HfERDf EK5:4PL.t\ARMAG ONDÊR 1
GEF.· O,1STA DIGHEDE t irr DiE I
~
f.~iaUOTEEI{ VERWYDER WORp NIF.J
University Free State
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII~I~ II~IIIIII III~IIIIIIIIIIIIIIIII~1111
34300002547762
Universiteit Vrystaat
FIELD COMPARISON OF RESOURCE UTILIZATION AND PRODUCTIVITY OF
THREE GRAIN LEGUME SPECIES UNDER WATER STRESS
KINDlE TESFAYE FANTAYE
FIELD COMPARISON OF RESOUCE UTILIZATION AND PRODUCTIVITY
OF THREE GRAIN LEGUME SPECIES UNDER WATER STRESS
by
KINDlE TESFAYE FANTAYE
A dissertation submitted
in accordance with
the requirement for the degree of
Doctor of Philosophy in Agrometeorology
In the Faculty of Natural and Agricultural Sciences
Department of Soil, Crop and Climate Sciences
University of the Free State
Supervisor: Professor Sue Walker
Bloemfontein
March 2004
I Univers1teit van d1e
Oranje-Vrystaat
BLOEMfONTEIN
1 6 FEB 2005
~u-o-vs-S-A-SO-L-B.IB-L-IO-TE-E-K -~
ii
Declaration
I declare that the dissertation hereby submitted by me for the degree of Doctor of Philosophy
at the University of the Free State is my own independent work and has not previously been
submitted by me at another university or faculty. I, furthermore, cede copyright of the
dissertation in favour of the University of the Free State.
Kindie~a'I)47aYi~.sfa-ye
Date: March 2004
Place: Bloemfontein, Republic of South Africa
iii
Dedication
To my grandmother
the late MANAHILOSH GESSESSE
who devoted her life to children and the needy.
iv
Contents
Title
Declaration ii
Dedication . . . . . . . . . . . . . . . . . . . . .. . . . . . . . .. . .. . . . . . . . . . . .. . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . iii
Acknowledgement v
List of Tables . . . . . . . . .. . .. . .. . . . . .. . . . . .. . . . . .. . .. . . . . . . . . . . . . . . . . .. . . . . . . . .. . . . . . . . . . . . . vi
List of Figures .. x
List of Symbols and Abbreviations . . . . . . . . . . . . . . . . . . . . .. . . . . .. . .. . .. . . . . . . . . . . . xv
Abstract XIX
Uittreksel XXll
Chapter 1: General Introduction . 1
Chapter 2: Agroclimatic Potential of Selected Locations in Ethiopia: Analysis
of Variability and Onset of Rainfall, Probability of Dry Spells and
Length of Growing Season. . . . . . . . . . . . . . . . . . . . .. . . . . . . . .. . . . . . . . .. . . . . . . . .. 13
Chapter 3: Phenology, Growth and Dry Matter Allocation in Three Grain Legume
Species Grown Under Three Water Regimes in A Semi-Arid
Environment. 38
Chapter 4: Resource Utilization of Three Grain Legume Species in a Semi-Arid
Environment. I.Water Use And Water Use Efficiency . 74
Chapter 5: Resource Utilization of Three Grain Legume Species in a Semi-Arid
Environment. II. Canopy Development, Radiation Interception and
Radiation Use Efficiency.. 92
Chapter 6: Comparative Water Relations, Leaf Gas Exchange and Assimilation
of Three Grain Legumes Under Water Deficit. . 106
Chapter 7: Comparison of Yield and Yield Components Response of Three Grain
Legumes Species to Variable Water Supply During the Reproductive
Stages........................................................................... 146
Chapter 8: Evaluation ofCROPGRO-Dry Bean and Chickpea Model in a
Semi-Arid Environment.. 165
Chapter 9: Summary and Recommendations . 182
References: 190
Appendices: 212
v
Acknowledgement
I wish to express my sincere thanks and gratitude to my supervisor, Professor Sue Walker,
who provided immeasurable support, guidance and critical comments throughout the study
period.
The contributions of the following organizations and individuals towards the success of the
research project are sincerely and gratefully acknowledged:
Alemaya University, Ethiopia, for purchasing the most expensive meteorological and
micrometeorological instruments for the field experiment and for financial support;
The National Meteorology Service Agency (NMSA) of Ethiopia for providing the long-
term meteorological data of 10stations free of charge, and its staff at Bole for their
friendly help,
The South University at Awassa for allowing its instruments to be used during the soil
profile description study at Awassa; the Ethiopian Agricultural Research Organization
(EARO) in Addis Ababa, and the South East Rangeland Project (SERP) at Jijiga for
providing meteorological data,
Dr. Eylachew Zewde, for his help during the soil profile description at Jijiga, and Mr.
Dereje Tamirat for his kind and humble support during the soil laboratory analysis at
Alemaya and Addis Ababa,
Prof. Mesifin Abebe for his extraordinary inspirational thoughts and invaluable advises.
Mr. Taddesse Mengistu and his family and staff at the Tony farm, Dire Dawa, for their
invaluable support and encouragement throughout the field experiment,
Dr. Mitsuru Tsubo for valuable comments during the write up of the thesis, and Dr. Elijah
Mukhala and Dr. Maleolm Hensley for their help during the early stage of the project,
Linda, Belmarie, Stephan, Ronell, Daniel (Agromet staff at UFS), Dr. Harun Ogindo,
Solomon, Semere, Mehari, Mike (fellow students) and other people who supported me
directly or indirectly.
My special thanks goes to Hanna Seifu, who took the painful task of typing the long-term
meteorological data and for her unfailing encouragement and support during the study period.
I owe special thanks to my family without their sacrifice and patience, this could not have
happened.
Finally, I thank with all my heart the Almighty God, who made everything possible!
VI
List of Tables
Table 2.1. Geographical description and rainfall database of ten stations used in the
study.
Table 2.2. Annual rainfall statistics of ten locations in the different ecoregions of
Ethiopia for the period 1970-2001.
Table 2.3. Average potential (PPD) and successful (SPD) planting dates, dates of end of
season calculated using ETa of site (ESa) and ET of crops planted on day
numbers indicated in parenthesis (ESc), length of growing season (LGS) and
risk of first planting for the first (if any) and the second rainy seasons for ten
locations in Ethiopia.
Table 2.4. Successful plating dates (SPD), dates of end of season (ES) and length of
growing season (LGS) at 20, 50 and 80 percentiles expressed in day of year
(DOY).
Table 2.5. Correlations between successful plating date (SPD), date of end of season
(ESa) and length of growing season (LGS).
Table 3.1a. Soil water regimes applied in the experiments and the lowest available soil
water (ASW) maintained at a depth of 300-600 mm before irrigation in each
water regime.
Table 3.1b. Duration of stress periods for the water stress treatments in each species
during the three seasons.
Table 3.2. Monthly weather conditions of the three seasons at Dire Dawa, Ethiopia.
Table 3.3. Time to emergence (TE), flowering (TF), pod initiation (TP) and maturity
(TM), and pod filling period (PFP) in the 2001/2002 and 2002/2003 seasons.
Table3.4. Time to emergence (TE), flowering (TF), pod initiation (TP) and maturity
(TM), and pod filling period (PFP) in the 2002 season.
Table 3.5. Comparisons of total above ground dry matter (ADM), leaf dry matter
(LDM), stem dry matter (SDM) and pod dry matter (PDM) production and
leaf area (LA) expansion during the post-flowering period using two-sample
t-test (t) and Kolmogorov-Smirnov test (KS) for three seasons.
Table 3.6. Allocation ratio (AR) calculated just before physiological maturity in three
grain legume species grown under three water regimes in 2002 and
2002/2003 seasons.
Table3.7. The contribution of post-flowering stem and leaf reserves to grain yield in
beans, chickpea and cowpea under well-watered (C) condition and mid-
season (MS) and late season (LS) water stress in three seasons.
vn
Table 3.8. Specific leaf area (SLA, cm2 il) of three grain legumes obtained from a
linear regression of leaf area vs. leaf dry matter for three seasons.
Table 3.9. Specific leaf area (SLA, cm2 g.l) of three grain legumes based on a linear
regression of leaf area vs. leaf dry matter for all three seasons data combined.
Table 4.1. Daytime mean vapour pressure deficit (kPa) above the canopy of three grain
legumes grown under three water regimes in three seasons for the period
between emergence and maturity.
Table 4.2. Seasonal (ETs, mm), pre-flowering (ETb• mm), post flowering (ETa. mm) and
ratio ofpre- to post flowering (ETa: ETb) water use and seasonal transpiration
(Ts, mm) and soil evaporation (Es. mm) of three grain legume species for
2002 and 2002/2003 seasons.
Table 4.3. Water use efficiency (kg ha" mm") for pre-flowering (WUEb), post
flowering WUEa), above ground dry matter at harvest (WUEd) and grain
yield (WUEg) and transpiration efficiency for grain yield (TEg, g mm") of
three grain legume species for 2002 and 2002/2003 seasons.
Table 4.4. Correlation coefficients for water use, water use efficiency and HI in three
grain legumes.
Table 5.1. Test of homogeneity of regression coefficients for K and RUE pooled over
the two seasons.
Table 6.1. Leaf water potential (MPa) of three grain legume species under well-watered
(C) and water stressed conditions during flowering (MS) and pod filling (LS)
periods in the 2001/2002 season.
Table 6.2. Leaf water potential (MPa) of three grain legume species under well-watered
_._- "---
(C) and water stressed conditions during flowering (MS) and pod filling (LS)
periods in the 2002 season.
Table 6.3. Leaf water potential (MPa) of three grain legume species under well-watered
(C) and water stressed conditions during flowering (MS) and pod filling (LS)
periods in the 2002/2003 season.
Table 6.4. Stomatal resistance (s cm") of three grain legume species under well-watered
(C) and water stressed conditions during flowering (MS) and pod filling (LS)
periods in the 2002 season.
Table 6.5. Stomatal resistance (s cm") of three grain legume species under well-watered
(C) and water stressed conditions during flowering (MS) and pod filling (LS)
periods in the 2002/2003 season.
Table 6.6. Rate of photosynthesis (Jl mol m-2 S-I) of three grain legume species under
well-watered (C) and water stressed conditions during flowering (MS) and
pod filling (LS) periods in the 2002 season.
Vlll
Table 6.7. Rate of photosynthesis (Jl mol m·2 S'I) of three grain legume species under
well-watered (C) and water stressed conditions during flowering (MS) and
pod filling (LS) periods in the 2000/2003 season.
Table 6.8. Rate of transpiration (m mol m·2 S'I) of three grain legume species under
well-watered (C) and water stressed conditions during flowering (MS) and
pod filling (LS) periods in the 2002 season.
Table 6.9. Rate of transpiration (m mol m-2 S-I) of three grain legume species under
well-watered (C) and water stressed conditions during flowering (MS) and
pod filling (LS) periods in the 2002/2003 season.
Table 6.10. Correlation coefficients among diurnal measurements of leaf water potential
(\jl, MPa), stomatal conductance (rs, s cm'), rate of photosynthesis (A, umol
m-2 sl), rate of transpiration (E, mmol m-2 S-I), air temperature (TA, 0C),
vapour pressure deficit (vpD, kPa) and photosynthetic ally active radiation
(pAR, umolm? S-I).
Table 6.11. Correlation of VPD measured at different heights of crop canopy with
available soil water (ASW, %), leaf water potential (\jiL, MPa), stomatal
resistance (rs, s cm'), rate of photosynthesis (A, umol m-2 S-I), rate of
transpiration (E, mmol m-2 S-I) and leaf temperature (Lr, 0C) in the three grain
legumes grown under water stress and non-stress conditions for two seasons.
Table 6.12. Recovery of physiological processes upon re-watering after MS stress in
2002/2003.
Table 6.13. Estimation of midday rate of photosynthesis (A, umol m-2 s"), transpiration
(E, mmol m-2 s"), leaf water potential (\jiL, MPa) and available soil water
(ASW, %) from weather, soil and plant parameters in three grain legumes
using stepwise regression for 2002 season.
Table 6.14. Estimation of midday rate of photosynthesis (A, umol m-2 s"), transpiration
(E, mmol m-2 s-\ leaf water potential (\jiL, MPa) and available soil water
(ASW, %) from weather, soil and plant parameters in three grain legumes
using stepwise regression for 2002/2003 season.
Table 6.15. Estimation of midday rate of photosynthesis (A, umol m-2 S-I), transpiration
(E, mmol m-2 s"), leaf water potential (\jiL, MPa) and available soil water
(ASW, %) from weather, soil and plant parameters in three grain legumes
using stepwise regression for data combined over two seasons.
Table 7.1. Crop evaporative deficit (1-(ETIETo)) of beans, chickpea and cowpea plants
under mid-season (MS) and late season (LS) water stress and well-watered
(C) conditions at a semi arid environment.
IX
Table 7.2. Mean squares in the analysis of variance of biomass, seed yield, number of
pods (NP) and number of seeds (NS) per meter square, 100 seed mass (SW)
and harvest index (Ill) for three grain legume species grown under three
water regimes in three seasons.
Table 7.3. Mean biomass production at harvest, grain yield, number of pods (NP) and
number of seeds (NS) per meter square, 100 seed mass (SW) and harvest
index (Ill) of three grain legumes under three water regimes in 2001/2002
and 2002.
Table 7.4. Mean biomass production at harvest, grain yield, number of pods (NP) and
number of seeds (NS) per meter square, 100 seed mass (SW) and harvest
index (Ill) of three grain legumes under three water regimes in 2002/2003.
Table 7.5. Mean crop growth rate (Cr), pod growth rate (Cp) and partitioning coefficient
(P) of three grain legumes grown under three water regimes in three seasons.
Table 7.6. Correlation (Pearson) of the grain yield of three-grain legume species with
some plant parameters under three water regimes for three seasons.
Table 8.1. Soil parameters for the experimental site at Dire Dawa, Ethiopia.
Table 8.2. Genetic coefficients of cultivars 'Roba-l ' and 'ICC4958' obtained from
"Gencalc" of DSSAT using 2001/2002 season and previous experiment data
from Dire Dawa.
Table 8.3. Statistical indexes of measured and simulated parameters of beans for data
combined over three water regimes and three seasons (n = 9).
Table 8.4. Regression coefficient for beans and chickpea from simulated and observed
data combined over three water regimes and three seasons (n = 9).
Table 8.5. Statistical indexes of measured and simulated parameters of chickpea for data
combined over three water regimes and three seasons (n = 9).
x
List of Figures
Figure 2.1. Map of Ethiopia showing the location of the meteorological station sites.
Figure 2.2. Trends of annual rainfall using 5-year moving average analysis in eight
stations in Ethiopia for the year 1970-2000.
Figure 2.3. Seasonal soil water balance of 10stations ID Ethiopia using reference
evapotranspiration (site water use) and maximum crop evapotranspiration
(crop water use).
Figure 2.4. Probabilities of receiving rainfall exceeding 0, 10, 20, 30, 40, 50, 100 and
150mm per decade at ten locations in Ethiopia.
Figure 2.5. Probabilities of maximum dry spells exceeding 5, 7, 10, 15 and 20 days
within 30 days after starting date at 10locations in Ethiopia.
Figure 2.6. Probability of dry spells exceeding 5, 7, 10, 15, and 20 days after onset
(sowing) at ten locations in Ethiopia.
Figure 2.7. Cumulative probabilities of potential (PPD) and successful (SPD) planting
dates and end of season (ESo) of the growing seasons at 10locations in
Ethiopia.
Figure 3.1. Daily maximum (Tmax) and minimum (Tmin) temperatures during the three
seasons (2001/2002, 2002 and 2002/2003).
Figure 3.2. Thermal time from planting to emergence (E), from emergence to flowering
(E-F), from flowering to podding (F-P), from podding to maturity (P-M) and
from flowering to maturity (F-M) for three grain legumes grown under well-
watered (C) and mid-season (MS) and late season (LS) water stress in three
seasons.
Figure 3.3. The seasonal course of total above ground dry matter (ADM), leaf dry matter
(LDM), stem dry matter (SDM) and pod dry matter (PDM) in beans,
chickpea and cowpea under water stress (MS, LS) and non-stress (C)
conditions in 2002.
Figure 3.4. The seasonal course of total above ground dry matter (ADM), leaf dry matter
(LDM), stem dry matter (SDM) and pod dry matter (PDM) in beans,
chickpea and cowpea under water stress (MS, LS) and non-stress (C)
conditions in 2002/2003.
Figure 3.5. The relationship between leaf area duration (LAD) and above ground dry
matter at maturity (ADM) in beans (BN), chickpea (CHP) and cowpea (COP)
for data combined over three water regimes and three seasons (n = 9).
xi
Figure 3.6. Dry matter allocation between leaf (LDM), stem (SDM) and pod (PDM) in
beans for data combined over three seasons under well-watered conditions
(C) and mid-season (MS) and late-season (LS) water stress.
Figure 3.7. Dry matter allocation between leaf (LDM), stem (SDM) and pod (PDM) in
chickpea for data combined over three seasons under well-watered conditions
(C) and mid-season (MS) and late-season (LS) water stress.
Figure 3.8. Dry matter allocation between leaf (LDM), stem (SDM) and pod (PDM) in
cowpea for data combined over three seasons under well-watered conditions
(C) condition and mid-season (MS) and late-season (LS) water stress.
Figure 3.9. An example of specific leaf area (SLA) determination by regression of leaf
area vs. leaf dry matter in beans (BN), chickpea (CHP) and cowpea grown
under well-watered (C) condition for data combined over three seasons.
Figure 3.10. The relationship between water use efficiency (WUE) and specific leaf area
(SLA) in beans (BN), chickpea (CHP) and cowpea (COP) under well-watered
conditions in two seasons.
Figure 3.11. The relationship between water use efficiency (WUB) and specific leaf area
(SLA) in beans (BN), chickpea (CHP) and cowpea (COP) under mid-season
water stress (MS) during the reproductive period in two seasons.
Figure4.1. Seasonal cumulative ET of three grain legumes species under water stress
(MS, LS) and non-stress (C) conditions in 2002 (left) and 2002/2003 (right)
seasons.
Figure 4.2. Seasonal irrigation plus rainfall received by each of the three water regimes
in 2002 and 2002/2003 seasons in a semi-arid environment.
Figure 5.1. Seasonal course of leaf area index (LAl) in beans, chickpea and cowpea
under mid-season (MS) and late season (LS) water stresses and well-watered
(C) conditions in 2002 (left) and 2002/2003 (right) seasons.
Figure 5.2. Measured fraction of PAR intercepted (F) in beans, chickpea and cowpea
under mid-season (MS) and late season (LS) water stress and well-watered
(C) conditions during 2002 (left) and 2002/2003 (right) seasons.
Figure 5.3. Illustration of canopy extinction coefficient (K) for beans, chickpea and
cowpea under mid-season (MS) and late season (LS) water stress and well-
watered (C) conditions for data combined over two seasons.
Figure 5.4. Seasonal cumulative intercepted PAR (MJ m-2 day -1) in beans, chickpea and
cowpea under mid-season (MS) and late season (LS) water stress and well-
watered (C) conditions in 2002 (left) and 2002/2003 (right) seasons.
XlI
Figure 5.5. Radiation use efficiency (RUE, g Mr') of beans, chickpea and cowpea under
mid-season (MS) and late season (LS) water stress and well-watered (C)
conditions for data combined over two seasons.
Figure 6.1. Diurnal variation of leaf water potential in beans, chickpea and cowpea under
mid-season water stress (MS) for 14 days and well-watered (C) conditions at
a semi-arid environment.
Figure 6.2. Diurnal variation of stomatal resistance in beans, chickpea and cowpea under
mid-season water stress (MS) for 14 days and well-watered (C) conditions at
a semi-arid environment.
Figure 6.3. Diurnal variation of the rate of photosynthesis in beans, chickpea and cowpea
under mid-season water stress (MS) for 14 days and well-watered (C)
conditions at a semi-arid environment.
Figure 6.4. Diurnal variation of the rate of transpiration in beans, chickpea and cowpea
under mid-season water stress (MS) for 14 days and well-watered (C)
conditions at a semi-arid environment.
Figure 6.5. Diurnal variation of photosynthetic ally active radiation (pAR, umol m-2 s-'),
air temperature (TA, 0C) and weather station vapour pressure deficit (vpD,
kPa) on 10 and 16 December 2002.
Figure 6.6. Relation of diurnal differences in leaf temperature (TL) to diurnal differences
in rate of photosynthesis (A, umol m-2 s-'), transpiration (E, mmol m-2 SO') and
stomatal resistance (r, s cm") between well-watered and mid-season water
stressed plants of beans (BN), chickpea (CHP) and cowpea (COP) for two
measurement dates (10and 16 December 2002) at a semi-arid environment.
Figure 6.7. Relation of diurnal differences in leaf temperature (TL) to diurnal differences
in leaf water potential between well-watered and mid-season water stressed
plants of beans (BN), chickpea (CHP) and cowpea (COP) for two
measurement dates (10 and 16 December 2002) at a semi-arid environment.
Figure 6.8. The relationship between available soil water (ASW), stomatal resistance (rs)
and leaf water potential (LWP) in three grain legumes under water stress (MS
&LS) and well-watered (C) conditions in 2002 (left) and 2002/2003 (right)
seasons.
Figure 6.9. The relationship between leaf water potential and stomatal resistance during
the reproductive period of three grain legumes under water stress (MS &LS)
and well-watered (C) conditions in 2002 (top) and 2002/2003 (bottom)
seasons.
Xlll
Figure 6.10. The relationship between available soil water (ASW), rate of photosynthesis
(A, umol m·2 S'I) and transpiration (E, mmol m·2 S'I) in three grain legumes
under water stress (MS &LS) and well-watered (C) conditions in 2002 (left)
and 2002/2003 (right) seasons.
Figure 6.11. The relationship between leaf water potential (LWP), rate of photosynthesis
(A, umol m·2 S'I) and transpiration (E, mmol m·2 S'I) in three grain legumes
under water stress (MS &LS) and well-watered (C) conditions in 2002 (left)
and 2002/2003 (right) seasons.
Figure 6.12. The relationship between stomatal resistance (rs, s cm"), rate of
photosynthesis (A, umol m·2 S'I) and transpiration (E, mmol m·2 S'I) in three
grain legumes under water stress (MS &LS) and well-watered (C) conditions
in 2002 (left) and 2002/2003 (right) seasons.
Figure 6.13. The relationship between rate of photosynthesis (A) and vapour pressure
deficit of the air (VPD) measured at 2 m height for the well-watered (left)
and stressed (right) plants of beans, chickpea and cowpea in 2002 and
2002/2003 seasons.
Figure 7.1. Relative yield reduction of three grain legumes due to mid-season (MS) and
late season (LS) water stress with respect to well-watered conditions in three
seasons.
Figure 7.2. Number of flowers and pods per plant at the end of the mid-season (MS)
stress (left) and number of pods per plant at maturity in the late season (LS)
stress (right) as compared to the control (C) for three grain legume species in
three seasons.
Figure 7.3. Number of primary and secondary braches per plant after the mid-season
stress (MS) and maturity of the late-season stress (LS) as compared to the
control (C) for three grain legume species in 2002/2003.
Figure 8.1. Seasonal course of measured (0) and simulated (P) LAl of beans under water
stress (MS, LS) and well-watered (C) conditions (left), and regression of
simulated vs. measured values (right) in 2002 and 2002/2003 seasons.
Figure 8.2. Seasonal course of measured (0) and simulated (P) above ground dry matter
production (ADM) of beans under water stress (MS, LS) and well-watered
(C) conditions (left), and regression of simulated vs. measured values (right)
in 2002 and 2002/2003 seasons.
Figure 8.3. Seasonal course of measured (0) and simulated (P) cumulative crop
evapotranspiration (En of beans under water stress (MS, LS) and well-
XIV
watered (C) conditions (left), and regression of simulated vs. measured
values (right) in 2002 and 2002/2003 seasons.
Figure 8.4. Comparison of simulated and measured maximum leaf area index (LAl),
above ground biomass at harvest (ADM), grain yield (Y) and harvest index
(Hl) of beans for three water regimes over three seasons.
Figure 8.5. Seasonal course of measured (0) and simulated (P) LAl of chickpea under
water stress (MS, LS) and well-watered (C) conditions (left), and regression
of simulated vs. measured values (right) in 2002 and 2002/2003 seasons.
Figure 8.6. Seasonal course of measured (0) and simulated (P) above ground dry matter
production (ADM) of chickpea under water stress (MS, LS) and well-watered
(C) conditions (left), and regression of simulated vs. measured values (right)
in 2002 and 2002/2003 seasons.
Figure 8.7. Seasonal measured (0) and simulated (P) cumulative crop evapotranspiration
(ET) of chickpea under water stress (MS, LS) and well-watered (C)
conditions (left), and regression of simulated vs. measured values (right) in
2002 and 2002/2003 seasons.
Figure 8.8. Comparison of simulated and measured maximum leaf area index (LAl),
above ground biomass at harvest (ABM), grain yield and harvest index (Hl)
of chickpea for three water regimes over three seasons.
xv
List of symbols and Abbreviations
A rate of photosynthesis (umol m-2 S-I)
AC height above canopy (cm)
J
ADM total above ground dry matter (g m-2or kg ha"; subscripts b for before
flowering, a for after flowering)
AR dry matter allocation ratio
ASW available soil water (mm or mm m")
BN common bean
C well-watered (control) treatment
CEC cation exchange capacity of soil (mmhos cm")
CHP chickpea
COP cowpea
Cp mean pod growth rate (g m-2°Cd-l)
Cr mean crop growth rate (g m-2 °Cd-I)
CV coefficient of variation (%)
d day
D drainage (mm)
index of agreement
DAP time after planting (d)
DAW time after withholding water (d)
DLL drained lower limit of soil (cm 3 cm -3 or mm m')
DM total dry matter (g m-2 or kg ha")
DOY day of year
duration of reproductive growth (d or CCd)
DSSAT Decision Support System for Agrotechnology Transfer
DUL drained upper limit of soil (cnr' cm" or mm m")
e vapour pressure of air (kPa)where subscripts s for saturation; a for actual and
superscript 0 for value calculated at a given temperature T
e; water use ratio or transpiration efficiency coefficient (g kPa kg")
E rate of transpiration (mmol m-2 S-I;subscript s for soil evaporation, mm)
ES end of season (DOY)
ET crop evapotranspiration (mm; superscripts s for seasonal, b for pre-flowering;
a for post-flowering; 0 for reference evapotanspiration)
E, cumulative transpiration (mm)
F the fraction of radiation intercepted (subscript i for experimental treatments)
FAO Food and Agriculture Organization of the United Nations
XVI
GDD growing degree day (OCd)
gs stomatal conductance (mol m-2 S-I)
ID harvest index
I PAR measured below canopy at soil surface (umol m-2s-1 or MJ m-2 d-I)
ID PAR measured above canopy (umol m-2s-1 or MJ m-2 d-I)
Ir irrigation (mm)
K canopy extinction coefficient
kc crop coefficient (subscripts prevand next for previous and next stage,
respectively and i day number with in the growing season)
KS Kolmogorov-Smirnov test
L length of crops stage (subscripts stage and prev for current and previous stage
respectively)
LA leaf area (cm2m-i or nr' m")
LAD leaf area duration (d)
LAl leaf area index
LDM leaf dry matter (g m-2)
LGS length of growing season (d)
LP lower half of plant canopy
LS late season/pod-filling period water stress
LSD least significant difference
MD mean deviation (mean bias error)
mc measured seed moisture content (%)
ME modelling efficiency
MS mid-season (flowering period) water stress
N maximum possible sunshine duration (hour)
n measured sunshine duration (hour)
NMSA National Meteorology Service Agency of Ethiopia
NP number of pods (per m-2)
NS number of seeds (per m-2)'
o observed data
o mean of observed data
OC organic carbon (%)
p dry matter partitioning coefficient
P resource use (Chapter 1)
P rainfall (mm, Chapter 2 and 4; subscript n for rainfall on a given day)
P simulated data (Chapter 8)
XVII
PAR photosyntheically active radiation (MJ m-2 d-I)
PDM pod dry matter (g m-2)
PFP pod filling period (d)
PPD potential planting date (DOY)
PTD photothermal days
R runoff (mm), superscript 2 for coefficient of determination
ROF root growth factor
RH relative humidity (%; subscripts max for maximum; min for minimum)
RI intercepted radiation (MJ m-2 d-I)
RMSE root mean square error
stomatal resistance (s cm")
RUE radiation use efficiency (g Mrl)
S soil water (mm; subscripts n for water stored on day n and n-J for previous
day)
SAT semi-arid tropics
SD standard deviation
SDM stem dry matter (g m -2)
SLA specific leaf area (g cm" )
species
SPAC soil-plant-atmosphere continuum
SPD successful planting date (d)
SR incoming solar radiation (MJ m-2 d-I)
SR, extraterrestrial radiation (MJ m-2 d-I)
SW hundred seeds mass (g)
t time during growing season (d) where subscript T for thermal time eCd)
T temperature eC) where subscripts A for air; b for base, max for maximum,
min for minimum, L for leaf
transpiration (mm; subscript s for seasonal; Chaper 4)
TDR time domain reflectometry
TE transpiration efficiency (g mm"; subscriptg for grain yield)
time to emergence (days; Chapter 3)
TF time to flowering (d)
TM time to maturity (d)
TP time to pod initiation (d)
Us mass flow of air per m2 ofleaf area (mol m-2 S-I)
UP upper half of plant canopy
VPD vapour pressure deficit of air (kPa)
xviii
w crop water use (mm)
WMO World Meteorological Organization
water regime (treatments)
weather station
water use efficiency (kg hamm") where subscripts d for total above ground
dry matter; g for grain yield; b for pre-flowering; a for post-flowering period
y grain yield (g m-2 or kg ha-I)
a canopy PAR absorption coefficient
ratio of PAR to global solar radiation
resource use efficiency
soil bulk density (mg m" or g m")
Ps soil particle density (mg m")
carbon isotope discrimination
L\c difference in CO2 concentration through measuring chamber (IJ. mol mol")
L\S change in soil water storage (mm)
L\w differential water vapour concentration (mmol mol")
'liL leaf water potential (MPa)
es soil water content at saturation (mm m" or cnr' cm")
XIX
Abstract
FIELD COMPARISON OF RESOUCE UTILIZATION AND PRODUCTIVITY OF
THREE GRAIN LEGUME SPECIES UNDER WATER STRESS
by
KINDIE TESFAYE FANTAYE
PhD in Agrometeorology at the University of the Free State
March2004
Grain legumes play a major role in low input agricultural systems by providing quality
protein to the poor communities and improving the natural resource base used for the
production of other rainfed cereal crops. The yield of the crops, however, is low mainly due to
water shortage. This study had a major aim of comparing the resource use and productivity of
beans, chickpea and cowpea under water stress and well-watered conditions in a seini-arid
environment so as to facilitate crop choice and management practices in different legume
producing environments.
Resource utilization and productivity studies for a given crop or cropping system involve both
the crop and its growing environment. In this study, therefore, resource utilization and
productivity were studied through field experimentation with three grain legume species and
analysis of rainfall/water supply behaviour of ten representative grain legume growing
regions in Ethiopia. The field experiments were conducted at Dire Dawa, Ethiopia. The
station lies in the semi-arid belt of the eastern Rift Valley escarpment with a long-term mean
annual rainfall of 612 mm and a soil dominated by Eutric Regosol. The field experiments
were conducted for three seasons in 200112002, 2002 and 2002/2003. The treatments were
three water regimes, viz., well-watered (C), mid-season (MS) and late season (LS) water
stress and three species arranged in a randomised split plot design using water regimes as
main plot and the species as sub-plot. The experiments involved measurements of important
variables in the soil-plant-atmosphere continuum.
Analysis of the long-term rainfall of 10stations in chapter 2 indicated the existence of major
regional differences in water supply. In some of the regions (e.g. Bahir Dar, Bako and Bole)
excess water is a problem while in other areas (e.g. Dire Dawa and Jijiga) water shortage is a
major bottleneck for crop production. Based on water supply, the regions were grouped as
ample water supply, intermediate water supply and poor water supply regions. The study
indicated the need to adjust crop choice and management practices based on site and seasonal
conditions.
xx
The resource utilization and productivity of the three species was studied based on a
micrometeorological approach involving phenology, growth and dry matter partitioning
(Chapter 3), water use and water use efficiency (Chapter 4), radiation and radiation use
efficiency (Chapter 5), water relations and carbon assimilation (Chapter 6) and yield and its
components (Chapter 7). Analysis of phenology and growth indicated a reduction of leaf area
and dry matter only in the MS treatment and a shortened growth period only in the LS
treatment in all species. However, species differences were observed in that the reduction in
leaf area due to MS stress was the least in cowpea compared to beans and chickpea. Both the
timing of water supply and species influenced dry matter allocation among aboveground
parts. The LS stress hastened dry matter allocation to the pod while the MS depressed it in all
species. In the LS stress, beans allocated a higher percentage of the above ground dry matter
to the seed than chickpea and cowpea during the mild temperature seasons while cowpea
allocated the highest percentage during the high temperature season. Such high dry matter
allocation to the pod is important to maintain high harvest index (HI) under water-limited
environments.
Water use varied across water regimes, the highest being in the C treatment followed by the
MS and LS treatments in descending order in all species. However, the MS treatments
resulted in the lowest water use efficiency (WOE) in all species due to low leaf area index
(LAl) and high soil evaporation. Despite differences in water use, the C and LS treatments
had similar WOE in all species indicating that some periods of water stress during the late
stage of crop growth may increase WOE and improve water saving in water-limited
environments. WOE was also strongly negatively correlated with specific leaf area (SLA)
under well-watered conditions in all species and in both seasons suggesting that it could be
used as a selection criterion for high WUE in the species. The MS treatment reduced
extinction coefficient (K) and thereby reduced fractional radiation interception (F) in all
species. Radiation use efficiency (RUE) was also negatively affected by the MS stress in
beans and chickpea whereas it was not affected by any of the water stress treatments in
cowpea.
The relationship among soil water, leaf water potential, stomatal resistance, rate of
photosynthesis (A) and transpiration (E), vapour pressure deficit and leaf temperature are
described in Chapter 6. Cowpea, followed by beans, closes its stomata at higher level of soil
water content and leaf water potential as compared to chickpea. Cowpea also has a capacity to
photosynthesise and transpire at a higher rate under favourable water supply and also to
maintain a slower rate of decline in A and E under low soil water status when compared with
beans and chickpea. The magnitude and rate of A decline was higher and faster in the MS
XXI
than in the LS stress, and among species, it was faster in chickpea than in beans and cowpea.
Stepwise regressions of data indicate that, unlike transpiration, photosynthesis could be
estimated from a few weather and physiological parameters with reasonable accuracy in all
the three species.
In contrast to cowpea, which is less and almost equally sensitive to both stress periods, the
grain yield of beans and chickpea was found to be more sensitive to the MS than the LS stress
during all seasons. The high sensitivity of beans and chickpea grain yield to the MS stress was
associated with reductions in LAl, WUB, RUE and dry matter partitioning to the pod as a
result of the stress. The lower grain yield reduction of cowpea under water stress is attributed
to the crop's ability to adjust its stomata promptly and maintain its LAl, photosynthesis and
RUE at a higher level than beans and chickpea.
Simulation of grain yield with CROPGRO in beans and chickpea gave a satisfactory result
with some limitations in simulating yield components. The model has shown a promising
potential to be used as a decision support tool in the semi-arid regions after further calibration
and testing.
The results generally show that cowpea is more productive and resource efficient than beans
and chickpea under water-limited conditions while beans is more productive and has higher
resource efficiency than cowpea and chickpea under well-watered conditions. It is concluded
that better productivity and optimum resource utilization can be achieved through proper
crop-environment matching. Moreover, crop management and breeding practices should
focus on increasing the WUB, RUE and HI of grain legumes to improve the yield of the crops
in mid-season drought prone environments.
Keywords: Beans, Chickpea, Cowpea, Gas exchange, Radiation use efficiency, Resource
utilization, Productivity, Semi-arid environment, Water deficit, Water use efficiency.
XXlI
Uittreksel
VELD-VERGELYKING VAN HULPBRONVERBRUIK EN PRODUKTIWITEIT VAN
DRIE GRAANBOONSPESIES ONDER WATERSTREMMING.
deur
KINDIE TESFA YE FANTA YE
PhD in Landbouweerkunde by die Universiteit van die Vrystaat
Maart 2004
Graanbone speel 'n belangrike rol in lae-inset landboustelsels deurdat dit kwaliteit proteïene aan
die arm gemeenskappe verskaf en die natuurlike hulpbronbasis vir die produksie van ander
droëland graangewasse verbeter. Die opbrengs van die gewasse is egter laag, hoofsaaklik a.g.v.
watertekorte. Hierdie studie het die hoofdoel gehad om die hulpbronverbruik en produktiwiteit
van bone, keker-ertjies en swartbekboontjies tydens toestande van waterstremming en geen
waterstremming te vergelyk in 'n semi-ariede omgewing om sodoende gewaskeuse te vergemaklik
en bestuurspraktyke in verskeie peulgewas-omgewings te bepaal.
Die bestudering van hulpbronverbruik en produktiwiteit binne 'n gegewe gewasstelsel behels
beide die gewas en sy groei-omgewing. Inhierdie studie is hulpbronverbruik en produktiwiteit dus
bestudeer deur veldeksperimentering waarin drie peulgewas-spesies gebruik is en die
reënval/watertoevoer van tien verteenwoordigende peulgewas groeistreke in Ethiopië ontleed is.
Die veldeksperimente is uitgevoer by Dire Dawa, Ethiopië. Die stasie is in die semi-ariede gordel
van die oostelike Skeurvallei-platorand geleë. Die langtermyn gemiddelde jaarlikse reënval is 612
mm, terwyl die grond deur 'n Eutric Regosol gedomineer word. Die veldeksperimente is vir drie
seisoene in 200112002,2002 en 2002/2003 uitgevoer. Die behandelings was drie waterverdelings,
nl. goed-gewater (C), middel-seisoen (MS) en laat-seisoen (LS) waterstremming en drie spesies in
'n ewekansige verdeelde perseelontwerp waarin waterverdelings as hoofpersele en die spesies as
sub-persele gebruik is. Die eksperimente het die meting van belangrike veranderlikes in die grond-
plant-atmosfeer kontinuum behels.
Ontleding van die lang-termyn reënval van 10 stasies in hoofstuk 2 het groot streeksverskille in
die watertoevoer uitgewys. In sommige van die streke (bv. Bahir Dar, Bako en Bole) is oortollige
water 'n probleem, terwyl watertekorte 'n groot demper op gewasproduksie plaas in ander gebiede
(bv. Dire Dawa en Jijiga). Die gebiede is aan die hand van watertoevoer gegroepeer as streke met
genoegsame watertoevoer, intermediêre watertoevoer en swak watertoevoer. Die studie het die
behoefte uitgewys om gewaskeuse en bestuurspraktyke na aanleiding van die perseel en seisoenale
toestande aan te pas. Die hulpbronverbruik en produktiwiteit van die drie spesies is ondersoek
deur gebruik te maak van 'n mikrometeorologiese benadering wat fenologie, groei en droëmassa-
skeiding (Hoofstuk 3), waterverbruik en waterverbruikseffektiwiteit (Hoofstuk 4), straling en
stralingsverbruikseffektiwiteit (Hoofstuk 5), waterverhoudinge en koolstof-assimilasie (Hoofstuk
6) en opbrengs en die komponente daarvan (Hoofstuk 7) behels. Ontleding van fenologie en groei
het gedui op 'n verlaging van blaaroppervlak en droëmassa in die MS behandeling alleenlik en 'n
verkorte groeitydperk in slegs die LS behandeling onder alle spesies. Verskille tussen spesies is
egter waargeneem aangesien die vermindering in blaaroppervlak a.g.v. MS stremming kleiner was
onder swartbekboontjies in vergelyking met bone en keker-ertjies. Beide die tydsberekening van
watertoevoer en die betrokke spesie het die droëmassa allokering onder bogrondse plantdele
beïnvloed. Die LS stremming het droëmassa allokering na die peul versnel, terwyl MS stremming
dit onder alle spesies vertraag het. Bone het meer droëmassa as keker-ertjies en swartbekboontjies
na die peul geallokeer. Sodanige droëmassa allokering na die peul is belangrik om 'n hoë
oesindeks (Hl) in waterbeperkte omgewings te onderhou.
xxiii
Waterverbruik het oor waterverdelings verskil; die hoogste verbruik het in die C behandeling
voorgekom gevolg deur die MS en LS behandelinge in dalende volgorde onder alle spesies. Die
MS behandelings het egter gelei tot die laagste waterverbruiksdoeltreffendheid (WUB) onder alle
spesies a.g.v. die blaararea-indeks (LAl) en hoë grondverdamping. Ten spyte van verskille in
waterverbruik het die C en LS behandelinge soortgelyke WUB onder alle spesies gehad wat dui
dat sommige tydperke van waterstremming gedurende die latere stadium van gewasgroei WUB
mag laat toeneem en waterbesparing in waterbeperkte omgewings bevorder. Daar was ook 'n sterk
negatiewe korrelasie tussen WUB en spesifieke braararea (SLA) onder goed-gewaterde toestande
onder alle spesies in beide seisoene wat daarop dui dat dit gebruik kan word as 'n seleksie-
kriterium vir hoë WUB onder die spesies. Die MS behandeling het die uitdunningskoëffisiënt (K)
verlaag en daardeur die gedeeltelike stralings-onderskepping (F) by alle spesies verlaag.
Stralingsverbruikseffektiwiteit (RUE) was ook negatief geaffekteer deur die MS stremming in
bone en keker-ertjies waarteen dit nie deur een van die waterstremmingsbehandelinge in
swartbekboontjies geaffekteer is nie.
Die verband tussen grondwater, blaar waterpotensiaal, huidmondjie-weerstand, fotosintese-(A) en
transpirasietempo (E), dampdrukdepressie en blaartemperatuur word in Hoofstuk 6 beskryf.
Swartbekboontjies, gevolg deur bone, sluit hul huidmondjies by hoër vlakke van grondwater-
inhoud en blaar-waterpotensiaal in teenstelling met keker-ertjies. Swartbekboontjies besit ook die
vermoë om teen 'n hoër tempo te fotosinteer tydens gunstige watertoevoer en om 'n stadiger
tempo van afuame in A en E te onderhou ten tye van lae grondwaterstatus in vergelyking met
bone en keker-ertjies. Die grootte en tempo van afuame in A was hoër en vinniger in die MS- as
in die LS-stremming, en tussen die spesies was dit vinniger in keker-ertjies as in bone en
swartbekboontjies. Stapsgewyse regressie van die data toon dat, anders as in die geval van
transpirasie, kan fotosintese met redelike akkuraatheid geskat word aan die hand van 'n paar weer-
en fisiologiese parameters onder die drie spesies.
In teenstelling met swartbekboontjies wat minder en amper ewe sensitief vir stremmingsperiodes
is, is daar gevind dat die graanopbrengs van bone en keker-ertjies meer sensitief is vir die MS- as
die LS-stremming in al die seisoene. Die hoë sensitiwiteit van boon en keker-ertjie graanopbrengs
vir die MS-stremming was geassosieer met afuames in LAl, WUB, RUE en droëmassa allokering
na die peul a.g.v. die stremming. Die laer afuame in graanopbrengs van swartbekboontjies onder
waterstremming kan toegeskryf word aan die vermoë van die gewas om sy huidmondjies vinnig
aan te pas en sy LAl, fotosintese en RUE by 'n hoër vlak as dié van bone en keker-ertjies te
onderhou.
Simulering van graanopbrengs met CROPGRO vir bone en keker-ertjies het bevredigende
resultate gelewer met 'n paar tekortkominge in die simulering van opbrengskomponente. Die
model het 'n belowende potensiaal getoon om na verdere kalibrasie en toetsing as 'n
ondersteunende besluitnemingshulpmiddel in die semi-ariede streke gebruik te word.
Die resultate toon oor die algemeen dat swartbekboontjies meer produktief en hulpbron-effektief
is as bone en keker-ertjies onder waterbeperkte toestande, terwyl bone meer produktief en 'n hoër
hulpbronverbruikseffektiwiteit as swartbekboontjies en keker-ertjies openbaar ten tye van
genoegsame watertoevoer. Die gevolgtrekking kan gemaak word dat beter produktiwiteit en
optimale hulpbronverbruik bereik kan word deur middel van gepaste gewas-omgewing passing.
Bowenal behoort die fokus van gewasbestuur en teelpraktyke te val op die verhoging van WUB,
RUE en Hl van peulgewasse om sodoende die opbrengs van die gewasse in middel-seisoen
droogte-omgewings te verhoog.
Sleutelwoorde: Bone, Keker-ertjies, Swartbekboontjies, Gas-uitruiling, Stralings-
verbruikseffektiwiteit, Hulpbronverbruik, Produktiwiteit, Semi-ariede omgewing,
Waterstremming, Waterverbruikseffektiwiteit.
1
CHAPTER 1
General Introduction
"And our water, the universal solvent, present in the air, in the soil, in plants, animals and
man. Without it life could not endure."
J.A. Toogood
Our Soil and Water
1.1. Introduction
The amount of water involved in food production is significantly higher than the amount
used in other sectors. Most of this water is provided directly by rainfall. Rainfed
agriculture depends entirely on rainfall, and it accounts for about 60% of production in
developing countries (FAO, 2003). Since the yield potential of most crops is attained
under favourable water supply environments, the potential to improve non-irrigated
yields is mainly restricted to areas where rainfall is subject to large seasonal and
interannual variations. With a high risk of yield reductions or complete crop loss from
dry spells and droughts, farmers praeticing rainfed agriculture are reluctant to invest in
inputs such as plant nutrients, high-yielding seeds and pest management (FAO, 2003).
For resource-poor farmers in semi-arid regions, the overriding requirement is to harvest
sufficient foodstuff to ensure sustained nutrition of the household through to the next
harvest.
More than 20 countries in the world, the majority of them in the arid and semi-arid
regions, are considered to be either water-scarce or water-stressed because of their
growing population and increased demand for water which is more than the
hydrological system can provide on a sustainable basis (Watson et al., 1998). As a
result, 800 million people are food-insecure, and 166 million pre-school children are
malnourished in the developing world (Rosegrant et al., 2002). Despite the increasing
demand for water in these countries, the supply is diminishing due to human activities
that degrade watersheds and threaten natural ecosystems (Goodrich et al., 2000).
Although water shortage and desertification affect all dryland areas, developing
countries are particularly vulnerable to the economic and social costs associated with the
decline of agricultural and natural ecosystem productivity (Goodrich et al., 2000). The
semi-arid tropics (SAT), which are severely affected by water shortage and
environmental degradation, include parts of 49 countries in South Asia, northern
Australia, sub-Saharan Africa, parts of eastern and southern Africa and some countries
2
of Latin America (Kumar and Abbo, 2001). One sixth of the world population inhabits
these areas, and about half of the population earns less than U.S $ 1 per day (Kumar and
Abbo, 2001). Grain legumes are among the 'major vital crops that can produce
sustainable grain yield and biomass in these harsh environments and provide quality-
protein to the inhabitants. These crops also play a major role in low input agricultural
systems. The prime advantage is their ability to fix atmospheric nitrogen and thereby
contribute positively towards the nitrogen balance of the cropping system (Subbarao et
al., 1995). Their contribution of biologically fixed nitrogen is a key factor in sustaining
long-term soil fertility in cereal production both in the developed and developing world
(Jayasundara et al., 1998). They also affect the cropping system positively by breaking
disease cycles, improving soil physical conditions and mobilization of unavailable soil
phosphorus (Hoshikawa, 1991). Therefore, a major rationale for including grain
legumes such as chickpea in the cropping system of the SAT environments is their
potential to contribute to the enhancement of the natural resource base used for the
production of other crops. These other crops are mostly staple foods of the poor
communities who rely on marginal rainfed lands. Enhancement of the natural resource
base is achieved through an increase in soil nitrogen amount which reduces the need for
fertilizer and thereby increases the saving of a household and decreases environmental
degradation (Kumar and Abbo, 2001).
Grain legumes occupy about 12.58 million ha of land in Africa and accounted for an
annual production of 5.56 million tons per annum during the 1980s (Saxena et al.,
1987).However, yield of grain legumes is generally lower and more variable than those
of many other crop species (Jeuffroy and Ney, 1997), and specifically even lower in
developing countries than in the developed ones (Oram and Agcaoili, 1992), being the
lowest in Africa when compared with other developing countries (Al-Jibouri and
Kassapu, 1987). Thus, there is a need to increase the performance of pulse crops,
particularly in developing countries, where most grain legume production is for human
consumption and demand is increasing due to increasing population pressure. Warm-
season grain legumes like common bean and cowpea and some cool-season grain
legumes such as chickpea are the most important pulses in the semi-arid and sub-humid
areas of sub-tropical Africa.
3
Common bean (Phaseolus vulgaris L.) is the major dietary protein source in East Africa
and Latin America (Graham and Ranalli, 1997). In Ethiopia, it occupies an area of 112
810 ha with a total production of94 764 tons (CSA, 1997). The crop is grown as a sole
crop or intercropped with other crops and usually receives less agricultural inputs under
multiple cropping systems. The yield ranges from 500 kg ha-l under farm conditions in
less developed countries up to 5000 kg ha-l under experimental conditions (Graham and
Ranalli, 1997). About 60% of the bean production worldwide occurs under drought
stress conditions (Graham and Ranalli, 1997), and this could be an even greater
percentage in the semi-arid regions such as East Africa where the growing season is
short and the rainfall is erratic.
Chickpea (Cieer arietinum L.) occupies an area of 11.1 million ha land worldwide with
a total annual production of 9.1 million tons, and ranks third among the worlds food
legumes (FAO, 1994). Chickpea, unlike other legumes such as grasspea, faba bean and
soybean, does not contain any major anti-nutritional chemicals and hence provides high
quality protein and starch to developing countries (Kumar and Abbo, 2001). Ethiopia is
designated as a secondary center of chickpea diversity and is the largest producer of this
crop in East Africa (Kumar and Abbo, 2001). The crop is mainly grown at an altitude of
between 1400 and 2300 m in the northern and central highlands of the country. It is
planted during August/September (van den Maesen, 1972) when the rainfall is
diminishing and hence the growth of the crop is mainly dependent on stored soil water.
About 90% of the world's chickpea is grown under rainfed conditions in a post rainy
season, on marginal lands, often without monetary inputs (Kumar and Abbo, 2001).
Drought is, therefore, the major constraint to increase the productivity of chickpea
(Kumar et al., 1996; Kumar and Abbo, 2001), the alleviation of which could lead to
50% increase in production with a value of ca. U.S. $ 900 million (Ryan, 1997).
Cowpea (Vigna anguieulata L. Walp.) is one of the most widely adapted and versatile
grain legume crops, grown on about 7 million ha of land in warm to hot regions of the
world (Rachie, 1985; Ehlers and Hall, 1997). The largest production of cowpea comes
from sub-Saharan Africa where it occupies 75% of the area of cowpea production while
the rest of the production is spread over Europe, Asia, and North America (Ehlers and
Hall, 1997). As indicated in the report of Singh (1987), the area allocated to cowpea in
Ethiopia is estimated to be 136 000 ha with a corresponding production of 34 000 tons.
4
Although dry seed is the major product of cowpea for human consumption, leaves, fresh
peas and fresh green pods are also consumed by people in different parts of the world
(Ehlers and Hall, 1997). The nutritional quality of cowpea is similar to that of common
bean but with higher levels of folic acid and lower levels of anti-nutritional and
flatulence producing factors and with a fast cooking time (Bressani, 1985; Ehlers and
Hall, 1997). Although, cowpea is intercropped with sorghum, pearl millet, maize,
cassava or cotton in many areas (Blade et al., 1997), it is also sole cropped in some
areas (Ehlers and Hall, 1997). Compared to other crop species, cowpea has considerable
adaptation to high temperature, drought and adverse edaphic factors (Hall and Patel,
1985; Ehlers and Hall, 1997). Therefore, because of its numerous attributes such as
adaptability, versatility, productivity and nutritional quality, cowpea has been chosen by
the US National Aeronautical and Space Administration (NASA) as one of few crops to
be studied for cultivation on space stations (Ehlers and Hall, 1997). Although it is
tolerant to numerous environmental constrains, cowpea is also responsive to favourable
growing environments (Ehlers and Hall, 1997). Drought is still one of the major
constraints that reduce the yield potential of cowpea in many regions (Turk et al.,
1980a).
Despite increasing demand and their vital role in sustaining the farming system, the
expansion of cereal cropping is pushing grain legume production to smaller and more
marginal areas in developing countries (e.g. Kumar and Abbo, 2001). The relegation of
these crops to marginal lands together with the ever increasing water shortage results in
low productivity and yield instability of the crops which further increases the demand
(Kumar and Abbo, 2001). Generally, producing enough food and generating adequate
income to feed the poor in the developing world is a great challenge. This challenge is
likely to intensify, with a global population that is projected to increase to 7.8 billion by
2025, putting even greater pressure on world food production, especially in developing
countries where more than 80% of the population increase is expected to occur
(Rosegrant et al., 2002). This challenge has to be tackled by increasing the productivity
of rainfed agriculture in the developing countries. One of the options to meet this
objective is integrated use of crop, weather and agroclimatic information so as to use the
available resource efficiently and maximize productivity.
5
1.2.Motivation
Water deficit limits global food productivity more severely than any other
environmental factor (Boyer, 1982; Fischer and Turner, 1978) and is the major abiotic
stress in many parts of the world (Johansen et al., 1992). As observed on many
occasions, drought remains the single most important factor threatening the food
security of many developing countries. Most developing countries that grow grain
legumes have large arid regions and, in addition, several countries have experienced
drought for extended periods. In severely affected areas, there appears to be a
widespread malnutrition problem and unless some long-term measures are taken to
enhance the cultivation of drought-resistant crops, which can provide a balanced diet,
this problem will continue.
Although the demand for grain legumes is increasing from time to time, cereal-based
production systems do not yet encourage the cultivation of these crops on the more
productive soils (Saxena et al., 1993b). As a result of many biotic and abiotic stresses,
there is a large yield gap between potential and realized yields of the legume crops
(Subbarao et al., 1995).Constraint analysis has showed that large yield and productivity
losses in grain legumes are due to water deficit (Subbarao et al., 1995). There is room,
however, to minimize and to a certain extent alleviate such losses through appropriate
scientific research. Sustainable grain legume production in water-limited environments
can be achieved through knowledge generation on agro-climate of crop growing sites,
resource capture and utilization efficiency of crops, crop-weather relations and
physiological adaptation mechanisms and integrating this knowledge into the decision
making process.
1.3. Derming the drought environment
Although drought is a common and recurring phenomenon, it lacks a single universal
definition mainly because the concepts and criteria of drought are relative and
dependent on each water user's needs and circumstances (Whitemore, 2000). According
to Wilhite and Glantz (1985) and Whitemore (2000), four commonly used definitions of
drought are identified as follows:
6
Meteorological drought is defined as a period when rainfall is significantly less than
the long-term average or some designed percentages thereof, or less than some fixed
value.
Agricultural drought occurs when soil water is reduced to levels that cause reductions
in yield of crops and/or pasture. Agricultural drought is further divided into early
season, mid-season, terminal or intermittent drought depending on the time of its
occurrence relative to the stage of crop growth.
Hydrological drought refers to a rainfall deficit capable of seriously reducing surface
and sub-surface hydrological levels.
Socio-economic drought occurs when water supply is insufficient to meet water
consumption for human activities such as industry, urban supply, irrigation, etc.
In the agronomic sense, drought refers a severe reduction in grain yield attributable to
plant water deficit (Subbarao et al., 1995). Although the magnitude or predominance of
a particular type of drought is region specific, grain legumes grown under rainfed
production system are prone to drought at any stage during their growth cycle (Subbarao
et al., 1995). Therefore, grain legumes grown under rainfed agricultural conditions can
be exposed to multiple drought stresses during the vegetative or reproductive phase of
growth. When drought occurs during the vegetative stage, the crop's recovery from the
drought depends on subsequent rainfall. On the other hand, terminal drought is the most
critical stress factor for grain legume crops grown on stored soil water during post-rainy
season (Subbarao et al., 1995), and under conditions when the seasonal rainfall is not
sufficient to recharge the soil water for reproductive growth.
Therefore, characterization of the drought pattern of the target environment is the first
step in designing strategies to alleviate drought stress (Subbarao et al., 1995). As
pointed out by the same authors, this step has been inadequately addressed in drought
research programs, mainly because of the complexity of the task. However, there is now
opportunity to deal with the problem because of the development of water balance
models and GIS (to assist in spatial visualization of the drought pattern) (Subbarao et
al., 1995) as well as progress made in developing models for analysis of daily rainfall.
This knowledge has the potential to allow estimation of long-term crop losses due to
drought stress, and the potential gains from alleviating drought stress through genetic
and management options (Subbarao et al., 1995). Since a characteristics of drought
7
resistance that is useful in one environmentmay not be useful in another, identifying the
drought behaviour of a given environment also has a potential advantage in fitting
specific drought resistance traits to specific environments and production systems (e.g.
Ludlowand Muchow, 1990). In general, " Identifying the climatic risks in the target
environment, identifying the functional components of yield affected by the
environment in the selected crops, and understanding the physiological processes
affected are important prerequisites to a successful crop improvement for drought prone
environments" (Turner et al., 2001).
1.4. Resource Utilization
Water and radiation, together with temperature, are the major natural resources that
govern the growth, development and productivity of crop plants. The capture and
.utilization of these resources by plants has been the subject of many studies in the
tropics and other environments (e.g. Monteith, 1977a, b; Squire, 1990; Morris and
Garrity, 1993;Monteith, 1994;Monteith et al., 1994; Ong et al., 1996; Williams, 2000;
Black and Ong, 2000).
In the resource capture approach, the productivity of a process is the product of the
amount of resources captured and the efficiency with which the resources are used in
producing the required product (Williams, 2000). This can be explained as
Y=PB (1.1)
where Y is the product, P is the resource used and B is the resource use efficiency. The
importance of this model in crop production is that it expresses productivity based on
resource acquisition, its conversion to biomass and the distribution of this biomass to
grain yield (Williams, 2000). Williams (2000) also indicated that the amount of resource
captured by crops depends on the availability of the resource and crop management
practices.
Radiation capture and utilization depend on the fraction of intercepted
photosynthetically active radiation (PAR) and its efficiency in producing dry matter
(Black and Ong, 2000). Though the method is criticized for its technical and theoretical
difficulties, intercepted radiation is commonly measured as the difference between the
8
total quantity of incident radiation and the quantity transmitted through the canopy to
the soil surface (Sinc1air and Muchow, 1999; Black and Ong, 2000). The amount of
radiation intercepted greatly depends on the quantity received at top of canopy, canopy
size and duration and fractional interception (Squire, 1990; Black and Ong, 2000).
Seasonal changes in fractional interception depend mainly on canopy architecture and
phenology of a given crop species. For example, the increase of fractional interception
is more rapid in cereals than in legumes because of differences in leaf initiation and
expansion (Squire, 1990). However, variation in fractional interception between crops is
smaller than the variation in green leaf area index. This is mainly because the extinction
coefficient is larger in those crops with slow canopy expansion, and as a result
maximum fractional interceptions differ little between crops grown under non-limiting
conditions (Black and Ong, 2000). Because of the difference in the duration of ground
cover, mean seasonal fractional interception values are generally lower in short-duration
cereals and legumes than perennial species (Squire, 1990; Black and Ong, 2000). In any
crop stand growing under optimal conditions (with adequate soil water, sufficient
nutrient supply, free from weed or insect infestation, and free of harmful pathogenic
activities), the dry matter (DM) production will increase linearly with the cumulative
amount of photosynthetically active radiation (PAR; 0.4-0.7 urn) that is intercepted (or
absorbed) by the canopy (Green, 1987). The efficiency of converting the intercepted
PAR into stand dry matter is defined by Monteith's (1977a) integral function:
DM
RUEj = -,2---- (1.2)
f or, (PSR)dt
Il
where RUE is the radiation-use efficiency (the subscript i denotes the experimental
treatments), t is the time of the growing season, Fi is the fraction of radiation intercepted
by the stand canopy and is a function of canopy development and stand duration, a is
the canopy absorptivity of PAR, and ft (= 0.50) is the ratio of PAR to global solar
radiation (SR). This model is well known as Monteith's "resource capture concept".
RUE can be affected by adverse environmental factors such as water stress which affect
photosynthetic activity. Therefore, RUE can be used to quantify the impact of stress
factors by comparing the observed values with those obtained under non-stress
conditions (Arkebauer et al., 1994).
9
Similar to radiation, DM production also depends on the capture and utilization of
water. The ratio of dry matter produced to water transpired or lost as evapotranspiration
is known as water use ratio or water use efficiency (Sinclair et al., 1984; Cooper et al.,
1988; Turner, 1997; Black and Ong, 2000). Therefore, dry matter production can be
expressed as
(l.3)
where ew is water use ratio, :EEw is cumulative transpiration and DM is dry matter (Black
and Ong, 2000). As observed in several studies, DM is linearly related to the quantity
of water transpired suggesting that ew is conservative (Azam-Ali, 1983; Connor et al.,
1985; Copper et al., 1987). This close relationship between DM and E; results from the
close linkage between CO2 and H20 vapor fluxes through the stomata in opposite
directions. Atmospheric vapor pressure deficit (VPD) affects the flux of C02 and H20
and it is considered as one of the most important factors that limit the productivity of
dryland areas (Squire, 1990). Although an active growing plant under well-watered
condition transpires at a rate determined by the prevailing atmospheric demand,
transpiration under water-stress condition is dictated by both plant (stomata adjustment,
rooting characteristics and leaf movement) and environmental factors (air humidity,
temperature and radiation load). In annual crop plants the canopy conductance (or its
reciprocal resistance) influences transpiration, particularly in stressed or senescent
canopies (Black and Ong, 2000). According to Ong et al. (1996), water use efficiency
during sustained drought is mainly controlled by the regulation of canopy size rather
than leaf conductance. The balance between transpiration and water absorption depends
on soil and atmospheric conditions, and a reduction in transpiration usually result in
decreased assimilation and growth (Black and Ong, 2000). C3 species have a far lower
WUE and RUE than C4 species (Squire, 1990; Sinclair and Muchow, 1999; Black and .
Ong 2000), and hence there is a need to improve the water and radiation use efficiencies
of C3 species, particularly under dry environments.
1.5. Drought resistance framework
Drought resistance in crop plants can be studied using the "drought resistance
framework" and the "resource capture or yield component framework" (Turner, 2000;
Turner et al., 2001). The drought resistance framework involves the identification of
specific morphological, physiological and biochemical characteristics that lead to
10
improved yield in dry environments. The resource capture or yield component
framework involves yield variation in terms of characteristics affecting water use and
water use efficiency, radiation use and radiation use efficiency, partitioning of
assimilates and the harvest index (Passioura, 1977; Turner, 2000; Turner et al., 2001).
The major components of the drought resistance framework are: (1) drought escape,
which involves earliness, (2) dehydration postponement, which involves maintenance of
turgor by stomatal regulation, accumulation of abscisic acid and/or osmotic adjustment,
and (3) dehydration tolerance, which involves membrane stability, tolerance to low leaf
water potential and accumulation of proline (Subbarao et al., 1995; Turner et al., 2001).
The resource capture or yield component framework involves the use of crop growth
models to study yield using physiological components that can effectively integrate a
number of complex processes into fewer biologically meaningful parameters (Turner et
al., 2001). As summarized in Turner et al., (2001), yield variation in grain legumes can
be analyzed using several resource capture models. Firstly, grain yield (Y) can be
explained using two components as follows:
Y=ADM * HI (1.4)
where ADM is total above-ground dry matter and HI is harvest index. Eq. (1.4) can be
further partitioned into functional components that can describe detailed physiological
processes for ADM and HI (Duncan et al., 1978) as follows:
(1.5)
where Cr is crop growth rate, Dr is duration of reproductive growth and p is the
partitioning coefficient (proportion of Cr portioned to yield). Y can also be analyzed as a
function of radiation interception and use as described by Monteith (1977a) as
Y = RI * RUE * HI (1.6)
where RI is cumulative intercepted radiation and RUE is radiation use efficiency. In
contrast, Passioura (1977) described yield in water deficit environments as a function of
water use and water use efficiency as
Y=W*WUE*HI (1.7)
where W is amount of water utilized by the crop and WUE is water use efficiency.
11
Each subcomponent of the various yield models represents an integrated function of a
number of physiological, morphological and biochemical characters (Hardwick 1988a;
Turner et al., 2001) and hence provide an integrated measure of crop performance in a
given environment. Any potential characters for drought adaptability can thus be
evaluated based on its functional relationship and strength of its correlation to one of the
yield components (Turner et al., 2001). Both the drought resistance and yield
component frameworks have been widely exploited in the improvement of yield in
cereal crops under drought prone environments (Ludlowand Muchow, 1990; Richards,
1996; Turner 1997; Turner et al., 2001). However, such information is sti11lacking for
most grain legumes.
1.6. Rationale
Water stress reduces crop growth on nearly all arable land (Solh, 1993) and severely
limits agricultural productivity (Boyer, 1982). Drought is probably the most important
stress factor limiting crop yields worldwide (Jones and Corlett, 1992). Furthermore, it is
often difficult to distinguish between direct effects of drought and its interactions with
other factors such as harmful pathogenic soil fungi, low soil fertility, and high air
temperatures. Drought affects every aspect of plant growth and the worldwide losses in
yield from drought probably exceed the loss from all other causes combined (Kramer,
1980). Therefore, drought at anyone stage of crop growth is the primary reason that
crop yields fall below their genetic potential and vary from year to year.
The drought-prone areas of Ethiopia cover about 60% of the total area of the country
(MoA, 1998) and account for 46% of the total cultivated land but contribute less than
10% of the total crop production in the country as a result of water stress (Reddy and
Kidane, 1994). These drought prone areas are characterized by erratic rainfall and a hot
dry climate with low annual precipitation amount and a short crop growing season
(Simane, et al., 1998; Reddy and Kidane, 1994). Although beans and cowpea are
usually grown by farmers in arid and semi-arid zones and chickpea is grown solely on
residual soil water in the relatively highland areas of the country, there is no scientific
data that support the choice of the crops for the stated environments. Although
information is available about the drought response of the individual crops in the
12
literature, it is difficult to compare the 'true' performance of the species because of
environmental, experimental and technical variations during the experiments. Moreover,
little research has been conducted on grain legumes that can be used to compare the
actual performance of the different crops and their resource utilization efficiency under
water stress conditions and to identify the most appropriate environmental conditions
for each crop.
1.7. Objectives
The major thrust of this study was to compare resource utilization and productivity of
common bean, cowpea and chickpea under water-stress conditions in the field and to
characterize their growing environments. The specific objectives of the study were:
1. To analyse yield-limiting weather conditions, particularly rainfall, in the grain
legume growing ecoregions of Ethiopia to generate information useful for
agricultural decisions making,
2. To compare the resource capture and utilization efficiency of the crops under water
stress and well-watered conditions,
3. To determine and compare the influence of water deficit on growth, yield and yield
components of the three species,
4. To examine each species' physiological response to drought during reproductive
growth stages and
5. To evaluate the DSSAT grain legume crop simulation model m a semi-arid
environment.
13
CHAPTER2
Agroclimatic Potential of Selected Locations in Ethiopia: Analysis of Variability and
Onset of Rainfall, Probability of Dry Spells and Length of Growing Season
2.1. Introduction
Agriculture is always under the influence of different weather and climatic challenges.
The degree of influence, however, depends on space, time and the type of agricultural
commodity considered which make one weather element more important than another.
For example, rainfall is the most important weather element that affects crop production
in the semi-arid tropics (e.g. Virmani et al., 1980). The rainfall in these regions is limited,
variable in space and time, and unpredictable (Stewart and Hash, 1982; Sivakumar,
1992). The Eastern Horn of Africa is one of the regions where rainfall is extremely
variable and unpredictable (Beltrando and Camberlin 1993; Beltrando, 1990; Ogallo,
1988). For example, in Ethiopia the rainfall is highly variable in amount and distribution
both in space and time (NMSA, 1996).
Analysis of rainfall events in a short time scale is indispensable for agricultural decision
making because of the fact that the seasonal rainfall distribution in the semi-arid tropical
regions is variable and as a result recommendations based on annual totals are misleading
(Simane and Struick, 1993; Virmani et al. 1980). Therefore, the start and end of the rains
and their distribution (Stem et al. 1982a; Sivakumar, 1988), and the length, frequency and
probability of dry spells (Sivakumar, 1992) in the growing season are key questions to be
addressed in the planning and management of dryland agriculture.
Predicting the start of the growing season (onset of the rains) is the most risky business in
agriculture because of the variability of the rainfall from year to year, from season to
season and from region to region. A "false" start of the rainfall prompts the farmer to
plant his crop early in the face of long dry spells after emergence. In most cases, this
results in poor crop stand and/or complete crop failure. This situation is a common
experience in many semi-arid tropical regions like Ethiopia. The subject has been the
topic of many studies resulting in many different definitions. Some of them include:
(1) the first occasion with more than 20 mm rainfall in one or two days after a certain
selected date (Virmani, 1975),
14
(2) the first ten-day period (decade) with more than 25 mm, provided that rainfall in the
next decade exceeded half the potential evapotranspiration (Kowal and Knabe, 1972),
(3) the first occasion when the 7 day total rainfall exceeds 25 mm and includes at least 4
rainy days (Raman 1974),
(4) when rainfall is greater than 0.35 ETo (reference Evapotranspiration) (Houérou et al.,
1993),
(5) the first occasion after a selected date when the rainfall accumulated over 2 days is at
least 20 mm and when no dry spell (exceeding 10 days) occurs within the following
30 days (Stern et al. 1982a), and
(6) the first occasion after a selected date when the rainfall accumulated over 3 days is at
least 20 mm and no dry spell of length more than 7 days occurs within the following
30 days (Sivakumar, 1988).
Although the definitions used by the different authors do vary, they show the importance
of analysing the start or onset of the rainfall in a given region so as to determine the
potential and risks involved in either planting early or late in the season.
The end of the growing season is mainly dictated by stored soil water and its availability
to the crop after the rain stops. In line with this, Stern et al. (1982a) defined the end of the
season as the first date on which soil water is depleted. Simane and Struik (1993) used a
threshold value of 20 mm total soil water to signify the end of the growing season.
--
Analysis of historical rainfall data to give information on the onset of rains, length of
growing season, probability and frequency of dry spells is used as an input to assess
cropping potential and risks in a given region. For example, in West Africa, where the
rainfall is also variable, analysis of historical data has been used to assess the potential
and risk of crop production in the region (Sivakumar, 1992; 1991; 1988). Except for a
few studies by Simane and Struik (1993) and Simane et al. (1999) on some selected sites
using decade (10 days) rainfall data and some reports by National Meteorology Service
Agency (NMSA) using monthly data (e.g. NMSA, 1996), the rainfall patterns of Ethiopia
and its agricultural implication has not yet been studied in detail using daily rainfall data.
The objectives are, therefore: (1) to analyse the pattern and spatial variability of rainfall at
selected stations which are in the different agroecological zones of Ethiopia, (2) to
determine the length of the growing season and to investigate the length and probability
15
of dry spells during the growing season, and (3) to evaluate the risk of planting with the
first rains of the growing season at the various sites.
2.2. Methodology
2.2.1. Site description and data acquisition
Ten meteorological stations, which lie in the different parts of Ethiopia, were selected for
the study. The choice of the stations was based on data availability, distribution of grain
legume production and representativeness of agroecological settings in the country. Daily
rainfall and temperature data were obtained from the National Meteorology Service
Agency (NMSA). In order to fill some missing values and years, data was also collected
from research stations at the respective locations as well as from the Ministry of
Agriculture. The different stations used in the study, their geographical descriptions and
the database considered are presented in Table 2.1 and Fig. 2.1.
Table 2.1. Geographical description and rainfall database of ten stations used in the
Study.
Station Latitude Longitude Altitude Data base period Number Source*
("N) eE) (m) of years
Alemaya 9.26 41.01 1980 1979-2001 23 NMSA
Awassa 7.05 38.29 1750 1970-1999 30 NMSA
BahirDar 11.36 37.25 1770 1970-1999 30 NMSA
Bako 9.07 37.05 1650 1970-1999 30 NMSA
Bole (A.A) 9.02 38.45 2408 1970-1999 30 NMSA
Debre Ziet 8.44 38.57 1900 1970-1999 30 NMSNDZARC
Dire Dawa 9.36 41.51 1260 1970-1999 30 NMSA
Jijiga 9.20 42.47 1775 1970-1999 30 NMSAlSERP
Mekeie 13.30 39.29 2070 1970-1988,1991-1999 28 NMSA
Melkassa 8.24 39.19 1540 1977-1999 23 NMSAlEARO
* EARO= Ethiopian Agricultural Research Organization, SERP = South East Rangeland Project, DZARC = Debre Zeit
Agricultural Research Centre.
Soil data were obtained from a previous study (Eylachew, 1994) as well as onsite soil
profile description, and analysis of samples collected from some of the sites (Awassa,
Dire Dawa and Jijiga) at the National Soil Laboratory in Addis Ababa. Reference
evapotranspiration (ET0) at lO-day intervals was taken from the NMSA report. Daily ET0
values were obtained by interpolation. Crop evapotranspiration (ET) of common bean (95
days maturing), chickpea (100 days maturing) and cowpea (100 days maturing) was
calculated using crop coefficients (kc) obtained from AlIen et al. (1998) and ETo of the
respective sites.
16
+Mekele
+
Bahir Dar
+ + D.Dawa+++Bole Alemaya Jijiga
Bako
D2e+~ +Melkasa
+
Awassa
One Centimeter = 166 Km
............w.a=Q=bid~ $>&& iJ¥M U w P! ....... u_. __ OO.. aM
Km 200 400 600 800 1000'200
Figure 2.1. Map of Ethiopia showing the location of the meteorological station sites.
Daily kc values were determined using the following relationship as described in Allen et
al. (1998).
(2.1)
Where i = day number within the growing season, kei = crop coefficient on day i, Lstage=
length of the stage under consideration (days), ~::CLpreJ = sum of the length of all
previous stages (days), kCnext= crop coefficient of the next stage, kCprev= crop coefficient
of the previous stage.
2.2.2. Analysis
The rainfall data were analysed using INSTA T Climatic Guide (Stem and Knock, 1998).
This statistical package allows summarization of daily rainfall data and further processing
of the data to obtain the starting date of planting, water balance calculation, determination
17
of the end of season, calculation of dry spell probabilities and length of growing season,
etc., based on information provided by the user. It also allows fitting of gamma
distribution and Markov chain probability models. For the purpose of comparing regions
on long and equal period of time, a Markov chain model was fitted to generate 100 years
data for each site for the calculation of onset of rains, end of season, and dry spell
probabilities. The advantage of this method is described in detail in Stern et al. (1982b)
and Stern and Coe (1982).
The annual, monthly and decadal patterns of the rainfall were examined for each site. The
coefficient of variation (CV) was calculated as the ratio of the standard deviation to the
mean rainfall. The annual rainfall trend was analysed using 5-year moving average for
sites that have more than 25 years of data. The probabilities of getting a rainfall exceeding
0, 10 20, 30, 40, 50 100 and 150 mm were estimated for each of the 36 decades. Decade
refers to the lO-day averaging periods of each month (WMO, 1966).
"Potential planting date" was defined as the first occasion with more than 20 mm rainfall
in three days after a selected date. The onset of the rains, explained by Sivakumar (1988)
as the first occasion after a selected date when the rainfall accumulated in 3 consecutive
days is at least 20 mm and no dry spells of more than 7 days in the next 30 days, was used
here as a "successful planting date". This criterion is chosen as the successful planting
date because it takes into account the potential dry spells at least in the following 30 days
after planting as in contrast with the potential planting date. The risk of first planting (the
failure of planting with first rains) was, therefore, calculated relative to the successful
planting date.
The daily rainfall data was processed to give maximum dry spell lengths in the next 30-
day periods from a starting date. Probabilities of the maximum dry spell lengths
exceeding 5, 7, 10, 15 and 20 days over the next 30 days from the first decade of
February (just before the start of the short rain period) to the last decade of November
(end of the main growing season) were calculated to get an overview of the drought
conditions throughout the year. The length of the dry spells (5-20 days) was selected in
such a way that both drought sensitive and drought tolerant crops are considered in the
growing season. Conditional dry spells (conditional on that the day before planting is
rainy) were also calculated in order to see whether a break in dry spell affects the length
of the following dry spells. Maximum and conditional dry spells were also calculated
18
starting from the onset of rains (successful planting date) to examine the probability of
short and long dry periods during crop growth period.
A simple water balance calculation was conducted for each location using rainfall, ET0
(reference evapotranspiration) or ET (crop evapotranspiration) and soil water content at
saturation. The water balance was calculated as described by Stem et al. (1982a) as
follows:
Sn= Sn-I+ Pn - ET (2.2)
where Sn = soil water on day n, Sn-I= soil water accumulated on previous day, Pn =
rainfall on day n and ET = reference or crop evapotranspiration. Water holding capacity
of the soil at saturation (es) was determined from bulk density as described by Williams et
al. (1992)
es =0.93*(1-(pt/ps» (2.3)
where Pb is bulk density (mg m") and p, is soil particle density (mg m') which is taken
as 2.65 whenever measurements are not available. The calculated soil water balance was
used to define the end of the growing season as the first date on which the soil water
drops to 10 mm m-I (i.e. <5.2 mm m" available water) after a predetermined date.
Available soil water was calculated as the difference between water content at drained
upper limit (DUL) and permanent wiling point or lower limit (DLL). The soil water
properties of each site are shown in Appendix lA. Once the start and end of the season
are known, the length of the growing season was obtained by subtraction.
2.3. Results and Discussion
2.3.1. Annual rainfall
The mean annual rainfall ranged from 601 mm (Mekele) to 1436 mm (Bahir Dar) and
was highly variable from year to year and location to location (Table 2.2). The coefficient
of variation (CV) ranged from 14% (Bako, Awassa) to 43% (Jijiga). Except for Jijiga,
high rainfall variability was observed in the low annual rainfall areas which agree with
previous reports (Brown and Cocheme, 1969, Simane and Struik, 1993, NMSA, 1996).
However, the high CV in Jijiga implies that such generalizations could be misleading, as
the rainfall of a given region could be variable despite its annual amount. Table 2.2 also
showed that the percentage of years with rainfall above X+SD ranged from 10 (Jijiga) to
22 % (Alemaya) and that of below X-SD ranged from 3 (Jijiga) to 23% (Dire Dawa).
19
Table 2.2. Annual rainfall statistics of ten locations in the different ecoregions of Ethiopia
for the period 1970-2001.
Station X Min Max SD CV >X+SD <X-SD Mar.-May Jun.-Sept.
(mm) (mm) (mm) (mm) (%) (%) (%) rainfall rainfall
(%)a (%)b
Alemaya 791 569 1053 145 18 22 22 35 53
Awassa 965 725 1226 132 14 17 20 30 50
Bahir Dar 1436 900 2042 250 17 10 10 8 84
Bako 1207 964 1652 170 14 17 13 22 69
Bole 1081 817 1552 172 16 13 13 22 69
DebreZiet 845 580 1123 145 17 13 17 19 74
Dire Dawa 612 357 965 157 26 17 23 39 44
Jijiga 682 389 1825 292 43 10 3 39 48
Mekeie 601 293 918 140 23 14 14 15 81
Melkassa 789 512 1276 163 21 13 14 18 68
X= mean annual rainfall
Min= minimum recorded value of annual rainfall
Max = maximum recorded value of annual rainfall
SD= standard deviation
CV= coefficient of variability
>X+SD= percent of years with annual rainfall greater than X+SD
<X-SD = percent of years with annual rainfall less than X+SD
a short-rain period, b long-rain period
Analysis of annual rainfall (recorded at each station) using 5-year moving average for the
period of 30 years showed a trend of both wet and dry years at most of the sites studied
(Fig. 2.2). The rainfall at Bahir Dar was decreasing steadily from the 1970s to the middle
of the 1980s (dry years). Nevertheless, it increased slightly again from that period to the
middle of the 1990s (wet years). The rainfall at Bako followed the same trend after the
middle of the 1990s. The trend at Jijiga was decreasing from the 1970s to the early 1980s
- +_._-
but remained almost stable after this period. In Debre Zeit, the rainfall was steadily
decreasing from the late 1980s to the early 1990s. On the other hand, the rainfall at Dire
Dawa was slightly increasing starting from the early 1980s (Fig. 2.2). The rainfall was
almost stable at Awassa and Bole from the 1980s to early the 1990s. Except Bahir Dar
and Bako, the annual rainfall showed an increasing trend for all stations after the middle
of the 1990s. Such variation among regions in long-term annual rainfall fluctuations
suggest the need to make agricultural decisions region specific to exploit opportunities
and minimize risks in each region.
20
2000
---+- Aw assa _ Bahlr Dar Bako _ • _ • _ Debre Zelt
1800 .•• _ •• Dire Daw a •••.••• Bole _ •• _. Jijlga _ . + . _ Mekeie
1600
1400
-E 1200
.§.
1000
~
.5 :":::::'...__ ......
ca 800 ...... \ ' ........;,,-....J .. ..._., ,_ .. _,:_II:: "-........ ..- . .,. .................
, ......p" ...._•..,
600 -f- \ _...._ .....- _- _ •.•• :(; _- -- •••• ~._,.- ~-.~.-f-;ob :jo
-+.--- ...+.-..!: ..•. :t-_-:;-....+..-_~'""'!f""""" -:-+c:.p.t::_ .>: _. -. _."=. - _./_."
400
200
0
1970 1975 1980 1985 1990 1995 2000
Year
Figure 2.2. Trends of annual rainfall using 5-year moving average analysis in eight stations
in Ethiopia for the year 1970-2000.
2.3.2. Rainfall distribution
The rainfall is characterized by its seasonality in all the locations studied (Appendix
IB&C). The pattern of the rainfall in Bahir Dar, Bako and Awassa is unimodal. The rest
of the locations studied have a bimodal or semi-bimodal pattern in which the rainfall has
a small peak in April/May and maximum peak in August. About 50-84% of the total
rainfall was received within four months (June-September) in all the locations except Dire
Dawa and Jijiga (Table 2.2). In the true bimodal rainfall areas in the eastern part of the
country (Alemaya, Dire Dawa, Jijiga), the first rainy season, which constitutes 35-39% of
the annual rainfall, extends from March to May while the second extends from July to
September (Table 2.2, Appendix 1C). The first (small) rain season in the bimodal regions
is very short and it is highly characterized by inter-annual variation (Simane and Struik
1994, NMSA, 1996). The period when P exceeds 50% ET0 in the main (big) rain season
extends from decade 19 to 29 in Alemaya, 9 to 29 in Awassa, 15 to 29 in Bahir Dar, 12 to
28 in Bako, 14 to 28 in Bole, 17 to 27 in Debre Zeit, 21 to 25 in Dire Dawa, 19 to 27 in
Jijiga, 18 to 25 in Mekele and 17 to 26 in Melkassa. Awassa and Bako have the longest
rainy periods followed by Bahir Dar and Bole (Appendix 1B&C). Similar to previous
reports (Simane 1990, Kassam, 1977; Kowal and Kassam, 1978), the relation between P
21
and 50% ET0 in the present study indicated a lower chance of false start of the rainfall
once the rainfall exceededhalf of the ET0 in the season.
Seasonal and annual rainfall variations in Ethiopia are results of the macro-scale pressure
systems and monsoon flows which are related to the changes in the pressure systems
(Haile, 1986; Beltrando and Camberlin, 1993; NMSA, 1996). The most important
weather systems that cause rain over Ethiopia include Sub-Tropical Jet (STJ), Inter
Tropical Convergence Zone (ITCZ), Red Sea Convergence Zone (RSCZ), Tropical
Easterly Jet (TEJ) and Somalia Jet (SJ) (NMSA, 1996). The spatial variation of the
rainfall is, thus, influenced by the changes in the intensity, position, and direction of
movement of these rain-producing systems over the country (Taddesse, 2000). Moreover,
the spatial distribution of rainfall in Ethiopia is significantly influenced by topography
(NMSA, 1996; Taddesse, 2000). STJ, ITCZ, RSCZ, TEJ and the SJ cause rainfall in the
bimodal and semi-bimodal areas. On the other hand, the rainfall in the unimodal areas is
mainly the result of the movement of the ITCZ though some influences of the other
weather systems still exist (NMSA, 1996).
2.3.3. Rainfall and evapotranspiration
The periods when rainfall (P) exceeds reference evapotranspiration (ET0) at 100, 50 and
35% are shown in Appendix 1B&C for the 10 locations. The period when P exceeds 35 %
ET0 is considered as the minimum water requirement for start of growing season
(Houérou et al., 1993). The maximum number of decades when P exceeds 35% ETo for
three consecutive decades during the main rain season are 11 in Alemaya, 22 in Awassa,
15 in Bahir Dar, 19 in Bako, 21 in Bole, 11 in Debre Zeit, 7 in Dire Dawa and Mekeie, 8
in Jijiga, and 10 in Melkassa (Appendix 1B&C). Comparison of monthly Pand 100%ETo
indicated that in some regions the rainfall exceeded the evaporative demand of the sites
for a period of as long as 4.7 months (Bako) whereas in other regions the rainfall could
not meet the evaporative demand at all (Dire Dawa, Jijiga). Regions like Alemaya,
Awassa, Bahir Dar, Bole, Debre Zeit, Mekeie and Melkassa satisfy their evaporative
demand for periods of at least 2 to 3 months. This period is the period of soil water
accumulation in the respective sites. According to Troll's (1965) climatic classification,
locations where rainfall exceeds reference evapotraspiration for 2 to 4.5 and 4.5 to 7
consecutive months are classified as dry semi-arid and wet-dry semi-arid, respectively.
Areas where P exceeds ET0 for a period of less than 2 months are classified as arid
22
tropics. Accordingly, Alemaya, Awassa, Bahir Dar, Bole, Debre Zeit, Mekeie and
Melkassa are classified as dry semi-arid whereas Bako is classified as wet-dry semi-arid
and Dire Dawa and Jijiga are classified as arid tropics.
2.3.4. Soil water
Results for the annual water balance studies using the long-term daily rainfall data are
shown in Fig. 2.3. The maximum water stored in the soil during the main rain season
ranged from 13 mm (Dire Dawa) to 176 mm (Bole). The period when soil water storage
remained above 50 mm was the longest at Bako (16 decades) followed by Bahir Dar (13
decades), Bole (12 decades), Debre Zeit (10 decades), Melkassa (9 decades), Mekeie (6
decades) and Alemaya (4 decades). The lowest soil water at Dire Dawa during the main
rain season was mainly due to its low annual rainfall, high evaporative demand and low
water holding capacity of the soil due to its sandy nature. The lower value at Jijiga is
predominantly associated with its high evapotraspiration and erratic nature of the rainfall.
In general, the results indicate the spatial variability of possible soil water accumulation
as influenced by the rainfall, soil characteristics, and the evaporative demand of a given
site (Huda, et al. 1990, Simane, 1990). The crop evapotranspiration tested in each site
showed that soil water accumulation was influenced by the type of crop grown and its
growing length (Fig. 3). Since the kc is less than one for most of the growing period, the
lower crop evapotraspiration resulted in a longer period of soil water accumulation than
when using the reference evapotraspiration at all of the sites.
2.3.5. Dependability of rainfall
In most semi-arid regions, the start, end and continuity of the rainfall are not reliable.
Therefore, information on probability of rainfall exceeding certain threshold values in a
given period is more important than the average rainfall. The probabilities of receiving
rainfall exceeding 0, 10,20,30,40,50, 100 and 150 mm per decade are shown in Fig. 2.4
for the 10 locations. The different rainfall values can be used as a threshold level for
different crops with different maturity, drought tolerance and water logging resistance as
well as for different soil types with different water holding capacities (Simane and Struik,
1993). In all the locations studied, the dependable rainfall (rainfall at 80% probability) is
higher, and the season is longer at low than at high rainfall thresholds. However, the
dependable rainfall is lower and the season is shorter in the bimodal rainfall areas
(Alemaya, Dire Dawa, Jijiga) even at lower rainfall thresholds when compared to the
23
200 200
Alemaya Awasa
E 150 ....... BN (201) 150 ....... BN (87)
.§. _. _. _CHP(201) _ . _ . _CHP( 245)
s...
COP(201) ____ COP (87)
---ETe
II 100 l 100 --_ETe
3::
e
ti) 50 50
o 0 +-~~~~--~~~--~~--~_r~~~
31 61 91 121 151 181211241 271 301 331361 31 61 91 121 151 181 211 241 271 301 331 361
200 200 ,-----------------~
Bahir Oar Bake
E 150 BN (118)
.§. ....... BN (152)
150 _ . _ . _CHP (245)
... _ . _ . _CHP(245) ____ COP(118)
! ____ COP(152)100 100 ---ETe
II ---ETe
3::
~ 50 50
o 0
31 61 91 121151 181 211241271 301 331361 31 61 91 121 151 181211 241 271 301 331 361
200 200 ,----------
Bole Debre Ziet
E 150 _. _. _CHP(245) 150 ....... BN (160)
.§.
--_ETe _ . _ . _CHP (245)...
! - - _ - COP (160)100
II 100 ---ETe
3::
~ 50
50
o
31 61 91 121 151181 211241 271301 331361 0
31 61 91 121 151 181211 241271 301331 361
200 ,------------------, 200.,----
Dire Dawa Jijiga
Ê 150 ....... BN (206) 150 BN (216)
.§. _ . _ . _CHP (206) _._._CHP(216)
... ____ COP (206) ____ COP(216)! ___ ETe100 ---ETe 100
II
3::
~ 50 50
o 0 ~~~~~~--~~~~_r--~~~~+
31 61 91 121 151 181211241271 301 331361 31 61 91 121 151 181211 241 271 301 331 361
200 200 ,---------------------,
Mekeie Melkasa
Ê 150 ....... BN (185) 150 BN(183)
_ .. _.CHP(185) ____ COP (183)
..§... _ . _ . _CHP (227)
s ____ COP(185)
---ETe
100 100
II ---ETe
3::
0 50 50
ti)
0
31 61 91 121 151 181211241271301 331361 31 61 91 121 151 181 211 241 271301 331 361
DOY DOY
Figure 2.3. Seasonal soil water balance of 10 stations in Ethiopia using reference
Evapotranspiration, ETo (site water use) and maximum crop evapotranspiration (crop water
use). Figures in parenthesis on the legend are days of planting (DOY) for the respective crops as
practiced by the farmers. BN = beans, CHP = chickpea, COP = cowpea.
24
unimodal and semi-bimodal regions. Among the unimodal rainfall areas, the period of
dependable rainfall at Mekeie was very short (80 days). The length of the dependable
rainfall in Awassa was exceptionally long at 0 and 10 mm thresholds (160-240 days) but
very short at 20 mm threshold (60 days).
Investigation of the daily rainfall data shows a higher number of rainy days with low
intensity of rainfall in the region. This nature of the rainfall may have significant impact
on the agricultural activities of the region because of its high infiltration into the soil as
well as its uniform distribution during the crop-growing season. The probability of
receiving a decadal rainfall exceeding 100 and 150mm reached 78 and 40%, respectively
for three consecutive decades (21-23) at Bahir Dar. These high rainfall conditions
necessitate the need to design alternative soil conservation practices and run-off
protection strategies particularly on soils with low water holding capacity and also to
design better drainage system in areas where the soil has high water holding capacity.
There is also a 16 to 30% chance of receiving 100mm rainfall in a decade at Debre Zeit,
Bole, Mekele and Melkassa during August. This high rainfall is problematic especially at
Bole and Debre Zeit where the soil has Vertic properties (high clay content) leading to
water logging conditions that hinder agricultural activities such as sowing, weeding,
fertilizer application etc., and also affects crop growth and development. The chance of
receiving rainfall exceeding 150 mm in a decade is very low in all regions except Bahir
Dar.
There is a good chance of getting high rainfall (100 mm per decade) in Mekeie during the
month of August such that a water harvesting technique can be practiced to capture the
rainfall, which can then be used later in the season to lengthen the growing period.
According to the study of Rockstrom and Falkenmark (2000), on average 30-40% of
seasonal rainfall is lost as runoff and deep percolation in many semi-arid cropping
systems of sub-Saharan-Africa. Harvesting and storing this water in reservoirs could help
in prolonging the crop growth period and also averting the effect of dry spells if applied
as supplemental irrigation (Barron et al., 1999; Rockstrëm, 2001).
25
100 100
Awassa;""'
80 80 ,r'
~ 60 lj
~ 60 .,' " ",
:ii 'J'
nl 40
.ti 40
0
ct 20 20
0 0
6 11 16 21 26 31 36 6 11 16 21 26 31 36
100
Bahir Dar 100
80 Bako•.•••.• Orrm 80
~
:>=- 60 60 .' ~, .J,
i 40 , 40 ,
.ti \ .0 '",
ct 20 ,; 20
0 0
6 11 16 21 26 31 36 6 11 16 21 26 31 36
100 100
80 80
"~Cl' 60 60
~
i 40
.ti 40
0
ct 20 20
"
0 0
6 11 16 21 26 31 36 6 11 16 21 26 31 36
100 100
Dire!lawa ..',80 , '.~, 80
~ 60
~ 60
i.. 40.ti 400
D. 20 20
0 0
6 11 16 21 26 31 36 6. 11 16 21 26 31 36
100 100
z 80 80s 60 60
i
.ti
0.. 40 40
D.
20 20
0 0
6 11 16 21 26 31 36 6 11 16 21 26 31 36
Time (decade) Time (decade)
Figure 2.4. Probabilities of receiving rainfall exceeding 0, 10, 20, 30, 40, 50, 100 and 150 mm
per decade at ten locations in Ethiopia.
26
2.3.6. Length of dry spells
2.3.6.1. Probability of dry spells computed on a calendar day basis
The probabilities of maximum dry spells exceeding 5, 7, 10, 15 and 20 days within a 30
day period after a specified starting date at the 10 locations is shown in Fig. 2.5. Dates are
presented from the beginning of February to the last decade of November to provide a
quick overview of the drought risks during the year at each location. In the bimodal
rainfall areas (Alemaya, Dire Dawa, Jijiga), the probability of maximum dry spell lengths
of 7 days or more never fall below 60% during the first rainy season (DOY 60-151).
Conditional dry spell probabilities (Appendix ID) also showed the same trend but had
lower values than the maximum probabilities indicating the influence of early rain on
shortening the length of subsequent dry spells.
The probability of dry spells of 5 days or more remained above 30% at Awassa during the
main rain season. On the other hand, the maximum and conditional probabilities
exceeding 15 and 20 days were below 5% from March to late September (DOY 60-262)
at the same site. The maximum dry spell probabilities decrease rapidly after DOY 180
(Jun 28) in Alemaya, Dire Dawa and Jijiga, DOY 130 (May 9) in Awassa, DOY 122
(May 1) in Bahir Dar, DOY 150 (May 29) in Debre Zeit, Mekeie and Melkassa. This
shows that the period after the indicated dates in each region is the period when there is
minimum risk to the emergence, establishment and subsequent growth of annual crops.
On the contrary, the probability of longer dry spell lengths (>20 days) increased rapidly
after DOY 256 (September 12) in Alemaya, DOY 272 (September 28) in Awassa, Bahir
Dar and Bako, DOY 260 (September 16) in Bole, DOY 252 (September 8) in Dire Dawa,
Debre Zeit and Melkassa, and DOY 232 (August 19) in Mekele. In all of the regions
studied except Awassa, the maximum and conditional dry spells with a length of 10 days
or more are above 60% after end of September (DOY 274) suggesting that standing crops
after this time will face increasingly greater risk of water shortage, particularly in areas
where the soil water holding capacity is low.
2.3.6.2. Dry spells computed on crop calendar basis
Because of the changing nature of sowing dates with the rainfall distribution of each year,
computations of dry spells on a calendar day basis have limited significance for specific
application in crop production (Sivakumar, 1992). Therefore, it is necessary to calculate
27
80
60
40
li. __ ... _ ..... \
20 20
o o+-~~~~~~~~ ..~~~~~~
o 30 60 90 120 150 180 210 240 270 300 330 360 0 30 60 90 120 150 180210 240 270 300330 360
100 100 ,--~~--=:,-- -,.....,.....-=...--,
~ 60
:.i,i __ 5
-.g.. 40 ....... 7__ 10
Go
20 ~__1_5 20
o 30 60 90 120 150 180 210 240 270 300 330 360 o 30 60 90 120 150 180 210 240 270 300 330 360
80
~
~ 60
.~a 40
o...
Go
20
o 30 60 90 120 150 180 210 240 270 300 330 360 o 30 60 90 120 150 180210 240270 300330 360
80 80
~ 60 60
:_igi 40 40
o...
Go
20 20
o+-~~~~~~~~~~~~~_J
o 30 60 90 120 150 180 210 240 270 300 330 360 o 30 60 90 120 150 180210240270 300 330 360
--. 80ot
o O+-~~--~~~~~~~~~~~~
o 30 60 90 120 150 180 210 240 270 300 330 360 0 30 60 90 120 150 180210240270 300330360
Time (DOY) Time (COY)
Figure 2.5. Probabilities of maximum dry spells exceeding 5, 7, 10, 15 and 20 days within 30
days after starting date (DOY 32) at 10 locations in Ethiopia.
28
the probabilities of dry spells after onset (successful planting dates) are established. The
probabilities of dry spells after onset of rain were computed for each station and are
shown in Fig. 2.6. The probability of long dry spells (15 and 20 days) increase rapidly 60
days after sowing (DAS) in Alemaya, 200 DAS in Awassa, 140 DAS in Bahir Dar, 160
DAS in Bako, 110DAS in Bole, 100 DAS in Debre Zeit, 50 DAS in Dire Dawa, 40 DAS
in Jijiga, 60 DAS in Mekeie, and 70 DAS in Melkassa. Although dry spell lengths of 15
days or more commence as early as 100 and 110DAS in Debre Zeit and Bole
respectively, standing crops will not be affected easily because of high water stored in the
soil during the peak rain months (Fig. 2.3). The soil of these regions is classified as
Vertisol and has high water holding capacity as compared to the rest of the regions
studied (Appendix lA). Bahir Dar and Bako also have high soil water storage capacity
that can prolong the growing period by a significant length.
Dry spell analysis helps in identifying the type of crop (short, medium or long maturing,
drought tolerant or susceptible, etc.) and management practices (supplemental irrigation,
fertilizer and insecticide application, etc.) (Sivakumar, 1992; Sirnane and Struik, 1993)
that is appropriate to the respective regions. For example, it is necessary to choose a
terminal drought tolerant variety if one wants to plant a crop variety with a maturity
length of more than 100 days at Mekeie and 120 days at Alemaya and Melkassa as longer
dry spells prevail after this period at these sites. Besides serving as a tool in the choice of
a suitable crop variety for a given site, this type of dry spell analysis could also be used as
a guide for breeding varieties of various maturity durations for the different locations
(Sivakumar, 1992).
An example of this application is shown for Dire Dawa for which crop varieties that
mature within 70 days are required whereas varieties which have a maturity period of 200
days or more are required for Awassa and Bako in order to fully utilize the regions'
resources. As pointed out by Sivakumar (1992), dry spell analysis could also be used as a
tool to study mismatches of phenology of a new crop to the rainfall regime of a given area
as well as to answer 'what if questions in decision-making. For example, one may ask
"what is the chance of dry spells longer than 15 days after 80 days of planting my crop?"
and get answers from the dry spell analysis like the one shown in Fig. 2.6. The length of
dry spell probabilities closely follows the annual rainfall amount in all the locations
studied except Jijiga.
29
100 100
__ Pr>=5 Awassa
80 .......Pr>=7
z _--_-_6_-- PPr>=10
80
r>=15
60 ---0-- Pr>=20
~ 60:s
lIS
..Q 40 40
s0: 20 20
0 0 ."
0 30 60 90 120 150 180 210 240 270 0 30 60 90 120 150 180 210 240 270
100 100
80 80
~ 60
~ 60:s
lIS 40
..Q 40
ae.. 20 20
0 0
0 30 60 90 120 150 180 210 240 270 0 30 60 90 120 150 180 210 240 270
100 100
80 80
~ 60
:~
60
s
lIS 40
..Q 40
0
Ir. 20 20
0 0
0 30 60 90 120 150 180 210 240 270 0 30 60 90 120 150 180 210 240 270
100 100
80 80
~ 60
~ 60:s
lIS 40
..Q 40
0
Ir. 20 20
0 0
0 30 60 90 120 150 180 210 240 270 0 30 60 90 120 150 180 210 240 270
100
100
80 80
~ 60
~ 60:s
.lIS 40..Q 400.
a.. 20 20
0 0
0 30 60 90 120 150 180 210 240 270 0 30 60 90 120 150 180 210 240 270
Days after onset Days after onset
Figure 2.6. Probability of dry spells exceeding 5, 7, 10, 15, and 20 days after onset (sowing)
at ten locations in Ethiopia.
30
In general, analysis of dry spells on calendar day basis provides an overview on the
frequency and distribution of dry spells in a given region. This is important to understand
the probability of different length of dry spells for the specific regions so as to make
appropriate recommendations such as time of planting, and also to match crop phenology
to the period of water availability. Moreover, the dry spell analysis after sowing dates are
established is an important step in addressing issues like: Which crop/variety is best for
which site? What is the probability of long dry spells on a certain day after planting the
crop in question? Is supplemental irrigation (if there is that option) needed and when
should it be applied? What kind of crop management should be practiced and during
which period?
2.3.7. Length of growing season
The average potential planting dates (PPD), risk of planting with early rains, the average
successful planting dates (SPD), dates of end of season (ES) and length of growing
season (LOS) calculated for the rainy seasons are shown in Table 2.3. Assuming normal
distribution of the 100 years data, the average SPD ranged from DOY 95 (April 4) in
Awassa to DOY 216 (August 3) in Jijiga. Thus, average SPD may start as early as April
and could be delayed until August 3 in the regions studied. The 20, 50 and 80 percentiles
of SPD, ES and LOS are shown in Table 2.4. The median successful planting dates (50%)
are similar to the average successful planting dates shown in Table 2.3 indicating the
normal distribution of the data. Compared to potential planting dates (PPD), average
~~ - --
SPDs were delayed by a range of 2 (Bako) t035 (Debre Zeit) days. Cumulative
probability curves shown in Fig. 2.7 were also constructed for the potential and successful
planting dates. The choice of any value on the cumulative probability curves depends on
the level of risk to be taken on the one hand and the length of the growing season needed
on the other. This in turn depends on the crop's drought sensitivity (particularly during
the early season) and its length of maturity (duration oftotal growth period).
The risk of potential planting dates was calculated with reference to the successful
planting dates. The risk ranged from 6% (Bako) to 68% (Debre Zeit) indicating high
chance of "false start" of the rainfall at Debre Zeit followed by Jijiga and Bole (Table
2.3). The 95% confidence interval shows that the risk of "false start" can vary from 2%
(Bako) to 80% (Jijiga) (Table 2.3).
31
Table 2.3. Average potential (PPD) and successful (SPD) planting dates, dates of end of season calculated using ETo of site (ESo) and ET of crops
planted on day numbers indicated in parenthesis (ESe), length of growing season (LGS) and risk of first planting for the first (if any) and the
second rainy seasons for ten locations inEthiopia.
Location First rainy season Second rainy season
PPD" SPD Risk of ESo ESc 6 LGS PPD" SPD Risk of ESo ESc 6 LGS
(DOY) (DOY) PPD (%)* (DOY) (DOY) (days) (DOY) (DOY) PPD (%)* (DOY) (DOY) (days)
ESo ESc ESo ESc
Alemaya 87(Marl) 121 52(43-64) 133 145 (97) - - 199(Jun21) 203 14(7-20) 304 311(197) 101 114
Awassa 82 (Marl) 95 29(22-38) 304 315 (115) 209 200
BahirDar 142 (MayI) 152 34 (26-46) 319 319 (152) 167 167
Bako 116 (Aprl) 118 6 (2-13) 331 331 (109) 213 222
Bole 112 (Aprl ) 139 61(51-71) 321 321 (245) 182 76
Debre Zeit 119 (Aprl) 154 68 (58-77) 310 320 (254) 156 66
Dire Dawa 81 (Marl) 141 74 (66-84) 199 (Jun21) 208 32 (24-44) 285 283 (207) 77 76
Jijiga 83 (Marl) 135 76 (68-85) 190 (Jun21) 216 63 (61-80) 215 268 (191) 4 77
MekeIe 181(Junl) 185 19 (10-27) 277 280 (185) 92 95
Melkassa 170(Junl) 183 46 (37-58) 285 297 (183) 102 114
a dates inparenthesis refer to starting dates of rainy season, * values in parenthesis are 95% confidence intervals of risk of potential planting date.
b numbers in parenthesis refer to dates of planning (DOY) of a 95-100 day maturing chickpea (Bole and Debre Zeit) and bean (the rest eight sites) following the local
farmers practice.
32
In some of the study areas like Bako, Alemaya and Mekele, the chance of dry spells
of 7 days or more was minimum after the rain had started. This is shown by the
narrow gap between PPD and SPD cumulative probability curves in Fig. 2.7.
Table 2.4. Successful planting dates (SPD), dates of end of season (ES) and length of
growing season (LGS) at 20, 50 and 80 percentiles expressed in day of year (DOY).
Station SPD ES LGS
20% 50% 80% 20% 50% 80% 20% 50% 80%
Alemaya 189 201 217 290 305 317 81 100 121
Awassa 70 87 113 278 314 326 177 213 247
Bahir Dar 139 152 165 312 319 326 154 168 181
Bako 102 118 133 323 330 340 194 214 230
Bole 110 143 165 316 321 326 157 179 213
Debre Zeit 133 160 176 305 310 316 133 151 180
Dire Dawa 191 206 220 272 287 300 63 82 lOO
Jijiga 196 216 233 214 215 230 4 37
Mekele 177 185 194 274 277 282 82 93 104
Melkassa 168 183 198 275 291 302 85 109 128
The analysis made to assess the potential of the first rain season (March-May) for
crop production in the bimodal rainfall areas (Alemaya, Dire Dawa, Jijiga) is shown
in Fig. 2.7. It was found that the rain during this period started early so that potential
planting was possible during the last week of March. However, the rain season ends
before successful planting dates were established. For example, the average risk of
potential planting was 52% in Alemaya, 74% in Dire Dawa and 76% in Jijiga (Table
2.3). The 95% confidence interval ranged from 43-85% for the same period. This
shows that this season is very risky and is not suitable for crop production purposes
unless crops are selected that can resist long and frequent dry spells. If this season
has to be used for production purpose in areas like Alemaya, crops are required that
can resist a dry spell length of more than one month (June and early July) and
continue growth during the second season without severe damage. Sorghum is the
only crop currently serving this purpose at Alemaya. In some areas like Jijiga, where
livestock rearing is predominant, the rain during the first season may be used for
growing pasture and grass for animal feed.
The end of the season was calculated using ET 0 of the respective sites and ET of the
pulse crops that are being grown at each site. The average dates of end of season are
shown in Table 2.3, and the cumulative probability curves are shown in Fig. 2.7.
33
100~ ~~ ~7- ~~ 100r---------~--_=== __ --------------I'"
Alemaya i .. Awassa••g" It-
' 80 .C"0" 80 I :'~•.0. ,I;>.' " ____ PPD
~ ;
,
, , _____ SPO
:ii 60 60I e
c I •
i ••••••• ESo
.aa , .
~ 40 , .. 40 ,
a.
Ë j :' 20
::I 20 .~
(.) .1.. - ,,
o +-~~~--r-~--~~--~--~'~--~-4
30 60 90 120 150 180 210 240 270 300 330 360 30 60 90 120 150 180 210 240 270 300 330 360
100 100~--------~~--------------------~ .'
~ Bahlr Dar Bako ,
~ 80 ,: 80
:ii I60
~ 60o
li. 40 40
8Ë 20 20
,
O+-~--~-- ___r--._~--._~--~~'_r~
30 60 90 120 150 180 210 240 270 300 330 360 30 60 90 120 150 180 210 240 270 300 330 360
100r- ~~--------------~~ 100.- ~~ __ ----------~--_,
Bole ,I
~ 80 80
~
:ii 60 60c ,.o.aa. 40 40 ,a.
Ë 20 20
o::I
I
o+-~--~~ __~--~~--~~~~·Ir__r~
30 60 90 120 150 180 210 240 270 300 330 360 30 60 90 120 150 180 210 240 270 300 330 360
100~-------,~.------------~9------ ..--~
100.- ~I----------~~~----~,~--~_. .: JIJlga;, JIj{.. •
80 80" Dire Dawa :~ : I •••1;>""/
" ..,t I.S··t;) I60 I • 60, 13-_.13 I " I ,I
,I 0'· •40 " ,l ,,I ~40 ,,I ...I
i : / I ,
E 20 ,~,; I 20
::I
U ..
, : ,
s3 , .,.
O+-~~~~ __~~r__T·~·-·~··-·~--~--r_~
30 60 90 120 150 180 210 240 270 300 330 360 30 60 90 120 150 180 210 240 270 300 330 360
100.-----------------7.Z~----~~----_, 100r-----------------~v_--------~--~ .'
Mekeie Melkassa
80 , 75
~
:ii 60
~. , 50o 40a..
Ë 20 25 ,
o::I •• J
O+-~--~--._~~._--~_r.--.-.. ~# ~--~~
30 60 90 120 150 180 210 240 270 300 330 360 30 60 90 120 150 180 210 240 270 300 330 360
Time (DOY) Time (DOY)
Figure 2.7. Cumulative probabilities of potential (PPD) and successful (SPD) planting
dates and end of season (ESo) of the growing seasons at 10 locations in Ethiopia. PPD1,
PPD2 and SPD1, SPD2 at the bimodal stations refer to potential and successful planting
dates for the first and second rainy seasons, respectively.
34
Using crop ET, the average date of end of the growing season ranged from DOY 288
(September 26) at Jijiga to DOY 331 (November 26) at Bako (Table 2.3). On the
other hand, the end of the growing season ranged from DOY 263 (September 19) at
Jijiga to DOY 331 (November 26) at Bako using ETo of the respective sites (Table
2.3). In areas like Alemaya, Awassa, Debre Zeit, Melkassa as well as Jijiga, the ET
resulted in longer growing period as compared to ET0 of the respective sites
suggesting that growing of grain legumes can extend the growing season of the sites
by reducing soil surface evaporation (lower actual water use than potential demand).
On the other hand, a 100 days maturing legume could not utilize the available water in
the high rainfall areas like Bako and hence soil water depletion (end of season) is
determined by evaporative demand of site.
Very strong association was found between onset of the rains and LGS and LGS and
ESo (Table 2.5). As expected, early plantings and late end of season result in longer
growing season. Using the sites' evaporative demand for calculating dates of end of
season, the median LGS ranged from 4 days (Jijiga) to 213 days (Bako) (Table 2.4).
When crop ET was used during the growing season, the LGS was found to range from
66 days (Debre Zeit) to 222 days (Bako). The ET of common bean (95 days
maturing), chickpea and cowpea (100 days maturing) were used in this study
assuming that the crops were planted on dates currently practiced by the local farming
community. It was found that in some of the locations studied (Dire Dawa, Jijiga,
Debre Zeit, Bole), the crops were prone to early and late season drought in areas like
Jijiga and to terminal drought in Dire Dawa, Debre Zeit, and Bole due to the current
planting practice and wrong choice of cultivars in terms of maturity length.
Therefore, it is advisable to choose the appropriate crop variety that can best fit the
actual length of the growing season in areas like Jijiga and Dire Dawa whereas in
areas like Bole and Debre Zeit, where heavy water logging conditions prevent early
sowing, it is essential to choose chickpea varieties which can mature within 70-80
days or varieties that can tolerate terminal droughts.
Awassa, Bako, Bole and Bahir Dar have one long growing period (longer than 5
months) which is suitable for late maturing crops. Grain legume crops, which usually
mature in less than 150 days, would not fully utilize the soil water in these regions.
Therefore, unless intercropped with other late maturing cereal crops like maize and
35
sorghum, sole cropping of grain legumes should not be recommended for these sites.
On the other hand, Alemaya and Melkassa were found suitable for grain legumes that
mature within 110days and Mekeie was found appropriate for grain legumes crops
that mature between 90 and 95 days.
Table 2.5. Correlations between successful planting date (SPD), date of end of season
(ESo) and length of growing season (LGS).
Station SPD vs ESo SPDvsLGS ESo vsLGS
Alemaya -0.05 -0.75 0.70
Awassa 0.02 -0.71 0.69
BahirDar 0.02 -0.85 0.50
Bako -0.02 -0.88 0.61
Bole -0.16 -0.98 0.37
Debre Zeit -0.09 -0.96 0.37
Dire Dawa -0.02 -0.76 0.67
Jijiga -0.10 -0.85 0.61
Mekeie -0.16 -0.89 0.59
Melkassa -0.24 -0.67 0.88
The cumulative probability curves in Fig. 2.7 indicate that in 80% of the years, the
season ends before day 340 (December 5) in the high rainfall areas (Awassa, Bahir
Dar, Bako and Bole) and before day 317 (November 12) in the low to intermediate
rainfall areas (Alemaya, Debre Zeit, Melkassa, Dire Dawa, Mekeie and Jijiga).
2.4. Summary and Conclusion
Ethiopia's economy is highly dependent on subsistence agriculture. More than 85% of
the population is engaged in this sector of which the majority are involved in rainfed
crop production. Because of tremendous variability of the rainfall from year to year
and season to season, the country becomes vulnerable to recurrent droughts.
Therefore, analysis of historical rainfall data in conjunction with soil factors can be
used in assessing cropping potential and risks in different regions so as to make
appropriate recommendations for crop planning and disaster prevention schemes. The
role of such studies in planning agricultural development is indicated in many reports
(Dennet et al., 1984; Peacock and Sivakumar, 1986; Simane and Struik 1993; Stem et
al., 1982a, 1982b; Virmani et al., 1980).
The current study has revealed the existence of broad regional differences in water
supply. In some areas like Bahir Dar, Bako and Bole, management of excess wateris
the major concern unlike some other areas (Jijiga and Dire Dawa) where water supply
is limited and hence maximizing water use and water use efficiency are crucial. This
36
means that there is a need to adopt different management strategies for optimum
resource use for each region. The variability in mean annual rainfall between and
within regions and seasons means that farming practice recommendations should be
region and season specific.
Information on the length and frequency of dry spells is crucial in adjusting crop and
cropping practices to the environment and to some extent in adjusting the micro-
environment to the crop (e.g. irrigation, mulching and drainage). Grain legume crops
are more susceptible to drought than cereal crops like sorghum, millet and wheat in
the semi-arid tropics. Therefore, the information generated on probability of dry spells
in the present study is important to facilitate agricultural decision making to sustain
the production of grain legumes under these variable and fragile environments.
The onset of the rains is an important factor in determining agricultural activities such
as sowing, choice of crop and division of labour. The study indicated strong
association between onset of rains and length of growing season. As shown by the
strong negative correlations, early sowing resulted in longer growing season and vice
versa. Therefore, relatively longer duration varieties can be used when the rain starts
early whereas short maturing varieties are needed whenever the onset is delayed from
the expected period. Sivakumar (1988) reported similar results in West Africa. This
again implies the need to adjust tactical decisions following the onset of the rains in
the current Ethiopian rainfed crop production system. In a given rain season, false
start of rainfall is less likely in all the regions once the decade precipitation exceed
half of the decade evapotraspiration. Therefore, this relationship between rainfall and
reference evapotranspiration can be used as a general guide in planning sowing dates
at the respective locations.
The dependability of first rains varied across locations. In many of the locations
studied such as Bahir Dar, Bole, Debre Zeit, Dire Dawa, Jijiga and Melkassa, the
chance of false start is very high resulting in high risk for early planting. At the other
locations (Alemaya, Bako and Mekele), the delay in successful planting after the
potential planting date is very narrow and hence the risk of first planting was very low
suggesting the reliability of the season once the rain has started. Regional differences
were observed for the length of growing season. Bako, Awassa, Bole and Debre Zeit
have a long growing season that is suitable for crops which have a maturity length of
37
5 to 7 months. Alemaya, Melkassa and Mekele have an intermediate growing season
which can support a crop maturing within 3 to 3.5 months. Dire Dawa has a relatively
short growing season. There was not any successful start of the season in 75% the
years at Jijiga and the median length of the growing season was only 4 days.
Therefore, this region is unsuitable for crop production despite the current practice of
growing many crop species, which often fail.
In the current Ethiopian agriculture system and its rainfall condition, matching of crop
phenology to the water availability period seems to be the best option to promote
pulse crop production in the country. A similar conclusion was reached by Simane
and Struik (1993) with respect to wheat production in the same country. Selection of
crop species/varieties for the drought prone areas (Jijiga, Dire Dawa, Mekele) should
be done with great caution so that farmers could get reasonable yields even in dry
years. For such regions, species/varieties that are drought-tolerant and adaptable to the
erratic nature of the rainfall are deemed necessary. Moreover, developing an
appropriate land use system is imperative for utilization of a given site to its full
capacity. For example, Jijiga is more suitable for livestock than crop production
implying that many of the impacts of drought events in the different parts of the
country are results of not only natural calamities but also improper land use system.
This has led to deforestation (for the sake of charcoal making to bake daily bread),
bush encroachment, and finally degradation of natural resources, which in turn
affected the climate of the respective regions. Studies like this one are believed to
have immense contribution to develop suitable land use system in the country.
Moreover, the information generated from this study will have valuable contribution
in advising the farmer, extension agent, agronomist and other groups involved in crop
production.
38
CHAPTER3
Phenology, Growth and Dry Matter Allocation in Three Grain Legume Species
Grown Under Three Water Regimes in a Semi-Arid Environment
3.1. Introduction
Grain legumes are a major source of plant protein in the developing and developed world
(Duranti and Gius, 1997). Being the only source of protein for a number of poor farming ,
communities in the semi-arid tropics (Singh, 1997a), these crops are given the nickname,
'poor man's meat' (Duranti and Gius, 1997). In the semi-arid tropical regions, the crops
are traditionally grown under rainfed conditions in marginal environments, and their
growth and development is usually affected by drought, which can occur at any time
during the growing period (Adams et al., 1985; Rachie, 1985; Graham and Ranalli, 1997).
Phenology plays an important role in plant growth and productivity. Under drought
conditions, it affects plant productivity through various simple or complex pathways
(Blum, 1996).Drought may hasten or delay phenological periods depending on the time it
occurs, its severity, rate of onset of stress and type of species involved (Blum, 1996). For
example, in wheat, mild water stress caused advanced flowering (Angus and Moncur,
1977) while severe water stress caused delayed flowering (Dwyer and Stewart, 1987).
Phenology is one of the major factors that control water use, duration of exposure to
stress, leaf area development and its duration, tissue juvenility, and also the degree of
stomatal response in plants (Blum, 1996; Turk and Hall, 1980a; Turk et al., 1980b).
Plants have a suite of morphological and physiological adaptations that allow them to
survive water stress (Monneveux and Belhassen, 1996). The degree of adaptation,
however, varies greatly within genera and species (Torrecillas et al., 1996). Patterns of
biomass allocation between different plant organs have been used to explain the response
of plants to variations in resource availability (Ninkovic, 2003). Allocation of
photosynthate among various plant parts is a mechanism by which plants modify their
growth in response to environmental conditions in a way that maximizes growth.
Therefore, when water supply is variable in the growing season, dry matter is partitioned
to plant parts depending on the time in the life cycle (Boyer, 1996). The redistribution of
assimilates accumulated during the vegetative and early reproductive periods to the seed
39
during the seed-filling period is considered as a potential source of yield stability in
terminal drought environments (Turner et al., 2001). Dry matter reallocation to grain has
been reported for a number of grain legumes including chickpea (Saxena, 1984; Singh,
1991; Leport et al., 1999), mungbean (Bushby and Lawn, 1992), groundnut (Wright et
al., 1991), and soybean (Westgate et al., 1989) under water deficit conditions. The
commonly observed rapid senescence and abscission of leaves in grain legumes under
water deficit is suggested to be a means of reallocating carbon and nitrogen from the
senescing leaves to the seed (Turner et al., 2001). Therefore, it is important to be able to
calculate some type of index of assimilate partitioning in order to relate yield and dry
matter allocation under water deficit conditions.
Information on pattern of growth and dry matter partitioning between various plant parts
is an essential step in the development of crop growth simulation models (Royo and
Blanco, 1999; Sheng and Hunt, 1999). Moreover, data on growth and its partitioning
would allow better interpretation of results within the context of processes and resource
exploitation (Williams et aI., 1996). However, such information is sparse for grain
legumes particularly under water deficit conditions in the field. The objective is,
therefore, to study the growth, phenology and dry matter allocation of common bean
(Phaseolus vulgaris L.), chickpea (Cicer arietinum L.) and cowpea (Vigna anguiculata
L.) grown under water stress and non-stress conditions in a semi-arid environment in the
field.
3.2. Materials and Methods
3.2.1. Field experiments
Three field experiments were conducted at the fruit farm and research centre of Alemaya
University in Dire Dawa, Ethiopia (latitude 9°6'N, longitude 41°8' E, altitude 1197 m
above sea level) during the periods from early December 2001 to late March 2002 (first
season), late March until the end of Jun 2002 (second season) and from mid-October
2002 to early February 2003 (third season). The station lies in the semi-arid belt of the
eastern rift valley escarpment with a long-term average rainfall of 612 mm. The soil is
classified as Eutric Regosol with a gentle slope (3-8%) (Amede, 1998). The texture and
structure of the topsoil (0-30 cm) are sandy loam and sub angular blocky, respectively.
The soil has an average pH (H20 1:2.5) of 8.52 and organic matter content of 1.18%
(Appendix 6D).
40
Seeds of common bean (ev. Roba-l ), Kabuli chickpea (cv. ICC-4958,) and cowpea (ev.
Black eye bean) were planted on 7 December 2001,27 March 2002 and 17 October 2002
for the first second and third seasons, respectively. All cultivars had semi-indeterminate
growth habit. Roba-I is an improved bean variety released by the Institute of Agricultural
Research, Ethiopia ten years ago. Blackeye bean is a well-adapted cowpea variety used as
a check by the lowland Pulse Improvement Research Program at Alemaya and Melkasa
research stations. ICC-4958 is a registered drought resistant chickpea cultivar (Saxena et
al., 1993) currently grown in Ethiopia.
Nitrogen and phosphorus fertilizers were applied to the soil before planting in the form of
urea and di-ammonium phosphate at a rate of 30 kg ha" each. Hand weeding was done
throughout the growing periods to .keep the plots free of weeds. Sumathion (20 ml ai/IO L
water) and Maneb (IOg ai/l OL water) were applied twice during each season to control
insect pests and fungal diseases, respectively.
3.2.2. Experimental design
The experiments had two water deficit treatments (referred here after as water stress
treatments) and a well-watered control treatment as shown in Table 3.1. The experimental
treatments, each replicated three times, were arranged in a randomized split plot design
using the water regimes as main plot and the crop species as sub-plot. The total
experimental area was 22.8 m x 40.2 m.
Table 3.1a. Soil water regimes applied in the experiments and the lowest available soil water
(ASW) maintained at a depth of 300-600 mm before irrigation in each water regime.
Water regime Stress period ** Minimum ASW at re-watering
Mid-season stress (MS) Flowering 23-25%
Late season stress (LS)* Pod filling until maturity 23-35%
Control (C) No water stress >60%
Table 3.1b. Duration of stress periods for the water stress treatments in each species during
the three seasons.
Water regime Species Seasons
2001/2002 2002 2002/2003
MS Beans 52-66 45-56 45-62
Chickpea 52-66 45-56 42-57
Cowpea 63-77 45-56 53-72
LS Beans 62 to maturity 58-71 62 to maturity
Chickpea 62 to maturity 58-71 62 to maturity
Cowpea 64 to maturity 58-71 62 to maturity
* Treatment received no water after the stress was induced unless rain occurred. ilj)AP~ days after planting.
41
Each sub-plot (4m x 6m) had 10 rows, and the inter- and intra-row spacing was 0.4 m and
0.1 m, respectively (25 plants m"), The 4 m length of the central four rows in each plot
were used for final yield determination whereas the other rows were used for destructive
measurements excluding the outer rows.
3.2.3. Irrigation schedule
Plots were irrigated (33.3 mm) immediately after planting to ensure uniform seedling
establishment. A measured amount of water from a 1000 m3 capacity water tank was applied
to each furrow using a 100 m long plastic hose. The tank was recharged each time from a
nearby well. Non-stressed treatment plots received irrigation whenever available soil water
at 30 cm depth (58 mm) reached 60-70%. Stress was imposed in the MS treatment by
withholding irrigation and rainfall until the available soil water depletion reached 23-25%
(Table 3.1). In the LS treatment, plots were not irrigated and also protected from rainfall for
the period from pod filling to maturity. A simple rain shelter, constructed on site from
transparent plastic (Gundle-plastall, South Africa) and wooden poles, was used to protect
stress plots from rainfall during the stress periods. The stress plots were covered with the
shelter only during an event of rainfall. Lateral movement of water between plots was
prevented by a polythene plastic sheet buried in the soil to a depth of 1.2 m. The soil water
content at 300-600 mm soil depth was monitored every day throughout the growing period
using Time Domain Reflectometery (TDR) (Soil Moisture Equipment Corp., CA, USA).
3.2.4. Experimental measurements
Leaf area was measured throughout the growing period destructively using a portable leaf
area meter (Model CI-202, CID, Inc., USA) from five randomly selected plants (0.2 m2 area)
in a plot at an interval of 10 days starting at 35, 17 and 20 days after planting (DAP) for the
first, second and third experiments, respectively until physiological maturity. Above ground
dry matter (ADM) was determined from the same five plants for the same interval of time.
The plant samples were separated into leaf, stem, pod and seed, and then dried in an oven for
72 hours at 60°C to determine mass of dry matter. Final harvest was done from March 1 to
April 6, May 30 to Jun 26 and January 13 to February 5 in the first, second and third
seasons, respectively, according to maturity dates of the crops in the different treatments.
42
3.2.4.1. Phenology
Calendar days and thermal time
Date of emergence, flowering, pod initiation and maturity were recorded whenever 50% of
plants in a plot show the character. Time to each phenological stage was determined starting
from the date of planting. Daily thermal time (h, °c d) was calculated as:
tT =(Tmax ~Tmin -t; )*M (3.1)
where Tmax and Tmin are daily maximum and minimum temperatures eC), T, is base
temperature eC) and .::1its time interval (day). The base temperatures used in the calculation
were 10 °c for beans (Guyer and Kramer, 1952; Hardwick, 1988b) and 8 -c for both
chickpea (Singh, 1991) and cowpea (Craufurd et al., 1997). The daily Tmax is made equal to
30°C if it is above this threshold (Cross and Zuber, 1972; Mauromicale et al., 1988). Thus,
the thermal time accumulated (GDD) for a given time interval (.::1tw) as calculated as:
GDD = t(T;mac +T;min - Tb) *M; (3.2)
;=1 . 2
3.2.4.2. Growth
3.2.4.2.1. Comparison of dry matter production
Because of ontogentic differences between the species, data on total above ground dry
matter (ADW) was compared after fitting the data with a Richards function. The Richards
Function (Richards, 1959) is expressed as:
a
y = (1+ (3.3)exp(b - CX)(I/d))
where y is In-transformed dry matter, x is time and a, b, c and d are constants. 'a' represents
the upper asymptote of the curve while 'd' controls inflection and its position on the curve.
Second or third degree polynomial functions were fitted to leaf dry matter (LDM) and stem
dry matter (SDM) while linear, logarithmic or exponential functions were fitted to pod dry
matter (pDM) depending on the water stress treatments. The curve fitting process was
conducted using Curve Expert Version 1.37 curve fitting system
(http://home.comcast.netl-curveexp~.
43
3.2.4.2.2. Specific leaf area (SLA)
Specific leaf area (SLA, m2 g-l or cm2 g-l) was calculated as explained by Hunt (1982) and
Causton and Venus (1981) as shown below:
SLA=~ (3.4)
LDM
The mean seasonal SLA of the species was calculated as the slope of the linear regression
between leaf area and leaf dry matter.
3.2.4.2.3. Leaf area duration (LAD)
Leaf area duration (LAD, days) was calculated as follows:
12
LADII_12 = f LAIdt (3.5)
II
where LAl is leaf area index calculated as the ratio of leaf area to the area of ground, and dt
is the change in time.
3.2.4.2. Dry matter allocation
The dry matter allocation among the shoot components was calculated by the method of
BOITellet al. (1989) as follows:
dWparl/dt
ARparl = / (3.6)
dWlolal dt
where AR is allocation ratio, dw is the change in dry mass and dt is the change in time. This
ratio enables one to assess the relative sink strength of the leaf, stem and pod by comparing
the growth rate of each part with the whole plant shoot growth rate.
3.2.5. Statistical analysis
Analysis of variance and mean separation (LSD) were conducted for the phenological
data using MSTATC program (Michigan State University, Michigan). Regression
analysis, t-tests and Kolmogorov-Smimov (KS) tests were made using NCSS 2000
(Number Cruncher Statistical Systems; Hintze, 1997). The Pearson correlation
coefficients were obtained using MINITAB for windows (Minitab Inc.).
44
3.3. Results and Discussion
3.3.1. Weather conditions
Daily maximum (Tmax) and minimum (Tmin) temperatures and monthly weather
conditions during the three seasons are shown in Fig. 3.1 and Table 3.2, respectively.
Both Tmax and Tmin were higher after 30 DAP in the second season than in the first and
third seasons.
45
40
35
u 30e.....G.I.::: 25ta..I
GI
Cl. 20
E
GI
I- 15
10 ---er- Tmax-2001/2002__+ __Tm in-2001/2002
__ Tmax-2002
5 _______Tmin-2002
---lIf-- Tmax-2002/2003
______ Tm in-2002/2003
0
0 20 40 60 80 100 120
Time (OAP)
Figure 3.1. Daily maximum (Tmax) and minimum (Tmin) temperatures during the three
seasons (2001/2002, 2002 and 2002/2003).
Table 3.2. Monthly weather conditions of the three seasons at Dire Dawa, Ethiopia.
Month Tmax Tmin RH SR Rainfall
(0C) eC) (%) (MJ m-2d-l) (mm)
Range Mean Range Mean Range Mean Range Mean Total
2001/2002
Dec. 28.2-31.7 29.8 12.1-18.5 15.0 32-70 52 12.8-20.7 19.4 6.0
Jan. 24.4-31.0 28.0 12.0-19.2 15.7 35-89 64 7.8-21.7 17.1 34.8
Feb. 29.8-34.6 31.6 11.6-19.8 15.4 25-65 51 8.9-23.0 21.2 0.0
Mar. 22.3-36.4 32.0 14.8-25.9 19.2 34-88 57 7.5-24.0 20.0 80.9
2002
Apr. 28.0-36.7 33.4 16.5-24.8 20.6 33-71 52 14.0-25.5 21.4 83.1
May 34.8-38.1 36.8 18.0-25.2 23.5 20-59 39 16.7-25.6 23.0 33.3
Jun. 33.8-38.2 36.4 19.8-24.8 22.9 21-55 38 12.1-24.5 22.1 9.7
2002/2003
Oet. 30.0-36.8 34.1 16.5-22.5 19.5 16-47 26 13.2-23.1 20.5 19.0
Nov. 30.2-34.1 32.0 14.0-19.8 16.4 16-42 26 16.2-22.0 20.7 0.0
Dec. 21.4-31.4 27.9 11.9-18.7 16.2 48-89 65 3.5-21.4 15.8 12.4
Jan. 25.3-32.4 28.3 8.3-21.2 14.5 38-73 60 10.0-22.7 19 19.8
45
The third season had higher Tmax and Tmin than the first season in the first half of the
season but lower Tmax and Tmin in the second half of the season. As shown by the relative
humidity (RH) values, the second season was less humid than the first but both seasons
received higher rainfall than the third season. Solar radiation (SR) was similar for all the
seasons. The length of the growing period was shorter in the second (94 days) than in the
first and third seasons (107 to 115 days).
3.3.2. Phenology
3.3.2.1. Calendar days
The phenological data is presented in Table 3.3 and 3.4. Cowpea had a mean emergence
period of 3-5 days in all the seasons, which was significantly shorter than both beans (6-9
days) and chickpea (7-9 days). Chickpea flowered 8-10 days earlier than beans and 10-12
days earlier than cowpea in 2001/2002 and 2002/2003 seasons.
However, the time from emergence to flowering was similar for the three species in 2002
due to the high temperature condition in this season, which hastened early flowering in
beans and cowpea. Chickpea also started pod formation earlier than the other species by
an average of 3 days. Early pod set in chickpea is considered as a prime strategy for
avoiding drought stress in environments prone to end of season water stress (Sedgley et
al., 1990). As stated by Kumar et al. (1996), development of early maturing varieties of
chickpea that escape drought can increase productivity and facilitates the production of
this crop in more drought prone areas.
The importance of earliness for better adaptation in drought prone environments has also
been shown for cowpea, pea and other grain legume crops (Hall and Patel, 1985; Sharma
and Khan, 1997). The number of days to pod initiation between beans and cowpea were
similar in 2001/2002 and 2002 but significantly different in 2002/2003 (Table 3.3). When
considered across seasons, the time to pod initiation was shorter in 2002 (48-50 days)
followed by 2002/2003 (50-57 days) and 2001/2002 (59-64 days). Higher temperature
hastened both the time to flowering and podding in 2002. The rate of progress towards
flowering in crop plants usually increases with increases in temperature up to an optimum
temperature (Summerfield et al., 1991; Roberts and Summerfield, 1987; Squire, 1990).
However, the period to pod initiation was not significantly affected by the MS treatment
(Table 3.3 and 3.4).
46
Table 3.3. Time to emergence (TE), flowering (TF), pod initiation (TP) and maturity (TM),
and pod filling period (PFP) in the 200112002and 2002/2003 seasons.
Sp WR Time (days)
200112002 2002/2003
TE TF TP TM PFP TE TF TP TM PFP
Beans C 9 58 62 93 31 6 49 54 102 48
MS 9 57 64 81 17 6 50 55 102 47
LS 9 58 63 77 14 6 48 54 84 30
Chickpea C 9 47 60 91 31 7 42 51 105 54
MS 8 48 60 78 18 7 42 50 106 56
LS 8 48 59 78 19 7 42 51 85 34
Cowpea C 6 61 64 90 26 4 53 57 104 47
MS 5 60 64 78 14 4 52 56 95 39
LS 5 59 64 78 14 4 51 55 84 29
LSD WR n.s n.s n.s 5.84** 6.07** n.s n.s. n.s 3.47*** 3.19***
(P<0.05) Sp 0.703*** 2.73*** 1.04*** n.s 3.28* 0.10** 2.21*** 0.98 n.s 4.57**
WR n.s. n.s. n.s n.s. n.s n.s. n.s. n.s n.s n.s.
x Sp
CV(%) 10.7 4.8 1.6 3.0 15.5 0.00 4.5 1.8 4.3 10.4
***, **, *:Treatment differences significant at 0.1, 1 and 5% probability level respectively, n.s: treatment
not significant at 5% probability level., ++ Sp = species, WR = water regime.
Table 3.4. Time to emergence (TE), flowering (TF), pod initiation (TP) and maturity (TM),
and pod filling period (pFP) in the 2002 season,"
Sp WR Time (days)
TE TF TP TM PFP
Beans C 6 43 48 89 41
MS 6 43 48 80 32
LS 6 43 48 73 25
Chickpea C 7 41 46 90 44
MS 7 41 46 65 19
LS 7 41 46 76 30
Cowpea C 3 43 50 91 41
MS 3 43 50 80 30
LS 3 43 50 76 26
measurements were not replicated.
There was no significant difference among species in the length of time from planting to
physiological maturity (Table 3.3 and 3.4) in any of the seasons. Nevertheless, the length
of time to physiological maturity was significantly affected by the water stress treatment
in all the seasons. The time to maturity in the C treatment was significantly longer than
both the MS and LS treatments in 2001/2002 and 2002 seasons and the LS treatment in
2002/2003. There was no significant difference between the MS and LS treatments in the
time to reach physiological maturity in 2001/2002 and 2002. However, in 2002/2003, the
MS treatment had similar length of time to mature to the C treatment (except in cowpea)
47
and was significantly different from the LS treatment in which plants matured 17 days
earlier. The longer maturity period in the MS treatment in 2002/2003 was due to
favourable temperature conditions «32°C) after re-watering that promoted the growth of
juvenile vegetative organs. The period after re-watering of the MS treatment in 2002 was
characterized by high temperatures (>34 °C) resulting in lack of vegetative re-growth of
organs in any of the species, and thus the length of maturity significantly reduced from
the C. Therefore, water deficit during the pod filling period significantly reduced the
length of time to physiological maturity whereas the effect in the MS treatment is
dependent on the temperature conditions after re-watering. Generally, the present results
indicated that water stress during the reproductive stage of grain legumes significantly
reduced the period of physiological maturity, particularly when it was coupled with high
temperatures. This is in line with the observation made by many authors for many crops
including chickpea (Singh, 1991), beans (Tedeschi and Zerbi, 1984), cowpea (Hall and
Patel, 1985), and wheat (Simane et al., 1993). Since it occurs towards the end of the rainy
season, end of season drought is usually associated with increasing temperature
(Calcango and Gallo, 1993; Singh, 1997b; Kumar and Abbo, 2001). Therefore, the
mechanism of shorter development period under late season water deficit has been related
to increases in leaf or canopy temperature (Slatyer, 1969; Sandhu and Horton, 1978). On
the contrary, severe water deficit is reported to delay developmental events in many
cereal crops because of the inhibition of growth resulting from the stress (see Blum,
1996).
Pod filling period (PFP) was shorter in 200112002 while it was longer in 2002/2003.
When daily temperatures were not too high (e.g. in 2001/2002 and 2002/2003), chickpea
had significantly longer PFP than cowpea in the first season and both beans and cowpea
in third season (Table 3.3). Cowpea had the shortest PFP during the same period
(2001/2002 and 2002/2003). Pooled over the species, the PFP in the C treatment was
significantly longer (13-15 days) than the MS and LS treatments in 200112002 and 2002
and the LS treatment (19 days) in 2002/2003. Except in 2002/2003, the MS and LS had
the same PFP when pooled over the species. Previous reports indicate that the duration of
pod filling varies greatly according to field conditions and growth type (Jeuffroy and Ney,
1997). For example, seed filling under different environmental conditions ends when
remobilizable nitrogen in the plant is exhausted (Munier-Jolian et al., 1996; Jeuffroy and
Ney, 1997). Similar to the observation in this study, water shortage during the
48
reproductive stage has shortened the period of seed filling in many other grain legumes
(e.g. Korte et al., 1983; Turc, 1995). There was no significant interaction between water
regime treatments and species for period of podding, physiological maturity and PFP in
the present study (Table 3.3) indicating similar response of the crops to the timing of
water stress for these characters.
3.3.2.2. Thermal time
Thermal time is an important phenological variable widely used in crop growth
simulation modelling. Therefore, determination of the thermal time of a certain growth
stage in field crops such as grain legumes is essential to develop models for a given site
and crop and/or calibrate existing models to suit a new environment. The thermal time
from planting to emergence (E), from emergence to flowering (E-F), from flowering to
pod initiation (F-P), from pod initiation to maturity (P-M) and from flowering to maturity
(F-M) was determined for beans, chickpea and cowpea for all three seasons, and the data
are presented in Fig. 3.2 and Appendix 3A&3B. Chickpea had longer thermal time
requirement (average of 132 cCd with a base temperature of 8°C) to emerge than beans
and cowpea in all the seasons. Cowpea needed an average of 76 cCd for emergence in
three seasons using a base temperature of 8 °c while beans needed an average of 101 cCd
using a base temperature of 10°C (Fig. 3.2). According to Squire (1990), differences in
the rate of germination at any temperature could be attributed to differences in optimum
and maximum (ceiling) temperatures in many determinate growth grain legumes. Rapidly
germinating seeds have a higher ceiling temperature than the slowly germinating ones
(Squire, 1990). Thus, the fast emergence of cowpea seeds would possibly shows a higher
ceiling temperature for germination in cowpea than in beans and chickpea. Early
germination is associated with early vigour and ground cover which are valuable drought
resistance traits in drought prone areas (Subbarao et al., 1995).
The average thermal time elapsed between emergence and flowering ranged from 600-
608 -ca in beans, 568-570 -ca in chickpea and 745-755 cCd in cowpea (Fig. 3.2). As
expected, there was no significant difference in thermal time between the treatment plots
before flowering. Compared to the control, the MS treatment in chickpea shortened the
period between flowering and pod initiation by an average of 13 cCd while the difference
between the C and MS for the same period was less than 5 cCd in beans and cowpea in
2002/2003 (Appendix 3A). However, there was no significant difference when data was
49
F-M Beans
~III
Cl P-M
~
cv
u F-P
oCl
o
c~ E-F
s::.
Il.
E
o 100 200 300 400 500 600 700 800
F-M
~III
Cl P-M
~
cv
~ F-P Chickpea
oCl
'c0~ E-F
s::.
Il.
E
o 100 200 300 400 500 600 700 800
F-M Cowpea
~III
Cl P-M
~
cuv F-P
Col
o
~c E-F
s::.
Il.
E
o 100 200 300 400 500 600 700 800
Thermal time (OCd)
Figure 3.2. Thermal time from planting to emergence (E), from emergence to flowering (E-
F), from flowering to podding (F-P), from podding to maturity (P-M) and from flowering to
maturity (F-M) for three grain legumes grown under well-watered (C) and mid-season (MS)
and late season (LS) water stress in three seasons. (Data are pooled over three seasons with
n =7: three replications for each of the two seasons (2001/2002 and 2002/2003) and one
replication for one season (2002». Horizontal bars refer to standard error of means.
50
pooled over seasons (Fig. 3.2). Significant differences were observed among the water
regime treatments for the thermal time elapsed between pod initiation and maturity and
between flowering and maturity (Appendix 3A). Compared to the C, the LS treatment
significantly reduced the thermal time required between pod initiation and maturity and
between flowering and maturity in all species while the effect of the MS treatment was
not consistent across seasons because of vegetative re-growth upon re-watering, which
was mainly influenced by the temperature conditions after re-watering. The average
thermal time elapsed between pod initiation and maturity was 509, 630 and 548 °Cd in the
C treatment, 403, 491, 322 °Cd in the MS treatment and 286, 372 and 322 °Cd in the LS
treatment for beans, chickpea and cowpea, respectively. On the other hand the average
thermal time elapsed between flowering and maturity was 553, 763 and 604 °Cd in the C,
461,624 and 462 °Cd in the MS, and 340,501 and 399 °ed in the LS for beans, chickpea
and cowpea, respectively (Fig. 3.2).
Generally, beans had lower thermal time than chickpea and cowpea. This may be related
to the higher base temperature used in the calculation for beans compared to the other two
species. Despite the same base temperature used for the two species, cowpea and
chickpea showed variability in thermal time requirements at different phenological stages.
For example, cowpea had higher thermal time requirement for the period between
emergence and flowering while it had lower requirements for the rest of its phenological
stages unlike chickpea which had lower thermal time requirements for the period between
emergence and flowering and higher requirements for the rest of its phenological stages.
Knowledge of such important differences is, thus, important for modelling the growth of
these crops for semi-arid areas. The use of thermal time to describe responses of plants
has been emphasized (Squire, 1990) and has been widely used to describe the progress of
crop development in grain legumes and other crops (Wilhelm and McMaster, 1995;
Jeuffroy and Ney, 1997).
3.3.3. Comparison of dry matter production
Besides the seeds which are used as a source of human and animal food, residues of grain
legumes are an important source of animal feed and nitrogen-rich fertilizer for the farmer
(e.g. Jayasundara et al., 1998) in many developing countries. Therefore, high dry matter
production is as important as the seed production for the subsistence farmer in sub-
Saharan Africa. In this study, therefore, the dry matter production of the three grain
legumes was compared both under well-watered and water stress conditions.
51
1000~ ~ 1ooo,- ~
C-BN 1000 MS-BN LS-BN
f 1200 __ AOM____ LOM 1200 1200
Cl
-:- 900 ••..... SOM_._ .• POM 900
~ 900
nl
E 600
,r:": 600 600~
o 300 .,,_~~-I.:_----
o I ~ ;... 300I ......""
f·· :_-»:" , <o ,
,
300 60' 0 900 1200'
, I
1000 1800 ':1 , !o 300 900 1200 1500 1800 , , I600 o 300 600 900 1200 1000 1800
1000,--- ~ 1500-r--------------, 1500,- -,,.. MS-CHPC-CHP
E 1200 1200 1200
LS-CHP
.~E! 900 900 900
~
nl
E 600 600 600
o~ 300 ,_--- ti'o I 300 300,.....;.<.-.;.-.;.:.::~-.~, -_._-/
~ .
, , .
... I o I ,.-=- 1 "'-"""'Y , , I o I ,...-="''C''j;'''--,. , , I
o 300 600 900 1200 1500 1800 o 300 600 900 1200 1000 1800 o 300 600 900 1200 1500 1800
1500,- _ 1000.----- ~ 1500,--- ,
,.. G-GOP MS-COP LS-COPE 1200 1200 1200
.~E! 900 900 900
~
nl
_j E 600 600 600
~
o~ 300 ~_.-:-...<..-..:- 300
cc:f:::>
I ..."..:'" .- .......... ". 300
O o ~ ~ _,.-t:6",'-- , ..., o I """"""-=:', .- " 1200 1000 18I00 0 I ~..--...-"~c--:,.;:.~;.,.. , I
vJ 300 600 900 1200 1000 1800 o 300 600 900 o 300 60' 0 900 1200 1500 1800
Thermal time (OCd)from planting Thermal time (OCd)from planting Thermal time (OCd)from planting
Figure 3.3. The seasonal course of total above ground dry matter (ADM), leaf dry matter (LDM), stem dry matter (SDM) and pod dry matter (pDM)
in beans, chickpea and cowpea under water stress (MS, LS) and non-stress (C) conditions in 2002 (BN = bean, CHP = chickpea, COP = cowpea).
Thermal time to flowering was 651, 790 and 867°C d for BN, CHP and COP, respectively.
52
1500.,------'-------:--------, 1500 MS-BN 1500 LS-BN
C-BN
~ 1200 __ ADM 1200____ LDM 1200
E
.s 900 ••••••. SDM... _._._PIl'vI 900 900
!
ft!
E 600
,
e /"
600 600
c 300 /' " 300
A-"'--' 300
o I ,~:-.~ , .. ---:_:., .... , I o I ,~~:-~-', , , I o I ,..-?? ~i , , , I
o 300 600 900 1200 1500 1800 o '300 600 900 1200 1500 1800 o 300 600 900 1200 1500 1800
1500,---------------,
C-CHP 1500r--------------. 1500.,--------------,
... 1200 MS-CHP LS-CHP1200 1200
.E.s.. 900 900 900
! 600
ft! 600 600
Ee 300 _. _'"c 300 300
o I ,~~"' ...~ ,I o I , .........;- -;-'-, , I o I ,--'"'T''' -ï-, , I
o 300 600 900 1200 1500 1800 o 300 600 900 1200 1500 1800 0 300 600 900 1200 1500 1800
1500 1500,----------------,... MS-COP 1500,---------------,C-COP 1200 1200 LS-COP'E 1200
..s..900 900 900
!
Ë 600 600 600
e
c 300 300 300
Ol ,~ .• ,. ,', I o I , ........:-.:.:.... - ,", , I o I , ~r" ,', , I
o 300 600 900 1200 1500 1800 o 300 600 900 1200 1500 1800 o 300 600 900 1200 1500 1800
Thermal time (OCd) after planting Thermal time (OCd) after planting Thermal time (OCd) after planting
Figure 3.4. The seasonal course of total above ground dry matter (ADM), leaf dry matter (LDM), stem dry matter (SDM) and pod dry matter (pDM)
in beans, chickpea and cowpea under water stress (MS, LS) and non-stress (C) conditions in 2002/2003 (BN = bean, CHP = chickpea, COP =
cowpea). Thermal time to flowering was 749,711 and 854 °Cd for BN, CHP and COP, respectively.
53
Under well-watered conditions, beans had faster growth than chickpea; and chickpea had
faster growth than cowpea during all the seasons (Fig. 3.3 and 3.4). Chickpea had the
lowest final dry matter in the C treatment under the relatively higher temperature season
(2002) but had the highest dry matter in the relatively milder temperature season
(2001/2002, data not shown). This shows the sensitivity of chickpea to high temperatures
which agrees with the report of Summerfield et al. (1981) that the reproductive growth of
chickpea suffered considerably more in hot (35/18 °c, day/night) environments compared
with that in a milder environment (30/10 -c, day/night).
In 2001/2002, bean accumulated the highest dry matter under the MS treatment because
of its long period of growth after re-watering followed by chickpea and cowpea in that
order. In this season, no significant differences were observed among water regime
treatments (Table 3.5).
In 2002, cowpea produced higher total dry matter than beans and chickpea (Fig. 3.3 and
3.4) indicating its better performance under high temperature conditions. Cowpea has
been considered to be one of the crop species most adapted to high temperatures and
grows better in hot environments such as the Sahel (Ntare, 1992; Ehlers and Hall, 1997).
In beans and chickpea, post-flowering dry matter production was significantly reduced by
mid-season water stress in 2002 whereas it was not significantly affected in cowpea
(Table 3.5, Fig. 3.3).
In 2002/2003, bean had the highest dry matter whereas chickpea and cowpea produced
equal amount of final dry matter. Dry matter in the MS treatment was significantly lower
than the C in all species in the 2002/2003 season.
The t- and KS- tests for differences in above ground dry matter production for the period
after flowering are shown in Table 3.5 for each season, species and water regime. Both
tests showed no significant differences in ADM among the water regime treatments in all
species in 2001/2002. The absence of significant difference between the C and MS
treatments in this season was due to vegetative re-growth in the MS treatments after re-
watering while senescence was accelerated in the C treatment. The tests showed a
different scenario in the other two seasons in which the MS treatment significantly
reduced post-flowering dry matter in beans and chickpea in 2002 and in all species in
2002/2003.
54
Table 3.5. Comparisons of total above ground dry matter (ADM), leaf dry matter (LDM),
stem dry matter (SDM) and pod dry matter (pDM) production and leaf area (LA)
expansion during the post-flowering period using two-sample t-test (t) and Kolmogorov-
Smirnov test (KS) for three seasons.
Species Water Regime* ADM LDM SDM PDM LA
KS t KS KS KS t KS
2001/2002
Bean C a a a a a a a a a a
MS a a a ab a a a a a a
LS a a a ac a a a a a a
Chickpea C a a a a a a a a a a
MS a a a ab a a b b b b
LS a a a b a a a a b b
Cowpea C a a a a a a a a a a
MS a a ab ab a a b a a a
LS a a b b a a b a a a
2002
Bean C a a a a a a a a a a
MS b b a a a a b b b b
LS a a a a a a ab ab a ab
Chickpea C a a a a a a a a a a
MS b b a a a a b b b b
LS a a a a a a b b ab ab
Cowpea C a a a a a a a a a a
MS a a a a a a a a a a
LS a a a a a a a a a a
2002/2003
Bean C a a a a a a a a a a
MS b b b b b b b b a a
LS ab ab a a ab ab ab ab a a
Chickpea C a a a a a a a a a a
MS b b b b b a b b b b
LS ab b a a a a ab ab a ab
Cowpea C a a a a a a a a a a
MS b b b b b b a b a a
LS a a a a a a a ab a a
* Water regime treatments designated by the same letter within a species are not significantly different from each other
at 5% t-test and the respective D values of the KS test.
The significant difference between the two treatments in the two seasons was because of
lack of re-growth of vegetative parts in the MS treatment after re-watering in 2002 and
2002/2003 because of high temperature and high VPD experienced by the plants after re-
watering, respectively.
On many occasions, alleviation of water stress is followed by a rapid rise in leaf water
potential and recovery of turgor (Simpson, 1981). The extent of recovery, however,
depends on the duration of stress, species involved, and subsequent environmental
conditions (Simpson, 1981). Except one case in chickpea, no significant differences were
observed between the C and LS treatments in post flowering dry matter production across
55
all species and seasons (Table 3.5). This is explained by the fast maturity of plants in the
LS treatment before a significant reduction in dry matter could occur.
Compared to beans and cowpea, chickpea was relatively sensitive to late season stress in
terms of dry matter production. Reduction in total dry matter of chickpea towards end of
season drought was also observed in other studies (Leport et al., 1999). According to the
KS-test, significant differences in post-flowering leaf dry matter production were
observed between the C and LS in chickpea and cowpea, and between LS and MS in
beans in 200112002, and between the C and MS treatments in 2002/2003 in all species
(Table 3.5). No significant differences were observed in 2002 among any of the water
stress treatments because of the depressing effect of high temperature on leaf growth,
particularly in chickpea and beans. In general, the results show the sensitivity of leaf dry
matter production to late-season water supply in chickpea and cowpea, and to mid-season
water supply in all species (Fig. 3.3 & 3.4). However, extremely high temperature
throughout the growing period, as observed in 2002, affected leaf dry matter
accumulation in all treatments and hence masked difference among the water regimes.
Besides temperature and species differences, leaf dry matter production during late season
water stress is also influenced by varietal differences. For example, Leport et al. (1999)
observed reduction ofleafand stem dry matter from 27-60% in many of the desi chickpea
genotyes while there was no reduction in the kabuli variety in a Mediterranean-type
environment.
A decrease in stem dry matter (SDM) production was observed towards the end of the
growing season in the C and LS treatments in beans and cowpea in all the seasons (Fig.
3.3&3.4). In chickpea, however, decline in stem dry matter production towards the end of
the season was not observed in any of the water regime treatments. A continuous increase
in stem dry matter was also observed for beans and cowpea in the MS treatment. There
was no significant difference in post-flowering stem dry matter production among the
water regime treatments and species except the C and MS treatments in beans and cowpea
in 200212003 (Table 3.5).
Maximum pod dry matter (PDM) was recorded in the C treatment followed by the LS
treatment in all species and seasons (Fig. 3.3 & 3.4). The MS treatment resulted in the
lowest PDM in all species and seasons. Among species, bean had the highest PDM while
chickpea and cowpea had similar PDM in all seasons under well-watered conditions.
56
PDM production by the species in the MS treatment was variable among seasons and as a
result there was no particular trend observed (Fig. 3.3&3.4). The KS test showed a
significant difference between the C and MS treatments in all species in 2002 and
2002/2003 and between the C and LS in chickpea in 2002 (Table 3.5). Since re-watering
resulted in growth of new leaves in most of the cases, the low PDM in the MS treatment
could be due to sink limitation rather than source limitation. As observed on many
occasions, early reproductive stage water tress (MS treatment) affects flowering through
inhibition of floral induction and development (Simpson, 1981; Saini and Westgate,
2000) and possibly flower drop. The stage of meiosis is believed to be the most stress
sensitive period of reproduction in a number of crops (Saini and Westgate, 2000), and
damage at this stage can limit the subsequent yield of the crop, particularly in determinate
plants (Simpson, 1981). As observed in chickpea in the present study, drought stress at
seed filling (LS treatment) can also limit yield despite favourable water condition earlier
in the development period (Simpson, 1981).
Leaf area growth was faster in beans than in chickpea and cowpea (Appendix 3C to 3E).
A decline in green leaf area was observed after pod initiation in all species, water regimes
and seasons except in the MS treatment of beans in 2001/2002. However, the rate of
decline was variable among species, water regimes and seasons. Rapid rate of total leaf
area decline was observed in the LS treatment in all species and seasons, and in the MS
treatment in chickpea. An increase in total leaf area was observed in beans after re-
watering of the MS treatment in 2001/2002 and 2002/2003 seasons. As shown in Table
3.5, significant differences in post flowering total leaf area were observed in chickpea
between the C and MS treatments in all the seasons and in beans in 2002. Reductions in
leaf expansion and subsequent decline in total dry matter production are associated with
the effect of water stress on cell expansion and division, which are usually reduced before
photosynthesis is affected (Hsiao, 1973; Simpson, 1981; Jeuffroy and Ney, 1997).
According to Blum (1996), the reduction in leaf area index and intercepted radiation
under water stress before flowering are largely the result of impaired leaf expansion and
changes in leaf display whereas the reduction after flowering is mainly due to early leaf
senescence. The same effect of water stress on leaf growth was observed in this study in
chickpea in all seasons and in beans only in the high temperature season. No significant
difference was observed among the water regimes in cowpea during any of the seasons
indicating its ability to maintain high total leaf area under different water supply and
57
temperature conditions. This agrees with previous reports that cowpea is less sensitive to
mild water deficit, and also has both drought avoiding and tolerating attributes (Grantz
and Hall, 1982; Squire, 1990).
Since leaf area is determined by phenology, stem morphology (e.g. stem length), rate of
leaf emergence and potential leaf size, the effect of drought on these factors also affects
leaf area (Blum, 1996). Therefore, lack of significant differences in the course of post-
flowering total leaf area among the water regime treatments in beans (2001/2002 and
2002/2003) and cowpea could be explained by the shorter stress accelerated maturity time
in the LS treatment (i.e. before many of the leaves senescence), and an increase in leaf
area in the MS treatment upon relief from the stress. Differences among species in
sensitivity of leaves to water stress are common (Simpson, 1981), and the current
differences in leaf area maintenance among the species indicate the differential response
of the crops to reproductive period water stress. Therefore, as observed in the present
study and in many other reports (e.g. Blum, 1996 and references there in), plasticity in
leaf area is an important mechanism by which crops under drought stress regulates water
use.
3.3.4. Leaf area duration (LAD) and fmal dry matter production
Three seasons' pooled data for LAD and final dry matter is presented in Fig. 3.5. The
results showed the dependence of dry matter production on green LAD in all the three
species. The LAD explained 75, 72 and 64% of the final dry matter variability in beans,
chickpea and cowpea, respectively. A t-test indicated that the slope of the regression line
in chickpea was significantly lower than that of beans and cowpea. Being the
photosynthetic site and primary source of assimilates, the leaf determines the performance
and productivity of plants in a given environment. Photosynthetic capacity of the leaf (net
assimilation rate, NAR) and its duration (LAD) are related to final yield (Hunt, 1982).
Similar to the present study, other workers also noted a positive correlation of longer
green leaf area duration with final yield in many legume crops (e.g. Kumudini et al.,
2001, Thomson and Siddique, 1977). As observed in this study in the LS treatment,
accelerated leaf senescence and reduced leaf growth are the major causes of final dry
matter reduction under drought environments. Therefore, species or cultivars that
maintain green leaf area under variable water supply, as observed in cowpea, could also
maintain yield under drought prone environments. It should also be noted that LAD is
58
influenced by temperature, ontogeny and nutrient availability besides water supply
(Squire, 1990).
In all species and seasons, much of the dry matter was allocated to the leaf until the plants
reach flowering when allocation to stems increased (Appendix 3F&3G). This IS lil
agreement with the observations made by Siddique et al. (1998) in lentil lil a
Mediterranean-type environment. The dry matter allocation after flowering was
influenced by species, water regime treatments and seasons.
1800
o BN (y = 5.86x + 33.9; R2 = 0.75)
1600
o CHP (y = 3.53x + 286.6, R2 = 0.72)
1400 o
t::. COP (y = 5.39x + 113.4, R2 = 0.64)
1200
~-
-E 1000Cl:cE 800«
600
400
200
0
0 50 100 150 200 250
LAD (days)
Figure 3.5. The relationship between leaf area duration (LAD) and above ground dry
matter at maturity (ADM) in beans (BN, -), chickpea (CHP, ----) and cowpea (COP, ---)
for data over three water regimes and three seasons (n = 9).
3.3.5. Dry matter (biomass) allocation
3.3.5.1. Dry matter allocation in 2002
The seasonal course of dry matter allocation ratio (AR) in 2002 is shown in Appendix 3F.
In this season, ARpod was increasing while ARleaf and ARstem were decreasing in beans
and cowpea under the C treatment. In chickpea, however, allocation to the pod and stem
increased slightly at the expense of allocation to the leaf in the same treatment. In the MS
treatment, allocation to the pod and stem increased while allocation to the leaf remained
59
constant (beans) or declined (chickpea). In cowpea, ARpod reached maximum of 1.21
(Appendix 3F) during the period of MS stress but declined after re-wateringand attained
a value of 0.37 at maturity (Table 3.6). Regardless of species, ARpod in the LS treatment
increased until maturity whereas ARteaf and ARstem decreased. The ARleaf decreased much
faster than ARstem in all species under late season stress.
Table 3.6. Allocation ratio (AR) calculated just before physiological maturity in three grain
legume species grown under three water regimes in 2002 and 2002/2003 seasons.
Species Plant part Seasons and water regimes
2002 2002/2003
C MS LS C MS LS
Beans AR1eaf -0.93 -0.49 -0.82 -0.25 -0.45 -1.28
ARm.m -0.77 0.96 -0.62 -0.24 0.00 -0.66
ARpoo 0.10 1.78 3.07 0.88 0.24 1.11
Chickpea AR1eaf -2.54 -1.77 -0.50 -0.31 -0.36 -0.85
ARm.m 0.28 0.65 0.75 0.39 0.01 0.30
ARpoo 0.71 1.79 1.82 0.03 1.30 1.17
Cowpea AR1eaf -0.36 -3.64 -0.94 -0.13 -0.52 -0.27
ARst.m -2.13 -1.36 0.36 0.03 0.12 -0.17
ARpoo 3.02 0.37 0.61 0.00 0.19 0.98
At around maturity, the ARpod in the LS treatment were 3.07, 1.82 and 0.61 for beans,
chickpea and cowpea respectively whereas the ARteaf were -0.82, -0.50 and 0.04 and
ARstem was -0.62, 0.75 and 0.36 for beans, chickpea and cowpea, respectively (Table 3.6).
The ARpod in the C treatment was 0.01, 0.71 and 3.00 for beans, chickpea and cowpea,
respectively for the same period (Table 3.6). Results show that under terminal drought,
crops allocate most of the assimilates to reproductive development whereas under C,
there is a tendency to distribute them to different above ground parts.
3.3.5.2. Dry matter allocation in 2002/2003
The seasonal course of AR among above ground parts in 2002/2003 is shown in
Appendix 30. Unlike 2002, ARpod under the C treatment in 2002/2003 increased with
time in beans whereas it decreased in chickpea and cowpea suggesting that allocation of
dry matter to the pod was also influenced by other environmental factors (like
temperature) besides water supply. Allocation of biomass to different plant parts usually
depends on species, growth habit (determinate vs. indeterminate), ontogeny and
environmental conditions such as temperature and water supply (Poorter and Nagel, 2000;
Squire, 1990). The MS treatment decreased ARstem in all species in this season,
60
particularly in beans and cowpea. The pattern of leaf, stem and pod AR in the LS
treatment was similar to that of the 2002 season. The ARpod at maturity was 1.11, 1.17 and
0.98 in the LS treatment and 0.88, 0.03 and 0.00 in the C treatment for beans, chickpea
and cowpea, respectively (Table 3.6). The seasonal AR in 200112002 was similar to that
of the 2002/2003 season (data not shown).
The contribution of post-flowering stem reserves to final grain yield is shown in Table
3.7. The reduction of stem dry matter at maturity relative to its maximum in the three
seasons ranged from 13-47%,2.7-3.2% and 0-23% in the C and 26-76%, 20-32% and 10-
34% in the LS treatments in beans, chickpea and cowpea, respectively. Such reduction of
stem dry matter towards the end of the growing season was not observed in the MS
treatment in any species or season except for cowpea in 2002 (Table 3.7). The
contribution of stem dry matter to the final grain yield in the three seasons ranged from 9-
37% in beans, 0-5% in chickpea, and 0-32% in cowpea in the C treatment and 19-64% in
beans, 28-40% in chickpea and 14-33% in cowpea in the LS treatment (Table 3.7).
Table 3.7. The contribution of post-flowering stem and leaf reserves to grain yield in beans,
chickpea and cowpea under well-watered (C) conditions and mid-season (MS) and late
season (LS) water stress in three seasons.
Calculated parameters* Seasons and water regimes
200112002 2002 2002/2003
C MS LS C MS LS C MS LS
Beans
1 86.4 80.5 38.3 71.0 71.9 64.9
2 47% 76% 13% 26% 28% 32%
3 37% 64% 9% 19% 22% 26%
4 93.0 88.8 198.4 97.8 186.0 127.7 84.7 128.2
5 54% 85% 69% 56% 71% 60% 59% 63%
Chickpea
1 6.94 28.8 76.7 9.2 39.0
2 3.2% 27% 32% 2.7% 20%
3 3% 30% 40% 5.2% 28%
4 53.0 14.4 162.3 85.6 29.1 65.9 38.0 68.4
5 26% 14% 42% 40% 11% 31% 28% 54%
Cowpea
1 53.7 23.2 24.6 51.1 24.0 112.4 5.2 21.1
2 23% 18% 17% 14% 7% 34% 3% 10%
3 32% 35% 33% 13% 13% 48% 4% 14%
4 15.1 20.8 166.0 97.0 134.6 57.3 60.0 38.4
5 9% 19% 47% 50% 54% 32% 45% 21%
* 1. Decreases in stem mass between period of its maximum and maturity (g m- )
2. Decreases in stem mass as percentage of maximum total stem mass
3. Decreases in stem mass as percentage of final grain mass (GY)
4. Decreases in leaf mass between period of its maximum and maturity (g m·2)
5. Decreases in leaf mass as percentage of maximum total leaf mass
61
The reductions in leaf mass between its maximum and maturity in the three seasons
ranged from 54-69%, 26-42% and 9-47% in the C treatment, 0-59%, 0-40% and 0-50% in
the MS treatment and 63-85%, 11-54%, and 19-54% in the LS treatment for beans,
chickpea and cowpea, respectively (Table 3.7). This indicated that the leaf had also
contributed a high percentage of its dry matter to the pod in each of the water regimes.
The current quantification of dry matter loss in stem and leaves to the grain is a gross
estimate and the calculated values seems to be higher because of the fact that the
observed reduction in dry mass was not adjusted for losses in respiration. In other crops
like wheat, for example, Borrell et al. (1989) used a factor of 0.33 (from a previous
report) to account for the loss of dry matter to respiration. However, if actual data is not
available for a given crop at a given site, the use of reported values might create further
complication. Therefore, the use of gross dry matter loss from a plant part as an index to
determine the magnitudes of assimilate translocation from a given organ, and to compare
species/varieties for a given purpose is preferable.
3.3.5.3. Relationship between leaf, stem and pod dry matter with total above ground
dry matter
For practical applications of assimilate translocation between plant organs in crop growth
simulation models, a quantitative relationship between leaf, stem, pod and root growth
has to be established. Although root growth was not determined in the present study
because of measurement difficulties under field conditions, the relationship between leaf,
stem and pod dry matter with the total above ground dry matter was determined using a
least square regression by combining the data of the three seasons as shown in Fig. 3.6,
3.7 and 3.8. Lower degree polynomial functions were found to be appropriate to explain
the partitioning of dry matter to the leaf and stem in all species and water regimes. On the
other hand, pod dry matter accumulation was explained by a power function in beans and
chickpea under all water regimes (Fig. 3.6 and 3.7) and by a linear function in cowpea
under the MS and LS treatments (Fig. 3.8).
The regression equations in Fig. 3.6 indicated that much of the dry matter of beans
tranlocated to the pod was from the leaves under the C and LS treatments, and that the
translocation was higher in the LS than in the C in beans. Except PDM in the MS
treatment, the fitted equations were able to explain more than 70% of the observed
in LDM, SDM and PDM accumulation of beans (Fig. 3.6).
62
900
LOM = -0.0003x2 + O.SOx + 11.24
R2 = 0.72
S 600 SOM = -0.0002x2 + 0.S1x - 14.94
:cii: R2 = 0.91
II.
:ii:
Q
I/) 300
:ii:
Q
...J
o
o 200 400 600 800 1000 1200 1400 1600
900 ,- --,
MS-BN LOM = 2x10-7X3 - 0.001x2 + 0.72x - 2.63
R2 = 0.74
S 600 SOM = -2x10-6X3 + 0.002x2 + 0.006x + 11.60
:ii:
Q R2 = 0.91
II.
:ii:
Q
300 POM = 0.0017x
1.656
I/)
:ii: R2 = 0.4925
Q
...J
o
o 200 400 600 800 1000 1200 1400 1600
900
LS-BN LOM = -0.0004x2 + 0.S4x + 6.89
R2 = 0.75
S 600 SOM = -0.0003x2 + O.SOx - 8.4
:ii:
Q R2 = 0.90
II.
:ii:
Q POM = 4X10-7X3.0687
I/) 300 2
:ii: R = 0.90
Q
...J
o
o 200 400 600 800 1000 1200 1400 1600
Total above ground dry matter (g m-2)
Figure 3.6. Dry matter allocation between leaf (LDM, ---), stem (SDM, ---) and pod (PDM,
-) in beans for data combined over three seasons under well-watered conditions (C) and
mid-season (MS) and' late-season (LS) water stress (n = 24, 24 and 18 for leaf and stem dry
matter and 13, 13, 7 for pod dry matter in the C, MS and LS treatments, respectively).
63
600
LOM = -7x10-7X3 + 0.0007x2 + 0.19x + 34.73 C-CHP
~ 500 R2 = 0.75
E
~ 400 SOM = 0.34x - 2.392
:!: R = 0.96
c
a.. 300
c:!:
fn 200
:!:
c
..J 100 oSOM
o .POM
o 200 400 600 800 1000 1200
600 .-----------------------------------~
MS-CHP
500 LOM = -0.0008x
2 + 0.70x - 0.23
R2=0.76
.2! 400 SOM = 0.33x + 5.81
:!:
c
e, R
2 = 0.83
300
c:!:
fn 200 POM = 3x 10-6X2.8315:c!: R2 = 0.76
..J
100
o
o 200 400 600 800 1000 1200
600
LS-CHP LOM = -0.0003x2 + 0.57x + 3.888
~ 500 R2 = 0.83
'E
~ 400 SOM = 0.3926x - 5.4912
:c!: R2 = 0.97
e, 300
:c!:
fn 200 POM = 0.0001X2.19272
:!: R = 0.75c
..J 100
o
o 200 400 600 800 1000 1200
Total above ground dry matter (g m-2)
Figure 3.7. Dry matter allocation between leaf (LDM'---), stem (SDM, ---) and pod (pDM,
-) in chickpea for data combined over three seasons under well-watered conditions (C)
and mid-season (MS) and late-season (LS) water stress (n = 24, 21 and 18 for leaf and stem,
and 16, 12,9 for pod in the C, MS and LS treatments, respectively).
64
600
LOM = -0.0003x2 + 0.52x + 13.74
.- 500 R2 = 0.72
"I A-
E
400 SOM = -0.0003x
2 + 0.58x - 16.17 A
2.! oSOM
:5 R
2 = 0.93 A
Q olPOM
Il. 300 POM = 6x10-5X2.2546
:5 R2 = 0.79 ......, 0•• - 0 ---- .0...---.Q -
fI) 200 . ....-.i,"' ..E)' ••• -~-~"e 0oe." :::: ~. ~ 0 0.' .. '"
:5 'ij 0 "~.
Q <0 ~ • 0 ,
..J ,~. 0100 0~~~ ..
•; tJ'UIl
0
0 200 400 600 800 1000 1200 1400
600
LOM = 3x10·6X3 - 0.0037x2 + 1.26x - 3.98 MS-COP
500 R2 = 0.74
or
E 400 SOM = 0.0001x
2 + 0.28x + 10.72
2.! R2 = 0.92
:5
Q
Il. 300 POM = 0.42x - 69.29
:5 R2Q = 0.74
fI) 200 ,
:5 0'
Q
..J
100
0
0 200 400 600 800 1000 1200 1400
600
LS-COP LOM = -0.0006x2 + 0.65x + 11.10
.r 500 R2=0.75
'E
2.! 400 SOM = -0.0003x2 + 0.53x + 5.19
:5
Q
Il. 300 8 R
2 = 0.80
:5
Q 200 PDM = 0.48x - 97.71fI)
:5 R2 = 0.68
Q
..J 100
0
0 200 400 600 800 1000 1200 1400
Total above ground dry matter (g m-2)
Figure 3.8. Dry matter allocation between leaf (LDM, ---), stem (SDM, ---) and pod (PDM,
-) in cowpea for data combined over three seasons under well-watered conditions (C)
condition and mid-season (MS) and late-season (LS) water stress (n = 24, 21 and 18 for leaf
and stem, and 13, 11, 8 for pod in the C, MS and LS treatments, respectively).
65
Regression of above ground parts on total dry matter in chickpea indicated a linear
growth of SDM in all water regimes (Fig. 3.7), and the PDM was best explained by the
power function under all water regimes (R2 >0.75). Un1ike beans and cowpea where leaf
re-growth was observed after re-watering, the regression equation in chickpea indicated
higher dry matter allocation to the pod under mid-season water stress which was
translocated on1y from the leaves (Fig. 3.7). The regression equations also showed higher
leaf dry matter translocation to the pod in the LS than in the C treatment. In chickpea, the
equation explained more than 75% of the variability in LDM, SDM and PDM under all
water regimes.
The regression equations in cowpea indicated that dry matter allocation between leaf and
stem was similar under well-watered conditions while allocation to- and translocation
from- the leaf was higher than the stem under late season water stress (Fig. 3.8). The
growth of PDM was explained by a linear function (R2>0.67) in both the water stress
treatments while it was explained by the power function (R2 = 0.79) under well-watered
condition. The equations under water stress conditions indicated that PDM growth was
higher in the LS than in the MS treatment in cowpea. The regression coefficients for
SDM in the MS were positive suggesting that dry matter was not translocated from the
stem to the pod in cowpea under mid-season water stress.
Generally, the regression coefficients in the three species indicated that dry matter
translocation between above ground parts was higher under conditions of late season
water stress followed by well-watered conditions similar to the data shown in Table 3.7.
Under these conditions, much of the translocation goes to pod growth, and the
translocation from the leaf is higher than from the stem in most cases. Under mid-season
water stress, translocation from the leaf to the pod was observed in chickpea whereas the
translocation in beans and cowpea was minimal. Regression of the three seasons
combined data indicated that dry matter translocation from the stems to the pods under
MS condition was un1ikely in all species.
When water supply is variable in the growing season, dry matter is partitioned differently
to the different parts of the plant in a way that maximizes growth (Boyer, 1996; Ninkovic,
2003; Poorter and Nagel, 2000). This translocation of assimilates, which enables the plant
to capture more of the resources that limit growth most, is considered as an adaptive
66
mechanism (Poorter and Nagel, 2000). Leaves, stems, roots and nodules are the major
sinks for assimilate produced prior to pod initiation in grain legumes (Singh, 1991).
Water stress after flowering resulted in allocation of greater proportion of assimilates to
pods than water stress during the whole growth phases in chickpea and most of the
translocation of assimilates to pods was reported to come from leaves (Singh, 1991).
About 15-20% of assimilates produced prior to pod initiation was translocated to pods in
chickpea (Singh, 1991; Saxena 1984) and the amount of translocation was directly
proportional to the intensity of water stress during pod and seed growth (Singh, 1991).
This fully agrees with the present result observed in chickpea. Leport et al. (1999)
observed higher redistribution of above ground dry matter during seed filling in Desi
chickpea than in Kabuli chickpea. Despite higher total dry matter, the latter produced the
lowest yield. Therefore, final grain yield is determined by total biomass production and
the proportion allocated to grain in both legume (Muchow et al., 1993; Leport et al.,
1999) and cereal crops (Squire, 1990; Van den Boogaard et al., 1997). Among grain
legumes, the translocation of reserves in chickpea is reported to be higher than faba bean,
lentil and field pea (Wery et al., 1993). Chickpea, however, had the lowest AR in the
present study as compared to beans and cowpea which could be due tg varietal
differences. In line with the present study, allocation of dry matter from vegetative parts
to the reproductive organs under water stress has been reported for many grain legumes
including lupins (French and Turner, 1991), mungbean (Bushby and Lawn, 1992), peanut
(Wright et al., 1991), soybean (Westgate, et al., 1989) and pigeonpea (Robertson et al.,
2001). As observed in the present study and also stated by Squire (1990), the effect of
drought on partitioning of assimilate depends on its severity, stage of crop development,
sink type and duration of sink growth.
3.3.6. Specific leaf area (SLA) and its relation with WUE
SLA plays a significant role in the growth and development of a given crop species.
Easier and less expensive measurements make it a desirable parameter in crop physiology
studies (Nageswara Rao and Wright, 1994). The seasonal mean SLA was determined for
the three grain legumes by regressing leaf area against leaf dry mass. In 200112002, the
highest SLA was recorded in the MS for bean arid in the LS for chickpea and cowpea
(Table 3.8). In 2002, SLA was highest in the LS treatment for all species (Table 3.8).
67
The SLA in 2002 ranged from 125-219 in beans, 99-144 in chickpea and 90-216 cm2 g"
in cowpea. SLA was generally higher and less variable among treatments in 2002/2003
than the previous two seasons (Table 3.8). The highest SLA was observed in the MS
treatment in chickpea and cowpea in 2002/2003. The values ranged from 155-182 in
beans, 114-136 in chickpea and 155-172 cm2 g-l in cowpea. Pooled over water regime
treatments, bean had the highest SLA followed by chickpea in the first two seasons and
by cowpea in the third season.
Table 3.8. Specific leaf area (SLA, cm2 g-l) of three grain legumes obtained from a linear
regression of leaf area vs. leaf dry matter for three seasons.
Species Water 2001/2002 2002 2002/2003
regime SLA R n SLA R n SLA n
Beans C 42 ± 26.8 0.83 7 183± 21.0 0.93 8 182 ± 13.1 0.96 9
MS 168 ± 38.3 0.76 8 125 ± 62.4 0.45 7 155 ± 8.2 0.98 9
LS 83 ± 64.3 0.29 5 219±14.1 0.98 6 170±5.9 0.99 7
Chickpea C 63 ± 12.9 0.83 7 112±13.5 0.92 8 119± 7.0 0.98 9
MS 54 ± 14.3 0.82 5 99±35.5 0.66 6 136 ± 4.3 0.99 9
LS 103 ± 8.1 0.99 4 144± 34.7 0.81 6 114±l1.2 0.95 6
Cowpea C 69± 17.5 0.79 6 90±28.3 0.63 8 172±15.0 0.95 9
MS 50±40.2 0.25 6 I69±24.4 0.91 8 183±12.4 0.97 9
LS 98 ±22.5 0.86 5 216 ± 15.3 0.98 6 155±9.7 0.98 7
Table 3.9. Specific leaf area (SLA, cm2 g-l) of three grain legumes based on a linear
regression of leaf area vs. leaf dry matter for all three seasons data combined.a
Species Water regime SLA R n P
Beans C 169 ±15.8 0.84 24 0.000
MS 150 ± 22.2 0.68 24 0.000
LS 208 ± 13.7 0.93 19 0.000
Chickpea C 114 ± 8.4 0.89 24 0.000
MS 107 ± 16.1 0.71 20 0.000
LS 143 ± 16.0 0.83 18 0.000
Cowpea C 114 ± 12.3 0.81 22 0.000
MS 160 ± 15.1 0.86 21 0.000
LS 186 ± 10.9 0.88 19 0.000
il SD = standard deviation of the regression line, n = number of observations, P = probability level.
Data was combined over three seasons in order to get a representative SLA for each
species and water regime. An example of the regression analysis for the combined data is
shown in Fig. 3.9 and the SLA for each water regime and species is presented in Table
3.9. When data was combined over the three seasons, the correlation (r) between LW and
SLA was greater than 0.90 in the C and LS treatments and greater than 0.80 in the MS
treatment (Table 3.9). The pooled data showed that SLA was significantly higher in the
68
60000 ~ ~
BN-C
50000
02002/03
40000 t:. 2002
E 02001/02
~ 30000
ca
Q)
.:c..
;. 20000 y = 169.1 x + 89a o 2
~ 10000 o R = 0.84
o 100 200 300 400 500
60000 ~ ~ ~
CHP-C
50000
E 40000
"!
E
~ 30000
ca
Q)
c"-....a 20000ca y = 114.2x + 1623
Q)
..J 10000 o R2 = 0.89
o 100 200 300 400 500
60000 ~ ~
50000 cOP-C
NE 40000
E
~ca 30000
Q)
.c
"-
...a 20000 t:.ca Y = 114.1x + 4810
~ 10000
R2 = 0.81
o Ol
o 100 200 300 400 500
Leaf dry matter (g m-2)
Figure 3.9. An example of specific leaf area (SLA) determination by regression of leaf area
vs. leaf dry matter in beans (BN), chickpea (eHP) and cowpea grown under well-watered
(C) conditions for data combined over three seasons.
69
LS than the other two treatments in all species, and bean had the highest SLA followed by
cowpea. Higher SLA in the LS shows thinner leaves which is indicative of dry matter
translocation from leaves to reproductive organs.
The relation of SLA with water use efficiency (WUE, see Chapter 4) under well-watered
and mid-season stress conditions for the 2002 and 2002/2003 seasons is presented in Fig.
3.10 and 3.11, respectively. SLA was strongly negatively correlated with WUE in the C
treatment in all species in both seasons and the MS treatments in beans in 2002/2003.
5
2
C-2002 oBN (y = -0.0227x + 7.3, R = 0.73)
oCHP (y = -0.0347x + 7.7, R2 = 0.84)4
ACDP (y = -0.0156x + 4.9, R2 = 0.91)
- ,.-. "E A
E 3
~ ~
E "\'\..
.C_l 2 '
W " -, '\.
::I
3: -IJ "
1 '-
(]I\
A"'''()
0 0 ..
0 100 200 300 400 500 600
5
C-2002/2003 oBN (y = -0.0403x + 8.8, R2 = 0.92)
4
- o CHP (y = -0.0303x + 5.4, R2 = 0.89).,-.E ACDP (y = -0.0383x + 7.5, R2 = 0.84)
E 3
~
E ,0
.C_l 2 A'
W ~ ~
::I 0 'A
3: 0 01 , ',A0' ,£
0'. •
0
0 d"., 0
0 100 200 300 400 500 600
SLA (ern" s')
Figure 3.10. The relationship between water use efficiency (WUE) and specific leaf area
(SLA) in beans (BN), chickpea (CHP) and cowpea (COP) under well-watered conditions in
two seasons (n = 6 and 8 for 2002 and 2002/2003 seasons, respectively).
70
Figure 3.11. The relationship between water use efficiency (WUE) and specific leaf area
(SLA) in beans (BN), chickpea (CHP) and cowpea (COP) under mid-season water stress
(MS) during the reproductive period in two seasons (n = 5 for each season). The data
considered was only for the period after flowering during the time of water stress.
Leaf area and leaf dry matter, from which SLA is derived, are closely related to radiation
interception, photosynthesis, transpiration, growth rate and final yield (Ma et al. 1992).
SLA determines the physiological cost of producing leaf area (Dingkuhn, et al., 2001).
High SLA is a major factor for early ground cover and high radiation interception and
hence determines the growth of plants in many situations (Dingkuhn et al., 2001). SLA is
an important component of crop growth simulation models as it relates dry matter
71
production to leaf area expansion, and consequently to radiation interception and
photosynthesis (Gray et al., 1993; Manschadi et al., 1998b). A sugarcane variety with
greater SLA showed a more rapid leaf area expansion and growth, and as a result SLA
was recommended as an index for improving the early growth of sugarcane (Terauchi et
al., 2001). Therefore, the high SLA in beans and cowpea in the present study suggests
better performance of the species in the utilization of resources such as radiation and
water. The highest and lowest SLA values observed in the LS and MS treatments
respectively suggest thinner leaves in the former (possible mobilization of assimilates
from the leaves) and thicker leaves in the latter (accumulation of high leaf mass).
As reported by Wright et al. (1994) and Nageswara Rao and Wright (1994), SLA is
positively correlated to carbon isotope discrimination (ó) and negatively to WUE (based
on transpiration) in peanut. It was also found to be stable across genotypes and
environments (Nageswara Rao and Wright, 1994). Therefore, SLA can be used as a
surrogate for ~ to identify genotypes with high WUE and total biomass (Hubick et al.,
1986; Wright et al., 1994; Nageswara Rao and Wright, 1994). A strong negative
correlation between WUE (based on evaporanspiration) and SLA was also observed in
the present study under well-watered conditions in all species (Fig. 3.10) suggesting the
potential use of the character as selection criteria for high field water use efficiency.
However, the relation under mid-season water stress was neither strong nor consistent
between seasons and species (Fig. 3.11) indicating the limitation of using SLA as
indicator of high field WUE under conditions of high soil surface evaporation.
3.4. Summary and conclusion
Information on phenology, pattern of growth and dry matter partitioning under different
environmental conditions is essential for agricultural decision-making and in the
development and/or calibration of crop growth simulation models. The phenology,
growth and dry matter partitioning behaviour of three grain legumes were investigated in
this study. Compared to the control, late season water stress significantly shortened the
time to maturity of the three grain legumes in all seasons while the effect of the mid-
season stress was season dependent. The thermal time requirement of the different
phenological stages of the species were determined under water stress and non-stress
conditions so that these values could be used to predict the phenological stages for
management decisions and in crop simulation models.
72
Leaf area growth and above ground dry matter production were significantly reduced by
mid-season water stress but not by late season stress. Re-growth of vegetative parts such
as leaves was observed in beans and cowpea upon re-watering after mid-season stress.
However, the degree of re-growth and its duration after stress relief was affected by
temperature conditions. High temperature conditions inhibited re-growth. LAD is highly
correlated with final dry matter yield in all species suggesting that radiation based crop
growth models can be effective in simulating the dry matter production of these grain
legume crops.
Allocation of assimilates among above ground parts was influenced by both the timing of
water supply and species. Allocation of dry matter to the pod was higher under the late
season water stress followed by the well-watered condition in all species. Both the leaves
and stem contributed to the growth of the pods although much of the dry matter was
translocated from the former. Allocation of dry matter under the MS stress was observed
only from the leaves and it was small when compared to the one in the LS and C
treatments. Among the species, beans had higher dry matter allocation to the pod than
chickpea and cowpea under C and LS conditions, whereas chickpea had higher pod
allocation than beans and cowpea under the MS conditions. Allocation of dry matter to
the pods under well-watered conditions seems to be season dependent in beans and
cowpea. Under milder temperature conditions, bean tend to allocate more dry matter to its
pod at the expense of allocation to its stem and leaf whereas cowpea allocate dry matter to
all parts with no preference. This condition reversed at high temperature conditions. The
combined data over the three seasons indicated that dry matter allocation among
aboveground parts was best explained by 1st to 3rd degree polynomial functions for the
stem and leaf growth and by a power function for the pod growth. The regression fit was
excellent and the coefficients determined can be used in calibrating existing crop growth
models as well as to develop new ones.
SLA was significantly negatively correlated with WUE under well-watered conditions in
all the species in both seasons, and, thus, it could be used as a selection criterion for high
WUE in high rainfall environments. However, the relationship under mid-season water
stress conditions was not strong suggesting further investigation in the stability of the
relationship between the two parameters under field conditions.
73
Generally, differences between chickpea on the one hand and cowpea and beans on the
other are wider than difference between cowpea and beans for many of the characters
studied. This kind of information is, therefore, essential to facilitate crop choice for a
given environment and also adjust to management practices.
74
CHAPTER 4
Resource Utilization of Three Grain Legume Species in A Semi-Arid Environment.
I. Water Use and Water Use Efficiency
4.1. Introduction
Dry matter production of a crop depends on the amount of water used and its efficiency of
use (Black and Ong, 2000). Plant water use depends mainly on water supply and
evaporative demand of a given environment. Therefore, crop water use is determined by
the prevailing atmospheric evaporative demand of the environment under well-watered
conditions and by both evaporative demand and crop factors under water deficit
conditions (Baldocchi et al., 1985; Singh et al., 1990; Black and Ong, 2000). Among crop
factors, regulation of canopy size can be more important than leaf conductance in
controlling water use during sustained drought (Ong et al., 1996). Therefore, based on the
type of water deficits (transient or long), water use at canopy level is controlled by long-
and short-term regulatory mechanisms in which reductions in transpiration at any level
results in reduced assimilation and growth (Blum, 1996; Black and Ong, 2000).
Water use efficiency (WUE) can be defined as the amount of dry matter (DM) produced
per unit of evapotranspiration (ET) (Sinc1air et al., 1984; Cooper et al., 1988; Turner,
1997). Both the numerator and denominator of this ratio are defined in many different
ways. The numerator could be expressed as the mass of CO2 that enters the stomata, or
the tofal 'dry matter produced by the crop, or the above ground dry matter or the grain
yield of the crop, and the denominator as the amount of water leaving the stomata
(transpiration) or the amount lost as evapotranspiration from the crop and soil (Cooper et
al., 1988). As it can be predicted readily from physiological principles and is relatively
conservative for a given location, the ratio of mass of C02 fixed as carbohydrate to mass
of water transpired from leaves, which is commonly termed as transpiration efficiency
(TE), is a useful quantity to evaluate crop performance (Tanner and Sinclair, 1983;
Copper et al., 1988). While the ratio of dry matter to transpiration shows the total biomass
productivity relative to water used by the plant, the ratio of dry matter to
evapotranspiration shows the agronomic yield of the system relative to total water use
(Loomis, 1983).
75
WUE is influenced by many factors including water supply, saturation vapor pressure
deficit of the air, CO2 concentration in the air, air temperature, plant factors (carbon
metabolism, stomatal behavior, canopy size and structure) and soil properties (Stanhill,
1986; Copper et al., 1988). Therefore, unlike TE, which is conservative within a given
species, WUE is affected by management practices. Environmental factors (e.g. drought)
that lower the leaf area and thereby increase soil surface evaporation (Es) are known to
reduce the WUE (Tanner and Sinclair, 1983). Comparisons of plant species show that the
WUE values for tropical C4 cereals are more than twice that of C3 species under similar
conditions, although drought tolerant C3 species (e.g. cowpea and cotton) show similar
WUE values as drought sensitive cultivars of C4 species such as maize and sorghum
(Squire, 1990; Black and Ong, 2000).
Supply of water is a major constraint in the semi-arid environments, and the pattern of
water deficit during the season varies across locations and years and with soil types
(Singh et al., 1990; Turner et al., 2001). Grain yield is a combined result of many
physiological and biochemical processes. Therefore, the study of yield determining
processes provides a better mechanistic assessment of the performance of a given cultivar
or environment than yield per se. Passioura (1977) expressed cereal yield in the dry
environments as shown in equation 1.7. Based on this relationship, the study ofW, WUE
and HI has been used in assessing the adaptation and yield of a number of cereal crops
(Turner et al., 2001) and recently grain legumes (Siddique et al., 2001). The components
of this model can, therefore, be used to study the performance of grain legumes under the
semi-arid regions where water shortage is prevalent.
As reported in many studies (e.g. French and Schultz, 1984; Perry, 1987; Loss et aI.,
1997; Siddique et al., 2001), the efficiency of water use by a crop can be used as a
benchmark for evaluating crop performance as well as for comparing environments.
Accordingly, the water use and WUE of many grain legumes have been studied
individually under various environments (e.g. Copper et al., 1988; Pannu and Singh,
1993; Silim et al., 1993a; Loss et al., 1997; Ashok, et al., 1999; Collino et al., 2000;
Siddique et al., 2001). However, there is little information available on the comparative
water use and WUE of grain legumes under water-stress and non-stress conditions in the
low rainfall semi-arid areas of Ethiopia. The objective of this study was, therefore, to
compare water use and WUE of common bean (Phaseolus vulgaris L.), chickpea (Cicer
76
arietinum L.) and cowpea (Vigna anguiculata L.) grown under different water regimes in
a semi-arid region.
4.2. Materials and Methods
4.2.1. Field experiments
Details of experimental site, material and design, cultural practices and irrigation
schedule are given in Chapter 3 and will be explained here in brief. Three field
experiments were conducted at Dire Dawa, Ethiopia during the periods from early
December 2001 to late March 2002 (first season), from late March to early July 2002
(second season) and from mid October 2002 to early February 2003 (third season). Seeds
of common bean (ev. Roba-I), chickpea (ev. ICC-4958) and cowpea (ev. black eye bean)
were planted on December 7, 2001, March 27, 2002 and October 17, 2002, for the first,
second and third seasons, respectively. The experiments had three water regime
treatments as shown in Table 3.1. The experimental treatments, replicated three times,
were arranged in a randomized split plot design using the water regime treatments as
main plot and the crop species as sub-plot. The total experimental area was 22.8m x
40.2m.
4.2.2. Experimental measurements
The soil water content to a depth of 300 mm in 2001/2002 season and 600 mm in 2002 and
2002/2003 seasons was monitored every day throughout the growing period using Time
Domain Reflectometery (TDR) (Soil Moisture Equipment Corp., CA, USA). Above ground
dry matter (ADM) was measured from five plants (0.2 m2 area) per plot at intervals of 10
days starting on 35, 17 and 20 days after planting (DAP) for the first, second and third
experiments, respectively until physiological maturity. The plant samples were separated
into leaf, stem, pod and seed and dried in an oven for 72 hours at 60°C for dry matter
determination. Based on the maturity period of plants in the different water regimes, the final
harvest was done from March 1 to April 6, from June 7 to July 4 and from January 13 to
February 5 in the first, second and third seasons, respectively.
4.2.3. Calculations
4.2.3.1. Seasonal evapotranspiration (ET)
The ET was calculated on a daily basis using the water balance equation as follows:
77
ET = P + Ir - R - D ± L\S. (4.1)
where P is rainfall, Ir is irrigation, R is runoff, D is deep percolation/drainage, and L\S is
change in soil water stored within in 600 mm depth (300 mm for the first season). Since
water was applied to reach field capacity during irrigation and there was no heavy rainfall
during the experimental periods, R and D were considered negligible in this experiment.
ET was calculated for different periods: (i) pre-flowering (ETb), (ii) post-flowering (ETa)
and (iii) seasonal total (ETs).
4.2.3.2. Water use efficiency
The WUE was calculated using total above ground dry matter at different stages for the pre-
flowering period (WUEb), post-flowering period (WUEa), whole season (WUEd) and for
grain yield at harvest (WUEg) as follows:
WUE = ADMb (4.2)
b ET.
b
WUE = ADMa (4.3)
a ET
a
WUE =ADM (4.4)
d ET
s
y
WUE =- (4.5)
g ET
s
where ADMb, ADMa and ADM refer above ground dry matter (kg ha") before flowering,
after flowering and at harvest, respectively, Y is grain yield (kg ha").
4.2.3.3. Transpiration (T)
Transpiration was calculated from DM using the following relationship (Tanner and
Sinc1air, 1983; Singh et al., 1990; Squire, 1990).
= DMT --(es -ea) (4.6)
ew
where T is transpiration (kg m-2), DM is total dry matter (g m-2), es and ea are mean
daytime saturation and actual vapour pressure of the air respectively (kPa) and ew is crop
specific transpiration efficiency coefficient (g kPa kg"). The ew values (g kPa kg") used
were 4.8 for chickpea (ICRISAT, 1988; Singh et al., 1990; Singh and Sri Rama, 1989),
78
4.2 for beans (Ogindo and Walker, 2003) and 3.5 for cowpea (Barnard et al., 1998; Ashok
et al., 1999). Because root dry matter was not measured, the total dry matter was
estimated using top/root ratio of 5: 1 for the well-watered treatment as shown for soybeans
(Barnard et al., 1998) and 4: 1 for the water stress treatments because of increased root
density and dry matter allocation to roots during water deficit (e.g. Husain et al., 1990;
Manschadi et al., 1998a). Transpiration efficiency for grain yield at harvest (TEg) was
calculated as the ratio of grain yield (Y) to seasonal transpiration (Ts). Seasonal soil
surface evaporation (Es) was calculated as the difference between seasonal
evapotranspiration (ETs) and T; Harvest index (HI) was calculated as the ratio of grain
yield to total above ground try matter at harvest.
4.2.3.4. Vapour pressure (e)
Saturation vapour pressure (es) was calculated as described by AlIen et al. (1998) as
follows:
eO(T ) = 0.6108ex ( 17.27Tmax J (4.7)
max p T + 237.3
max
eO(T. ) = 0.6108ex ( 17.27Tmin J (4.8)
mm p Tmin + 237.3
(4.9)
eO(T. )* RHmax +eO(T )* RHmin
=---m-me ----10~0~-2----
ma-x ---1-00~=- (4.9)
a
(4.11)
Tmax, Tmin,RHmax, RHminrefers to daytime maximum temperature, minimum temperature
eC), maximum relative humidity and minimum relative humidity (%), respectively and
VPD is mean daytime vapour pressure deficit of air (kPa).
4.3. Results and Discussion
4.3.1. Seasonal irrigation and water use
Because of the shallow depth of soil water measurement in the first season, detail results
are presented only for the second and third seasons. Cumulative ET was higher in the
79
third season than in the second because of longer growing period in the 2002/2003 season
(Fig. 4.1). However, the rate of water use (data not shown) was higher in the second than
in the third season. This could be due to higher vapour pressure deficit (VPD) of the air in
the second season as compared to the third one (Table 4.1). As reported on many
occasions (e.g. Bierhuizen and Slatyer, 1965; Tanner and Sinclair, 1983; Stanhill, 1986;
Squire, 1990), the VPD of the air is the most important driving force that controls the rate
of water vapour exchange between a plant canopy and its boundary layer. Irrigation and
water use were the highest in the C treatment followed by the MS treatment in both
seasons in all the species (Fig.4.1 and 4.2). The cumulative water use in the MS and LS
treatments was lower than the C treatment during the respective treatment periods and
remained below the control for the rest of the season (Fig. 4.1). As shown in the same
figure, an increase in water use was observed in the MS treatment upon re-watering. The
first season has lower water use (Appendix 5) compared to the other two seasons, which
could be due to low VPD of the season (Table 4.1) and/or shallow soil water
measurement.
4.3.2. Comparison of pre-flowering, post-flowering and seasonal water use
Significant differences in total seasonal water use were observed among the water stress
treatments and among the species in both seasons (Table 4.2). The LS treatment had
lower seasonal water use (ETs) than the MS and the difference in ETs between the MS
and LS treatments was significant (P<0.05) in both seasons (Table 4.2). Significant
differences (P<0.05) were also observed among the species.
Under well-watered conditions, cowpea had the highest ETs (403 mm) compared to the
lowest in chickpea (375 mm) in the second season whereas the three species had similar
ETs values (422-430 mm) in the third season (Table 4.2). Chickpea had the lowest ETs
and Es in the second season while it had the highest in the third season. This seasonal
difference in ETs was, therefore, mainly due to differences in leaf area growth (Chapter 3)
which affected percent ground cover and soil surface evaporation. Beans had the lowest
ETs under the LS treatment in both seasons while chickpea and cowpea had similar values
(Table 4.2). Similar to the present study, a significant difference in seasonal water use
was also reported for chickpea (Siddique and Sedgley, 1987) and faba bean over four
sowing dates (Loss et al., 1997) in the low rainfall Mediterranean environments. On the
other hand, Siddique et al. (2001) did not find any significant difference in seasonal water
80
use in a range of erect and prostrate gram legumes under rainfed conditions in the
Mediterranean environment which agreed with the present observation under well-
watered conditions.
600.- ~ 600.- -,
Beans 2002 Beans 2002/2003 .
500 ....... e 500 ""
____ MS r' ,
E 400 ·__ LS .' ,"
.§. 400
tu "
g! 300 300
;;
:ft;!
E 200 200
::::I
(.)
100 100
o +----r--~----~--~--~--~ 0+-----.-----,.----,----.....,------,-----1
o 20 40 60 80 100 120 o 20 40 60 80 100 120
600 ,..-- -, 600,..----------------,
Chickpea Chickpea .
500 500 "
.tt:,/
:,
Ê.§. 400 400
tu "
g! 300 300
;;
:ft;!
E 200 200
::::I
(.)
100 100
o +---~--~----~--,_--___._--~ o +----,------,----r--~--_r--~
o 20 40 60 80 100 120 o 20 40 60 80 100 120
600~-----------------------~ 600,..-- --,
Cowpea Cowpea
,
500 500
- ,..,
,.
,,
.E§. 400
tu .'
400
g! 300 300
;;
ft!
:E; 200 200
o::::I
100 100
o +----r----r---~--_.----~--~
o 20 40 60 80 100 120 0 20 40 60 80 100 120
Time after planting (days) Time after planting (days)
Figure 4.1. Seasonal cumulative ET of three grain legumes species under water stress (MS,
LS) and non-stress (C) conditions in 2002 (left) and 2002/2003 (right) seasons.
81
Table 4.1. Daytime mean vapour pressure deficit (kPa) above the canopy of three grain
legumes grown under three water regimes in three seasons for the period between
emergence to maturity.
Season Water regime Beans Chickpea Cowpea
200112002 C 1.60 1.62 1.61
MS 1.74 1.75 1.64
LS 1.56 1.75 1.67
2002 C 2.15 2.15 2.14
MS 2.11 2.02 2.09
LS 2.10 2.09 2.07
2002/2003 C 1.76 1.75 1.77
MS 1.76 1.75 1.82
LS 1.83 1.82 1.85
&0,- -.
.2002
.200212003
500
- 400-EE
S 300
.c~
(I)
::J
C. 200
e
0
:c=u
Cl) 100
'E
o
C-BN MS-BN LS-BN C-CHP MS-CHP LS-CHP C-COP MS-COP LS-COP
Treatments
Figure 4.2. Seasonal irrigation plus rainfall received by each of the three water regimes (C =
well-watered, MS = mid-season water stress, LS = late season water stress) and three grain
legumes (BN= beans; CHP = chickpea, COP = cowpea) in 2002 and 2002/2003 seasons in a
semi-arid environment.
There was no significant difference (P>O.05) among species for pre-flowering water use
(ETb) in all seasons (Table 4.2 and Appendix 5A) suggesting that vegetative stage water
use is similar for the three species under well-watered conditions. On the other hand,
differences in post-flowering water use (ETa) among the water regime treatments were
significant (P<O.05) in all seasons (Table 4.2 and Appendix 5A). The C and LS
82
treatments had the highest and the lowest post-flowering water use, respectively. The
corresponding values for the second and third seasons were 237.1 and 220.4 mm in the C
treatment and 139.6 and 123.7mm in the LS treatment, respectively. The large variation
in post-flowering water use between the water regime treatments is mainly due to the
effect of water deficit on canopy development and water availability in the stressed
treatments which influence soil cover by leaves and root water uptake, respectively
(Chapter 3 of this document; Siddique and Sedgley, 1987). Moreover, water stress
hastened physiological maturity in the LS treatment and resulted in lower seasonal water
use. The difference in post-flowering water use among species was significant (P<0.05) in
the first and second seasons but not in the third season (Table 4.2 and Appendix SA).
Pooled over water regimes, cowpea had the highest (189.1 mm) post-flowering water use
in the second season followed by beans (186.5 mm), whereas chickpea had the highest
(185.2 mm) in the third season followed by cowpea (174.4 mm). The lowest ETa,
however, was recorded for chickpea (162.2mm) in the second season and for beans (165.7
mm) in the third season. This shows the inconsistency of post-flowering water use by the
species over seasons which is mostly due to differences in canopy ground cover among
species which varied with season.
Water use by crops can be influenced by the stage of crop development and, thus,
information on water use for a given phenological period is important to adjust water
supply and other management practices (e.g. Simane, 1993). This study attempted to
determine the water use for vegetative and reproductive periods in three grain legume
species under three water regimes. Significant variation in the ratio of pre- to post-
flowering water use (ETb:ETa) was observed among water regime treatments but not
among species in both seasons (Table4.2). The ratio ranged from 0.61 to 1.11 in the
second season and from 0.93 to 1.73 in the third season. The ratio was the highest in the
LS treatment (1.04 in the second and 1.66 in the third seasons) while it was the lowest in
the C treatment (0.64 and 0.94 in the second and third seasons, respectively).
The lack of significant difference among species for pre- and post- flowering water use
under the C treatment indicate the similarity of the species potential to use the same
amount of water under non-limiting water supply conditions. However, the ratios of pre-
to post-flowering water use indicate that post-flowering water use in terminal drought
83
Table 4.2. Seasonal (ETs, mm), pre-flowering (ET b, mm), post flowering (ET a, mm) and ratio of pre- to post flowering water use (ETa:
ETb) and seasonal transpiration (Ts, mm) and soil evaporation (Es,mm) of three grain legume species for 2002 and 2002/2003 seasons,"
Water Regime Species 2002 2002/2003
(Wit) (Sp) ET. ETb ET. ETb:ET. Ts E. ET. ETb ETa ETb:ETs Ts Es
C BN 388.5 147.4 241.1 0.61 341.3 47.2 421.9 203.8 218.1 0.93 264.5 157.5
ClIP 375.3 146.4 228.9 0.64 201.5 173.8 430.1 207.4 222.7 0.93 226.4 203.7
COP 402.6 161.2 241.4 0.67 364.0 38.7 426.2 205.7 220.6 0.97 246:5 179.7
MS BN 321.1 138.5 182.6 0.76 94.1 227.1 354.2 196.4 157.7 1.43 173.7 180.5
ClIP 244.9 128.8 116.1 1.11 125.8 119.1 416.5 203.1 213.2 1.03 165.1 251.5
COP 332.6 148.1 184.5 0.80 164.9 167.7 381.1 206.9 174.2 1.27 255.8 125.4
LS BN 278.9 143.1 135.8 1.10 245.8 33.2 287.2 196.2 121.3 1.63 193.7 93.4
ClIP 288.9 147.4 141.5 1.04 160.7 128.3 321.3 201.5 119.8 1.73 202.9 118.4
COP 283.9 142.4 141.5 1.01 164.1 119.8 326.3 197.9 128.4 1.60 196.6 129.7
LSD (PitO.05) WR 14.3** n.s. 13.5** 0.01** 33.9** 33.9** 36.5** n.s. 38.3** 0.39* 17.4** 32.3**
Sp 10.2** n.s. 10.7* n.s. 25.6** 23.6** 26.4* n.s. n.s. n.s. n.s. 36.1*
WRxSp n.s. n.s n.s n.s 40.8** 40.8** n.s. n.s. n.s n.s. 5.25* 62.5*
CV(%) 10.0 5.5 10.0 0.84 1l.l 19.6 6.9 6.7 10.6 14.4 13.8 21.9
*, **: Significant at 5 and 1% probability level respectively; s: Soil water was measured to a depth of 60 cm. BN: beans; ClIP: chickpea; COP: cowpea.
84
conditions was much lower than the one under favorable water supply conditions in all
species in both seasons as expected. Although ETa is slightly higher than ETb in the high
temperature season (the second season), the ETb to ETa ratio of the three species was
generally closer to unity under well-watered conditions indicating similar water use in the
vegetative and reproductive growth stages of the species. Higher values of the ratio in
2002/2003 than in 2002 season indicate that post-flowering water use is dependent not
only on water supply but also on other environmental conditions of the season such as air
temperature and humidity. In the Mediterranean environment, the ratio of pre- to post-
flowering water use was found to be 1.3 or less for early sown faba bean (Loss et al.,
1997), and it ranged from 1.19 to 1.71 for erect- and from 1.27 to 2.27 for prostrate- grain
legume species under rainfed conditions (Siddique et al., 2001). The values found in the
present study under the water stress treatments are within the range of the reported values.
The ratios reported here for the three species are useful to adjust water supply based on
phenological stages in different water supply environments.
4.3.3. Water use efficiency (WUE)
The pre-flowering water use efficiency (WUEb) ranged from 17.9-18.8 kg ha" mm-I in
the second season and from 10.9-13.4 kg ha" mm-I in the third season (Table 4.3).
Differences in WUEb were not significant among species (Table 4.3). Post-flowering
water use efficiency (WUEa) was higher than pre-flowering water use efficiency (WUEb)
in all seasons. This is explained by higher water loss through soil surface evaporation
- -
than plant transpiration during the early period of the vegetative growth compared to the
post-flowering period when the canopy fully covers the ground and reduces soil
evaporation. Pre-flowering water use efficiency (WUEa) was also higher than seasonal
total above ground dry matter (WUEd) and grain yield (WUEg) water use efficiency in
both seasons which could be due to dry matter loss by respiration until harvest during
which vegetative growth had already ceased. It is important to note that respiration
consumes up to 50% of the photosynthate produced and thereby lowers the seasonal
WUE as compared to the WUE calculated on the short-term basis (Ong et al., 1996).
Difference in WUEa was significant among species in the second season and among water
regimes in the third season (Table 4.3). In the second season, the WUEa ranged from 17.1
kg ha" mm-I in the MS to 49.9 kg ha-I mm" in the LS treatments in beans and in the third
season, it ranged from 11.9 kg ha-I mm" in chickpea in the MS to 37.8 kg ha" mm" in
85
beans in the C treatment. The WUEa in the third season was significantly (P<0.05)
reduced by the MS treatment in all species when compared to the C and LS treatments
which had similar values (Table 4.3).
Table 4.3. Water use efficiency (kg ha" mm") for pre-flowering (WUEb), post flowering
WUEa), above ground dry matter at harvest (WUEd) and grain yield (WUEg) and
transpiration efficiency for grain yield (TEg, g mm") of three grain legume species for 2002
and 2002/2003 seasons.
Water Species 2002 . 200212003
regime (Sp) WUE" WUF..! WUEg TEg wuft, WUE" WUF..! TEg
(WR)
C BN 18.5 29.6 14.2 6.0 0.69 12.8 37.8 14.8 5.7 1.14
CHP 18.5 29.7 9.9 3.2 0.58 10.9 25.0 12.4 4.7 0.86
COP 18.5 26.2 12.2 3.3 0.37 13.4 31.2 13.6 5.8 0.77
MS BN 18.5 17.1 4.8 0.6 0.22 12.3 14.3 11.1 3.9 1.10
CHP 18.7 34.1 10.0 2.0 0.41 11.6 11.9 9.0 3.2 0.63
COP 17.9 26.1 6.8 1.6 0.31 12.5 13.7 12.2 5.1 0.71
LS BN 18.3 49.9 14.5 6.2 0.71 11.9 29.3 14.0 6.0 1.13
CHP 18.4 36.4 10.5 3.8 0.70 12.2 35.2 14.7 5.5 1.02
COP 18.8 28.4 13.5 2.6 0.46 1l.3 33.1 14.3 5.3 0.76
LSD WR n.s. n.s. 1.7·· 1.1.. 0.15" n.s. 9.3·· 1.8· l.I· n.s.
(DIID.OS) SP n.s. 5.6· 1.3. 0.5"· 0.07·· n.s. n.s. n.s. n.s. 0.12
WRxSp n.s. 9.7"· 2.3"· 1.9"· 0.14·· n.s. n.s. n.s. n.s. n.s
CV(%) 3.0 17.7 12.7 16.0 15.9 13.3 28.7 16.6 18.3 12.5
., .": Significant at 5 and I% probability level respectively; ': Soil water was measured to a depth of 600 mm. BN: beans; CHP:
chickpea; COP: cowpea.
WUEdwas significantly (P<0.05) affected by the water regime treatments in both seasons
(Table 4.3). The values ranged from 4.8-14.5 kg ha" mm" (between MS and LS
treatment in beans) in the second season and from 9.0-14.8 kg ha-I mm" (between MS
and C treatments in chickpea) in the third season (Table 4.3). Among the water regimes,
the lowest WUEd was recorded in the MS treatment while the highest was recorded in the
LS treatments in both seasons for all species except in chickpea in the second season. The
_ lowest WUE recorded in the MS treatment is a result of low dry matter production and
possibly a high soil surface evaporation resulting from reduction in leaf area in contrast to
the LS treatment where the plants mature before a significant loss of foliage. Beans and
cowpea have similar WUEd across the water regimes in both seasons. Although chickpea
had a higher or similar value of WUEd to that of beans and cowpea in a few cases, it
generally tended to have lower values across the water regimes in both seasons. This
shows that beans and cowpea are more efficient than chickpea in producing dry matter
under water limited environments. Therefore, WUEd can be used as a criterion to select
crops for water limited environments.
86
The WUEg differed significantly (P<0.05) among the water regime treatments in both
seasons (2nd and 3rd) and among the species in only in the second season (Table 4.4). The
average WUEg were 4.1, 1.4 and 4.2 kg ha-I mm-I in the second season and 5.4, 4.1, and
5.6 kg ha-I mm" in the third season for the C, MS and LS treatments, respectively
indicating that the MS treatment had the lowest WUEg. Among the species, beans had the
highest (4.3 kg ha" mm") and cowpea had the lowest (2.4 kg ha-I mm") WUEg in the
second season while beans and cowpea had the highest (5.3 kg ha" mm") and chickpea
had the lowest (4.8 kg ha-I mm") WUEg in the third season. The WUEd and WUEg found
in the present study are comparable to previous reports in chickpea under drought stress
and irrigation conditions (Silim and Saxena, 1993a), lentil (Silim et al., 1993b), and
mungbean (Pannu and Singh, 1993) under rainfed and irrigated conditions. In many
environments, chickpea is reported to use 100-450 mm of water in a season depending on
yield produced with a WUEd ranging from 5.2-35.2 kg ha" mm" and a WUEg from 1.1-
15.7 kg ha" mm" (Sandhu et al., 1978; Singh and Bhushan, 1980; Sivakumar and Singh,
1987; Siddique and Sidgley, 1987; Keatinge and Cooper, 1983; Singh and Virmani, 1990;
Silim and Saxena 1993a). The current WUEd are also within the range of values reported
for beans (8.5-24.8 kg ha-I mm") over different seasons and population densities (Tsubo
et al., 2003). The present WUEd recorded are generally lower than the values reported for
a number of erect- (up to 30 kg ha" mm") and prostrate- (up to 38 kg ha" mm") grain
legumes in the Mediterranean environment (Siddique et al., 2001) suggesting genotypic
differences in WUE. However, the values reported for WUEg in the same study, with the
exception of fababean and pea, are within the range of values found in the present study.
The current values are far below the maximum WUEd (30-36 kg ha-I mm") reported for
V.faba and V. narbonensis and WUEg(14-16 kg ha-I mm") for V.faba and P. sativum in
the Mediterranean environments (Loss et al., 1997; Siddique et al., 2001). This indicates
that the species in the current study have lower maximum WUE than other grain legume
species such as faba bean and field pea, suggesting that the current species may a have a
lower potential for water use than the mentioned cool-season food legumes.
4.3.4. Seasonal transpiration and transpiration efficiency (TE)
The calculated seasonal transpiration (Ts) values were significantly different among water
regime treatments (P<O.Ol) in both the second and third seasons and among species in the
second season (P<0.05) (Table 4.2). The seasonal transpiration ranged from 94.1 to 364.0
mm in the second season and from 165.1 to 264.5 mm in the third season. Pooled over the
87
species, the seasonal transpiration values were 302.3, 128.2, and 190.2 mm in the second
season and 245.8, 198.2 and 197.8 mm in the third season for the C, MS and LS
treatments, respectively. These transpiration values comprised 78, 43 and 67% of the ETs
in the second season and 58, 52 and 63% in the third season for the C, MS and LS
treatments, respectively. Subtraction of T, from ETs shows that water loss due to soil
surface evaporation can be as high as 57% in the mid-season drought prone semi-arid
areas and as low as 22% in those semi-arid areas where water supply is favourable.
Pilbeam et al. (1995) found that 49-83% of the rainfall in the semi-arid area of Kenya was
lost as soil surface evaporation, and the seasonal average transpiration was only 23% of
the evapotranspiration. Cooper et al. (1987) estimated soil surface evaporation in the
range of 40-50% of the seasonal crop evaporanspiration, which is within the range of the
present values found in a semi-arid environment. The pattern of water use in the form of
transpiration among the species was different over the seasons. Beans and cowpea had the
highest transpiration in the second season whereas cowpea had the highest in the third
season. The transpiration in beans, chickpea and cowpea constitute 69,54 and 68% of the
ETs in the second season and 59, 51 and 62% of the ETs in the third season, respectively.
In most of the cases, soil evaporation was the highest in chickpea across the water
regimes (Table 4.2) mainly because of its slow leaf area growth (Chapter 3) and poor
ground cover. Therefore, management practices that increase ground cover and reduce the
soil surface evaporation are expected to increase the WUE of this crop. Except for a few
instances, beans and cowpea had similar transpiration values in both seasons under both
well-watered and water stress conditions.
The TE calculated based on grain yield (TEg) was significant (P<0.05) among the water
regime treatments in the second season but not in the third (Table 4.3). The highest TEg
was recorded in the LS while the lowest was recorded in the MS. The values ranged from
0.22-0.71 g mm" in the second season and from 0.63-1.14 g mm" in the third season. TEg
values for grain legumes are not available in the literature. However, the TE values (based
on total dry matter) reported in other studies include 1.89-2.33 g mm" for chickpea
(Cooper et al., 1988), 2.21-3.57 g mm-I for cowpea (Ashok et al., 1999),2.2-3.7 g mm"
for beans (Pilbeam et al., 1995; Ogindo, 2003), and 1.5-5.2 g mm" for groundnut (Ong et
al., 1987; Matthews et al., 1988; Azam-Ali et al., 1989). The LS treatments had slightly
higher TEg values as compared to the control in 2002 which cou1d be explained by
reductions in the mean transpiration rate (Ashok et al., 1999) as a result of partial
88
stomatal closure while the lower values in the MS stressed treatments could be explained
by reductions in transpiration as a result of high soil surface evaporation (Table 4.3).
4.3.5. Relationship between harvest index, water use and water use efficiency
The correlation between HI, water use and WUE parameters in the second and third
seasons is shown in Table 4.4. In beans, correlations among the water use and WUE
parameters were not strong under well-watered conditions except a significant (P<O.05)
positive correlation between WUEd and WUEa. HI was also not strongly correlated with
any of the water use and WUE parameters under the control water regime. Except
significant positive correlations between WUEd and WUEa and HI and WUEg, similar
conditions were observed in the LS treatments. However, WUEd, WUEg and HI were
significantly positively correlated with ETb and negatively with WUEb in the MS
treatment (Table 4.4). This suggests the importance of high water use during the
vegetative stage for high WUE and HI at harvest in mid-season drought stressed beans.
Moreover, WUEg was strongly (P<O.Ol) correlated with WUEd, and HI was significantly
and positively correlated with both WUEd and WUEg. On the other hand, WUEa and
WUEg are important for high HI of beans in terminal drought environments. Thus, high
reproductive period WUE is an important mechanism that enables the plant to maximize
dry matter production under limited water supply during the period of grain growth.
In chickpea, correlations among the water use and WUE parameters and HI were weak
except that HI was positively significantly (P<O.05) correlated with WUEg in the MS and
LS treatments, and WUEg was positively correlated with WUEb in the C treatment. Weak
correlations among water use, WUE and HI across the water regime treatments in
chickpea suggest that HI in this crop may not depend on water use and water use
efficiency.
On the other hand, for cowpea WUEd and HI in the C treatment and WUEg and HI in the
LS treatment were significantly positively correlated with vegetative stage water use
(ETb). WUEg and HI were significantly negatively correlated with WUEb in both the MS
and LS treatments of cowpea (Table 4.4). In addition, WUEg was positively correlated
with WUEd in both the MS and LS treatments. HI was positively correlated (P<O.05) with
WUEg in the water regimes but the correlation was significant only in LS treatments. HI
was significantly negatively correlated with WUEd in the C treatment while it had weak
89
positive correlation in the MS and LS treatments. The results in cowpea indicated that
high WUEd and HI at harvest are achievable through high water use during the vegetative
stage under well-watered and mid-season water stress conditions which could result in
high dry matter and seed yield at harvest. Therefore, management practices in mid-season
drought environments should focus on increasing the water use of cowpea during the
vegetative period in order to increase HI of the crop.
Table 4.4. Correlation coefficients for water use, water use efficiency and HI in three
gram. Iegumes. +
Species Water ETb ETa WUEb WUEa WUEd WUEg
regime
------(mm)------ ----------------kg ha' mm' ---------------
Beans C WUEd -0.52 -0.76 0.70 0.84*
WUEg -0.14 -0.23 0.29 0.42 0.80
HI 0.30 0.20 -0.15 0.11 0.15 0.45
MS WUEd 0.89* 0.67 -0.84 -0.31
WUEg 0.95** 0.74 -0.92** -0.27 0.97**
HI 0.98** 0.82* -0.99** -0.34 0.87* 0.96**
LS WUEd -0.69 -0.53 0.77 0.91 **
WUEg 0.21 -0.03 0.32 0.64 0.65
HI -0.16 0.52 0.26 0.52 0.42 0.89*
Chickpea C WUEd 0.09 0.00 -0.05 0.45
WUEg 0.04 0.04 0.18 0.35 0.84*
HI 0.41 0.44 -0.09 -0.35 0.32 0.68
MS WUEd -0.41 -0.60 0.53 0.20
WUEg 0.22 0.08 0.09 0.30 0.22
HI 0.37 0.32 -0.16 0.25 -0.44 0.81*
LS WUEd 0.17 0.30 -0.50 0.10
WUEg -0.36 -0.15 0.45 0.38 0.68
HI -0.56 -0.31 0.64 0.48 0.18 0.81 *
Cowpea C WUEd 0.95** -0.44 0.99*** 0.27
WUEg 0.16 0.61 -0.04 0.26 -0.09
HI 0.85* 0.40 -0.83* -0.12 -0.86* 0.59
MS WUEd 0.41 -0.37 -0.49 -0.51
WUEg 0.77 -0.30 -0.84* -0.77 0.88*
HI 0.97* -0.05 -0.98** -0.80 0.37 0.77
LS WUEd 0.62 0.17 -0.54 -0.34
WUEg 0.88* -0.03 -0.83* 0.19 0.85*
HI 0.88* -0.16 -0.88* 0.65 0.40 0.82*
+ n = 6 (three observations and two seasons).
*, **, *** = values significant at 5, 1 and 0.1% P levels.
Comparison of species under well-watered condition indicate that high WUEd at harvest
was strongly positively correlated with high WUEa in beans and high WUEb in cowpea
while it is not dependent on any of the water use parameters in chickpea. High HI was
strongly positively associated with high WUEg in all species under terminal water stress
as well as in chickpea under mid-season stress.
90
WUEg and HI were significantly negatively correlated with WUEb in the MS treatment in
beans and in the MS and LS treatments in cowpea but significantly positively correlated
with ET b in the same species and treatments indicating the importance of high vegetative
period water use for high HI and grain yield. However, although maximum utilization of
water is important when it is available for high dry matter production, it will cause early
depletion of soil water and will have a negative effect on yield if there is water shortage
during the reproductive period. As a result the advantage of high vegetative water use
depends on the availability of water later in the season. Comparison of the water regime
treatments over species indicated that HI was negatively correlated with WUEa in the MS
treatment (r = -0.34 to -0.51) in beans and cowpea and positively (r = 0.48 to 0.65) in the
LS treatment in all species while the relationship was weak in the C treatment (r = -0.35
to 0.11). This shows that water applied after stress in the MS treatment is mainly utilized
in the production of vegetative matter rather than contributing to grain yield. This was
observed in the field when plants of beans and cowpea started producing new leaves and
flowers upon re-watering while previously developed pods dried out faster. When pooled
over the species, WUEg was strongly correlated with HI in all water regimes and
explained 60, 81 and 82% of the variability in HI in the C, MS and LS treatments,
respectively suggesting that high grain WUE of a crop is important in increasing HI of
grain legumes in different environments with respect to water supply. As indicated by
Passioura (1977) for cereal crops, the present results show that HI is strongly correlated
with grain yield in all water regimes (Chapter 7) but the association of HI with the
components of WUE in grain legumes is environment dependent. For example, selection
for high reproductive period water use and WUE could result in high HI in terminal
drought environments in beans and cowpea unlike mid-season drought and high rainfall
environments where selection for the same may not bring any success.
HI depends largely on the relative proportion of pre- and post-flowering dry matter and
the mobilization of pre-flowering assimilates to the grain (Ludlowand Muchow, 1990;
Chapter 3), and is considered as a potential source of yield stability in terminal-drought
environments (Turner et al., 2001). In crops like chickpea that are usually grown on
stored soil water, HI is related to the amount of water available after flowering (Passioura,
1977), and greater water use in the post flowering period was associated with higher HI
and WUEg in grain legumes (Siddique et al., 2001). Therefore, as shown in the present
91
work and stated by Ludlowand Muchow (1990), the study of HI in relation to water use
and use efficiency is the best avenue for improving grain yield of crops by increasing the
amount of water transpired so as to maintain high HI and yield.
4.4. Conclusion
Water use and water use efficiency are important components in the study of resource
utilization by plant species. The present study indicated that water use and its efficiency
were greatly influenced by the time (growth stage) and amount of water supply in grain
legumes. The mid-season water stress treatment had intermediate seasonal water use but
the lowest crop WUE resulting from low leaf area index, high soil surface evaporation,
(particularly in chickpea) and reduced sink size. Therefore, crop management practices
that decrease soil evaporation are expected to increase the WUE of these crops under
mid-season drought environments. On the other hand, water use was much higher for the
well-watered treatment compared to the late season stress treatment while the two
treatments had similar WUE values at harvest. Therefore, in water limited environments,
applying a certain period of water stress towards the end of the growing season is
advantageous inincreasing water saving while maintaining high WUE. Among species,
beans and cowpea had similar WUE values while chickpea had the lowest record during
all the seasons. Therefore, grain legume species that efficiently utilize the available water,
such as beans and cowpea, are important in drought prone crop growing environments.
The information generated in this study will be valuable in modeling the water use and
WUE of the species for the semi-arid regions and also in facilitating crop choice for a
given environment regarding water supply.
92
CHAPTERS
Resource Utilization of Three Grain Legume Species in a Semi-Arid Environment.
Il,Canopy Development, Radiation Interception and Radiation Use Efficiency
5.1. Introduction
Under non-stressed environmental conditions, the amount of dry mass produced by a crop
is linearly related to the amount of solar radiation (SR), specifically photosynthetically
active radiation (PAR), intercepted by the crop (RI) (Monteith, 1977a; Gallagher and
Biscoe, 1978; Russell et al., 1989). The slope of the linear regression between biomass
and cumulative radiation intercepted by a crop has been used to determine the radiation
use efficiency (RUE) (e.g. Muchow et al., 1993; Muchow and Sinclair, 1994; Ceotto and
Castelli, 2002). This relationship is also employed to develop simple crop models. For
example, Monteith (1977a) and Russell et al. (1989) expressed yield (Y) as a function of
RI, RUE and harvest index (HI) as shown in equation 1.6.
Radiation interception is variable throughout the crop growing period (Sivakumar and
Virmani, 1984; Watiki et al., 1993) influenced mainly by the green leaf area duration and
canopy extinction coefficient (K) (Thomson and Siddique, 1997; Jeuffroy and Ney,
1997). Many studies on grain legumes indicated the variability of K values for a given
species resulting from the effect of environmental constraints (like drought) on its canopy
through the modification of angle, spatial distribution and optical properties of leaves
(Jeuffroy and Ney, 1997). These indicate that radiation interception and use are
influenced by both genetic and environmental factors. Comparisons of species with
respect to photosynthetic processes indicate that C4 species have higher RUE than C3
species, and within C3 species non-leguminous C3 species have higher RUE than
leguminous species (Goss et al., 1986). Large variation in RUE was also reported among
grain legume species mainly due to a variety of environmental conditions (Sinclair and
Muchow, 1999). Reductions in RUE due to water deficits have been reported in many
studies on grain legumes (e.g. Hughes and Keatinge, 1983; Muchow, 1985a; Green et al.,
1985; Singh and Sri Rama, 1989).
RUE is a major component of the radiation-based crop growth models, which integrate
several developmental, morphological, physiological, and biochemical processes at a
higher level of plant functions (Turner et al., 2001). Therefore, RUE can be used to
93
evaluate crop performance and yield limitations under different seasonal and climatic
conditions (Sinclair and Muchow, 1999). Since the RUE of grain legumes (with low
energy content) is lower than that of other non-leguminous C3 species (Sinclair and
Muchow, 1999), further research is needed to investigate whether the reported low value
is due to the inherent characteristic of these species or the result of environmental effects.
Moreover, little information on the radiation interception and use of grain legumes are
available for the semi-arid regions in Ethiopia where drought is a major crop yield
limiting factor. This study was, therefore, initiated to compare radiation capture and
utilization in common bean (Phaseolus vulgaris L.), chickpea (Cicer arietinum L.) and
cowpea (Vigna anguiculata L.) grown under different water regimes at a semi-arid region
in Ethiopia.
5.2. Materials and Methods
5.2.1. Field experiments
Detail explanations of the experimental site, experimental material, experimental design,
cultural practices and irrigation schedule are given in Chapter 3 and will only be
explained here in brief. Three field experiments were conducted at Dire Dawa, Ethiopia
during the periods from early December 2001 to late March 2002 (first season), from late
March to early July 2002 (second season) and from mid October 2002 to early February
2003 (third season). Radiation data was collected in the second and third seasons only.
Seeds of common bean (cv. Roba-1), chickpea (cv. ICC-4958) and cowpea (ev. black eye
bean) were planted on March 27, 2002 and October 17, 2002 for the second and third
seasons, respectively. The experiments had three water regime treatments as shown in
Table 3.1. The experimental treatments, replicated three times, were arranged in a
randomized split plot design using the stress treatments as main plot and the crop species
as sub-plot. The total experimental area was 22.8 m x 40.2 m.
5.2.2. Experimental measurements
Incident and transmitted PAR (0.4 to 0.7 urn wavelength) were measured using the SunScan
Canopy Analysis System, BF2 type (Delta-T Devices, U.K.) at 10 day intervals from 17 and
20 days after planting (DAP) in the second and third seasons, respectively until
physiological maturity. The instrument has a single quantum sensor and a one-meter long
probe with 64 photodiodes equally spaced along its length. While the former measures
94
incident PAR above the canopy, the latter measures transmitted PAR beneath the canopy
placed at right angle to the crop rows on the soil surface. The PAR measurement was taken
between 12:00 to 13:30 local time (GMT+3) for a period of about 5 minutes in each plot.
In the second season, daily total incident SR was determined using the Angstrom's equation
(Angstrom, 1924) as:
SR = SRc (a + b nIN) (MJ m-2 d-I) (5.1)
where SRo is extraterrestrial radiation (MJ m-2 d-I), N is maximum possible sunshine
duration (hour), n is measured sunshine duration (hour) and a and b are constants. The data
for n was obtained from the class A weather station at Dire Dawa International Airport
(latitude 9°36' N, longitude 41°51' E, and altitude 1260m above sea level; 500m away from
the experimental area). In the third season, SR was measured by an automatic weather
station (CMlO, Campbell Scientific, USA) installed adjacent to the experimental field. The
measured SR and n in 2002/2003 were used to determine a and b for the experimental site.
The values of a and b found by a linear regression of SRlSRc vs nIN were 0.212 and 0.479
respectively with R2 = 0.94. These figures are very close to the values recommended by
Allen et al. (1998) for areas where measured SR data is not available. Daily PAR was
determined by multiplying the value of SR by 0.5 (Campbell and Norman, 1989; Monteith
and Unsworth, 1990). Cumulative intercepted PAR for each plot was calculated by
multiplying daily PAR by the value of the fractional interception (F) obtained on each
measurement date and by the slope of the regression line (F versus time) for the period
between measurement dates, and finally by summing up all the daily values in the season.
Above ground dry matter (ADM) was measured from five plants (0.2 m2 area) per plot at
intervals of 10 days starting at 17 and 20 DAP for the second and third seasons, respectively.
The plant samples were separated into leaf, stem, pod and seed and dried in an oven for 72
hours at 60°C. After drying, each component was weighed separately and the total above
ground dry matter was obtained by adding the weight of each component. Leaf area was
measured throughout the growing period using a portable leaf area meter (Model CI-202,
CID Inc., USA) from five plants per plot. Leaf area index (LAl) was calculated as the ratio
of total leaf area to area of ground. The final harvest took place between Jun 7 and July 4
and between January 13 and February 5 in the second and the third seasons, respectively
for the different treatments.
95
A linear regression of accumulated biomass versus cumulative PAR was used to calculate
seasonal RUE. Canopy extinction coefficient (K) was determined from the slope of the
regression line between the natural logarithm of PAR transmission and leaf area index
(LAl) (Monteith, 1965) assuming zero interception.
5.3. Results and Discussion
5.3.1. Leaf area index and fractional PAR interception
The two growing seasons were characterized by similar pattern of LAl development (Fig.
5.1). The second season, however, had a higher peak than the third season. In all the species,
maximum LAl was achieved between 50 and 60 DAP in the second season and between 60
and 70 DAP in the third season. Once the maximum value was attained, LAl declined in
both seasons. However, the rate of LAl decline was not similar across the species; for
example, the rate of LAl decline was faster in beans and cowpea than in chickpea in the C
treatment. LAl development was similar among the water regimes until commencement of
stress. Greater reduction in LAl was observed when the plants were stressed during the
flowering period (MS). However, an increase in LAl was observed in beans for the MS
treatment after the plants were re-watered and recovered from the stress in both seasons.
This adjustment of canopy development is one type of developmental plasticity which is
important for the crop to perform well in dry environments. The importance of
developmental plasticity towards better adaptation to water deficits was reported for many
crops, including beans (Singh and White, 1988), chickpea (Saxena et al. 1993a), and
groundnut (Harris et al., 1988). Compared to the MS, the rate of LAl decline due to water
stress was faster in the LS treatment in all species and seasons mainly due to the shedding of
leaves. Muchow et al. (1993) reported high rate ofleaf shedding and decline of specific leaf
nitrogen in soybean, mungbean and cowpea under rainfed condition. Comparison of species
indicated that beans had the highest maximum LAl in both seasons followed by cowpea and
chickpea in that order.
Seasonal fractional PAR interception (F) was similar over the two seasons (Fig. 5.2). The
fraction of PAR intercepted in the MS treatment never reached the same maximum value as
the C treatment, and the difference was slightly higher in the second than in the third season.
However, compared to the MS treatment, the difference in fractional PAR interception
between the C and LS treatments was small, indicating that water stress in early reproductive
stage is more important in reducing the fraction of PAR intercepted by the respective crops
Figure 5.1. Seasonal course of leaf area index (LAl) in beans, chickpea and cowpea under
mid-season (MS) and late season (LS) water stresses and well-watered (C) conditions in
2002 (left) and 2002/2003 (right) seasons.
97
fraction of PAR intercepted in the MS treatment could be attributed to the effect of water
stress on leaf area development (Lecoeur et al., 1995; Blum, 1996) and canopy extinction
coefficient (Jeuffroy and Ney, 1997). Since dry matter production is linearly related to PAR
interception, low post-flowering dry matter production in the MS treatment (Chapter 3) is
mainly attributed to low PAR interception.
5.3.2. Canopy extinction coefficient
The fitted regression lines between the natural logarithm of transmitted PAR and LAl are
shown in Fig. 5.3 (combining the two seasons). The regression lines were forced through
the origin assuming that I = lo when LAl = O. The slopes of the regression lines were
tested for similarity using the t-test. The K values were generally higher in the non-
stressed than in the stressed treatments and in the third than in the second season (data not
shown). The K values for beans, chickpea and cowpea in the C treatment were found to
be 0.84, 1.02, and 0.86, respectively (Fig. 5.3). As compared to the C and LS treatments,
the MS treatment resulted in the lowest values of K (0.45, 0.63 and 0.53 for bean,
chickpea and cowpea, respectively). Intermediate values of K between the C and MS
treatments were recorded in the LS treatment. In comparing the species under well-
watered conditions, chickpea had the highest K value followed by cowpea and beans. K is
a function of leaf size and orientation (Saeki, 1960) ranging from 0.3 to 1.3 where K less
than 1 refers to non-horizontal or clumped leaf distributions and K greater than 1 refers
horizontal or regular leaf distributions (Szeicz, 1974, Jones, 1992). In the present study,
therefore, chickpea had more horizontal leaves than beans and cowpea in the C treatment.
There was significant difference in K between C and MS and between MS and LS for
beans (Table 5.1). As shown in Table 5.1, difference between C and MS was also
significant (P<0.05) in chickpea and cowpea whereas differences between C and LS and
between MS and LS were not significant (P>0.05). The lower values of K recorded in the
MS treatment cou1d be attributed to the modification of leaf angle and orientation by the
water deficit (Jeuffroy and Ney, 1997). The values of K observed in the C treatment are
higher than previous reports on beans (0.4 by Gardner et al., 1979 and 0.64 by Tusbo et
al., 2001), chickpea (0.4-0.61 by Hughes et al., 1987), pea (0.33-0.49 by Heath and
Hebblethwaite, 1985) but within the range of values reported by Thomson and Siddique
98
2002 Bean . 'n, 2002/2003
0.8 "\ [3··El 0.8~ .. [!)
\
0.6 \, ~ \
\ 0.6 \
'Q_o \
\
0.4 0.4 \b
0.2 0.2
0.0 -I----r--~--,.__-_r_-__,__-__l 0.0 -I---=r----,--_r_----,--,.__--f
o 20 40 60 80 100 120 o 20 40 60 80 100 120
1.0 -r-r- --,
Chickpea Chickpea
0.8 0.8
0.6 0.6
u..
0.4 0.4
0.2 0.2
0.0 -1----,------,---r----,-----,---I 0.0 -I----T---r---...,.-----,--,.__---1
o 20 40 60 80 100 120 o 20 40 60 80 100 120
Cowpea Cowpea
0.8 ,,0.8
co
0.6 0.6 "'<il
\
u.. \\
0.4 0.4 \ \
\
(!)
0.2 0.2
'/
0.0 -I-----.---r-----r-----,--...,.--~ 0.0 -I-----.--..-----r------,----,---....,
o 20 40 60 80 100 120 0 20 40 60 80 100 120
Time after planting (days) Time after planting (days)
Figure 5.2. Measured fraction of PAR intercepted (F) in beans, chickpea and cowpea under
mid-season (MS) and late season (LS) water stress and well-watered (C) conditions during
2002 (left and 2002/2003 (right) seasons.
99
LAl
o 1 2 3 4 5 6
o +-Sn-=~~-----L------~----~------~----~
~ Bo Beans
-1 0 --~--~ 0 0 l:i
-;:;'0 -Ar_
""J(J - __
- ~ --_ --0LI....I.. -3 l:i
DC (K = 0.84, R2 = 0.91)
-4
o MS (K = 0.45, R2 = 0.77) o
-5 l:i LS (K = 0.81, R2 = 0.84)
-6 ~ _J
o 1 2 3 4 5 6
o ~a-~--~----~------~------~------~----~
Chickpea
-1
-2
ii:"
-...I.. -3 o '.c ~"i3- ..
-4 DC (K =1.02, R2 = 0.89) •o
o MS (K =0.63, R2 = 0.64)
-5
l:i LS (K =0.89, R2 = 0.89)
-6
o 1 2 3 4 5 6
o
-1 Cow pea
- -2LI..
-...I.. -3c DC (K =0.86, R2 = 0.80)
-4
o MS (K = 0.53, R2 = 0.84)
-5
l:i LS (K = 0.75, R2 = 0.81)
-6 ~--------------------------------------------~
Figure 5.3. IDustration of canopy extinction coefficient (K) for beans, chickpea and cowpea
under mid-season (MS ---) and late season (LS -) water stress and well-watered (C --)
conditions for data combined over two seasons (LAl = leaf area index,
F = fractional PAR interception).
(1997) for many grain legumes. The previously reported K value for cowpea was 0.93
(Varlet-Grancher and Bonhommme, 1989 in Jeuffroy and Ney, 1997) which is higher
than the value found in this study. As indicated in some reports (Jeuffroy and Ney, 1997;
Thomson and Siddique, 1997), K varies depending on environmental differences such a
water deficit and season and crop species.
100
Table 5.1. Test of homogeneity of regression coefficients for K and RUE pooled over the two
seasons.
Species Water regime K* RUE (g M.r!)*
C a a
Bean MS b b
LS a ab
C a a
Chickpea MS b b
LS ab ab
C a a
Cowpea MS b a
LS ab a
* Similar letterswithin species in a colwnn are not significant(P> 0.05). Values are
shown in the respectivegraphs.
Compared to the respective controls, the highest reduction in K (46%) was observed for
beans in the MS treatment whereas the reduction in chickpea and cowpea was similar (38%).
The lowest value of K in the MS treatment showed better canopy adjustment (such as leaf
movement) of the crops in response to water deficit which is an important mechanism by
which grain legumes adapt to drought stress (Begg, 1980). Unlike the MS treatment,
however, beans had higher K value in the LS treatment when compared with chickpea and
cowpea indicating its poor leaf adjustment to water deficit occurring late in the season. Such
differences among species in canopy adjustment to the timing of water stress suggest that K
could be used as a selection criterion in grain legumes to identify cultivars that are capable of
adjusting their canopy in response to water deficit at different stages of growth.
5.3.3. Dry matter production and interception of PAR
As shown in Chapter 3, above ground dry matter (ADM) production was higher in
2002/2003 than in 2002 season because of longer growing period in the third season. Under
non-limiting water conditions, ADM production was similar for all species in the second
season but it was higher for beans and cowpea in the third season. In all the species, ADM
increase was fastest during the initial exponential growth phase, as expected, and declined
towards the end of the growing season under well-watered conditions. Compared to the C
treatment, ADM production was severely affected by the MS treatment in all of the species
in both seasons (Chapter 3). Reduction in ADM due to the LS treatment was variable
between species and seasons. The reduction in ADM of beans and chickpea due to the LS
was higher in the third season than in the second season, in contrast the reduction in cowpea
was smaller in the third season than in the second season (Appendix 4B & C).
101
700 .- ~ 700~------------------------~
600 Beans .[] 2002 600 2002/2003
e-te:: 500
JJj
••• (;J.. •• C 500
a.
400 _-O-_MS
.'
"C 400
.! --er--LS
Q...B 300 300.41. 200 200
oE 100 100
0+--4~~~--~--~--~--~ O+---~--~--_,----r_--._--~
o 20 40 60 80 100 120 0 20 40 60 80 100 120
700 ,- ~ 700~------------------------~
600 Chickpea 600 Chickpea
<et:: 500 500a.
"C 400 400
.!
.Q.. 300 300.8. 200 20041
oE 100 100
0-t---fiT"'-=---,-----,----.----,------iO+----li;I:----r-------.-----.-----.---~
o 20 40 60 80 100 120 o 20 40 60 80 100 120
700 ,- -, 700 -,-- ---,
600 Cowpea .IJ 600 Cowpea
et:: 500 .rif· 500
~
.. 400 400"C41
Q. 300 300
B.. 200 200
.c! 100 100
o +-__~~~_.__--__,_--_,----_,__--~ O+----f!II:---_,__---,----__,_----r-----!
o 20 40 60 80 100 120 o 20 40 60 80 100 120
Tim e after planting (days) Time after planting (days)
Figure 5.4. Seasonal cumulative intercepted PAR (MJ m-2 day -1) in beans, chickpea and
cowpea under mid-season (MS) and late season (LS) water stress and well-watered (C)
conditions in 2002 (left) and 2002/2003(right) seasons.
Seasonal changes in PAR intercepted by the species are presented in Fig. 5.4. The cumulative
intercepted PAR was higher in the second than in the third season. The increase in cumulative
PAR in the C treatment after 60 DAP in 2002/2003 was not as fast as in 2002 because of a
long period (>20 days) of cloudy weather that reduced the incoming radiation and thereby
hindered photosynthesis and canopy growth.
There was no significant difference in the PAR intercepted between the C and LS treatments
in all the species and both seasons. Although an increase in cumulative PAR interception
was observed after re-watering of the MS treatment in the third season, the cumulative
102
intercepted PAR was slightly lower than that of the C treatment in all the species and in both
seasons. The cumulative intercepted PAR was higher in 2002 than in 2002/2003 season and
significantly correlated with F in the third season (r = 0.83* to 0.96**). However, there was
poor correlation between incident PAR and the seasonal cumulative PAR intercepted in both
seasons. This agrees with the result by Thomson and Siddique (1997) in which differences
among crops in PAR interception is not a function of the length of time that crops intercept
radiation.
5.3.4. Radiation use efficiency
RUE of the C treatment was higher in the second season than in the third season (Fig. 5.5).
In the LS treatment, the RUE values were higher in the second season than in the third
season whereas the values for the MS treatment were variable among species and seasons.
So, the best fit of the linear regression was obtained by pooling the data of the two seasons
(Fig. 5.5). The RUE values for the C treatment were higher than those for the MS and LS
treatments. Under non-limiting water conditions, the RUE was 2.44 (1.22), 2.07 (1.04) and
2.16 (1.08) g Mr! PAR (SR, values in brackets) for beans, chickpea and cowpea,
respectively (Fig. 5.5). Beans, chickpea and cowpea had the RUE values of2.00 (1.00), 1.68
(0.84), and 1.80 (0.90) g Mr! PAR (SR) respectively in the LS treatment. The values in the
LS treatments were not significantly different (P>O.OS)from the values recorded in both the
C and LS treatments in beans and chickpea (Table 5.1). As compared to well-watered
conditions, water stress during the flowering period resulted in significantly lower (P<O.OS)
RUE values in beans and chickpea but not in cowpea (Table 5.1). The figures in the MS
treatment were 1.50 (0.75), 1.45 (0.73) and 1.59 (0.80) gW! PAR (RS) for beans, chickpea
and cowpea, respectively (Fig. 5.5). RUE was not significantly affected by any of the
treatments in cowpea (Table 5.1). Although the values were not significantly different from
each other (P>0.05), beans had the highest RUE, followed by cowpea and chickpea in that
order under well-watered conditions.
The maximum RUE values found in this study are slightly higher than the range of values
(in g Mr! SR) reported for several grain legumes in different environments, including 0.30-
0.93 for chickpea (Hughes et al., 1987; Singh and Sri Rama, 1989; Leach and Beech, 1988),
0.15-0.78 for beans (Tsubo et al., 2003), 0.72 for pea (Martin et al., 1994), 0.92 for mung
bean (Muchow and Charles-Edwards, 1982),0.96 for lentil (McKenzie and Hill, 1991),0.58
for lupin (Gregory and Eastham, 1996) and 0.41-0.99 for various grain legumes, including
103
1600 ~ ~
.. .• Bean1400 o C (RUE = 2.44, R2 = 0.80) o
- 1200 o MS (RUE = 1.50, R2 = 0.78)~ ..
-§, 1000 t:::. LS (RUE = 2.00, R2 = 0.84) o .. . .0. o--...Cl) 800ca
E 600
.>..-
C 400
200
o ~~~~ -. .- -. ~ ,- ~
o 100 200 300 400 500 600 700
1600 ~ ~
Chickpea
1400 oe (RUE = 2.07, R2 = 0.91)
- 1200 o MS (RU E = 1.45, R2 = 0.89) ..~
-; 1000 t:::. LS (RUE = 1.68, R2 = 0.86)--...
o
Cl) 800 o
ca
E 600
>-
c... 400
200
o
o 100 200 300 400 500 600 700
1600 ~~ ~
Cowpea
1400 (RUE = 2.16, R2 = 0.81)
cf' o1200 (RUE = 1.59, R2 = 0.85)
-§, .'1000 (RUE = 1.80, R2 = 0.82) o... o o o
:Cl:) 800 t:::.
ca
E 600 ••D··· _--o .A...
>
c...
- 400 G-ê• ._.._. -- -oo--~ 0
200 ... ;.....0.. Ot:::.,..,..._."............ ~
o .. ~~~,- -. ~ .- -. -. ~
o 100 200 300 400 500 600 700
Cumulative intercepted PAR(MJ m")
Figure 5.5. Radiation use efficiency (RUE, g M.rl) of beans, chickpea and cowpea under
mid-season (MS - - -) and late season (LS -) water stress and well-watered (C --)
conditions for data combined over two seasons.
104
chickpea, lentil, faba bean, pea, and some others (Thomson and Siddique, 1997). However,
the present maximum values (g Mr! SR) are within the range of other reports for groundnut
(1.02-1.37, Bell et al., 1987; Marshall and Willey, 1983), cowpea (1.09, Muchow et al.,
1993), pigeon pea (1.23, Hughes and Keatinge, 1983), pea (0.96-1.46, Heath and
Hebblethwaite, 1985) and faba bean (2.04, Fasheun and Dennet, 1982). Variations in the
reported RUE values among experiments cou1d be due to differences in crop variety and
other environmental factors.
As summarized in Sinclair and Muchow (1999), the maximum RUE values (g Mr! SR)
reported for C4 species were 2.00, 1.77 and 1.40 for sugarcane, maize and sorghum
respectively and for non-leguminous C3 species were 1.75, 1.56, 1.46, 1.39, and 1.30 for
potato, sunflower, wheat, rice and barley respectively. Therefore, the values for cereals and
non-legume C3 species are higher than the maximum values reported for grain leguminous
species (Sinclair and Muchow, 1999) including those found in present study.
Relative to well-watered conditions, reduction in RUE due to water stress in the flowering
period was 39, 30 and 26% in beans, chickpea and cowpea respectively. Reduction in RUE
under water stress conditions was also reported for many grain legumes such as soybean,
cowpea and other grain legumes (Muchow, 1985a), chickpea (Singh and Sri Rama, 1989),
beans (Ogindo, 2003), faba bean (Green et al., 1985), pea (Keatinge et al., 1985; Heath and
Hebblethwaite, 1985) and pigeon pea (Hughes and Keatinge, 1983). This cou1dbe due to the
depressing effects of water deficit on leaf photosynthesis, such as high leaf temperature, leaf
senescence, stomatal closure, restricted leaf expansion, and poor leaf area development. The
latter two mostly apply to the MS treatment. Therefore, the magnitude of reduction in RUE
in grain legumes can be dependent on the growth stages at which the stress is imposed, and
its severity.
5.4. Conclusion
The results from the present study indicate that dry matter production is highly associated
with the fraction of PAR intercepted, which in turn is highly and positively correlated with
green LAl. Therefore, species that intercept a large fraction of the PAR are important in the
dry environments. Under non-limiting water supply, the efficiency of radiation conversion
into dry matter is comparable in the three species, indicating the conservative nature of RUE
in grain legumes under well-watered conditions. The RUE values found in the present study
105
are within the range of previous reports, confirming that low RUE in grain legumes could be
the inherent characteristics of these species. RUE is more sensitive to water stress during
early than late stage reproductive period in some species (e.g. beans and chickpea) while it is
not significantly affected by any of the stress treatments in others (e.g. cowpea). This species
variability could be exploited in a crop breeding programme to develop cultivars that have
stable RUE under variable soil water conditions in dry environments. Although a high K
value under well-watered conditions is important for high F and RUE, species with high K
values during early stage reproductive water stress have low RUE, suggesting the
importance of canopy modification in response to water deficits (which helps decrease leaf
temperature) in grain legumes to maintain photosynthesis and RUE in dry environments.
The information obtained from this study will be valuable in developing radiation-based
crop growth models suitable for the dry areas.
106
CHAPTER6
Comparative Water Relations, Leaf Gas Exchange and Assimilation of Three Grain
Legumes Under Water Deficit
6.1. Introduction
Water shortage is a major constraint to crop production as crops are usually exposed to
drought periods of varying duration and intensity during their growth (Sadras and Milroy,
1996), particularly in the semi-arid regions of the world (Squire, 1990). Semi-arid
climates are characterized by fluctuating rainfall both in amount and distribution, and
plants grown under these climates are prone to frequent atmospheric drought, even when
the soil water reserves are adequate (Maroco et al., 1997). Grain legumes are grown
under rainfed conditions in the semi-arid tropical regions and their yield depends on the
amount of water transpired and the seasonal pattern of soil water availability (Adams et
al., 1985; Rachie, 1985; Turk et al., 1980b, Cooper et al., 1988). Because of the erratic
nature of the rainfall during the season, these crops can experience intermittent or
continuous water deficits during their growth.
When plants are exposed to water deficits, they often exhibit physiological responses that
can result in adaptation to the environment. The plants that are usually grown under dry
environments have evolved their own adaptation strategies which can be categorized as
drought escape, dehydration postponement and dehydration tolerance (Levitt, 1980;
Turner,.1986; Laffary and Louguet, 199Q;Turner, 1991; Turner et al., 2001). "Escaping"
drought involves completion of the life cycle after a significant rainfall and before the
onset of the drought period. Dehydration postponement involves maintenance of plant
water status in the presence of environmental drought (drought avoidance) while
dehydration tolerance involves maintenance of plant function in the presence of drought
(drought tolerance). Both these include whole plant mechanisms that provide the plant
with the ability to respond and survive drought (Turner, 1986; Blum, 1988; Laffary and
Louguet, 1990; Turner, 1991). Therefore, different plant responses induced by drought
should reflect the different adaptations, or the lack of them.
Plant productivity generally depends on the rate of C02 assimilation (Srivasta and
Strasser, 1996; Costa Franca et al., 2000), and transpiration, which serves as a major
cooling mechanism for plant leaves through the evaporation process (Jalali-Farahani et
al., 1993). Stomatal pores act as the exchange pathway for both CO2 and H20 between
107
the atmosphere and the plant cells and hence control the rate of photosynthesis and water
use (Cowan and Troughton, 1971; Farquhar and Sharkey, 1982; Collatz et al., 1991).
Stomatal adjustment is one of the prominent examples of plant responses to drought, and
stomata can be considered to be integrators of all environmental factors that affect plant
growth (Mori son, 1998). Stomata regulate water use and the development of water stress,
and influence plant growth rates through effects on availability of CO2 assimilation (e.g.
Baldocchi et al., 1985). Thus, stomatal responses to environmental drought have a
substantial influence on plant adaptation in dry climates (Bates and Hall, 1982b).
Leaf water status and humidity of the air are reported to have a major influence on
stomatal conductance in the field (Turner, 1991). Stomatal conductance usually decreases
when plants are subjected to soil water deficits (Bates and Hall, 1982b; Lopez et al.,
1988), and differences in stomatal conductance in response to leaf water potential have
been reported in many grain legumes (Lawn, 1982; Muchow, 1985b; Flower and Ludlow,
1986). Maintenance of stomatal conductance and photosynthesis during leaf water deficit
has been associated with favorable seed yield in soybean genotypes (Solane et al., 1990).
In general, water flow in the soil-plant-atmosphere system is governed by differences in
the water potential of the three systems and the resistances in the water flow pathway.
The soil water potential usually determines the upper limit of leaf water potential while
the lower limit is set by the combined action of atmospheric variables, soil water potential
and the resistances to flow (Choudhury, 1985). Therefore, proper understanding and
modelling of plant processes and reactions to water deficit requires the determination of
the quantitative relationships between soil-plant water relations, growth, gas exchange
and assimilation rate (Ritchie, 1981; Sadras and Milroy, 1996). Thus, actual plant
responses to soil water deficit in the field can be obtained by a simultaneous study of soil
and plant water status, stomatal resistance and its effect on gas exchange and assimilation.
Several studies have attempted to understand the different response of grain legumes to
water deficits under controlled and some field conditions (Sinclair and Ludlow, 1986;
Muchow, 1985b; Lawn 1982; Angus et al., 1983; Turk et al., 1980a,b; Parson and Howe,
1984; Markhart, 1985; Kuppers et al., 1988; Vasquez-Tello et al., 1990; Cruz de
Cravalho et al., 1998; Leport et al., 1998; 1999). Next to pigeonpea, many studies
indicate cowpea as a drought tolerant crop among grain legumes (Sinclair and Ludlow,
1986; Vasquez-Tello et aI., 1990; Cruz de Carvalho et al., 1998) while beans is
108
considered to be susceptible (Vasquez-Tello et al., 1990; Cruz de Carvalho et al., 1998).
Chickpea is also considered as a drought tolerant crop among the cool-season food
legumes (Singh, 1993; Leport et al., 1999). However, the underlying physiological
responses of these three species have not been investigated in the field under the same
seasonal, environmental and experimental conditions. Knowledge of a particular response
by each species under the same conditions is important for evaluation and development of
crop simulation models, and to develop guidelines for crop choice for a specific
environment in areas like Ethiopia where the species are grown in diverse environments
with poor yield.
Therefore, the objective of this investigation was to determine and compare the
relationship between soil water, leaf water potential, stomatal resistance, rate of
photosynthesis and transpiration in common bean (Phaseolus vulgaris L.), chickpea
(Cieer arietinum L.) and cowpea (Vigna anguieulata L. Walp) under water stressed and
non-stressed conditions during the reproductive stages in the field under a semi-arid
environment.
6.2. Materials and Methods
6.2.1. Field experiments
Descriptions of experimental site, agronomic information, irrigation schedule,
experimental layout and treatments and weather conditions during the experimental
periods are given in Chapter 3. Except leaf water potential, all physiological data were
collected in the 2002 and 2002/2003 seasons only.
6.2.2. Measurements
6.2.2.1. Soil water
The soil water content to a depth of300 mm in 2001102 and 600 mm in 2002 and 2002/03
was monitored every day using Time Domain Reflectometery, TDR (Soil Moisture
Equipment Corp., CA, USA) starting from planting. The available soil water during the
measurement period was above 60% in the control plots while it reached up to the lowest
23% in the stressed plots.
109
6.2.2.2. Leaf water potential
Midday (12:00-14:00 local time) leaf water potential of the stressed and non-stressed plots
for the three seasons was measured on upper fully exposed leaves of five plants per plot
using a pressure chamber (PMS Instrument Company, Oregon, U.S.A) every other day
throughout each stress period. The leaves were covered with a white polythene bag before
excision and then water potential measurement of each leaf was completed within the next 2
minutes to avoid evaporative water loss that affects the readings.
6.2.2.3. Stomatal resistance
The stomatal resistance of the stressed and non-stressed plots was measured in parallel with
the water potential measurement using diffusion porometer (Model AP4, Delta-T Devices
Ltd. U.K) in 2002 and 2002/03 on similar leaves to those used for the water potential
measurement. The measurement was made from three leaves per plant and five plants per
plot giving a total of 15 measurements per plot. The abaxial surface of the upper fully
expanded leaves was considered for this purpose. The measurement was made between
12:00-14:00 local time (GMT +3) every other day during each stress period. The porometer
was calibrated regularly against a calibration plate depending on changes in relative
humidity and temperature of the environment during the measurements dates.
6.2.2.4. Rate of photosynthesis and transpiration
Leaf gas exchanges and leaf temperatures were measured using an Infrared Gas Analyser,
IRGA, Type LCA-4 (ADC Bio Scientific Ltd., U.K.) between 12:00-13:00 local time
every other day during each stress treatment. The IRGA was recalibrated every week
against a standard gas (compressed gas taken from unpolluted area with CO2
concentration of 360 ppm) to obtain accurate readings of CO2 within the acceptable
range. The IRGA calculates the rate of photosynthesis (umol m-2s-1) from the measured
parameters using the following equation:
A = Us * ~c (6.1)
where ~c is difference in CO2 concentration through chamber (umol mol") and Us is mass
flow of air per m2 ofleaf area (mol m-2 sol).
Transpiration rate (mmol m-2 sol)was also calculated as
E=us * ~w (6.2)
110
Eis leaf transpiration rate; !:lwis the difference in water vapour concentration (mmol
mol") within and out of the chamber.
6.2.2.5. Diurnal measurements
Diurnal measurements of leaf water potential, stomatal resistance, and gas exchange were
made in the third season on 10 December 2002 for chickpea and 16 December 2002 for
beans and cowpea on stressed and non-stressed plots of each species. Leaf water potential
was measured from 6:00 to 18:00 local time while the other measurements were made
from 6:00 to 16:00 because of the low level of daylight at 18:00 during the measurement
periods. The measurements were taken from the upper two fully expanded leaves of five
randomly selected plants per plot.
6.2.3. Data analysis
Mean and standard error calculations and t-tests were made using Number Cruncher
Statistical System, NCSS 97 (Hintze, 1997). Correlation and regression analyses were
performed using MINITAB for Windows, release 12.21 (Minitab Inc., 1998). Some linear
regressions were also fitted using Microsoft Excel (Microsoft Corporation). The increase
or decrease of parameters with time was determined by linear regression. Threshold
(cutoff) values for a given variable were calculated as intersection point of two linear
regression lines obtained from data points clustered based on similarity of trend.
6.3. Results and Discussion
6.3.1. Leaf water potential (\jiL)
The midday \jiL under well-watered conditions remained above -1.50 MPa in beans, -1.58
MPa in chickpea and -1.25 MPa in cowpea during all the seasons (Tables 6.1, 6.2 & 6.3).
The values observed under well-watered conditions in the present study are lower than
the value (-0.6 MPa) reported for six unstressed cool-season grain legumes including
chickpea in a Mediterranean-type environment (Leport et al., 1998). The midday \jiL
recorded at the end of the stress treatments never fell below -1.70 and -1.60 MPa in beans
and cowpea, respectively, and the range of variations in \jiL between the control and
stressed plants were small in these two species. This agrees with previous observations
made by Bates and Hall (1982a), Turk et al., (1980b), Nwalozie and Annerose (1996) and
Diallo et al. (2001) for cowpea. The lowest midday \jiL, which ranged from -3.98 to -3.02
111
MPa between the three seasons, was recorded for chickpea (Tables 6.1, 6.2 & 6.3).
Midday \jiL values ofless than -3.0 MPa were also reported for water stressed chickpea in
a Mediterranean-type environment (Leport et al., 1998; 1999). Relative to the controls,
the average decline in \jiL was 3.4, 7.7, and 4.8% per day in the MS treatments and 1.6,
10.1, and 2.8% per day in the LS treatments in beans, chickpea and cowpea, respectively.
Therefore, the relative rate of leaf water potential decline due to water deficit is faster in
chickpea than in beans and cowpea. The fastest decline in chickpea could be attributed to
its slower and lower stomatal adjustment relative to the decline in \jiL as explained below.
Table 6.1. Leaf water potential (MPa) of three grain legume species under well-watered (C)
and water stressed conditions during flowering (MS) and pod filling (LS) periods in the
200112002season.
Species and water regime treatments
DAW* BN CHP COP
CB DAW
BN CHP COP
MS C MS C MS C LS C LS C LS
8 -1.28 -1.57 -1.17 -1.92 -1.09 -1.10 3 -1.19 -1.10
(0.05) (0.02) (0.07) (0.04) (0.01) (0.02) (0.04) (0.02)
10 -1.34 -1.67 -1.51 -2.45 -1.10 -1.15 6 -1.23 -1.54 -1.49 -2.11 -1.17 -1.30
(0.03) (0.02) (0.07) (0.07) (0.02) (0.04) (0.03) (0.01) (0.01 ) (0.05) (0.02) (0.04)
14 -1.38 -1.68 -1.55 -2.93 -1.19 -1.27 8 -1.36 -1.57 -1.24 -3.62 -1.19 -1.34
(0.02) (0.02) (0.03) (0.02) (0.04) (0.02) (0.04) (0.05) (0.08) (0.10) (0.04) (0.01)
16 -1.30 -1.69 -1.58 -3.72 -1.19 -1.27 10 -1.39 -1.68 -1.42 -3.70 -1.15 -1.37
(0.02) (0.06) (0.05) (0.18) (0.47) (0.04) (0.03) (0.05) (0.05) (0.16) (0.03) (0.02)
19+ -1.36 -1.25 -1.24 -1.55 -1.21 -1.28 12 -1.28 -1.64 -1.29 -3.89 -1.24 -1.37
(0.04) (0.00) (0.08) (0.10) (0.19) (0.12) (0.06) (0.04) (1.02) (0.11 ) (0.02) (0.02)
22+ -1.39 -1.28 -1.42 -1.55 15 -1.29 -1.60 -1.29 -3.98 -1.24 -1.38
(0.03) (0.07) (0.05) (0.00) (0.05) (0.04) (0.09) (0.121 (0.011 (0.02)
.. DAW- days after withholding water, +- measurement after re-watering·
B Numbers in parenthesis refer to standard error of means.
Table 6.2. Leaf water potential (MPa) of three grain legume species under well-watered (C)
and water stressed conditions during flowering (MS) and pod filling (LS) periods in the
2002 season.
Species and water regime treatments
DAW BN CHP COP BN CHP COP
CB MS C MS C MS C LS C LS C LS
4 -1.34 -1.44 -1.52 -2.34 -1.25 -1.48 -1.28 -1.47 -1.53 -1.73 -1.15 -1.32
(0.04) (0.04) (0.10) (0.04) (0.07) (0.03) (0.03) (0.04) (0.02) (0.02) (0.06) (0.03)
6 -1.26 -1.45 -1.53 -2.44 -1.19 -1.38 -1.18 -1.40 -1.27 -2.35 -1.10 -1.43
(0.02) (0.04) (0.04) (0.08) (0.05) (0.04) (0.04) (0.03) (0.04) (0.09) (0.03) (0.04)
8 -1.20 -1.48 -1.55 -2.74 -1.20 -1.57 -1.23 -1.57 -1.35 -2.50 -1.07 -1.53
(0.05) (0.02) (0.02) (0.07) (0.02) (0.04) (0.04) (0.04) (0.03) (0.15) (0.06) (0.02)
10 -1.15 -1.57 -1.52 -3.08 -1.05 -1.57 -1.33 -1.63 -1.55 -2.70 -1.02 -1.55
(0.05) (0.04) (0.03) (0.12) (0.03) (0.04) (0.09) (0.02) (0.03) (0.03) (0.04) (0.03)
12 -1.07 -1.58 -1.53 -3.37 -1.02 -1.57 -1.33 -1.70 -1.61 -2.90 -1.13 -1.60
!0.021 (0.03) (0.06) (0.10) (0.03) (0.021 (0.021 (0.031 (0.071 (0.031 (0.021 (0.03)
B Numbers in parenthesis refer to standard error of means.
112
Table 6.3. Leaf water potential (MPa) of three grain legume species under well-watered (C)
and water stressed conditions during flowering (MS) and pod filling (LS) periods in the
2002/2003 season.
Species and water regime treatments
DAW BN CHP COP BN CHP COP
CB MS C MS C MS C LS C LS C LS
2 -1.07 -1.12 -1.23 -1.23 -0.96 -1.00
(0.03) (0.03) (0.02) (0.02) (0.02) (0.03)
4 -1.25 -1.23 -1.30 -1.37 -0.91 -0.95 -1.20 -1.18 -1.00 -1.20 -1.18 -1.18
(0.02) (0.01 ) (0.02) (0.01) (0.02) (0.02) (0.03) (0.04) (0.03) (0.02) (0.04) (0:02)
6 -1.29 -1.30 -1.30 -1.76 -0.61 -1.10 -0.88 -1.12 -0.63 -1.28 -0.84 -0.92
(0.02) (0.02) (0.02) (0.01) (0.02) (0.03) (0.07) (0.02) (0.07) (0.07) (0.04) (0.02)
8 -1.16 -1.38 -1.37 -2.11 -0.97 -1.20 -1.02 -1.13 -0.52 -1.48 -0.70 -0.85
(0.02) (0.01) (0.01) (0.02) (0.02) (0.03) (0.02) (0.03) (0.04) (0.04) (0.05) (0.03)
10 -1.10 -1.43 -1.53 -2.37 -0.96 -1.35 -0.78 -1.05 -0.53 -1.28 -0.57 -1.02
(0.01) (0.01) (0.03) (0.05) (0.02) (0.02) (0.02) (0.03) (0.02) (0.02) (0.04) (0.02)
12 -1.05 -1.49 -1.10 -2.45 -1.18 -1.50 -0.76 -1.07 -0.82 -1.17 -0.73 -0.87
(0.02) (0.02) (0.02) (0.02) (0.04) (0.02) (0.02) (0.03) (0.04) (0.02) (0.04) (0.02)
14 -1.00 -1.59 -1.23 -2.57 -0.84 -1.46 -0.99 -1.21 -0.81 -1.55 -0.83 -1.05
(0.02) (0.02) (0.03) (0.03) 90.04) (0.07) (0.07) (0.02) (0.06) (0.07) (0.03) (0.02)
16 -1.10 -1.62 -1.29 -3.02 -0.57 -1.53
(0.03) (0.03) (0.04) (0.02) (0.04) (0.02)
18 -1.07 -1.66
(0.03) !0.02)
B Numbers in parenthesis refer to standard error of means.
6.3.2. Stomatal resistance (rs)
The rs in the control plots ranged from 0.4 to 6.0 s cm" in 2002 and 1.2 to 3.6 s cm" in
2002/2003 during the measurement period (Tables 6.4 & 6.5). The maximum r, at the end
of theMê treatment was 12.5, 11.0, and 9.2 in 2002 and 16.5, 15.9, and 37.9 s cm" in
2002/2003 in beans, chickpea and cowpea, respectively (Tables 6.4 & 6.5). The
maximum r, recorded in the LS treatment was 11.8, 10.4 and 9.3 in 2002 and 2.3,2.6 and
3.0 s cm" in the three species in the same order. Although both temperature and VPD
were higher in 2002 than 2002/2003, the latter season had higher rs values than the former
in the MS treatment. This could be due to the effect of high temperature (in 2002) on the
physiological mechanisms that control stomatal adjustment during water stress. On the
other hand, the lower rs values in the LS treatments in 2002/03 were due to low intensity
of the stress because of cloudy weather (lower VPD, which ranged from 0.68 to 2.2 kPa
compared to 2.4 to 4.2 kPa in 2002 for the same treatment period and lower temperature).
Although there was a significant difference in both midday \jiL and r, among the species in
2002/2003, there was no significant difference in the maximum r, recorded at the end of
113
Table 6.4. Stomatal resistance (s cm") of three grain legume species under well-watered (C)
and water stressed (S) conditions during flowering (MS) and pod filling (LS) periods in the
2002 season.
Speciesand waterregimetreatments
DAW BN CHP COP BN CHP COP
CB MS C MS C MS C LS C LS C LS
4 2.78 5.65 1.56 5.22 1.73 5.45 3.44 4.63 2.10 3.43 1.43 3.99
(0.05) (0.25) (0.09) (0.36) (0.11) (0.40) (0.15) (0.30) (0.09) (0.18) (0.08) (0.24)
6 2.52 7.05 1.52 7.18 1.64 5.53 4.59 6.51 1.98 4.10 1.24 4.18
(0.13) (0.37) (0.22) (0.86) (0.013) (0.37) (0.35) (0.40) (0.13) (0.36) (0.09) (0.44)
8 1.86 7.55 1.80 8.25 1.88 8.35 6.27 8.49 2.41 5.02 2.22 6.27
(0.20) (0.39) (0.12) (0.61) (0.09) (1.00) (0.37) (0.34) (0.13) (0.29) (0.16) (0.47)
10 1.94 9.58 1.68 10.98 1.57 8.40 3.24 9.24 2.08 5.51 1.77 7.11
(0.09) (0.64) (0.09) (1.35) (0.08) (1.06) (0.27) (0.88) (0.24) (0.89) (0.15) (0.52)
12 2.90 12.54 2.00 11.02 1.73 9.22 3.57 12.8 1.82 6.35 2.84 9.29
(0.14) (0.69) (0.15) (1.41) (0.04) (1.13) (0.28) (0.77) (0.10) (0.74) (0.20) (0.88)
ijNumberisnparenthesrisefertostandaredrrorofmeans.
Table 6.5. Stomatal resistance (s cm") of three grain legume species under well-watered (C)
and water stressed (S) conditions during flowering (MS) and pod filling (LS) periods in the
2002/2003 season.
Speciesand waterregimetreatments
DAW BN CHP COP BN CHP COP
CB MS C MS C MS C LS C LS C LS
2 1.74 2.25 1.56 1.41 1.01 1.45
(0.21) (0.17) (0.10) (008) (0.11 ) (0.09)
4 2.43 1.96 0.87 1.17 1.65 3.77 1.52 1.77 0.54 1.33 1.47 1.09
(0.18) (0.13) (0.23) (0.10) (0.18) (0.34) (0.38) (0.31 ) (0.13) (0.22) (0.29) (0.36)
6 3.80 3.48 1.60 4.96 1.61 8.61 0.61 0.89 0.32 0.40 1.10 2.38
(0.40) (0.48) (0.13) (0.60) (0.18) (0.92) (0.04) (0.12) (0.02) (0.02) (0.24) (0.54)
8 2.09 6.71 3.27 5.11 3.49 11.92 0.45 1.40 0.58 2.60 0.98 3.03
(0.22) (0.53) (0.19) (0.56) (0.52) (2.01) (0.06) (0.21 ) (0.04) (0.45) (0.16) (0.34)
10 0.56 5.26 6.05 9.68 1.01 32.32 0.45 0.64 0.68 1.15 0.71 0.95
(0.03) (0.50) (0.06) (1.69) (0.11 ) (8.13) (0.02) (0.06) (0.05) (0.06) (0.05) (0.05)
12 2.29 9.36 1.95 10.95 1.90 37.94 0.51 1.65 0.72 1.60 0.93 1.89
(0.25) (0.23) (0.06) (1.63) (0.29) (12.77) (0.07) (0.10) (0.09) (0.20) (0.12) (0.31)
14 1.21 9.63 1.17 13.8 1.10 20.98 1.73 5.05 0.59 2.92 1.48 2.96
(0.04) (1.85) (0.07) (2.92) (0.24) (4.00) (0.38) (0.79) (0.09) (1.06) (0.13) (0.26)
16 1.45 13.75 2.62 25.94 0.98 5.07
(0.12) (1.86) (0.31 ) (6.31) (0.16) (0.35)
18 2.40 16.54
(0.21l (2.66)
bNumberisn parenthesriesfertostandard errorofmeans.
MS treatment in 2002 despite significant differences In \jiL among the species.
Differences in rs among species were not significant in the LS treatment in both seasons.
Nevertheless, significantly higher rs values in the stressed plants than in controls indicate
the role of stomatal adjustment to the drought adaptation of grain legumes as reported in
114
other studies (Bates and Hall, 1982b; Baldocchi et al., 1985; Trejo and Davies, 1991;
Barradas et al., 1994; Cruz de Carvalho et al., 1998; Costa Franca et al., 2000).
The trigger for stomatal closure (adjustment) under periods of water stress is reported to
be associated with root-to-shoot communication via Abscisic acid (ABA) translocation in
many crops (Davies et al., 1990; Davies and Zhang, 1991; Ribaut and Pilet, 1991; Blum
and Johnson, 1993) and decreases in hydraulic conductance of the soil-leaf continuum
(Sperry, 2000). Relative to the respective control measurements, the reductions in A at the
end of the LS treatment ranged from 61-81% in beans, 36-81 % in chickpea and 48-66%
in cowpea between the two seasons, and the average reduction rate during the whole
stress period ranged from 3.5-4.3, 3.9-4.0 and 3.2-3.4 % per day, respectively. This shows
that reduction in the rate of net photosynthesis is slightly higher and faster in the mid-
season than in the late-season stressed grain legumes.
Fast closure of stomata (even before detection of leaf water deficit) has been measured in
beans in response to soil water deficit (Trejo and Davies, 1991; Barradas et al., 1994).
Under severe water stress, complete stomata closure at lower \jiL was reported in beans
when compared to cowpea (Cruz de Carvalho et al., 1998). However, cowpea has a better
stomatal adjustment than beans by maintaining partial stomatal opening as stress
increases and simultaneously avoiding drought by early regulation of stomatal closure
(Cruz de Carvalho et al., 1998).
6.3.3. Rate of photosynthesis (A) and transpiration (E)
The maximum photosynthesis rate recorded under well-watered conditions was 20.5, 23.6
and 23.9 in 2002 and 19.7, 20.2 and 21.1 umol m-2 S-1 in 2002/2003 in beans, chickpea
and cowpea, respectively (Tables 6.6 & 6.7). The average values for the same treatment
were 15.9,20.6 and 17.0 in 2002 and 16.7, 16.3 and 16.3 in 2002/2003 in beans, chickpea
and cowpea, respectively. There was no significant difference in A between the species
under favourable water supply conditions. The average values found here are lower than
the values reported for chickpea in a Mediterranean-type environment (Leport et al.,
1998) but higher than the values reported for beans and cowpea under non-stress
conditions in a controlled experiment (Cruz de Carvalho et al., 1998).
115
Table 6.6. Rate of photosynthesis (umol m-2 S-l) of three grain legume species under well-
watered (C) and water stressed conditions during flowering (MS) and pod filling (LS)
periods in the 2002 season.
Species and water regime treatments
DAW BN CHP COP BN CHP COP
CG MS C MS C MS C LS C LS C LS
4 13.57 5.48 18.23 8.12 10.90 5.80 20.50 10.44 23.55 16.49 23.94 13.11
(2.25) (0.79) (3.01) (0.32) (2.78) (1.10) (0.61) (1.02) (0.59) (1.68) (0.87) (0.59)
6 12.91 3.46 16.60 3.12 11.76 5.60 16.38 7.01 17.94 10.54 19.06 10.68
(1.37) (0.60) (0.70) (0.46) (1.35) (1.15) (0.91) (0.52) (1.29) (1.04) (1.03) (1.29)
8 14.37 4.22 20.79 2.66 20.69 6.20 16.51 4.62 22.60 8.78 16.58 7.36
(1.73) (0.35) (2.56) (0.91) (0.70) (1.15) (0.73) (0.46) (1.47) (1.12) (0.93) (1.36)
10 16.68 2.99 20.73 2.44 15.50 4.70 15.61 2.97 23.27 8.27 18.26 7.29
(2.28) (0.53) (2.50) (0.75) (1.86) (1.66) (0.59) (0.25) (1.02) (1.45) (1.20) (0.77)
12 17.81 2.50 22.88 2.27 17.37 5.15 14.19 2.74 18.93 3.53 16.85 5.70
(0.79) (0.59) (2.70) (0.81) (2.85) (0.85) (0.94) (0.86) (1.67) (1.06) (0.57) (1.42)
Table 6.7. Rate of photosynthesis (umol m-2 s") of three grain legume species under well-
watered (C) and water stressed conditions during flowering (MS) and pod filling (LS)
periods in the 2002/2003 season.
Species and water regime treatments
DAW BN CHP COP BN CHP COP
CG MS C MS C MS C LS C LS C LS
2 16.46 15.56 20.20 18.80 14.80 14.70
(2.42) (0.57) (0.44) (0.73) (1.20) (1.19)
4 15.50 17.00 13.29 16.90 14.27 14.70 16.60 13.30 18.80 17.70 16.50 18.80
(1.78) (0.71) (2.33) (1.58) (1.02) (1.47) (0.87) (0.57) (1.57) (1.94) (1.71 ) (0.88)
6 12.80 11.40 15.49 8.40 16.40 10.62 17.00 10.10 10.20 11.40 11.95 10.40
(1.19) (0.96) (1.74) (1.23) (0.57) (0.37) (0.65) (0.28) (1.20) (1.07) (0.47) (1.20)
8 16.50 7.40 18.70 5.82 14.91 7.83 16.31 9.14 4.05 4.16 6.17 4.22
(0.69) (0.46) (0.89) (0.42) (1.44) (0.29) (0.15) (0.58) (0.44) (0.36) (0.30) (0.51)
10 18.10 7.80 16.22 3.70 14.82 4.64 16.60 9.50 17.00 7.54 18.80 14.20
(0.54) (0.37) (0.57) (0.46) (0.57) (0.28) (0.57) (0.55) (2.51 ) (0.50) (0.99) (2.04)
12 16.41 3.90 14.59 3.10 13.50 2.20 19.70 13.35 18.40 11.70 21.10 15.87
(0.81) (0.50) (1.31) (0.28) (1.71 ) (0.35) (0.60) (1.14) (0.95) (2.08) (1.44) (0.96)
14 18.03 5.49 16.19 2.33 16.95 2.22 14.97 5.78 19.68 12.52 18.61 9.66
(1.01) (0.61) (1.67) (0.36) (0.47 (0.43) (1.06) (0.54) (1.06) (1.67) (1.19) (0.86)
16 18.90 2.25 12.47 2.66 17.17 2.80
(1.44) (0.47) (0.81) (0.61) (0.30) (0.41)
18 16.46 1.75
F·42~ (0.30)
b Numbers inparenthesisrefer to standarderror of means.
The A values recorded at the end of the MS treatment were 2.50,2.27, and 5.15 umol m-2
S-1 in 2002 and 1.75, 2.66 and 2.80 umol m-2 S-1 in 2002/2003 in beans, chickpea and
cowpea, respectively indicating a reduction of photosynthesis by 86-89% in beans, 79-
90% in chickpea and 70-84% in cowpea during the two seasons when compared to the
116
respective control measurements. As shown by similar A values, species difference
during the severe stage of water stress was not significant. The average relative rate of
decline in A during the MS treatment was 4.6, 5.5, and 4.0% per day in beans, chickpea
and cowpea, respectively. The minimum values of A recorded at the end of the LS
treatment were 2.74, 3.53, and 5.70 in 2002 and 5.78, 12.52 and 9.66 umol m-2 S-1 in
2002/2003 for beans, chickpea and cowpea, respectively (Tables 6.6 & 6.7).
Relative to the respective control measurements, the reductions in A at the end of the LS
treatment ranged from 61-81% in beans, 36-81 % in chickpea and 48-66% in cowpea
between the two seasons, and the average reduction rate during the whole stress period
ranged from 3.5-4.3, 3.9-4.0 and 3.2-3.4 % per day, respectively. This shows that
reduction in the rate of net photosynthesis is slightly higher and faster in the mid-season
than in the late-season stressed grain legumes. Among the species, reduction in A was
much faster in chickpea than beans, and in beans than in cowpea in all the seasons and
treatments. Relative to the control, however, reductions in A were higher in beans and
chickpea than in cowpea under both seasons and stress treatments. As indicated by Cruz
de Carvalho et al. (1998), the lower decrease rate of A in cowpea, compared to beans and
chickpea, could be either due to its ability to maintain partial stomatal opening under
stress or due to less sensitivity of its photosynthetic activity to the stress. Reductions in A
due to water stress have been reported for a number of grain legumes including chickpea,
beans and cowpea (Leport et al., 1998; Cruz de Cravalho et al., 1998). The major reason
for the reduction of A in beans and cowpea under severe water stress has been reported to
be stomatal closure (Cruz de Carvalho et aI., 1998; Costa Franca et al., 2000) and reduced
biochemical capacity such as reduced rubisco activity and increases in internal CO2
concentration (Sage and Reid, 1994; Bordribb, 1996; Vu et al., 1998). The high values of
A in the LS treatment in 2002/2003 were due to low intensity of the water stress.
The reduction of E as a result of the MS stress in the two seasons ranged from 64-87% in
beans, 87-88% in chickpea and 72-73% in cowpea at the end of the stress period. The
average relative reduction rate ofE during the MS stress was 2.6, 6.4 and 1.8% per day in
beans, chickpea and cowpea, respectively. The minimum E values in the LS treatment
ranged from 1.7-3.4,4.0-4.4 and 3.2-3.4 mmol m-2 S-1 between the two seasons in beans,
chickpea and cowpea, respectively (Tables 6.8 & 6.9). Relative to the respective control
measurements at the end of the stress, the LS treatment reduced E by 50-65% in beans,
117
22-49% in chickpea and 29-51% in cowpea in the two seasons. The relative decline rate
of E during the LS stress was 4.7-4.9, 3.0-5.5 and 2.8-3.7% per day. While the rate of
decline in E was higher in chickpea in the MS stress, it was higher in beans in the LS
stress.
Table 6.8. Rate of transpiration (mmol m-2 S-l) of three grain legume species under well-
watered (C) and water stressed conditions during flowering (MS) and pod ïtIling (LS)
periods in the 2002 season.
Species and water regime treatments
DAW BN CHP cOP BN CHP COP
Cll MS C MS C MS C LS C LS C LS
4 8.13 3.48 8.66 6.58 5.80 3.94 10.83 7.27 11.87 9.73 9.95 7.76
(1.32) (0.41) (1.32) (0.25) (1.53) (0.69) (0.50) (0.46) (0.38) (0.99) (1.03) (0.69)
6 6.40 3.40 7.90 3.58 6.47 3.58 9.37 5.00 11.63 7.59 9.47 7.09
(0.54) (0.46) (0.82) (0.67) (0.49) (0.40) (0.54) (0.29) (0.45) (0.48) (0.50) (0.63)
8 7.35 2.73 7.56 2.84 9.10 3.28 8.79 3.49 10.34 6.82 8.57 4.93
(0.89) (0.18) (0.39) (0.71) (0.59) (0.52) (0.31) (0.24) (0.59) (0.55) (0.40) (0.65)
10 8.42 2.98 11.20 2.44 7.21 2.72 8.75 3.23 11.92 5.98 8.40 4.95
(0.92) (0.38) (1.36) (0.64) (0.87) (0.72) (0.39) (0.19) (0.78) (0.51) (0.32) (0.33)
12 6.83 2.54 12.41 1.65 8.66 2.32 9.58 3.39 8.69 4.44 6.47 3.19
(0.51) (0.37) (1.40) (0.28) (0.72) (0.49) (0.57) (0.65) (0.55) (0.22) (0.31) (0.51)
ijNumbers in parenthesis refer to standard error of means.
Table 6.9. Rate of transpiration (mmol m-2 s") of three grain legume species under well-
watered (C) and water stressed conditions during flowering (MS) and pod ïtIling (LS)
periods in the 2002/2003season.
Species and water regime treatments
DAW BN CHP COP BN CHP COP
C MS C MS C LS C LS C LS
2 5.19 5.22 7.73 5.28 4.91 4.75
(0.70) (0.56) (1.21) (0.40) (0.54) (0.61)
4 6.68 6.74 5.96 6.30 3.89 4.80 5.60 4.90 7.00 5.60 3.80 4.10
(0.70) (0.56) (0.64) (0.66) (0.48) (0.44) (0.41) (0.35) (0.33) (0.56) (0.76) (0.51)
6 5.60 4.30 5.60 3.20 4.83 3.26 6.61 3.27 7.05 3.92 4.02 3.70
(0.53) (0.71) (0.43) (0.39) (0.20) (0.07) (0.12) (0.18) (0.19) (0.14) (0.17) (0.21)
8 6.80 2.90 3.86 2.38 4.43 3.35 2.50 1.54 7.40 1.77 5.81 3.66
(0.66) (0.33) (0.29) (0.49) (0.32) (0.21) (0.06) (0.05) (0.04) (0.07) (0.08) (0.08)
10 9.25 3.30 2.91 1.54 5.81 1.90 3.63 1.67 8.40 2.34 5.10 3.38
(0.46) (0.23) (0.32) (0.29) (0.54) (0.12) (0.10) (0.16) (0.27) (0.19) (0.12) (0.17)
12 5.93 1.66 6.20 2.50 4.00 1.50 5.67 2.57 7.41 2.74 4.73 3.33
(0.60) (0.19) (0.39) (0.31) (0.76) (0.12) (0.30) (0.21) (0.53) (0.47) (0.09) (0.18)
14 8.56 1.74 7.60 1.20 6.30 1.28 3.45 1.73 5.09 3.99 4.81 3.40
(0.21) (0.14) (0.79) (0.21) (0.17) (0.18) (0.43) (0.22) (0.35) (0.49) (0.35) (0.13)
16 7.54 0.98 4.81 0.57 5.60 1.58
(0.45) (0.19) (0.23) (0.14) (0.08) (0.07)
18 5.19 0.69
(0.70) (0.10)
Numbers in parenthesis refer to standard error of means.
118
Rate of E decline was lower in cowpea than in beans and chickpea in both the MS and LS
stresses. High E values in cowpea, despite high rs, suggest that the crop may have the
ability to maintain partial stomatal opening under water stress. Similar results were also
reported under controlled conditions (Cruz de Carvalho et al., 1998). Transpiration serves
as a major mechanism for cooling plant leaves through the evaporation process. As soil
water becomes limiting, however, the evaporation cooling is reduced leading to an
increase in leaf or canopy temperature (Jalali-Farahani et al., 1993) and a decrease in CO2
assimlation. Thus, a crop like cowpea that maintains its E under water deficit also
maintains better CO 2 assimilation.
6.3.4. Diurnal measurements
The hourly changes in \jiL, rs, A and E measured on December 10, 2002 (chickpea) and
December 16, 2002 (beans and cowpea). The two measurement dates have similar
photosynthetically active radiation (PAR), air temperature and VPD (Fig. 6.1). The
maximum PAR, air temperature and VPD recorded were 2145 J.Lmolm-2S-I, 29.6 °c and
2.58 kPa on 10 December 2002 and 2145 umol m-2 S-I, 28.6 oe and 2.90 kPa on 16
December 2002, respectively.
The minimum \jiL in chickpea was observed around 14:00 local time in both the stressed
and control plants while in beans it was observed between 12:00-14:00 in the control
plants and around 14:00 in the stressed bean plants (Fig. 6.2). In cowpea, the minimum
was observed between 12:00-14:00 and at 16:00 for the control and stressed plants,
respectively (Fig. 6.2). The hourly \jiL of cowpea plants declined from early morning until
16:00 and only started to recover after 16:00 in both the stressed and well-watered plants
unlike the case in beans and chickpea where the 'ilL started to recover just after 14:00.
Late afternoon recovery of \jiL was faster in the stressed than in the well-watered plants in
beans and cowpea while it was faster in the well-watered plants in chickpea. The hourly
rate of \jiL decline in the stressed plants was faster between 6:00-8:00 in beans and
chickpea while the change between 8:00-14:00 was very small. Difference in \jiL between
well-watered and stressed plants was higher in chickpea but less in cowpea and beans
(Fig. 6.2). The decline of diurnal \jiL with time until late afternoon in both well-watered
and stressed plants could be mainly due to the lag in water absorption vis-á-vis
transpiration as a result of rising radiation load and increasing VPD (Ehrler et al., 1978).
119
35 2500 4
30 Dec.102000
25 3
20 1500 ct: c
I-'" 15 oe( Q. 21000 Q. >
10 --iIE- VPD
5 500
0 0 0
600 800 1000 1200 1400 1600 1800 600 800 1000 1200 1400 1600 1800
35 2500 4
30 2000 Dec.16
25 3
20 1500 ct:
oe( c 2
l-'" 15 1000 0-Q. >
10
5 500
0 0 0
600 800 10001200 1400 1600 1800 600 800 1000 1200 1400 1600 1800
Time (hour) Time (hour)
Figure 6.1. Diurnal variation of photosynthetically active radiation (PAR, umol m·2 S·l), air
temperature (TA' "C) and weather station vapour pressure deficit (VPD, kPa) on 10 (top)
and 16 (bottom) December 2002.
The diurnal change of \jiL in the three species was highly associated with the diurnal
changes ofVPD (r = -0.81 * to -0.98**) and air temperature (r = -0.89* to -0.98**) both
of which are responsible for increased water loss from the plant (Squire, 1990).
The diurnal change in r, was very high in the stressed plants while it was very small in the
well-watered plants in all species (Fig. 6.3), that is, the stomata remained open in the
well-watered plants but closed in the stressed plants. This indicated that the decline in \jiL
in the control plants did not cause a major increase in r, in all the species suggesting the
existence of a threshold \jiL value beyond which the r, increases. On the other hand, the
decline in \jiL in the stressed beans and cowpea was significantly correlated with an
increase in rs (r = -0.87* to -0.91 *) (Table 6.10). While differences in r, were observed
by 8:00 in beans and cowpea, there was no any difference between the control and
stressed plants in chickpea despite the fact that there was big difference in \jiL between the
water regimes at the same time. Therefore, the weak correlation of hourly values of \jiL
and rs in chickpea (r = -0.32 to -0.59) in both the control and stress plants suggested that
120
low leaf water potential was not the trigger of stomatal closure in chickpea. Therefore, the
differential response of the species is mainly attributed to the relative nature of water
stress in that a water potential which induces stomatal closure in one species may have
little effect on another (Sperry, 2000).
0.0
l -0.5 Bean
~
i.ii -1.0..c. • "X: .......CD
0 -1.5 ....... .. "Z 2i::. .'
.C.L. ..•... =.... _•.• 'Z'"
CD
'Iii
;: -2.0
....... MS
_-C
'la
CD
....I -2.5
-3.0
600 800 1000 1200 1400 1600 1800
0.0
ac.a. -0.5
Chickpea
~.ii.i -1.0 :1:,.CCD. "
0 -1.5 "
.C..L. "CD.
la -2.0
;: -, • .:z:'l:, .' .
'la .....' .
CD . -2.5 ' •••••• ;L •••••• ' •• ·I......... <IC' •
....I
-3.0
600 800 1000 1200 1400 1600 1800
0.0
I~
l -0.5 Cowpea
~ I .... -....:t.. ............
i.ii .... =-=_ - ...cCD.. -1.0 ":Ii: ...... ' .. ' •• x: •••• _ .... _~ .......... .. " .."
0
.C.L. -1.5
CD
.'I;.ii.: -2.0
la
CD
....I -2.5
-3.0
600 800 1000 1200 1400 1600 1800
Time (hour)
Figure 6.2. Diurnal variation of leaf water potential in beans, chickpea and cowpea under
mid-season water stress (MS) for 14 days and well-watered (C) conditions at a semi-arid
environment. Vertical bars indicate standard errors.
121
25
Bean
20 C
•..••.. MS
r . . ...- "15 .,";' ".......... .E " ,u
-(I..-)... 10 . ,.r-' >1,.'
.s:
5 ,.
,, .'
..: ~
0
600 800 1000 1200 1400 1600 1800
25
Chickpea
20
-~ 15
-u(I).-....10
O5 ~"r' •••=••• ::z::" ••:••• ~~==~==~~~~~
600 800 1000 1200 1400 1600 1800
25 ,--------------- --,
- Cowpea
20 ..-r .... ., ... ·-·I·-~.. -..~....-.... l-'
.'.'
.'
- 15 ....I" -'" - ...... I"~u
-(I-).. 10 ........ ...
5 .....
.
:i. OE I I I I
~
0
600 800 1000 1200 1400 1600 1800
Time (hour)
Figure 6.3. Diurnal variation of stomatal resistance in beans, chickpea and cowpea under
mid-season water stress (MS) for 14 days and well-watered (C) conditions in a semi-arid
environment. Vertical bars indicate standard errors.
122
25
Beans
20
...... ••••••• fv1S
';'
I/) 15 C
"I
E
"0
E.0. 10
-•5:! .E .5 .,,<C .·1...... '. r. ....... ·1:- ...-....!
0
600 800 1000 1200 1400 1600 1800
25
Chickpea
...... 20
';'
I/)
"I
E 15
"0
E
0
.2 10-E<C
5
.. .. ···I--. ...... .':I:
.... . .·I·. ... ... . ..'
.:E•••••••••• l:
0
600 800 1000 1200 1400 1600 1800
25
Cowpea
20
.r-
'I/)
N
'e 15
"0
§... 10
•-5e:!<C .. ..I-·.5 • ... • .._ -li:. .....•••• ' •. ;r:.-" ····.r.·····..·..
o +-------~----~------~------~------~------~~
600 800 1000 1200 1400 1600 1800
Time (hour)
Figure 6.4. Diurnal variation of the rate of photosynthesis in beans, chickpea and cowpea
under mid-season water stress (MS) for 14 days and well-watered (C) conditions at a semi-
arid environment. Vertical bars indicate standard errors.
123
10
Beans
8
-- ....... MS___ C';'
1/1 6
Ol
E
'0
E 4
-Ew
2 _ .::J::- ••
":a;:_ - •••• -. :E••••
0 -' ••• :a;: ••• -.-. _ • .:t:
600 800 1000 1200 1400 1600 1800
10
Chickpea
8
'-fn- 6
Ol
E
'0
E 4
-Ew
2
-- .·I-'··-,-.;[.... ..-'-~
........ - ...... "2:_
0 -'- •• _;r:
600 800 1000 1200 1400 1600 1800
10
Cow pea
8
'-fn- 6
Ol
E
'0
-E 4Ew
2 .... .. ~_ .. _-. .. .... T~ .. - ,.·-·I- ...."",
0
600 800 1000 1200 1400 1600 1800
Time (hour)
Figure 6.5. Diurnal variation of the rate of transpiration in beans, chickpea and cowpea
under mid-season water stress (MS) for 14 days and well-watered (C) conditions in a semi-
arid environment. Vertical bars indicate standard errors.
124
The diurnal change of r, was positively correlated with VPD (r = 0.79 to 0.94*) and air
temperature (r = 0.61 to 0.96**) (Table 6.10). Bates and Hall (1982b) also found strong
correlation of stomatal conductance with VPD in cowpeas. Stomatal response to the leaf-
to-air VPD was found to be strongly associated with the survival strategies adopted by
dryland C4 grasses and C3 species (Maroco et al., 1997). Increases in air temperature
results in greater water vapour gradients between leaf and air which promotes stomatal
closure (Sage and Reid, 1994). Stomata response to both air temperature and vapour
pressure gradient between leaf and air under non-stress conditions is variable in C3 plants.
Table 6.10. Correlation coefficients among diurnal measurements of leaf water potential (""
MPa), stomatal conductance (rs, s cm"), rate of photosynthesis (A, umol m-2 s'I), rate of
transpiration (E, mmol m-2 s"), air temperature (TA'DC),vapor pressure deficit (vpD, kPa)
and photosyntheticaUy active radiation (PAR, umolm? S-1).8
Species Water regime S'l' rs A VPD TA PAR
Bean C rs -0.46
A -0.86* 0.37
VPD -0.92** 0.63 0.63
TA -0.98** 0.61 0.80* 0.96**
PAR -0.76 0.25 0.98** 0.49 0.69
E -0.71 0.25 0.92* 0.45 0.64 0.97**
MS rs -0.91*
A -0.12 -0.84*
VPD -0.86* 0.82* -0.23
TA -0.95** 0.89* -0.15 0.96**
PAR -0.76 0.81* -0.11 0.54 0.70
E -0.06 -0.17 0.93* -0.39 -0.24 -0.10
Chickpea C rs -0.32
A -0.83 -0.77
VPD -0.81* 0.94** 0.67
TA -0.89* 0.98** 0.79 0.98**
PAR -0.65 0.53 0.88* 0.31 0.48
E -0.69 0.75 0.93* 0.40 0.55 0.99**
MS rs -0.59
A -0.34 -0.69
VPD -0.80* 0.79 0.28
TA -0.88* 0.77 0.23 0.98**
PAR -0.55 0.20 -0.17 0.19 0.38
E -0.34 -0.52 0.68 0.00 -0.09 0.40
Cowpea C rs -0.88*
A -0.31 0.47
VPD -0.97** 0.94** 0.33
TA -0.91 * 0.96** 0.55 0.96**
PAR -0.46 0.56 0.92* 0.46 0.65
E -0.44 0.28 0.94** 0.13 0.38 0.78
MS rs -0.87*
A -0.57 -0.43
VPD -0.98** 0.88* 0.53
TA -0.94** 0.93** 0.64 0.96**
PAR -0.51 0.72 0.47 0.57 0.74
E -0.001 0.44 0.75 0.02 0.23 0.43
an = 6; *, ** Correlation coefficients significant at 5 and 1% P level, respectively. gThe VPD is a mean of
half hourly measurements between 10:00 to 13:30 local time.
125
For example, stomatal conductance increases with temperature at constant VPD between
leaf and air in many species, while in some species such as sweet pepper and caster bean,
stomata open with increasing temperature under high and low VPDs, respectively (Sage
and Reid, 1994 and references there in). There was also a positive correlation between r,
and PAR, particularly under stress conditions. A similar condition was reported in wheat
where diurnal differences in stomatal conductance between stressed and unstressed plots
were related to differences in net radiation (Choudhury, 1985).
There was no significant difference in the hourly changes of A between the control and
stressed plants before 8:00 in the morning in any of the species (Fig. 6.4). Rate of
photosynthesis attained its maximum between 10:00-14:00 in beans and at midday in
chickpea and cowpea in the well-watered plants while it attained its lowest in the stressed
plants at the same time (Fig. 6.4). Therefore, the maximum difference in A between the C
and MS treatments was observed at midday in all the species. A was slightly higher in the
morning and late afternoon than the midday values in the stressed plants, particularly in
chickpea. In. beans, the stressed plants had similar A values to that of the C early in the
morning (8:00) indicating the adaptation strategy of the plant to maintain its
photosynthesis under water limited conditions. The pattern in the diurnal change of E was
similar to that of A in chickpea and cowpea, except that recovery in the late afternoon
was slow (Fig. 6.5). Beans showed a drastic decline in E after midday in contrast to the
observation in A where reduction was slow between 12:00-14:00.
Diurnal change of A and E in the control plots was strongly correlated with the change of
air temperature and PAR and moderately with 'JIL and VPD (Table 6.10). However, both
A & E were not correlated with any of the these parameters in the stressed plants because
photosynthesis in the stress plants responded favourably to these weather variables until
10:00 in the morning but declined afterwards when the weather elements increase to the
highest. These results indicate the different response of the stressed plants to diurnal
change of weather conditions from that of well-watered plants. A was positively
correlated with rs in the controls whereas the correlation was negative in the stressed
plants indicating that there was a threshold r, beyond which A declined.
One of the effects of water stress is an increase in leaf temperature (TL) of the stressed
plants as a result of stomatal closure and reduced transpiration. Most of the diurnal
126
20
Photosynthe sis
16
« ...
.!: 12
III
Q) .,
u
c 8 00 '.
Q..). - ..... ,_
Q)
!I: .................o 4 .. , .....o -...._
-.._,o ..o ..0......
-4
-8.0 -6.0 -4.0 -2.0 0.0 2.0 4.0
Differences In LT
4
Transpiration t::. BN (y = -0.81x + 1.85, R2 = 0.67)
2 t::. .CHP (y = -0.99x + 1.76, R2 = 0.90)
.... (y = -1.75x - 0.33, R2 = 0.67)
I-
.!: 0
III .....
Q) -2 .....u ...
e
.Q.). " .o 0 ........ ' ..
Q)
!I: -4 ...
' .. ........ ...
' ..
0 ...........0...
-6 o •"',
-8
-1 o 2 3 4 5 6 7
Differences In E
4
(y = 0.40x + 2.93, R = 0.84)
2 (y = 0.28x + 1.12, R2 = 0.40).-.
0
~ (y = 0.33x + 1.40, R2 = 0.63)....
I- 0
.= o ..,...-__""- .
III -2 _. • .' ••....:.t. ........Q)u o ._ _.•.• .,;: ...;:e; :;...:';'" - Stom ata I resista neeeQ..). -4 .
Q)
!I:
0 o-6 o •
-8
-20 -15 -10 -5 o
Differences in rs
Figure 6.6. Relation of diurnal differences in leaf temperature (TL, DC) to diurnal differences
in rate of photosynthesis (A, umol m-2 sol), transpiration (E, mmol m-2 sol) and stomatal
resistance (r, s cm") between well-watered and mid-season water stressed plants of beans
(BN,-), chickpea (CHP, ----) and cowpea (COP, ---) at a semi-arid environment.
127
variation in TL difference between the control and stressed plants was explained by
differences in r, (84, 40 and 46%) and E (67, 90 and 67%) for beans, chickpea and
cowpea, respectively (Fig. 6.6). Therefore, TL in the stressed plants integrates the other
effects of water stress, and it explains most of the differences in photosynthesis.
Differences in diurnal TL between the C and MS treatments explained 80, 98, and 74% of
the diurnal variation of the difference in A between the two treatments in beans, chickpea
and cowpea, respectively (Fig. 6.6). The graphs indicate that when leaf temperatures of
the well-watered and stressed plants are equal, there is no difference in A between the
stress and control treatments. Therefore, the major diurnal difference in the rate of
photosynthesis between well-watered and stressed plants is a drastic increase in TL in the
stressed plants which results in disruption of enzymatic activities which are responsible
for C02 assimilation.
It has been indicated that thermal extremes directly damage biochemical systems through
protein denaturation, loss of membrane integrity, photoinhibition and ion imbalance (Sage
and Reid, 1994). An increase in air temperature also results in greater water vapour
gradients between leaf and air and promotes stomatal closure and TL increase which in
turn limit photosynthesis (Sage and Reid, 1994, and references there in). Leaf
temperatures are determined by the energy exchange processes involving radiation,
convection, and transpiration (Sivakumar, 1986). When soil water becomes limiting,
stomatal closure occurs resulting in reduced transpiration, increased heat load on the
canopy and a consequent rise in leaf temperatures which can reach up to 10°C above air
temperature in some C4 species (Pearcy et al., 1971). High temperature, which occurs
during periods of water stress, can affect C02 assimilation and reduce water use
efficiency (Baldocchi et al., 1981b; Baldocchi et al., 1985). Species may have different
responses and mechanisms to regulate their leaf temperature. For example, differences in
\jiL explained 77 and 76% of the variability in TL difference between the stressed and
unstressed plants in beans and chickpea, respectively while there was no significant
correlation in cowpea (Fig. 6.7). Comparison of the species indicated that cowpea had a
better mechanism of maintaining its photosynthesis under high leaf temperature as
compared to beans and chickpea.
128
4,- --,
Le af w ate r pote ntial
2 6
~i 0 +- o__ ~_._,~----
/--,_--------~.-.'-'-.------~
-2 ":"/.- 66 • e. .'
;~ _9' 6 .
~ -4 .'
c '
o o 6BN (y=7.6x-11.3,R2=0•.76)-6
• CHP ( y = 14.2x - 8.8, R2= 0.77)
o COP ( y = 11.3x - 5.7, R2= 0.23)
-8 ~----------------------------------------~
0.0 0.5 1.0 1.5 2.0
Differences in leaf water potential (MPa)
Figure 6.7. Relation of diurnal differences in leaftemperature (TL)to diurnal differences in
leaf water potential between well-watered and mid-season water stressed plants of beans
(BN), chickpea (CHP) and cowpea (COP) for two measurement dates (10 and 16 December
2002) at a semi-arid environment.
6.3.5. Relationship between ASW, 'JIL, rs, A and E
The response of plants to the environmental conditions is complex and interrelated.
Therefore, it is necessary to understand the interrelationships of, at least, the major
environmental and physiological variables to explain and model the response of plants to
environmental constraints such as water deficit. Here an attempt is made to investigate the
relation of soil and leaf water status, stomatal resistance, photosynthesis and transpiration
in beans, chickpea and cowpea. To do so, data were combined for the well-watered (C)
and stressed (MS and LS) treatments to investigate the relationships during the whole
reproductive period of the three species.
There was a strong linear relationship (R2>0.70 except in chickpea in 2002/2003)
between 'JIL and ASW (%) in all seasons and species (Fig. 6.8). On average, 'JIL declined
by 0.01 MPa per a percent decline in ASW in beans and cowpea and by 0.03 MPa per a
percent decline ASW in chickpea during both seasons. The consistent relationship
between 'JIL and ASWacross seasons suggests that 'JIL can be easily determined from a
measurement of soil water in grain legumes. Since water flows from a higher energy to a
lower energy level, water movement in the soil-plant-atmosphere continuum (SPAC) is a
function of soil water status, plant water status, atmospheric vapour pressure deficit and
129
50 -4.0 50 -4.0
BN-2002 LWP = 0.01ASW - 1.84 BN-2002/03
-3.5 -3.5
40 • MS & C-rs
R2 = 0.86 40
• LS & C-rs -3.0 rs = 24.50e-O·04ASw 'iii-3.0 D..
oMS & C-lWP
& r, =
~
25.92e-O·o2ASw R2 = 0.71
6.LS C-LWP
30 2 = -2.5 30
-2.5 :më
...... R 0.93 CD
';"E -2.0 LWP = 0.01ASW - 1.80 -2.0 8.
u R2 = ...0.71 .CD.
.!.!.!.
20 oA.€JA 20 -1.5 cv
O.l o~·&l:i' •• A -1.5 ~.Q,Q ~·····t.·.o~~ • ••. ~ .•00 Q) ..
• éfIF-G.GG.':_O c
.v.
-1.0 6. -1.0 ~
10 10 6. 6."~
-0.5 -0.5
0 0.0 0 0.0
0 20 40 60 80 100 0 20 40 60 80 100
50 -4.0 50 -4.0
CHP-2002 rs = 73.75e-O·05ASW CHP-2002/03= -3.5 -3.50 LWP 0.03ASW - 3.92
40 40 R2 = 0.76
0
" R2 = 0.94 -3.0 -3.0 'i6. '. ~
'á"o 0 LWP = O.02ASW - 3.11
"
~ 30 41·~ . -2.5
0
30 .'..0 0 R2 = 0.54 -2.5 ~'E
u rs = 26.53e-O·03ASW . • .. -2.0 '. 0" -2.0 i
.!.!.!.
.
= .
.. ...
. R2 0.94 6. 0 .CD.lO 20 -1.5 20 66 o- •• -1.5 cv
6 ê ''S,5,'' ~ ..~..6 CV
-1.0 • 6 '-0•• 0• -1.0 s10 10 6 ~' •
-0.5 6 66 -0.5
0 0.0 0 • • 0.0
0 20 40 60 80 100 0 20 40 60 80 100
50 -4.0 50 -4.0
COP-2002 = COP-2002/2003LWP 0.01ASW - 1.89 -3.5 -3.5
40 R2 = 0.88 40
-3.0 LWP = 0.01ASW - 1.93 -3.0 'iiiQ.
rs = 26.56e-O·03ASW -2.5 R2 = 0.81 ~30 30 -2.5~ R2 = 0.94 • ~
'E -2.0 -2.0
u i
.!!!. 20
... -1.5
20 ...
lO ~,?, "C6 -1.5 .CD.
-1.0 ''''·.0,.,., ~ .s
CV
0 -1.0
6 ..
~..
10 10 rs = 5E+07ASwa·95 tc CV•• A .~'h CD
-0.5 ...J
R2 = -0.50,73
0 0.0 0 • 0.0
0 20 40 60 80 100 0 20 40 60 80 100
/!ISW (o/~ /!ISW (o/~
Figure 6.8. The relationship between available soil water (ASW), stomatal resistance (rs) and
leaf water potential (LWP) in three grain legumes under water stress (MS &LS) and weU-
watered (C) conditions in 2002 (left) and 2002/2003 (right) seasons. (BN= beans, CHP =
chickpea, COP = cowpea).
130
the resistance encountered at each level. Therefore, the above mentioned values could be
useful in relating soil water to plant water status so that an efficient system of irrigation
scheduling could be devised for these crops.
On the other hand, the relation between ASWand r, was not linear and was more
appropriately explained by an exponential function (Fig. 6.8). With a decline in ASW, the
r, increased exponentially at rate of 0.02-0.04, 0.03-0.05 and 0.03 s cm-I per ASW (%) in
beans, chickpea and cowpea, respectively in the two seasons. The increase in r, with the
decline in ASW (%) was higher and faster in cowpea in 2002/2003 and the relationship
was best explained by a power function (Fig. 6.8). The relationship of rs to ASW
indicated that stomata closure was triggered at higher soil water in 2002 than in
2002/2003. The exponential relationship between rs and ASW was broken down into two
linear regressions to find the threshold value above which r, increased drastically with a
decrease in ASW. Linear regressions of the data points and finding the intersection point
of the regression lines indicated that stomatal closure was initiated when ASW reached
62.3, 62.4 and 86.2% in the high temperature season (2002) and 55.0, 45.5 and 65.4% in
the more mild temperate season (2002/2003) in beans, chickpea and cowpea, respectively.
The closure of stomata at higher ASW in 2002 could be due to the high temperature
prevailing during that season as compared to 2002/2003.
Linear regressions of the data shown in Fig. 6.9 between the control and stressed plants
indicated that a trigger of stomatal closure occurred at an average threshold \jiL value of
-1.48, -2.08 and -1.11 MPa in beans, chickpea and cowpea, respectively. This shows a
more rapid closure of stomata at higher \jiL in cowpea than beans and chickpea which also
agrees with other reports on the same crop (Shackel and Hall, 1983; Diallo et al., 2001).
However, the value obtained for beans is higher than the values reported (-0.6 to -0.9
MPa) for complete stomatal closure of this crop in another study (Costa Franca et al.,
2000). The lower \jiL thresholds obtained here compared to the previous reports could be a
result of cultivar and/or environmental differences between the studies. On the contrary,
Cruz de Carvalho et al. (1998) observed a complete stomatal closure in beans at higher
plant water status than a drought tolerant cowpea cultivar under severe water stress
conditions. In addition, the relation of r, with ASW indicates that stomatal closure in
cowpea occurs at higher soil water status than in beans and chickpea (Fig. 6.8). This
better stomatal adjustment behaviour of cowpea in response to water deficit makes it one
131
of the droughts avoiding C3 species (Squire, 1990; Cruz de Carvalho et al. (1998). As
mentioned earlier, the lower 'l'L for the initiation of stomatal closure in chickpea is mainly
associated with the low stomatal adjustment behaviour of the species. In general, stomatal
closure in chickpea was observed at lower ASWand 'l'L than beans and cowpea in both
seasons suggesting that the contribution of stomatal closure to drought avoidance of the
crop is limited compared to the other two species.
40
°oBN
(y = 0.04e-3.44lWP, R2 = 0.86) 2002
CHP (y = 0.44e-1.02lWP, R2 = 0.92) ~
[).COP (y = 0.06e-3.14lWP, R2 = 0.87) 30 E
-u(I)Cl)
-cuCIS20 .I!!
(.IC.)l).
- , CIS~ ~o -CIS
° _\.Q 0\ 10 E0
,-D,-~ ° 3. -VJ
~- .
-B- - - - _0l:'lD-- '
0
-4 -3 -2 -1 0
40
[).
= 2 = 2002/03oBN (y 0.02e-4·01lWP, R 0.84)
°CHP (y = 0.17e-1.70lWP, R2 = 0.83) [). ~30 E
[).COP (y = 0.0ge-3.41lWP, R2 = 0.62) -u(I)Cl)
u
[). 20 -cCIS0 .I!!(I)
-C..l).10 CIS-CISE0(I)
0
-4 -3 -2 -1 0
Leaf water potential (MPa)
Figure 6.9. The relationship between leaf water potential and stomatal resistance during the
reproductive period of three grain legumes under water stress (MS &LS) and well-watered
(C) conditions in 2002 (top) and 2002/2003 (bottom) seasons. BN= beans, CHP = chickpea,
COP = cowpea.
132
A and E were linearly related to ASW for beans and chickpea in both seasons whereas the
relation was best explained by a power function in cowpea (Fig. 6.10). There was no
significant difference between the two seasons in the response of A and E to ASW in any
species. Therefore the average decline in A per a percent decline in ASW was 0.24 and
0.29 umol m-2 S-1 and that of E was 0.10 and 0.12 mmol m-2 S-1 in beans and chickpea
respectively. The relationship between photosynthesis and transpiration with available
soil water in cowpea indicate that the crop has a capacity to photo synthesise and transpire
at a higher rate under favourable water supply and also maintain a slower rate of decline
in A and E under low soil water conditions. As shown in Fig. 6.10, rate of photosynthesis
had decreased by 5 fold when the ASW reached 47.9, 43.3 and 39.0% in the high
temperature season (2002) and 30.8, 43.1 and 36.2% in the mild temperature season
(2002/2003). The photosynthesis rate of beans was affected at higher ASW in the high
temperature season compared to its performance in the mild temperature season. This
may suggest that high temperatures under conditions of water stress could affect the
productive capacity of the plant when it is coupled with water deficit.
The present results indicated that the rate of photosynthesis was more sensitive to the
decline of soil water than the rate of transpiration in all species. This could be explained
by the sensitivity of photosynthesis to increased leaf temperatures even before the stomata
dose completely. It is suggested that limitations to C02 assimilation in beans was caused
by metabolic restrictions that can be differentiated between those occurring in the range
of20 to 30°C and 30 to 35 oe (Pastenes and Horton, 1996) ..
Both A and E were exponentially related to \jiL in all species and seasons (Fig. 6.11). Rate
of photosynthesis declined exponentially with \jiL at a rate of 4.00, 1.18 and 2.31 umol m-2
S-1 in 2002 and 3.36, 0.82 and 1.88 umol m-2 S-1 in 2002/2003 in beans, chickpea and
cowpea, respectively. Similarly, E declined exponentially with \jiL at a rate of 2.60, 0.80,
1.94 mmol m-2 S-1 in 2002 and 3.30, 0.82, 1.07 mmol m-2 S-1 in 2002/2003 in the three
species order as above (Fig. 6.11). The data shows that the decline in A and E with \jiL
was higher in beans while it was lowest in chickpea. Although chickpea had the fastest
declining \jiL, it showed the lowest rate of exponential decline in A and E. This implies
that either \jiL may not be a good indicator of plant water status in chickpea or the crop
may have other mechanisms to maintain its A and E at a lower plant water status. Most
133
25~---------------------------, 25.- ~
• M3&C-A 13N-2002 13N-2002l2003
.LS&C-A
20 6 M3&C-E 20oLS&C-E •
A = 0.25ASW - 2.71
6
15 A = 0.24ASW - 6.51
= •
R2= 0.84
15 " 6
R2 0.90 E = O.OOASW-1.37 .""
E = 0.11ASW -1.51 • o R2= 0.00 •10 § 10
R2= 0.85 o ... <>.1\ "lo" o 0
6 • • o
6 'X6 9.···
5 5 • " clp6.•.• ' .·ti·6 6
• • 0 •.0" "Li66 ti>
@.00..····· 6'o".u
O+-----.------r-----.-----.----~ O+-~--,------r----~----~----~
o 20 40· 00 80 100 o 20 40 00 80 100
25~----------------------------__, 25.-----------------------------~
Of'-2000 " Of'.2002l2003
20 A = 0.31ASW - 8.40 20 A = O.26ASW- 6.21
R2= 0.89 • R2= 0.80
15 15
w E = 0.13ASW -1.33 E = 0.10ASW - 2.49
« 10 R2= 0.81 6 .• ' 10 R2= 0.82
" " .. ....•. 6 0' 0
5 5
O+-----,------r-----.----~r_--~
o 20 40 00 80 100 o 20 40 00 80 100
25~--------------------------__. 25.-----------------------------~
COP-2002 COP-2002l2003
20 A= 0.04ASW·31 • 20 A= 0.01ASw·m
R2= 0.88 R2= 0.82 • •
15 15 ."• •
E = 0.04ASW·16 " ,,'\
w E = 1.1ge°.02ASW •" "
';10 R2= 0.76 10 R2= 0.82 ". .o
6 6 ~•
•• ,..- 0
•• • .. ..,•.• '- 0 0
5 • 6 60 ••·•... 5
1\-.....6 0
• ~"'V
o
O+-----.-----.------r-----r----~ O+-----.------.-----.-----,r---~
o 20 40 00 80 100 o 20 40 00 80 100
ASW(~
Figure 6.10. The relationship between available soil water (ASW), rate of photosynthesis (A,
umol m·l S'l) and transpiration (E, mmol m·2 S'l) in three grain legumes under water stress
(MS &LS) and well-watered (C) conditions in 2002 (left) and 2002/2003 (right) seasons. BN=
beans, CHP = chickpea, COP = cowpea.
134
~---------------------- ~ 25 25
• MS&C-A BN-2002 BN-2002l03
• LS&C-A 36LWP
oMS&C-E 20 A = 764.35e3- 20
6LS&C-E ~ = 0.85
15 E = 274.53e3.30LWP 15
A = 2172.28e4.00LWP w
R2 = 0.82 ~ = 0.84 .(10 10
E= 202.21 e2.60LWP
R2 = 0.71 5 5
o o
-4 -3 -2 -1 o -4 -3 -2 -1 o
,-------------------,---------, 25 25
A = 113.35e1.18LWP CHP-2000 CHP-200212003
R2 • 20 A = 35.36eo.82LWP= 0.88 • 20~ = 0.67
E = 34.314eo.80LWP 15
E = 13.60eo.82LWP 15 w
~=O.84 cf
10 ~ =0.68 10
5 5
o
-4 -3 -2 -1 o -4 -3 -2 -1 o
25 25
A = 223.29e2·31LWP • COP-2002 COP-200212003
R2 = 0.82 • 20 A = 70.23e1.88LWP 20
• ~ =0.71
E = 70.30e1.94LWP • • 15 15
R2 = 0.64 • E = 10.93e1.07LWP w
10 ~=0.73 10 <C
5 5
~----~-------'--o-----r----~ 0 o
-4 -3 ~ ~ 0-4-3 -2 -1 o
Leaf water potential (MPa) Leaf water potential (MPa)
Figure 6.11. The relationship between leaf water potential (LWP), rate of photosynthesis (A,
umol m-:Zs'') and transpiration (E, mmol m-:Zs") in three grain legumes under water stress
(MS &LS) and well-watered (C) conditions in 2002 (left) and 2002/2003 (right) seasons.
BN= beans, CHP = chickpea, COP = cowpea.
probably, the maintenance of A and E under low 'ilL in chickpea could be related to the
ability of the crop to adjust osmotically in response to decreasing \jiL (Morgan et al., 1991;
Leport et aI., 1998) compared to cowpea in which osmotic adjustment is absent (Diallo et
al., 2001). The first decline in A was observed at an average 'ilL of -1.2, -1.5, and -1.1
MPa and it was affected by as much as five fold when \jiL reached -1.5, -2.7 and -1.5 MPa
in beans, chickpea and cowpea, respectively.
135
25~ ~ 25,- -,
• tv'S &C-A BN-2002
.. LS&C-A BN-2002l2003
20 6 o tv'S &C-E 20 A= 22.07e.o·16rs
liLS&C-E
R2 = 0.97
15 A = 25.1ge.o·2Ors 15
R2 = 0.86 E = 8.35e.o·16rs
we 10 R2 = 0.95E = 11.20e.o·13rs 10
R2 = 0.75
5 5 •
o+- .- .- ~
o 5 10 15 o 5 10 15 20 25 30 35 40
25,- , 25,- -,
I.. CHP-2000 CHP-2002l2003
20 20 ..
A = 31.04e.o·24rs \.. A = 17.5ge.o·14rs
2
15 R =0.91 15 R2 = 0.91
w
.{, E = 14.05e.o·16rs E = 6.80e.o·14rs
10
R2=0.84 10 R2 = 0.91
5 5
o+- .- ~--------~ •O+---.---.----r~_r--_r--_.--_.--~
o 5 10 15 0 5 10 15 20 25 30 35 40
25,----------------------- -, 25,- __
.. COP-2002 COP-2002l2003
20 • A = 23.16e.o·18rs 20 ..... A = 15.91e.o·05rs
~=0.87 R2 = 0.81
15 15 ~
E = 10.96e.o·16rs
w E = 4.67e.o·03rs
.{, ~=0.78
10 10 R2 = 0.76
5
5 ~l----'0 -.•.0. _. _. •
O+- .- .- ~ . - - .. e-· . - -..._-... .JJ __O+---.---.----r--_.---,---,---,--~
o 5 10 15 0 5 10 15 20 25 30 35 40
Stomatal resistance Stomatal resistance
Figure 6.12. The relationship between stomatal resistance (rs, s cm"), rate of photosynthesis
(A, umol m-2 s'') and transpiration (E, mmol m-2 sol) in three grain legumes under water
stress (MS &LS) and well-watered (C) conditions in 2002 (left) and 2002/2003 (right)
seasons. BN= beans, CHP = chickpea, COP = cowpea.
136
The present values are lower than the threshold \jiL at which a marked decline in the rate
of photosynthesis and stomatal conductance was observed in six cool-season food
legumes in a Mediterranean-type environment (Leport et al., 1998) as well as in narrow
leafed lupin (Turner and Henson, 1989). Marked decrease in A is reported to be
associated with \jiL at which ABA concentration increased in the leaves causing stomata to
close (Turner and Henson, 1989).
A and E were also exponentially related to r~in all the species in both seasons (Fig. 6.12).
Similar relationships were reported in soybean (Baldocchi et al., 1985). The rate of
decline in A and E with an increase in r, was higher in 2002/2003 than in 2002. The
exponential rate of decline in A and E with an increase in r, was similar in beans and
chickpea while it was lower in cowpea (Fig. 6.12).
The rate of decline in A with an increase in rs was higher than that of E suggesting that A
is more sensitive to stomatal closure than E does. This could be as a result of cuticular gas
exchange under conditions of stomatal closure which more favours the passage of water
vapour than C02 as observed in grape (Boyer et al., 1997). Linear regressions of the data
in Fig. 6.12 indicated that both A and E were detrimentally affected when the r, increased
on average above 8.0, 8.1 and 6.1 s cm" in beans, chickpea and cowpea, respectively.
This and other similar studies (Cruz de Carvalho et ai, 1998; Costa Franca et al., 2000)
indicate that early stomatal closure is the main reason for the decline of C02 assimilation
under water stress.
6.3.6. Relationship of ASW, A, E, r, \jiL and TL to VPD
The correlation coefficients between VPD (measured within canopy and at different
heights above crop canopy) and the other parameters are presented in Table 6.11. It was
found that ASW was negatively correlated with VPD measured within canopy and half a
meter above canopy but not with the VPDs measured above this height (Table 6.11).
Although not correlated to VPD directly, ASW is reported to influence the sensitivity of
stomata to air humidity (Calvet, 2000). The correlations of \jiL and rs with the VPD
measured at the different heights in the canopy were sigirificantly negative and positive,
respectively in all three species in both seasons (Table 6.11). A similar relation between
VPD and stomatal behaviour (conductance or resistance) has been reported in a number
of crops and environments (Schulez and Hall, 1982; Grantz and Meinzer, 1990; Aphalo
137
and Jarvis, 1991; Maroco et al., 1997). The response of stomata to changes in air VPD
can be related to the changes in the rate of transpiration (Mott and Pankhrust, 1991;
Schulze, 1994). Therefore, the degree of stomatal response to leaf-to-air VPD depends on
a number of factors including water stress (Tewolde et al., 1993; Schulze, 1994), air
humidity (Kawamitsu et al., 1993; Hinckley and Braatne, 1994), radiation (Grantz et al.,
1987; Hinckley and Braatne, 1994), plant phenological stage (Jones, 1992), concentration
of CO2 and plant growth substances (Hinckley and Braatne, 1994).
Table 6.11. Correlation ofVPD measured at different heights of crop canopy with available
soil water (ASW, %), leaf water potential (\jiL, MPa), stomatal resistance (r., s cm"), rate of
photosynthesis (A, umol m-2 S-1), rate of transpiration (E, mmol m-2 S-1) and leaf temperature
(LT, °C) in the three grain legumes grown under water stress and non-stress conditions for
two seasons,"
Parameter 2002 2002/2003
VPD
Within 0.5m (AC)D lrn (AC) 2m(WS) 0.5m (AC) lm (AC) 2m(WS)
canopy
Bean
ASW -0.60** -0.58** -0.32 -0.37 -0.14 -0.14 -0.14
'l'L -0.62** -0.61 ** -0.46* -0.55* -0.52** -0.52** -0.60***
r, 0.67** 0.64** 0.55* 0.65** 0.44* 0.39* 0.41 *
A -0.48* -0.52* -0.43 -0.50* 0.15 0.17 0.18
E -0.42 -0.52* -0.45* -0.48* 0.37* 0.41 * 0.48**
TL 0.64** 0.58** 0.45* 0.57* 0.73*** 0.70*** 0.74***
Chickpea
ASW -0.72** -0.64** -0.36 -0.39 -0.28 -0.26 -0.21
'l'L -0.62** -0.55* -0.47* -0.53* -0.66*** -0.63*** -0.60***
r, 0.52* 0.41 0.44* 0.46* 0.50** 0.48** 0.39*
A -0.51 * -0.42 -0.41 -0.43 -0.08 -0.08 -0.09
E -0.36 -0.35 -0.37 -0.35 0.11 0.17 0.25
TL 0.50* 0.28 0.25 0.32 0.76*** 0.71 *** 0.68***
Cowpea
ASW -0.47* -0.38 -0.35 -0.41 -0.09 -0.05 -0.01
'l'L -0.32 -0.29 -0.37 -0.43 -0.36* -0.31 -0.33
r, 0.39 0.26 0.43 0.49* 0.36* 0.28 0.27
A -0.35 -0.51 * -0.60** -0.64** 0.16 0.25 0.16
E -0.25 -0.38 -0.56* -0.56* 0.41 * 0.44* 0.39*
TL 0.44* 0.62** 0.66** 0.69** 0.78*** 0.80*** 0.74***
+.n = 20 and 30 in 2002 and 2002/2003 respectively; *, **, *** value significant at 5, 1 and 0.1% Plevel,
respectively. bAC= above crop canopy, WS = weather station. "The VPD used here is a mean of hourly
measurements between 10:00 and 2:00 local time.
A was significantly negatively correlated with VPD (r = -0.60*) under well-watered
conditions for beans and cowpea while there was no correlation for chickpea (Fig. 6.13).
This suggests that the stomata of beans and cowpea are sensitive to changes in VPD of
the air even under high soil water status. Under water stress conditions, however, A was
significantly correlated with VPD (r = -0.58*) in cowpea while there is no correlation in
beans and chickpea (Fig. 6.13). This could be due to the overriding role of water stress in
controlling the stomatal response of beans (stomata remain close irrespective of changes
138
25 25
..... Beans Beans
1/1 20 0 20
'1 02002 y= -0.99x+ 7.77E 15
"0 t:. 2002/2003 15 R2 = 0.05
E ~
0... 10 10 Ot:.
.2 0 t:.y = -2.75x + 23.07
! 5 m 0R2= 0.36 5 Q
cC Ó t:.t:. 000 0-8
. 0 t:.t:.0
0 2 3 4 0 2 3 4
25 25
..... Chickpea 0 00
20 0 Chickpea y = -2.77x + 14.95
1/1 t:. t:.t:. t:. 0
'1 at:. 0
0 20
0 t:.t:.
R2=0.19
E 15 ~t:. 15 Ot:.
"0 t:. ti>
E 10 10
0...
.2 y = 0.29x + 16.78 5
! 5 R2= 0.00
cC 0 0
0 2 3 4 0 2 3 4
25 25
..... Cow pea Cow pea y= -2.82x+ 16.9520 20
in t:.t:. R2 = 0.34
'1
E 15 15 t:.
"0 10 0 0E 10
0... y = -1.83x + 20.61
.2 5
! R2= 0.35
5
cC 0 0
0 2 3 4 0 2 3 4
VPD at 2m height (kPa) VPD at 2m height (kPa)
Figure 6.13. The relationship between rate of photosynthesis (A) and vapour pressure deficit
of the air (VPD) measured at 2 m height for the well-watered (left) and stressed (right)
plants of beans, chickpea and cowpea in 2002 and 2002/2003 seasons. Data presented is for
non-cloudy day measurements and for both the mid-season and late season stress periods
combined over the two seasons. The VPD is a mean of hourly measurements between 10:00
and 2:00 local time.
in VPD) unlike cowpea where the stomata still remained responsive to changes in VPD
and soil water deficit.
Lack of any correlation between A and VPD under both well-watered and stressed
conditions in chickpea suggest that the stomata of this crop are not responsive to changes
in VPD. Generally, increased VPD is associated with a decrease in A though the
relationship depends on other factors like leaf temperature, leaf conductance and water
139
stress (Sage and Reid, 1994). TL was significantly positively correlated with the VPDs in
all crops and seasons (Table 6.11). Increases in leaf temperature are primarily responsible
for increases in VPD of the air in the vicinity of the canopy (Sage and Reid, 1994). Since
the VPD at weather station (2 m) is correlated with most of the parameters, it could be
used as representative of canopy VPD in crop-weather relation studies.
6.3.7. Post stress recovery
The extent to which photosynthetic capacity is maintained during periods of water stress
and the ability for rapid recovery after re-watering is important in crop adaptation in dry
environments. Thus, an understanding of the recovery of photosynthesis and other
physiological processes from water stress may aid in identifying drought resistance
mechanisms in crop plants. Measurements of A, E and stomatal conductance (&) made at
the end of mid-season stress and three days after re-watering are presented in Table 6.12
as percent of the control at each measurement time. Recovery was calculated as the
difference between the two measurements. While the recovery of transpiration was very
high in chickpea, the recovery of photosynthesis was higher in beans and cowpea (Table
6.12). On the other hand, the recovery of & was higher in beans and chickpea compared
to cowpea. Higher recovery of A while the recovery of & is lower in cowpea indicates the
existence of reversible non-stomatal factors that could reduce photosynthesis during
periods of water deficit in this crop. One of such factors is leaf temperature (Table 6.12).
Table 6.12. Recovery of physiological processes upon re-watering after MS stress in
2002/20~0~3.-+Param-e-te-r -------M-e-as-ur-em-e-nt-------B-ea-n---C-hi~ck~pe~a---C-ow-p-ea----
E(%) Before re-watering 87 91 83
After re-watering 43 24 41
Recovery (%) 44 67 42
Before re-watering 95 96 97
After re-watering 50 53 67
Recovery (%) 45 43 30
A(%) Before re-watering 90 74 86
After re-watering 44 40 41
Recovery (%) 46 34 45
Tddifference, oe) Before -1.8 -7.2 -4.3
After 2.6 -4.9 0.7
+, The values indicated are percent of reductions (or difference for temperature between C and MS) relative
to the control just before re-watering and three days after re-watering. Percent of recovery is the difference
between the relative percentage values before and after re-watering.
140
The leaves of the stressed plants were hotter than the controls by -1.8, -7.2 and -4.3 -c in
beans, chickpea and cowpea, respectively during the stress but became cooler than the
controls by 2.6 "C in beans and by 0.7 °c in cowpea after three days of re-watering (Table
6.12). However, the previously stressed leaves of chickpea remained hotter than their
control counter parts by 4.9 °c after re-watering. Comparison of the temperature
differences between stressed and control plants before and after re-watering showed that
the decline of leaf temperature up on re-watering was higher in cowpea (5°C) than in
beans (4 °C) and in beans than in chickpea (2.3 °C). Therefore, fast recovery of
photosynthesis from stress in cowpea is mainly due to fast declining leaf temperatures
while the recovery in beans is due to both lowering of temperatures in the stressed leaves
and rapid recovery of stomatal conductance. Chickpea, which had slower reduction of
leaf temperature upon re-watering, showed the lowest recovery of its photosynthesis,
while the recovery of its stomatal conductance resulted in the fast recovery of
transpiration. Since chickpea had the lowest leaf water potential during the stress, the
lowest recovery of A could also be explained by possible damage in the photosynthesis
apparatus caused by the drought. In others studies, slower and partial recovery of
photosynthesis following rehydration was observed in beans (Cruz de Carvalho et al.,
1998). In comparing beans and cowpea, Cruz de Carvalho et al. (1998) found higher
recovery of photosynthesis in cowpea than in beans. The low recovery in beans was
suggested to be due to the damage of the photosynthetic apparatus. However, since the
recovery of photosynthesis of cowpea and beans is similar, one may assume that there has
no damage to photosynthetic apparatus of beans in the present study. In pigeonpea, a
complete recovery of photosynthesis has been observed after seven days of re-watering
(Lopez et al., 1988).
6.3.8. Estimation of photosynthesis and transpiration from other measured
parameters
Studies which involve field measurement of photosynthesis, transpiration and plant and
soil water status are big challenges in many developing countries due to lack of
equipment. Therefore, simple equations that can relate these variables with other
parameters, which are cheaper to measure, could help scientific and educational activities
in these countries. In this study, a stepwise regression was employed to estimate A and E
from weather and other physiological parameters in the reproductive stage of the three
grain legume species.
141
Table 6.13. Estimation of midday rate of photosynthesis (A, umol m-2 ft), transpiration (E,
mmol m-2 s-t), leaf water potential (\jiL, MPa) and available soil water (ASW, %) from
weather, soil and plant parameters in three grain legumes using stepwise regression for 2002
season.
Season 2002 R
(n= 20)
Beans
A = -11.5+ 0.0824ASW +1.72E + 0.105TL + 0.367VPD(lm) 0.95
A = -6.50 + 0.223ASW + 0.15VPD(2m) 0.80
E = -2.59 + O.l04ASW + 0.45VPD (canopy) 0.75
E = -0.68 + 0.092ASW + 0.015VPD(2m) 0.73
'l'L = -1.42- 0.035rs + 0.019E + 0.084VPD (2m) - 0.06VPD (canopy) 0.92
'l'L = -1.80 + 0.0061ASW-0.0072VPD (2m) 0.86
ASW = 79.0-5.70rs + 2.83E + 2.19VPD (2m) 0.92
ASW = 105-7.60rs +2.66VPD (2m) 0.89
Chickpea
A = 4.13 + 0.120ASW + 1.28E-0.232TL 0.94
A = -12.8 + 0.299ASW +1.50VPD (2m) 0.80
E = 28.6-0.077ASW+7.45LWP-1.15 VPD (lm) + 1.20VPD (canopy) 0.75
E = -3.00 + O.l13ASW + 0.701VPD (2m) 0.56
'l'L = -3.56 + 0.0199ASW- 0.045rs + 0.028E + 0.04VPD (O.5m) 0.97
'l'L P = -4.26 + 0.028ASW + 0.0781VPD (2m) 0.95
ASW =148.0 + 32.3'1'L- 4.23VPD (0.5m) 0.96
ASW = 152 + 33.2'1'L - 4.16VPD (2m) 0.95
Cowpea
A = 0.68 + 3.39 VPD (canopy) - 4.21 VPD (0.5m) + 0.132 ASW (%) + 0.871 E 0.97
A = 5.76 + 0.170ASW-1.97VPD (2m) 0.77
E = 11.8-0.748rs-0.897VPD (lm) + 2.15VPD (canopy)- 1.84VPD (0.5m) 0.86
E = 2.26 +0.074ASW- 0.543VPD (2m) 0.67
'l'L = -2.04 + 0.009ASW + 0.041VPD (canopy) 0.91
'l'L = -2.04 +0.0086ASW + 0.036VPD (2m) 0.89
ASW = 191.0 + 29.8 LWP-5.35rs -1.43TL 0.96
ASW = 307-5.69 TL - 3.74VPD(2m) 0.53
Regression équationsfrorii 1lie be-st variables identified withthéstepwise regression and
from easily available variables such as ASWand VPD are shown in Tables 6.13, 6.14 &
6.15 for each season and for data pooled over the two seasons. The regression equations
were highly significant (P< 0.01) for all species. The coefficient of determination (R2)
was higher for each individual season compared to the pooled data (Table 6.13 to 6.15).
However, the R2 for the pooled data of A was also higher (~0.80) indicating that A could
be determined from the parameters with reasonable accuracy. On the other hand, the 1- for
E (for each season and pooled data) was low, particularly in cowpea, suggesting that the
success of estimating E from the parameters indicated in Table 6.13 to 6.15 was low.
Estimation of A and E from ASWand VPD (at weather station) gave a better result in
beans (R2 ~ 0.79) compared to chickpea (R2 ~ 0.42) and cowpeatk" ~ 0.16).
142
Table 6.14. Estimation of midday rate of photosynthesis (A, umol mo2 s"), transpiration (E,
mmol mo2 s"), leaf water potential (\jiL, MPa) and available soil water (ASW, %) from
weather, soil and plant parameters in three grain legumes using stepwise regression for
2002/2003 season.
Season 2002/2003 R
(n=30)
Beans
A = -5.85-2.26VPD(2m) + 0.091ASW + 1.83E + 0.248 TL 0.94
A = -8.82 + 2.44VPD (2m) + 0.251ASW 0.81
E = -4.20+ 2.14VPD (2m) + 0.072ASW-0.119rs 0.81
E = -5.40+ 1.94VPD (2m) + 0.090ASW 0.79
\jiL= -1.26-0.129VPD(2m)+ 0.006ASW-0.0017rs 0.89
\jiL= -1.43-0.156VPD (2m)+ 0.0082ASW 0.87
ASW = 120+ 1.99VPD (2m)+ 62.7\j1L+ 3.72E 0.85
Chickpea
A = -10.9-2.73VPD(2m) + 6.48\j1L+ 2.12E+ 0.774 TL 0.88
A = -3.19 + 033VPD(2m) + 0.202ASW 0.43
E = 2.09+ 1.77VPD (2m) + 0.016ASW+ 1.28LWP- O.l46rs 0.64
E = -2.78 + 1.05VPD(2m) + 0.071ASW 0.52
\jiL= -0.715-0.284VPD(2m) + 0.012ASW-0.0298rs-0.03TL 0.90
\jiL= -2.21-0.360VPD(2m) + 0.021ASW 0.83
ASW = 97.6+ 11.0VPD (2m) + 34.60\jlL 0.74
Cowpea
A = -15.9-2.32VPD(2m) + 0.0875ASW + 1.22E + 0.666TL -0.273rs 0.80
A = -2.73 + 1.38 VPD (2m) + 0.182ASW 0.36
E = 3.05+4.31 VPD(1m)-2.79VPD(0.5m) + 0.02rs-0.06ITL 0.22
E = 2.94 + 0.972VPD(2m)-0.0083ASW 0.16
\jiL= -1.65-0.0113VPD(2m) + 0.012ASW-0.0020rs 0.87
\jiL= -1.46-0.138VPD(2m) + 0.0101ASW 0.68
ASW = 123.00 + 7.89VPD(2m) + 65.6\j1L 0.85
Regression equations for determining \jiL and ASW from the other parameters are also
given in Table 6.13 to 6.15. A better estimate ofleaf water potential can be obtained from
ASWand VPD in the three crops (~ = 0.68-0.87). A reasonable estimate of ASW (%) can
also be found from VPD and \jiL in chickpea and cowpea (~= 0.74-0.85) and including E
on these variables in beans (~ = 0.85).
These equations are useful for determining plant and soil water status from other
parameters for the purpose of irrigation scheduling or adjusting other management
practices in these species for environments similar to the experimental site. The equations
are also important to understand the relationship of the different weather and
physiological parameters so as to model the growth and development of the species under
water deficit environments.
143
Table 6.15. Estimation of midday rate of photosynthesis (A, umol mo2so\ transpiration (E,
mmol mo2 sO\ leaf water potential ('!'L, MPa) and available soil water (ASW, %) from
weather, soil and plant parameters in three grain legumes using stepwise regression for data
combined over two seasons.
2002 and 2002/2003 seasons data combined R2
(n=50)
Beans
A = 13.5-1.37VPD (2m) + 11.9'1'L+ 1.65E + 0.217TL 0.88
A = -2.52 + 0.217 ASW- 0.419VPD (2m) 0.67
E = -4.56+ 0.098ASW + 1.16VPD(2m) 0.73
'l'L = -1.07-0.090VPD(2m)+ 0.0022ASW-0.031rs 0.85
'l'L = -l32- 0.I44VPD(2m)+ 0.0059ASW 0.78
ASW = 87.1 + 29.3LWP-1.45rs + 4.85E 0.83
Chickpea
A = 0.02 + 6.27'1'L + 1.49E + 0.397T L 0.86
A = -6.96 + 0.248ASW + 0.719VPD(2m) 0.61
E = -0.385 + 1.65VPD(2m) + 0.084ASW-0.133rs 0.61
E = -5.49 + 1.54 VPD (2m) + 0.103ASW 0.59
'l'L = -0.724-0.289VPD(2m) + 0.014ASW - 0.027rs 0.88
'l'L = -2.20-0.376VPD(2m)+ 0.020ASW 0.84
ASW = 85.7 + 12.8VPD(lm) - 8.70VPD (0.5m) + 21.0'l'L +1.97E-0.69rs 0.80
ASW = 100 + 11.0 VPD (2m) + 33.9'1'L 0.71
Cowpea
A = -0.50-0.88VPD(2m)-1.67VPD (Im) + 8.28'1'L-0.216rs + 1.42E + 0.604TL 0.79
A = -1.46-0.062VPD(2m) + 0.191ASW 0.49
E = -8.90 + 0.103ASW-5.44'1'L + 0.029rs 0.43
E= -0.60 + 0.049ASW + 0.8l3VPD (2m) 0.29
'l'L = -1.43-0.124VPD(2m)+ 0.0103ASW- 0.0321E 0.81
'l'L = -1.41-0.15VPD(2m) + 0.0088ASW 0.76
ASW = 110.0 + 5.51VPD(2m) + 58.6'1'L-0.355rs + 3.43E 0.77
ASW = 128.0 + 9.38VPD(2m) +69.2'1'L 0.61
6.4. Summary and Conclusion
Differences in leaf water potential between well-watered and stressed plants of beans and
cowpea were very small despite large variations in soil water, stomatal resistance,
photosynthesis and transpiration. However, a higher decline in leaf water potential with
declining soil water was observed in chickpea, and the decline was greater in the late-
season stress than the mid-season stress. The magnitude and rate of photosynthetic
reduction was higher and faster in the mid-season than in the late-season stress in all
species, and among the species the rate of photosynthesis declined faster in chickpea than
in beans and cowpea. Cowpea had the lowest reduction in its rate of photosynthesis under
severe water stress compared to the other two species.
The diurnal course of leaf water potential was different for the control and stressed plants
of the three species indicating the differential response of the species to the changing
weather conditions. Moreover, the diurnal physiological response of the stress plants to
144
the changing weather conditions (temperature, radiation and VPD) was different from that
of the controls. Such differential responses are partly because of the overriding effect of
water stress on other responses (e.g. A and VPD relationship in beans), and partly
because of plants' adaptation mechanism to survive and maintain productivity under
stress environment while maximizing productivity under favourable environment.
Increase in leaf temperature was found to be the major factor for the decline of net
photosynthesis in the stressed plants through its effect on stomatal adjustment,
transpiration and possibly enzymatic activities. Although most of the physiological
parameters considered were highly correlated with mean daytime VPDs measured closer
to the canopy, significant correlations were also obtained with the mean daytime VPD
measured at the weather station. Because of the ease of measurement and its availability,
the use of standard VPD can give good representation of canopy VPD in crop-weather
relation studies.
Cowpea closes its stomata at higher leaf and soil water potential followed by beans
whereas the stomata of chickpea remain open under lower leaf and soil water status. This
fast stomatal response makes cowpea a more drought-avoiding crop than beans and
chickpea. On the other hand, the rate of photosynthesis and transpiration decline with leaf
water potential was lower in chickpea than in beans and cowpea, suggesting that chickpea
may have other mechanisms (possibly osmotic adjustment) to maintain productivity
despite its fast declining plant water status. Therefore, two contrasting scenarios were
observed. Chickpea had the lowest and fastest decling leaf water potential while cowpea
had the highest leaf water potential and fastest closing stomata. Both species, however,
maintained similar rate of leaf photosynthesis under severe water stress conditions. This
indicates the different adaptation mechanism of the species though the goal (maintaining
leaf photosynthesis) remains the same. Nevertheless, post-stress recovery of
photosynthesis was lower in chickpea compared to the other two species. The three
species have one thing in common that their rate of photosynthesis can be estimated from
a few weather and physiological parameters and soil water content with reasonable
accuracy.
Chickpea, as a cool-season food legume, is considered to be more drought susceptible
than the warm-season food legumes such as beans and cowpea. In the current study,
however, chickpea showed considerable performance which is almost similar to the
145
warm-season grain legumes in terms of maintaining its photosynthesis under severe water
stress and high temperature. Therefore, this species has a promising potential to be grown
in dry environments with higher temperature and evaporative demand compared to the
environments where it is currently grown in Ethiopia.
146
CHAPTER 7
Comparison of Yield and Yield Components Response of Three Grain Legumes
Species to Variable Water Supply During the Reproductive Stages
7.1. Introduction
Drought is a major limiting factor to yield and increased productivity of tropical crops.
Grain legumes are commonly grown under rainfed conditions in the semi-arid tropics and
are subjected to drought as a result of water shortage at one or more of their growth stages
(e.g. Turk et al., 1980a; Kumar et al., 1996; Kumar and Abbo, 2001). Drought at various
stages in the crop cycle is the major reducer of yield in many grain legumes (e.g. Kumar
and Abbo, 2001; Jeuffroy and Ney, 1997) and cereals (e.g. Saini and Westgate, 2000).
Plant growth and development can be affected by water deficit at any time during the
crop life cycle. However, the extent and nature of damage, the capacity of recovery and
the impact on crop yield depends on the developmental stage at which the stress occurs
and the type of crop species or cultivars involved (French and Turner, 1991; Kirda and
Kanber, 1999; Simane et al., 1993; De Costa et al., 1999; Saini and Westgate, 2000).
Compared to the vegetative period, drought at some time during the reproductive phase is
responsible for more reduction in grain yield of many crops (Sinoit and Kramer, 1977;
Calvache and Reichardt, 1999; Kumar and Abbo, 2001; Jamieson et al., 1995; NeSmith
and Ritchie, 1992; Saini and Westgate, 2000). The sensitivity of yield reduction as a
result of drought also varies among stages within the reproductive period in grain
legunies. For "example, bean yield was reported to be more susceptible to water stress at
flowering than pod-filling stage (Calvache and Reichardt, 1999) while cowpea grain yield
was found to be more susceptible to pod-filling than flowering period water deficit (Turk
et al., 1980a). Therefore, characterization of developmental stages and the determination
of sensitive periods in a given environment are important to avoid or minimize the effect
of water stress on grain legumes (Jeuffroy and Ney, 1997).
The number of pods initiated and their rate of growth, the number of seeds per pod and
their mass are important determinants of harvest index in grain legumes (Lawn and Ahn,
1985; Siddique and Sedgley, 1986; Jiang and Egli, 1995). Moreover, yield is a function of
sink size and its subsequent filling by the source, both of which are affected by water
stress depending on its timing and severity with respect to plant growth stages (Blum,
1996). The ability to mobilize pre- and post- anthesis assimilates during the pod filling
147
period (especially during water deficit) is also an important mechanism in stabilizing
yield of grain legumes (Goldsworthy, 1984; Evans, 1993; Turner et al., 2001).
Although, there have been a number of studies that compared the effect of drought during
various stages of reproductive development on crop yield in different environments, it is
almost impossible to compare crops at an equivalent tissue water status (Saini and
Westgate, 2000) because of environmental, experimental and seasonal differences.
Therefore, the objectives of this study were (1) to compare the response of yield and
yield components of three grain legume species, viz., beans, chickpea and cowpea, to
different water regimes, (2) to determine the physiological cause of yield variation, and
(3) to identify the most drought-sensitive reproductive stage for each species under the
same environmental, seasonal and experimental conditions.
7.2. Materials and Methods
7.2.1. Field experiments
Detail explanations on experimental site, material, design, cultural practices, and
irrigation schedule are given in Chapter 3.
7.2.2. Measurements
7.2.2.1. Number of flowers and pods after water stress
The numbers of flowers and pods (NP) per plant were counted from five randomly
selected plants in the central rows of each plot for the stress treatment and its
corresponding control. For the LS treatment, counts were made at maturity. The numbers
of primary and secondary branches were also counted from the five randomly selected
plants in 2002/2003.
7.2.2.2. Yield and yield components
At final harvest time, all plants from an area of 6.4 m2 in the central four rows were
harvested by hand. Five plants were selected randomly from the harvested area and the
number of pods (NP) and number of seeds (NS) per pod were counted for each plant, and
then the NP and NS mo2 calculated. Above ground total biomass was determined after
drying the harvested plants in the open air for 10 days. After hand threshing, seeds were
separated from the straw, weighed and adjusted to 12.5% moisture after determining the
148
moisture content using a Grain Moisture Tester. The equation used for adjusting the yield
to the specified moisture content was as follows:
y . =((12.s-me)*YJ+y (7.1)
adj 100
where Yadj is moisture adjusted yield, Y is unadjusted yield and me is measured moisture
content (%). Hundred seed mass (SW) was determined by weighing 100 dried seeds of
each plot using a digital sensitive balance. Harvest index (HI) was calculated as the ratio of
grain yield to total above ground biomass.
7.2.2.3. Partition coefficient
Mean crop growth rate (Cr, g m-2 °Cd-I) was calculated as the linear rate of increase in
above ground biomass over the total growth period expressed in thermal time and pod
growth rate (Cp, g m-2 °Cd-I) was calculated as the linear rate of increase in pod dry matter
for the thermal time between 50% flowering and maturity of each species for each
treatment. The partitioning coefficient (P) was calculated as explained in Duncan et al.
(1978), Greenburg et al., (1992) and Williams and Saxena (1991) as
Cp
p=- (7.2)
Cr
using the above two growth rates.
7.2.2.4. Statistical Analysis
Analyses of Variance (ANOVA) and mean separation (Least Significant Difference,
LSD) were conducted using the NCSS statistical program (Hintze, 1997). Correlation and
regression analyses were also performed whenever necessary.
7.3. Results and Discussion
7.3.1. Water use and evaporative demand
In order to help the discussions in this chapter, a short summary of the magnitude of
water stress in each treatment is presented in Table 7.1 as the seasonal crop evaporative
deficit (l-ET/ETo) which relates the crop water demand (ET) to the evaporative demand
(ET0) of the site. The seasonal crop evaporative deficit was less than 0.15 in the controls
in all species, but ranged between 0.17-0.27 in beans, 0.18-0.39 in chickpea and 0.16-0.33
in cowpea in the MS, and between 0.14-0.37 in beans and 0.15-0.34 in chickpea and 0.14-
149
0.35 in cowpea in the LS treatment in the three seasons (Table 7.1). This shows that the
plants in the LS treatment were generally more stressed in 2002/2002 and 2002 seasons
than in 2002/2003 season in all species. As mentioned previously, the low stress intensity
in the LS treatment in 2002/2003 was due to cloudy weather conditions during the stress
period that reduced the evaporative demand.
Table 7.1. Crop evaporative deficit (l-(ETIETo» of beans, chickpea and cowpea plants
under mid-season (MS) and late season (LS) water stress and well-watered (C) conditions at
a semi arid environment.
Season Species l-(ETIETo)
C MS LS
200112002 Beans 0.10 0.17 0.22
Chickpea 0.01 0.25 0.27
Cowpea 0.03 0.33 0.33
2002 Beans 0.12 0.27 0.37
Chickpea 0.15 0.39 0.34
Cowpea 0.09 0.24 0.35
2002/2003 Beans 0.07 0.22 0.14
Chickpea 0.05 0.18 0.15
Cowpea 0.06 0.16 0.14
7.3.2. Effect of water stress on yield and yield components
Biomass and yield were higher in 2002/2003 than in 200112002 and 2002 seasons.
Analysis of variance showed significant differences among water regimes for all
parameters in all seasons except for HI in 2002 and final biomass and SW in 2002/2003
(Table 7.2). Significant differences were also observed among species for all variables
except HI in 200112002 and 2002 and yield in 2001/2002. The water regime x species
interaction was also significant for NP m-2, NS m-2 and SW in 2001/2002 and 2002
seasons (Table 7.2). Pooled over species, yield reductions were 77,68 and 37% in the MS
treatment and 65, 30 and 23% in the LS treatment for the first, second and third seasons,
respectively indicating that water stress, irrespective of its timing, resulted in reduced
grain yield in all seasons.
Although the lowest grain yield was recorded in the MS treatment, there was no
significant difference in grain yield between the MS and LS treatments in any of the
species and seasons except in 2001/2002 (Tables 7.3 and 7.4). In 200112002, the MS
treatment resulted in significantly lower yield than the LS treatment in beans (Table 7.3).
Relative to the control, yield reduction of the species over season ranged from 34-92% in
150
beans, 26-87% in chickpea and 33-61 % in cowpea in the MS treatment, and 22-77% in
beans, 15-58% in chickpea and 19-56% in cowpea in the LS treatment (Fig. 7.1).
Table 7.2. Mean squares in the analysis of variance of biomass, seed yield, number of pods
(NP) and number of seeds (NS) per meter square, 100 seed mass (SW) and harvest index
(HI) for three grain legume species grown under three water regimes in three seasons,"
Source di NPm-7 NSm-7 SW Biomass Seed yield HI
200112002
REP 2 29789.1 404196.8 5.208 144721.6 88356.5 0.001
Water 2 621575.7*** 7956288.1 ** 81.8* 12463024.5* 6826001.8* 0.150*
regime(WR)
Error 4 9816.3 214890.6 6.5 1182788.4 437537.9 0.005
Species (Sp) 2 866587.8*** 1534872.4·· 283.7··· 1535775.8* 28723.9 0.020
WRxSp 4 313564.1·** 752279.1 ** 19.7* 1075428.3· 315170.9 0.015
Error 12 5242.8 131160.5 5.7 263670.1 181429.4 0.007
Total 26
2002
REP 2 217.9 84137.4 6.3 748600.3 6708.0 0.016
WR 2 410234.7·· 7047980.3··· 57.9** 16454573.1 *** 2704967.6· 0.025
Error 4 21705.9 92801.3 2.1 239192.1 252060.4 0.024
Sp 2 586748.5**· 6042233.4·*· 147.6··* 1222957.7*· 891557.1 ** 0.011
WRxSp 4 71563.8** 1975375.4*** 14.5*** 2397078.9*** 830180.5** 0.034
Error 12 11638.2 156819.1 0.60 175329.9 114650.8 0.019
Total 26
2002/2003
REP 2 104474.7 374127.1 21.2 596573.4 67948.9 0.000
WR 2 379524.5* 10948589.4· 1.4 6822149.0 1680161.2* 0.020*
Error 4 50326.3 691061.6 3.5 1591115.2 129919.1 0.002
Sp 2 2471188.6*** 87999971.7*** 150.7*** 4428756.3** 4202278.2*** 0.081***
WRxSp 4 133767.6 146463.8 7.9 265839.1 46027.7 0.001
Error 12 57177.4 325397.9 7.0 353142.5 148617.7 0.004
Total 26
., *, ••• variation significant at 5, I and 0.01% P levels, respectively.
1.0
0.9
aBN
0.8 .CHP
c 0.7 DCOP
0
;; 0.6
u
::::I
".0.. 0.5Cl)
~ 0.4
III
>= 0.3
0.2
0.1
0.0
MS LS MS LS MS LS
2001/2002 2002 2002/2003
Seasons and water stress treatments
Figure 7.1. Relative yield reduction of three grain legumes due to mid-season (MS) and late
season (LS) water stress with respect to well-watered conditions in three seasons.
151
Except for 2002, chickpea was more sensitive to flowering period water stress than beans
and cowpea. The lowest relative yield reduction in 2002 in chickpea was mainly due to
the high temperature condition which affected the yield of plants in the control treatment.
On the other hand, yield reduction in beans was the highest in this season (Fig. 7.1) when
flowering water stress was coupled with high temperature.
Both beans and chickpea were more sensitive to the MS than the LS stress across all
seasons. As compared to beans and chickpea, the gap in yield reduction due to the MS
and LS stresses in cowpea is closer suggesting that water stress at either of the stages
could have a similar effect on final grain yield. Moreover, except in the LS treatment in
2002, cowpea was less sensitive than beans to yield reductions due to both the MS and LS
stresses. Under the LS stress, beans was more sensitive than cowpea and chickpea in the
first season whereas cowpea was more sensitive than the other species in the second
season. The yield response of the three species to the LS stress was almost similar in the
third season because of low intensity of the stress as explained above.
Although the amount of yield reduction varied among species and time of water stress,
the present results indicated a significant amount of yield reduction due to both the MS
and LS stresses in all three grain legumes. A significantly higher reduction of grain yield
due to reproductive than vegetative season drought has also been reported in a number of
grain legumes including chickpea (Singh, 1987; Sivakumar and Singh, 1987), beans
(Acosta Gallegos and Shibata, 1989; Tesfaye, 1997), cowpea (Turk et al., 1980a), mung
bean (Pannu and Singh, 1993; De Costa et al., 1999) and peanut (Wright et al., 1991).
Compared to the control, both the MS and LS treatments reduced NP m-2 and NS m-2 in
all the species (Fig. 7.2 and 7.3). However, NP m-2 was more affected by the MS
treatment while NS m-2 was equally affected by the MS and LS treatments in the milder
temperature seasons (200112002 and 2002/2003). In the higher temperature season
(2002), nevertheless, the MS treatment had significantly lower NS than the LS.
Water stress also affected SW in 200112002 and 2002 but not in 2002/2003 which could
be as a result of low VPD of the air during this particular period. The reduction of SW
.from the control was significantly higher in the MS than in the LS treatment in 2002 but
similar for both the MS and LS treatments in 200112002 (Table 7.3).
152
Table 7.3. Mean biomass production at harvest, grain yield, number of pods (NP) and
number of seeds (NS) per meter square, 100 seed mass (SW) and harvest index (Hl) of three
grain legumes under three water regimes in 2001/2002 and 2002 seasons.
Variable" Water regime 200112002 2002
Bean Chickpea Cowpea Bean Chickpea Cowpea
Cll A488.0b· A 1489.3 a A 281.3 c A417.0b A 1162.7 a A 446.7 b
NPm-2 MS B 328.7 a C 246.0 ab A 173.3 b B 165.0 b C423.3 a B 159.0 b
LS B 250.0 b B 683.0 a A 179.0b A 450.7 b B 702.7 a A 295.3 b
C A 2906.3 a A 2979.0 a A 2643.3 a A 2849.3 b A 1162.7c A 4015.3 a
NSm-2 MS B 1944.3 a B 246.0b B 1382.7 a B 808.3 a B 846.0 a C 1137.0 a
LS B 1476.0 a B 683.0b B 1557.3 a A 3043.0 a B 848.3 c B 2281.0 b
C A20.7b A 34.5 a A 21.0 b A 22.7 b A 27.7 a A 20.7 c
SW (g) MS B 16.3 c B 25.7 a A 20.7 b C 12.8 c B 23.7 a A 19.3 b
LS B 14.7 c B 24.3 a A20.0b B 17.0 c B 25.0 a A20.3 b
C A4619.8a A 4057.3 a A 4182.3 a A 5690.2 a A 3748.6 b A 4960.7 a
Biomass MS C 1468.7b C 1348.9 b B 3213.0 a C 1560.0 b B 2491.6 a B 2300.8 a
(kg ha")
LS B 2546.9 a B 2453.1 a B 2890.6 a B 4096.5 a B 3075.2 b B 2312.0 c
C A 2361.3 A 2276.0 A 1704.2 A 2368.7 a A 1117.6 b A 1330.9 bGrain yield
MS B 460.9 B 292.7 B 733.2 C199.6 b A 822.9 a B 513.4 ab(kg ha")
LS B 540.6 B 953.5 B 746.0 B 1744.9 a A 949.1 b B 682.5 b
C AO.50 AO.56 A 0.41 0.41 0.29 0.27
HI MS BO.30 BO.22 BO.23 0.13 0.37 0.22
LS BO.21 BO.39 BO.26 0.43 0.30 0.29
.. Each value is the mean of three replicates; XWater regime: C = control, MS= mid-season stress, LS = late-season
stress
• Within species, values across water regimes preceded by the same capital letter are not significantly different at
P~0.05; within a water regime, values across species followed by the same lower case letter are not significantly
different at P~0.05.
Table 7.4. Mean biomass production at harvest, grain yield, number of pods (NP) and
number of seeds (NS) per meter square, 100 seed mass (SW) and harvest index (BI) of three
grain legumes under three water regimes in 2002/2003.* éJ
Variable Water regime Bean Chickpea Cowpea
C 651.0 ± 7.87 1794.7 ± 209.24 346.3 ± 42.61
NPm-2 MS 509.7 ± 53.49 1074.7 ± 298.93 258.0 ± 29.38
LS 468.3 ±34.l4 972.3 ± 206.09 196.7± 17.15
C 4819.7 ± 94.41 2879.0 ± 449.09 3957.7 ± 537.79
NSm-2 MS 3000.0 ± 538.69 1074.7 ± 298.93 1690.3 ± 314.40
LS 3238.3 ± 377.86 1261.7 ± 309.80 1599.3 ± 138.70
C 24.0±0.00 30.3 ± 1.20 22.7±0.33
SW MS 23.3 ±0.33 28.7±2.03 23.7 ± 1.20
LS 21.3 ±0.33 32.0±3.79 24.7 ± 1.33
C 6798.2 ± 603.69 5641.6 ± 104.17 4912.4± 187.79
Biomass MS 5235.7 ± 722.87 4006.2 ± 625.00 3840.2 ± 668.39
LS 4670.9 ± 660.19 4037.4 ± 700.71 3903.0 ± 180.42
C 3221.5 ± 350.14 1765.4 ± 98.00 1907.6 ± 163.11
Grain yield MS 2117.8 ± 379.02 931.9 ± 268.54 1274.3 ± 223.17
LS 2449.0 ± 308.05 1373.1 ± 269.11 1544.9 ± 94.96
C 0.47 ± 0.02 0.31 ± 0.02 0.39±0.02
HI MS 0.41 ± 0.01 0.23 ±0.02 0.33 ±O.OO
LS 0.54±0.05 0.31 ±0.04 0.40 ± 0.01
* Explanations as in Table 7.2. a According to the ANOVA analysis, the WR x Sp interaction was not significant and
hence mean separation was not conducted for the interaction means. Values indicated next to means are standard errors.
153
80,--------------------------, 80,--------------------------,
... 200112002 200112002
~ 60 ~ 60
:s ë.5c ...8.
140
~ !
...40
.C.G. §
~ 20 'e8 20
a: Q.
o o
BN CHP COP BN CHP COP
80~--------~~------------~ 80,--------------------------.
2002
i1:... 60
8.
j 40
E
c:s
'8 20
Q.
Bean CHP COP Bean CHP COP
80~------------------------~ 60~------------------------~
200212003 200212003
j 60 1:§ CG
c ë.5... 40
140 8....
'2 !
CG E:s 20li; 20 c
~ '8
JI: Q.o o
BN CHP COP BN CHP COP
Treatment Treatment
Figure 7.2. Number of flowers and pods per plant at the end of the mid-season (MS) stress
(left) and number of pods per plant at maturity in the late season (LS) stresses (right) as
compared to the control (C) for three grain legume species in three seasons. BN = beans,
CBP = chickpea, COP = cowpea, LP = lower half of the plant canopy, UP = upper half of
the plant canopy. When visible, vertical bars indicate standard error of means.
154
60
MS [J Primary
o Secondary
UI
QI
..c. 40uco
....
a...0
QI
.a
E 20
::l
Z
o
BN CHP COP
60
LS
UI
QI
.c
u.cea. 40
....
a..
0.
QI
.a
E 20
:::I
Z
0
BN CHP COP
Treatment
Figure 7.3. Number of primary and secondary braches per plant after the mid-season stress
(MS) and maturity of the late-season stress (LS) as compared to the control (C) for three
grain legume species in 2002/2003. When visible, vertical bars indicate standard error of
means.
Yield component response of the species to the water deficit treatments was not
consistent across seasons (Table 7.3 to 7.4). In cowpea, NP m-2 was affected by the MS
and LS treatments in 2002 and 2002/2003 seasons, respectively but not affected in
2001/2002. Inbeans and chickpea, however, NP was mostly sensitive to MS water stress
whereas NS was sensitive to both the MS and LS stresses.
155
In addition to data collected at harvest, number of flowers and pods were counted at the
end of the stress period in the MS and at maturity for the LS treatment. The count of
flowers and pods in the lower and upper half of the plant canopy indicated that the
majority of the reproductive organs were located in the upper half of the plants,
particularly in chickpea and cowpea. As shown in Fig. 7.2, the numbers of reproductive
organs were significantly reduced by the water stress treatments in beans and chickpea as
a result of flower abortion and/or dropping of flowers and pods. The reduction in the
number of flowers and pods, (6-21% in the MS and 2-48% in the LS treatment) was
minimal in cowpea in all the seasons. Beans was more sensitive to reduction of
reproductive organs as a result of stress during flowering (seasonal mean of 30%) than
stress during the pod filling period (19%) when compared to chickpea which is equally
sensitive to reduction of reproductive organs under both flowering (27%) and pod filling
(27%) period water deficits.
As the numbers of branches in many grain legume species determine the number of pods
that a plant can carry, a count of primary and secondary branches was made after each
stress period in 2002/20003 to investigate the effect of reproductive period water stress on
branch number. The data showed that cowpea had more primary branches than secondary
branches while chickpea and beans had higher number of secondary branches than
primary ones (Fig. 7.3). Un1ike secondary branches, primary branches were not affected
by either of the water stress treatments in all species suggesting that these branches were
mostly produced during the vegetative growth period. Dramatic reduction of secondary
branches was observed in chickpea due to the MS stress (42%) while the reduction in
beans was moderate (23%) and that of cowpea was minimum «10%). The reduction of
secondary branches (due to early drying and/or stunted growth) as a result of the LS
treatment was highest in beans (32%) followed by cowpea (28%) and chickpea (25%).
Generally, reduction in the number of flowers and secondary branches by the water stress
treatments was reflected in significantly lower NP m-2 in all species. Previous studies also
indicated reductions in NP due to reproductive stage water stress in cowpea (Wien, et al.,
1979; Turk et al., 1980a), beans (Acosta Gallegos and Shibata, 1989; Tesfaye, 1997),
peanut (Harris et a!., 1988; Wright et al., 1991), soybean (Wien et al., 1979) and mung
bean (Pannu and Singh, 1993). In addition to low number of pod bearing branches, the
negative effect of water stress on meiosis and pollen fertility could have contributed to the
reduction ofNP as observed in many cereals (Saini and Westgate, 2000).
156
SW of cowpea was not affected by any of the stress treatments in any the seasons. This
contradicts with the report of Wien et al. (1979) who found an increase in SW under
water deficit during the reproductive stage of cowpea. The contradiction could be either
due to cultivar or climatic differences between the studies. Chickpea normally had the
highest SW and NP m-2 under all water regimes in all seasons which are compensatory
for its low number of seeds per pod. As compared to the control, SW in beans and
chickpea was significantly reduced by the MS and LS treatments in the first two seasons.
Although beans and cowpea had similar SW under well-watered conditions, bean had
lower SW but significantly higher NP and NS m-2 than cowpea under the MS and LS
treatments in two of the three seasons (Table 7.3 to 7.4). The influence of water stress on
seed number and mass, generally, depends on its timing and intensity during the
reproductive growth. For example, in pea, seed abortion occurred when water stress
coincided with the initiation of linear seed filling while seeds that reached this stage
before water stress maintained normal growth (Ney et al., 1994). Therefore, in agreement
with the results of Ney et al. (1994), the present species respond to seed-filling water
stress either by reducing seed number (e.g. cowpea) or mobilizing reserves to maintain a
constant seed growth rate (e.g. beans). Generally, environmental stresses such as water
deficit can induce a compensation growth between yield components indicating the
developmental plasticity of plant yield systems under environmental stress (Adams et al.,
1967).
Biomass at harvest was significantly affected by the water stress treatments in 2001/2002
and 2002 seasons but not in 2002/2003 (Table 7.3 and 7.4). It was severely reduced by
the MS treatments in beans and chickpea in 2001/2002 and in beans in 2002. Biomass
production ranged from 3749 (chickpea) to 6798 (beans), 1349 (chickpea) to 3840
(cowpea) and 2312 (cowpea) to 4671 (beans) kg ha" in the C, MS and LS treatments,
respectively over the three seasons (Table 7.3 to 7.4). As shown in Chapters 3 and 4, the
lowest biomass in the MS treatment is a result of low LAl, RUE and WUE. This is in
agreement with other reports on chickpea (Sivakumar and Singh, 1987), beans (Acosta
Gallegos and Shibata, 1989), cowpea (Turk, and Hall, 1980 a, b) and mung bean (Pannu
and Singh, 1993).
HI was significantly higher in the C treatment in 2001/2002 while it was similar between
the C and LS treatments in 2002 and 2002/2003 (Table 7.3 and 7.4). Differences between
157
species in HI was observed only in the 2002/2003 in which beans had significantly higher
HI than cowpea, and cowpea had significantly higher HI than chickpea (Table 7.4). The
lowest HI was recorded in the MS treatment in all the seasons. HI varies on the ability of
a genotype to partition current assimilate to the seed and the reallocation of stored or
structural assimilates to the same sink (Turner et al., 2001). The lack of significant
difference between the species in the two seasons could be explained by the different
assimilate partitioning pattern of the species as shown in Chapter 3. For example, beans
and cowpea had similar seed number and seed mass under well-water supply conditions,
but bean had relatively higher number of seeds than cowpea, and cowpea had higher seed
mass than bean in the LS treatment. Both beans and cowpea had higher partitioning (P) in
the LS treatment (Table 7.5). This shows that partitioning in bean is used to maintain seed
number (fill all the available seeds) rather than mass whereas in cowpea partitioning
seems to be used to maintain seed mass rather than number (i.e. fill most of the available
seeds).
Chickpea had low p in the latter two seasons but also had similar HI to that of beans and
cowpea under the different water regimes. This could be due to higher partitioning of
current assimilate to the seed than reallocation of stored dry matter in chickpea. In other
studies, differences in chickpea grain yield were associated with differences in crop
growth rate (Cr) rather than variation in dry matter partitioning, p (Williams and Saxena,
1991). Therefore, the lack of significant differences in HI among the species could be a
result of the different strategies used by the plants to maintain their HI at high level. The
lowest HI in the MS treatment could partly be a result of lower partitioning to the pods
(Table 7.5) and resumption of vegetative growth after re-watering at a cost of pod growth
and also generally low crop growth rate owing to low LAl and RUE recorded in the same
treatment.
7.3.3. Partitioning
Dry matter partitioning among plant parts was discussed in detail in Chapter 3. Here, the
mean total growth rate (Cr) and pod growth rate (Cp) and mean dry matter partitioning to
the pod (P) and its importance as a yield component is presented. As shown in Table 7.5,
the control treatment had higher Cr than the stress treatments in most of the cases though
the LS treatment excelled the control in some cases (e.g. beans) because of its shorter
growth period. In beans, both Cr and Cp in the MS were lower than the LS treatment in all
158
Table 7.5. Mean crop growth rate (Cr), pod growth rate (Cp) and partitioning coefficient (P)
of three grain legumes grown under three water regimes in three seasons.
Parameter Water regime Beans Chickpea Cowpea
200112002
C 0.831 ± 0.271 1.02 ± 0.174 1.020 ± 0.126
Cr (g m-2 °Cd-1) MS 0.692 ± 0.060 0.693 ± 0.098 0.520 ± 0.049
LS 0.890 ± 0.109 0.575 ± 0.069 0.509 ± 0.061
C 0.395 ± 0.197 0.582 ± 0.142 0.559 ± 0.093
c, (g m-2 °Cd-1) MS 0.190 ± 0.042 0.275 ± 0.065 0.269 ± 0.013
LS 0.690 ± 0.307 0.395 ± 0.043 0.318 ± 0.144
C 0.475 0.571 0.548
P MS 0.275 0.397 0.517
LS 0.775 0.687 0.625
2002
C 1.05 ± 0.132 0.822 ± 0.051 0.923 ± 0.137
Cr (g m-2 °Cd-1) MS 0.450 ± 0.098 0.633 ± 0.068 0.778 ± 0.146
LS 1.260 ± 0.092 0.855 ± 0.051 0.814 ± 0.188
C 0.940± 0.165 0.484± 0.094 0.445 ± 0.007
Cp (g m-2 °Cd-1) MS 0.163± 0.029 0.423 ± 0.004 0.358±0.171
LS 1.07± 0.225 0.355 ± 0.031 0.324 ± 0.023
C 0.895 0.589 0.482
P MS 0.362 0.668 0.460LS 0.849 0.415 0.398
2002/2003
C 1.600 ± 0.185 0.990 ± 0.067 1.1430.157
Cr(g m-2 °Cd-1) MS 0.799 ± 0.091 0.502 ± 0.047 0.501± 0.063
LS 0.991 ± 0.106 0.683 ± 0.062 0.987± 0.082
C 0.739 ± 0.491 0.393 ± 0.143 0.509 ± 0.159
Cp (g m-2 °Cd-1) MS 0.004 ± 0.144 0.154 ± 0.112 0.055 ± 0.142
LS 0.718 ± 0.286 0.456 ± 0.136 0.681 ± 0.097
C 0.462 0.397 0.445
P MS 0.005 0.307 0.110LS 0.725 0.668 0.690
the seasons. When compared to the LS, both chickpea and cowpea also had lower Cr in
the MS only in 2002 and 2002/2003 seasons. C, in both chickpea and cowpea was higher
in the LS than in the MS in the milder temperature seasons but lower in the higher
temperature season. This difference in Cr is mainly attributed to differences in canopy
development and energy interception between environments and the crops (Greenburg et
al., 1992). Therefore, Cr gives an integrated measure of the source use capacity of a crop
and can be further evaluated through the effect of radiation interception and RUE (Turner
et al., 2001). Under water-limited environments, Cr is a result of the crop's ability to
capture and transpire water and its efficiency of water use (Passioura, 1977). Therefore,
differences in Cr indicate differences in resource utilization among species in a given
environment. The correlation of Cr with grain yield (data not shown) was variable across
seasons. Unlike Cr, however, p was strongly positively correlated with grain yield in the
control (0.76 to 0.98) and LS (0.98) treatments in 2002 and 2002/2003 while the
159
correlation in the MS treatment was weak and variable among seasons (-0.56 to 0.79).
Beans, followed by chickpea in 2001/2002 and 2002 and by cowpea in 2002/2003, had
the highest p in the LS treatment (Table 7.5).
Remobilisation of assimilates from shoot to seed is reported for beans (Acosta Gallegos
and Shibata, 1989) and cowpea (Hall and Patel, 1985). Except in chickpea in 2002, p was
generally low in the MS treatment in all species and seasons. The lowest partitioning in
the MS treatment could be a result of reduced reproductive organ establishment which
decreased pod set (smaller sink strength) resulting in reduced partitioning, sink limitation
(Greenburg et al., 1992). This partitioning difference between the water regimes indicates
differences in the ability of the crops to initiate enough sink to utilize the carbon
assimilate available under different environments (Greenburg et al., 1992). As shown in
peanut, drought at the pod filling stage maximizes partitioning because established fruit
generally has priority for the available assimilate in the event of stress (Greenburg et al.,
1992).
The present study shows differences in p among species in that beans had higher p in the
LS and C treatments than the MS in all seasons unlike chickpea and cowpea where the p
in the water regimes was varied across seasons. The contribution of Cr and p towards
stress adaptation can vary among plant species. For example, tolerance of p to high
temperature is considered more important to peanut adaptation than Cr under severe water
deficit (Greenburg et ai, 1992) while in chickpea Cr was the major source of yield
variation under water stress rather than p or length of reproductive period (Williams and
Saxena, 1991). In the present study, however, p was consistently correlated with grain
yield while the correlation of Cr with grain yield was variable across seasons. Therefore, p
seems to be a good indicator of yield performance under varying water and temperature
conditions. In general, attainment of high biomass followed by high partitioning to the
seed is the major requirement of a high grain yield in many grain legumes including
chickpea (Saxena et al. 1990; Singh, 1991; Leport et al, 1999), cowpea (Wien et al.,
1979), lentil (Silim et al. 1993a), pea (Silim et al., 1985), soybean (Westgate et al., 1989),
mung bean (Bushby and Lawn, 1992), peanut (Wright et al., 1991) and narrow leafed
lupin (French and Turner, 1991). The enzyme sucrose synthase is reported to be
responsible for controlling the rate of seed filling, seed size and finally sink activity
(Mohapatra et al., 2000).
160
7.3.4. Yield component framework
Grain yield is determined by both reproductive components (e.g. NP per plant, NS per
pod and SW) and components that integrate many plant functions at a higher level (e.g.
WUE, RUE, HI, length of phenological development periods). In this study, some of
these parameters were correlated with grain yield (Table 7.6) in order to find the major
yield determining components in grain legumes under different water supply conditions
during the reproductive periods.
Table 7.6. Correlation (pearson) of the grain yield of three-grain legume species with some
plant parameters under three water regimes for three seasons,"
Water regime Parameters* 200112002++ 2002 2002/2003
Well-watered Biomass -0.02 0.87 0.89
III 0.86 0.96 0.91
Number of seeds per m2 0.95 0.26 0.88
Days to flowering -0.57 0.63 0.24
Days to podding -0.80 0.20 0.09
Days to maturity 0.83 -0.78 0.97
Reproductive period 0.73 -0.98 -0.44
Pod filling period 0.99 -0.63 -0.46
WUEd 0.47 0.92 0.91
WUEg 0.98 0.99 0.51
RUE 0.90 0.99
Mid-season stress Biomass -0.99 0.95 0.92
III 0.24 0.99 0.95
Number of seeds per m2 0.55 0.11 0.99
Days to podding 0.79 -0.50 0.61
Days to maturity -0.14 -0.86 -0.13
Reproductive period -0.99 -0.86 -0.36
Pod filling period -0.99 -0.93 -0.34
WUEd 0.92 0.99 0.45
WUEg 0.99 0.97 0.13
RUE 0.69 0.98
Late season stress Biomass -0.57 0.98 0.93
HI 0.97 0.93 0.98
Number of seeds perm2 -0.42 0.59 0.99
Days to podding -0.76 -0.24 0.47
Days to maturity -0.87 -0.97 -0.67
Reproductive period 0.82 -0.80 -0.44
Pod filling period 0.87 -0.45 -0.52
WUEd 0.20 0.50 -0.92
WUEg 0.99 0.99 0.88
RUE 0.99 0.94
an = 9 (three species with three replications)
*WUEd = water use efficiency of total above ground dry matter, WUEg = water use efficiency of seed yield, RUE =
radiation use efficiency. ++ Correlation coefficients with values greater than 0.72 or less than -0.72 are significant at 5
%P level.
161
Under well-watered conditions, grain yield was strongly positively correlated with HI, NS
m", WUEd, WUEg and RUE (Table 7.6). Days to maturity was positively correlated with
grain yield in 2001/2002 and 2002/2003 seasons but both days to maturity and pod filling
period correlated negatively to yield in 2002. This negative correlation indicated the
importance of early maturity and short period of pod filling for high grain yield under
high temperature conditions. In other studies, high temperature during the reproductive
period in late sown chickpea led to reduced seed size, lower yield and lower WUE
(Sivakumar and Singh, 1987).
Under mid-season water stress, grain yield was positively correlated with HI, NS m",
WUEd, WUEg and RUE although the strength of correlation varied across seasons. Except
for the high temperature season, the period from sowing to podding was positively
correlated with seed yield. However, grain yield was strongly negatively correlated with
the period of pod filling in all the seasons suggesting the need for a short pod-filling
period under mid-season stress for high seed yield. Days to maturity was negatively
correlated with grain yield in 2002 indicating the importance of early maturity for high
grain yield when water stress at mid-season is combined with high temperature
conditions. Under late season water stress, grain yield was strongly positively correlated
with HI, WUEg and RUE in all the seasons (Table 7.6). Pod filling period was negatively
correlated with grain yield in 2002 and 2002/2003 but positively correlated in 2001/2002.
Length of maturity period was negatively correlated with grain yield in all the seasons
indicating the need to select early maturing cultivars for high grain yield in grain legumes
under terminal drought environments. The correlation of WUEd with grain yield under-
late season stress was variable across seasons, being positive in 2001/2002 and 2002 and
negative in 2002/2003.
Total above ground biomass at harvest was strongly positively correlated with grain yield
in 2002 and 2002/2003 in all the water regimes but correlated negatively in 2001/2002
suggesting the seasonality of its association with seed yield. This could be due to some
environmental factors that reduce initiation of reproductive primordia during the
transition period from vegetative to reproductive phase. However, similar to the
observations in 2002 and 2002/2003 seasons, several reports indicated strong positive
correlation of biomass with grain yield in many crops such as chickpea (Silim and
Saxena, 1993b), beans (Acosta Gallegos and Shibata, 1989) and a range of other legumes
162
(Thomson et al., 1997; Siddique et al., 1999). Similar to the present results, positive
correlation of grain yield with HI, seed number and mass, and early flowering has been
reported in chickpea under severe water deficit (Silim and Saxena, 1993b). Positive
correlations of grain yield with biomass at maturity (Thomson et al., 1997; Siddique, et
al., 1999) and harvest index (Thomson et al., 1997) were reported in a range of grain
legume species in Australia. In agreement with the present study, phenology was strongly
negatively correlated with grain yield under the dry year while it was weakly correlated
under the wet year in lentil (Silim et al., 1993a). Earliness is considered to be important in
cowpea, pea, and other grain legumes (Hall and Patel, 1985; Subbarao et al., 1995;
Sharma and Khan, 1997; Siddique et al., 1999). Strong positive correlation of grain yield
with HI and RUE is reported in mung bean (De Costa et al., 1999).
The analysis of yield components showed that high yield of grain legumes under different
water supply conditions was a result of high total biomass at harvest, high NS, high HI
and high RUE. In agreement with the present study, Chapman et al. (1993), Pannu and
Singh (1993) and De Costa et al. (1999) have observed a positive response of HI to
irrigation in mung bean that irrigation during flowering and pod filling periods increased
HI through greater pod initiation and higher pod growth. High biomass production is a
function of high radiation interception, RUE and WUE, all of which are influenced by
LAl which is a parameter very sensitive to water deficit in grain legumes (Acosta
Gallegos and Shibata, 1989; Chapman et al., 1993; De Costa et al., 1997). Favourable
water supply during flowering and pod filling stages is, therefore, required to maximize
RUE by maintaining high LAl, and thereby increase biomass and seed yield. RUE is a
measure of the efficiency of canopy photosynthesis (Norman and Arkebauer, 1991;
Loomis and Connor, 1992) and is highly sensitive to water deficits (Lawlor, 1995)
particularly during flowering and pod filling stages when LAl and transpiration are very
high (De Costa et al., 1999). Therefore, high LAl and high RUE and WUE are important
morphological and physiological component of seed yield, respectively in grain legumes.
Although, the genotypes used in the present correlation study are few in number, the
observed relationships between grain yield and yield determing parameters showed
interesting environment dependent (water regime in this case) relationships which insight
further investigation on grain legumes.
163
7.4. Conclusion
Water deficit during the reproductive period is a major factor for the low yield of grain
legumes. In this study, the most water stress sensitive stages of each species and yield
determining parameters in each water regime were identified.
Beans and chickpea are more sensitive to flowering than pod-filling water stress while the
yield response of cowpea is similar under the two periods of stress. Therefore,
minimizing the water stress at these most sensitive phenological stages through
management or breeding methods could help maximize the yield of these crops.
Moreover, such information can also be used to set irrigation priorities based on critical
growth stages and thereby increase water saving and yield in areas where the crops are
grown under irrigation conditions. In most of the cases, the yield of cowpea is less
sensitive to both MS and LS stresses than that of beans and chickpea. This information is
useful in the practice of crop choice based on environmental constraints such as water
stress.
Grain yield was greatly reduced when stress occurred during the flowering period which
is mainly associated with the adverse effect of the stress on growth of reproductive
organs, LAl, efficiency of radiation and water use, and partitioning of dry matter to the
seed. The reduction of yield under pod-filling water stress is mainly due to a shorter
reproductive period and to some extent reduction of the numbers of reproductive organs
such as number of seeds per pod and numbers of pods per plant. Therefore, management
or breeding activities which improve one or more of these characters are expected to
improve the yield of grain legumes under drought prone environments.
Unlike the mean crop growth rate, dry matter partitioning to the pod during the
reproductive periods was strongly correlated with grain yield. High dry matter
partitioning during late season water stress is found to be an important mechanism by
which species maintain high HI and grain yield under terminal drought environments.
Accordingly, crops such as beans, which have high dry matter partitioning capacity
during late season water stress, could be grown in areas where water is available during
the vegetative and early reproductive season while it is scarce towards the end of the
season.
164
Yield determining factors in grain legumes varied along water regimes and temperature
conditions. Yield, however, was strongly positively correlated with HI and RUE across
all water regimes in both seasons. High NS m-2 and high WUE and longer maturity time
are important for high grain yield in well-watered and mild temperature environments. On
the other hand, short reproductive period and high WUE are important in mid-season
drought environments with intermediate temperature. Under less extreme temperatures,
yield was strongly positively correlated with WUEg, p and early maturity in the LS
treatment indicating the importance of these characters in terminal drought environments
to improve grain yield. Moreover, as observed in the 2002 season, short pod filling
period, early maturity, high biomass, HI and WUE are important for obtaining high grain
yield under high temperature environments with variable water supply. The relationships
found between yield and yield component characters in each water regime can be used by
breeders as a guide to improve the yield of grain legumes in the different environments.
165
CHAPTER 8
Evaluation of the CROPGRO-Dry Bean and Chickpea Model in a Semi-Arid
Environment
8.1. Introduction
Dynamic crop growth simulation started in the early 1960s, with successful application to
well defined growth processes such as canopy photosynthesis (e.g. de Wit, 1965; Duncan
et al., 1967). Since then, the rapid development of computing power has led to models on
many aspects of crop growth and development. Since there exist an almost infinite
number of combinations of soil type, weather and agricultural practices, experimenting in
all desirable situations is impossible. Therefore, there is a need to use models to increase
the human capacity in understanding the different possible interaction of these factors
(Penning de Vries et al., 1988). A crop model can be used as a quantitative scheme for
predicting the growth, development and yield of a crop, given a set of genetic coefficients
and relevant environmental variables (Monteith, 1996).
Crop growth models are increasingly being used to support field research and extension
in many countries (Carberry et al., 2002; Jagtap et al., 2002). Applying models can lead
to a more effective way of using existing knowledge for extension, agronomic and
cropping systems research and breeding. It also leads to a more effective experimentation
and integration of the scientific disciplines involved in crop production (Penning de Vries
_et al., 1988). As outlined by Boote et al. (1996), models can assist in synthesis of research
undertaking about the interactions of genetics, physiology and environment, as well as
integration across disciplines, and organization of data. They can assist in pre-season and
in-season management decisions on cultural practices, fertilization, irrigation, cultivar
choice and pesticide use. Crop models _can also assist policy makers by predicting soil
erosion, leaching of agrochemicals, effect of climate change, and also by giving large area
yield forecasts (Boote et al., 1996). Moreover, variability in yields of sensitive crops or
cropping sequences due to variations in weather can be tested with long-term weather
data, which speeds up crop assessment. Analysis of yield variability across season has
been undertaken for certain crops such as faba bean (Grashoff et al., 1987 as cited in
Penning de Varies et al., 1989), rice (Morris, 1987) and wheat (Aggrawal and Kalra,
1994). This type of analysis helps establish the impact of year-to-year weather variability
on the crop much faster than with more conventional field experimentation methods (e.g.
166
Penning de Varies et al., 1989; Matthews et al., 2002; Jones et al., 2003). In general,
models are currently being applied to solve several agricultural problems. Matthews et al.
(2002) described a number of models and their applications in tropical agricultural
systems.
The DSSAT (Decision Support System for Agrotechnology Transfer) comprises a set of
annual crop simulation models and a data base management system, together with utilities
analysis program (Tsuji et al., 1994; Thornton et ai, 1994), and is being used in many
countries. For example, in Africa, the DSSAT models have been applied in the study of
crop management (e.g. Vos and Mallett, 1987; Mbabaliye and Wojtkowski, 1994;
Wafula, 1995), irrigation management (e.g. Kamel et al., 1995; MacRobert and Savage,
1998), fertilizer management (e.g. Singh et al., 1993; Thornton et al., 1995; Jagtap et al.,
1999), climate change (Muchena and Iglesias, 1995), climate variability (Phillips et al.,
1998), food security (Thornton et al., 1997) and so on. The model can be used for storing
information concerning field trials, extracting data from crop models for the purpose of
.validation or comparing management strategies and performing simple analysis of the
results of simulation runs (Thornton et ai, 1994; Tsuji et al., 1994).
Therefore, it seems that simulation of crop growth can help strengthen regional
development and agricultural planning in many countries. An obvious question, however,
is whether simulation modelling has a real role to play in developing countries. To clarify
such queries, it is necessary to test and evaluate the performance of already developed
models with experimental data collected in developing countries. Along this line, Hensley
et al. (1997) and Botes (1994) made comparisons between the Putu model (de Jager et aI.,
1981) and other models. In comparing DSSAT3 and Putu maize and wheat, Hensley et al.
(1997) reported that these models mostly gave reliable yield predictions although they
were sometimes unreliable, and further pointed out weaknesses observed in both models.
Therefore, crop models proposed for broader crop management applications should be
tested widely and against diverse field experiments to assess their ability of answering
practical questions (Boote et al., 1996).
In general, in sub-Saharan Africa where unpredictable fluctuations of weather and climate
has made many crop field trial experiments a risky exercise, the involvement of crop
models in the decision support system of a given production system becomes more
imperative than before. However, candidate crop models should be validated under the
167
specific environments before they are adapted to solve practical problems in such
countries (Hoogenboom et al. 1994). Therefore, the main objective of this study was to
validate the CROPGRO grain legume model of DSSAT in a semi-arid environment using
experimental data collected for beans and chickpea during three seasons in Ethiopia.
8.2. Input Data and Methodology
8.2.1. CROPGRO model
Crop growth models, which share a common input and output data format, have been
developed and embedded in a software package called DSSAT (Tsuji et al., 1994; Jones
et al., 1994; Hoogenboom et al., 1994). The models under DSSAT umbrella include
CROPGRO which is a mechanistic, process-oriented model for grain legumes with
weather, crop development, carbon balance, crop and soil N balance and soil water
balance subroutines (Boote et al., 1998a, 1998b; Hoogenboom et al., 1994b). Crop
development includes processes like vegetative and reproductive development, duration
of root and leaf growth, onset and duration of reproductive organs, and dry matter
partitioning to plant organs over time. The crop carbon balance includes daily simulation
of photosynthesis, conversion and incorporation of carbon into crop tissues, carbon losses
to abscised parts, and growth and maintenance respiration. The carbon balance also
includes simulation of leaf area expansion, growth of vegetative tissues, pod and seed
addition, shell and seed growth, nodule growth, senescence and carbohydrate
mobilization (Boote et al., 2002). The crop N balance includes daily soil N uptake, N2
fixation, mobilization of N from vegetative tissues to reproductive organs, rate of N use
for new tissue growth and rate of N loss in abscised parts. Soil water balance processes
include infiltration of rainfall and irrigation, runoff, soil surface evaporation, root water
uptake, drainage of water below the root zone, and crop transpiration (Hoogenboom et al,
1994b; Boote et al., 1998a; 1998b; 2002). Model state variables are simulated and output
on a daily basis for crop, soil water, and soil N balance processes.
The generic CROPGRO model (v3.5) uses a common FORTRAN code to simulate the
growth of grain legumes such as dry bean, soybean, peanut and chickpea using input files
that define species traits and cultivar attributes (Boote et al., 1998a; 1998b). DSSAT also
provides a seasonal analysis system (including crop rotation and different management
options) to simulate possible long-term adaptation measures so as to analyse those
management scenarios that can decrease potential agricultural productivity under
168
expected climate change conditions (Thornton and Hoogenboom, 1994). In the present
study, CROPGRO was used to simulate the growth and yield of common bean (Phaseolus
vulgaris L.) and chickpea (Cicer areitinium L.).
8.2.2. Input data
The DSSAT crop models are designed to use a minimum set of soil, weather, crop and
management information. The models integrate at daily time steps, and hence require
daily weather data, consisting of maximum and minimum temperature, solar radiation,
and rainfall as input. The models also require a soil profile description.
8.2.2.1. Soil data
Soil profile description was made from a 2 m deep pit on the experimental site following
the USDA soil classification system' (Appendix 6A & 6B). Soil texture, bulk density,
colour, pH, organic matter and organic carbon, total nitrogen, and cation exchange
capacity (CEC) were analysed at the Soil Laboratory of Alemaya University following
established methods while soil water at field capacity and permanent wilting point were
determined at the National Soil Laboratory of Ethiopia (Appendix 6C & 6D). The soil
parameters used in the model are listed in Table 8.1.
Table 8.1. Soil parameters for the experimental site at Dire Dawa, Ethiopia.
Soil Depth DLL a DULa esa pH RGFa BD OC Total Clay Silt
horizon (cm) 3 (cm3(cm%m /cm3) (cm3/cm3) ) (H2O) (g/rrr') (%) N (%) (%)
(%)
Ap 0-10 0.181 0.305 0.377 8.5 1.00 1.24 1.18 0.10 33.0 36.0
A 11-40 0.211 0.336 0.384 8.6 0.75 1.23 1.36 0.12 40.0 37.0
Ba 41-70 0.205 0.335 0.409 8.6 0.20 1.27 1.20 0.10 38.0 44.0
Bt 71-90 0.218 0.348 0.406 8.4 0.10 1.41 1.14 0.10 41.0 43.0
BC 91-180 0.196 0.317 0.364 8.5 0.00 1.36 1.00 0.09 36.0 27.0
DUL = drained upper limit, DLL = Drained lower limit, es = saturation (upper limit), RGF = root growth factor, BD =
bulk density, and OC = organic carbon.
• calculated by DSSA T from other input soil parameters
8.2.2.2. Weather data
Three field experiments with three water regime treatments were conducted in 2001/2002,
2002, and 2002/2003 season at Dire Dawa, Ethiopia (see previous chapters for detail
description of site and experimental design). Daily maximum and minimum temperatures,
solar radiation and rainfall were collected from a class A weather station at Dire Dawa
airport (latitude = 9°4'N, longitude = 41°5'E, altitude = 126Om), which is 500 m away
from the experimental site, in 2001/2002 and 2002 seasons and from an automatic
169
weather station placed at the experimental site (latitude = 9°6'N, longitude = 41°8'E,
altitude = 1197m) in 2002/2003 starting from 1 October 2002(Appendix 7A to 7C).
8.2.2.3. Crop genetic coefficients
The bean cultivar planted was Roba-1 and the chickpea cultivar was ICC-4958 (see
Chapter 3 for description). In 2001/2002, both beans and chickpea were planted on a
lOm x 10m plot each adjacent to the main experimental area for the determination of
minimum crop data sets (phenology, growth and yield) for calibration. Previous data
collected for beans at the same site (Tesfaye, 1997) was also used for the calibration. The
genetic coefficients of each variety were estimated by repeated iterations using the
"Gencalc" program of DSSAT as described by Hunt and Pararajasingham (1994) until a
close match between simulated and observed phenology, growth and yield was obtained.
The calculated genetic coefficients are shown in Table 8.2.
Table 8.2. Genetic coefficients of cultivars 'Roba-l' and 'ICC4958' obtained from
"Gencalc" of DSSAT using 2001/2002 season and previous experiment data from Dire
Dawa.
Description Genetic coefficient
Roba-l ICC4958
Developmental aspects
Critical short day length (h) 12.2 11.0
Slope of relative response of development to photoperiod (h) 0.0 -0.143
Time between plant emergence and flower appearance in photothermal days (PTD) 36.0 30.0
Time between first flower and first pod (PTD) 2.0 8.0
Time between fust flower and fust seed (pTD) 9.0 15.0
Time between first seed and physiological maturity (PTD) 29.0 35.0
Time between first flower and end of leaf expansion (PTD 18.0 42.0
Seed filling duration for pod cohort at standard growth conditions (PTD 14.0 29.0
Time required for cultivar to reach final pod load under optimal conditions (PTD) 8.0 18.0
Growth aspects
Maximum leaf photosynthesis rate at 30°C and high radiation (mg CO2 m"~s") 1.0 1.7
Specific leaf area of cultivar under standard growth conditions (cm" g' ) 220.0 150.0
Maximum area of full leaf (three leaflets, cm") 133.0 10.0
Maximum fraction of daily growth that is partitioned to seed + shell 1.0 1.0
Maximum weight per seed (g) 0.247 0.403
Average seed per pod under standard growing conditions 6.20 1.30
8.2.3. Model evaluation (validation)
The cultivars were grown under three water regimes namely, mid-season (stage RI)
stress, late season (stage R4) stress and well-watered condition (see previous chapters for
details) for three seasons. Data collected on crop evapotranspiration, phenology, growth,
yield and yield components from all three seasons were used for evaluation of the models.
170
The data included in the evaluation part of the model for 2001/2002 was from a different
data set used for the calibration.
Model performance was evaluated based on five statistical indexes: the index of
agreement (da), the mean deviation (MD), the root mean square error (RMSE), the
coefficient of variation (CV) and the modelling efficiency (ME) following Willmot
(1982), Hoogenboom et al. (1999), Gabrielle et al. (1995), and Loague and Green (1991).
d ;-1a =1- (8.1)
n
MD=n-I~:<p; -0;) (8.2)
;=1
RMSE =[n-It(~ _0;)' r (8.3)
CV = RMSE *100 (8.4)
o
(8.5)
;=1
where 0 is observed value, P is model-predicted value, 0 is observed mean value, i is
observation and n is number of observations.
The values of MD indicate whether there is a systematic bias in the simulated values. The
RMSE reflects the magnitude of the mean difference between predicted and measured
values. The CV is a relative measure of the amount of unexplained variation and is
independent of the unit of measurement used (Wang et al., 2003). The simulation is
considered excellent if the CV is < 10%, good if it is within 10-20%, fair ifit is between
20-30% and poor if it is >30% (Jamieson et al., 1991). The maximum value of ME is 1
(optimal value) and it compares modelling variability with experimental variability. A
negative value of ME refers to the fact that the modelling variability is greater than the
experimental variability, and hence the simulation is not satisfactory (Rinaldi et al.,
2003). Higher values of da indicate better model performance whereas lower values
171
indicate poor performance. Evaluation of model performance was also performed by
plotting "simulated vs. measured" values and calculating the parameters of the regression
line (slope, intercept, and R2j and comparing them with al: 1 line.
8.3. Results and Discussion
8.3.1. Model validation
8.3.1.1. CROPGRO-DRY BEAN
The seasonal course of measured and simulated leaf area index (LAl), above ground
matter production (ADM), and cumulative evapotranspiration (ET) of beans for two
seasons are shown in Fig. 8.1 to 8.3. As indicated by the closer match between the
regression and the 1:1 lines and a high da value (0.97), the model predicted LAl well in
2002 but over estimation was observed after flowering in 200212003.
7 7
2002 .o-c
6 • oP-c
2002
.O-MS 6 1 :1oc
ei oP-MS
5 • oMSi .O-LS 56 PolS 3 6 LS
3 4 ".tiG. 4I
I 0 nl 03 • "30 3E Y = 0.90x + 0.48
• ii)2 • 2
R2 = 0.91
0
6 ~ • da = 0.97I
t 0 0
0 - • 0
0 20 40 60 80 100 120 0 2 3 4 5 6 7
6 7
200212003 200212003
5 6 1 :1
0 0 5 .-
4 ê1&
•lil I 0 3• "tiGI 4ca: 3 •• • 0 i...J ei :s
2 • •
3
• • Q .ëCl)0 y = 1.32x - 0.03
• • 2• R2 = 0.791 • • da = 0.87
0 III ~ 0
0 20 40 60 80 100 120 0 2 3 4 5 6 7
Time (days after planting) Measured LAl
Figure 8.1. Seasonal course of measured (0) and simulated (P) LAl of beans under water
stress (MS, LS) and well-watered (C) conditions (left), and regression of simulated vs.
measured values (right) for 2002 and 2002/2003 seasons.
172
12000 -,--- --, 12000 -,-- --,
2002 a _ • - 2002 1 :110000 - .O-C ~ DC
- o P-C
.A. 'ns
..c: 9000- oMS
.O-MS Cl
~ns 8000- o P-MS ::. ALS
..I<: bo O-LS ::!:
Cl
::. 6000 A P-LS o ~ 6000::!: 'C o
~ I 0 .! ... 0ns •• ' AD
4000 - 8 6 • 0o '3 o .iI 0
o 0 ~ 3000- o 8?O-'y'= 0.46x + 421.74
2000 -
R2 = 0.92
o ~ • I
o 20 40 60 80 100 120 o 3000 6000
12000 -,------- --, 12000 -,----- ----,
2002/2003 ~- 2002/2003 1 :1
10000 - .'.ncs:
Cl 9000
r'ns 8000-
• ::.
• • ::!:..c: o
:C:.l 6000- A A ~ 0lil .A. ~ 6000 & c:·b.! ~.' 0::!: ns
~ 4000 ~.i•0 •0 • i '3 y = 0.78x + 1212.56~ 30002000 - ~ R2 = 0.89
D • da = 0.95
0- ... • • o-.~,---r---r---r--~
o 20 40 60 80 100 120 o 3000 6000 9000 12000
Time (days after planting) Measured ADM (kg ha-1)
Figure 8.2. Seasonal course of measured (0) and simulated (P) above ground dry matter
production (ADM) of beans under water stress (MS, LS) and well-watered (C) conditions
(left), and regression of simulated vs. measured values (right) for 2002 and 2002/2003
seasons.
On the other hand, ADM was underestimated in 2002 whereas the estimation is closer to
the 1:1 line in 200212003 with a high index of agreement (0.95). Although the same da
value (0.90) was obtained in both seasons, cumulative ET after flowering was
overestimated in 2002 while it was underestimated in 200212003. The results indicate that
the model overestimates LAl and cumulative ET but underestimates ADM under higher
temperature conditions (2002) compared to the more milder temperature season
(2002/2003 ).
The main statistical indexes used to evaluate the accuracy of the models are reported in
Table 8.3. Regression coefficients, R2 and probability values for the test of the regression
are also presented in Table 8.4 for phenology, yield and yield component variables. The
173
CROPGRO-DRYBEAN model was able to simulate time to anthesis (flowering) but not
the time to maturity. Despite a lower CV (14%), the value of da was 0.46 and the ME was
negative (-0.36) for maturity date indicating more variability in the simulation than in the
actual measurements. As shown by small MD and RMSE values and a high da value
(0.86) and a positive ME, the model simulated maximum LAl with reasonable accuracy.
Simulation of above ground dry matter at harvest was also fair with a RMSE of 21.4%, a
da value of 0.92 and ME of 0.73, which is closer to the optimal value (1.0). Regression of
the ADM data indicated a slope closer to the 1:1 line with R2 value of 0.77, which is
significant at 1% probability level (Fig. 8.4, Table 8.5). The model also simulated mass
per seed and HI fairly well with a CV of < 30% (Table 8.3).
600~-----------------------. 600-,--------------,
2002 2002
500- IIO-C 500- DC
o P-C -"
1:1
Ê o
oMS
g 400- .O-MS o 0 Ê 6 LS gIJ" 0o P-MS o • g 400- Cl,'
Iii .O-LS ~ ~ Iii r;fX
~ 300- 60-LS ~ .. " 300- .!Sl,';: .s 0'
nl
:; !!li 0 I • :n;l !il..
E 200- I E 200 y= 1.17x+36.13
::l
CJ !!li • en .fJIi" R2 = 0.98
100 !!li • 100 ·m
!!li • ~ da = 0.90
!!li I
o +---,---,.----,------,--,---l o-~,--.---.---,---,--~
o 20 40 60 80 100 120 o 100 200 300 400 500
600,----------------------. 600-,------------------------,
2002/2003 2002/2003
500 • 500- 1:1
Êg Ê400- I • g 400-Iii I
~ 300- I Ië I• IiiI I•iI1 69B ".s 300;:
:n;l
n
• !!li :;
l
E 200 !!li E 200::l 0Z1
CJ I en Y= 0.71x+ 30.89
100 • 100 R2 = 0.90da = 0.90
O-l---.,----.,----.,----.,----.-----l o-.~,--.---,--,---,--~
o 20 40 60 80 100 120 o 100 200 300 400 500
Time (days after planting) Measured ET (mm)
Figure 8.3. Seasonal course of measured (0) and simulated (P) cumulative crop
evapotranspiration (ET) of beans under water stress (MS, LS) and well-watered (C)
conditions (left), and regression of simulated vs. measured values (right) for 2002 and
2002/2003 seasons.
174
Table 8.3. Statistical indexes of measured and simulated parameters of beans for data
combined over three water regimes and three seasons (n = 9).
Variable Observed Simulated Statistical indices
Mean S.D* Mean S.D MD RMSE CV ME da
(%)
Anthesis date (DAP) 48.4 7.50 44.3 3.60 -4.10 5.50 ll.5 0.39 0.76
Maturity date (DAP) 86.4 10.8 84.0 5.50 -2.90 11.8 13.7 -0.36 0.41
Max. LAl (m2m-2) 4.20 1.37 4.80 1.52 0.50 0.99 23.2 0.41 0.86
Grain yield (kg ha") 1699.0 1014.2 1631.1 953.00 -67.9 212.1 12.5 0.95 0.99
ADM (kg ha-Iy+-+ 4113.9 1798.6 3819.2 1700.0 -294.7 878.8 21.4 0.73 0.92
HI 0.42 0.08 0.40 0.12 -0.02 0.09 20.8 0.09 0.85
Weight per seed (g) 0.19 0.04 0.17 0.05 -0.04 0.05 26.2 -0.76 0.69
No. of seeds per m2 2254.0 1295.7 948.3 411.3 -1305.8 1613.3 71.6 -0.74 0.55
No. of seeds per pod 6.31 0.72 5.99 0.43 -0.32 0.75 12.0 -0.24 0.44
* S.D = standard deviation, MD = mean deviation, RMSE = root mean square error, CV = coefficient of
variation for RMSE, and ME = modelling efficiency. ++above ground dry matter at harvest.
7~------------------------~ 8000 .---------------,
Maximum LAl ADM 1:1.
6 /1:1 ...-.
oC ! 6000
5 oMS
3 to LS l
"C 4 s
~ ct 4000"C
"5 3
E ~
iii 2 y = 0.91x + 0.92 "5E
R2 = 0.67 iii 2000 y = 0.83x + 415.15
1 R2 = 0.77
O~-----'------r------r----~
o 2 4 6 8 o 2000 4000 6000 8000
Measured LAl Merasured ADM (kg ha-1)
4000..,--------------, 0.8 ..,..-------------------------,
Grain yield HI 1 :1
1 :1
.-.
~ 3000
! 0.6
Cl) 3:
~ "C
~ 2000 ~
"5 0.4
~
"5 .Ë(/)
E
iii 1000 y = 0.92x + 69.51 0.2
R2 = 0.96 y = 1.23x - 0.12
R2 = 0.67
o 1000 2000 3000 4000 0.0 0.2 0.4 0.6 0.8
Measured Y (kg ha") Measured HI
Figure 8.4. Comparison of simulated and measured maximum leaf area index (LAl), above
ground biomass at harvest (ADM), grain yield (Y) and harvest index (HI) of beans for three
water regimes over three seasons. The broken lines refer the regression line (equations
indicated) while the solid lines refer the 1:1 line between the simulated (with CROPGRO-
DRY BEAN model) and measured values with n = 9.
175
Grain yield was simulated well with a CV between 10-20% and values of da and ME very
close to the optimal value (Table 8.3). A regression equation, which is significant at 0.1%
P level with slope of 0.92, confirms that the model performed well in simulating bean
grain yield (Fig. 8.4, Table 8.4). However, the model underestimated number of seeds per
m2 and number of seeds per pod with a very poor ME (Table 8.3). Therefore, further
calibration and testing of the model may be required to increase its accuracy in simulating
seed number.
Table 8.4. Regression coefficient for beans and chickpea from simulated and observed data
combined over three water regimes and three seasons (n = 9).
Variable a b R Probability
Beans
Anthesis date (DAP) 21.2±0.69 0.48±0.01 0.99 0.000
Maturity date (DAP) 84.0±16.96 -0.01±0.20 0.00 0.982
Maximum LAl (m2/m2) 0.92±1.06 0.91±0.24 0.67 0.007
Grain yield (kg/ha) 69.5±142.9 0.92±0.07 0.96 0.000
Biomass at harvest (kg/ha) 415.1±768.10 0.83±0.17 0.77 0.002
ill -0.12±0.14 1.23±0.33 0.67 0.007
Weight per seed (g) -0.014±0.06 0.74±0.33 0.77 0.002
Number of seeds per m2 385.8±189.9 0.25±0.07 0.62 0.012
Number of seeds per pod 4.9±1.36 0.17±0.21 0.08 0.456
Chickpea
Anthesis date (DAP) 31.5±0.55 0.13±0.01 0.94· 0.000
Maturity date (DAP) 72.8±6.73 0.lo±O.08 0.20 0.226
Maximum LAl (m2/m2) 0.36±0.66 0.78±0.24 0.60 0.015
Grain yield (kg/ha) -101.6±230.0 1.11±0.17 0.87 0.000
Biomass at harvest (kg/ha) 704.6±837.3 0.74±0.22 0.61 0.013
HI 0.01±0.13 1.02±0.35 0.56 0.021
Weight per seed (g) -0.29±0.19 1.67±0.66 0.48 0.039
Number of seeds per m2 1674.7±495.8 0.06±0.38 0.04 0.876
Number of seeds per pod 0.95±0.18 0.22±0.14 0.26 0.158
8.3.1.2. CROPGRO-CHICKPEA
The simulated values of LAl during the crop cycle were generally close to the measured
values in the 2002 and 2002/2003 seasons with underestimation towards the end of the
growing period in 2002 (Fig. 8.5). Underestimation of seasonal LAl by the model was
higher under the well-watered conditions than the stressed ones in both seasons (Fig. 8.5).
However, as shown by high da values greater than 0.88 for the combined data over the
water regimes, the model simulated seasonal growth of LAl fairly well under both well-
watered and water stress conditions during the RI and R4 growth stages. Although the
model tended to underestimate seasonal ADM in the 2002 and 2002/2003 seasons,
particularly under well-watered conditions (Fig. 8.6), the index of agreement was high
(>0.85) indicating that the simulation of above ground dry matter production during the
176
(>0.85) indicating that the simulation of above ground dry matter production during the
crop growth cycle was acceptable. The model simulated cumulative crop ET with a better
accuracy (da >0.95) in both seasons (Fig. 8.7) indicating that the simulation of cumulative
ET was very good.
6-.------------------. 6-r--------------.,
2002 2002 1 :1
5 .o-c 5
oP-c oC
.O-MS oMS
4 o P-MS :5 4 c.LS
3 A Q-LS
Ac.. • .-l>
c. P-LS • C. "'0 ..3 ii 8 8 .I!V! 3 o .. tri 0
~ o • 3 .. -b 0
A
2 • E• Ui 2
y = 0.69x + 0.42
o R2 = 0.74
1 da = 0.90
o IlO III
o 20 40 60 80 100 120 o 2 3 4 5 6 7
6-.--------------~ 7-.--------------~
2002/2003 2002/2003
5 6 1:1
4 5
3 ~ 43 !
"3 3
.5 y = 0.77x + 0.362
ti) 2 R2 = 0.63
da = 0.89
o ft. I 1O~--.---,_-,_-r_-r_-r_~
o 20 40 60 80 100 120 o 2 3 4 5 6 7
Time (days after planting) Measured LAl
Figure 8.5. Seasonal course of measured (0) and simulated (P) LAl of chickpea under water
stress (MS, LS) and well-watered (C) conditions (left), and regression of simulated vs.
measured values (right) for 2002 and 2002/2003 seasons.
Unlike in beans, the model for chickpea simulated maturity date well with a CV of <20%
and a positive ME value (0.06) whereas the simulation of anthesis date was very poor
(ME = -2.87). The da value, on the other hand, is the same (0.47) for both maturity and
anthesis dates suggesting that ME may be a more sensitive indicator of model
performance than da.
177
12000 ,.----------------, 12000 ,.-- -,
2002 2002 1 :1
10000 .O-C • ~...... DC
[J P-C • 'n• .cs 9000 oMSs: 8000 .O-MS Cl I:;, LS
'ns oP-MS ot. ~
.c
Cl Á O-LS
:E
~ 6000 I:;, polS • [J o .'
•.... • ~ 6000[J:E 'ti.! 4·ci [J• 0 0 ns ol:;, ••• '~ 4000 I:;,8 '3 0,'-0~ 3000
I ~ .. rf" Y = 0.65x + 132.07
2000 • R2 = 0.95II
o ..... O~--._-_.-da-=_0..89--~
o 20 40 60 80 100 120 o 3000 6000 9000 12000
12000 ,.--------------, 12000 -,-------------,
2002/2003 2002/2003 1 :1
10000 ...... Y = 0.56x + 605.69
~ 9000
2
...... 8000 • • R = 0.78~
~ :E da = 0.86
Cl
~ 6000 . • .... ~ 6000• ! 'ti:E I:;, .!~ 4000 ns• • [J [J '3 .' . álo• .5 30002000 8 8 ~ fn~
O +-....~. ~-.--,--,,--.-~
o 20 40 60 80 100 120 o 3000 6000 9000 12000
Time (days after planting) Measured ADM (kg ha")
Figure 8.6. Seasonal course of measured (0) and simulated (P) above ground dry matter
production (ADM) of chickpea under water stress (MS, LS) and well-watered (C) conditions
(left), and regression of simulated vs. measured values (right) for 2002 and 2002/2003
seasons.
The model simulated maximum LAl and above ground dry matter at harvest fairly well
with a CV of <30% and ME of value 0.49 and 0.37, respectively (Table 8.5). The
regression equation between simulated and measured values of maximum LAl and final
ADM were significant (P< 0.05) with R2 value greater than 0.60 (Table 8.4, Fig. 8.8).
The model was successful in simulating grain yield at harvest (CV <20%, ME = 0.79, da
= 0.96) and HI (CV <25% and ME = 0.55, da = 0.84) (Table 8.5). The slopes of the
regression lines were very close to the 1:lline with an R2 value ofO.87 and 0.56 for grain
yield and HI, respectively (Fig. 8.8) indicating the best fit between observed and
simulated values. The regression equations were highly significant (P<O.Ol) for grain
yield and significant (P<0.05) for HI (Table 8.4).
178
500,---------------, 500~-----------~
2002 2002
400 .O-C 400 aC 1 :1
aP-C o • Ê oMS
Ê • O-MS .§. 6LS
.§. 300 o P-MS ~ 181 ti 300
ti .6. O-LS
ê i &
o • "C6 P-LS .!
200 a I ~ 200 y = 1.09+x17.33
E R2 = 0.97
100 I • iii 100 da = 0.96
I I
O+--II-r--.---.-~~-~
o 20 40 60 80 100 o 100 200 300 400 500
500,-------------~ 500~-----------~
200212003 200212003
• 400 1:1
Ê
•• 0
6 II !. 300
i 0 tio "C
200 • .!III 200"SE y = 0.78+x32.55
100 I iiiI 100 R2 = 0.98
da = 0.97
O+-----r-----.----,....----,--~
o 20 40 60 80 100 o 100 200 300 400 500
Time (days after planting) Measured ET (mm)
Figure 8.7. Seasonal measured (0) and simulated (P) cumulative crop evapotranspiration
(ET) of chickpea under water stress (MS, LS) and well-watered (C) conditions (left), and
regression of simulated vs. measured values (right) in 2002 and 2002/2003 seasons.
Table 8.5. Statistical indexes of measured and simulated parameters of chickpea for data
combined over three water regimes and three seasons (n = 9).
Variable Observed Simulated Statistical indices
Mean S.D* Mean S.D MD RMSE CV ME da
(%)
Anthesis date (DAP) 43.00 3.61 37.00 0.50 -6.00 6.70 15.4 -2.87 0.47
Maturity date (DAP) 86.00 14.12 82.00 3.24 -4.10 12.90 15.1 0.06 0.47
Max. LAl (m2m-2) 2.60 0.98 2.37 0.98 -0.22 0.66 25.5 0.49 0.86
Grain yield (kg ha") 1284.20 570.63 1323.90 680.72 39.70 245.70 19.1 0.79 0.96
ADM (kg ha-It+ 3540.10 1323.7 3305.4 1247.5 -234.67 841.00 23.8 0.37 0.87
HI 0.37 0.09 0.38 0.12 0.01 0.08 21.5 0.55 0.84
Weight per seed (g~ 0.28 0.04 0.18 0.10 -0.10 0.12 43.1 -9.30 0.45
No. of seeds perm 1076.30 809.7 1740.30 806.3 664.0 1237.1 115.0 -1.63 0.32
No. of seeds per pod 1.30 0.31 1.23 0.13 -0.04 0.25 19.9 0.23 0.54
* S.D = standard deviation, MD = mean deviation, RMSE = root mean square error, CV = coefficient of
variation for RMSE, and ME =modelling efficiency. ++ above ground dry matter at harvest.
179
5 6000
Maximum LAl -. ADM
0 1:1
4 oC 61:1 -'nl0 .c
:3 oMS
4500
Cl 0
"-0 3 6LS
::!5.
QI 0 ::E
0
ca ~ 3000
'3 "0
E QI
iii 2 -ca'3
y = 0.78x + 0.36 E 1500 Y = 0.73x + 704.58
= iii 2R2 0.60 R = 0.61
0
0
0 1500 3000 4500 60000 1 2 3 4 5
Measured LAl Measured ADM (kg ha")
2500 0.8
Grain yield (Y) HI
-. 1:1
'm 2000.c
Cl 0
0.6
~ :E
>- 1500 "0
"0 S
S ca'3 0.4.!
::::I 1000
0
E
E iii 0
iii
500 y = 1.11x -101.56 0.2
R2 = 0.87 Y = 1.04x - 0.00
R2 = 0.56
0 0.0
0 500 1000 1500 2000 2500 0.0 0.2 0.4 0.6 0.8
Measured Y (kg ha") Measured HI
Figure 8.8. Comparison of simulated and measured maximum leaf area index (LAl), above
ground biomass at harvest (ADM), grain yield and harvest index (BI) of chickpea for three
water, regimes over- three seasons. The broken lines refer to the regression line (equations
indicated) while the solid lines refer to the l:lline between the simulated (with CROPGRO-
CmCKPEA model) and measured values with n = 9.
The model failed to simulate weight per seed and number of seeds per m2 as shown by the
high RMSE and CV and low da and negative values of ME (Table 8.5). Moreover, the
regression equations had an R2 value of less than 0.50 and were not significant (P>0.05,
Table 8.4). Therefore, both the bean and chickpea models showed weakness in simulating
yield components.
8.3.2. General model performance
CROPGRO was generally successful in simulating grain yield at harvest, maximum leaf
area index, aboveground biomass at harvest and harvest index under both well-watered
and water deficit conditions during the reproductive period. The success of CROPGRO in
180
simulating grain yield has also been reported for many grain legumes including beans
(Hoogenboom, et al., 1994b; White et al., 1995), soybean (Ruiz-Nogueira et al., 2001;
Wang et al., 2003; Mall et al., 2004), and peanut (Singh et al., 1994) under different
environmental conditions including water deficit.
Calculation of the crop ET using the "Ritchie's" model (as provided in DSSAT)
overestimated the seasonal ET during the high temperature season while a slight
underestimation was observed in the milder temperature season for both beans and
chickpea (Fig. 8.3&8.7). However, the general trend, as shown by high da and ME values,
is acceptable. Despite a higher LAl and crop ET, CROPGRO underestimated
accumulation of above ground dry matter during the growth period of both beans and
chickpea in the high temperature season which could be as a result of the lower
optimum/maximum temperatures set in the models for optimum/maximum
photosynthesis. For example, the optimim (Optl , Opt2) and maximum temperatures set
for pod addition in CROPGRO for beans are 13, 25 and 36°C (Boote et al., 2002). The
maximum value set in the model is, thus, by far lower than the maximum air temperature
(>39 °C) observed during the experimental period in 2002 such that the model may not
have accounted for the photosynthesis above 36°C during pod initiation. Therefore,
further improvement in the parameterisation of cardinal temperatures in the models is
desirable .:The model also clearly underestimated maturity date, number of seeds per m2
and number of seeds per pod in beans as well as anthesis date and weight per seed in
chickpea. Number of seeds per m2, however, was overestimated in chickpea. Therefore,
further improvement as well as calibration with log-term crop data is necessary to correct
the simulation of these components by the model.
8.4. Conclusion
Grain legumes are grown predominantly under rainfed conditions where water is a major
limiting factor and the interannual rainfall variability is high. Crop modelling has proven
a valuable tool to evaluate the long-term consequences of weather patterns and
environmental conditions. However, the candidate crop models must be tested and
calibrated for new regions prior to their use as decision support tools in agriculture.
In the present study, the CROPGRO model simulated LAl, crop ET and aboveground
biomass at harvest with reasonable accuracy while the simulation for grain yield at
181
harvest was very good for both beans and chickpea. The models showed good
performance in simulating these variables under well-watered and water stress conditions
during the reproductive period of the crops. However, the simulation of yield components
at harvest in both seasons, and biomass accumulation during crop cycle under high
temperature conditions was poor in both crops suggesting further improvements of the
models to suit for the study environment. Model weakness was also observed on
simulation of maturity date in beans and flowering date in chickpea.
Generally, the performance of CROPGRO in simulating the final yield of beans and
chickpea is very good. Therefore, it is concluded that with further calibration using multi-
season crop data, the model has good potential to be used as a decision support tool in the
semi-arid areas of Ethiopia where long-term field experimentation is costly and less
effective due to fluctuating weather conditions.
182
CHAPTER9
Summary and Recommendations
9.1. Summary
Natural calamities, overpopulation and extreme poverty are threatening the food security
of many developing countries, particularly in Africa. The problem is even worse in the
SAT regions where water shortage is a recurrent occurrence. Grain legumes are among
the major vital crops that can produce sustainable grain yield and biomass in these harsh
environments and provide quality protein for the inhabitants besides serving as source of
cash income for the households. These crops also play a major role in low input
agricultural systems, and have the potential to contribute to the enhancement of the
natural resource base used for the production of other raifed crops which are the staple
foods of the poor communities. However, the yield of grain legumes is generally lower in
developing countries than the developed ones, and is the lowest in Africa (Al-Jibouri and
Kassapu, 1987; Oram and Agcaoili, 1992; Jeuffrroy and Ney, 1997). Water deficit is one
of the major constraints that contribute to the low yield of grain legumes in many regions
(e.g. Turk et al., 1980a; Graham and Ranalli, 1997; Kumar et al., 1996).
Common bean, chickpea and cowpea are the major grain legume crops traditionally
grown in Ethiopia either in the same or different environments. The yield of these crops is
very low either because of wrong crop choice or rigid management practices (such as
planting the same time every year) in the different growing environments. Successful
crop production in the SAT regions entails detail agroclimatic information, proper crop
choice and flexible options for management practices according to actual climatic
conditions. However, there is no such information in Ethiopia that can assist or advice on
choices among these grain legumes for a specific environment in terms of resource
utilization and productivity. Therefore, this study was initiated to compare the resource
utilization (water and radiation) and productivity of three grain legumes under water
stress and non-stress conditions, and to analyse the rainfall behaviour of selected grain
legume producing regions of Ethiopia. Information that has been reported in this study
will make a valuable contribution to agricultural scientists and extension officers with
regard to practical decisions for grain legume production in the semi-arid regions.
183
Chapter 2 has presented the historical rainfall analysis at ten weather stations in the grain
legume producing regions of Ethiopia. The study showed the existence of regional
differences in water supply for crop production. In some areas like Bahir Dar, Bako and
Bole, management of excess water is the major concern unlike other areas (Jijiga and
Dire Dawa) where water shortage is critical. Therefore, maximizing water use and water
use efficiency are crucial at Dire Dawa and Jijiga. The regions were categorized into three
groups in terms of the length of water availability period (length of effective growing
season) for crop production, namely, high water supply (Awassa, Bahir Dar, Bako, Bole,
. Debre Zeit), intermediate water supply (Alemaya, Mekele, Melkasa) and low water
supply (Dire Dawa and Jijiga). In this way, the need for adopting different management
practices (strategic and tactical) for optimum resource utilization in each region is
emphasized.
The major effects of water stress are reflected in altered growth, phenology and dry
matter distribution. Chapter 3 is devoted to investigate and describe possible differences
among species in their response to mid-season (MS) and late season (LS) water stress so
that the favourable traits identified can be exploited in the improvement of the respective
crops. Late season water stress significantly shortened the time to maturity of all three
grain legumes in all seasons while the effect in the mid-season stress was dependent on
temperature conditions after re-watering. The thermal time requirements of the different
phenological stages determined under water stress and non-stress conditions can be used
to predict each phenological stage for management decisions in the field and for
improvement of crop simulation models. Leaf area growth and above ground dry matter
production were only significantly affected by the mid-season stress. Among species, leaf
area was least affected in cowpea. Dry matter production is strongly positively correlated
with leaf area duration (LAD) in all the threes species indicating that factors that reduced
the LAD (water stress in this case) also reduced dry matter production. In the present
case, the reduction in dry matter in the LS, though not significant, is partly due to a
shorter LAD while the reduction in the MS is largely attributed to lower mid-season LAl,
and also a shorter LAD under high temperature conditions when compared to the control.
Dry matter allocation among above ground parts was influenced by both the timing of
water supply (growth stages) and species. High dry matter from the leaves and stem was
allocated to the pods during late season water stress followed by well-watered condition.
Under LS stress, dry matter allocation to pods was the highest in beans while it was
184
similar in chickpea and cowpea. Dry matter allocation was lower in the MS treatment
(only from leaves), and under such conditions chickpea allocated more dry matter to its
pods than beans and cowpea. The relationship between dry matter distribution among
above ground parts during the whole growth period was explained using regression
equations. These values are expected to be useful in modelling the growth of these crops.
Moreover, information on dry matter allocation is helpful to breeders to select and breed
crops for specific environments such as the one mentioned here (mid-season and terminal
drought environments). Although the relationship was not significant during mid-season
water stress conditions, a strong negative linear relationship between specific leaf area
(SLA) and water use efficiency (WUE) was found under well-watered conditions in all
the three species. Since the determination of WUE involves measurement of soil water
which requires expensive instruments, low SLA could be used to select cultivars for high
WUE under-well watered conditions in these species.
In Chapters 4 and 5, the water and radiation use and the respective use efficiencies of the
three grain legumes are presented. Dry matter production in grain legumes is found to be
strongly and positively correlated with the fraction of PAR that is intercepted (F) which is
also strongly positively correlated with LAL As a result, factors that reduce LAl also
reduce F and consequently result in low dry matter production. In line with this, water
stress during the flowering period reduced extinction coefficient (K), F and RUE when
compared to well-watered conditions. Species differences were observed on the effect of
mid-season water stress- on RUE in that water stress during the flowering period
significantly reduced RUE in beans and chickpea whereas it had no effect on cowpea. On
the other hand, water stress during the late season did not significantly affect these
parameters in any of the three species.
Water use was higher under well-watered conditions followed by mid-season stress in all
three species than the late season stress. The lower water use in the stress treatments is
mainly a result of low seasonal water supply and stomatal closure. However, WUE was
the highest in the LS stress and the lowest in the MS due to lower water use in the former
and due to lower dry matter production and higher soil evaporation in the latter when
compared to the control. Owing to its high seasonal water use as well as high dry matter
production, the WUE in the control is intermediate between the MS and LS treatments in
chickpea and cowpea while it is similar to the LS in beans. The species' have similar
185
water use but different WUE (lower in chickpea but similar in beans and cowpea) under
well-watered conditions which is mainly as result of differences in canopy cover and
resultant soil evaporation.When tested across seasons, the three species had similar WUE
under the pod filling period water stress. Nevertheless, chickpea had the lowest WUE
when water stress occurred during the flowering period stress. Beans and cowpea had
similar WUE under all water regimes when tested across seasons. Unlike chickpea, WUE
in beans and cowpea is strongly and positively correlated with HI so that improving the
WUE in beans and cowpea is expected to improve the grain yield of these crops. In
general, in terms of WUE, beans and cowpea are more productive than chickpea under
high water supply and mid-season drought environments while all the species may have
similar productivity under terminal drought environments. The differences between the
water stress treatments in affecting the radiation and water parameters indicate the
importance of the timing of water supply in affecting the resource utilization of grain
legumes. In this study, stress during flowering period was found to be detrimental in
reducing radiation and water use efficiency. Therefore, the implication of these results is
that crop management and breeding practices should focus on increasing the RUE and
WUE to improve the yield of grain legumes in sporadic mid-season drought prone
environments.
In Chapter 6, the physiological response of the three species to variable water supply and
weather conditions, and the inter-relationships among the physiological parameters were
investigated. Itwas found that the leaf water potential of chickpea was more responsive to
the decline in soil water than beans and cowpea. This was due to slow stomata closure
mechanism as indicated by low stomatal resistance in chickpea. Cowpea closes its
stomatal faster than both beans and chickpea which makes it a more drought-avoiding
crop than the other two species. The magnitude and rate of photosynthesis decline was
higher and faster in the mid-season than in the late-season stress in all species, and among
the species the rate of photosynthesis declined faster in chickpea than in beans and
cowpea. Cowpea had the lowest reduction in the rate of photosynthesis under severe
water stress (available soil water <32%) compared to the other two species. Increase in
leaf temperature was found to be the major factor for the decline of diurnal rate of
photosynthesis in the stressed plants through its effect on stomatal adjustment,
transpiration and possibly enzymatic activities.
186
Despite low rate of stomatal closure and low level of leaf water potential, chickpea
maintained a similar rate of photosynthesis to that of cowpea at the severe stage of the
water stress. This indicates that chickpea could have some molecular and cellular
adaptation mechanisms that enable it to maintain its photosynthesis similar to the species
with high leaf water potential under low available soil water conditions.
The rate of photosynthesis in the three species can be estimated from a few weather and
physiological parameters with reasonable accuracy. This will help those who do not have
the instruments to measure the photosynthesis of these crops directly. The estimation of
transpiration, however, was not encouraging probably because it is affected by more
variables than photosynthesis. The relationships established by the regression equations
between photosynthesis, transpiration, soil water, leaf water potential and stomatal
resistance are useful for modelling or calibrating existing crop growth models for the
three species. However, it should be noted that these relationships are cultivar specific.
In chapter 7, the growth stages that are most sensitive to water deficit for each species,
and those crop parameters that determine grain yield in each water regime treatment were
identified. The grain yield of beans and chickpea is more sensitive to flowering than pod-
filling water stress while the yield of cowpea is almost equally sensitive to both periods of
water stress, particularly under milder temperature seasons. Therefore, minimizing the
effect of water stress at these sensitive phenological stages for each crop through
management or breeding methods could_help maximize the yield of these crops. When
comparing the species, the yield of cowpea is less sensitive to both MS and LS stresses
than that of beans and chickpea in most of the seasons. This information can be useful in
the practice of crop choice based on environmental constraints such as water stress.
Water deficit during the flowering period resulted in the highest reduction in grain yield
which is attributed to the adverse effect of the stress on growth of reproductive organs,
LAl, efficiency of radiation and water use, and partitioning of dry matter to the seed. The
reduction of grain yield under pod-filling water stress is mainly due to a shorter
reproductive period and to some extent a reduction of reproductive organs such as
number of seeds per pod and numbers of pods per plant. Therefore, any management or
breeding activity which improves one or more of these characters is expected to improve
the yield of grain legumes under drought prone environments. Grain yield is positively
correlated with HI and RUE under all water supply conditions while its correlation with
187
WUE and p varied with water supply. Therefore, it is necessary to design environment
dependent selection strategies to improve the yield of grain legumes in a wide of range
environments with respect to water supply.
Crop modelling has proven a valuable tool to evaluate the long-term consequences of
weather patterns and environmental conditions. However, the candidate crop models must
be tested and calibrated for new regions and cultivars prior to their use as decision
support tools in agriculture. The DSSAT grain legume crop model (CROPGRO) was
evaluated in Chapter 8 for its ability in simulating evapotranspiration, LAl and yield of
beans and chickpea using three seasons observed data. Under both well-watered and
reproductive period water stress conditions, the model simulated LAl, crop ET and above
ground biomass at harvest with reasonable accuracy. The simulation for grain yield at
harvest was very good for both beans and chickpea. However, the simulation of yield
components at harvest in all seasons, and biomass accumulation during crop cycle under
high temperature conditions was very poor in both crops suggesting further improvements
of the models to make it suitable for the study environment. Model weakness was also
observed on simulation of maturity date in beans and flowering date in chickpea. It is
concluded that with further calibration using long-term data, the model has good potential
to be used as a decision support tool in the semi-arid areas of Ethiopia where long-term
field experiment is costly and less effective as a result of fluctuating weather conditions.
9- .2. Recommen--d-ations
Based on resource availability (in the various environments) and its use for productivity
(among species), the following recommendations are made:
• Site and season specific adjustment of crop choice and other management
practices for the regions studied. This will allow proper utilization of resources
and minimization of risks in each region unlike the traditional way of employing
the same farming practices every year.
• Jijiga, which has a very short growing season with high risk of planting failure, is
more suitable for livestock than crop production.
• Chickpea varieties that mature within 80 days or less are needed for Bole and
Debre Zeit where the crop is grown as a post-rainy season crop. On the other
188
hand, there is a huge potential to grow medium maturing chickpea varieties twice
in a season at Awassa where the crop is not yet widely grown.
• Common bean, which has a higher RUE and WUE, is more productive than
chickpea and cowpea under well-watered conditions. Therefore, whenever high
yield is desired from this crop (e.g. large scale production for export), it is
advisable to grow beans in the relatively high rainfall areas (with warm
temperature).
• Both beans and chickpea are more sensitive to flowering than pod-filling period
water stress. Therefore, it is necessary to match the flowering period of these
crops to the water availability period in order to maintain yield and/or minimize
yield reduction in water limited-environments.
• Cowpea is more adapted to mid- and late- season water limited environments than
beans and cowpea because of its ability to maintain LAl, RUE and photosynthesis
under water stress conditions. Therefore, cowpea is recommended over the other
two species for areas such as Dire Dawa where the growing period is short with
possible water shortage during the early reproductive periods.
9.3. Future studies
Firstly, agroclimatic information is one of the key elements that are needed for successful
crop production in semi-arid regions. However, such information is either lacking or
available only as large area recommendation in Ethiopia. The country is very large
covering about 1,221,900 sq. km with great terrain diversity and wide variations in
climate, soils, natural vegetation and settlement patterns. The country currently has more
than 530 first to fourth class weather stations. Of these, only 10stations were considered
in the present study. Therefore, further research on water supply, length of growing
season and water balance of the rest of the stations in a GIS (Geographical Information
Systems) environment is required to support agricultural decision-making in the country.
Secondly, the current study emphasized on the response of the three species to the
different water regimes based only on above ground parts of the plants. This is due to the
obvious reason that root measurement under field conditions is very difficult. However,
root size, morphology, length, density and hydraulic conductance are basic root attributes
to meet the transpiration demand of the shoot (Passioura, 1982). Therefore, in order to
189
supplement the current study with information about the below ground response of the
crops, root growth study, which requires specialized equipment, is suggested for each
species under controlled or semi-controlled conditions in the same environment and soil
conditions.
Thirdly, crop modeling is now serving as a decision support system in many developed
countries. Such a support system is deemed necessary for semi-arid regions in developing
countries where long-term field experiments are becoming unaffordable because of
economic reasons. Therefore, development of new grain legume crop growth models and
also intensive calibration of existing ones with long-term data to suit for semi-arid regions
is an immediate area of research to focus on.
"If you speak of development you have to start with water, it's as simple as that."
Carel de Rooy, UNICEF
190
References
Acosta Gallegos, J.A., Shibata, J.K., 1989. Effect of water stress on growth and yield of
indeterminate dry bean (Phaseolus vulgaris) cultivars. Field Crops Res. 20, 81-93.
Adams, M.W., 1967. Basis of yield component comparison in crop plants with special reference
to the field bean, Phaseolus vulgaris. Crop Sci. 7, 505-510.
Adams, M.W., Coyne, D.P., Davis, J.H.C., Graham, P.H., Francis, C.A. 1985. Common bean
(Phaseolus vulgaris L.). In: Summerfield, R.J., Roberts, E.H. (Eds.), Grain Legume
Crops. Collins, London, pp. 433-476.
Aggrawal, P.K., Kalra, N., 1994. Analysing the limitations set by climatic factors, genotype, and
water and nitrogen availability on productivity of wheat. II. Climatically potential yields
and management strategies. Field Crops Res. 38, 93-103.
Al-Jibouri, H.A., Kassapu, S., 1987. FAO grain legumes program in east and central Africa. In:
ICRISAT (International Crops Research Institute for the Semi-Arid Tropics). Research
On Grain Legumes in Eastern and Central Africa. Summary Proceedings of the
Consultative Group Meeting For Eastern and Central African Regional Research on
Grain Legumes (Groundnut, Chickpea and Pigeonpea), 8-10 December 1986,
International Livestock Center For Africa (ILCA), Addis Ababa, Ethiopia. Patancheru,
India: ICRISAT.
Allen, R.G., Pereira, L.S., Raes, D., Smith, M., 1998. Crop Evapotranspiration: Guidelines for
Computing Crop Water Requirements. FAO Irrigation and Drainage Paper, 56. Rome,
Italy. 300 pp.
Amede, K., 1998. Soil of the Dire Dawa Administrative Council. Dire Dawa Administrative
Council Agriculture Office, Addis Ababa. 80 pp.
Angstrom, A., 1924. Solar and terrestrial radiation. Q. J. R. Meteorol. Soc. 50, 121-125.
Angus, J.F., Moncur, M.W., 1977. Water stress and phenology in wheat. Aust. J. Agrie. Res. 28,
177-181.
Angus, J.F., Hasegawa, S., Hsiao, T.C., Liboon, S.P., Zandstra, H.G., 1983. The water
balance of'post-monsoon dryland crops. J. Agric..Sci. 101,699-710.
Aphalo, P.J., Jarvis, P.G., 1991. Do stomata respond to relative humidity? Plant Cell Environ.
14, 127-132.
Arkebauer, T.J., Weiss, A., Sinclair, T.R., Blum, A., 1994. In defence of radiation use
efficiency: A response to Demetriades-Shah et al. (1992). Agrie. For. Meteorol. 68,
221-227.
Ashok, l.S. Aftab Hussien, Prasad, T.G., Udaya Kumar, M., Negeswara Rao, R.C., Wright,
G.C., 1999. Variation in transpiration efficiency and carbon isotope discrimination in
cowpea. Aust. J. Plant Physiol. 26, 503-510.
Azam-Ali, S.N., 1983. Seasonal estimates of transpiration from a millet crop using aporometer.
Agrie. For. Meteorol. 30, 13-24.
Azam-Ali, S.N.,Simonds, L.P., Negeswara Rao, R.C., Williams, J.H., 1989. Population, growth
and water use efficiency of groundnut (Arachis hypogea) maintained on stored soil
water. m. Dry matter, water use and light interception. Exp. Agrie. 25, 77-86.
Baldocchi, D.D., Verma, S.B., Rosenberg, N.J., 1981a. Environmental effects on the CO2 flux
and CO2-water flux ratio of alfalfa. Agrie. Meteorol. 24, 175-184.
Baldocchi, D.D., Verma, S.B., Rosenberg, N.J., 1981b. Mass and energy exchange ofa soybean
canopy under various environmental regimes. Agron. J. 73, 706-710.
191
Baldocchi, D.D., Verma, S.B., Rosenberg, N.l., 1985. Water use efficiency in a soybean field:
Influence of plant water stress. Agrie. For. Meteorol. 34, 53-65.
Barnard, R.O., Rethman, N.F.G., Annandale, J.G., Mentz, H.W., Jovanovic, N.C., 1998. The
screening of crop, pasture, and wetland species for tolerance of polluted water
originating in coal mines. Water Research Commission Report, No. 582/1/98, South
Africa, pp. 259-262.
Barradas, V.L., Jones, H.G., Clark, J.A, 1994. Stomatal response to changing irradiance in
Phaseolus vulgaris L. J Expt. Bot. 45, 931-936.
Barron, J., Rockstrëm, J., Gichuki, F., 1999. Rainwater management for dry spell mitigation in
semi-arid Kenya. E. Afr. Agrie. Forestry. J. 65, 57-69.
Bates, L.M., Hall., AE., 1982a. Relationship between leaf water status and transpiration in
cowpea with progressive soil drying. Oeeologia (Berl.) 53, 285-289.
Bates, L.M., Hall., AE., 1982b. Diurnal and seasonal response of stomatal conductance for
cowpea plants subject to different levels of environmental drought. Oeeologia (Berl.)
54, 304-308.
Begg, J.E., 1980. Morphological adaptations ofleaves to water stress. In: Turner, N.C., and
Kramer, PJ. (Eds.), Adaptations of Plants to Water and High Temperature Stress.
Wiley, New York, pp. 33-42.
Bell, MJ., Muchow, R.C., and Wilson, G.L. 1987. The effect of plant population on peanuts
(Arachis hypogaea) in a monsoonal tropical environment Field Crops Res. 17, 91-107.
Beltrando, G., 1990. Space-time variability of rainfall in April and October-November over East
Africa during the period 1932-83. Int. J Climatol. 10,691-702.
Beltrando, G., Camberlin, P., 1993. Inter annual variability of rainfall in the eastern horn of
Africa and indicators of atmospheric circulation. Int. J Climatol.13, 533-546.
Bierhuizen, J.F., Slatyer, R.O., 1965. Effect of atmospheric concentration of water vapour and
CO2 determining transpiration-photosynthesis relationship. Agrie. Meteorol. 2, 259-170.
Black C., Ong, C., 2000. Utilization of light and water in tropical agriculture. Agrie.
For. Meteorol. 104,25-47.
Blade, S.F., Shetty, S.V.R., Terao, T., Singh, B.B., 1997. Recent developments in cowpea
cropping systems research. In: Singh, B.B., Raj, M. (Eds.), Advances in Cowpea
Research. International Institute of Tropical Agriculture: Ibadan, Nigeria; Japan
International Research Center for Agricultural Sciences: Tsukuba, Japan.
Blum, A, 1988. Plant Breeding for Stress Environments. CRC Press, Boca Raton, FL.
Blum, A, 1996. Crop responses to drought and interpretation of adaptation. Plant Growth
Regul. 20, 135-148.
Blum, A, Johnson, J.W., 1993. Wheat cultivars respond differently to drying top soil and a
possible non-hydraulic root signal. J Exp. Bot. 44, 1131-1139.
Boote K.J., Jones, J.W., Pickering, N.B., 1996. Potential uses and limitations of crop models.
Agron. J. 88,704-716.
Boote, K.J., Jones, J.W., Hoogenboom, G., 1998a. Simulation of crop growth: CROPGRO
model. In: Peart, R.M., Curry, R.B. (Eds.), Agricultural Systems Modelling and
Simulation. Marcel Dekker, New Youk, pp.780-788.
Boote, K.J., Jones, J.W., Hoogenboom, G., Pickering, N.B., 1998b. The CROPGRO model for
gain legumes. In: Tsuji, G.Y., Hoogenboom, G., Thornton, P.K. (Eds.), Understanding
Options for Agricultural Production. Kluwer Academic Publishers, Dordrecht, the
Netherlands, pp.99-128.
192
Boote, KJ., Minguez, M.I., Sau, F., 2002. Adapting the CROPGRO legume model to simulate
growth of faba bean. Agron. J. 94, 743-756.
Bordribb, T., 1996. Dynamics of changing intercellular CO2 concentration (Ci) during drought
and determination of minimum functional Ci. Plant Physiol. 11, 179-185.
Borrell, AK., Incoll, L.D. Simpson, R.J., Dalling, MJ., 1989. Partitioning of dry matter and the
deposition and use of stem reserves in a semi-dwarf wheat crop. Ann. Bot. 63, 527-539.
Botes, J.H.F., 1994. A simulation and optimisation approach to estimate the value of irrigation
information for decision makers under risk. Ph.D. Thesis, University of the Orange Free
State, South Africa.
Boyer, J.S., 1982. Plant productivity and environment. Science 218, 443-448.
Boyer, J.S., 1996. Advances in drought tolerance in plants. Adv. Agron. 56, 187-219.
Boyer, J.S., Wong, S.C., Farquhar, G.D., 1997. CO2 and water vapour exchange across leaf
cuticle (epidermis) at various water potentials. Plant Physiol. 114, 185-191.
Bressani, R., 1985. Nutritive value of cowpea. In: Singh, S.R., Rachie, K.O. (Eds.), Cowpea
Research, Production and Utilization. Wiley, New York, pp. 353-359.
Brown, L.H., Cocheme, J., 1969. A study of the agroclimatology of the highlands of Eastern
Africa. FAO, Rome.
Bushby, H.V.A, Lawn, R.J., 1992. Accumulation of partitioning of nitrogen and dry matter by
contrasting genotypes of mungbean (Vigna radiata L. Wilezek) Aust. J Agric Res. 43,
1609-1628.
Calcagno, F., Gallo, G., 1993. Physiological and morphological basis of abiotic stress resistance
in chickpea. In: K.B. Singh, M.C. Saxena (Eds.), Breeding for Stress Tolerance in Cool-
Season Food Legumes, ICARDA, Wiley, Chichester, UK, pp 293-309.
Calvache, M., Reichardt, K., 1999. Effect of water stress imposed at different plant growth
stages of common bean (Phseolus vulgaris) on yield and N2 fixation. In: Kirda, C.,
Moutonnet, P., Hera, C., Nielsen, D.R. (Eds.), Crop Yield Response to DefIcit
Irrigation. Developments in Plant and Soil Sciences, Vol. 84, Kluwer Academic
Publishers, pp. 121-127.
Calvet, J.C., 2000. Investigating soil and atmospheric plant water stress using physiological and
micrometeorological data. Agric. For. Meteorol. 103,229-247.
Campbell, G.S., and Norman, J.M., 1989. The description and measurement of plant canopy
structure. In: Russell, G., Marshall, B., and Jarvis, P.G. (Eds.), Plant Canopies: Their
Growth, Form and Function. Cambridge University Press, Cambridge, pp. 1-19.
Causton, R.D., Venus, J.C., 1981. The Biometry of Plant Growth. Edward Arnold, 307 pp.
Carberry, P. S., Hochman, Z., McCown, R. L., Dalgliesh, N. P., Foale, M. A, Poulton, P.
L.,Hargreaves, J. N. G., Hargreaves, D. M. G., Cawthray, S., Hillcoat N., Robertson M.
J., 2002. The FARMSCAPE approach to decision support: Farmers', advisers',
researchers' monitoring, simulation, communication and performance evaluation. Agric.
Syst. 74, 141-177.
Ceotto, E., and Castelli, F., 2002. Radiation use efficiency in flue-cured tobacco (Nicotiana
tabacum L.): Response to nitrogen supply, climate variability and sink limitations. Field
Crops Res. 74,117-130.
Chapman, S, C., Ludlow, M. M., Blamey, F.P.C., 1993. Effect of drought during early
reproductive development on the dynamics of yield development of cultivars of
groundnut (Arachis hypogaea L.) Field Crops Res. 32,227-242.
Choudhury, BJ, 1985. Evaluating plant and canopy resistance of field grown wheat from
193
concurrent diurnal observations of leaf water potential, stomatal resistance, canopy
temperature, and evaporation flux. Agric. For. Meteorol. 34, 67-76.
Collatz, GJ., Ball. LT., Grivet, C., Berry, lA., 1991. Physiological and environmental
regulation of stomatal conductance, photosynthesis and transpiration. A model that
includes laminar boundary layer. Agric. For. Meteorol. 54, 107-136.
Collino, DJ., Dardanelli, lL., Sereno, R. Racca, R.W., 2000. Physiological response of
argentine peanut varieties 0 water stress. Water uptake and water use efficiency. Field
Crops Res. 68, 133-142.
Connor, DJ., Jones, TJ., Palta, lA., 1985. Response of sunflower to strategies of irrigation. II.
Growth, yield and efficiency water use. Field Crops Res. 10, 15-26.
Cooper, PJ.M., Campbell, G.S., Heath, M.C., Hebblethwaite, P.D., 1988. Factors which affect
water use efficiency in rainfed production of food legumes, and their measurement. In:
R.J. Summerfield (Ed.), World Crops: Cool season Food Legumes". Kluwer Academic
Publishers, pp. 813-829.
Cooper, PJ.M., Gregory, PJ., Tully, D., Harris, H.C., 1987. Improving water use efficiency of
annual crops in rainfed farming systems of West Asia and North Africa. Exp. Agric. 23,
113-158.
Costa Franca, M.G., Thai, A.T.P., Pimentel, C., Rossiello, R.O.P., Zuily-Fodil, Y., Laffray, D.,
2000. Differences in growth and water relations among Phaseolus vulgaris cultivars in
response to induced drought stress. Env. Exp. Bot. 43, 227-237.
Cowan, LR., Troughton, J.H., 1971. The relative role of stomata in transpiration and
assimilation. Planta 97, 325-336.
Cown, I.R., 1977. Stomatal behavior and environment. Adv. Bot. 4, 117-128.
Craufurd, P.Q., Summerfield R.J., Ellis, R.H., and Roberts, E.H., 1997. Photoperiod,
temperature, and the growth and development of cowpea. In: Singh, B.B., Mohan Raj,
D.R., Dashiell, K.E., and Jackai, L.E.N. (Eds.) Advances in Cowpea Research. Co-
Publication of International Institute of Tropical Agriculture (IITA) and Japan
International Research Centre for Agricultural Sciences (JIRCAS). !ITA, Ibadan,
Nigeria, pp. 75-86.
Cross, H.Z., Zuber, M.S., 1972. Prediction of flowering dates in maize based on methods of
estimating thermal time units. Agron. J. 64, 351-355.
Cruz de Carvalho, M.H., Laffray, D., Louguet, P., 1998. Comparison of the physiological
response of Phaseolus vulgars and Vigna unguiculata cultivars when submitted to
drought conditions. Env. Exp. Bot. 40, 197-207.
CSA (Central Statistical Authority), 1997. Agricultural sample survey 1996/97 (1989 E.C).
. Report on area and production for major crops (private peasant holdings, 'meher'
season). Vol. 1, No. 171. Addis Ababa, Ethiopia.
Davies, W.l, Mansfield, T.A., Hetherington, A.M., 1990. Sensing for soil water status and the
regulation of plant growth and development. Plant Cell Environ. 13,709-719.
Davies, W.J., Zhang, L, 1991. Root signals and the regulation of growth and development of
plants in drying soil. Plant Physiol. Plant Mol. Biol. 42, 55-76.
De Costa, W.A.J.M., Shanmugathasan, K.N., Joseph, K.D.S.M., 1999. Physiology of yield
determination of mung bean (Vigna radiata (L.) Wilczek) under various irrigation
regimes in the dry and intermediate zones of Sri Lanka. Field Crops Res. 61, 1-12.
de Jager, lM., Botham, D.P., van Vuuren, C.J.J., 1981. Effective rainfall and the assessment of
potential wheat yields in shallow soil. Crop Production 10, 51-56.
de Wit, C.T., 1965. Photosnthesis of leaf canopies. Inst. Biol. Chemo Res. Field Crops Herb.
194
Agric. Res. 663. Wageningen, Netherlands.
Dennet, M.D., Keating, J.D.H., Rodgers, J. A, 1984. A comparison of rainfall regimes at six
- sites in Northern Syria Agric. For. Meteorol. 31, 319-328.
Diallo, AT., Samb, PJ., Roy-Macauley, H., 2001. Water status and stomatal behaviour of
cowpea, Vigna anguiculata (L.) Walp, plants inoculated with two Glomus species at
low moisture levels. Eur. J Soil Bioi. 37, 187-196.
Dijkstra, P, Lambers, H., 1989. A physiological analysis of genetic variation in relative growth
rate within Plantago major L. Functional Eco/. 3, 577-587.
Dingkuhn, M, Tivet, F., Siband, P., Asch, F., Audebert, A, Sow, A., 2001. Varietal differences
in specific leaf area: A common physiological determinant of tillering ability and early
growth vigor? In: Peng S., and Hardy B. (Eds.), Rice Research for Food Security and
Poverty Alleviation. Proceedings of the International Rice Research Conference, 31
March to April 2000, Los Banos, Philippines. International Rice Research Institute, Los
Banos, Philippines, pp. 95-108
Duncan, W.G., Roomis, R.S., Williams, W.A, Hannu, R., 1967. A model for simulating
photosynthesis in plant communities. Hilgardia 38, 181-205.
Duncan, W.G., McCloud, D.E., MCaw, R.L., Boote, J.J., 1978. Physiological aspects of peanut
yield improvement. Crop Sci. 18, 1015-1020.
Duranti, M., Gius, C., 1997. Legume seeds: protein content and nutritional value. Field Crops
Res. 53, 31-45
Dwyer, L.M., Stewart, O.W., 1987. Influence of photoperiod and water stress on growth, yield
and development rate of barley measured in heat units. Can. J Plant Sci. 67, 21-34.
Ehlers, J.D., Hall, AE., 1997. Cowpea (Vigna unguiculata L. Walp.). Field Crops Res. 53, 187-
204.
Ehr1er, W.L., Idso, S.B., Jackson, R.D., Reginato, R.J., 1978. Diurnal changes in water potential
and canopy temperature of wheat as affected by drought. Agron. J. 70, 999-1004.
Elowad, H.O. A, Hall, AE., 1987. Influence of early and late nitrogen fertilization on yield and
nitrogen fixation of cowpea under well-watered and dry field conditions. Field Crops
Res. 15,229-244.
Evans, L.T., 1993. Crop Evolution, Adaptation and Yield. Cambridge University Press,
Cambridge, UK, 500 pp.
Eylachew Z., 1994. Properties of major soils of Ethiopia. Report on National Soil Reference
Collection. FAOIUNDP, Addis Ababa, Ethiopia.
FAO, 1994. Food and Agriculture Organization of the United Nations Year Book. FAO, Rome,
Italy.
FAO, 2003. Agriculture, Food and Water. Food and Agricultural Organization of the United
Nations, Rome, Italy.
Farquhar, G.O., Sharkey, T.D., 1982. Stomatal conductance and photosynthesis. Ann. Rev. Plant
Physiol. 33,317-345.
Fasheun, A, Dennett, M.D., 1982. Interception of radiation and growth efficiency in field
beans (Viciafaba L.) Agric. Meteorol. 26, 221-229.
Fery, R.L., 1990. The cowpea: Production, utilization, and research in the United States. Hort.
Rev. 12, 197-222.
Fischer, R.A, Turner, N.C., 1978. Plant productivity in the arid and semiarid zones. Ann. Rev.
PlantPhysiol. 29,277-317.
195
Flower, D.J., Ludlow, M.M., 1986. Contribution of osmotic adjustment to the dehydration
tolerance of water-stressed pigeonpea (Cajanus cajan L. millsp.) leaves. Plant Cell
Environ. 9, 33-40.
French, R.J., Schultz, J.E., 1984. Water use efficiency of wheat in Mediterranean type
environment. I. The relationship between yield, water use and climate. Aust. J. Agric.
Res. 35,743-764.
French, R.J., Turner, N.C., 1991. Water deficit change dry matter partitioning and seed yield in
narrow-leafed lupins (Lupinus angustifolius L.) Aust. 1. Agric. Res. 42, 471-484.
Gabrielle, B., Menasseri, S., Houot, S., 1995. Analysis and field evaluation ofCERES model
water balance components. Soil Sci. Soc. Am. 1. 59, 1403-1412.
Gallagher, J.L., Biscoe, P.V., 1978. Radiation absorption, growth and yield of cereals. 1.
Agric. Sci. (Camb.) 91,47-60.
Gardiner, T.R., Vietor, D.M., Craker, L.E., 1979. Growth habit and row width effects on
leaf area development and light interception of field beans. Can. 1. Plant Sci., 59, 191-
199.
Goldsworthy, P.R., 1984. Crop growth and development: The reproductive phase. In:
Goldsworthy, P.R., Fisher, N.M. (Eds.), The Physiology of Tropical Field Crops. John
Wiley, UK, pp. 163-212.
Gollan, T., Turner, N.C., Schuze, E.D., 1985. The response of stomata and leaf gas exchange to
vapor pressure deficit and soil water content. Ill. In the sclerophyllous species Nerium
oleander. Oecologia (Berl.) 65, 356-362.
Goodrich, D.C. et al., 2000. Preface paper to the Semi-Arid-Land-Surface-Atmosphere
(SALSA) program special issue. Agric. For. Meteorol. 105,3-20.
Goss, G., Varlet-Grancher, C., Bonhomme, R., Chartier, M., Allirand, J.M., and Lemaire, G.,
1986. Maximum dry matter production and solar radiation intercepted by a canopy.
Agronomie 6, 47-56.
Graham, P.H., Ranalli, P. 1997. Common bean (Phaseolus vulgaris L.). Field Crops Res.
53, 131-146.
Grantz, D.A., Hall, A.E., 1982. Earliness of an indeterminate crop Vigna unguiculata (L.)Walp,
as affected by drought, temperature and plant density. Aust. 1. Agric. Res. 33, 531-540.
Grantz, D.A., Meinzer, F.C., 1990. Stomatal response to humidity in a sugarcane field:
Simultaneous porometric and micrometeorological measurements. Plant Cell Environ.
13,27-37.
Grantz, D.A., Moore, P.H., Zeiger, E., 1987. Stomatal response to light and humidity in
sugarcane: Predication of daily time courses and identification of potential selection
criteria. Plant Physiol. 10, 197-204.
Gray, C., Jones, J.W., Longuenesse, J.J., 1993. Modelling daily changes in specific leaf area of
tomato: The contribution of the leaf assimilate pool. Acta Hort. 328,205-210.
Green, C.F., 1987. Nitrogen nutrition and wheat growth in relation to absorbed solar radiation.
Agric. For. Meteorol. 41,207-248.
Green, F.C., Hebblethwaite, P.D., Ison, D.A., 1985. A quantitative analysis of varietal and
moisture status effects on the growth of Vicia faba in relation to radiation absorption.
Ann. Appl. Bioi. 106, 143-145.
Greenburg, D.C., Williams, J.H., Ndunguru, B.J., 1992. Differences in yield determining
processes of groundnut (Arachis hypogaea L.) genotypes in varied drought
environments. Ann. Appl. Bioi. 120,557-566.
196
Gregory, P.J., Eastham, 1., 1996. Growth of shoots and roots, and interception of radiation by
wheat and lupin crops on a shallow, duplex soil in response to time of sowing. Aust. J.
Agrie. Res. 47, 427-447.
Guyer, R.B., Kramer, A., 1952. Studies off actors affecting the quality of green and wax beans.
University of Maryland Bull. A68.
Haile, T., 1986. Climatic variability and support feedback mechanism in relation to the
Sahelo-Ethiopian droughts. M.Sc. Thesis, Department of Meteorology, University of
Reading, UK, pp.l19-137.
Hall, A.E., Patel, P.N., 1985. Breeding for resistance to drought and heat. In: Singh, S.R.,
Rachie, K.O. (Eds.), Cowpea: Research, Production and Utilization. Wiley, New York,
pp. 137-151.
Hammer, G.L. Woodruff, D.R., Robinson. 1.B., 1987. The effect of climatic variability and
possible climatic change on reliability of wheat cropping: a modelling approach. Agrie.
For. Meteorol. 41, 13-142.
Hardwick, R.e., 1988a. Critical physiological traits in pulse crops. In: Summerfield, R.J. (Ed.),
World Crops: Cool Season Food Legumes. Kluwer, Dordrecht, The Netherlands, pp.
885-896.
Hardwick, R.e., 1988b. Review of recent research on navy bean (Phaseolus vulgaris) in the
United Kingdom. Ann. App. Bioi. 113, 205-227.
Harris, D., Matthews, R.B., Nageswara Rao, R.C., and Williams, 1.H., 1988. The physiological
basis for yield differences between four genotypes of groundnut (Arachis hypogaea) in
response to drought. Ill. Developmental processes. Exp. Agrie. 24, 215-226.
Haterlein, A. J., 1983. Bean. In: Tearce, T.P., Peet, M. M. Jr (Eds.), Crop Water Relations.
Wiley, New York, pp. 157-185.
Heath, M.C., and Hebblethwaite, P.D., 1985. Solar radiation interception by leafless, semi-
leafless and leafed peas (Pisum sativum) under contrasting field conditions. Ann. App.
Bioi. 107,309-318.
Hensley, M., Anderson, J.J., Botha, J.1., van Staden, P.P., Singels, A., Prinsloo, M., Du Tiot, A.,
1997. Modelling the water balance on benchmark ecotopes. WRe Report No. 508/1/97,
Pretoria, South Africa.
Hinckley, T.M., and Braatne, J.H., 1994. Stomata. In: Wilkinson, R.E. (Ed.), Plant-
Environment Interactions. Marcel Dekker, Inc., New York, pp. 323-355.
Hintze, J.L., 1997. NCSS 97, Statistical systems for windows: User guide. Number Cruncher
Statistical Systems, Kaysville, Utah, USA.
Hoogenboom, G., Jones, J.W., Wilkens, P.W., Batchelor, W.D., Porter, C.H., Boote, K.J., Hunt,
L.A., Bowen, W.T., Hunt, L.A, Pickering, N.B., Singh, U., Godwin, D.e., Baer B.,
Boote, K.1., Ritchie, J.T., White, J.W., 1994a. Crop models. In: Tsuji, G.Y., Uehara, G.,
Balas, S. (Eds.), DSSAT v3, vol. 2. University of Hawaii, Honolulu, Hawai, pp. 95-244.
Hoogenboom, G., White J.W., Jones, 1.W., Boote, KJ., 1994b. BEANGRO: A process-
oriented dry bean model with a versatile user interface. Agron. J. 86, 182-190.
Hoogenboom, G., Wilens, P.W., Tsuji, G., 1999. International Benchmark Sites Network for
Agrotechnology Transfer, DSSAT version 3, Vol. 4. University of Hawaii, Honolulu,
Hawaii.
Hoshikawa, K. 1991. Significance of legume crops in improving the productivity and stability of
cropping systems. In: Johansen, C.; Lee, KK., and Sahrawat, KL. (Eds.), Phosphorus
Nutrition of Grain Legumes in the Semi-Arid Tropics. ICRISAT, Patancheru, India, pp.
173-181
197
Houérou, H.N., Popov, G.F., See, L., 1993. Agro-Bioclimatic Classification of Africa.
Agrometeorology Series Working Paper, No. 6., FAO, Rome, Italy. 204 pp.
Hsiao, T.C., 1973. Plant resources to water stress. Ann. Rev. Plant Physiol. 24, 519-570.
Hubick, K.T., Farquhar, G.o., Shorter, R., 1986. Correlation between water use efficiency and
carbon isotope discrimination in diverse peanut (Arachis) germplasm. Aust. J Plant
Physiol. 13, 803-816.
Huda, A.K.S., Cogle, A.L., Miller, C.P., 1990. Agroclimatic analysis of selected locations in
North Queensland. International Crops Research Institute for the Semi-Arid Tropics
(ICRISAT), Patancheru, Andhra Pradesh, India and Queenslands Department of
Primary Industries Mareeba, Queensland, Australia, 181 pp.
Huda, A.K.S., Virmani, S.M., 1987. Effect of variations in climatic and soil water on
agricultural productivity of selected locations in India. In: Parry, M. L., Carter, T. R.
and Konijn, N.T. (Eds.), The Impact of Climatic Variation on Agriculture, Vol. 2.
Assessment in Semi-Arid Regions. International Institute for Applied Systems Analysis
and the United Nations Environment Program. Reidel, Dordrecht, The Netherlands, pp.
36-55.
Hughes, G., and Keatinge, J.D .H. 1983. Solar radiation interception, dry matter production and
yield in pigeon pea (Cajanus eajan (L.) Millspaugh). Field Crops Res.6, 171-178.
Hughes, G., Keatinge, J. D.H., Scott, Cooper J.B.M, and Dee, N.F., 1987. Solar radiation
interception and utilization by chickpea (Cieer arietinum L.) crops in northern Syria. J.
Agrie. Sci. (Camb.) 108,419-424.
Hunt, L.A., Pararajasingham, S.; 1994. Genotype coefficient calculator. In: Tsuji, G.Y., Uehara,
G., Balas, S. (Eds.), DSSAT v3, vol. 3. University of Hawaii, Honolulu, Hawai, pp.
201-223.
Hunt, R. 1982. Plant Growth Curves. The Functional Approach to Growth Analysis. Edward
Arnol. London.
Husain, M.M., Reid, J.B., Othman, H., Gallagher, J.N., 1990. Growth and water use offaba
beans (Vieiafaba) in a sub-humid climate I.Root and shoot adaptations to drought
stress. Field Crops Res. 23, pp. 1-17.
ICRISAT (International Crops Research Institute for the Semi-Arid Tropics). 1988. Annual
report for the year 1987. ICRISAT, Patancheru, India, pp. 327-330.
Jagtap, S.S., Abamu, F.J., Kling, J.G., 1999. Long-term assessment of nitrogen and variety
technologies on attainable maize yields in Nigeria using CERES-maize. Agrie. Syst. 60,
77-86.
Jagtap, S. S., Jones, J. W., Hildebrand, P., Letson, D., O'Brien, J. J., Podestá, G., Zierden, D.,
Zazueta F., 2002. Responding to stakeholder's demands for climate information: from
research to applications in Florida. Agrie. Syst. 74,415-430.
Jalali-Farahani, H.R., Slack, D.C., Kopec, D.M., Matthias, A.D., 1993. Crop water stress index
for Bermuda grass turf: A comparison. Agron. J. 85, 1210-1217.
Jamieson, P.D., Martin, R.J., Francis, G.S., 1995. Drought influences on grain yield of barley,
wheat and maize. N Z. J. Crop Hort. Sci. 23, 55-66.
Jamieson, P.D., Porter J.R., Wilson D.R., 1991. A test of the computer simulation model ARC-
WHEAT! on wheat crops grown in New Zealand. Field Crops Res. 27, 337-350.
Jayasundara, H.P.S., Thomson, B.D., Tang, C., 1998. Response of cool season food legumes
to soil abiotic stress. Adv. Agron. 63, 78-153.
Jeuffroy, M.-H., Ney, B., 1997. Crop physiology and productivity. Field Crops Res. 53,3-
16.
198
Jiang, H.F., Egli, D.B., 1995. Soybean seed number and crop growth rate during flowering.
Agron J. 87, 264-267.
Johansen, C., Baldev, B., Brouwer, J.B., Erskine, W., Jermyn, W.A, Li-Juan, L., Malik, B.A,
Ahad Miah, A, Silim, S.N., 1992. Biotic and abiotic stresses constraining productivity
of cool season food legumes in Asia, Africa, and Oceania. In: Muehlbauer, FJ., Kaiser,
W.J., (Eds.), Expanding Production and Use of Cool Season Food Legumes. Kluwer
Dordrecht, pp. 175-194.
Johansen, C., Singh, D. N., Krishnamurthy, L., Saxena, N.P., Chauhan, Y.S., Kumar Rao,
J.V.D.K, 1997. Options for alleviating stress in pulse crops. In: Asthana, AN., Ali, M.
(Eds.), Recent Advances in Pulse Research. Indian Society of Pulse Research and
Development, Indian Institute of Pulse Research (IIPR), Kanpur, India, pp. 425-442.
Jones, H.G., 1992. Plants and Microclimate. A Quantitative Approach to Environmental Plant
Physiology. Second Edition, Cambridge University Press, Cambridge.
Jones, H.G., Corlett, J.E., 1992. Current topics in drought physiology. J. Agrie. Sci.
(Camb.) 119,291-196.
Jones, J.W., Hoogenboom, G., Porter, C.H., Boote, KJ., Batchelor, W.D., Hunt, L.A., Wilkens,
P.W., Singh, U., Gijsman, AJ., Ritchie, J.T., 2003. The DSSAT cropping system
model. Eur. J. Agron. 18,235-265.
Jones, J.W., Hunt, L.A, Hoogenboom, G., Godwin, D.C., Singh, u., Tsuji, G.Y., Pickering,
N.B., Thornton, P.K, Bowen, W.T., Boote, KJ., Ritchie, J.T., 1994. Input and output
files. In: Tsuji, G.Y., Uehara, G., Balas, S. (Eds.), DSSAT v3, Vol. 2. University of
Hawaii, Honolulu, Hawai, pp. 1-94.
Kamel, A, Schroeder, K., Sticklen, J., Rafea, A, Salah, A, Schulthess, U., Ward, R., Ritchie,
J.T., 1995. Integrated wheat crop management based on generic knowledge-based
systems and CERES numerical-simulation. AI Applications 9, 17-28.
Kassam, AH., 1977. Agroc1imatic Suitability Assessment of Rainfed Crops in Africa by
Growing Period Zones. FAO, Rome.
Kawamitsu, Y., Yoda, S., Agata, W., 1993. Humidity pre-treatment affects the response of
stomata and CO2 assimilation to vapour pressure difference in C3 and C4 plants. Plant
Cell Physiol. 34, 113-119.
Keatinge, J. D. H., Cooper, P.J.M., 1983. Kabuli chickpea as a winter-sown crop in northern
Syria: Moisture relations and crop productivity. J. Agrie. Sci. (Camb.) 100,667-680.
Keatinge, J.D.H., Cooper, PJ.M., and Hughes, G., 1985. The potential of peas as a forage in the
dry land cropping rotations of Western Asia. In: Hebblethwaite, P.D., Heath, M.C. and
Dawkins, T.C.K. (Eds.), The Pea Crop. Butterworths, London, pp.l85-192.
Kirda, C., Kanber, R., 1999. Water, no longer a plentiful resource, should be used sparingly in
irrigated agriculture. In: Kirda, C., Moutonnet, P., Hera, C., Nielsen, D.R. (Eds.), Crop
Yield Response to Deficit Irrigation. Developments in Plant and Soil Sciences, Vol. 84,
Kluwer Academic Publishers, pp.l-20.
Korte, L.L., Specht, J.E., Williams, J.H. and Sorensen, R.C., 1983. Irrigation of soybean
genotypes during reproductive ontogeny. il: Yield component response. Crop Sci. 23,
528-533.
Kowal, J.M., Kassam, AH., 1978. Agricultural Ecology of Savanna. Oxford University Press,
Oxford, UK, 403 pp.
Kowal, J.M., Knabe, D.T., 1972. An agroc1imatological atlas of the northern states of Nigeria.
Ahmadu Bello University Press, Zaria.
Kramer, P.J., 1980. Drought, stress and the origin of adaptations. In: Turner, N.C. and Kramer
199
P'J. (Eds.), Adaptation of Plants to Water and High Temperature Stress. John Willey &
Sons Inc., New York, pp. 7-20.
Kumar, J., Abbo, S., 2001. Genetics of flowering time in chickpea and its bearing on
productivity in semiarid environments. Adv. Agron. 72, 107-138.
Kumar, J., Sethi, S,C., Johansen, C., Kelley, T.G., Rahman, M ..M., van Rheenen, H.A., 1996.
Potential of short-duration chickpea varieties. Ind. 1. Dryland Agrie. Res. Dev. 11, 28-
32.
Kumudini, S., Hume, D.J., Chu, G., 2001. Genetic improvement in short-season soybeans: I.
Dry matter accumulation, partitioning, and leaf area duration. Crop Sci. 41, 391-398.
Kuppers, B.I.L., Kuppers, M., Schulze, E.D., 1988. Soil drying and its effect on leaf
conductance and C02 assimilation of Vigna anguiclata (L.) Walp. I. The response to
climate factors and to the rate of soil drying in young plants. Oecologia 75, 99-104.
Laffary, D., Loguet, P., 1990. Stomata response and drought resistance. Bulletin de la
Société Botanique de France 137,47-60.
Lawlor, D.W., 1995. The effect of water deficit on photosynthesis. In: Smirnoff, N. (Ed.),
Environment and Plant Metabolism. Bioscientific Publishers, Oxford, UK, pp. 129-160.
Lawn, RJ., 1982. Response of four grain legumes to water stress in southern queensland.I.
Physiological response mechanisms. Aust. 1.Agrie. Res. 33, 482-496.
Lawn, R.J., Ahn, C.S., 1985. Mung bean (Vinga radiata (L.) Wilczek/Vigna Mungo (L.)
Hepper). In: Summerfleld, RJ., Roberts, E.H. (Eds.), Grain legume Crops, Collins,
London, pp. 584-623.
Lecoeur, J., Wery, J., Turc, 0., Tardieu, F., 1995. Expansion of pea leaves subject to short
water deficit: Cell number and cell size are sensitive to stress at different periods of leaf
development. 1. Exp. Bot. 46, 1093-1101.
Leach, G.J., and Beech, D.F., 1988. Response of chickpea accessions to row spacing and plant
density on a vertisol on the Darling Downs, South-eastern Queensland. II. Radiation
interception and water use. Aust. 1.Exp. Agrie. 28, 377-383.
Leport, L., Turner, N.C., French, RJ., Barr, M.D., Duda, R, Davies, S.L., Tennant,D., Siddique,
K.H.M., 1999. Physiological response of chickpea genotypes to terminal drought in a
Mediterranean-type environment. Eur. J. Agron. 11,279-291.
Leport, L., Turner, N.C., French, RJ., Tennant, D., Thomson, B.D., Siddique, K.H.M., 1998.
Water relations, gas exchange and growth of cool-season food legumes in a
Mediterranean-type environment. Eur. 1.Agron. 9, 295-303.
Levitt, J., 1980. Response of Plants to Environmental Stress. Vol. II. Water, Radiation, Salt and
other Stresses. Academic Press, New York, pp. 395-434.
Loague, K., Green, R.E., 1991. Statistical and graphical methods for evaluating solute transport
models: Overview and application. 1. Contam. Hydrol. 7,51-73.
Loomis, R.S., 1983. Crop manipulation for efficient use of water: An overview. In: Taylor,
H.M., Jordan, W.R, and Sinclair, T.R. (Eds.), Limitations to Efficient Water Use in
Crop Production. American Society of Agronomy, Crop Science Society of America,
Soil Science Society of America Inc., Madison, USA, pp. 345-374.
Loomis, R.S., Connor, DJ., 1992. Crop Ecology: Productivity and Management in Agricultural
Systems. Cambridge University Press, Cambridge, UK, pp. 538.
Lopez, F.B., Setter, T.L., McDavid, C.R., 1988. Photosynthesis and water vapour exchange of
pigeonpea leaves in response to water deficit and recovery. Crop Sci. 28, 141-145.
Loss, S.P., Siddique, K.H.M., Dennant, D., 1997. Adaptation offaba bean (Viciafaba
200
L.) to dryland Mediterranean-type environments. ill. Water use and water use
efficiency. Field Crops Res. 54, 153-162.
Ludlow, M. M., Muchow, RC., 1990. A critical evaluation of traits for improving crop yields
in water limited environments. Adv. Agron. 43: 107-153.
Ma, L., Gardner, F.P., Selamat, A., 1992. Estimation of leaf area from leaf and total mass
measurements in peanut. Crop Sei. 32,467-471.
MacRobert, J.F., Savage, M.J., 1998. The use of a crop simulation model for planning wheat
irrigation in Zimbabwe. In: Tsuji, G.Y., Hoogenboom, G., Thornton, P.K. (Eds.),
Understanding Options for Agricultural Production. Kluwer Academic Publishers,
Dordrecht, the Netherlands, pp. 205-220.
Mall, RK., Lal, M., Bhatia, V.S., Rathore, L.S., Singh, R, 2004. Mitigating climate change
impact on soybean productivity in India: A simulation study. Agrie. For. Meteorol. 121,
113-125.
Manschadi, A.M., Sauerborn, L, Stiitzel, H., Gabel, W., Saxena, M.C., 1998a. Simulation of
faba bean (Vieiafaba L.) root system development under Mediterranean conditions.
Eur. J. Agron. 9, 259-272.
Manschadi, A.M., Sauerborn, J., Stiitzel, H., Gabel, W., Saxena, M.C., 1998b. Simulation of
faba bean (Vieiafaba L.) growth and development under Mediterranean conditions:
Model adaptation and evaluation. Eur. J. Agron. 9,273-293.
Markhart, A.H., 1985. Comparative water relation of Phaseolus vulgaris L. and Phaseolus
aetifolius Gray. Plant Physiol. 77, 113-117.
Maroco, J.P., Pereira, J.S., Chaves, M.M., 1997. Stomatal response to leaf-to-air vapour
pressure deficit in Sahelian species. Aust. J. Plant Physiol. 24, 381-387.
Marshall, B., Willey, RW., 1983. Radiation interception and growth in intererop of pearl
millet/groundnut. Field Crops Res. 7, 141-160.
Martin, I., Tenorio, J.L., Ayerbe, L., 1994. Yield, growth, and water use of conventional and
semi-leafless peas in dry environments. Crop Sei. 34, 1576-1583.
Matthews, R.B., Harris, D., Nageswara Rao, RC., Williams, J.H., Wadia, K.D.R., 1988. The
physiological basis for yield differences between four genotypes of groundnut (Aarehis
hypogaea) in response to drought. I. Dry matter production and water use. Exp. Agrie.
24, 191-202.
Matthews, R., Stephens, W., Hess, T., Middleton, T., Graves, A., 2002. Application of crop/soil
simulation models in tropical agricultural systems. Adv. Agron. 76, 31-124.
Mauromicale, G., Cosentino, S., Copani, V., 1988. Validty ofthermal unit summations for
purpose of prediction in Phaseolus vulgaris L. cropped in Mediterranean environment.
Aeta Hort. 229,321-331.
Mbabaliye, T., Wojtkowski, P.A., 1994. Problems and perspectives on the use ofa crop
simulation model in an African research station. Exp. Agrie. 30, 441-446.
McKenzie, B.A., Hill, G.D., 1991. Intercepted radiation and yield oflentils (Lenis eulinaris) in
Canterbury, New Zealand. J. Agrie. Sci. (Camb.) 117,339-346.
MoA (Ministry of Agriculture), 1998. Agroecological zones of Ethiopia. Natural Resources
Management and Regulatory Department and GTZ (German Agency for Technical
Cooperation). Addis Ababa, Ethiopia.
Mohapatra, P.K., Turner, N.C., Siddique, K.H.M., 2000. Assimilate partitioning in chickpea
(Cieer arietinum) in drought-prone environments. In: Saxena, N.P., Johansen, C.,
Chahan, Y.S., Rao, R.C.N. (Eds.), Management of Agricultural Drought; Agronomic
and Genetic Options, Oxford and mH, New Delhi.
201
Monneveux, P., Belhassen, E., 1996. The diversity of drought adaptation in the wide. Plant
Growth Regul. 20, 85-92.
Monteith, J.L., 1965. Light distribution and photosynthesis in field crops. Ann. Bot. 29, 17-37.
Monteith, J.L, 1977a. Climate and efficiency of crop production in Britain. Philos. Trans. R.
Soc. London B. 281,277-294.
Monteith, J.L., 1977b. Climate. In: Alvim, P. de T., Kozlowski, T.T. (Eds.), Ecophysiology of
Tropical Crops. Academic Press, London, pp. 1-27.
Monteith, J.L., 1994. Validity of the correlation between intercepted radiation and biomass.
Agric. For. Meteorol. 68, 213-220.
Monteith, G .L., 1996. The quest for balance in crop modelling. Agron. J. 88, 695-697.
Monteith, J.L., Unsworth, M., 1990. Principles of Environmental Physics, 2nd Edition.
Edward Arnold, London.
Monteith, J.L., Scott, R.K., Unsworth, M.U. (Eds.), 1994. Resource Capture by Crops.
Nottingham University Press, Loughborough, UK, 469 pp.
Morgan, J.M., Rodriguez-Maribona, B., Knights, E.J., 1991. Adaptation to water deficit in
chickpea breeding lines by osmoregulation: Relationship to grain yields in the field.
Field Crops Res. 27, 61-70.
Morison, J.I.L., 1998. Stomatal response to increased CO2 concentration J. Exp. Bot. 49, 443-
452.
Morris, R.A., 1987. Characterizing rainfall distributions at IRRI cropping systems research sites
in the Philippines. In: Impact of Weather Parameters on the Growth and Yield of Rice.
International Rice Research Institute, Los Banos, pp. 189-214.
Morris, R.A., Garrity, D.R., 1993. Resource capture and utilization in intercropping: Water.
Field Crops Res. 34, 303-317.
Mott, K.A., Pankhurst D.F., 1991. Stomatal response to humidity in air and helox. Plant
Cell Environ. 14,509-515.
Muchena, P., Iglesias, A., 1995. Vulnerability of maize yields to climate change in different
farming sectors in Zimbabwe. In: Rozenzweig, C., Allen L.H., Harper, L.A., Hollinger,
S.E., Jones, J.W. (Eds.), Climate Change and Agriculture: Analysis of Potential
International Impacts (ASA Special Publication 59). American Society of Agronomy,
Madison, Wisconsin, pp. 229-239.
Muchow, R.C., 1985a. An analysis of the effect of water-deficits on grain legumes grown in a
semi-arid tropical environment in terms of radiation interception and its efficiency of
use. Field Crops. Res.Il, 309-323.·
Muchow, R.C., 1985b. Stomatal behavior in grain legumes under different soil water regimes in
a semi-arid environment. Field Crops Res. 11,291-307.
Muchow, R.C., Charles-Edwards, D.A., 1982. An analysis ofthe growth ofmung beans at a
range of plant densities in tropical Australia. 1. Dry matter production. Aust. J. Agric.
Res. 33,41-51.
Muchow, R.C., Sinclair, T.R., 1994. Nitrogen response ofleafphotosynthesis and canopy
radiation use efficiency in field-grown maize and sorghum. Crop Sci. 34,721-727.
Muchow, R.C., Robertson, M.J., Pengelly, B.C., 1993. Radiation use efficiency of soybean,
mungbean, and cowpea under different environmental conditions. Field Crops Res. 32,
1-16.
Mukhala, E., 1998. Radiation and Water Utilization Efficiency by Mono-Culture and Intererop
202
to Suit Small Scale Irrigation Farming. Ph.D. Thesis, Department of Agrometeorology,
University of the Orange Free Sate, South Africa, 240 pp.
Munier-Jolian, N.M., Ney, B., Duthion, C., 1996. Termination of seed growth in relation to
nitrogen content of vegetative parts in soybean plants. Eur. J. Agron. 5,219-225.
Nageswara Rao, R.C., Wright, G.C., 1994. Stability of the relationship between specific leaf
area and carbon isotope discrimination across environments in peanut. Crop Sci. 34, 98-
103.
NeSmith, D.S., Ritchie, J.T., 1992. Maize (Zea mays L.) response to a severe water-deficit
during grain filling. Field Crops Res. 29,23-35.
Ney, B., Duthion, C. Turc, 0., 1994. Phenological response of pea to water stress during
reproductive development. Crop Sci. 34, 141-146.
Ninkovic, V., 2003. Volatile accumulation between barley plants affects biomass allocation. J.
Exp. Bot. 54, 1931-1939.
NMSA (National Meteorology Service Agency). 1996. Climatic and Agroclimatic Resources of
Ethiopia. Vol. 1, No. 1. National Meteorology Service Agency of Ethiopia, Addis
Ababa, 137 pp.
Norman, J. M., Arkebauer, T.J., 1991. Predicting light use efficiency from leaf characteristics.
In: Hanks. J., Ritchie, J.T., (Eds.), Modelling Plant and Soil Systems. American Society
of Agronomy, WI, USA, pp. 125-143.
Ntare, B.R., 1992. Variation in reproductive efficiency and yield of cowpea under high
temperature conditions in the Sahelian environment. Eupytica 59, 27-32.
Nwalozie, M.C., Annerose, D.J.M., 1996. Stomatal behavior and water status of cowpea and
peanut at low soil moisture levels. Acta Agron. Hung. 44, 229-236.
Ogallo, L.J., 1988. Relationship between seasonal rainfall in East Africa and the Southern
oscillation. J. Climatol. 8,31-43.
Ogindo, H.O., 2003. Comparing the Precipitation Use Efficiency of Maize-Bean Intereropping
with Sole Cropping in a Semi-Arid Ecotope. Ph.D. Thesis in Agrometeorology,
University of the Free State, South Africa, pp. 36-45.
Ong, C.K., Simonds, L.P, Mathews, R.B., 1987. Response to saturation deficit in a stand of
groundnut (Arachis hypogea L.). 2. Growth and development. Ann. Bot. 59, 121-128.
Ong, C.K., Black, C.R., Marshall, F.M., Corlett, J.E., 1996. Principle of resource capture and
utilization of light and water. In: Ong, C.K., Huxley, P.A. (Eds.), Tree-crop
Interactions: A Physiological Approach. CAB International, Wallingford, UK, pp. 73-
158.
Oram, P.A., Agcaoili, M., 1992. Introduction. In: Muehlbauer, F.J., Kaiser, W.J., (Eds.),
Expanding Production and Use of Cool Season Food Legumes. Kluwer Dordrecht, pp.
3-52.
Pannu, R.K., Singh, D.P., 1993. Effect of irrigation on water use, water use efficiency, growth
and yield ofmungbean. Field Crops Res. 31,87-100.
Parson, L.R., Howe, T.K., 1984. Effect of water stress on the water relation of Phseolus vulgaris
and the drought resistant Phaseolus actifolius. Science 60, 197-202.
Passioura, J.B., 1977. Grain yield, harvest index, and water use of wheat. J. Aust. Inst. Agric.
Sci. 43, 117-120.
Passioura, J.B., 1982. The role of root system characteristics in the drought resistance of crop
plants. In: International Rice Research Institute (IRRI), Drought Resistance in Crops
With Emphasis on Rice, IRRI, Philippines, pp. 71-82.
203
Pastenes, C., Horton, P., 1996. Effect of high temperature on photosynthesis in beans .11.
CO2 assimilation and metabolite contents. Plant Physiol. 112, 1253-1260
Peacock, J.M., Sivakumar, M.V.K., 1986. An environmental physiologist's approach to
screening for drought resistance in sorghum with particular emphasis to Sub-Saharan
Africa. In: Proceedings of an International Drought Symposium, 19-23 May, 1986.
Nairobi, Kenya, pp. 101-102.
Pearcy, R.W., Bjorkman, D., Harrison, AT., Mooney, H.A, 1971. Photosynthetic performance
of desert species with C-4 photosynthesis in death valley, California. Carnegie Inst.
Wash. Year Book. 70, pp. 540.
Penning de Vries, F.W.T, Jansen, D.M., ten Berge, H,F.M., Bakema, A, 1989. Simulation of
Ecophysiological Processes of Growth in Several Annual Crops. Pudoc, Wageningen,
The Netherlands, 271 pp.
Penning de Vries, F.W.T., Rabbinge, R., Jansen, D.M., Bakema, A, 1988. Transfer of systems
analysis and simulation in agriculture to developing countries. Agrie. Admin. Exten. 29,
85-96.
Perry, M.W., 1987. Water use efficiency of non-irrigated field crops. Proceeding of the fourth
Australian Agronomy Conference, La Trobe University, Melbourne, pp. 83.
Phillips, J.G., Cane, M.A., Rosenzweig, C., 1998. ENSO, seasonal rainfall patterns and
simulated maize yield variability in Zimbabwe, Agrie. For. Meteorol. 90, 39-50.
Pilbeam, C.J., Simmonds, L.P., Kavilu, AW., 1995. Transpiration efficiencies of maize and
beans in semi-arid Kenya. Field Crops Res. 41, 179-188.
Poorter, H., Nagel, 0., 2000. The role ofbiomass allocation in the growth response of plants at
different levels of light, CO2, nutrients and water: A quantitative review. Aust. J. Plant
Physiol. 27, 595-607.
Rachie, K.O., 1985. Introduction. In: Singh, S.R. Rachie, K.O. (Eds.), Cowpea Research,
Production and Utilization. Wiley, New York, pp. 205-213.
Raman, C.V.R, 1974. Analysis of commencement ofmonsoon rains over Maharashtra state for
agricultural panning. Scientific report 216, India Meteorological Department, Poona,
India.
Reddy M.S., Kidane, G., 1994. Overview of dryland farming research ofIAR: Objectives and
focus. In: Reddy, M.S. and Kidane G. (Eds.), Development of Technologies for the
Dryland Farming Areas of Ethiopia. Proceeding of the First National Workshop on
Dryland Farming Research in Ethiopia, 26-28 November 1991, Nazareth, Ethiopia.
Ribaut, J.M., Pilet, P.E., 1991. Effect of water stress on growth, osmotic potential and abscisic
acid content of maize roots. Physiol. Plant. 81, 156-162.
Richards, FJ., 1959. A flexible growth function for empirical use. J. Exp. Bot. 10,290-300.
Richards, RA, 1996. Defining criteria to improve yield under drought. Plant Growth Regul.
20, 157-166.
Rinaidi, M., Losavio, N., Flagella, Z., 2003. Evaluation and application of OIL CROP-SUN
model for sunflower in southern Italy. Agrie. Syst. 78, 17-30.
Ritchie, J.T., 1981. Water dynamics in the soil-plant-atmosphere system. Plant Soil 58, 81-96.
Roberts, E.H., Summerfield, RJ., 1987. Measurement and prediction of flowering in annual
crops. In: J.G. Atherton (Ed.), Manipulation of Flowering, London, Butterworths, pp.
17-50.
Robertson, MJ., Silim, S., Chauhan, Y.S., Ranganathan, R, 2001. Predicting growth and
204
development of pigeonpea: Biomass accumulation and partitioning. Field Crops Res.
70,89-100.
Rockstrëm, J., 2001. Green water security for the food makers of tomorrow: Windows of
opportunity in drought-prone savannas. Water Sci. Technol. 43, 71-78.
Rockstrëm, J., Falkenmark, M., 2000. Semi-arid crop production from a hydrological
perspective: Gap between potential and actual yields. Crit. Rev. Plant Sci. 19,319-346.
Rosegrant, M., Cai, x., Cline, S., Nakagawa, N., 2002. The role ofrainfed agriculture in the
future of global food production. Discussion Paper, Environment and Production
Technology Division (EPTD) IFPRI, Washington, D.C., U.S.A.
Royo, C., Blanco, R., 1999. Growth analysis of five spring and five winter triticale genotypes.
Agron. J. 91, 305-311.
Ruiz-Nogueira, B., Boote, K.J., Sau, F., 2001. Calibration and use ofCROPGRO-soybean
model for improving soybean management under rainfed conditions. Agrie. Syst. 68,
151-173.
Russell, G., Jarvis, P.G., Monteith, J. L., 1989. Absorption ofradiation by canopies and
stand growth. In: Russell, G., Marshall, B., and Jarvis, P.G. (Eds.), Plant Canopies:
Their Growth, Form and Function. Cambridge University Press, Cambridge, pp. 21-40.
Ryan, J.G., 1997. A global perspective on pigeonpea and chickpea sustainable production
systems: Present status and future potential. In: Asthana, A.N., Ali, M. (Eds.), Recent
Advances in Pulse Research. Indian Society of Pulse Research and Development, Indian
Institute of Pulse research (IIPR), Kanpur, India, pp. 1-31.
Sadras, V.O., Milroy, S.P., 1996. Soil-water thresholds for the response ofleaf expansion and
gas exchange: A review. Field Crops Res. 47, 253-266.
Saeki, T., 1960. Interrelationships between leaf amount, light distribution and total
photosynthesis in a plant community. Bot. Mag. Tokyo 73, 55-63.
Sage, R.F., Reid, C.D., 1994. Photosynthetic response mechanisms to environmental change in
C3 plants. In: Wilkinson, R.E. (Ed.), Plant-Environment Interactions. Marcel Dekker,
Inc., New York, pp. 413-500.
Saini, H.S., Westgate, M.E., 2000. Reproductive development in grain crops during drought.
Adv. Agron. 68, 59-95.
Sandhu, B.S., Horton, M.L., 1978. Temperature responses of oats to water stress in the field.
Agrie. Meteorol. 19,329-336.
Sandhu, B.S., Prihar, S.S., Khera, K.L., Sandhu, K.S., 1978. Scheduling irrigation to chickpea.
Indian J. Agrie. Sci. 48, 486-492.
Saxena, N.P., 1984. Chickpea. In: Goldsworthy, P.R. and N.M. Fischer (Eds.), The Physiology
of Tropical Crops. Wiley, New York, pp 419-452.
Saxena, M.C., Beniwal, S.P.S., Malhorta, R. S., 1987. Significance of Cool-Season Legumes
and ICARDA's Role in Their Improvement in Sub-Saharan Agriculture. In: ICRISAT
(International Crops Research Institute for The Semi-Arid Tropics). Research on Grain
Legumes in Eastern and Central Africa. Summary Proceedings of the Consultative
Group Meeting for Eastern and Central African Regional Research on Grain Legumes
(Groundnut, Chickpea and Pigeonpea), 8-10 December 1986, International Livestock
Center for Africa (ILCA), Addis Ababa, Ethiopia. Patancheru, India: ICRISAT.
Saxena, M.C., Silim, S.N., Singh, K.B., 1990. Effect of supplementary irrigation during
reproductive growth on winter and spring chickpea (Cieer arietinum) in a
Mediterranean environment J. Agrie. Sci. (Camb.) 114,285-293.
Saxena, N.P., Johansen, C., Saxena, M.C., Silim, S.N., 1993a. Selection for drought and
205
salinity tolerance in cool season food legumes. In: Singh, KB., and Saxena, M.C.,
(Eds.), Breeding for Stress Tolerance in Cool Season Food Legumes. Wiley, UK, pp.
245-270.
Saxena, N.P., Krishnamurthy , L., Johansen, C.,1993b. Registration of drought resistant
chickpea germplasm. Crop Sci. 33, 1424.
Schulze, E. -D., 1994. The regulation of plant transpiration: Interactions offeedforward,
feedback and futile cycles. In: Schulze, E. -D. (Ed.), Flux Control in Biological
Systems: From Enzymes to Populations and Ecosystems. Academic Press, New York,
pp. 203-235.
Schulze, E.-D., Hall, AE., 1982. Stomatal response, water loss, and C02 assimilation rates of
plants in contrasting environments. In: Lange, O.L., Nobel, P.S., Osmond, C.B.,
Zeigler, H. (Eds.), Physiological Plant Ecology. II. Water Relations and Carbon
Assimilation. Encyclopaedia of Plant Physiology New Series, Vol. 12, Springer-
Verlag:Berlin, pp. 181-230.
Sedgley, R.H., Siddique, KH.M., Walton, G.H., 1990. Chickpea ideotypes for Mediterranean
environments. In: H. A van Rheenen and M.C. Saxena (Eds.), Chickpea in the
Nineties, ICRISAT, India, Patancheru, India, pp. 87-91.
Shackel, KA, Hall, AE., 1983. Comparison of water relation and osmotic adjustment in
sorghum and cowpea under field conditions. Aust. J Plant Physiol. 10,423-435.
Sharma, B, Khan, T., 1997. Creating higher genetic yield potential in field pea: Present status
and future potential. In: Asthana, AN., Ali, M. (Eds.), Recent Advances in Pulse
Research, Indian Society of Pulse Research and Development, Indian Institute of Pulse
Research (IIPR), Kanpur, India, pp. 199-215.
Sheng, Q., Hunt, L.A, 1991. Shoot and dry weight and soil water in wheat, triticale and rye.
Can. J Plant Sci. 71,41-49.
Siddique, KH.M., Loss, S.P., Pritchard, D.L. Regan. K.L., Tennant, D., Tettner, R.L.,
Wilkinson, D., 1998. Adaptation of lentil (Lens eulinaris Medik.) to Mediterranean-type
environments: Effect of time of sowing on growth, yield and water use. Aust. J Agrie.
Res. 49, 613-626.
Siddique, K.H.M, Loss, S.P., Regan, KL., Jettner, RL., 1999. Adaptation and seed yield of cool
season grain legumes in Mediterranean environments for south-western Australia. Aust.
J Agrie. Res., 50, 375-387.
Siddique, KH.M., Regan, KL., Tennant, D., Thomson, B.D., 2001. Water use and water use
efficiency of cool season grain legumes in low rainfall Mediterranean-type
environments. Eur. J Agron. 15,267-280.
Siddique, K H.M., Sedgley, RH., 1986. Chickpea (Cicer arietinum L.), a potential grain
legume for South-Western Australia: Seasonal growth and yield. Aust. J Agrie. Res.
37,245-261.
Siddique, K.H.M., Sedgley, RH., 1987. Canopy development modifies the water economy of
chickpea (Cieer arietinum L.) in South-western Australia. Aust. J Agrie. Res. 37, 599-
610.
Silim, S.N., Hebblethwaite, P.D., Heath, M.C., 1985. Comparison of the effect of autumn and
spring sowing date on growth and yield of climbing peas. J Agrie. Sci. (Camb.) 104,
35-46.
Silim, S.N., Saxena, M.C., 1993a. Adaptation of spring-sown chickpea to the Mediterranean
basin. I.Response to moisture supply. Field Crops Res. 34, 121-136.
Silim, S.N., Saxena, M.C., 1993b. Adaptation of spring-sown chickpea to the Mediterranean
basin. II. Factors influencing yield under drought. Field Crops Res. 34, 137-146.
206
Silim, S.N., Saxena, M.C., Erskine, W., 1993a. Adaptation of lentil to the Mediterranean
environment. I. Factors affecting yield under drought conditions. Exp. Agrie. 29, 9-19.
Silim, S.N., Saxena, M.C., Erskine, W., 1993b. Adaptation oflentil to the Mediterranean
environment. II. Response to moisture supply. Exp. Agrie. 29, 21-28.
Simane, B., 1990. Agroclimatic analysis to assess crop production potentials in Ethiopia.
Department of Plant Sciences, Alemaya University of Agriculture, Ethiopia, and
International Crops Research Institute for the Semi-Arid Tropics (ICRISAT),
Patancheru, India.
Simane, B.1993. Drought resistance in durum wheat. PhD Thesis, Wageningen, The
Netherlands, pp. 63-79.
Simane, B., Struik, P.C. 1993. Agroclimatic analysis: a tool for planning sustainable wheat
(Triticum turgidum var. durm) production in Ethiopia. Agrie. Eeo. Env. 47, 31-46.
Simane, B., Struik, P.C., Nachit, M.M., Peacock, J.M., 1993. Ontogenetic analysis of yield
components and yield stability of drum wheat in water-limited environments. Euphytiea
71,211-219.
Simane, B., Tanner, D.J., Amsal T., Asefa T., 1999. Agro-ecological decision support systems
for wheat improvement in Ethiopia: Climatic characterization and clustering of wheat
growing regions. Afr. Crop Sci. J. 7, 9-19.
Simane, B., Worthman, C., Hoogenboom, G., 1998. Haricot bean agroecology in Ethiopia:
Definition using agroclimatic and crop growth simulation models. Afr. Crop Sci. J. 6, 9-
18.
Simpson, G.N., 1981. Water Stress on Plants. Praeger Publishers, New York, pp. 89-139.
Sinclair, T.R., Ludlow, M.M., 1986. Influence of soil water supply on plant water balance of
four tropical grain legumes. Aust. J. Plant Physiol. 13, 9-341.
Sinclair, T.R., Muchow, R.C., 1999. Radiation use efficiency. Adv. Agron. 65, 215-265
Sinclair, T.R., Seligman, N.A.G., 1996. Crop modelling: From infancy to maturity. Agron. J. 88,
698-703.
Sinclair, T.R., Tanner C.B., Bennet, J.M., 1984. Water use efficiency in crop production.
Bioscience 34, 36-40.
Singh, D.P., 1997b. Tailoring the plant type in pulse crops. Plant Breed. Abstr. 67, 1213-1220.
Singh, G., Bhushan, L.S., 1980. Water use, water-use efftciency, and yield of dryland chickpea
as influenced by P fertilization, stored soil water and crop season rainfall. Agrie. Water
A1anage.2,299-305.
Singh, KB., 1993. Problems and prospects of stress resistance breeding in chickpea. In: Singh,
KB., Saxena, M.C. (Eds.), Breeding for Stress Tolerance in Cool-Season Food
Legumes, Wiley, Chichester, pp. 17-35.
Singh, KB. 1997a. Chickpea (Cieer arietinum L.). Field Crops Res. 53, 161-170.
Singh, P., 1991. Influence of water deficit on phenology, growth and dry matter allocation in
chickpea (Cieer arietinum). Field Crops Res. 28, 1-15.
Singh, P., Boote, KJ., Rao, Y.A, Iruthayaraj, M.R., Sheikh, A.M., Hundal, S.S., Narang, R.S.,
1994. Evaluation of the groundnut model PNUTGRO for crop water response to water
vailability, sowing dates, and seasons. Field Crops Res., 39, 147-162.
Singh, P., Saxena, N.P., Monteith, J.L., Huda, A.KS., 1990. Defining, modeling, and managing
water requirement of chickpea. In: ICRISAT (International Crops Research Institute for
the Semi-Arid Tropics), Chickpea in the Nineties: Proceedings of the Second
207
International Workshop on Chickpea Improvement, 4-8 Dec 1989. ICRISAT Center,
India, Patancheru, India.
Singh, P., Sri Rama, Y.V., 1989. Influence of water deficit on transpiration and radiation use
efficiency of chickpea (Cieer arietinum L.). Agrie. For. Meteorol. 48, 317-330.
Singh, P., Virmani, S.M., 1990. Evaporation and yield of irrigated chickpea. Agrie. For.
Meteorol. 52, 333-345.
Singh, S.P., White, J.W., 1988. Breeding common beans for adaptation to drought conditions.
In: White, J.W., Hoogenboom, F., Ibarra, F., and Singh, SP. (Eds.), Research on
Drought Tolerance in Common Bean. Bean Program, Cali, Colombia, pp. 261-285.
Singh, S.R. 1987. lITA's grain legume improvement program in relation to eastern and central
Africa. In: ICRISAT (International Crops Research Institute for the Semi-Arid Tropics).
Research on Grain Legumes in Eastern And Central Africa. Summary Proceedings of
the Consultative Group Meeting for Eastern and Central African Regional Research on
Grain Legumes (Groundnut, Chickpea and Pigeonpea), 8-10 December 1986,
International Livestock Center for Africa (ILCA), Addis Ababa, Ethiopia. Patancheru,
India: ICRISAT.
Singh, U., Thornton, P.K, Saka, A.R., Dent, J.B., 1993. Maize modeling in Malawi: A tool for
soil fertility research and development. In: Penning de Vries, F.W.T., et al. (Eds.),
Systems Approaches for Agricultural Development. Kluwer Academic Publishers,
Dordrecht, The Netherlands, pp. 253-273.
Sinoit, B.K, Kramer, P.J., 1977. Effect of water stress during different stages of growth of
soybean. Agron. J. 69,274-278.
Sivakumar, M.V.K, 1986. Canopy-air temperature differentials, water use and yield of chickpea
in a semiarid environment. Irrg. Sci. 7, 149-158.
Sivakumar, M.V.K, 1988. Predicting rainy season potential from the onset of rains in southern
Sahelian and Sudanian climatic zones of West Africa. Agrie. For. Meterol. 42, 295-305.
Sivakumar, M.V.K, 1991. Agrometeorology research at the ICRISAT Sahelian centre. In:
Influence of the Climate on the Production of Tropical Crops. Burkina Faso, Sept.,
1991, International Foundation for Science, Stockholm, Sweden, pp. 146-171.
Sivakumar, M.V.K, 1992. Empirical analysis of dry spells for agricultural application in West
Africa. J. Climate 5,532-539.
Sivakumar, M.V.K, and Virmani, S.M., 1984. Crop productivity in relation to interception of
photosynthetically active radiation. Agrie. For. Meteorol. 31,131-141.
Sivakumar, M.V.K, Singh, P., 1987. Response of chickpea cultivars to water stress in semi-arid
environment. Exp. Agrie. 23, 53-61.
Slatyer, R.O., 1969. Physiological significance of internal water relations to crops yield. In: J.D.
Eastin, F.A. Haskins, C.Y. Sullivan and C.H.M. van Bavel (Eds.), Physiological
Aspects of Crop Yield. ASA, Madison, USA, pp. 53-83.
Solane, R.J., Patterson, R.P. Carter Jr, T.E., 1990. Field drought tolerance of soybean plant
introduction. Crop Sci., 30, 118-123.
Solh, M.B., 1993. New approaches to breeding for stress-environments- Discussion. Int. Crop.
Sci. 1,579-581. .
Sperry, J.S., 2000. Hydraulic constraints on plant gas exchange. Agrie. For. Meteorol. 104, 13-
33.
Squire, G. R., 1990. The Physiology of Tropical Crop Production. CAB International,
Willingford, UK, 236 pp.
208
Stanhill, G., 1986. Water use efficiency. Adv. Agron. 39, 53-85.
Srivasta, A, Strasser, RJ., 1996. Stress and stress management ofland plants during a regular
day. J. Plant Physiol. 148,445-455.
Stem, RD. and Knock, J. 1998. INSTAT climatic guide. Statistical Service Centre, University
of Reading, UK.
Stem, RD., Coe, R., 1982. The use of rainfall models in agricultural planning. Agrie. Meteorol.
26,35-50.
Stem, RD., Dennet, M.D., Dale, l.C., 1982a. Analysing daily rainfall measurements to give
agronomically useful results. I. Direct methods. Exp. Agrie. 18,223-236.
Stem, RD., Dennet, M.D., Dale, l.C. 1982b. Analysing daily rainfall measurements to give
agronomically useful results. II. A modelling approach. Exp. Agrie. 18,237-253.
Stewart, JJ., Hash, C.T., 1982. Impact of weather analysis on agricultural production and for the
semi-arid areas of Kenya. J. App. Meteorol. 21, 479-495.
Subbarao, G.V., Johansen, C., Slinkard, AE., Rao, R.C.N., Saxena, N.P., Chauhan, Y.S., 1995.
Strategies for improving drought resistance in grain legumes. Crit. Rev. Plant Sci. 14,
469-523.
Summerfield, RJ., Minchin, F.R, Roberts, E.H., Hadley, P., 1981. Adaptation to contrasting
aerial environments in chickpea (Cieer arietinum L.) Trop. Agrie. (Trinidad) 58, 97-
113.
Summerfield, RJ., Roberts, E.H., Ellis, R.H., 1991. Towards the reliable prediction of time to
flowering in six annual crops. I. The development of simple models for fluctuating field
environments. Exp. Agrie. 27,11-31.
Szeicz, G., 1974. Solar radiation in crop canopies. J. App. Eeol. 11, 1117-1156.
Taddesse, T., 2000. Drought and its predictability in Ethiopia In: Wilhite, D.A (Ed.), Drought:
A Global Assessment, Vol.I. Routledge, pp. 135-142.
Tanner, C.B., Sinclair, T.R., 1983. Efficient water use in crop production: Research or re-
research? In: Taylor, H.M., Jordan, W.R, and Sinclair, T.R. (Eds.), Limitations to
Efficient Water Use in Crop Production. American Society of Agronomy, Crop Science
Society of America, Soil Science Society of America Inc., Madison, USA, pp. 1-27.
Tedeschi, P. Zerbi, G., 1984. Plant development, flowering course and yield of kidney bean
plants, Phaseolus vulgaris L. grown in lysimeters with relation to different water
regimes. Riv. Ortoflorofrutt. It. 68, 1-10.
Terauchi, T. Matsuoka, M., Nakagawa, H., Nakano, H., 2001. A breeding index for sugarcane.
JIRCAS Research Highlights, pp. 40-41.
Tesfaye, K.F., 1997. The Response of Haricot Bean (Phaseolus vulagaris L.) to Water Deficit
Under Different Stages of Growth. M.Sc. Thesis, Alemaya University of Agriculture,
Alemaya, 90 pp.
Tewolde, H., Dobrenz, K., Voight, RL., 1993. Seasonal trends in leaf photosynthesis and
stomatal conductance of drought stressed pearl millet is associated to vapour pressure
deficit. Photosynthesis Res. 38,41-49.
Thomson, B.D, Siddique K.H.M., 1997. Grain legumes in low rainfall Mediterranean-type
environments II. Canopy development, radiation interception, and dry matter
production. Field Crops Res. 54, 189-199.
Thomson, B. D, Siddique, K. H. M., Barr, M. D., Wilson, J. M., 1997. Grain legume species in
low rainfall Mediterranean-type environments I. Phenology and seed yield. Field Crops
Re. 54, 173-187.
209
Thornton, P .K., Hoogenboom, G., 1994. A computer program to analyse single-season crop
model outputs. Agron. J., 86, 860-868.
Thornton, P.K, Hoogenboom, G., Wilkens, P.W., Jones, J.W., 1994. Seasonal analysis. In:
Tsuji, G.Y., Uehara, G., Balas, S., (Eds.), DSSAT version 3, Vol. 3. University of
Hawaii, Honolulu, Hawai.
Thornton, P.K., Saka, AR., Singh, u.,Kumwenda, J.D.T., Brink, J.E., Dent, J.B., 1995.
Application of a maize crop simulation model in the central region of Malawi. Exp.
Agrie. 31, 213-226.
Thornton, P.K., Bowen, W.T., Ravelo, AC., Wilkens, P.W., Farmer, G., Broek, J., Brink, J.E.,
1997. Estimating millet production for famine early warning: An application of crop
simulation modelling using satellite and ground-based data in Burkina Faso. Agrie. For.
Meteorol. 83,95-112.
Torrecillas, A, Alarcon, J.J., Domingo, R., Planes, J., Sanchez-Blanco, M.J., 1996. Strategies
for drought resistance in leaves of two almond cultivars. Plant Sci. 118, 135-143.
Trejo, C.L., Davies, W.J., 1991. Drought-induced closure of Phaseolus vulgaris L. stomata
precedes leaf water deficit and any increase in xylem ABA concentration. J. Exp. Bot.
42, 1507-1515.
Troll, C., 1965. Seasonal climate of the earth. In: Rodenwaldt, E. and Jusatz, H. (Eds.), World
Map of Climatology, Berlin, Springer-Verlag, 28 pp.
Tsubo, M., 2000. Radiation interception and use in maize and bean intereropping system. Ph.D.
Thesis, University ofthe Orange Free State, South Africa, 143 pp.
Tsubo, M., Mukhala, E., Ogindo, H.O, and Walker, S., 2003. Productivity of maize-bean
intereropping in a semi-arid region of South Africa. Water SA 29, 381-388.
Tsuji, G.Y., Uehara, G., Balas, S., (Eds.), 1994. DSSAT version 3. University of Hawaii,
Honolulu, Hawai.
Turc, 0., 1995. Drought stress accelerates seed growth and development in pea. In: Proceedings
of the 2nd European Conference on Grain Legumes, Copenhagen, pp. 106-107.
Turk, K.J., Hall, AE., 1980a. Drought adaptation of cowpea. ill. Influence of drought on plant
growth and relations with seed yield. Agron. J. 72,413-420.
Turk, K.J., Hall, AE., 1980b. Drought adaptation of cowpea. IV. Influence of drought on water
use and relations with seed yield. Agron. J. 72,421-427.
Turk, K.J., Hall, AE., Asbell, C.W., 1980a. Drought adaptation of cowpea. I. Influence of
drought on seed yield. Agron. J. 72, 428-434.
Turk, K.J., Hall, AE., Asbell, C.W., 1980b. Drought adaptation of cowpea. ll. Influence of
drought on plant water status and relations with seed yield. Agron. J. 72,421-427.
Turner, N.C., 1986. Crop water deficits: A decade of progress. Adv. Agron. 39, 1-51.
Turner, N.C., 1991. Measurement and influence of environmental and plant factors on stomatal
conductance in the field Agrie. For. Meteorol. 54, 137-154.
Turner, N. C., 1997. Further progress in crop water relations. Adv. Agron. 58,292-339.
Turner, N. C., 2000. Drought resistance: A comparison of two frameworks. In: Saxena, N.P.
Johansen, C., Chauhan, Y.S., Rao, R.C.N. (Eds.), Management of Agricultural Drought:
Agronomic and Genetic Options. Oxford and mH, New Delhi.
Turner, N.C., Henson, I.E., 1989. Comparative water relations and gas exchange of wheat and
lupins in the field. In: Kreeb, K.H., Ritcher, H., Hinckley, T.M. (Eds.), Structural and
Functional Responses to Environmental stresses. SPB Academic Publishing, The
Hague, pp. 293-304.
210
Turner, N.C., Wright, G.C., Siddique, K.H.M., 2001. Adaptation of grain legumes (pulses) to
water-limited environments. Adv. Agron. 71, 193-231.
Tsuji, G.Y., Uehara, G., Balas, S., (Eds.), 1994. Decision Support System for Agrotechnology
Transfer (DSSAT) version 3. University of Hawaii, Honolulu, Hawai.
Van den Boogaard, R Alewijnse, D., Veneklaas, E.I., Lambers, H., 1997. Growth and water use
efficiency of 10 Triticum estivum cultivars at different water availability in relation to
allocation ofbiomass. Plant Cell Environ. 20, 200-210.
Van der Maesen, L.J.G., 1972. Cieer L., a Monograph ofthe Genus, with Special Reference to
the Chickpea (Cieer arietinum L.), lts Ecology and Cultivation. Mededelingen Land
bouwhoge School Wagenungen, The Netherlands, pp. 342.
Vasquez-Tello, A, Zuily-Fodil, Y., Pham Thi, AT., Vieira da Silva, J.B. 1990. Electrolyte and
PI leakages and solute sugar content as physiological tests fore screening resistance to
water stress in Phasolus and Vigna species. J. Exp. Bot. 41, 827-832.
Venus, J.C. Causton, D.R., 1979. Plant growth analysis: the use of the Richards function as an
alternative to polynomial exponentials. Ann. Bot. 43, 623-632.
Virmani, S.M., 1975. The agricultural climate of the Hyderabad region in relation to crop
planning (a simple analysis). International Crops Research Institute for the Semi-Arid
Tropics (ICRISAT), Patancheru, India.
Virmani, S.M., Sivakumar, M.V.K., Reddy, S.J., 1980. Climatological features of the semi-arid
tropics in relation to the farming system research program. In: Proceedings,
International Workshop on the Agroclimatological Research Needs of the Semi-Arid
Tropics, Nov., 1978. ICRISAT, India, pp. 22-24.
Virmani, S.M., Sivakumar, M.V.K., Reddy, S.J., 1982. Rainfall probability estimates for
selected locations in semi-arid India. Research Bulletin NO.l (2nd enlarged edition).
International Crops Research Institute for the Semi-Arid Tropics (ICRISAT),
Patancheru, India, 170 pp.
Von Caemmerer, S., Farquhar, G.D., 1981. Some relationship between the biochemistry of
photosynthesis and the gas exchange ofleaves. Planta 153,376-387.
Vos, de R.N., Mallett, J.B., 1987. Preliminary evaluation of two maize (Zea mays L.) growth-
.simulation models. South AfricanJ. Plant SoiI4(3), 131-136.
Vu, J.C.V., Baker, J.T., Pennanen, AH., Allen, L.H., Bowes, G., Boote, K.J., 1998. Elevated
CO2 and water deficit effects on photosynthesis, ribulose bisphosphate Carboxylase-
oxygnase, and carbohydrate metabolism in rice. Physiol. Plant 103, 327-339.
Wafula, B.M., 1995. Application of crop simulation in agricultural extension and research in
Kenya. Agrie. Syst. 49, 399-412.
Wang, F., Fraisse, C.W., Kitchen, N.R., Sudduth, K.A., 2003. Site-specific evaluation of the
CROPGRO-soybean model on Missouri claypan soils. Agrie. Syst. 76, 985-1005.
Watiki, J.M., Fukai, S., Banda, J.A, and Keating, B.A, 1993. Radiation interception and growth
of maize/cowpea intererop as affected by maize plant density and cowpea cultivar. Field
Crops Res. 35, 123-133.
Watson, RT., Zinyowera, M.C., Moss, RH., Dokken, D.J. (Eds.), 1998. The Regional Impacts
of Climate Change. An Assessment of Vulnerability. A Special Report of !PCC
Working Group II. Published for the Intergovernmental Panel on Climate Change.
Cambridge University Press, Cambridge, 517 pp.
Wery, J. Turc, O. Lecoeur, J., 1993. Mechanisms of resistance to cold, heat, and drought in
211
cool-season legumes, with special reference to chickpea and pea. In: K.B. Singh, MC.
Saxena (Eds.), Breeding for Stress Tolerance in Cool-Season Food Legumes. Wiley,
Chichester, pp. 271-291.
Westgate, M.E., Schussler, J.R., Reicosky, D.C., Brenner, M.L., 1989. Effect of water deficits
on seed development in soybeans. II. Conservation of seed growth rate. Plant Physiol.
91,980-985.
White, J.W., Hoogenboom, G., Jones, J.W., Boote, KJ., 1995. Evaluation of the dry bean
model BEANGRO V1.01 for crop production research in a tropical environment. Exp.
Agric. 31, 241-254.
Whitemore, J.S., 2000. Drought Management on Farmland. Kluwer Academic Publishers, pp. 1-
9.
Wien, H.C., Littleton, E.J., Ayanaba, A., 1979. Drought stress of cowpea and soybean under
tropical conditions. In: MusseU, H., Staples, R.C., (Eds.), Stress Physiology in Crop
.Plants. John Wiley & Sons, N.Y., pp. 284-301.
Wilhelm, W.W., McMaster, G.S., 1995. The importance of the phyllochron in studying the
development of grasses. Crop Sci. 35,1-3.
Wilhite, D.A., Glantz, M.H., 1985. Understanding the drought phenomenon: The role of
definitions. Water. Int. 10, 111-120.
Williams, J.H., 2000. The implication and application of resource capture concepts to
crop improvement by plant breeding. Agric. For. Meteorol. 104,49-58.
Williams, J.H., Nageswara Rao, R.C., Dougbedji, F., Talwar, H.S., 1996. Radiation interception
and modelling as an alternative to destructive samplings in crop growth measurements.
Ann. Appl. Bioi. 129, 151-160.
Williams, J., Ross, P.J., Bristow, K.L., 1992. Prediction of the Campbell water retention
function from texture, structure and organic matter. In: van Genuchten, M.T., Leij, F.J.,
and Lund, L.J. (Eds.), Proc. Int. Workshop on Indirect Methods for Estimating the
Hydraulic Properties of Unsaturated Soils. University of California, Riverside, CA, pp.
427-441.
Williams, J.H., Saxena, N.P., 1991. The use of non-destructive measurement and physiological
models of yield determination to investigate factors determining differences in seed
yield between genotypes of "desi" chickpeas (Cicer arietum). Ann. Appl. Bioi. 119, 105-
112.
Willmott, C.J., 1982. Some comments on the evaluation of model performance. Bull. Amer.
Meteorol. Soc. 63, 1309-1313.
WMO (World Meteorological Organization), 1966. International Meteorological Vocabulary.
WMO-No 182, TP.91. WMO, Geneva.
Wright, G.C., Hubick, K.T., Farquhar, G.D., 1991. Physiological analysis of peanut cultivar
response to timing and duration of drought stress. Aust. J. Agric. Res. 42, 453-470.
Wright, G.C., Nageswara Rao, R.C., Farquhar, G.D., 1994. Water use efficiency and carbon
isotope discrimination in peanut under water deficit conditions. Crop Sci. 34, 92-97.
Zhang, J., Davies, W.J., 1987. Increased synthesis of ABA in partially dehydrated root tips and
ABA transport from root to leaves. J. Exp. Bot. 38, 2015-2023.
212
Appendix 1
Appendix lA. Soil type and soil water relations of the study areas (Eylachew, 1994 and
sample analysis)
Location Soil Type ST DUL DLL Available es
(mm m") (mm m") (mm m") (mm m")
FAOfUNESCO USDA
Alemaya Eutric Regosol Typic Ustorhent C 88 55 33 137
Campus
Awassa Regosol NA CL 108 52 56 150
(Profile-I )
Bahir Dar Ferric Luvisol Typic Rhodustalf C 132 84 48 150
Bako Humic Ferralosol Rhodic Haplustox C 123 93 30 162
Bole/Akaki Vertisol Vertisol C 157 92 65 185
Debre Zeit Vertisol Vertisol C 139 60 79 180
Dire Dawa Eutric Regosol Typic Ustorhent SL 78 39 38 46
Jijiga NA NA CL 90 63 27 103
Mekeie Calcic Cambisol Typic Eutrochrept L 94 64 30 143
Melkassa
(Nazareth) Haplic Andosol Typic Haplustand SL 72 38 34 168
DUL= drained upper limit; DLL=drained lower limit; C= clay; CL= clay loam; L= loam; SL= sandy
loam; ST= surface soil texture; NA= information not available.
180
160 -___--PETo
E 140 ____ 50%ETo
E 120 _-+-_35%ETo
0
tu 100
"C
C 80
IV
.i.i. 60
C
IV 40
0::
20
0
1 6 11 16 21 26 31 36
Decade
Appendix lB. Comparison of long-term decade rainfall (P) and reference
evapotranspiration (ETo) at 100, 50 and 35% levels at Bahir Dar, Ethiopia.
213
100 100 100
Ale maya _-_-P ETa Awassa
eo ••••••• 5O%ETa eo eo
Ê -+--35%ETa
.§. eo eo eo
~...
i 40 40 40
!c:
;ï! 20 20 20
0
6 11 16, 21 26 31 36 0 0
6 11 16 21 26 31 36 6 11 16 21 26 31 36
100
100 100
eo Debr. ZOR [ire Daw.
Ê eo eo
.§. eo
~ eo eo...
i 40
40 40
~
;ï! 20 20 20
0
6 11 16 21 26 31 36 0 0
6 11 16 21 26 31 36 6 11 16 21 26 31 36
100
Jlpg. 100 100 Melkassa
eo Mekele
eo eo
Ê
.§. eo
~ eo eo...
i 40 40 40
~
~ 20 20 20
0 0 0
6 11 16 21 26 31 36 6 11 16 21 26 31 36 6 11 16 21 26 31 36
Decade Decade Decade
Appendix IC. Comparison of long-term decade rainfall (P) and reference
evapotranspiration (ETo) at 100,50 and 35% levels for nine locations in Ethiopia.
214
100 100
80 80
i! 60 60
~
j 40 ---b--1O 40
IJ:. ----*"-15
20 _20 20
o O~-.-,--~~~~~~~~~-r--~
o 30 60 00 120 150 180 210 240 ZTO 300 330 360 0 30 60 00 120 150 180 210 240 ZTO 300 330 360
80 80
i! 60 60
I40 40
20 20
o O~~~ __~~~~~~~~~ __~~~
o 30 60 00 120 150 180 210 240 ZTO 300 330 360 0 30 60 00 120 150 180 210 240 270 300 330 360
80 80
i! 60 ~ 60
~ ~
I40 j 40IJ:.
20 20
o 30 60 00 120 150 180 210 240 ZTO 300 330 360 o 30 60 00 120 150 180 210 240 ZTO 300 330 360
80 80
60
40
20
o O~-r~--~-r--~~~~~~-'--~~
o 30 60 00 120 150 180 210 240 ZTO 300 330 360 0 30 60 00 120 150 180 210 240 ZTO 300 330 360
COY OOy
Appendix ID. Probabilities of conditional dry spells exceeding 5, 7, 10, 15 and 20 days
within 30 days after starting date at 10 locations in Ethiopia.
215
Appendix 2
Appendix 2A. Meteorological, micrometeorological and physiological instruments used in
the study. From left to right are: leaf area meter, pressure chamber, porometer, Infrared
Gas Analyzer (LAC4), Time Domain Retlectometry (TDR) and Sun scan canopy analysis
system at the back (top), and automatic weather station (bottom).
216
Appendix 2B. Plot layout and an example of soil and micrometeorological data collection
during the experimental periods on fields of beans, chickpea and cowpea.
217
Appendix3
Appendix 3A. Thermal time from planting to emergence (E), from emergence to
flowering (E-F), from flowering to podding (F-P), from podding to maturity (P-M) and
from flowering to maturity (F-M) for three grain legumes grown under well-watered (C)
and mid-season (MS) and late season (LS) water stresses in two seasons.
Spa WR 2001/2002 2002/2003
E E-F F-P P-M F-M E E-F F-P P-M F-M
Beans C 113 581 48 399 457 101 650 23 547 570
MS 118 559 81 224 305 101 792 28 533 561
LS 113 577 55 189 245 lOl 635 39 338 376
Chickpea C 131 553 146 466 613 132 579 134 725 859
MS 121 554 160 271 431 132 579 121 754 875
LS 126 554 141 275 416 132 579 134 451 585
Cowpea C 86 758 41 402 442 76 792 59 598 657
MS 81 749 64 252 270 76 778 63 510 572
LS 75 741 69 210 279 76 764 67 381 447
LSD WR n.s n.s. n.s. 93.3** 72.1** n.s n.s. n.s. 46.1*** 62.7***
(P<0.05) SP 13.4*** 38.7*** 36.5*** 48.5** 70.7*** 0.03*** 35.7*** 34.5*** 60.3*** 57.8***
WRxSP n.s. n.s. n.s. n.s n.s. n.s. n.s. n.s. n.s. n.s.
CV(%) 12.1 6.0 39.7 16.0 10.3 0.00 5.2 25.3 10.9 9.2
***, **, *:Treatment significanatt0.1,I and 5% probabilitylevelrespectivelyn,.s:treatmentnot significanatt
5% probabilitlyevel.•Sp= species,W R = waterregimes.
Appendix 3B. Thermal time from planting to emergence (E), from emergence to
flowering (E-F), from flowering to podding (F-P), from podding to maturity (P-M) and
from flowering to maturity (F-M) for three grain legumes grown under well-watered (C)
and mid-season (MS) and late season (LS) water stress in 2002°
Sp wR E-F F-P P-M F-M
Beans C 575 83 696 779
MS 575 83 549 632
LS 575 83 422 505
Chickpea C 594 92 836 928
MS 594 92 358 450
LS 594 92 561 653
Cowpea C 700 132 778 910
MS 700 132 576 708
LS 700 132 484 616
measurements were not replicated.
218
Appendix 3C. The time course of mean leaf area (cm' mo2) expansion in three grain
legumes under three water regimes in 200112002.
Species DAP Water regimes with standard errors (SE)
C MS LS
Mean SE Mean SE Mean SE
Beans 39 6864 230 8486 1767 7696 1127
49 10356 1202 8342 1363 8833 735
57 20098 1141 9415 3984 15087 189
68 19764 903 11356 888 14584 1098
78 13237 1640 9670 2880 5429 2337
88 11527 1965 12396 686
98 11854 1111 23802 2103
Chickpea 38 4977 970 3071 498 4832 481
48 5725 926 4186 513 5305 567
56 7159 2540 8101 1780 8235 3243
67 12131 1912 11675 1543 10128 2881
77 9944 3418 8477 2588 15408 8623
87 3366 2147 12049 0 30312 0
97 2137 2137 11437
Cowpea 35 8119 863 8081 1698 8325 132
45 9741 2185 9212 415 8930 165
55 17693 3064 18855 3930 18737 120
66 22446 5746 15008 578 20051 826
76 16736 1906 10826 1371 8916 1426
86 15674 3285 1608 868
96 14517 3821
Appendix 3D. The time course of mean leaf area (cm" mo2) expansion in three grain
legumes under three water regimes in 2002.
Species DAP Water regimes with standard errors (SE)
C MS LS
Mean SE Mean SE Mean SE
Beans 17 3800 20 2765 19 2196 21
27 11830 136 10725 132 10444 126
37 33938 765 29537 651 32337 375
47 47486 894 46051 1102 50694 692
57 57806 587 23688 1107 59682 1105
67 48391 462 5507 1251 4194 1169
77 18186 144 12640 112
87 14710 136
Chickpea 17 1571 114 1705 112 2274 114
27 5294 131 6177 126 6973 130
37 18689 807 20052 798 19130 710
47 34618 732 27541 1125 38548 1142
57 46875 1213 19815 1259 49188 1185
67 39882 409 8028 1179 23641 1214
77 38615 175
87 26370 181
Cowpea 17 3057 32 3189 21 2529 28
27 6973 127 7925 124 6803 132
37 22359 1108 18291 1175 21298 1169
47 36044 1238 37130 1201 40487 1192
57 48902 2100 39909 1437 56090 1289
67 34857 877 17986 1103 10768 1104
77 18015 256 12909 709
87 11410 343
219
Appendix 3E. The time course of mean leaf area (cm" m-2) expansion in three grain
legumes under three water regimes in 2002/2003.
Species DAP Water regimes with standard errors (SE)
C MS LS
C SE MS SE LS SE
Beans 20 2264 131 2220 44 2387 61
30 6623 436 6138 517 6782 551
40 19085 506 17938 717 18071 593
50 29947 2181 23518 1801 27265 833
60 39350 4272 22304 3894 36612 2476
70 41016 1981 21634 1304 28108 1688
80 33756 719 23265 1674 16496 2942
90 16655 736 11804 840
100 12638 340 12305 1215
Chickpea 20 2199 86 2159 146 2159 89
30 6255 278 5727 156 5870 197
40 14827 607 16599 1315 14802 2580
50 18129 1379 21164 2065 22276 2260
60 29759 1111 15268 3314 25703 619
70 27460 2682 21385 4213 20889 2996
80 25691 2915 17438 843 16732 803
I 90 19862 2140 12813 74
l 100 17338 1734 13524 79" Cowpea 20 2283 78 2307 24 2288 430 5976 58 5998 67 6371 28
40 11850 655 12882 288 13828 465
50 25888 1710 25287 2183 22296 1138
60 32062 1087 30736 2320 33285 387
70 36508 3387 20147 1334 26227 1084
80 28346 2100 18199 2844 20009 2133
90 16207 882 12291 942
100 17058 2379 13999
220
4,- ---,4.,---------------. 4,- -,
BN-C BN-MS BN-LS .A
! 2 ~~ ••.A
tU • ' 2 I; 2
~ ... 6
o • '
1; 0 !:,.__ o I !" _~_._i.t·/Q·-.oI ..... A!~" ~:..~!~. II I...... :I . I I o t----rU I .::.~ I I
:.2;: • __.... AR·leaf •
-2 ... 0- .. AR·stem "., -2 -2
••• I!J.••• AR·pod "
4~ '. ~ -4 .....__--------------' 4L- ~
o 300 600 900 1200 1500 1800 o 300 600 900 1200 1500 1800 o 300 600 900 1200 1500 1800
4 ,- , 4,- --, 4 ,- -,
CHP-C CHP-MS CHP-LS
.2 2 2
~ .A-
2
••A
~e 0 I I ~::;e... !!,,:e;;;~:,::é:;:8 I o I I e::;8:"!~;·t:_: _·0 I O-f-r ~::;~:"r--= i-~·:0' I
i '", I i I -; :, I
U
o
~ -2 •• '. -2 -. -24~ ~ 4~ ~ ~ 4~ ~
o 300 600 900 1200 1500 1800 0 300 600 900 1200 1500 1800 0 300 600 900 1200 1500 1800
4.0 -r+- ~ 4 ,- -,
4 ~----------_.
COP-C A COP-MS
COP-LS
j .' 2.0 2~tU 2 .-~ , I •• A-0 8" '","0, "I' •.,','. 0.0 I 0·::8 ... !_ï!_.,:"-':G:-". .e. I 0+-_,... ~:;:~:.- :;i..-'e I I ,::.0~:.:.•.1:1._lo{
~ i i i i •• , '.' li. " I
oU ~~•• .. "0
~ -2 -..'e -2,0 -2
4.0 ~ 4~ ~ ~
4 o 300 600 900 1200 1500• 1800 o 300 600 900 1200 1500 1800 o 300 600 900 1200 1500 1800
Thermal time (OCd)from planting Thermal time (OCd)from planting Thermal time (OCd)from planting
Appendix 3F. The time course of calculated allocation ratios (AR) of leaf, stem and pod in beans, chickpea and cowpea under stress and non-stress
conditions in 2002. Thermal time to flowering was 651, 790, and 867 "Cd for BN, CHP and COP respectively.
221
2.0 I 2.0 T -------------------~,-LS. II
....... AR-Ieaf BN-MS I BN-LS
... 0. .. AR-stem 1.5 _:.'": 1.5 ~
... 6... AR-pod
-:; 1.0 •
I '1.0 ••
.
ti 1.0 i
S... ••.• .Ii O 5 ···~::t-._;O"A..
~ 05 J. G··g::·_··i.-.e:: A ' •• , Q. . G.- •• ~0.5 s.-0" .... '~:. : f1
..~2 ,.. 0 I 0.0 ~\0.0 I I .... ::. 0.0 •••.k.. ..()
C( -0.5 -0.5 •.-0.5 -1.0 ~
-1.0 -'-- ---' -1.5
o 300 600 900 1200 1500 1800 -1.0 oL---=--=:-~~;_;;_O300 600 900 1200 1500 1800 o 300 600 900 1200 1500 1800
2.0 I I
C-CHP 2.0 I I
1.5 MS-CHP
2.0 I LS-CHPI
.2 1.5 1.5
'E 1.0 ~1.0 • :_ ~g •... .-A. -, A
;; 0.5 •...
G... 8., $,:'.', 0 O.SS' :-i::-
1.0 .. I
."
~ Q .. Q •• ~_ Q •• .Q~~ .A 0.5 G··.Q.. .Q::f:::ê.;o I
,g 0.0 ••• '6·. 0.0 •••. 0.0 _ I I I +"
cC ' ••••
-0.5 -0.5 0..'. ••• .• -0.5 j '-.
-1.0 .L._ --' -1.0 -1.0L .•..::..._---:-::~-:
o 300 600 900 1200 1500 1800 o 300 600 900 1200 1500 1800 o 300 600 900 1200 1500 1800I ~'u.
2.0 I I 2.0 TI--------------------------,I
2.0
MS-COP LS-COPc-COP
1.5 1.5 1.51.0 !
.'E2 1.0 1.0 • A' .A- .6
1:. •.... A. . ~ A',
~ 0.5 0.5 .•.• '.
0.5 j •...•
"'::~:'Q"Q ~ 0I\)__j S··.G .. ~ .. G•• Q " .• o
'u" ..... ',. '·g~t··. I 0.0 ". '."s- 0.0 G' .. G: :....-: -0. IQ .. ê..2 0' I I ". '._ .. .05 I I . I ••• ~ ••• ..~ IC( -0.5 -0.5 .... _ ....
-1.0 -'-- -' -1.0 .L._ ___J -1.0 -'-- -'
o 300 600 900 1200 1500 1800 0 300 600 900 1200 1500 1800 0 300 600 900 1200 1500 1800
Thermal time (Oed) from planting Thermal time (Oed) from planting Thermal time (Oed) from planting
Appendix 3G, The time course of calculated allocation ratios (AR) of leaf, stem and pod in beans, chickpea and cowpea under stress and
non-stress conditions in 2002/2003. Thermal time to flowering was 749, 711 and 854 °Cd for BN, CSP, and COP respectively,
--_-
222
Appendix4
Appendix 4A. The time course of mean leaf (LSM), stem (SDM), pod (pDM) and total above ground (ADM) dry matter production of three grain
legumes under three water regimes (C, MS, LS) in 2001/2002++
DAP LDM SDM PDM ADM
C MS LS C MS LS C MS LS C MS LS
Beans
39 15.7±3.9 19.7 ±3.9 15.7±3.9 11.8± 0.0 11.8±O.0 11.8±O.0 27.5±3.9 31.5±3.9 27.5±3.9
49 52.7±4.8 34.2 ±6.5 43.1±2.7 26.8±2.6 20.7±4.5 25.8±2.3 79.5±3.1 54.9±11.0 68.9±3.2
57 91.1±4.4 62.2±6.0 69.8±3.4 54.3± 8.9 43.9±5.2 56.7±13.9 I45.4±4.5 106.0±8.1 126.5±17.2
68 126.5±l1.8 85.2± 10.9 107.0±3.0 107.4±10.3 78.1±20.5 118.0±9.3 81.6±l7.8 28.1±16.7 85.4±10.1 315.6±35.1 191.4±47.6 310.5±8.7
78 138.7±10.4 81.6±16.3 95.4±16.l 117.5±9.6 86.8±14.7 119.0±5.3 220.5±16.5 78.9± 17.3 202.6±57.2 476.7±34.8 247.3±41.6 417.1±71.3
88 192.4±9.5 161.7±10.3 217.2±21.9 169.0±15.4 407.4±l52.0 150.5±12.6 817.1±180.3 481.2±37.6
98 75.0± 9.1 195.8±9.9 99.2±21.4 170.0±28.9 213.7±43.5 151.1±46.2 387.8±67.1 464.3±84.3
108 86.8 243.4±50.2 I 78.6±33.7 617.8±122.7
Chickpea
38 21.4±2.8 22.0±1.0 23.6±8.8 11±3.1 13.4±4.l 11.0±2.9 32.5±5.9 35.4±5.0 34.6±5.0
48 32.1±1.9 25.8±1.0 46.6±7.09 19±2.2 19.3±3.2 27.l±4.4 51.3±3.3 45.l±3.9 73.8±3.9
56 77.5±17.2 70.8±29.0 68.7±2.2 37±8.7 77.9±29.0 45.6±3.8 2.8±1.7 5.9±5.9 1.6±O.9 117.6±27.2 I 86.7±61.5 I I5.9±61.5
67 107.2±30.3 112.9±29.6 94.0±13.3 63±18.6 95.4±22.6 74.4±9.3 55.3±25.2 29.7±11.5 48.6±19.3 225.3±73.9 238.1±59.1 217.0±59.1
77 I63.7±33.l 135.6±39.4 108.4±23.5 148±35.2 117.1±1.9 115.3±26.8 I95.6±67.5 55.3±13.4 121.6±33.5 506.8±131.3 307.9±28.2 345.3±28.2
87 250.l±73.4 162.3* 257±54.4 79.9±79.9 362.2±59.l 136.9* 868.8±180.0 538.9±
97 157.6±46.6 248±120.2 295.9±96.9 701.6±263.2
107 118.0* 197* 534.1 * 621.8*
Cowpea
35 22.8±3.2 23.6±7.3 21.4±3.7 2.8±1.6 3.9±2.05 4.5±2.2 25.6±4.8 27.5±9.4 26.0±1.7
45 39.3±1O.4 35.4±6.8 31.5±3.9 19.7±7.9 11.8±O.0 15.7±3.9 59.0±18.0 47.2±6.8 47.2±6.8
55 98.4±12.9 89.5±9.6 88.9±2.3 53.l±9.8 58.8±9.03 60.8±3.1 151.5±22.7 148.3±18.5 149.7±4.7
66 133.8±29.1 96.2±O.9 118.8±11.6 104.3±31.7 90.3±7.5 121.8±15.5 9.4±3.1 13.0±3.5 10.6±4.9 241.2±63.9 190.0±11.3 249.3±25.1
76 142.8±1O.9 110.6±15.9 104.1±11.4 174.3±l1.3 119.6±14.0 I22.8±3.4 147.9±37.4 73.2±6.4 97.8±26.0 465.1±40.7 328.0±35.8 298.8±27.6
86 175.3±19.2 149.6±34.0 73.4±8.9 251.8±35.8 131.3±51.3 141.7* 375.0±40.2 I26.3±36.0 108.0±42.l 802.1±90.4 389.0±97.8 162.7±16.7
96 161.7±6.0 204.6±44.5 470.6±112.9 836.9±161.2
each value, with the respective standard errors, represents a mean of three replications, * values for only one replication
223
Appendix 4B. The time course of mean leaf (LSM), stem (SDM), pod (pDM) and total above ground (ADM) dry matter production of three grain
legumes under three water regimes (C, MS, LS) in 2002
OAP LOM SOM POM ADM
C MS LS C MS LS C MS LS C MS LS
Beans
17 18.9±1.0 11.2±O.3 10.0±1.1 5.9±1.5 3.5±O.5 3.0±1.0 25.0±0.9 14.6±O.5 13.0±O.3
27 60.2±O.7 46.0±2.0 50.8±O.8 24.8±O.9 24.8±1.8 24.2±1.2 85.0±0.3 64.9±3.4 75.2±1.2
37 128.7±2.4 117.5±2.1 129.3±2.3 86.8±1.6 73.8±2.8 83.2±3.6 358.0±2.6 308.7±2.2 345.7±6.0
47 266.2±1.2 180.6±2.7 253.2±2.4 253.2±1.5 152.9±2.6 240.2±2.2 68.5±O.2 33.l±O.3 71.4±O.3 519.4±1.7 334.1±3.3 495.6±5.1
57 282.7±1.7 168.2±3.3 265.0±1.6 333.5±1.7 191.8±2.2 265.0±3.2 193.6±O.6 66.l±O.7 179.4±2.5 924.7±1.7 445.8±2.1 830.2±2.0
67 255.0±3.1 122.2±4.5 167.6±2.5 259.1±2.8 210.7±2.6 208.3±4.4 556.6±O.7 108.0±1.0 545.9±5.9 1069.3±3.6 440.1±1.4 982.5±3.4
77 158.8±2.3 77.9±6.2 251.4±1.3 233.1±1.2 563.7±1.1 112.7±4.9 1063.6±O.7 423.8±O.3
87 93.3±2.2 283.9±1.3 681.1±2.7 1057.9±O.7
Chickpea
17 9.4±O.1 10.6±O.1 11.8±O.2 3.0±0.1 4.1±O.1 4.7±O.1 12.4±O.3 15.0±0.2 16.5±1.0
27 31.3±O.2 34.8±O.2 37.2±O.3 13.6±O.3 15.3±O.1 17.7±O.2 44.7±O.2 48.2±3.2 54.9±0.3
37 76.7±1.2 83.2±1.2 85.6±1.3 42.5±O.4 50.2±O.5 45.4±1.2 277.7±2.1 208.5±2.3 254.3±3.1
47 234.9±2.8 182.4±3.5 263.2±2.9 178.8±1.8 123.4±1.2 177.7±3.1 34.2±3.2 7.1±3.5 34.8±3.2 416.5±7.6 304.6±6.8 440.5±2.1
57 279.2±3.3 213.7±2.1 268.5±3.2 244.9±2.1 185.9±3.8 243.8±2.0 132.2±3.0 87.4±1.8 90.3±3.4 640.0±6.2 521.7±3.1 588.1±3.4
67 387.8±1.5 129.8±8.4 240.2±3.9 250.3±1.0 227.2±3.6 285.7±3.0 208.9±1.5 174.1±9.6 174.7±5.7 834.4±2.4 532.4±19.4 764.7±4.1
77 346.5±1.9 280.9±1.0 227.8±2.2 853.6±4.8
87 216.6±2.0 335.8±1.2 451.5±2.9 1006.7±4.3
Cowpea
17 9.4±O.4 15.3±1.2 10.6±1.1 3.5±O.2 2.4±O.4 4.1±O.3 13.0±0.7 17.7±1.7 14.8±1.7
27 28.3±O.3 36.6±1.9 36.0±1.7 14.8±O.3 14.8±O.6 15.9±O.3 43.3±O.5 50.8±3.3 51.7±4.3
37 108.0±1.1 86.8±1.2 91.5±1.5 54.9±0.7 41.3±O.8 44.3±O.9 161.5±2.6 130.2±1.9 135.7±5.5
47 168.8±2.3 168.2±2.8 175.9±1.8 206.2±1.6 137.5±1.8 213.7±1.2 360.4±5.1 306.5±4.8 390.5±1.9
57 324.0±3.2 240.2±3.5 290.4±5.8 292.2±1.3 286.3±2.9 305.7±3.8 192.4±1.1 155.8±5.2 193.0±8.1 915.8±3.6 784.4±7.9 774.8±10.4
67 367.7±4.1 129.3±2.2 112.1±4.7 309.9±O.8 299.2±1.8 221.3±5.3 291.0±1.5 286.8±5.8 252.6±5.2 968.0±3.3 717.1±2.4 587.6±2.5
77 240.8±2.1 103.3±2.6 373.6±O.8 321.7±1.5 367.7±1.8 293.3±6.7 979.6±1.5 710.6±3.9
87 212.5±2.0 327.6±1.1 449.7±4.7 988.4±2.3
224
Appendix 4C: The time course of mean leaf (LSM), stem (SDM), pod (pDM) and total above ground (ADM) dry matter production of three grain
legumes under three water regimes (C, MS, LS) in 2002/2003++
DAP LDM SDM PDM ADM
C MS LS C MS LS C MS LS C MS LS
Beans
20 10.8±O.5 6.8±2.3 11.8±1.2 4.5±O.7 7.5±1.2 5.9±O.9 15.3±O.7 19.9±1.9 17.7±O.6
30 32.1±3.2 32.9±2.8 31.3±2.2 14.8±1.0 14.4±O.9 14.0±0.5 46.8±4.1 47.2±3.6 45.2±2.6
40 96.6±6.8 87.7±3.5 92.9±4.8 52.5±5.1 47.6±3.l 54.l±5.6 149.l±12.0 135.4±5.2 147.0±8.8
50 152.3±13.8 143.4±18.0 148.3±15.5 11O.2±10.0 96.6±8.2 92.5±4.6 262.4±23.7 240.0±26.0 240.8±14.8
60 232.7±18.8 154.4±16.7 196.7±5.3 203.8±24.1 170.2±18.4 207.6±17.6 485.5±43.6 365.1±38.1 505.4±28.6
70 208.9±14.3 114.9±13.5 168.4±12.0 238.5±9.8 148.9±37.4 198.1±20.3 297.5±24.6 61.8±121.4 227.8±18.9 744.9±28.9 385.2±84.7 594.3±30.7
80 197.9±39.8 137.5±25.0 76.9±15.6 250.3±38.8 155.4±22.8 168.0±62.2 449.4±98.8 113.0±223.9 341.1±63.2 1039.6±110.6 705.9±35.7 722.0±63.3
90 119.2±13.2 82.2±4.8 251.8±21.6 144.0±20.3 630.7±53.2 49.56±231.4 1432.6±115.3 715.7±13.9
100 98.0±3.1 62.8±10.2 195.0±28.3 152.9±32.6 509.4±120.4 82.4±230.0 1352.4±152.6 655.1±80.2
Chickpea
20 12.4±1.2 I 1.8±1. 7 11.8±1.4 4.7±O.6 5.7±1.6 5.5±O.8 17.l±1.2 17.5±2.3 17.3±1.6
30 36.6±O.9 33.4±1.6 31.3±1.0 15.9±1.2 14.2±O.0 14.4±O.5 52.5±2.0 47.6±1.6 45.6±O.8
40 94.6±1.3 100.7±3.4 84.0±1O.4 55.5±2.5 54.1±1.l 49.4±5.5 150.1±1.9 154.8±4.3 133.4±15.9
50 158.0±7.7 150.3±15.0 136.7±6.5 101.7±9.1 99.0±10.7 88.1±6.5 259.7±26.6 249.3±25.7 224.9±1O.4
60 227.0±3.6 146.4±26.7 207.0±16.4 190.4±9.7 150.9±15.5 160.3±15.7 82.4±9.9 54.3±10.8 50.2±12.3 499.9±12.7 351.6±36.2 417.5±23.6
70 212.5±19.8 I 16.9±21.0 I 87.7±37.4 I 97.3±21.7 159.6±26.2 I 97.5±47.4 220.0±31.1 145.0±55.0 143.0±42.l 629.8±72.6 488.1±72.0 461.6±73.4
80 178.2±16.0 112.7±6.8 131.4±16.3 268.2±27.2 135.2±8.6 220.9±50.1 283.9±65.3 77.9±21.3 168.6±25.7 841.0±51.1 485.2±26.4 591.2±30.7
90 193.8±26.5 95.6±1O.3 369.3±10.8 191.4±45.9 214.1±1O.0 174.1±67.2 928.2±32.9 474.8±54.6
100 154.0±16.l 103.9±12.0 364.9±67.3 186.1±24.8 331.3±33.0 138.7±21.0 1028.2±68.9 524.7±82.3
Cowpea
20 12.6±O.5 12.0±0.7 12.8±O.7 4.5±O.4 4.3±O.2 4.5±O.4 17.l±O.6 16.3±O.9 17.3±1.l
30 33.6±2.1 34.6±1.9 35.6±1.3 14.6±O.4 16.3±2.2 15.5±O.5 48.2±2.3 51.0±4.0 51.2±1.5
40 85.2±2.1 93.8±7.5 95.2±6.1 37.4±2.0 42.7±2.3 49.2±O.8 122.6±2.6 I 36.5±9.8 144.4±6.9
50 147.2±14.3 153.7±16.9 147.6±13.0 80.1f3.2 87.7±4.l 86.6±5.0 227.2±17.4 241.4±19.5 234.1±14.9
60 195.0±2.9 165.l±10.0 209.1±9.8 I 84.7±4.6 156.6±12.6 21O.5±7.7 19.7±4.6 22.6±4.9 19.5±O.7 399.4±9.7 344.3±18.4 439.1±16.3
70 184.5±15.9 106.2±7.4 160.3±4.7 232.2±22.3 149.3±18.7 203.6±11.5 222.1±33.l 88.3±7.2 148.3±18.3 638.8±43.3 343.9±24.1 571.3±33.4
80 151.5±2.3 88.9±15.7 I49.5±1.4 218.6±7.1 171.2±24.2 201.7±14.9 184.9±12.4 161.3±20.3 223.5±29.2 709.2±20.8 541.8±60.9 707.1±16.8
90 128.1±8.1 57.4±28.7 231.2±39.7 85.4±43.1 304.9±99.7 76.1±38.1 1292.0±37.8 417.6±12.3
100 135.2±13.l 85.6* 248.5±23.5 72.8±72.5 302.8±39.0 61.8±61.8 1142.7±77.6
++ each value, with the respective standard errors, represents a mean of three replications * values for only one replication
225
AppendixS
Appendix SA. Seasonal (ETs), pre-flowering (ETb), post flowering (ETa) and ratio of pre- to post flowering (ETa: ETb) water use (mm),
seasonal transpiration (Ts, mm), soil evaporation (Es, mm), water use efficiency (kg ha" mm") for pre-flowering (WUEb), post flowering
WUEa), above ground dry matter at harvest (WUEd) and grain yield (WUEg) and transpiration efficiency for grain yield (TEg, g mm") of
three grain legume species for 2001/2002 seasons"
Species (Sp) Water ETs ETb ETa ETb:ETa r, Es WU~ WUEa WUEd WUEg TEg
Regime(WR)
C BN 379.3 224.0 155.2 1.45 175.2 204.3 14.2 17.6 12.2 6.3 1.32
CHP 416.0 203.9 212.0 0.98 136.9 279.0 11.1 17.6 9.8 5.6 1.66
COP 410.7 246.1 164.5 1.49 192.4 218.0 6.3 33.4 10.2 4.2 0.88
MS BN 351.3 224.0 134.3 1.63 60.6 290.7 8.7 16.9 4.2 1.3 0.73
CHP 318.0 203.9 111.0 1.99 49.2 269.0 11.2 21.9 4.3 0.9 0.61
COP 283.0 246.1 77.7 2.87 150.6 132.3 6.7 25.3 11.4 2.5 0.49
LS BN 328.3 217.0 95.6 2.31 94.2 234.3 14.2 18.3 7.9 1.7 0.57
CHP 307.0 207.1 86.7 2.31 89.5 217.7 9.6 25.4 8.4 3.3 1.06
COP 282.7 218.5 56.5 4.01 137.9 144.7 6.7 29.8 10.3 2.7 0.54
LSD (POO.05) WR 32.9·· n.s. 15.3·· 0.31·· 54.6· n.s. n.s. n.s. n.s. 2.5· 0.22··
Sp n.s. 14.1·· 18.9·· 0.34·· 19.1·· 38.8·· 3.2·· 4.6·· 1.8·· n.s. 0.24··
WRxSp 46.2· n.s. 32.7· 0.59·· 33.4·· 67.1· n.s. n.s. 3.2·· n.s. n.s.
CV(%) 7.6 6.3 15.1 15.8 15.4 17.1 31.7 19.6 20.5 40.5 26.5
*, **: Significant at 5 and 1% probability level respectively; 3: Soil water was measured to a depth of 30 cm. BN: beans; ClIP: chickpea; COP: cowpea
226
Appendix 6
Appendix 6A. Experimental site (Dire Dawa) soil profile description, general
1. Implementing unit: PhD thesis project
2. 2. Region: Dire Dawa
3. Village: TonyFarm
4. Profile No. 1 (200 m south of the main Office of Tony farm administration)
5. Altitude: 1176m
6. Latitude 09° 36.8'N
7. Longitude 41°50.4' E
8. Physiography: Alluvial plain
9. Geology: Alkaline olivine basalts and tuff of the Ashange group
10. Parent material: Colluvium derived from basalt
11. Topography: Level (flat)
12. Rainfall: 612mm
13. Evaptranspiration: 1964 mm
14. Growing period 60-70 days
15. Slope: a) Gradient: 8% (11°) b) Length: lOOm
16. Erosion: None
17. Drainage: Well drained
18. Water table: a) dry season: 24 m b) Wet season: 10m
19. Flooding: Medium
20. Gravels, stones, rock outcrops: None
21. Quality of ground water: Good
22. Degree of degradation: Non degraded
23. Surface cracks: None
24. Nature of soil formation: Alluvial deposition
25. Natural vegetation: Cultivated land surrounded by different tree species and
orange plantation
26. Important crops grown: cabbage, carrot, egg plant, pepper, cucumber, onion,
sorghum, maize, bean
27. Soil Class (FAO): Eutric Regosol
28. Date: Jun 19,2002
227
Appendix 6B. Profile description-physical properties
Depth Horizon p Colour Mottle Structure Consistence Presence of Porosity Effere- Cracks and
(cm) (g/cnr') roots vescence nodules
Dry Wet Size Type Dry Moist Wet
0-10 AP 1.24 4/3 3/4 None Fine Blocky Soft Friable Non Very high Very High None
5YR 5YR Granular sticky fine
10-41 A 1.23 3/4 3/2 None Very Sub-angular Slightly Very Slightly High Fine High None
5YR 7.5YR fine blocky hard friable sticky
41-71 BI 1.27 4/3 3/2 None Medium Angular Very hard Very Sticky Very few Medium High None
5YR 5YR blocky friable
71-91 BIl 1.41 3/2 3/47.5 None Medium Prism like Very hard Friable Slightly Very fine Medium High None
5YR YR columnar sticky roots
91+ BIll 1.36 3/3 *5YR None Fine Granular Very hard Very Non None Fine High None
5YR friable sticky
Colour descriptions: 4/3 5YR = Reddish brown; 3/4 5YR, 3/3 5YR, 3/2 5YR = Dark reddish brown; 3/2 7.5YR, 3/4 7.5YR = Dark brown
228
Appendix 6C. Soil water relations*
Depth Thickness (mm) DUL(%) DLL(%) DUL DLL ASW(mm)
0.33 bar 15 bar (mm) (mm)
0-10 100 32.01 15.43 32 15 17
10-41 3lO 36.24 17.90 112 56 56
41-71 300 36.83 19.10 III 57 54
71-91 300 34.58 18.12 lO4 54 50
>91 100 32.37 16.19 32 16 16
Total (mm m") 391 198 193
* DUL = drained upper limit of soil water, PWP =drained lower limit of soil, ASW = available soil water.
Appendix 6D. Soil chemical (and some physical) properties*
Lab.No. Soil Depth pH(H20) EC Org.C OM TotalN P Exchangeable cations (meq/l00gm soil) Texture (%) Texture
(cm) (mmhos/cm) (%) (%) (%) (ppm) Class
K Na Ca Mg CEC Sand Clay Silt
15195 0-10 8.48 0.634 1.177 2.029 0.101 19.87 0.75 0.50 26.22 10.26 58.6 31 33 36 CL
15196 11-40 8.60 0.533 1.357 2.339 0.117 14.15 1.08 1.11 29.92 9.33 58.6 23 40 37 CL
15197 41-70 8.61 0.496 1.197 2.064 0.lO3 6.91 0.80 1.37 31.81 10.10 58.8 19 38 44 SCL
15198 71-90 8.45 0.601 1.137 1.961 0.098 5.47 0.56 1.19 33.42 8.82 58.6 16 41 43 SC
15199 >91 8.48 0.477 0.998 1.720 0.086 5.38 0.53 1.23 31.71 5.76 58.2 36 36 27 CL
* CL= clay loam, SCL= silty clay loam, SC = silty clay
229
Appendix 7
Appendix 7A. Daily weather conditions at Dire Dawa «latitude 9°6'N, longitude 41°8' E,
altitude 1197 m), Ethiopia in the 2001/2002 season.
Day DAP Tomea• Tome1n
p+ Ir** Wind speed n RHmax RHmln SR* ETo
mm Mm m s' % % MJ m-<day" mm
I-Dec 30.2 15.5 0 0.88 10.7 68 23 20.7 4.3
2-Dec 31.2 15 0 0.83 10.7 56 18 20.7 4.5
3-Dec 30.6 13.2 0 0:93 10.4 41 22 20.3 4.2
4-Dec 31 13.6 0 0.93 10.2 58 27 20.1 4.2
5-Dec 30.7 15.8 0 0.93 10.3 48 21 20.2 4.2
6-Dec 31.7 16.4 0 0.83 9.6 60 30 19.3 4.4
7-Dec 30.2 16 0 0.82 7.5 84 39 16.5 3.8
8-Dec 1 28.6 18.5 6 0.6 4.7 90 50 12.8 3.5
9-Dec 2 28.8 15.2 0 0.75 9.3 96 37 18.9 4.1
lO-Dec 3 29.8 15.7 0 0.79 8.5 86 30 17.8 4.1
Il-Dec 4 31.4 17.8 0 16.7 0.78 10.6 70 18 20.6 4.7
12-Dec 5 31.3 17 0 0.93 10.2 56 25 20.1 4.3
13-Dec 6 30 16.1 0 0.78 7.8 89 42 16.9 3.9
14-Dec 7 30.2 17.6 0 0.92 10.3 90 32 20.2 4.1
IS-Dec 8 30.1 15.2 0 16.7 0.82 10.2 81 26 20.1 4.3
16-Dec 9 30.5 15.1 0 0.79 10.1 60 32 19.9 4.4
17-Dec 10 30.7 16.8 0 0.79 10 74 32 19.8 4.4
18-Dec Il 29.7 15 0 0.78 9.4 95 38 19.0 4.1
19-Dec 12 30.1 14.4 0 0.93 10.1 76 27 19.9 4.1
20-Dec 13 28.4 13.6 0 0.87 10.2 80 24 20.1 4.1
21-Dec 14 28.2 13.7 0 0.81 10 80 30 19.8 4.1
22-Dec 15 28.5 14.2 0 16.7 0.8 10 77 30 19.8 4.2
23-Dec 16 29 14.2 0 0.83 10.1 87 22 19.9 4.2
24-Dec 17 28.7 12.1 0 0.83 10.1 71 27 19.9 4.2
25-Dec 18 28.2 13.8 0 0.83 10.1 78 28 19.9 4.1
26-Dec 19 29.4 14.2 0 0.73 10.1 73 27 19.9 4.5
27-Dec 20 29 13.6 0 0.68 9.5 61 28 19.1 4.6
28-Dec 21 28.7 13.5 0 0.75 10.1 80 29 19.9 4.3
29-Dec 22 29 14.9 0 0.73 10.1 72 31 19.9 4.4
30-Dec 23 29.8 14.1 0 0.73 10.1 76 21 19.9 4.6
31-Dec 24 29.9 14 0 0.89 10.1 65 33 19.9 4.1
I-Jan 25 31 14.8 TR 16.7 0.67 8 92 39 17.0 4.2
2-Jan 26 26.3 19.2 5 0.72 2.9 95 58 10.3 2.7
3-Jan 27 24.4 18 13.9 0.45 1.6 98 80 8.6 2.2
4-Jan 28 24.6 18.2 12.7 0.36 1.2 96 68 8.0 2.8
5-Jan 29 26.7 18.4 3.2 0.53 3.4 96 60 11.0 3.1
6-Jan 30 25.5 18.4 TR 0.64 2.8 98 67 10.2 2.5
7-Jan 31 27 16.8 TR 0.66 6.1 100 59 14.6 3.3
8-Jan 32 27.5 17.1 0 0.8 7.3 96 46 16.2 3.5
9-Jan 33 28.9 16.8 0 0.75 10 85 45 19.8 4.2
lO-Jan 34 28.7 16 0 0.71 8.8 92 47 18.2 4.0
Il-Jan 35 29.6 16.6 0 25.0 0.74 10.2 95 48 20.1 4.2
12-Jan 36 27.5 17.1 TR 1.21 5 92 51 13.2 2.7
13-Jan 37 24.7 17.2 TR 0.75 0.9 98 65 7.8 2.1
14-Jan 38 28 17.5 0 0.76 5.1 96 46 13.4 3.3
IS-Jan 39 28.8 15.2 0 0.72 8.8 94 41 18.4 4.1
16-Jan 40 27.7 16.4 0 0.91 7.5 94 41 16.7 3.5
17-Jan 41 28.5 15.3 0 16.7 0.79 9.2 86 41 19.0 4.0
18-Jan 42 28.6 15.8 0 0.89 9.4 83 38 19.3 3.9
19-Jan 43 29.5 15.6 0 0.85 10.4 79 35 20.7 4.3
20-Jan 44 28.2 14.5 0 0.88 10.6 82 34 21.0 4.1
21-Jan 45 27.8 14.4 0 0.89 10.5 79 38 20.9 4.1
22-Jan 46 27.4 14.3 0 0.73 6 87 42 14.9 3.6
23-Jan 47 28.1 14.7 0 0.85 10.7 92 32 21.3 4.2
230
24-Jan 48 29.1 13.6 0 0.97 10.7 68 25 21.3 4.2
25-Jan 49 29 13 0 25.0 0.99 10.4 53 16 21.0 4.1
26-Jan SO 28.3 12 0 0.98 10.8 61 30 21.6 4.1
27-Jan SI 28.5 12.3 0 0.85 10.8 56 27 21.6 4.4
28-Jan 52 30 14.2 0 0.83 10.4 64 27 21.1 4.5
29-Jan 53 29.1 13.2 0 16.7 0.89 10.5 77 29 21.3 4.3
30-Jan 54 28.9 14.5 0 0.78 10.7 75 28 21.7 4.6
31-Jan SS 29.1 14.6 0 0.87 10.5 78 33 21.4 4.3
I-Feb 56 29.8 12 0 16.7 0.96 10.5 81 20 21.5 4.3
2-Feb 57 29.8 11.7 0 0.81 10.6 63 18 21.7 4.7
3-Feb 58 30.1 12.4 0 0.68 10.6 60 27 21.7 5.0
4-Feb 59 30.8 12.8 0 0.79 10.4 69 29 21.5 4.7
5-Feb 60 29.9 IS 0 16.7 0.86 10.4 72 26 21.6 4.5
6-Feb 61 30.5 14.7 0 0.84 10.4 70 24 21.6 4.7
7-Feb 62 30 13.2 0 0.76 10.5 66 23 21.8 4.8
8-Feb 63 30.6 13.1 0 25.0 0.83 10.6 74 19 22.0 4.7
9-Feb 64 30.3 13.1 0 0.77 10.7 53 20 22.2 5.0
lO-Feb 65 31 13 0 0.75 10.6 62 14 22.1 5.1
Il-Feb 66 30.5 11.8 0 0.67 10.3 66 21 21.8 5.2
12-Feb 67 31 11.6 0 33.3 0.66 10.3 79 26 21.8 5.2
13-Feb 68 30.6 12.8 0 0.82 10.4 90 33 22.0 4.6
14-Feb 69 30.7 16.8 0 0.69 9.9 91 39 21.4 4.9
IS-Feb 70 30 18.8 0 16.7 0.8 7 87 41 17.2 4.1
16-Feb 71 30 19.8 0 0.76 6.4 80 43 16.4 4.1
17-Feb 72 32 18 0 16.7 0.87 9.9 85 31 21.5 4.8
18-Feb 73 33.9 18 0 0.92 lO 69 29 21.7 4.9
19-Feb 74 32.1 16.6 0 0.99 10.3 95 33 22.2 4.6
20-Feb 75 31.6 17.4 0 25.0 1.04 10.5 92 37 22.6 4.5
21-Feb 76 34.6 18.4 0 0.95 10.4 61 25 22.5 5.0
22-Feb 77 33.5 16.5 0 0.89 10.1 92 31 22.1 4.9
23-Feb 78 33 16.4 0 16.7 0.88 9.4 95 26 21.1 4.8
24-Feb 79 33.2 14.8 0 1.08 10.6 89 20 22.9 4.7
25-Feb 80 34.2 19.7 0 0.88 10.6 32 17 23.0 5.3
26-Feb 81 34.1 17.7 0 25.0 0.93 10.6 67 23 23.0 5.1
27-Feb 82 33.3 17.2 0 1.03 10.4 71 25 22.8 4.8
28-Feb 83 33.5 17.8 0 1.17 10.4 70 25 22.9 4.6
l-Mar 84 32.9 17.5 0 1.04 10.3 SO 31 22.8 4.8
2-Mar 85 33.3 17.3 TR 25.0 1.09 8.1 96 32 19.5 4.3
3-Mar 86 22.3 18.8 58 0.9 0 97 79 7.5 1.8
4-Mar 87 25.5 17 TR 0.57 0.4 96 52 8.1 2.8
5-Mar 88 27 18.7 1.7 0.9 4.7 96 62 14.6 3.2
6-Mar 89 29.5 16.5 0 0.81 9.9 95 51 22.4 4.6
7-Mar 90 30.9 17 0 1.05 10.4 84 42 23.2 4.6
8-Mar 91 33.1 18.6 0 1.03 10.7 45 28 23.7 5.0
9-Mar 92 32.2 20 0 1.14 10.4 47 33 23.3 4.8
lO-Mar 93 32.3 21.2 0 0.7 9.9 60 39 22.6 5.6
li-Mar 94 31.7 17.7 0 0.82 9.6 83 30 22.2 5.1
12-Mar 95 32.4 16 0 0.86 10.4 89 22 23.4 5.2
13-Mar 96 32 15.6 0 0.71 10.6 77 25 23.8 5.6
14-Mar 97 30.9 14.8 0 0.69 10.6 77 29 23.8 5.5
IS-Mar 98 32.1 17 0 1.04 10.4 79 35 23.6 4.9
16-Mar 99 31.9 20.5 TR 2.95 3.8 56 33 13.6 2.5
17-Mar lOO 33.8 24.5 TR 3.2 9.2 54 33 21.8 3.6
18-Mar lOl 31.1 22 TR 0.98 1.7 88 41 10.4 3.0
19-Mar 102 29.8 21.6 3.4 1.23 5.1 96 45 15.6 3.4
20-Mar 103 30.8 ,18.8 14.6 0.95 5.1 93 41 15.6 3.8
• calculated using Angstrom s equatton from sunshme duration, TR - trace, calculated using the Penman-Montieth
Equation .•• Irrigation amounts are for the control treatments. Water stressed treatments did not receive water for the
duration shown in Table 3.1b.
L- __
231
Appendix 7B. Daily weather conditions at Dire Dawa «latitude 9°6'N, longitude 41°8' E,
altitude 1197 m), Ethiopia in the 2002 season.
Date DAP Tm .. Tm1n p lr** Wind speed n RHmax RHm1n SR* ET.
oC oC mm mm m s' hour % % MJ m-'day- mm
20-Mar 30.8 18.8 14.6 0.95 5.1 93 41 15.6
21-Mar 33.3 17.4 1.3 0.84 7 87 31 18.6
22-Mar 31.4 19.2 0 0.74 9.1 87 44 21.8
23-Mar 33.8 20.5 TR 0.74 9.5 82 35 22.5
24-Mar 33.6 21.4 0 0.74 9 82 33 21.7
25-Mar 35.2 19.1 0 1.01 10.4 74 25 23.9
26-Mar 36.4 20.2 TR 2.55 7.3 46 21 19.2
27-Mar 36.4 25.9 0 1.56 10.4 44 23 24.0
28-Mar 1 35.4 23.6 1.9 1.03 6.1 72 30 17.4
29-Mar 2 33.3 18.2 0 25.0 0.89 9.6 90 32 22.8
30-Mar 3 33.1 19 0 1.04 9.9 76 34 23.3
31-Mar 4 35.8 20 0 1.03 10.1 74 27 23.6
I-Apr 5 35.4 21.3 TR 25.0 1 8 49 27 20.4 5.1
2-Apr 6 35.1 21.2 0 1.57 7.4 51 31 19.5 4.3
3-Apr 7 35.4 24.6 1 1.17 6 64 33 17.3 4.4
4-Apr 8 34 21.3 1.4 0.98 6.2 86 41 17.6 4.3
5-Apr 9 34.7 20.5 18.3 25.0 1.18 10.3 92 32 24.0 5.3
6-Apr 10 32.4 19.2 TR 1.03 8.2 83 39 20.8 4.6
7-Apr II 33.4 20.1 15.5 1.03 7.5 88 33 19.7 4.6
8-Apr 12 34.4 19.7 TR 1.6 8.8 71 32 21.7 4.5
9-Apr 13 34.7 22.6 0 1.35 8 60 28 20.5 4.7
lO-Apr 14 32.4 20.4 0 3.9 8.9 94 32 21.9 3.7
II-Apr IS 34.5 24.8 28.8 2.78 10.4 46 28 24.2 4.4
12-A~r 16 30.1 18.1 5 1.49 6.1 85 SO 17.6 3.6
13-Apr 17 28 19 2.6 0.63 3.8 95 64 14.0 3.5
14-Apr 18 29 18.2 TR 0.82 6.2 96 SI 17.7 4.0
IS-Apr 19 30.2 18 0 0.59 8.5 94 44 21.3 5.0
16-Apr 20 29.8 20.7 6.8 1.14 8.4 92 47 21.1 4.5
17-Apr 21 33.4 21 0 2.06 10 64 36 23.6 4.5
18-Apr 22 33.6 23 0 2.28 11.2 61 32 25.5 4.8
19-Apr 23 35.5 23.5 0 0.69 11.2 63 32 25.5 6.5
20-Apr 24 33.5 20.7 0 0.95 11.1 86 33 25.3 5.6
21-Apr 25 31.4 21.1 0 1.16 9.4 88 41 22.7 4.9
22-Apr 26 32 18.1 0 1.1 10.6 69 25 24.6 5.2
23-Apr 27 33.2 17.4 0 33.3 0.76 11 49 23 25.2 6.0
24-Apr 28 34.9 18.5 0 0.97 9.8 45 19 23.3 5.4
25-Apr 29 34.6 17.5 0 1.08 9.3 56 22 22.5 5.1
26-Apr 30 34.3 16.5 0 16.7 1.16 9.3 44 24 22.5 4.9
27-Apr 31 33.2 21 TR 1.12 5.5 57 30 16.6 4.2
28-Apr 32 36.7 22 0 1.71 9.1 43 24 22.2 4.7
29-Apr 33 36.1 24 0 16.7 1.73 10.1 57 23 23.8 5.0
30-Apr 34 36.2 24 3.7 2.1 7.9 88 26 20.3 4.4
I-May 35 35.8 22.4 0 1.69 7.5 SS 24 19.7 4.3
2-May 36 36.8 24.2 0 1.7 11.1 57 18 25.3 5.3
3-May 37 36.6 24 0 1.16 11.2 60 22 25.4 5.8
4-May 38 34.8 23.2 0 25.0 1.21 10 64 27 23.6 5.3
5-May 39 35.7 23 TR 1.23 8.8 59 28 21.7 5.0
6-May 40 35.3 22.6 0 1.01 10.8 64 25 24.8 5.8
7-Mav 41 35.5 22.7 0 16.7 1.51 8.7 62 24 21.5 4.7
8-Mav 42 34.9 24.5 0 1.03 7.4 58 19 19.5 5.0
9-Mav 43 36.5 21.2 0 1.12 11 SS 15 25.0 5.7
10-Mav 44 37.7 22.6 0 1.9 11 59 11 25.0 5.1
I I-May 45 38.1 24 0 16.7 2.31 11 57 13 25.0 4.9
12-May 46 37.8 24.4 0 1.65 10.8 56 16 24.7 5.3
13-May 47 37.8 24 0 2.05 11.1 58 II 25.1 5.1
14-May 48 38 24.6 0 16.7 2.53 8.2 SS 21 20.7 4.2
IS-May 49 37.7 25.2 3.1 3.36 10.2 SS 15 23.7 4.3
232
16-May 50 37.9 21.2 1.3 1.73 Il 57 20 24.9 5.2
17-May 51 37.3 23.8 0 2.03 9.3 52 18 22.3 4.6
18-May 52 35.3 23.6 0 25.0 2.71 5.6 55 21 16.7 3.4
19-May 53 36.6 24.2 0 2.95 11.3 56 26 25.4 4.6
20-May 54 38.1 24.3 0 16.7 2.11 9.6 53 18 22.7 4.7
2I-May 55 37.2 23.7 0 2.95 9.6 57 25 22.7 4.3
22-May 56 35.7 24.4 25.8 2.39 8.6 90 30 21.2 4.4
23-May 57 35.7 18 0 1.47 11.3 71 27 25.3 5.2
24-May 58 36.6 23.4 0 1.67 10.9 59 17 24.7 5.2
25-May 59 37.1 23.8 0 2.35 11.5 54 16 25.6 4.9
26-May 60 37.5 23.6 0 2:42 9.9 58 18 23.1 4.6
27-May 61 37.1 24.2 TR 25.0 2.77 9.3 50 21 22.2 4.2
28-May 62 35.7 23.8 0.4 3.04 8.3 55 27 20.7 3.9
29-May 63 38.1 24.6 TR 25.0 2.15 9.4 45 18 22.3 4.5
30-May 64 37.7 24.2 TR 2.21 8.8 59 19 21.4 4.5
31-May 65 37.5 23.7 2.7 2.74 7.8 65 16 19.9 4.0
I-Jun 66 36.9 24 TR 2.03 10 53 23 23.2 4.7
2-Jun 67 38.2 24 TR 2.63 9.5 51 16 22.4 4.4
3-Jun 68 36.8 24.2 TR 2.26 10.9 50 24 24.5 4.8
4-Jun 69 38.2 24.8 0 25.0 2.03 9.8 41 17 22.9 4.6
5-Jun 70 36.5 24.6 0 2.91 10 46 17 23.2 4.2
6-Jun 71 34.9 23.5 TR 2.83 8.6 55 27 21.0 3.9
7-Jun 72 33.8 23.2 1.8 2.57 8.2 60 31 20.4 3.9
8-Jun 73 33.8 20.6 5 25.0 2.42 9.8 73 29 22.8 4.3
9-Jun 74 35.6 19.8 2.9 2.6 6.5 65 26 17.9 3.6
lO-Jun 75 36 20 TR 3.08 8 53 24 20.1 3.7
Il-Jun 76 36.9 23.6 0 3.2 10.3 57 18 23.6 4.3
12-Jun 77 36.7 21.5 0 2.95 9.6 56 19 22.5 4.1
13-Jun 78 36.7 22.8 0 3.7 9 55 16 21.6 3.8
14-Jun 79 34.2 23 TR 3.03 2.7 56 27 12.1 2.7
IS-Jun 80 36.5 23.2 TR 1.84 8.3 56 24 20.5 4.4
16-Jun 81 36.6 23 TR 2.3 10.2 57 19 23.4 4.6
17-Jun 82 36.8 23.1 0 3.52 9.3 55 17 22.0 3.9
18-Jun 83 35.6 22.2 0 3.33 7.9 56 27 19.9 3.6
19-Jun 84 37.3 23 0 2.84 10.2 52 18 23.4 4.3
20-Jun 85 37.1 22.3 0 3.21 10.5 50 18 23.8 4.2
21-Jun 86 34.8 23 0 3.62 10.7 62 27 24.1 4.1
22-Jun 87 36.5 22.7 0 2.18 10.5 61 20 23.8 4.7
23-Jun 88 36.7 22.6 0 2.65 10.3 52 19 23.5 4.4
24-Jun 89 36.5 22.7 0 2.61 10.6 56 19 24.0 4.5
25-Jun 90 36.4 22.5 0 2.23 10.1 53 18 23.2 4.5
26-Jun 91 36.6 23 0 2.08 9 56 15 21.6 4.4
27-Jun 92 37.4 23.4 0 2.89 10.7 53 16 24.2 4.4
28-Jun 93 38 23 0 3.52 10 49 16 23.1 4.1
29-Jun 94 37.5 23.2 0 3.41 10.4 51 18 23.7 4.2
30-Jun 95 36.8 2,4.2 TR 2.39 8 50 18 20.1 4.1...calculated using Angstrom s equation from sunshine duration, TR -- trace, calculated using the Penman-Montieth
method ....... Irrigation amounts are for the control treatments. Water stressed treatments did not receive water for the
duration shown in Table 3.1b.
233
Appendix 7C Daily weather conditions at Dire Dawa «(latitude 9°6'N, longitude 41°8' E,
altitude 1197 m), Ethiopia in the 2002/2003 season.
Day DAP TmD• Tmin p Ir** Wind speed RHmax RHmln SR ET.
oC oC mm mm m s' 010 % MJ m'<day' mm
5-0ct 30.0 17.4 0.7 0.97 14.1 3.9
6-0ct 33.1 20.0 0 1.13 17.9 4.4
7-0ct 33.6 17.2 0 1.14 22.2 4.3
8-0ct 35 16.9 0 1.01 20.2 4.4
9-0ct 35.4 18.6 0 0.95 19.4 4.3
10-0ct 34.1 18.2 0 1.33 17.8 4.2
11-0ct 35 22.5 0 1.16 18.8 4.4
12-0ct 34.8 19 0 1.35 22.7 4.2
13-0ct 35.2 22.2 0 1.6 23.1 4.3
14-0ct 35.7 21.3 0 1.36 21.4 4.3
15-0ct 35.5 21.9 0 1.26 22.1 4.3
16-0ct 35.8 20.4 0 1.42 19.1 4.2
17-0ct 35.5 21 0 1.3 22.9 4.3
18-0ct 1 35.4 20 0 1.28 21.5 4.0
19-0ct 2 34.6 20.9 0 1.18 22.8 4.1
20-0ct 3 34.7 20.9 0 1.25 22.9 4.5
21-0ct 4 34.4 22 0 1.48 22.8 3.8
22-0ct 5 31.5 22.4 0 25.0 1.59 15.2 4.4
23-0ct 6 34.4 19.5 0 1.03 22.5 4.3
24-0ct 7 33.4 20.5 0 1.14 20.9 4.2
25-0ct 8 33.4 16.8 0 25.0 1.14 22.3 4.3
26-0ct 9 33.4 15.5 0 1.19 20.8 4.4
27-0ct 10 32.2 15.8 0 0.86 21.2 4.3
28-0ct 11 31.9 20.3 TR 25.0 1.05 13.2 4.1
29-0ct 12 33.4 21.4 0 1.06 21.0 4.2
30-0ct 13 34.3 21.5 0 1.28 22.0 4.1
31-0ct 14 33.4 21.1 0 25.0 1.21 21.5 3.4
I-Nov 15 34 17.5 0 1.21 21.6 3.9
2-Nov 16 33.7 17.8 0 1.14 21.7 3.3
3-Nov 17 33.4 17 0 1.07 22.0 3.7
4-Nov 18 32 15.7 TR 0.9 19.3 3.8
5-Nov 19 30.3 19.8 0 33.3 1.04 17.2 4.0
6-Nov 20 30.8 15.3 0 0.83 17.8 3.9
7-Nov 21 32.6 15.4 0 1.11 21.7 3.8
8-Nov 22 32.9 17.1 0 1.07 21.6 3.7
9-Nov 23 33.5 16.8 0 25.0 1.13 20.9 3.6
lO-Nov 24 32.2 17.8 0 1.11 20.8 3.9
Il-Nov 25 32.2 16.2 0 1.00 21.5 3.7
12-Nov 26 32.2 16.2 0 1.15 21.5 3.8
13-Nov 27 32.5 17 0 25.0 1.08 21.6 3.4
14-Nov 28 32.5 17.6 0 1.07 21.4 3.0
15-Nov 29 33 17.2 0 1.06 21.5 3.8
16-Nov 30 32.4 16 0 1.11 21.3 4.0
17-Nov 31 32.1 14.6 0 1.03 21.7 3.6
18-Nov 32 31.3 14.4 0 25.0 1.14 21.4 3.6
19-Nov 33 30.5 14 0 1.06 21.3 3.4
20-Nov 34 30.6 15.4 0 0.87 21.2 3.2
21-Nov 35 30.2 17.2 0 0.93 16.2 2.1
22-Nov 36 32.2 17 0 0.96 21.1 2.3
23-Nov 37 32 14.3 0 1.00 20.9 2.6
24-Nov 38 31.5 16 0 1.00 56 22 20.7 1.8
25-Nov 39 31.6 17.1 0 1.00 61 19 21.2 2.2
26-Nov 40 31.5 17.5 0 0.95 64 20 21.2 2.4
27-Nov 41 31.6 17.6 0 1.00 76 21 20.7 2.7
28-Nov 42 31.5 17 0 29.2 1.03 78 21 20.0 2.5
29-Nov 43 30.6 15.8 0 0.98 65 24 20.8 2.6
30-Nov 44 31.5 14.8 0 1.02 85 33 20.5 3.9
234
I-Dec 45 31.39 17.07 0 16.7 1.39 86 34 18.7 3.3
2-Dec 46 30.32 17.73 0 0.99 93 30 19.4 3.3
3-Dec 47 30.8 16.26 0 1.09 98 32 15.8 3.6
4-Dec 48 30.31 15.03 0 1.37 79 26 21.2 3.l
5-Dec 49 30.47 14.4 0 1.26 90 25 20.8 2.7
6-Dec 50 29.01 14.59 0 1.14 85 26 20.5 3.4
7-Dec 51 28.93 11.91 0 25.0 1.24 81 14 21.4 3.9
8-Dec 52 28.63 14.04 0 1.20 72 27 20.6 4.0
9-Dec 53 29.99 16.1 0 12.5 1.12 77 29 20.2 3.7
lO-Dec 54 29.96 l3.38 0 1.11 93 32 17.8 3.7
Il-Dec 55 30.l4 16.11 0 1.17 84 27 20.1 3.9
12-Dec 56 29.76 16.99 0 1.35 69 29 20.1 3.9
13-Dec 57 29.99 16.06 0 1.10 80 31 19.4 3.5
14-Dec 58 29.42 16.26 0 0.82 92 41 12.5 4.2
IS-Dec 59 27.63 18.01 0 1.38 98 51 16.0 4.0
16-Dec 60 29.83 16.24 0 1.07 96 27 19.6 4.2
17-Dec 61 28.86 14.02 0 1.13 83 23 20.5 3.8
18-Dec 62 28.67 15.95 0 25.0 1.21 69 34 19.8 3.1
19-Dec 63 29.04 17.31 0 1.22 74 38 19.7 3.3
20-Dec 64 29.85 17.16 0 1.02 82 37 17.9 3.5
21-Dec 65 28.31 16.88 0 1.16 82 35 16.7 3.2
22-Dec 66 22.43 17.5 0 0.89 97 67 6.7 3.0
23-Dec 67 25.4 18.69 0 0.99 91 54 8.7 3.8
24-Dec 68 26.55 17.97 0 25.0 0.84 92 55 11.3 3.6
25-Dec 69 22.11 18.1 0 0.72 99 73 3.8 3.9
26-Dec 70 21.41 17.46 0 0.59 99 80 3.5 3.6
27-Dec 71 23.94 16.78 3.6 0.88 99 71 9.9 3.7
28-Dec 72 25.l1 16.62 2.6 0.84 99 65 11.4 3.2
29-Dec 73 24.76 18.26 0 1.01 98 60 10.3 3.1
30-Dec 74 24.69 15.12 0 0.93 99 62 9.4 4.0
31-Dec 75 26.72 14.49 5.2 0.73 98 49 16.5 3.7
I-Jan 76 25.35 15.98 0 0.88 96 50 13.1 4.4
2-Jan 77 26.44 15.88 0 0.82 92 45 l3.7 4.8
3-Jan 78 27.9 l3.91 0 0.86 97 39 18.2 4.5
4-Jan 79 25.52 l3.74 0 25.0 0.89 99 48 14.0 4.5
5-Jan 80 25.31 14.65 0 0.98 93 50 10.0 3.6
6-Jan 81 26.55 13.55 0 1.32 97 30 20.1 4.2
7-Jan 82 25.55 8.31 0 1.15 88 26 21.9 4.6
8-Jan 83 26.15 11.16 0 1.06 90 22 21.1 4.8
9-Jan 84 27.8 12.6 0 1.26 65 25 21.0 4.9
lO-Jan 85 27.86 14.17 0 25.0 1.28 65 18 21.7 3.9
Il-Jan 86 27.29 12.63 0 1.28 69 7 22.7 4.4
12-Jan 87 26.56 12 0 1.21 77 13 22.7 4.4
13-Jan 88 26.63 13.l7 0 1.32 76 14 21.8 4.2
14-Jan 89 26.7 10.18 0 16.7 1.06 78 20 22.0 4.3
IS-Jan 90 27.89 11.36 0 1.09 76 31 21.4 4.5
16-Jan 91 28.73 11.08 0 0.89 88 33 20.9 4.0
17-Jan 92 28.53 14.49 0 1.05 87 37 20.6 4.5
18-Jan 93 26.65 15.88 0 16.7 0.74 90 46 10.5 4.4
19-Jan 94 28.78 15.78 0 1.00 95 44 15.3 4.6
20-Jan 95 31.01 15.09 5.8 1.10 99 34 19.1 4.5
21-Jan 96 30.18 15.5 0.2 1.16 99 37 16.6 4.4
22-Jan 97 29.08 16.03 12.6 1.45 99 42 16.9 3.3
23-Jan 98 29.38 14.57 0.2 1.06 99 38 21.1 4.8
24-Jan 99 28.65 17.01 0 1.11 87 43 19.6 4.4
25-Jan 100 31.18 15.54 0 1.04 95 33 21.3 4.5
26-Jan 101 30.18 16.04 0 1.07 92 38 20.l 4.2
27-Jan 102 30.85 15.77 0 1.23 98 37 21.2 4.8
28-Jan 103 32.36 21.16 0 1.61 85 32 18.l 3.5
29-Jan 104 31.53 18.85 0 2.09 79 27 19.9 4.6
30-Jan 105 29.93 14.29 0 1.11 93 32 21.7 4.4
31-Jan 106 30.19 18.48 0 1.39 76 30 21.7 4.4