Printer Friendly

ASSESSMENT OF PARAMETRIC AND NON-PARAMETRIC METHODS FOR SELECTING STABLE AND ADAPTED SPRING BREAD WHEAT GENOTYPES IN MULTI - ENVIRONMENTS.

Byline: Hasan Kilic

ABSTRACT

The objectives of this research were to assess genotype-environment interaction (GEI) and to determine stable bread wheat (Triticum aestivum L.) genotypes for grain yield in the South-Eastern Anatolia region of Turkey during the 2004-2007 growing seasons. Twenty five bread wheat genotypes were evaluated in a randomized complete block design with 4 replications. Genotypes, environments main effects and GEI were significant at P less than 0.01. Both parametric (bi, S2di, Ri2, Wi2, Si2, CVi, si2, P59, ai, li and Pi) and non-parametric (Si(1), Si(2) Si(3) Si(6), TOP and RS) univariate stability statistics were used to determine stability of the bread wheat genotypes. Genotypes G16, G6, G3, G5 and G20 were the most stable based on parametric and non-parametric stability measures used. The level of association among the stability measures was assessed using Spearman's rank correlation.

The rank correlation matrix indicated that most n on-parametric measures were significantly inter-correlated with parametric measures and therefore can be used as alternatives.

Key words: Bread wheat, GE interaction, grain yield, parametric and non-parametric measures and stability.

INTRODUCTION

Wheat is the first important cereal crop of Turkey and now accounts for about 75% of the total cereal production with coverage of 11.9 million hectares (Anonymous, 2008). Bread wheat occupies first place and is grown at about 85% all the wheat cultivated area in Turkey. In South-Eastern Anatolian region, drought occurs when there is irregular precipitation, which causes yield losses in varieties sensitive to drought. Some diseases, such as root rot, yellow and leaf rust negatively affect wheat yield. Furthermore, grain quality is deteriorated significantly by Sunn Pest (Eurygaster spp. Heteroptera:Scutelleridae).

The development of varieties which can be adapted to a wide range of environments is the ultimate goal of plant breeders in a crop improvement program. A low level of interaction with unpredictable variables such as adverse weather conditions would give more uniform and stable yields, whereas a high level of interaction with a controllable variable such as fertilizer application, is desirable (Mohebodini et al., 2006).

High yield stability usually refers to a genotype's ability to perform consistently, whether at high or low yield levels, across a wide range of environments (Annicchiarico, 2002). Although several models for the statistical measurement of the stability have been suggested, each of which show different aspects of stability but not a single method can adequately explain cultivar performance across environments, such as regression slope value (Finlay and Wilkinson, 1963) and deviation from regression (Eberhart and Russell, 1966). Though there are well- recognized statistical and biological limitations in the regression approach (Lin et al., 1986; Westcott, 1987; Crossa, 1990; Flores, 1993), it ensures beneficial parameter estimates when number of genotypes and environments considered in the analyses are sufficiently large and when there are no extreme environments that affect regression slopes (Flores, 1993).

Some other univariate stability parameters are; environmental variance (S2xi) (Roemer, 1917 attributed in Becker and Leon, 1988), superiority index (PI) (Lin and Binns, 1988); Plaisted and Peterson's (1959) mean variance component for a pair-wise genotype-environment interaction (P59), Plaisted's (1960) variance component for GE interaction (Oi), Wricke's (1962) ecovalence (Wi2), Shukla's (1972) stability variance (si2), Francis and Kanenberg's (1978) coefficient of variability (CVi), Freeman and Perkins (1971) stability method, Hanson's (1970) genotypic stability(Di2). All these mentioned methods are parametric. When parametric methods are used for stability, estimations are made about the range of data and the uniformity of variance.

However, non-parametric stability measures are not generally affected by data distribution. As these methods are based on ranks and not on values, a genotype is considered stable if its ranking is relatively constant across environments (Flores et al., 1998). Several non- parametric methods have been improved to explain and interpret the responses of genotypes to environmental variation (Huehn, 1979; Kang, 1988; Ketata et al., 1989; Fox et al., 1990).

The level of association among adaptability or stability estimates of different models is indicative of whether one or more estimates should be obtained for reliable prediction of cultivar behaviour, and also helps the breeder to choose the best adjusted and most informative stability parameter(s) to fit his/her concept of stability (Duarte and Zimmermann, 1995).

The objectives of this study were to (i) compare parametric and nonparametric stability statistics of 25 bread wheat genotypes' data collected from a number of different agro-climatic conditions, (ii) determine promising genotypes with high yield and stability (iii) evaluate the level of associations among the parametric and nonparametric stability parameters.

MATERIALS AND METHODS

Twenty-five bread wheat genotypes were grown in 10 environments, Diyarbakir rainfed, Diyarbakir irrigated, Hazro, Ceylanpinar and Mardin during the 2004-2007 growing seasons at the South-Eastern Anatolian in Turkey. The 25 genotypes comprised 20 advanced lines from National program, CIMMYT (International Maize and Wheat Improvement Center) and ICARDA (International Center for Agricultural Research in the Dry Areas) and 5 Turkish cultivars. The experimental layout was a randomized complete block design with 4 replications. Sowing was done with an experimental drill in 1.2 m x 7 m plots, consisting of 6 rows spaced 20 cm apart. The seeding density was 450 seeds per m-2. Harvesting was done in 1.2 m x 5 m plots by combine harvester and yield was determined and expressed in t ha-1. Details of the 25 genotypes and means of grain yield are given in Table 1.

The growing seasons, environments, soil properties together with details of fertilizer application are given Table 2. Amounts of rainfall, together with supplementary irrigation applied at each location during the growing period are also given Table 2.

Code Pedigree and selection history of cultivars-lines###Code###Pedigree and selection history of cultivars-lines

G1###ZG-1004.82/KATYA-A.1###G14###SHUHA-7/4/NIF/3/SOTY//NAD63/CHRIS

###GD 2807-0D-0D-0D-0D-0D###ICW92-0671-4AP-0L-2AP-0L-1AP-0AP

G2###COQ/NAC 2//F.12-71/COC###G15###GONEN-98

###CMBW 89-Y-01530-OTOPM-1D-0D-0D-0D-0D

G3###MANGO 2/5/SU.92/CI.13465//PPEN/3/PHO/4/###G16###SHUHA-6/3/RMN F12-71/SKA//CA8055

###YMH/TOB//BEZ###ICW92-0717-1AP-0BR-4AP-0AP

###CMBW 89-Y-01588-OTOPM-3D-0D-0D-0D-0D###G17###HP 1731 - (RAJLAXMIN)- OIND

G4###BOW"S"/MOR"S"//OPATO/BOW"S"

###CMBW 89-Y-2584-5D-0D-0D-0D-0D###

G5###NURKENT###G18###F6.74/BUN//SIS/3/LIRA

###CM90561-21Y-0M-0Y-2M-0Y-2M-0Y

G6###KT/BAPE//FN/GU/3/BZA/4/TRM/5/ALDEN/6/S###G19###OPATA/RAYON//KAUZ CMBW90Y3180-0

###ERI/7/BOW###TOPM-3Y-010M-010M-010Y-1M-...

###CMBW 89-Y-00830-OTOPM-4D-0D-0D-0D-0D###

G7###AROONA 3/YR 15 (S)###G20###ADANA-99

G8###BEI JING 411###G21###CAZO/KAUZ//KAUZCMBW90Y3284-0

###TOPM-14Y-010M-010M-010Y-6M- ...

G9###440259/AKHALTSIKHIS TSITELI DOLI###G22###SKAUZ 2/SRMA CMBW91M02694F-0

###TOPY-12M-010Y-010M-010Y- ...

G10###PEHLIVAN###G23###SW89.3064 2/BORL65 CMBW91M03786F-0

###TOPY-17M-010Y-010M-015Y- ...

G11###MOMCHIL/KATYA 1 LS316###G24###BOW/PRL//BUC/3/LUAN

###CMSS93Y00118S-105Y-3B-3Y-0100B

G12###TUI//CMH76-252/PVN"S"###G25###BASRIBEY-95

###ICW92-0214-0AP-1AP-2AP-0AP

G13###TUI//CMH76-252/PVN"S"ICW92-0214-0AP-

###1AP-3AP-2AP-0AP

Table 2. Some information on experiment, soil properties and climate for environments where the experiments were conducted

Code###Growing###Environments###Soil properties###Fertilization###Rain-fall Irrigation###Mean

###seasons###kg ha-1###mm###mm###yield t ha-1

###N###P2O5

E1###2004-05###Diyarbakir (rainfed)###pH = 7.43 clay-silt###60a + 60b###60a###389.4###-###5.74

E2###2004-05###Diyarbakir (irrigated)###pH = 7.61 clay-silt###80 + 60###80###389.4###100###7.65

E3###2004-05###Hazro (rainfed)###pH = 7.50 clay-silt###60 + 60###60###NA###-###5.81

E4###2004-05###Ceylanpinar (rainfed)###pH = 7.80 clay-silt###60 + 60###40###323.5###-###2.72

E5###2005-06###Diyarbakir- (rainfed)###pH = 7.50 clay-silt###60 + 60###60###534.2###-###7.64

E6###2005-06###Diyarbakir- (irrigated)###pH = 7.69 clay-silt###80 + 60###80###534.2###100###8.15

E7###2005-06###Hazro (rainfed)###pH = 7.56 clay-silt###60 + 60###60###NA###-###4.77

E8###2006-07###Diyarbakir- (rainfed)###pH = 7.30 clay-silt###60 + 60###60###534.2###-###6.09

E9###2006-07###Diyarbakir- (irrigated)###pH = 7.43 clay-silt###80 + 60###80###534.2###100###6.93

E10###2006-07###Mardin (rainfed)###pH = 7.58 clay-silt###60 + 60###50###319.0###-###3.16

a: Sed bed b: Stem elongation , data not available.

Statistical analyses: A combined analysis of variance was first undertaken across the test environments. Then ten parametric stability parameters were performed in accordance with Eberhart and Russel's (1966) the regression coefficient (bi) and deviation from regression (S2di), Pinthus's (1973) coefficients of determination (R2), Wricks's (1962) ecovalance (Wi2), Shukla's (1972) stability variance (si2), Francis and Kannenberg's (1978) coefficient of variability (CVi) and genotypic variance (S2i); Tai's (1971) environmental effects (ai) and deviation from the linear response (li); Plaisted and Peterson's (1959) mean variance component for pair-wise GEI (P59).

Lin and Binn's (1988) superiority index (Pi) which the genotypes of greatest interest would be those with the lowest Pi values, Two sets of non-parametric statistics to estimate stability were used in this study. One of them (Huehn, 1979; Nassar and Huehn, 1987) consisted of four nonparametric stability statistics (Si(1), Si(2), Si(3) and Si(6)) combining mean yield and stability (see also Becker and Leon, 1988; Akcura and Kaya, 2008). We also used the methodology described by Fox et al., (1990), who proposed a nonparametric superiority measure for general adaptability using stratified ranking of cultivars. A genotype that occurred mostly in the top third (high TOP- value) was considered a widely adapted cultivar. Kang's (1988) rank-sum (RS) is another non-parametric stability procedure where both yield and Shukla's (1972) stability variance were used as selection criteria. This index assigns a weight of one to both yield and stability statistics to identify high-yielding and stable genotypes.

The genotype with the highest yield is given a rank of one and a genotype with the lowest stability variance is assigned a rank of one. All genotypes are ranked in this manner, and the ranks of yield and stability variance are summed for each genotype. Genotypes with the lowest rank-sum are the most desirable (Mohammadi and Amri, 2008).

Rank measures and adjusted means of grain yields were used to depict plot by SAS PLOT procedure (Lu, 1995? Akcura et al., 2008). Besides, the stability parameters were compared using Spearman's rank correlation (Steel et al., (1997). All analysis was performed using SAS Software (SAS Institute, 1999).

RESULTS AND DISCUSSION

Combined analysis of variance for grain yield showed that genotypes' and environments' main effects and GEI were significant at P less than 0.01 (Table 3). In the case GEI effect suggest that there are significant differences in responses of genotypes to environments, and hence sensitivity and instability (Akcura et al., 2009). Genotypic rank differences over environments showed the existence of crossover GEIs (Crossa, 1990). This was fitted by the significant effect of GEI in the joint analysis of variance (Table 3) and showed the necessity to assess the response of the genotypes to environmental variation.

Table 3. Analysis of variance for stability parameter for 25 bread wheat genotypes.

Source of variation###DF###Sum of###Mean

###Squares###Squares

Genotype (G)###24###99.916124###4.163172

Environments E + G E###225###3309.206###14.7

E (linear)###1###3112.73069

G x E (Linear)###24###159.879###6.661

Pooled Deviations###200###36.596###0.18298

Pooled Error###720###312.744940###0.434368

CV(%) : 11.23

P less than 0.01

Parametric measures: Considering environment, grain yield of environments over genotypes ranged from 3.16 t ha-1 for E10 to 8.15 t ha-1 for E6. Grain yield of genotypes over environments ranged from 5.29 t ha-1 to 6.31 t ha-1 (Table 4). Taking the grain yield over environments as the first parameter genotypes, G4, G5, G6, G7, G11, G13, G16, G17, G18, G19, G20, G22, G24 and G25 had higher grain yield than grand mean grain yields (5.87 t ha-1), while genotypes G1, G2, G3, G8, G9, G10, G12, G14, G15, G21 and G23 had lower values than mean grain yield.

Table 4. Mean yield of 25 bread wheat genotypes tested across 10 environments (t ha-1).

Code###E1###E2###E3###E4###E5###E6###E7###E8###E9###E10###Mean

G1###5.15###6.49###5.93###2.70###6.88###7.35###4.36###5.09###5.55###3.41###5.29 m

G2###5.49###8.05###5.36###3.08###7.08###7.50###5.43###6.08###7.20###1.80###5.71 g-k

G3###4.73###6.25###5.51###2.88###6.85###7.02###4.96###5.38###6.39###2.98###5.29 m

G4###5.46###7.27###5.41###3.20###8.26###8.12###5.27###6.23###7.19###3.36###5.98 b-g

G5###5.93###8.81###6.22###3.69###7.73###8.62###5.42###6.27###7.09###3.36###6.31 a

G6###5.72###7.30###6.11###2.34###7.19###7.58###4.52###5.93###6.25###3.04###5.60 i-l

G7###5.93###7.63###6.66###2.56###8.40###8.33###5.43###5.23###6.48###3.38###6.00 b-f

G8###3.90###7.10###5.67###3.29###7.18###7.90###4.55###5.30###6.68###2.09###5.37 lm

G9###6.11###7.52###6.20###3.26###7.64###8.04###3.92###6.01###6.67###3.14###5.85 e-i

G10###6.21###7.04###6.42###2.47###7.05###7.53###4.90###6.01###6.57###3.28###5.75 f-j

G11###6.13###7.84###5.84###2.96###7.29###8.44###5.01###5.70###6.64###3.28###5.91 c-h

G12###5.81###7.79###5.66###2.86###7.21###8.31###4.03###6.48###7.47###2.78###5.84 e-i

G13###5.80###7.04###6.20###2.82###7.60###7.99###4.74###6.59###7.30###2.84###5.89 d-h

G14###5.57###6.61###5.56###2.17###6.60###7.97###5.28###5.33###6.00###3.24###5.43 klm

G15###5.71###7.76###6.37###2.08###6.60###7.82###4.49###6.01###6.78###2.94###5.66 h-l

G16###6.30###8.34###5.72###3.36###7.67###8.61###5.05###6.52###7.51###2.90###6.20 abc

G17###6.26###8.13###6.03###2.53###8.07###8.62###4.80###7.08###7.94###3.65###6.31 a

G18###6.51###8.54###5.96###2.63###8.95###9.05###4.88###6.24###7.46###2.91###6.31 a

G19###6.01###8.43###5.78###2.15###8.64###8.66###4.56###6.69###7.74###3.69###6.23 ab

G20###6.19###7.76###5.97###2.53###7.30###8.20###4.98###6.22###6.49###3.42###5.91 d-h

G21###5.27###7.18###4.34###2.88###7.16###7.75###4.79###6.03###6.80###3.23###5.54 j-m

G22###5.54###8.33###5.96###2.29###8.42###8.61###5.13###6.72###7.94###3.42###6.24 ab

G23###5.78###7.58###5.51###1.84###8.06###9.01###4.45###5.76###6.93###3.12###5.80 e-j

G24###5.80###7.47###5.57###2.68###8.70###8.30###3.77###7.13###7.45###4.59###6.15 a-d

G25###6.24###8.88###5.39###2.78###8.52###8.32###4.61###6.11###6.80###3.11###6.08 a-e

Mean###5.74###7.65###5.81###2.72###7.64###8.15###4.77###6.09###6.93###3.16###5.87

LSD (0.05)###0.29

: Means in a column with the same letter are not significantly different at the 5% level.

The results of eleven parametric stability parameters and mean grain yield are given in Table 5.

The measures of adaptability and stability are necessary for its suggestion to target environments for preferring genotypes. According to Eberhart and Russell (1966), regression coefficients (b i ) approximating 1.0 coupled with deviation from regression (S 2 di ) of zero indicate average stability. Genotypes have general adaptability when associated with high mean yield while genotypes are poorly adapted to environments when associated with low mean yield. Regression coefficient (b i ) values above 1.0 define genotypes with higher sensitivity to environmental alteration. Regression coefficients decreasing below 1.0 ensure a measure of greater resistance to environmental variation, and hence, increasing specificity of adaptability to low yielding environments. Genotypes G17, G18, G19, G22 and G25 had higher grain yields and a coefficient values greater than 1.0 (Table 5).

These genotypes are sensitive to environmental variations and would be suggested for cultivation under favourable conditions, whereas G1, G3, G14 and G21 with b i less than 1 and lowest average yields were poorly adapted across environments and might have specific adaptation to harsh conditions. On the contrary, G5, G7, G16 and G24 had higher grain yields and a coefficient values near 1.0. These genotypes showed average stability. Among these cultivars G16 was the most appropriate one, because it had higher yield value than the mean, b i values near 1.0 and low S 2 di (i.e. regression coefficient not significantly different from 1.0 with grain yields above grand mean). Pinthus's (1973) stability parameter or coefficient of determination (R 2 i ) values which are the predictability of response estimates (R 2 i =1), ranged from 0.87 to 0.99, in which a variation of mean grain yield was explained by genotype response across environments.

None of values of coefficient of determinations was significantly different from 1.0. In terms of this parameter, all of genotype could be considered stable for grain yield (Table 5). Wricke (1962) suggested using ecovalance (Wi 2) as a stability parameter. According to this stability parameter, genotypes with the smallest ecovalance (Wi 2) values are considered stable. The W 2 i was lowest for genotypes G6, G13, G16, G20 and highest for G1, G2, G3, G7, G8 and G24 (Table 5). The good correlation between Wi 2 and Si2 (r = 0.85 ) showed that these two measures led to similar results. According to the environmental variance (S2 i), G3 followed by G1, G21, G19 and G10 had the lowest variation across environments and G18 followed by G19, G23, G25, G22 and G17 showed the largest variation (Tables 5).

According to Francis and Kannenberg's (1978) coefficient of variation stability parameter (CV i), genotypes G3, G1, G5, G10 and G4 were considered to be stable although they had low performance except G5, and the genotypes G23, G18, G19, G25, and G22 with the highest yield performance were considered unstable. An unbiased estimate using stability variance (s2 i) of genotypes was determined according to Shukla (1972). The stability variance (s2i) revealed that the genotypes G6, G11 G20, G16, G13, G4 and G5 had the smallest variance across the environments, while the genotypes G24, G8 and G2 had the largest s2i. Hence, the genotypes G6, G11 G20, G16, G13, G4 and G5 were stable while the genotypes G24, G8 and G2 were unstable.

When the stability parameter of Plaisted and Peterson (1959) was used, the genotypes G6, G11, G20, G13, G16, G4 and G5 had lower P59 and stable genotypes Tai's model (1971) is based upon the principle of structural relationship analyses, which the GEI effect of genotype is portioned into two components. They are the linear response to environmental effects, which is measured by statistic (a i ) and deviation from the linear response, which are measured by (l i ) statistic. A perfectly stable genotype is that in which (a i , l i )= (-1, 1). According to these stability statistics, durum wheat genotypes G2, G1 and G10 could be considered as stable (Table 5) because these genotypes were near to (a i , l i )=(- 1, 1) and G12 had average stability with (a i , l i ) = (0, 1). The superior genotype should be the one with the lowest superiority index (P i ) value, thus the negative (r = -0.98 ) correlation between yield and P i -value (Table 6).

In such cases, the genotypes of greatest interest would be those with the lowest P i values, most of which would be attributed to genetic deviation (Lin and Binns, 1988). Accordingly, genotypes G17, G5, G19, G22, G18 and G16 have the greatest mean yield and the lowest P i values (Table 5). Non - parametric measures: The result of 6 different non-parametric stability statistics and genotype mean grain yields are presented in Table 5. Two rank stability measures (S i (1) and S i (2) ) from Nassar and Huehn (1987) are based on ranks of genotypes across environments and they give equal weight to each environment. Genotypes with fewer changes in ranking are considered to be more stable (Becker and Leon, 1988). Zero variance is an indication of maximum stability. Accordingly, S i (1) and S i (2) of the tested genotypes showed that genotypes G6, G20 and G16 had the lowest values; therefore, these genotypes were regarded as the most stable genotypes according to S i (1) and S i (2).

On the other hand, G24, G1, G3, G2, G18 and G10 had the highest S i (1) and S i (2) values; therefore, they were determined to be unstable Two other nonparametric statistics (S i (3) and S i (6)) combine yield and stability based on yield ranks of genotypes in each environment (Nassar and Huehn, 1987). Si (3) and S i (6) ranged from 9.4 to 59.22 and 1.54 to 7.37, respectively. Genotypes G6, G21, and G3 had the lowest S i (3) and S i (6) values hence, these genotypes were characterized as the most stable genotypes, as well as with regard to Si (3) and Si (6) statistics (Table 5). However, mean yields of G17, G1 and G23 were lower than total mean. While genotypes G13, G19 and G24 were the 3 highest mean yielding genotypes, they were characterized as unstable genotypes according to S i (3) and S i (6) parameters (Tables 5).

According to the non-parametric (TOP) superiority index (Fox et al., 1990), G17 was an adapted genotype, because it ranked in the Top third of genotypes in a high percentage of environments (high TOP value, 80%), and was followed by G16 (70%), G22 (70%), G5 (60%) and G19 (60%) (Table 5). The superiority parameter of Fox et al. (1990) consists of scoring the percentage of environments in which each genotype ranked in the top, middle, and bottom third of trial entries. The undesirable genotypes identified by this method were G3 and G21.

According to rank-sum (RS) statistics (Kang, 1988), genotypes with a low rank-sum are regarded as the most desirable. This parameter revealed that genotypes G5, G16 and G17 had the lowest values, and were stable genotypes, whereas genotypes G1, G3 and G8, which had the highest values, were undesirable (Table 5).

Interrelationship among parametric and non-parametric measures: The rank correlations between grain yield and stability measures are given in Table 6. Grain yield is significantly correlated with bi, Si (2), ai, Si (6) , TOP (P less than 0.01) and with the measures of Si (3) and R2 (P less than 0.05) and showed a negative and significant correlation with Pi and RS (P less than 0.01). The coefficient of regression (b i ) is significantly correlated with S i (2) , CVi, Di 2, ai, Si (3), S i (6) , TOP (P less than 0.01) and R 2 (P less than 0.05), and showed a negative and significant correlation with P i (P less than 0.01). Coefficient of determination (R2 i) had negative and significant correlations with S di 2, Wi 2, si 2 , P59, li, Si (2), RS (P less than 0.01) and with the measures of ai, Pi and Si (1) (P less than 0.01). Deviation from Regression (S di 2) is significantly correlated with W i 2, si 2, P59, li, (P less than 0.01) and with the measures of S i (1) S i (2) and RS (P less than 0.05).

Ecovalance (Wi 2) was positively associated with si 2, P59, l i, S i (1), S i (2), S i (3) and RS (P less than 0.01).

Table 5. Mean grain yield values (t ha-1) and stability parameters of 25 bread wheat genotypes across 10 environments.

###Parametric measures###Non-parametric measures

Code###X a###bi b###R2i c###S di2###Wi2###Si 2###CVi###si2###P59###si###li###Pi###Si(1)###Si(2)###Si(3)###Si(6)###TOP###RS###F

G1###5.29###0.77###0.94###0.15###2.74###2.20###28.05###0.32###0.27###-0.23###1.22###1.43###10.24###74.68###20.74###2.84###10.00###47.00###5

G2###5.71###1.01###0.91###0.42###3.36###3.92###34.67###0.40###0.31###0.01###3.58###0.95###10.60###80.54###43.29###4.98###30.00###41.00###2

G3###5.29###0.77###0.95###0.11###2.57###2.13###27.54###0.30###0.26###-0.24###0.92###1.44###10.51###78.10###16.03###2.30###0.00###44.00###8

G4###5.98###0.94###0.96###0.14###1.25###3.19###29.85###0.14###0.19###-0.06###1.21###0.49###7.89###43.79###30.54###4.50###30.00###18.00###4

G5###6.31###0.97###0.96###0.15###1.24###3.41###29.23###0.14###0.19###-0.03###1.30###0.24###7.33###40.49###20.34###5.36###60.00###7.00###8

G6###5.60###0.95###0.98###0.08###0.72###3.17###31.80###0.08###0.16###-0.05###0.67###0.90###6.36###29.33###9.40###1.54###10.00###22.00###10

G7###6.00###1.00###0.92###0.33###2.62###3.80###32.44###0.31###0.27###0.01###2.79###0.56###9.56###63.73###48.94###5.58###40.00###30.00###2

G8###5.37###0.96###0.89###0.45###3.68###3.59###35.###0.43###0.33###-0.04###3.86###1.38###9.31###63.66###19.17###2.17###10.00###47.00###2

G9###5.85###0.95###0.95###0.17###1.43###3.27###30.89###0.16###0.20###-0.05###1.44###0.62###7.89###46.50###24.27###3.46###30.00###24.00###2

G10###5.75###0.88###0.95###0.17###1.78###2.83###29.24###0.21###0.22###-0.12###1.42###0.80###10.11###70.77###32.05###4.16###20.00###30.00###5

G11###5.91###0.96###0.98###0.08###0.74###3.24###30.43###0.08###0.16###-0.04###0.72###0.56###7.82###45.60###17.28###3.38###40.00###14.00###6

G12###5.84###1.06###0.97###0.16###1.41###4.05###34.44###0.16###0.20###0.06###1.37###0.64###8.53###51.29###25.59###3.75###20.00###24.00###1

G13###5.89###0.99###0.97###1.29###1.04###3.48###31.66###0.12###0.17###-0.01###1.11###0.60###8.16###46.28###24.19###3.49###30.00###18.00###5

G14###5.43###0.86###0.93###0.23###2.42###2.78###30.66###0.28###0.25###-0.14###1.95###1.26###11.02###88.00###19.74###2.50###10.00###41.00###2

G15###5.66###1.01###0.95###0.20###1.64###3.71###34.07###0.19###0.21###0.01###1.74###0.89###8.22###48.49###20.37###2.49###10.00###30.00###2

G16###6.20###1.04###0.97###0.12###0.99###3.87###31.70###0.11###0.17###0.04###0.10###0.33###6.76###34.40###40.30###5.77###70.00###10.00###11

G17###6.31###1.10###0.98###0.12###1.25###4.29###32.83###0.14###0.19###0.10###0.99###0.23###7.67###44.46###35.13###5.85###80.00###10.00###7

G18###6.31###1.23###0.99###0.08###2.33###5.34###36.60###0.27###0.25###0.24###0.64###0.30###10.53###81.29###50.53###6.86###50.00###19.00###3

G19###6.23###1.19###0.98###0.14###2.24###5.01###35.91###0.26###0.24###0.19###1.20###0.29###9.42###68.00###59.22###7.37###60.00###21.00###3

G20###5.91###0.96###0.99###0.08###0.71###3.30###30.74###0.08###0.16###-0.03###0.71###0.55###6.67###33.16###18.63###3.26###20.00###13.00###8

G21###5.54###0.89###0.94###0.21###2.02###2.94###30.92###0.23###0.23###-0.11###1.77###1.01###8.04###46.40###13.34###2.50###0.00###36.00###4

G22###6.24###1.17###0.98###0.12###1.77###4.79###35.10###0.20###0.22###0.17###0.99###0.29###8.98###56.93###47.49###6.22###70.00###16.00###5

G23###5.80###1.18###0.98###0.11###1.97###4.96###38.36###0.23###0.23###0.19###0.95###0.69###7.42###41.51###25.68###3.18###10.00###30.00###2

G24###6.15###1.00###0.87###0.61###4.87###4.02###32.61###0.58###0.40###0.00###5.19###0.42###11.04###92.50###54.06###6.07###40.00###32.00###2

G25###6.08###1.13###0.96###0.23###2.36###4.63###35.40###0.28###0.25###0.13###1.94###0.42###9.16###62.67###35.68###4.56###30.00###26.00###1

Mean###5.87

KEY: X=Grain yield (t ha -1 ); bi = Regression coefficient, R 2 = Pinthus's coefficient of determination; S di 2 ; W i 2 = Wrick's ecovalance; Si 2 = Environmental variance; CV i =Coefficient of variation; s i 2 =Shukla's stability; P59= Plaisted and Peterson's stability parameter; a and l= Tai's stability parameters; PI=Lin and Binns superiority measure; Genotype absolute rank difference mean over n environments (Si 1); between ranks variance over n environments (S i 2 ); sum of the absolute deviations of the squares of ranks for each genotype (S i 3 ); the sum of squares of ranks for each genotype relative to the mean of ranks(S i 6 ); Kang's rank-sum(RS); number of sites at which the genotype occurred in the top third of the ranks (TOP). F= Frequency of number of stability parameters over all of stability parameters for each genotype, if a genotype had nine values of F, it could be as stable. a Values in italics of are significantly higher than the mean at P less than 0.05. b Values in italics of are non significantly different from the unity at P less than 0.05. None of values of coefficient of determinations was significantly different from 1.0

Table 6. Spearman's rank correlation coefficients between yields and stability parametric and non-parametric measures for 25 bread wheat genotypes across 10 environments

Measures###YLD###Parametric measures###Non-parametric measures

###bi###R2###Sdi2###Wi2###Si2###CVi###si2###P59###ai###li###Pi###Si(1)###Si(2)###Si(3)###Si(6)###TOP

###bi###0.67

###R2###0.46###0.46

###Sdi2###-0.22###-0.11###-0.81

###Wi2###-0.30###-0.02###-0.67###0.51

###Si2###0.66###0.98###0.38###-0.04###0.09

Parametric measures###CVi###0.30###0.86###0.24###0.06###0.24###0.89

###si2###-0.30###-0.02###-0.67###0.51###1.00###0.09###0.24

###P59###-0.30###-0.02###-0.67###0.51###1.00###0.09###0.24###1.00

###ai###0.67###1.00###0.46###-0.11###-0.02###0.98###0.86###-0.02###-0.02

###li###-0.23###-0.14###-0.90###0.91###0.62###-0.05###0.07###0.62###0.62###-0.14

Non-parametric###Pi###-0.98###-0.65###-0.50###0.25###0.37###-0.63###-0.28###0.37###0.37###-0.65###0.27

###Si(1) -0.21###-0.06###-0.58###0.47###0.86###0.02###0.10###0.86###0.86###-0.06###0.50###0.29

###Si(2) -0.19###-0.05###-0.55###0.42###0.86###0.04###0.12###0.86###0.86###-0.05###0.47###0.27###0.99

###Si(3)###0.63###0.69###0.02###0.17###0.33###0.73###0.53###0.33###0.33###0.69###0.19###-0.59###0.39###0.39

###Si(6)###0.87###0.66###0.21###-0.01###0.05###0.68###0.34###0.05###0.05###0.66###-0.01###-0.84###0.17###0.18###0.88

###TOP###0.95###0.66###0.37###-0.11###-0.24###0.65###0.30###-0.24###-0.24###0.66###-0.15###-0.94###-0.14###-0.12###0.69###0.89

###RS###-0.79###-0.37###-0.69###0.47###0.81###-0.29###0.03###0.81###0.81###-0.37###0.55###0.82###0.66###0.65###-0.17###-0.51###-0.74

###-0,7854###-0,3738###-0,6869###0,4692###0,8054###-0,2946###0,0315###0,8054###0,8054###-0,3738###0,5508###0,82###0,6592###0,6477###-0,1662

P less than 0.05,###P less than 0.01

Environmental variance (S2i), is significantly correlated with CVi, ai, Si(3), Si(6) TOP (P less than 0.01) and showed a negative and significant correlation with Pi (P less than 0.01). The coefficient of variation (CVi) was positively and significantly associated with ai and Si(3) (P less than 0.01). Stability variance (si2) had positive and significant correlations with li, Si(1), Si(2) and RS (P less than 0.01). P59 is significantly correlated with li Si(1), Si(2) and RS (P less than 0.01). Alpha (ai) is significantly correlated with Si(3), Si(6) and TOP (P less than 0.01), and showed a negative and significant correlation with Pi (P less than 0.01). Lamda (li) had positive and significant correlations with Si(1), Si(2) and RS (P less than 0.05). The superiority index (Pi) was negatively associated with Si(3), Si(6) and TOP (P less than 0.01) and positively correlated with RS (P less than 0.01).

Non-parametric measure of Si(1), is significantly correlated with Si(2) and RS (P less than 0.01). Si(2) had positive and significant correlations with RS (P less than 0.01). Si(3), as well as Si(6) parameters were positively correlated with TOP (P less than 0.01). The TOP was negatively and significantly associated with RS (P less than 0.01).

Environmental variations are important in determining performance, and hence, evaluations based on several years and locations are a good strategy to pursue in breeding for varying environments. Agricultural producers in developing countries which use restricted inputs and grow cereals under harsh and unstable environments require stable varieties. In these cases, genotypes with good performance and stability should be recommended.

The most severe limitation of the regression approach is the poor repeatability of both bi and S2di (Jalaluddin and Harrison, 1993); its usefulness in measuring genotype adaptability depends largely on the assumption that genotypes respond linearly to the environments. In such cases, the results of nonparametric estimation and testing procedures, which are based on ranks, can be more reliable (Mut et al., 2009). Several multivariate methods which have been proposed allow a more detailed analysis of GEI, but the complexity of these methods, sometimes regarded as their main advantage, paradoxically is the main obstacle to their widely usage in plant breeding (Flores et al., 1998). Though, different stability methods are indicative of high, intermediate or low stability performance, the stability values do not provide information for reaching definitive conclusions (Mohammadi and Amri, 2008).

Many stability measures that have been used in this study considered that stability of genotypes related to yield and stability. Hence, it is essential that both yield and stability should be considered simultaneously to reveal the beneficial effects of GEI and to select the genotypes.

According to the matrix correlation analysis, parametric and non-parametric measures used in this study revealed that these parameters can be used for evaluating the responses of bread wheat genotypes to changing environments. Also, mean grain yield ( C ) and coefficient of regression (bi) of genotypes were significantly positively correlated to the non-parameters of Si(3), Si(6) and TOP used. Similarly, Akcura and Kaya (2008); Kilic et al. (2010) reported high rank correlations between grain yield and these measures of stability. Although they do not supply information about genotype adaptability, nonparametric stability measurements seem to be useful alternatives to parametric measurements (Yue et al., 1997). The repeatability, reliability and suitability of parametric and non-parametric methods to select the best genotypes in different crops need to be further investigated.

Conclusion: The study of genotypic stability identified new suitable advanced lines and revealed why some varieties are grown in the South Eastern Anatolia region. According to both parametric and non-parametric stability parameters, bread wheat genotypes G16, G6, G3, G5 (Nurkent) and G20 (Adana-99) were more stable varieties, which had 11, 10, 8, 8 and 8 stability parameters out of 17 stability statistics used, respectively. In fact, Adana-99, Turkish commercial variety, has high grain quality. This variety was released by the Cukurova Agricultural Research Institute (one of the costal regions). It not only appears to have a specific adaptation to this region but can also be grown successfully in other spring zones of Turkey.

Nurkent, another Turkish commercial variety, also showed high stability and may still be of interest for growers in South Eastern Anatolia. From the advanced lines, genotypes G16, G6 and G3 could be used success fully as progenitors in breeding programmes for the production of high grain yield and stability bread wheat in the costal zones of Turkey

REFERENCES

Akcura, M. and Y. Kaya (2008). Nonparametric stability methods for interpreting G x E interaction of bread wheat genotypes (Triticum aestivum L.). Genetics and Molecular Biology, 31(4):906-913.

Akcura, M., Y. Kaya and S. Taner (2009). Evaluation of durum wheat genotypes using parametric and non parametric stability statistics. Turkish J. Field Crops, 14(2): 111-122.

Annicchiarico, P (2002). G x E interactions. Challenges and opportunities for plant breeding and cultivar recommendations. FAO Plant Production and Protection, 174 - 115 p. Rome.

Anonymous (2008). Statistics of agricultural production. less than http://www.tarimsal.com/tarim_istatistikleri.htm greater than [accessed 15 01 2010]

Becker, H. C. and J. Leon (1988). Stability analysis in plant breeding, Plant Breed 101: 1-23.

Crossa, J (1990). Statistical analyses of multi-location trials. Adv. Agron, 44 : 55-85.

Duarte, J. B. and M. J. Zimmermann (1995). Correlation among yield stability parameters in common bean. Crop Sci., 35: 905-912.

Eberhart, S. A. and W. A. Russell (1966). Stability parameters for comparing varieties. Crop Sci., 6: 36 - 40.

Finlay, K. W., Wilkinson, G.N. 1963. The analysis of adaptation in a plant breeding programme. Aust. J. Agric. Res., vol. 14, p. 742-754.

Flores, F (1993). Genotype environment interaction of Vicia faba L. Ph.D. Thesis, Univ. of Cordoba Spain.

Flores, F., M. T. Moreno, J. I. Cubero (1998). A comparison of univariate and multivariate methods to analyze G x E interaction. Field Crops Research, 56: 271-286.

Fox, P. N., B. Skovmand, B. K. Thompson, H. J. Braun, R. Cormier (1990). Yield and adaptation of hexaploid spring triticale. Euphytica, 47: 57- 64.

Francis, T. R., Kannenberg, L. W. (1978). Yield stability studies in short-season maize. I. A descriptive method for grouping genotypes, Can J. Plant Sci., v. 58, p.1029-1034.

Freeman, G. H., Perkins, J. M. (1971). Environmental and genotype-environmental components of variability VIII. Relations between genotypes grown in different environments and measures of these environments. Heredity, 27: 15-23.

Hanson, W. D (1970). Genotypic stability. Theor. Appl. Genet., 40 : 226-231.

Huehn, M (1979). Contributions to the acquisition of phenotypic stability. I. Proposal based on some rank-sum stability parameters. EDV Medizin und Biologie, 10 : 112-117.

Jalaluddin, M. and S. A. Harrison (1993). Repeatability of stability estimators for grain yield in wheat. Crop Sci., 33: 720-725.

Kang, M. S. (1988) A rank-sum method for selecting high- yielding, stable corn genotypes. Cereal Res Comm, 16 : 113-115.

Ketata, H. Y., S. K. Yau. and M. Nachit (1989). Relative consistency performance across environments. International Symposium on Physiology and Breeding of Winter Cereals for stressed Mediterranean Environments, Montpellier. pp 391-400.

Kilic, H., M. Akcura and H. Aktas (2010). Assessment of parametric and non-parametric methods for selecting stable and adapted durum wheat genotypes in multi-environments. Not. Bot. Hort. Agrobot. Cluj, 38 (3): 271-279.

Lin, C. S., M. R. Binns and L. P Lefkovitch (1986). Stability analysis: where do we stand? Crop Sci., 26 : 894-900.

Lin, C. S. and M. R. Binns (1988). A method of analyzing cultivar x location x year experiments: A new stability parameter. Theor. Appl. Genet., 76: 425-430.

Lu, H. S (1995). PCSAS Program for estimating Huehn's nonparametric stability statistics. Agro. J., 87: 888891.

Mohammadi, R. and A. Amri (2008). A. Comparison of parametric and non-parametric methods for selecting stable and adapted durum wheat genotypes in variable environments. Euphytica, 159: 419-432.

Mohebodini, M., H. Dehghani and S. H. Sabaghpour (2006). Stability of performance in lentil (Lens culinaris Medik) genotypes in Iran. Euphytica, 149: 343-352.

Mut, Z., N. Aydin, H. O. Bayramoglu and H. Ozcan (2009). Interpreting genotype x environment interaction in bread wheat (Triticum aestivum L.) genotypes using non-parametric measures. Turk J Agric., 33: 127-137.

Nassar, R. and M. Huehn (1987). Studies on estimation of phenotypic stability: tests of significance for non- parametric measures of phenotypic stability. Biometrics, 43: 45-53.

Plaisted, R. L. and L. C. A. Peterson (1959). Technique for evaluating the ability of selections and yield consistency in different locations or seasons. Am Potato J., 36 : 381-385.

Plaisted, R. L. (1960). A shorter method of evaluating the ability of selection to yield consistently over seasons. Am Potato J., 37: 166-172.

Pinthus, M. J. (1973). Estimates of genotypic value: a proposed method. Euphytica 22: 345351.

SAS Institute (1999). SAS/STAT user's guide. 8. Version. SAS Institute Inc. Cary. NC.

Saini, D. P. and P. L. Gautam (1990). Note on G x E analysis in segregating population of durum wheat. Ind. J. Genetic Plant Breed, 50: 199-201.

Shukla, G. K. (1972). Some statistical aspects of partitioning G x E components of variability. Heredity, 29:237-245.

Steel, R. G. D., J. H. Torrie and D. A. Dickey (1980). Principles and procedures of statistics. Mc Graw Hill Book Company, New York, 612 p. (in English).

Tai, G. C. C. (1971): Genotypic stability analysis and its application to potato regional trials. Crop Sci., 11: 184-190.

Westcott, B. A. (1987). Method of assessing the yield stability of crop genotypes. J. Agric. Sci., 108: 267-274.

Wricke, G. (1962). On a method of understanding the biological diversity in field research. Z. Pflanzenzucht, 47: 92-96.

Yue, G. L., K. L. J. Roozeboom, W. T. Schapaugh and G. H Liang (1997). Evaluation of soybean genotypes using parametric and nonparametric stability estimates, Plant Breed., 116: 271275.

Department of Field Crops, Faculty of Agriculture, Bingol University, 12000, Bingol, Turkey, Corresponding author e-mail: hkilic@bingol.edu.tr
COPYRIGHT 2012 Asianet-Pakistan
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Kilic, Hasan
Publication:Journal of Animal and Plant Sciences
Article Type:Report
Geographic Code:9PAKI
Date:Jun 30, 2012
Words:7305
Previous Article:BIOLOGY AND MANAGEMENT OF PURPLE NUTSEDGE (CYPERUS ROTUNDUS L.).
Next Article:COMPARATIVE STUDY OF DEVELOPMENTAL AND REPRODUCTIVE CHARACTERISTICS OF CHRYSOPERLA CARNEA (STEPHENS) (NEUROPTERA: CHRYSOPIDAE) AT DIFFERENT REARING...
Topics:

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters