtained in this study: de Juan Valero et al.,(2005) obtained lower GDD values for the laststages, as the selected threshold temperatureswere considered constant for the entire cycle,which is restrictive to accumulating GDD at theend of the cycle. The thermal time that is re-quired to complete a growing stage may be con-stant and depends on the hybrid (Cutforth andShakewich, 1989). Differences in the GDD dur-ing the physiological maturity stage due to lowertemperatures are consistent with previous re-search (Dwyer and Stewart, 1986). The required GDD to reach the bulbing stage isconsistent with research by Lancaster et al.,(1996) who reported 600 GDD as the minimumvalue for onion cv. ‘Early Longkeeper’ to reachFAO Stage- III. The results that were obtained inthis study show higher values (Fig. 4) to reach thisstage (972 GDD). According to Lancaster et al.,(1996) the reason for this crop behaviour is theminimum day length that is required to start bul-bing (13.5 h). In this study onion varieties weresowed early (around the 15th of February), in-volving a higher number of GDD to reach bul-bing. Confidence intervals show good results forthe selected methods and threshold tempera-tures. Restrictive variables, such as weeds at thevegetative stage or frost, which are usual at thesestages and stop onion growth beyond GDD, couldbe the cause of the highest intervals that were ob-served during the vegetative and bulbing stages.

4.2.3. GrapeThe total amount of GDD has been traditionallyreported to be between 2800 and 4000 GDD toreach the physiological maturity of fruits (Branas,1974). These values are much higher than thosethat are required by these varieties.The thermal requirements are higher than thoseneeded for ripening in the cv. ‘Merlot’ (1693GDD) and similar to research by Amorós et al.,(2010) for cv. ‘Cencibel’. Similar data were ob-tained from 1987 to 1991 for 10 different varietiesin nearby locations such as Tomelloso (Jiménez,1993). The low impact of other climatic variables,such as precipitation or solar radiation, or culturalpractices even for rain-fed vineyards could ex-plain the low variability for every growing stageduration.

58

Ital

ian

Jour

nal o

f Agr

omet

eoro

logy

- 3/

2015

Riv

ista

Ita

liana

di A

grom

eteo

rolo

gia

- 3/2

015


4.3. Green canopy cover pattern throughoutthe growing cycle

4.3.1 MaizeDoorenbos and Pruitt (1977) reported that max-imum values should be reached at the end ofcrop development stage (Fig. 6a). Although,these maximum values were reached at 1208GDD at FAO Stage-III they were within therange of maximum values at the beginning of thisstage. The GCC began to decline at the start ofFAO Stage–IV, meaning that higher than usualmaximum values were tolerated at the end ofFAO Stage-III. During this time of maximumGCC, yield formation occurs, which involvesmaximum values being tolerated longer than ex-pected by the crop and declining at the middle ofFAO Stage-IV.Physiological maturity is the time that GCC de-clines to zero (Steduto et al., 2009). Neverthe-less, GCC continues decreasing after the end ofFAO Stage-IV; because mechanical harvest ofmaize is performed when kernels begin drying(less than 18% of water content), at 2199 GDDand not at kernel physiological maturity (35% ofwater content).The total number of GDD (2059) agrees with theinterval that was proposed by de Juan Valero etal., (2005): 2061 to 2210 as total required GDD inthe same area.

4.3.2. Onion Low obtained values of GCC through FAO Stage-I fit with Fig. 6b data from most studies (Stedutoet al., 2009). Other authors, such as López-Urreaet al., (2009), correlated GCC with Kc values,showing very similar behaviour of GCC to thepattern observed here for a similar variety (cv.‘Granero’). However, absolute values reported byLópez-Urrea et al., (2009) are higher than thosethat were collected in this study. This result maybe due to a lower plant density because the envi-ronment is quite similar and the varieties are con-sidered Long-Day onions.

5. CONCLUSIONSThe use of the proposed methodology, involvingthe proper selection of GDD calculation methodsand threshold temperatures for each phenologi-cal stage of the crop, is useful for characterizingcrop growth and development using thermal timeand can be used to advise farmers on water re-quirements. Although no significant differences were found

between the GDD methods, the same methodwere implemented across all phenological stageswith the limit of temperature variations in ordernot to limit the adoption of this methodology be-cause of its complexity. Most of the studied cropsrequired triangle-based methods except for maizecrop which required GDDSD to reach the best fitfor the different growth and development stages.For spring cereals, the selected threshold tem-peratures ranged from 0º to 5 ºC (TLMIN) and from20 to 36 ºC (TLMAX). In the case of maize, TLMIN

varied from 6 to 11 ºC and TLMAX varied from 26ºC at the establishment stage to 40 ºC at maturity.The selection of different threshold temperaturesat different stages involved an improvement inthe GDD calculation methods compared to a con-stant TLMIN of 8 ºC, which is used in most cropmodels. The selected threshold temperatureswere quite similar for the two considered varietiesof grape. These temperatures also presented uni-form behaviour throughout the phenologicalcycle: the values for TLMIN and TLMAX were ap-proximately 5 ºC and 27ºC, respectively, fromemergence to the end of growth and 10 and 35 ºCwere the minimum and maximum values to theend of the phenological cycle. The study of GCC patterns and their relationshipswith the progression of GDD are appropriatetools for characterizing crop growth and develop-ment stages based on the FAO methodology todetermine the crop water requirements.

6. ACKNOWLEDGMENTSThe authors would like to thank the EducationMinistry of Spain for financing a UniversityTeaching Scholarship (Formación de ProfesoradoUniversitario, FPU) from the Researching HumanResources Education National Program, includedin the Scientific Researching, Development andTechnological Innovation National Plan 2008e2011(EDU/3083/2009).We would like to thank to the Regional Governmentof Castilla-La Mancha for funding PEII-2014-011-P project. We also wish to thank the Water UserAssociation SORETA located in Tarazona de LaMancha, Albacete, Spain and the Irrigation Users’Association of “Eastern Mancha” for their supportof this work.

REFERENCESAENOR, 2004. UNE 500540. Redes de estaciones

meteorológicas automáticas: Directrices para lavalidación de registros meteorológicos proce-dentes de redes de estaciones automáticas. Val-

59

Ital

ian

Jour

nal o

f Agr

omet

eoro

logy

- 3/

2015

Riv

ista

Ita

liana

di A

grom

eteo

rolo

gia

- 3/2

015


idación en tiempo real. AENOR, Madrid. (InSpanish).

Allen J. C., 1976. A modified sine wave method forcalculating degree days. Environmental Ento-mology, 5: 388-396.

Allen R. G., Pereira L. S., Raes D., Smith M., 1998.Crop evapotranspiration—guidelines for com-puting crop water requirements. FAO Irrigationand Drainage, Paper 56, FAO, Rome.

Amorós J. A., Pérez C., García F. J., Campos J. A.,Muñoz F., 2010. Climate effects on ripeningprocess in Vitis vinifera, L. cv. Cencibel. Proc 8th

International Terroir Congress, Soave (Italy),June 14-18.

Andarzian B., Bannayan M., Steduto P., MazraehH., Barati M. E., Barati M. A., Rahnama A.,2011. Validation and testing of the AquaCropmodel under full and deficit irrigated wheat pro-duction in Iran. Agricultural Water Manage-ment, 100: 1-8.

Arnold C. Y., 1960. Maximum-minimum tempera-tures as a basis for computing heat units. Journalof the American Society for Horticultural Sci-ence, 76: 682-692.

Baskerville G. L., Emin P., 1969. Rapid estimationof heat accumulation from maximum and mini-mum temperatures. Ecology, 50: 514-517.

Bauer A., Frank A., Black A., 1992. A crop calendarfor spring wheat and for spring barley. NorthDakota Farm Research, 49: 21-25.

Birch C. J., Hammer G. L., Rickert K. G., 1998.Temperature and photoperiod sensitivity of de-velopment in five cultivars of maize (Zea maysL.) from emergence to tassel initiation. FieldCrop Research, 55: 93-107.

Branas J., 1974. Viticulture, Dehan, Montpellier,France. 990 pp.

Brown D., Bootsma A., 1993. Crop heat units forcorn and other warm-season crops in Ontario.Factsheet. Agdex/Ontario. Ministry of Agricul-ture and Food 111, 31.

Carbonneau A., Riou C., Guyon D., Riom J.,Schneider C., 1992. Agrométéorologie de lavigne en France. Office des publications offi-cielles des Communautés européennes, Luxem-bourg. (In French).

Cesaraccio C., Spano D., Duce P., Snyder R. L.,2001. An improved model for determining de-gree-day values from daily temperature data.International Journal of Biometeorology, 45:161-169.

Córcoles J. I., Ortega J .F., Hernández D., Mo -reno M. A., 2013. Estimation of leaf area indexin onion (Allium cepa L.) using an unmanned

aerial vehicle. Biosystem Engineering, 115(1):31-42.

Crafts-Brandner S. J., Salvucci M. E., 2002. Sensi-tivity of photosynthesis in a C4 plant, maize, toheat stress. Plant Physiology, 129: 1773-1780.

Cutforth H. W., Shaykewich C. F., 1989. Relation-ship of development rates of corn from plantingto silking to air and soil temperature and to ac-cumulated thermal units in a prairie environ-ment. Canadian Journal of Plant Science, 69:121-132.

de Juan Valero J. A., Tarjuelo J. M., Ortega J. F.,2003. Sistemas de cultivo. Evaluación de itiner-aries técnicos. Mundi Prensa Libros SA, Madrid,Spain. 835 pp. (In Spanish).

de Juan Valero J. A., Maturano M., Artigao RamírezA., Tarjuelo J. M., Ortega J. F., 2005. Growthand nitrogen use efficiency of irrigated maize ina semiarid region as affected by nitrogen fertil-ization. Spanish Journal of Agricultural Re-search, 3: 134-144.

Derieux M., Bonhomme R., 1982. Heat unit re-quirements for maize hybrids in Europe. Resultsof the European FAO sub-network: II. Periodfrom silking to maturity. Maydica 27: 79-96.

Dofing S., Karlsson M., 1993. Growth and develop-ment of uniculm and conventional-tillering bar-ley lines. Agronomy Journal, 85: 58-61.

Doorenbos J., Pruitt W.O., 1977. Crop Water Re-quirements, FAO Irrigation Drainage, Paper 24,FAO, Rome.

Dwyer L., Stewart D., 1986. Leaf area developmentin field-grown maize. Agronomy Journal, 78:334-343.

Estévez, J., Gavilán, P., Giráldez, J. V., 2011. Guide-lines on validation procedures for meteorologicaldata from automatic weather stations. Journal ofHydrology, 402(1): 144-154.

FAO, 2013. Water uses. Aquastat. FAO’s Informa-tion System on Water and Agriculture. Availablein http://www.fao.org/nr/water/aquastat/water_use/index.stm. [30 October 2014].

Gitelson A. A., Kaufman Y. J., Stark R., RundquistD., 2002. Novel algorithms for remote estima-tion of vegetation fraction. Remote Sensing ofEnvironment, 80: 76-87.

Gladstones J. S., 1992. Viticulture and environ-ment: a study of the effects of environment ongrapegrowing and wine qualities, with empha-sis on present and future areas for growingwinegrapes in Australia. Winetitles, Adelaide,Australia: 310 pp.

Gutierrez A., Williams D., Kido H., 1985. A modelof grape growth and development: the mathe-

60

Ital

ian

Jour

nal o

f Agr

omet

eoro

logy

- 3/

2015

Riv

ista

Ita

liana

di A

grom

eteo

rolo

gia

- 3/2

015


matical structure and biological considerations.Crop Science, 25: 721-728.

Haun J. R., 1973. Visual Quantification of WheatDevelopment. Agronomy Journal, 65: 116-119.

Hsiao T. C., Heng L., Steduto P., Rojas-Lara B.,Raes D., Fereres E., 2009. AquaCrop-The FAOcrop model to simulate yield response to water:III. Parameterization and testing for maize.Agronomy Journal, 101: 448-459.

IGME, 1980. El Sistema hidrogeológico de Al-bacete (Mancha Oriental): Sus recursos en aguassubterráneas, utilización actual y posibilidadesfuturas. IGME. Avalaible in http://aguas.igme.es/igme/publica/libros1_HR/libro60/lib60.htm. [30October 2014]. (In Spanish).

ISO, 1994. ISO/TC 146/SC 5. Meteorology. Qual-ity managment and quality assurance standars:Guidelines for selection and use. ISO 9000.

Jiménez J., 1993. Adaptación de diez cultivares tin-tos de vid (Vitis Vinífera L.) a la región de LaMancha. Doctoral Thesis. University of Madrid,Spain. (In Spanish).

Juskiw P. E., Jame Y. W., Kryzanowski L., 2001. Phe-nological development of spring barley in ashort-season growing area. Agronomy Journal,93: 370-379.

Kato T., Kamichika M., 2006. Determination of acrop coefficient for evapotranspiration in asparse sorghum field. Irrigation Drainage, 55:165-175.

Kirby E., Appleyard M., Fellowes M., 1982. Effectof sowing date on the temperature response ofleaf emergence and leaf size in barley. Plant, Cell& Environment, 5: 477-484.

Lancaster J., Triggs C., De Ruiter J., Gandar P.,1996. Bulbing in onions: photoperiod and tem-perature requirements and prediction of bulbsize and maturity. Annals of Botany London, 78:423-430.

Lindsey A. A., Newman J. E., 1956. Use of OfficialWather Data in Spring Time: TemperatureAnalysis of an Indiana Phenological Record.Ecology, 37: 812-823.

Lopes J., Eiras-Dias J., Abreu F., Climaco P., CunhaJ., Silvestre J., 2008. Thermal requirements, du-ration and precocity of phenological stages ofgrapevine cultivars of the Portuguese collection.Ciência e Técnica Vitivinícola, 23: 61-71.

López-Bellido L., 1991. Cereales. Cultivos her-báceos. Mundi-Presa. Madrid, Spain. 539 pp.(In Spanish).

López-Urrea R., Martín de Santa Olalla F., Mon-toro A., López-Fuster P., 2009. Single and dualcrop coefficients and water requirements for

onion (Allium cepa L.) under semiarid condi-tions. Agricultural Water Management, 96:1031-1036.

Luo Q., 2011. Temperature thresholds and cropproduction: a review. Climatic Change, 109: 583-598.

McMaster G. S., Wilhelm W., 1997. Growing de-gree days: one equation, two interpretations.Agricultural and Forest Meteorology, 87: 291-300.

MARM, 2010. Estrategia para la modernizaciónsostenible de los de regadíos, Horizonte 2015.MARM. Avaliable in http://www.magrama.gob.es/es/calidad-y-evaluacion-ambiental/ par-ticipacion-publica/pp_2009_p_019.aspx. [30 Oc-tober 2014]. (In Spanish).

Martinez-Cob A., 2008. Use of thermal units to es-timate corn crop coefficients under semiarid cli-matic conditions. Irrigation Science, 26:335-345.

Meek D. W., Hatfield J. L., 1999. Data qualitychecking for single station meteorological data-bases. Agricultural and Forest Meteorology,69(1): 85-109.Meier U., 2001. Growth Stages ofmono-and dicotyledonous plants. BBCH Mono-graph, Federal Biological Research Centre ofAgriculture and Forest. Braunschweig, Ger-many.

Moncur M. W., Rattigan K., Mackenzie D. H., McIntyre G. N., 1989. Base Temperatures for Bud-break and Leaf Appearance of Grapevines.American Journal of Enology and Viticulture, 40:21-26.

Montoro A., López-Fuster P., Fereres E., 2011. Im-proving on-farm water management through anirrigation scheduling service. Irrigation Science,29: 311-319.

Ortega J. F., de Juan J. A., Tarjuelo J. M., 2005. Im-proving water management: The irrigation advi-sory service of Castilla-La Mancha (Spain).Agricultural Water Management, 77: 37-58.

Poni S., Cola G., Mariani L., Salinari F., Diago M.P., Tardaguila J., Rossi V., 2012. Simulation mod-els: plant growth and development. Proc.MoDeM_IVM Conference, Bordeaux (France),Nov 29.

Porter J. R., Gawith M., 1999. Temperatures andthe growth and development of wheat: a review.European Journal of Agronomy, 10: 23-36.

Raes D., Steduto P., Hsiao T. C., Fereres E., 2009.AquaCrop The FAO Crop Model to SimulateYield Response to Water: II. Main Algorithmsand Software Description. Agronomy Journal,101: 438-447.

61

Ital

ian

Jour

nal o

f Agr

omet

eoro

logy

- 3/

2015

Riv

ista

Ita

liana

di A

grom

eteo

rolo

gia

- 3/2

015


Rawson H. H. M., Macpherson H. G., 2000. Irri-gated Wheat: Managing Your Crop. FAO, Rome,Italy.

Ritchie J. T., Nesmith D. S., 1991. Temperature andCrop Development. In: Modeling Plant and SoilSystems (Ritchie JT, Nesmith DS, eds.). Ameri-can Society of Agronomy, Inc., Crop Science So-ciety of America, Inc., Soil Science Society ofAmerica, Inc., Madison, WI, (United States), pp:5-29.

Rodríguez-Díaz J. A., Pérez-Urrestarazu L., Cama-cho-Poyato E., Montesinos P., 2001.. The para-dox of irrigation scheme modernization: moreefficient water use linked to higher energy de-mand. Spanish Journal of Agricultural Research,9(4): 1000-1008.

Saiyed I. M., Bullock P. R., Sapirstein H. D., FinlayG. J., Jarvis C. K., 2009. Thermal time modelsfor estimating wheat phenological developmentand weather-based relationships to wheat qual-ity. Canadian Journal of Plant Science, 89: 429-439.

Sevacherian V., Stern V. M., Mueller A. J., 1977.Heat accumulation for timing Lygus controlmeasures in a safflower-cotton complex. Journalof Economic Enthomology, 70: 399-402.

Steduto P., Hsiao T. C., Raes D., Fereres E., 2009.AquaCrop-The FAO Crop Model to SimulateYield Response to Water: I. Concepts and Under-lying Principles. Agronomy Journal, 101: 426-437.

Stockle C. O., Nelson R. L., 1994. Cropsyst User’smanual (Version 1.0). Biological Systems Engi-neering Dept., Washington State University,Pullman, WA, USA.

Tei F., Scaife A., Aikman D. P., 1996. Growth of let-tuce, onion, and red beet. 1. Growth analysis,light interception, and radiation use efficiency.Annals of Botany-London, 78: 633-643.

Tollenaar M., Daynard T., Hunter R., 1979. Effectof temperature on rate of leaf appearance andflowering date in maize. Crop Science, 19: 363-366.

Viator R. P., Nuti R. C., Edmisten K. L., Wells R.,2005. Predicting cotton boll maturation periodusing growing degree days and other climaticfactors. Agronomy Journal, 97: 494-499.

Wang E., Engel T., 1998. Simulation of phonologi-cal development of wheat crops. AgriculturalSystems, 58(1): 1-24.

Wang J. Y., 1960. A critique of the heat unit ap-proach to plant response studies. Ecology, 41:785-790.

Wann M., Yen D., Gold H. J., 1985. Evaluation andcalibration of three models for daily cycle of airtemperature. Agricultural and Forest Meteorol-ogy, 34: 121-128.

Weikai Y., Hunt L. A., 1999. An equation for mod-elling the temperature response of plants usingonly the cardinal temperatures. Annals ofBotany, 84: 607-614.

Winkler A. J., 1962. General viticulture. Universityof California Press. 710 pp.

Wheeler T. R., Craufurd P. Q., Ellis R. H., PorterJ. R., Vara Prasad P.V., 2000. Temperature vari-ability and the yield of annuals crops. Agricul-ture, Ecosystems and Environment, 82:159-167.

White M. A., Asner G. P., Nemani R. R., Privette J.L., Running S. W., 2000. Measuring fractionalcover and leaf area index in arid ecosystems.Digital camera, radiation transmittance, andlaser altimetry methods. Remote Sensing of En-vironment, 74: 45-57.

Wright J. L., 1982. New evapotranspiration crop co-efficients. Proceeding of the American Societyof Civil Engineers, Journal of the Irrigation andDrainage Division 108(IRI2): 57-74.

Wong L. S. L., Baker R. J., 1986. DevelopmentalPatterns in Five Spring Wheat Genotypes Vary-ing in Time to Maturity. Crop Science, 26: 1167-1170.

Zadoks J. C., Chang T. T., Konzak C. F., 1974. Adecimal code for growth stages of cereals. WeedResearch, 14: 415-421.

62

Ital

ian

Jour

nal o

f Agr

omet

eoro

logy

- 3/

2015

Riv

ista

Ita

liana

di A

grom

eteo

rolo

gia

- 3/2

015


47-62 ballesteros:Layout 1 - Agrometeorologiaagrometeorologia.it/documenti/Rivista2015_3/... ·...

Documents

Transcript of 47-62 ballesteros:Layout 1 - Agrometeorologiaagrometeorologia.it/documenti/Rivista2015_3/... ·...