Heritability of Drought Resistance
RESEARCH
D
rought is the major abiotic constraint aff ecting peanut (Ara-chis hypogaea L.) productivity and quality worldwide. Two-thirds of the global production occurs in rain-fed regions of the semi-arid tropics where rainfall is generally erratic and insuffi cient, causing unpredictable drought stress, the most important constraint for peanut production (Wright and Nageswara Rao, 1994; Reddy et al., 2003). Even peanut grown under irrigation may experience drought because of limited water supply or because irrigation water is applied in amounts at frequencies less than optimal for plant growth. Improving water access and management are practically diffi cult since water is a scarce resource. Therefore, breeding for drought resistance is an important strategy in alleviating the prob-lem and off ers the best long-term solution. Selection of segregating populations under stress conditions has been a standard approach for developing cultivars with improved stress tolerance. While direct selection for yield under stressed conditions can be eff ec-tive, the limitations of this approach are high resource investment and poor repeatability of the results due to the large genotype ×
Heritability of Drought Resistance
Traits and Correlation of Drought Resistance
and Agronomic Traits in Peanut
P. Songsri, S. Jogloy,* T. Kesmala, N. Vorasoot, C. Akkasaeng, A. Patanothai, and C. C. Holbrook
ABSTRACT
Inheritance of traits is important for develop-ing effective breeding schemes for improving
desired traits. The aims of this study were to esti-mate the heritabilities (h 2) of drought resistance traits and the genotypic (r G ) and phenotypic (r P ) correlations between drought resistance traits and agronomic traits, and to examine the rela-tionships between drought resistance traits under stressed and nonstressed conditions. The 140 lines in the F 4:7 and F 4:8 generations from four peanut (Arachis hypogaea L.) crosses were tested under fi eld capacity (FC) and two-thirds available soil water (2/3 AW) in two fi eld experiments. Data were recorded for specifi c leaf area (SLA), SPAD chlorophyll meter read-ing (SCMR), and biomass, pod yield, harvest index, number of mature pods per plant, seed per pod, and seed size. The h 2 for biomass, pod yield, DTI (drought tolerance index) (pod yield), DTI (biomass), HI, SLA, and SCMR were high for all tested crosses (0.54–0.98). The r G (?0.61 and ?0.66) and r P (?0.61 and ?0.66) between SLA and SCMR were strong and negative under 2/3 AW and FC. Under 2/3 AW conditions, SCMR was positively correlated with pod yield and seed size. Compared to SLA, SCMR had higher r G and r P with pod yield, biomass, and other agronomics traits. Signifi cant correlations between FC and 2/3 AW conditions were found for pod yield, biomass, SCMR, and SLA, indicat-ing that these traits could be selected under FC or 2/3 AW conditions. SPAD chlorophyll meter reading, which is easy to measure, is potentially useful as a selection trait for drought resistance because of high h 2 and positive correlation with pod yield and agronomic traits.
P. Songsri, S. Jogloy, T. Kesmala, N. Vorasoot, C. Akkasaeng, and A. Patanothai, Dep. of Plant Science and Agricultural Resources, F act. of Agriculture, Khon Kaen University, Khon Kaen, 40002, Thailand; C.C. Holbrook, USDA-ARS, Coastal Plain Exp. Stn., P.O. Box 748, Tifton, GA, 31793. Received 25 Apr. 2008. *Corresponding author (sanun@kku.ac.th).
Abbreviations: 2/3 AW, two-thirds available soil water; BIO, biomass; DAS, days after sowing; DTI, drought tolerance index; E, environ-ment; FC, fi eld capacity; G, genotype; HI, harvest index; PY, pod yield; SCMR, SPAD chlorophyll meter reading; SLA, specifi c leaf area; TE, transpiration effi ciency; WUE, water use effi ciency; Y, year.
Published in Crop Sci. 48:2245–2253 (2008).doi: 10.2135/cropsci2008.04.0228? Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.
environment (G×E) interaction that results in slow breed-ing progress (Wright et al., 1996).
More rapid progress may be achieved by using physi-ological traits (Nigam et al., 2005) such as harvest index (HI) or water use effi ciency (WUE), specifi c leaf area (SLA), and SPAD chlorophyll meter reading (SCMR). Both SLA and SCMR have been used as surrogate traits for WUE (Wright et al., 1994; Nageswara Rao and Wright, 1994; Sheshshayee et al., 2006; Nigam et al., 2005). In a biological model, yield is explained to be a function of water transpired, WUE, and HI (ratio of economic yield to total biomass produced) (Passioura, 1986). Water use effi ciency, defi ned as total biomass production per unit of water transpired, is not an easy trait to measure and, therefore, is not practical for use in large-scale breeding programs for improving drought tolerance.
Water use effi ciency is negatively correlated with leaf carbon isotopic composition (Δ) in a range of crop spe-cies, including peanut (F arquhar et al., 1982; Hubick et al., 1986; Wright et al., 1988, 1994). While measurement of Δ is rapid, it is an expensive technique and may not be feasible in large segregating breeding populations, par-ticularly in developing countries. Specifi c leaf area, the ratio of leaf area to leaf dry weight, is negatively related to leaf thickness and Δ and hence WUE, over a wide range of cultivars and environments in peanut (Wright et al., 1994; Nageswara Rao and Wright, 1994). Signifi cant and high correlations between SLA and ribulose 1-5 bisphos-phate carboxylase (Rubisco) (Nageswara Rao et al., 1995) suggested that photosynthetic capacity per unit leaf area is the major factor contributing to variation in WUE in peanut. There are a few published reports suggesting the predominant role of additive gene eff ects in SLA inheri-tance (Nigam et al., 2001; Surihan et al., 2005). Heritabil-ity estimation of water transpired, transpiration effi ciency (TE), and HI has been reported that varied between crosses and traits (Cruickshank et al., 2004).
Nageswara Rao et al. (2001) reported signifi cant cor-relations among SCMR, SLA, and specifi c leaf nitrogen. A strong and positive relationship between SCMR and WUE was found in peanut (Sheshshayee et al., 2006). Specifi c leaf area and SCMR are negatively correlated (Nageswara Rao et al., 2001; Upadhyaya, 2005). Upadhyaya (2005) also reported genetic variation for SCMR in peanut.
Information on the inheritance of HI, SLA, and SCMR and the genetic correlations among these traits will be useful for planning a suitable breeding strategy for improving drought tolerance. Drought can alter the heritability estimates of these traits; therefore, genetic gain through conventional selection may be diff erent under drought and well-watered conditions. Genetic cor-relations between drought resistance traits and agronomic traits have to be studied in details under drought and well-watered conditions to evaluate correlated responses to selection of drought resistance traits on agronomic traits. Objectives of this study were to estimate (i) the heritabili-ties of drought resistance traits, (ii) the genotypic and phe-notypic correlations between drought resistance traits and agronomic traits in peanut under diff erent water levels, and (iii) the relationship between drought resistance traits under stressed and nonstressed condition.
MATERIALS AND METHODS
Genetics Materials
Four peanut F
1
hybrids (ICGV 98308 × ‘KK60-3’, ICGV 98324 × KK60-3, ICGV 98308 × ‘Tainan 9’, and ICGV 98324 × Tainan 9) were generated from the hybridization of two drought-resis-tant lines (ICGV 98308 and ICGV 98324; medium-maturing [110 d to maturity] and medium-seeded type), selected for low yield reduction, with two high-yielding cultivars, KK60-3 (late-maturing [120 d to maturity] and large-seeded type) and Tainan 9 (early-maturing [100 d to maturity] and medium-seeded type). ICGV 98324 and KK 60-3 are know to have high SCMR and low SLA, ICGV 98308 has moderate SLA and moderate SCMR, and Tainan 9 has high SLA and low SCMR under both stressed and nonstressed conditions. The F
1
seeds were planted and their seeds harvested in bulk for each cross. In the F
2
and F
3
gen-erations, two pods were kept for each plant and bulked for each cross. Line separation was performed in the F
4
generation. A total of 140 lines (35 lines for each cross) were randomly selected and multiplied in the F
5
and F
6
generation.
The 140 families from four crosses were evaluated in the F
4:7 and F
4:8
generations (F
4
–derived lines in the F
7
and F
8
genera-tions, respectively) under two soil moisture levels, fi eld capac-ity (FC) and two-thirds available soil water (2/3 AW), for 2 yr in dry season 2005–2006 and 2006–2007. A split-plot design with four replications was used for both years at the Field Crop Research Station, Faculty of Agriculture Khon Kaen Univer-sity, located in Khon Kaen province, Thailand (16°28′ latitude, 102°48′ longitude, 200 m above sea level) during November 2005 to March 2006, and repeated during November 2006 to April 2007. Soil type is Yasothon series (loamy sand, Ocix Paleustults), with an FC soil moisture of 11.0% and permanent wilting point of 4.6%. Two soil moisture levels, F C (11.0%) and 2/3 AW (8.8%), in 0 to 60 cm depth were assigned as main plots, and peanut lines were laid out in subplots. Each entry was planted in fi ve row plots 3.2 m long. Spacing was 50 cm between rows and 20 cm between hills within the row.
Crop Management
Land was prepared for planting by plowing three times. Lime (625 kg ha–1), phosphorus fertilizer as triple superphosphate (24.7 kg P ha–1), and potassium fertilizer as potassium chloride (31.1 kg K ha–1) were applied before planting. Seeds were treated with captan [3a,4,7,7a-tetrahydro-2-[(trichloromethyl)thio]-1H-isoin-dole-1, 3(2H)-dione] at the rate of 5 g kg–1 seed before plant-ing, and seeds of the large-seeded genotypes were also treated with ethrel (2-chloroethylphosphonic acid) 48% at the rate of 2 mL L–1 water to break dormancy. Three to four seeds were planted per hill, and the seedlings were thinned to two plants per hill at 14 d after sowing (DAS). Rhizobium was applied to the
decline until reaching the predetermined levels of 2/3 AW at 0
to 60 cm at 28 DAS, then held more or less constant until har-vest. In maintaining the specifi ed soil moisture levels, water was added to the respective plots by subsurface drip irrigation based on crop water requirement and surface evaporation, which were
calculated following the methods described by Doorenbos and
Pruitt (1992) and Singh and Russell (1981), respectively.Total crop water use for each water treatment was calcu-lated as the sum of transpiration and soil evaporation. Transpira-tion was calculated using the methods described by Doorenbos and Pruitt (1992):ET crop = ET o K c where ET crop is crop water requirement (mm d –1), ET o is evapo-transpiration of a reference plant under specifi ed conditions cal-culated by pan evaporation method, and K c is the crop water
requirement coeffi cient for peanut, which varies with genotype
and growth stage (Doorenbos and Kassam, 1986). Surface evap-oration (E s ) was calculated as (Singh and Russell, 1981)
E s = β(E o /t )where E s is soil evaporation (mm), β is light transmission coef-fi cient measured depending on crop cover, E o is evaporation
from class A pan (mm d –1), and t is days from the last irrigation or rain.Data Collection Weather Parameters
Weather data for both years were obtained from a meteorologi-cal station about 30 m away from the experimental site and are
presented in Fig. 1.
seed by applying a water-diluted commercial peat-based inocu-lum of Bradyrhizobium (mixture of strains THA 201 and THA 205; Department of Agriculture, Ministry of Agriculture and Cooperatives, Bangkok, Thailand) on the rows of peanut plants. Weeds were controlled by an application of alachlor [2-chloro-2’, 6’-diethyl-N -(methoxymethyl) acetanilide 48%, w v –1
, emulsi-fi able concentrate] at the rate of 3 L ha –1
at planting and hand weeding during the remainder of the season. Gypsum (CaSO 4) at the rate of 312 kg ha –1
was applied at 45 DAS. Carbofuran (2,3-dihydro-2, 2-dimethylbenzofuran-7-ylmethylcarbamate, 3% granular) was applied at the pod-setting stage. Pests and diseases were controlled by weekly applications of carbosulfan [2-3-dihydro-2,2-dimethylbenzofuran-7-yl (dibutylaminothio)
methylcarbamate 20% w v –1, water soluble concentrate] at 2.5
L ha –1, methomyl [S -methyl-N -((methylcarbamoyl) oxy) thio-acetimidate 40% soluble powder] at 1.0 kg ha –1 and carboxin
[5,6-dihydro-2-methyl-1,4-oxathiine-3-carboxanilide 75%
wettable powder] at 1.68 kg ha –1.
A subsoil drip-irrigation system (Super Typhoon, Netafi m
Irrigation Equipment & Drip Systems, Israel), with a distance
of 20 cm between emitters was installed with a spacing of 50 cm between drip lines at 10 cm below the soil surface mid-way between peanut rows and fi tted with a pressure valve and water meter to ensure a uniform supply of measured amounts of water across each plot. Soil moisture was initially maintained at fi eld capacity (102.63 mm in 60 cm depth) until 21 DAS in all treatments to support crop establishment. After 21 DAS, the
2/3 AW treatment was imposed by withholding irrigation until
the soil moisture at 0 to 60 cm of soil depth was reduced to the
predetermined levels of 82.57 mm at 60 cm depth. Afterward,
soil moistures for the stress treatment was allowed to gradually
Figure 1. Rainfall, evaporation (E O ), relative humidity (RH), maximum and minimum air temperature, and solar radiation (SR) in (a, b)
2005–2006 and (c, d) 2006–2007 in Khon Kaen, Thailand.
The fi eld trials were conducted during the dry seasons from November 2005 to March 2006 and November 2006 to April 2007. There was maximum rainfall (13.0 mm) at 95 DAS in the dry season 2005–2006, and (39 mm) at 97 DAS in the dry season 2006–2007 (F ig. 1). The seasonal mean maximum and minimum air temperature ranged between 32.0°C and 20.0°C in 2005–2006 and 33.0°C and 20.0°C in 2006–2007. Daily pan evaporation ranged from 2.8 to 9.6 mm in 2005–2006 and 2.9 to 9.8 mm in 2006–2007. Seasonal mean solar radiation was 16.7 MJ m–2 d–1 in 2005–2006 and, 18.8 MJ m–2 d–1 in 2006–2007. Soil Moisture Status
Soil moistures were measured by the gravimetric method at planting and harvesting at the depths of 0 to 5, 25 to 30, and 55 to 60 cm. The measurement at planting was for calculating the correct amount of water to be applied to the crop, and the measurement at harvest was for calculating the water use of the crop. The soil water status was also monitored at 7-d intervals using a neutron moisture meter (Type I.H. II SER. No. N0152, Ambe Didcot Instruments Co., Abingdon, Oxon, UK). Six-teen-second neutron moisture meter readings were made at least weekly from a depth of 0.3 to 0.9 m at 0.3-m intervals. SPAD Chlorophyll Meter
Reading and Speci? c Leaf Area
In each plot, fi ve plants were randomly selected to record SCMR and SLA at 52, 67, 82, and 97 DAS following the pro-cedure described by Nageswara Rao et al. (2001). Briefl y, the second fully expanded leaves were detached from the chosen plants between 8:30 and 9:30 a.m. and brought to the labora-tory in zipped polythene bags for recording observations. The SPAD chlorophyll meter (Minolta SPAD-502 m, Tokyo, Japan) reading was recorded twice on each leafl et of the tetrafoliate leaf along the midrib. In recording the SCMR, care was taken to ensure that the SPAD meter sensor fully covered the leaf lamina and that interference from veins and midribs was avoided.
After recording SCMR, the leaf area of all fi ve sampled plants was measured with a leaf area meter (LI 3100C Area Meter, LI-COR Inc., Lincoln, NE) after which leaves were dried in an oven at 80°C for at least 48 h to determine the leaf dry weight. Immediately after drying, the leaves were weighed and the SLA was derived as leaf area per unit leaf dry weight (cm2 g–1). The SLA was calculated using the following formula: SLA = leaf area (cm2)/leaf dry weight (g).
Agronomic Traits
For each plot, three rows with 2.8 m in length
(4.2 m2) were harvested at maturity (R8)
(Boote, 1982), and their pods were removed
before taking fresh shoot weight in the fi eld. A
2-kg random sample of shoots was oven-dried
at 80°C for 48 h and dry weight was measured.
Shoot dry matter content was then calculated
and used in determining shoot dry weight for a
plot. Pod yields were weighed after air drying
to approximately 8% moisture content.
The number of mature pods per plant
(mature pods was separated from immature
pods, which were identifi ed by dark internal pericarp color), number of seed per pod and 100 seed weight were also recorded at fi nal harvest.
Harvest index was computed by the following formula: HI = total pod weight at the fi nal harvest/total biomass at the fi nal harvest.
Drought tolerance index (DTI), as suggested by Nautiyal et al. (2002), was calculated for biomass—DTI (BIO)—and pod yield—DTI (PY)—as the ratio of each parameter under stressed treat-ments (2/3 AW) to that under well-watered (FC) condition. Statistical Analysis
Individual analysis of variance was performed for each year fol-lowed a split-plot design (Gomez and Gomez, 1984). Homo-geneity of variance was tested for all characters and combined analysis of variance of 2-yr data was performed. Calculation procedures were conducted using MSTAT-C package (Bricker, 1989). Because water regime × genotype interaction was signifi-cant, each water regime was analyzed separately according to a randomized complete block design (Gomez and Gomez, 1984).
Estimates of broad-sense heritability for the four crosses were calculated by partitioning variance components of family mean squares to pooled environmental variance (δ2
E
) and geno-typic variance (δ2
G
), and then broad-sense heritability estimates (h2
b
) were calculated as follows (Holland et al., 2003):
h2
b
= δ2
G
/δ2
P
δ2
P
= δ2
G
+ δ2
GE
/e + δ2
E
/re
where δ2
G
is genotypic variation, δ2
P
is phenotypic variation, r is number of replications, and e is number of environments. The standard error of heritability (Singh et al., 1993) for drought resistance traits was calculated to give a measure of the preci-sion of the estimate.
Because the evaluation of heritability estimates was conducted in late generations (F
7
and F
8
) of segregating materials when most genes were nearly fi xed in individual genotypes, it would be expected that additive genetic variances for the traits under study were fi xed through generation advance (Holland, 2001).
Phenotypic and genotypic correlations between drought resistance traits and agronomic traits were calculated following the methods of Falconer and Mackay (1996) as follows (Table 1):
Phenotypic correlation (r
P
) = (M*
3
M
3
)/[(M*
3
) (M
3
)] 1/2
Genotypic correlation (r
G
) = (M*
3
M
3
? M*
2
M
2
)/
[(M*
3
? M*
2
) (M
3
? M
2
)] 1/2
Table 1. Analysis of variance of cross and cross product.
Source of variation df
Mean square
of character MCP?EMS?EMCP?
X Y
Year (Y)Y – 1 Rep. within Y Y(r – 1)?
Families (F) F – 1M*
3M
3
M*
3
M
3
δ2
E
+ rδ2
FE
+ reδ2
F
δ
E*E
+rδ
FE*FE
+reδ
F*F
F × Y(F ? l)(r ? 1)M*
2M
2
M*
2
M
2
δ2
E
+ rδ2
FE
δ
E*E
+ rδ
FE*FE
Pooled error Y(r ? 1)(F ? 1)M*
1M
1
M*
1
M
1
δ2
E
δ
E*E
?MCP, mean square of cross product; EMS, expected mean square; EMCP, expected mean square of cross product.
?r, number of replications.
where M* is mean square of character X and M is mean square of character Y . Simple correlation was used to determine the relationship between biomass, pod yield, and drought resistance traits under well-watered and drought conditions to understand whether the performance of peanut genotypes was consistent across environments.
RESULTS AND DISCUSSION
Monitoring of Soil Moisture
Soil moisture was measured with a neutron moisture meter at 7-d intervals until harvest (Fig. 2). The results showed reasonable management of soil moistures. A clear distinction
among soil moisture levels was noted at 30 cm of soil depth. Soil moistures at 90 cm depth were similar among treat-ments because the amount of water applied in each treat-ment was calculated for 0 to 60 cm. Visual wilting was observed in the 2/3 AW treatment in the afternoon.
Combined Analysis of Variance
Combined analysis of variance showed signifi cant diff er-ences among 140 progenies (P ≤ 0.01) for biomass produc-tion, pod yield, and the drought surrogate traits HI, SCMR, and SLA (Table 2). This indicated that genetic variation
exists for these characters and, thus, that heritability could
Figure 2. Soil moisture volume fraction in two available soil water regimes [? eld capacity (FC), ●; and 2/3 available water (AW), ] at (a, b) 30 cm, (c, d) 60 cm, and (e, f) 90 cm of the soil level during the 2005–2006 and 2006–2007 dry seasons in Khon Kaen, Thailand.
be estimated. The interaction eff ects of Y × G were sig-nifi cant (P ≤ 0.01) for pod yield under well-watered and 2/3 AW conditions and signifi cant (P ≤ 0.05) for HI under 2/3 AW condition. In theory, pod yield is a complex trait in which multiple genes are involved, and high G × E interaction is expected (Wright et al., 1996). Based on low coeffi cient of variation and high F-ratio from analysis of variance, the best assessment times for SLA and SCMR was determined to be 67 DAS.
Heritability of Drought Resistance Traits Heritability estimates within four peanut crosses were cal-culated for SCMR and SLA at 67 DAS, and for HI, DTI (BIO), and DTI (PY) at harvest (Table 3). Heritability esti-mates for HI, SLA, and SCMR were high for all four pea-nut crosses under both nonstressed and stressed conditions, ranging from 0.81 to 0.97. Drought tolerance indexes for pod yield and biomass showed lower heritability estimates than those for pod yield and biomass themselves under nonstressed and stressed conditions. Heritability estimates for BIO and PY varied from 0.73 to 0.98, and DTI (BIO) and DTI (PY) varied from 0.54 to 0.96.
Most characters had similar heritability estimates when compared between diff erent water levels. This should make selection for drought tolerance easier. However, DTI is still useful in explaining how some genotypes had higher pod yield under drought. Previous reports on inheritance of drought resistance traits suggested a predominant role of additive gene eff ects in SLA and HI inheritance (Nigam et al., 2001; Surihan et al., 2005). In early generations (F
3 and F
4
), Cruickshank et al. (2004) reported that broad-sense heritability of transpiration, TE, and HI were varied among peanut crosses and traits depending on levels of genetic variation in parents. Information on heritability of drought resistance traits [DTI (BIO), DTI (PY), HI, SCMR, and SLA)] under both stressed and nonstressed are needed for predicting progress from selection. Most of the drought resistance traits in our study had high heritabil-ity estimates, indicating that breeding progress could be achieved for these characters.
Table 2. Mean squares from the combined ANOVA for pod yield, biomass, and drought tolerance index for biomass, DTI (BIO),?and pod yield, DTI (PY), and harvest index (HI) at harvest, SPAD chlorophyll meter reading (SCMR), and speci? c leaf area (SLA) at 67 d after sowing under nonstressed (Non) and stressed (Stress) conditions of 140 peanut genotypes in the dry season of 2005–2006 and 2006–2007 in Khon Kaen, Thailand.
Source of variation df
Pod yield Biomass HI SCMR SLA Non Stress DTI (PY)Non Stress DTI (BIO)Non Stress Non Stress Non Stress
Year (Y)119.0310.57**0.17209.4139.110.490.000.04* 12.580.43969.73596.56 Rep. within Y 6 3.550.520.2140.077.270.190.010.0133.5938.00409.12153.09 Genotypes (G)139 2.95** 2.17**0.39**14.98**14.60**0.16**0.03**0.02**57.48**56.58**124.53**89.11** Y × G1390.13**0.10**0.030.66 1.830.040.000.00* 2.79 2.4811.68 5.17 Pooled error 8340.090.070.030.78 1.730.040.000.00 4.08 3.5814.047.77
*Signi? cant at P ≤ 0.05.
**Signi? cant at P ≤ 0.01.
?DTIs were calculated by the ratio of stressed (2/3 available water)/nonstressed (? eld capacity) conditions.
Table 3. Estimates of heritability with standard error for biomass (BIO), pod yield (PY), drought tolerance index for biomass, DTI? (BIO), and pod yield, DTI (PY), and harvest index (HI) at harvest and speci? c leaf area (SLA) and SPAD chlorophyll meter reading (SCMR) at 67 d after sowing of four crosses of peanut under stressed and nonstressed conditions in the dry seasons of 2005–2006 and 2006–2007 in Khon Kaen, Thailand.
Cross
Heritability
BIO PY DTI (BIO)DTI (PY)HI SLA SCMR
Stressed
ICGV 98308 × ‘KK60-3’0.94 ± 0.060.93 ± 0.070.93 ± 0.070.86 ± 0.110.94 ± 0.050.93 ± 0.070.89 ± 0.10 ICGV 98308 × ‘Tainan 9’0.81 ± 0.160.95 ± 0.050.54 ± 0.250.92 ± 0.070.89 ± 0.080.81 ± 0.150.96 ± 0.03 ICGV 98324 × ‘KK60-3’0.73 ± 0.200.93 ± 0.070.67 ± 0.210.87 ± 0.110.95 ± 0.040.91 ± 0.080.92 ± 0.08 ICGV 98324 × ‘Tainan 9’0.96 ± 0.040.97 ± 0.030.86 ± 0.120.96 ± 0.030.89 ± 0.080.95 ± 0.050.96 ± 0.04 Nonstressed
ICGV 98308 × ‘KK60-3’0.89 ± 0.120.91 ± 0.08——0.94 ± 0.040.83 ± 0.150.89 ± 0.11 ICGV 98308 × ‘Tainan 9’0.98 ± 0.020.98 ± 0.02——0.97 ± 0.020.91 ± 0.090.97 ± 0.02 ICGV 98324 × ‘KK60-3’0.93 ± 0.070.93 ± 0.06——0.92 ± 0.060.91 ± 0.090.90 ± 0.08 ICGV 98324 × ‘Tainan 9’0.98 ± 0.020.98 ± 0.01——0.96 ± 0.030.95 ± 0.050.96 ± 0.04?DTIs were calculated by the ratio of stressed (2/3 available water)/nonstressed (? eld capacity) conditions.
Genotypic Correlation among
Drought Resistance Traits
Phenotypic and genotypic correlations provided similar information in this study, and only genotypic correlations are reported. Strong and negative genotypic correlations were found between SLA and SCMR under both stressed and nonstressed conditions (?0.61, P ≤ 0.01, and ?0.66, P ≤ 0.01, respectively) (Table 4). In previous studies, the simple correlation between SLA and SCMR was reported under nonstressed conditions (Wright et al., 1994; Nag-eswara Rao et al., 2001; Upadhyaya, 2005) and end-of-season drought conditions (Nigam and Aruna, 2008). In this study, we evaluated material in both stressed and nonstressed conditions in the same trials. Our fi nding show that genotypic and phenotypic correlations between SLA and SCMR were consistent under both FC and 2/3 AW conditions. The results show consistency of SLA and SCMR in a wide range of soil water levels and drought conditions. Drought tolerance index for pod yield had strong and positive genotypic correlation with DTI (BIO) (0.69, P ≤ 0.01). Harvest index was quite low correlated with DTI (PY) under stressed condition (0.37, P ≤ 0.01) and also was correlated with SCMR both under drought and well-watered conditions (0.13, P ≤ 0.01, and 0.33, P ≤ 0.01, respectively).
Genotypic Correlation between
Drought Resistance Traits and Yield
and Yield Components
Genetic correlations between drought resistance traits and yield and yield components provide information on expected responses in yield and yield components from selection for drought resistance traits. High genotypic cor-relations were found for HI and PY under drought (0.76, P ≤ 0.01) and nonstressed (0.79, P ≤ 0.01) conditions, and for HI with the number of mature pods per plant under both stressed and nonstressed treatments (0.62, P ≤ 0.01, and 0.49, P ≤ 0.01, respectively) (Table 5). The genotypic
correlations between HI and seed size were also moderate and positive under both stressed and well-watered condi-tions (0.50, P ≤ 0.01, and 0.47, P ≤ 0.01, respectively). The surrogate traits for WUE (SLA and SCMR) (Wright et al., 1994; Nageswara Rao and Wright, 1994; Sheshshayee et al., 2006) had low correlation with pod yield. However, SCMR had higher genotypic correlations with PY, BIO, and other agronomic traits under both stressed and well-watered conditions than did SLA. SCMR showed quite low positive correlations with biomass (0.18; P ≤ 0.01) and pod yield (0.21; P ≤ 0.01) under stressed and mod-erate positive correlations with BIO (0.41; P ≤ 0.01) and PY(0.51; P ≤ 0.01) under well-watered conditions. SPAD chlorophyll meter reading was moderate positively cor-related with seed size under stressed (0.43, P ≤ 0.01) and well-watered (0.48; P ≤ 0.01) conditions. DTI (BIO) and DTI (PY) had moderate positive correlations with bio-mass (0.47, P ≤ 0.01, and 0.52, P ≤ 0.01, respectively), with pod yield (0.34, P ≤ 0.01, and 0.57, P ≤ 0.01, respectively), and with number of mature pods per plant (0.34, P ≤ 0.01, and 0.45, P ≤ 0.01, respectively) under drought conditions. SPAD chlorophyll meter reading and SLA were strongly and negatively correlated at all evaluation dates (data not shown), and this association was relatively stable across environments (stressed and well-watered).
Among drought resistance traits [DTI (BIO), DTI (PY), HI, SCMR and SLA], HI had the highest correla-tion with PY, but the measurement of HI was more dif-fi cult, laborious, and costly than that of PY. Also, genetic correlations between SCMR and PY and HI were low. However, these traits have lower G × E interaction than do yield (Wright et al., 1996). It would be possible to Table 4. Genotypic (r
G
) correlation estimates among drought resistance traits for all four peanut crosses of 140 genotypes in the dry seasons of 2005–2006 and 2006–2007 in Khon Kaen, Thailand (df = 556).?
Stressed Nonstressed DTI? (PY)SCMR SLA HI SCMR SLA HI DTI (BIO)0.69**–0.34**0.050.06———DTI (PY)–0.28**0.060.37**———SCMR–0.61**0.13**–0.66**0.33** SLA0.11*–0.10*
*Signi? cant at P ≤ 0.05.
**Signi? cant at P ≤ 0.01.
?DTI, drought tolerance index; BIO, biomass, PY, pod yield, SCMR, SPAD chloro-phyll meter reading; SLA, speci? c leaf area; HI, harvest index.
?DTIS were calculated by the ratio of stressed (2/3 available water)/nonstressed (? eld capacity) conditions.
Table 5. Genotypic (r
G
) correlation estimates between drought resistance traits and agronomic traits for all four peanut cross of 140 genotypes in the dry seasons of 2005–2006 and 2006–2007 in Khon Kaen, Thailand (df = 556).?
Drought
resistance
traits
Agronomic traits
BIO PY
Seed
size
No. mature
pods/plant
Seed/
pod Stressed
DTI? (BIO)0.47**0.34**0.010.34**0.29** DTI (PY)0.52**0.57**0.25**0.45**0.14** SCMR0.18**0.21**0.43**–0.20**–0.04
SLA0.070.070.060.040.10*
HI0.19**0.76**0.50**0.62**0.16** Nonstressed
SCMR0.41**0.51**0.48**0.020.24** SLA0.01–0.09*–0.12**0.020.06
HI0.010.79**0.47**0.49**0.26** *Signi? cant at P ≤ 0.05.
**Signi? cant at P ≤ 0.01.
?DTI, drought tolerance index; BIO, biomass, PY, pod yield, SCMR, SPAD chloro-phyll meter reading; SLA, speci? c leaf area; HI, harvest index.
?DTI were calculated by the ratio of stressed (2/3 available water)/nonstressed (? eld capacity) conditions.
improve yield by selecting for high HI and SCMR. The SCMR is an indicator of the photosynthetically active light-transmittance characteristics of the leaf and posi-tive correlated with chlorophyll content (Akkasaeng et al., 2003) and chlorophyll density (Arunyanark et al., 2008) and WUE (Sheshshayee et al., 2006).
Nonetheless, the integration of physiological traits (or their surrogates) in the selection scheme would be advanta-geous in selecting genotypes that are more effi cient water utilizers (SCMR [surrogates trait]) or partitioners of pho-tosynthates into economic yield (HI) (Nigam et al., 2005). The SPAD chlorophyll meter provides an easy opportu-nity to integrate a surrogate measure of WUE with PY, in the selection scheme of a drought resistance breeding program in peanut.
Relationship of Drought Resistance
Traits under Well-Watered versus
Drought Conditions
A comparison of drought resistance traits under well-watered versus drought conditions should provided a better understanding of the most suitable conditions for selecting drought resistant genotypes. Signifi cant cor-relations between traits under stressed and nonstressed conditions were found in all four peanut crosses for HI, SCMR, SLA, PY, and BIO (Table 6), indicating that the-ses traits could be selected either under well-watered or water-stressed conditions. As heritability estimates were high under both well-watered and stress conditions and the traits under diff erent water regimes were correlated well, it is advisable to fi rst select peanut genotypes under well-watered conditions in large early segregating popula-tions because drought simulation is much more diffi cult; later, the selections can be refi ned under both drought and nonstressed conditions in advanced generations. CONCLUSION
In summary, most traits measured in these four peanut crosses had high heritability, indicating that breeding progress should be possible. The results of the present study indicated that harvest index, SPAD chlorophyll
meter reading, and specifi c leaf area observa-
tions can be recorded at both stressed and non-
stressed conditions. This gives peanut breeders
a large fl exibility to record these observations
in a large number of segregating populations
and breeding lines in the fi eld, thus making it
easy to incorporate these physiological traits associated with drought tolerance in breeding
and selection schemes in peanut. SPAD chloro-
phyll meter reading should be particularly use-
ful as a selection criterion for drought tolerance
in peanut because of high heritability and the simplicity in gathering.Acknowledgments
The authors are grateful for the fi nancial support of the Royal Golden Jubilee PhD Program (Grant no. PHD/0190/2544) and the Senior Research Scholar Project of Professor Dr. Aran Patanothai, and also partial support by the Basic Research for Supporting Groundnut Varietal Improvement for Drought Tol-erance Project under the Thailand Research Fund. We thank the many people who work in fi eld collecting data and process-ing samples.
References
Akkasaeng, C., N. Vorasoot, S. Jogloy, and A. Patanotai. 2003.
Relationship between SPAD readings and chlorophyll con-tents in leaves of peanut (Arachis hypogaea L.). Thai J. Agric.
Sci. 36(3):279–284.
Arunyanark, A., S. Jogloy, C. Akkasaeng, N. Vorasoot, T. Kes-mala, R.C. Nageswara Rao, G.C. Wright, and A. Patanothai.
2008. Chlorophyll stability is an indicator of drought toler-ance in peanut. J. Agron. Crop Sci. 194:113–125.
Boote, K.J. 1982. Growth stage of peanut (Arachis hypogaea L.).
Peanut Sci. 9:35–40.
Bricker, A.A. 1989. MSTAT-C user’s guide. Michigan State Univ., East Lansing.
Cruickshank, A., A. Dowkiw, G.C. Wright, C.R. Nageswara Rao, and S.N. Nigam. 2004. Heritability of drought–resistance traits in peanut. In T. Fischer, N. Turner, J. Angus, L. McIn-tyre, M. Robertson, A. Borrell, D. Lloyd (ed.) New directions for a diverse planet: Proceedings for the 4th Int. Crop Science Congress, Brisbane, Australia. 26 Sept.–1 Oct. 2004. Doorenbos, J., and A.H. Kassam. 1986. Groundnut. p. 97–100. In Yield response to water. FAO Irrigation and Drainage Paper
33. FAO, Rome.
Doorenbos, J., and W.O. Pruitt. 1992. Calculation of crop water requirements. p. 1–65. In Guidelines for predicting crop water requirements. FAO Irrigation and Drainage Paper 24.
FAO, Rome.
Falconer, D.S., and T.F.C. Mackay. 1996. Introduction to quanti-tative genetics. 4th ed. Longman, London.
F arquhar, G.D., M.H. O’Leary, and J.A. Berry. 1982. On the
relationship between carbon isotope discrimination and the intercellular carbon dioxide concentration in leaves. Aust. J.
Plant Physiol. 9:121–137.
Gomez, K.A., and A.A. Gomez. 1984. Statistical procedures for agricultural research. 2nd ed. John Wiley & Sons, New York.
Table 6. Correlation coef? cients of biomass (BIO), pod yield (PY), harvest index (HI), speci?c leaf area (SLA), and SPAD chlorophyll meter reading (SCMR) of four peanut crosses under stressed (d) and well-watered (w) con-ditions during 2005–2006 and 2006–2007 in Khon Kaen, Thailand (df = 33).
Correlation
Peanut cross
ICGV 98308 ×
‘KK60-3’
ICGV 98308 ×
‘Tainan 9’
ICGV 98324 ×
‘KK60-3’
ICGV 98324 ×
‘Tainan 9’
BIO
w vs. BIO
d
0.48**0.79**0.62**0.84**
PY
w vs. PY
d
0.35* 0.73**0.61**0.71**
HI
w vs. HI
d
0.75**0.62**0.58**0.46**
SCMR
w vs. SCMR
d
0.73**0.84**0.53**0.86**
SLA
w vs. SLA
d
0.59**0.35*0.52**0.86**
*Signi? cant at P ≤ 0.05. **Signi? cant at P ≤ 0.01.
Holland, J.B. 2001. Epistasis and plant breeding. Plant Breed. Rev.
21:27–92.
Holland, J.B., W.E. Nyquist, and C.T. Cervartes-Martinez. 2003.
Estimating and interpreting heritability for plant breeding: An update. Plant Breed. Rev. 22:2–112.
Hubick, K.T., G.D. Farquhar, and R. Shorter. 1986. Correlation between water-use effi ciency and carbon isotope discrimi-nation in diverse peanut (Arachis) germplasm. Aust. J. Plant Physiol. 13:803–816.
Nageswara Rao, R.C., H.S. Talwar, and G.C. Wright. 2001.
Rapid assessment of specifi c leaf area and leaf nitrogen in pea-nut (Arachis hypogaea L.) using a chlorophyll meter. J. Agron.
Crop Sci. 186:175–182.
Nageswara Rao, R.C., M. Udaykumar, G.D. F arquhar, H.S.
Talwar, and T.G. Prasad. 1995. Variation in carbon isotope discrimination and its relationship to specifi c leaf area and ribulose–1, 5–bisphosphate carboxylase content in groundnut genotypes. Aust. J. Plant Physiol. 22:545–551.
Nageswara Rao, R.C., and G.C. Wright. 1994. Stability of the rela-tionship between specifi c leaf area and carbon isotope discrimi-nation across environments in peanut. Crop Sci. 34:98–103. Nautiyal, P.C., R.C. Nageswara Rao, and Y.C. Joshi. 2002. Mois-ture-defi cit–induced changes in leaf water content, leaf carbon exchange rate and biomass production in groundnut cultivars diff ering in specifi c leaf area. Field Crops Res. 74:67–79. Nigam, S. N., and R. Aruna. 2008. Stability of soil plant analytical development (SPAD) chlorophyll meter reading (SCMR) and specifi c leaf area (SLA) and their association across varying soil moisture stress conditions in groundnut (Arachis hypogaea L.). Euphytica 160:111–117.
Nigam, S.N., H.D. Upadhyaya, S. Chandra, R.C. Nageswara Rao, G.C. Wright, and A.G.S. Reddy. 2001. Gene eff ects for specifi c leaf area and harvest index in three crosses of ground-nut (Arachis hypogaea). Ann. Appl. Biol. 139:301–306. Nigam, S.N., S. Chandra, K. Rupa Sridevi, A. Manoha Bhukta,
G.S. Reddy, R.C. Nageswara Rao, G.C. Wright, P.V. Reddy,
M.P. Deshmukh, R.K. Mathur, M.S. Basu, S. Vasundhara, P. Vindhiya Varman, and A.K. Nagda. 2005. Effi ciency of physiological trait-based and empirical selection approaches
for drought tolerance in groundnut. Ann. Appl. Biol.
146:433–439.
Passioura, J.B. 1986. Resistance to drought and salinity: Avenues for improvement. Aust. J. Plant Physiol. 13:191–201. Reddy, T.Y., V.R. Reddy, and V. Anbumozhi. 2003. Physiological responses of peanut (Arachis hypogaea L.) to drought stress and its amelioration: A critical review. Plant Growth Regul. 41:75–88. Sheshshayee, M.S., H. Bindumadhava, N.R. Rachaputi, T.G.
Prasad, M. Udayakumar, G.C. Wright, and S.N. Nigam.
2006. Leaf chlorophyll concentration relates to transpiration effi ciency in peanut. Ann. Appl. Biol. 148:7–15.
Singh, S., and M.B. Russell. 1981. Water use by maize/pigeon-pea intercrop on a deep Vertisol. p. 271–282. In Proc. Int.
Workshop on Pigeonpeas. 15–19 Dec 1980. Vol. 1. ICRISAT Cente. Patancheru, AP, India.
Singh, M., S. Ceccarelli, and J. Hamblin. 1993. Estimation of herita-bility from varietal trials data. Theor. Appl. Genet. 86:437–441. Surihan, B., A. Patanothai, and S. Jogloy. 2005. Gene eff ect for specifi c leaf area and harvest index in peanut (Arachis hypogaea L.). Asian J. Plant Sci. 4(6):667–672.
Upadhyaya, H.D. 2005. Variability for drought resistance related traits in the mini core collection of peanut. Crop Sci. 45:1432–1440. Wright, G.C., K.T. Hubick, and G.D. Farquhar. 1988. Discrimi-nation in carbon isotopes of leaves correlates with water-use effi ciency of fi eld-grown peanut cultivars. Aust. J. Plant Phys-iol. 15:815–825.
Wright, G.C., and R.C. Nageswara Rao. 1994. Peanut water relations. p. 281–325. In J. Smartt (ed.) The groundnut crop.
Chapman & Hall, London.
Wright, G.C., R.C. Nageswara Rao, and M.S. Basu. 1996. A physiological approach to the understanding of genotype by environment interactions: A case study on improvement of drought adaptation in groundnut. p. 365–380. In M. Cooper and G.L. Hammer (ed.) Plant adaptation and crop improve-ment. CAB Int., Wallingford, UK.
Wright, G.C., R.C. Nageswara Rao, and G.D. F arquhar. 1994.
Water use effi ciency and carbon isotope discrimination in peanut under water defi cit conditions. Crop Sci. 34:92–97.