Pattern analysis of international sorghum multi-environment trials for grain-yield adaptation

 Pattern analysis, which consists of joint and complementary use of classification and ordination techniques, was applied to grain-yield data of 12 sorghum genotypes in 25 environments to identify the grouping of genotypes and environments. The 12 genotypes represented a wide geographical origin, different genetic diversity, and three photoperiod-sensitive classes. The 25 environments represented a super population of widely different environments covering latitudes from 20°S to 45°N. The knowledge of environmental and genotype grouping helped reveal several patterns of genotype×environment (GE) interaction. The existence of two mega-environments – African and Asian – was indicated. Within these mega-environments, several subgroups were further discernible. The Asian-type subgroups of environments tended to be closer to one another, suggesting that they discriminated genotypes similarly. By contrast, the African-type sub-groups of environments were more divergent. Differential genotype adaptation patterns existed in the two mega-environments. The repeatability of the GE patterns seen in this multi-environmental trial, however, needs to be established over time. Received: 4 April 1997 / Accepted: 6 October 1997     

Crop diagnosis and probe genotypes for interpreting genotype environment interaction in winter wheat trials

Genotype*environment interaction has been analyzed with 12 genotypes and four probe genotypes in French wheat trials. An integrated approach was developed which combined crop diagnosis with the analysis of interaction by factorial regression. Crop diagnosis was helpful to characterize the environments and to select environmental variables. Such an approach succeeded in providing an agronomic explanation of genotype*environment interaction and in defining the responses or parameters for each genotype and each environment. Earliness at heading, susceptibility to powdery mildew and susceptibility to lodging were the three major genotypic covariates. Interaction could also be related to environment features, measured indirectly by the behavior of the four probe genotypes during the formation of yield, what we called the outputs of a simplified crop diagnosis, or described directly by indicators of yield-limiting factors. Two important crop diagnosis covariates were analyzed in order to characterize interaction during the formation of yield: the reduction in kernel number, which described the time-period until flowering, and the reduction in thousand kernel weight, which corresponded to the period after flowering. These variates were estimated for each probe genotype and allowed us to compare the behavior of the 12 genotypes to that of the probe genotypes. Both periods of the formation of yield contributed to the interaction, and ’Camp-Rémy’ was the probe of particular interest for the comparisons. When true environmental variates were used, factorial regression revealed that water deficits during the formation of grain number and level of nitrogen were predominant. Such an integrated approach could be exploited when varieties are tested in a network where numerous and diverse yield-limiting factors may occur. Received: 3 August 1998 / Accepted: 16 March 1999     

Empirical best linear unbiased prediction in cultivar trials using factor-analytic variance-covariance structures

 Results of multi-environment trials to evaluate new plant cultivars may be displayed in a two-way table of genotypes by environments. Different estimators are available to fill the cells of such tables. It has been shown previously that the predictive accuracy of the simple genotype by environment mean is often lower than that of other estimators, e.g. least-squares estimators based on multiplicative models, such as the additive main effects multiplicative interaction (AMMI) model, or empirical best-linear unbiased predictors (BLUPs) based on a two-way analysis-of-variance (ANOVA) model. This paper proposes a method to obtain BLUPs based on models with multiplicative terms. It is shown by cross-validation using five real data sets (oilseed rape, Brassica napus L.) that the predictive accuracy of BLUPs based on models with multiplicative terms may be better than that of least-squares estimators based on the same models and also better than BLUPs based on ANOVA models. Received: 18 October 1997 / Accepted: 31 March 1998     

Robustness of statistical tests for multiplicative terms in the additive main effects and multiplicative interaction model for cultivar trials

The additive main effects multiplicative interaction model is frequently used in the analysis of multilocation trials. In the analysis of such data it is of interest to decide how many of the multiplicative interaction terms are significant. Several tests for this task are available, all of which assume that errors are normally distributed with a common variance. This paper investigates the robustness of several tests (Gollob, F _GH1, F_GH2, F_R)to departures from these assumptions. It is concluded that, because of its better robustness, the F _Rtest is preferable. If the other tests are to be used, preliminary tests for the validity of assumptions should be performed.     

Interpreting genotype × environment interaction in tropical maize using linked molecular markers and environmental covariables

An understanding of the genetic and environmental basis of genotype×environment interaction (GEI) is of fundamental importance in plant breeding. In mapping quantitative trait loci (QTLs), suitable genetic populations are grown in different environments causing QTLs×environment interaction (QEI). The main objective of the present study is to show how Partial Least Squares (PLS) regression and Factorial Regression (FR) models using genetic markers and environmental covariables can be used for studying QEI related to GEI. Biomass data were analyzed from a multi-environment trial consisting of 161 lines from a F_3:4 maize segregating population originally created with the purpose of mapping QTLs loci and investigating adaptation differences between highland and lowland tropical maize. PLS and FR methods detected 30 genetic markers (out of 86) that explained a sizeable proportion of the interaction of maize lines over four contrasting environments involving two low-altitude sites, one intermediate-altitude site, and one high-altitude site for biomass production. Based on a previous study, most of the 30 markers were associated with QTLs for biomass and exhibited significant QEI. It was found that marker loci in lines with positive GEI for the highland environments contained more highland alleles, whereas marker loci in lines with positive GEI for intermediate and lowland environments contained more lowland alleles. In addition, PLS and FR models identified maximum temperature as the most-important environmental covariable for GEI. Using a stepwise variable selection procedure, a FR model was constructed for GEI and QEI that exclusively included cross products between genetic markers and environmental covariables. Higher maximum temperature in low- and intermediate-altitude sites affected the expression of some QTLs, while minimum temperature affected the expression of other QTLs. Received: 10 January 1999 / Accepted: 12 March 1999     

Assessing the impact of natural service bulls and genotype by environment interactions on genetic gain and inbreeding in organic dairy cattle genomic breeding programs

The objective of the present study was to compare genetic gain and inbreeding coefficients of dairy cattle in organic breeding program designs by applying stochastic simulations. Evaluated breeding strategies were: (i) selecting bulls from conventional breeding programs, and taking into account genotype by environment (G×E) interactions, (ii) selecting genotyped bulls within the organic environment for artificial insemination (AI) programs and (iii) selecting genotyped natural service bulls within organic herds. The simulated conventional population comprised 148 800 cows from 2976 herds with an average herd size of 50 cows per herd, and 1200 cows were assigned to 60 organic herds. In a young bull program, selection criteria of young bulls in both production systems (conventional and organic) were either ‘conventional’ estimated breeding values (EBV) or genomic estimated breeding values (GEBV) for two traits with low (h²=0.05) and moderate heritability (h²=0.30). GEBV were calculated for different accuracies (r_mg), and G×E interactions were considered by modifying originally simulated true breeding values in the range from r_g=0.5 to 1.0. For both traits (h²=0.05 and 0.30) and r_mg⩾0.8, genomic selection of bulls directly in the organic population and using selected bulls via AI revealed higher genetic gain than selecting young bulls in the larger conventional population based on EBV; also without the existence of G×E interactions. Only for pronounced G×E interactions (r_g=0.5), and for highly accurate GEBV for natural service bulls (r_mg>0.9), results suggests the use of genotyped organic natural service bulls instead of implementing an AI program. Inbreeding coefficients of selected bulls and their offspring were generally lower when basing selection decisions for young bulls on GEBV compared with selection strategies based on pedigree indices.     

Linear and non-linear regression analysis of genotype X environment interactions in pearl millet

Summary Regression analysis was computed on the grain yield of 15 single cross F₁ hybrids of pearl millet (Pennisetum typhoides (Burm.) S. & H.) evaluated in 20 environments at 19 sites in India to assess the nature of genotype X environment interactions. Linear, quadratic, cubic, twoand three-intersecting straight line models were examined for fit. The interactions of six hybrids viz. MH 110, MH 113, MH 114, MH 115, MH 120 and MBH 110 were explained by the linear regression model. The response of the remaining nine hybrids was largely non-linear. The two and three-intersecting straight line models fit better than the quadratic and cubic models and explained non-linearity of response. The two-intersecting straight line models fit for 6 hybrids MH 106, MH 107, MH 112, MH 116, MH 117 and BJ 104. The response of MH 109 was best explained by a three-intersecting straight line model, but there still existed a significant remainder variation. The truncation of environmental range by assuming moving division points was more efficient than the fixed division points for the segmental regression models. The stability of hybrid varieties on the best fitting model has been discussed.     

Genotype x environment interaction of crossover type: detecting its presence and estimating the crossover point

Genotype-environment interaction (GEI) introduces inconsistency in the relative rating of genotypes across environments and plays a key role in formulating strategies for crop improvement. GEI can be either qualitative (i.e., crossover type) or only quantitative (i.e., non-crossover type). Since the presence of crossover-type interaction has a strong implication for breeding for specific adaptation, it is important to assess the frequency of crossover interactions. This paper presents a test for detecting the presence of crossover-type interaction using the response-environment relationship and enumerates the frequency of crossovers and estimation of the crossover point (CP) on the environment axis, which serves as a cut-off point for the two environments groups where different/specific selections can be made. Sixty-four barley lines with various selection histories were grown in northern Syria and Lebanon giving a total of 21 environments (location-year combinations). Linear regression of the genotypic response on the environmental index represented a satisfactory model, and heterogeneity among regressions was significant. At a 5% level of significance, 38% and 19% of the pairs showed crossover interactions when the error variances were considered heterogeneous and homogeneous, respectively, implying that an appreciable number of crossovers took place in the case of barley lines responding to their environments. The CP of 1.64 t/ha, obtained as the CP of regression lines between the genotype numbers 19 and 31, provided maximum genotype x environment-group interaction. Across all environments, genotype nos. 59 and 12 stood first and second for high yield, respectively. The changes in the ranks of genotypes under the groups of environments can be used for selecting specifically adapted genotypes. Received: 25 January 1999 / Accepted: 16 March 1999     

Determining environmental covariates which explain genotype environment interaction in winter wheat through probe genotypes and biadditive factorial regression

Genotype-environment interaction has been analyzed in a winter-wheat breeding network using bi-additive factorial regression models. This family of models generalizes both factorial regression and biadditive (or AMMI) models; it fits especially well when abundant external information is available on genotypes and/or environments. Our approach, focused on environmental characterization, was performed with two kinds of covariates: (1) deviations of yield components measured on four probe genotypes and (2) usual indicators of yield-limiting factors. The first step was based on the analysis of a crop diagnosis on four probe genotypes. Difference of kernel number to a threshold number (DKN) and reduction of thousand-kernel weight from a potential value (RTKW) were used to characterize the grain-number formation and the grain-filling periods, respectively. Grain yield was analyzed according to a biadditive factorial regression model using eight environmental covariates (DKN and RTKW measured on each of four probe genotypes). In the second step, the usual indicators of yield-limiting factors were too numerous for the analysis of grain yield. Thus a selection of a subset of environmental covariates was performed on the analysis of DKN and RTKW for the four probe genotypes. Biadditive factorial regression models involved environmental covariates related to each deviation and included environmental main effect, sum of water deficits, an indicator of nitrogen stress, sum of daily radiation, high temperature, pressure of powdery mildew and lodging. The correlations of each environmental covariate to the synthetic variates helped to discard those poorly involved in interaction (with | correlation | <0.3). The grain yield of 12 genotypes was interpreted with the retained covariates using biadditive factorial regression. The models explained about 75% of the interaction sums of squares. In addition, the biadditive factorial regression biplot gave relevant information about the interaction of the genotypes (interaction pattern and sensitivities to environmental covariates) with respect to the environmental covariates and proved to be interesting for such an approach. Received: 8 March 1999 / Accepted: 29 July 1999     

Variation on reaction norm in lodgepole pine natural populations

Variations on the norm of reaction among ten natural lodgepole pine populations sampled from three lodgepole pine subspecies (Pinus contorta ssp. contorta, ssp. latifolia and ssp. murrayana) were studied by using 20 year heights measured in 57 provenance test sites across interior British Columbia (B.C.). There were significant population by site interactions. Concurrent joint regression and the AMMI model were used to dissect these population by environmental interactions. Joint regression analysis indicated that three populations (from the northwest) had a negative linear regression coefficient with environmental deviation, three (from central and southeast sites) had a positive regression coefficient and four (from the southwest) had a zero regression coefficient. The AMMI model revealed a similar pattern of reaction norm among the ten populations. But the three significant IPCA axes, which captured twice as much of the G × E sum of squares than joint regression, were more effective in separating the ten populations and associating their performance with the climate of test sites and their origin. The variation patterns of reaction norm in lodgepole pine populations demonstrated that adaptation of lodgepole pine natural populations to the various physical environments, at sub-species as well as at population level, was due largely to a balance between selection for high growth potential in less severe environments and selection for high cold hardiness in severe environments. Received: 4 May 2000 / Accepted: 10 November 2000     

The use of microsatellites for detecting DNA polymorphism, genotype identification and genetic diversity in wheat

A set of 20 wheat microsatellite markers was used with 55 elite wheat genotypes to examine their utility (1) in detecting DNA polymorphism, (2)in the identifying genotypes and (3) in estimating genetic diversity among wheat genotypes. The 55 elite genotypes of wheat used in this study originated in 29 countries representing six continents. A total of 155 alleles were detected at 21 loci using the above microsatellite primer pairs (only 1 primer amplified 2 loci; all other primers amplified 1 locus each). Of the 20 primers amplifying 21 loci, 17 primers and their corresponding 18 loci were assigned to 13 different chromosomes (6 chromosomes of the A genome, 5 chromosomes of the B genome and 2 chromosomes of the D genome). The number of alleles per locus ranged from 1 to 13, with an average of 7.4 alleles per locus. The values of average polymorphic information content (PIC) and the marker index (MI) for these markers were estimated to be 0.71 and 0.70, respectively. The (GT)_n microsatellites were found to be the most polymorphic. The genetic similarity (GS) coefficient for all possible 1485 pairs of genotypes ranged from 0.05 to 0.88 with an average of 0.23. The dendrogram, prepared on the basis of similarity matrix using the UPGMA algorithm, delineated the above genotypes into two major clusters (I and II), each with two subclusters (Ia, Ib and IIa, IIb). One of these subclusters (Ib) consisted of a solitary genotype (E3111) from Portugal, so that it was unique and diverse with respect to all other genotypes belonging to cluster I and placed in subcluster Ia. Using a set of only 12 primer pairs, we were able to distinguish a maximum of 48 of the above 55 wheat genotypes. The results demonstrate the utility of microsatellite markers for detecting polymorphism leading to genotype identification and for estimating genetic diversity. Received: 15 May 1999 / Accepted: 27 July 1999     

Additive main effects and multiplicative interaction (AMMI) analysis of genotype-location interaction in variety trials repeated over years

Genotype-location (GL) interaction effects are of special interest for breeding programmes to identify adaptation targets, adaptive traits and test sites. These effects, generally having relatively low repeatability between years, should be studied on a multiyear basis in annual crops. Their assessment by additive main effects and multiplicative interaction (AMMI) analysis is currently defined for this situation. Two procedures based on cross validations are proposed for testing the GL-interaction principal component axes, exploiting the utilities of the computer programme MATMODEL. The use of Gollob’s F test, F _GH2 test, F _R test and the heuristic criterion based on the signal-to-noise ratio is also envisaged. The consistency of results provided by the testing procedures was verified on four data sets of different cereal crops. Gollob’s test tended to be the most liberal, while the F _GH2 test appeared somewhat more liberal than the F _R test. The signal-to-noise ratio gave results consistent with the F _R test considered at a P?0.01 level of significance. These criteria disagreed in two data sets with the conclusions provided by the two cross-validation procedures which, in turn, also disagreed in one data set. Preference could be given to different testing procedures depending on the number of test years, locations and genotypes.     

Characterisation of terminal sire sheep farm systems,based on a range of environmental factors: a case study in the context of genotype by environment interactions using Charollais lambs

The objective of this study was to define different terminal sire flock environments, based on a range of environmental factors, and then investigate the presence of genotype by environment interactions (G×E) between the environments identified. Data from 79 different terminal sire flocks (40 Texel, 21 Charollais and 18 Suffolk), were analysed using principal coordinate and non-hierarchical cluster analyses, the results of which identified three distinct environmental cluster groups. The type of grazing, climatic conditions and the use of vitamins and mineral supplements were found to be the most important factors in the clustering of flocks. The presence of G×E was then investigated using data from the Charollais flocks only. Performance data were collected for 12 181 lambs, between 1990 and 2010, sired by 515 different sires. Fifty six of the sires had offspring in at least two of the three different cluster groups and pedigree information was available for a total of 161 431 animals. Traits studied were the 21-week old weight (21WT), ultrasound muscle depth (UMD) and log transformed backfat depth (LogUFD). Heritabilities estimated for each cluster, for each trait, ranged from 0.32 to 0.45. Genetic correlations estimated between Cluster 1 and Cluster 2 were all found to be significantly lower than unity, indicating the presence of G×E. They were 0.31 (±0.17), 0.68 (±0.14) and 0.18 (±0.21) for 21WT, UMD and LogUFD, respectively. Evidence of sires re-ranking across clusters was also observed. Providing a suitable strategy can be identified, there is potential for the optimisation of future breeding programmes, by taking into account the G×E observed. This would enable farmers to identify and select animals with an increased knowledge as to how they will perform in their specific farm environment thus reducing any unexpected differences in performance.     

Comparative mapping of QTLs for agronomic traits of rice across environments by using a doubled-haploid population

We report here the RFLP mapping of quantitative trait loci (QTLs) which affect some important agronomic traits in cultivated rice. An anther culture-derived doubled-haploid (DH) population was established from a cross between indica and japonica rice varieties. A molecular linkage map comprising 137 markers was constructed based on this population which covered the rice genome at intervals of 14.8 cM on average. The linkage map was used to locate QTLs for such important agronomic traits as heading date, plant height, number of spikelets per panicle, number of grains per panicle, 1 000-grain weight and the percentage of seed set, by interval mapping. Evidence of genotype-by-environment interaction was found by comparing QTL maps of the same population grown in three diverse environments. A total of 22 QTLs for six agronomic traits was detected which were significant in at least one environment, but only seven were significant in all three environments; seven were significant in two environments and eight could only be detected in a single environment. However, QTLs-by-environment interaction was trait dependent. QTLs for spikelets and grains per panicle were common across environments while traits like heading date and plant height were more sensitive to environment. Received: 22 February 1996 / Accepted: 10 May 1996     

Detection of QTLs for genotype × environment interactions in tomato seeds and seedlings

Seed quality and seedling establishment are the most important factors affecting successful crop development. They depend on the genetic background and are acquired during seed maturation and therefor, affected by the maternal environment under which the seeds develop. There is little knowledge about the genetic and environmental factors that affect seed quality and seedling establishment. The aim of this study is to identify the loci and possible molecular mechanisms involved in acquisition of seed quality and how these are controlled by adverse maternal conditions. For this, we used a tomato recombinant inbred line (RIL) population consisting of 100 lines which were grown under two different nutritional environmental conditions, high phosphate and low nitrate. Most of the seed germination traits such as maximum germination percentage (G_max), germination rate (t₅₀) and uniformity (U₈₄₁₆) showed ample variation between genotypes and under different germination conditions. This phenotypic variation leads to identification of quantitative trait loci (QTLs) which were dependent on genetic factors, but also on the interaction with the maternal environment (QTL × E). Further studies of these QTLs may ultimately help to predict the effect of different maternal environmental conditions on seed quality and seedling establishment which will be very useful to improve the production of high-performance seeds.     

Induced resistance in the indeterminate growth of aspen (<Emphasis Type="Italic">Populus tremuloides</Emphasis>)

Studies of induction in trees have examined rapid induced resistance (RIR) or delayed induced resistance (DIR), but have not examined induction that occurs in leaves produced by indeterminately growing trees subsequent to, but in the same season as, damage. We refer to induction that occurs during this time period as intermediate-delayed induced resistance (IDIR). We assessed the influences of genetic and environmental factors, and their interactions, on temporal and spatial variation in induction and on tradeoffs between induced and constitutive levels of resistance in indeterminately growing saplings of aspen (Populus tremuloides). We utilized a common garden of 12 aspen genotypes experiencing two levels of defoliation and two levels of soil nutrients. We assessed concentrations of phenolic glycosides and condensed tannins in damaged leaf remnants collected 1 week after defoliation to examine rapid and local induction, and in undamaged leaves produced 8 weeks after defoliation to assess intermediate-delayed and systemic induction. In general, tannins showed RIR, while phenolic glycosides expressed IDIR. For both classes of allelochemicals, we found high estimates of broad-sense heritability and genetic variation in both induced and constitutive levels. Genetic variation may be maintained by both direct costs of allelochemicals and by costs of inducibility (phenotypic plasticity). Such costs may drive the tradeoff exhibited between induced and constitutive levels of phenolic glycosides. IDIR may be important in reducing total-season tissue loss by providing augmented resistance against late summer herbivores in trees that have experienced damage earlier in the season. Herbivore-resistant compensatory growth is especially beneficial to young trees growing in competitive environments.     

Mapping of quantitative trait loci controlling adaptive traits in coastal Douglas-fir. I. Timing of vegetative bud flush

Thirty three unique quantitative trait loci (QTLs) affecting the timing of spring bud flush have been identified in an intraspecific mapping population of coastal Douglas-fir [Pseudotsuga menziesii (Mirb.) Franco var. menziesii]. Both terminal and lateral bud flush were measured over a 4-year period on clonal replicates at two test sites, allowing for the repeated estimation of QTLs. QTLs were detected on 12 linkage groups and, in general, each explained a small proportion of the total phenotypic variance and were additive in effect. Several QTLs influence the timing of bud flush over multiple years, supporting earlier evidence that the timing of bud flush through developmental stages is under moderate to strong genetic control by the same suite of genes through developmental stages. However, only a few QTLs controlling the timing of bud flush were detected at both test sites, suggesting that geographic location plays a major role in the phenology of spring growth. A small number of QTLs with year and site interactions were also estimated. Received: 20 July 2000 / Accepted: 19 October 2000     

Genomic regions involved in productivity of two interspecific poplar families in Europe. 1. Stem height, circumference and volume

Interspecific hybrids of Populus species are known for their superior growth. In this study, we examined the effect of the genetic background and contrasting environmental conditions on growth and searched for quantitative trait loci (QTL) affecting growth traits. To this end, two hybrid poplar families resulting from controlled crosses, Populus deltoides ‘S9-2’ × P. nigra ‘Ghoy’ (D × N, 180 F₁) and P. deltoides ‘S9-2’ × P. trichocarpa ‘V24’ (D × T, 182 F₁), were grown at two contrasting sites, Northern Italy and Central France. At the end of the second growing season, tree dimensions (stem height, circumference, and volume) were assessed. The performances of both families significantly differed within and between sites. Tree volume was significantly larger at the Italian site as compared to the French site. Genotype by environment interactions were significant but low for both families and for all growth traits. Tight correlations among the individual growth traits indicated that there may be a common genetic mechanism with pleiotropic effects on these growth traits. In line with previous studies, linkage groups I, VII, IX, X, XVI, XVII, and XIX appeared to have genomic regions with the largest effects on growth traits. This study revealed that (1) both families have high potential for selection of superior poplar hybrids due to the pronounced heterosis (hybrid vigor) and the large genetic variability in terms of growth and (2) the choice of site is crucial for poplar cultivation. Dillen and Storme contributed equally to the work. An erratum to this article can be found at     

Genotype × environment interaction of Hevea clones in traditional and non-traditional rubber growing regions of Vietnam

The main objective of this research was to assess genotype?×?environment interactions and the stability status of 24 Hevea clones along with 2 checks in terms of latex yield and girth growth under two different environmental conditions representing for traditional and non-traditional rubber growing regions of Vietnam. Genotype?×?environment interactions were found to be highly significant for both latex yield and girth at opening, indicating that the latex yield and girth growth of the studied rubber clones were significantly affected by the environmental conditions of the location. Stability analyses indicated that LH 94/359, LH 91/579 and LH 94/481 were the best clones for highest latex yield and adaptability to traditional, non-traditional and both rubber growing regions of Vietnam, respectively. In terms of girth at opening, the stability analyses revealed that these clones were also accepted as the average girth growth and adaptability to both rubber growing regions of Vietnam.