13. Tools/methods for genotype × environment interaction analysis

Analysis of GEI is important to obtain information on the performance of genotypes in terms of adaptability and stability. Analysis of variance is performed across environments in order to identify the presence of GEI in multilocation trials. When the GEI variance is found to be significant, then one of the various methods for measuring the stability of genotypes can be used to identify the most stable genotype(s). Several statistical methods have been proposed for analysis and interpretation of GEI [63–66]. The joint regression analysis [67–69] method has been widely used; nonetheless, several limitations of the method have been stated [70, 71]. For example, see [48]. The PCA method has the ability to overcome the limitations associated with the linear regression method by giving more than one statistic, that is, the scores on the principal component axes, to describe the response of a genotype. Another method which has been proposed for analysis of GEI is the cluster analysis which is a numerical classification technique that defines groups of clusters of individuals [48, 72]. Currently, the additive main effects and multiplicative interaction (AMMI) model [64, 71] and genotype main effect plus genotype × environment interaction (GGE) biplot methodology [66] are the two most powerful statistical tools used by many researchers for the analysis of multilocational trial data. The AMMI model combines the analysis of variance for the genotype and environment main effects with principal component analysis of the genotype × environment interaction. It also provides a better prediction assessment and a valuable approach for understanding GEI and obtaining better yield estimates. The interaction is described in the form of a biplot display, where PCA scores are plotted against each other and provides visual inspection and interpretation of the GEI components. Integrating biplot display and genotypic stability statistics enable genotypes to be grouped based on similarity of performance across diverse environments. Similarly, the GGE biplot analysis enables visual (graphical) presentation of interaction estimate. This method also combines analysis of variance and PCA by partitioning together sums of squares of genotypes and sums of squares of GEI (which are relevant in genotype evaluation) using PCA method. The biplot technique is used for the presentation and estimation of genotypes in different environments [73]. The GGE biplot shows the first two principal components (PC1 and PC2) which are obtained by decomposition of singular values of multilocation trials yield data. GGE biplot analysis enables the identification of the genotypes with the highest yields in different environments, comparison of their performances in different environments, identification of ideal genotype, as well as mega-environments (model of regional distribution or target environment) [74, 75].

trials are conducted to ascertain crop performance in a wide range of environments for

Genotype × Environment Interaction: A Prerequisite for Tomato Variety Development

http://dx.doi.org/10.5772/intechopen.76011

83

Living organisms are made up of genes whose expression are subject to modification by the environment; therefore, genotypic expression of a phenotype is environmentally dependent [84]. This is because genotypes exhibit different levels of phenotypic expression under different environmental conditions resulting in crossover performances [85]. Crossover performances by genotypes in different environments result from differential genotypic responses under varying environmental conditions [63, 86]. This results in genotype by environment interaction where one genotype gives its maximum performance in one environment by performing poorly in another environment. In G × E interaction, the magnitude of the observed genetic variation changes from one environment to another and tends to be larger in better environ-

The objective of most plant breeders is to develop new varieties that will perform consistently well across multiple environments. However, significant G × E interaction has been reported for most quantitative traits in tomato particularly for fruit yield and quality traits such as lycopene, total soluble solids, vitamin C, etc. [19, 88]. A tomato variety with improved fruit quality in one environment may not necessarily perform the same in another location due to differential responses to the different environmental conditions prevailing in the different locations. Environmental factors such as soil, moisture, temperature, light intensity, humidity, rainfall, photoperiod, and agronomic practices play important role in the expression of the genes controlling the trait of interest. This results in different phenotypic expression among locations. Genotype × environment interaction effect complicates the selection of suitable varieties by breeders because elite varieties developed for one location may not perform the same in different locations. In some cases, the quality of fruits of tomatoes is significantly influenced by genotype by environment interaction. Such interactions confound the selection of the superior cultivars by altering their relative productiveness in different environments. For instance, see [89]. Other studies [90] also reported significant G × E interaction effect on total sugars among six tomato varieties grown under field and screenhouse conditions. This problem implies that tomato varieties that were developed and selected under field conditions may not perform to its full potential when farmers grow them under controlled environments. Therefore, the extent of G × E interactions effect for most traits of economic importance needs to be taken into account during the selection process in order to obtain crop varieties that will

Breeding of crops involves different attributes of the genetic materials that are subject to variation in environmental conditions [91]. In some cases, direct selection is slow due to low heritability, polygenic control, epistasis, and significant G × E interaction on the trait of interest

14.2. Problems of genotype × environment interaction effect on selection

give consistent performance across environments and seasons.

14.3. Elimination of genotype × environment interaction

adaptability and stability in performance [47].

ments than poorer environments [87].

14.1. Causes of genotype × environment interaction

Several researchers have compared the efficiency of AMMI and GGE biplot for analyzing GEI. According to Yan and others, the major disadvantage of the AMMI model is that it is insensitive to the most important part of the crossover GEI [75]. Moreover, the AMMI model does not offer any advantage to the breeder for genotypic and site evaluation when analyzing METs data because there is no clear biological separation between the two terms, genotype and GEI. However, the GGE biplot is a powerful statistical model that takes care of some of the disadvantages of AMMI. The method is an effective statistical tool for identifying the best performing cultivar in a given environment and the most suitable environment for each cultivar, comparison of any pair of cultivars in individual environments, the best cultivars for each environment and mega-environment differentiation, average yield and stability of the genotypes, and the discriminating ability and representativeness of the environments [75–77]. Gruneberg and others indicated that AMMI was highly effective for the analysis of MET [78]. Kandus and others also revealed that the AMMI model is the best model for describing the GEI [79]. Stojaković and others [80] and Mitrovic and others [81] found that both models provided similar results. However, contrary to these reports, [75, 82, 83] concluded in their comparison of both models that the GGE biplot was superior to the AMMI biplot in mega-environment analysis and genotype evaluation.

#### 14. Prospects and problems of G × E

The phenomenon of genotype × environment interaction refers to the differential performance of genotypes in different environments that affect the efficiency of selection in a breeding program. G × E interaction arises due to the differences in the sensitivities of genotypes to the different environmental conditions. In order to mitigate the effect of G × E interaction, crops need to be tested in several environments to assess their specific and broad adaptation [53, 76]. Though tomatoes do well in both tropical and temperate climates, its performance can vary with respect to the environments [18]. Prior to the release of every crop variety, multilocation trials are conducted to ascertain crop performance in a wide range of environments for adaptability and stability in performance [47].

#### 14.1. Causes of genotype × environment interaction

effects with principal component analysis of the genotype × environment interaction. It also provides a better prediction assessment and a valuable approach for understanding GEI and obtaining better yield estimates. The interaction is described in the form of a biplot display, where PCA scores are plotted against each other and provides visual inspection and interpretation of the GEI components. Integrating biplot display and genotypic stability statistics enable genotypes to be grouped based on similarity of performance across diverse environments. Similarly, the GGE biplot analysis enables visual (graphical) presentation of interaction estimate. This method also combines analysis of variance and PCA by partitioning together sums of squares of genotypes and sums of squares of GEI (which are relevant in genotype evaluation) using PCA method. The biplot technique is used for the presentation and estimation of genotypes in different environments [73]. The GGE biplot shows the first two principal components (PC1 and PC2) which are obtained by decomposition of singular values of multilocation trials yield data. GGE biplot analysis enables the identification of the genotypes with the highest yields in different environments, comparison of their performances in different environments, identification of ideal genotype, as well as mega-environments (model of

Several researchers have compared the efficiency of AMMI and GGE biplot for analyzing GEI. According to Yan and others, the major disadvantage of the AMMI model is that it is insensitive to the most important part of the crossover GEI [75]. Moreover, the AMMI model does not offer any advantage to the breeder for genotypic and site evaluation when analyzing METs data because there is no clear biological separation between the two terms, genotype and GEI. However, the GGE biplot is a powerful statistical model that takes care of some of the disadvantages of AMMI. The method is an effective statistical tool for identifying the best performing cultivar in a given environment and the most suitable environment for each cultivar, comparison of any pair of cultivars in individual environments, the best cultivars for each environment and mega-environment differentiation, average yield and stability of the genotypes, and the discriminating ability and representativeness of the environments [75–77]. Gruneberg and others indicated that AMMI was highly effective for the analysis of MET [78]. Kandus and others also revealed that the AMMI model is the best model for describing the GEI [79]. Stojaković and others [80] and Mitrovic and others [81] found that both models provided similar results. However, contrary to these reports, [75, 82, 83] concluded in their comparison of both models that the GGE biplot was superior to the AMMI biplot in mega-environment

The phenomenon of genotype × environment interaction refers to the differential performance of genotypes in different environments that affect the efficiency of selection in a breeding program. G × E interaction arises due to the differences in the sensitivities of genotypes to the different environmental conditions. In order to mitigate the effect of G × E interaction, crops need to be tested in several environments to assess their specific and broad adaptation [53, 76]. Though tomatoes do well in both tropical and temperate climates, its performance can vary with respect to the environments [18]. Prior to the release of every crop variety, multilocation

regional distribution or target environment) [74, 75].

82 Recent Advances in Tomato Breeding and Production

analysis and genotype evaluation.

14. Prospects and problems of G × E

Living organisms are made up of genes whose expression are subject to modification by the environment; therefore, genotypic expression of a phenotype is environmentally dependent [84]. This is because genotypes exhibit different levels of phenotypic expression under different environmental conditions resulting in crossover performances [85]. Crossover performances by genotypes in different environments result from differential genotypic responses under varying environmental conditions [63, 86]. This results in genotype by environment interaction where one genotype gives its maximum performance in one environment by performing poorly in another environment. In G × E interaction, the magnitude of the observed genetic variation changes from one environment to another and tends to be larger in better environments than poorer environments [87].

#### 14.2. Problems of genotype × environment interaction effect on selection

The objective of most plant breeders is to develop new varieties that will perform consistently well across multiple environments. However, significant G × E interaction has been reported for most quantitative traits in tomato particularly for fruit yield and quality traits such as lycopene, total soluble solids, vitamin C, etc. [19, 88]. A tomato variety with improved fruit quality in one environment may not necessarily perform the same in another location due to differential responses to the different environmental conditions prevailing in the different locations. Environmental factors such as soil, moisture, temperature, light intensity, humidity, rainfall, photoperiod, and agronomic practices play important role in the expression of the genes controlling the trait of interest. This results in different phenotypic expression among locations. Genotype × environment interaction effect complicates the selection of suitable varieties by breeders because elite varieties developed for one location may not perform the same in different locations. In some cases, the quality of fruits of tomatoes is significantly influenced by genotype by environment interaction. Such interactions confound the selection of the superior cultivars by altering their relative productiveness in different environments. For instance, see [89]. Other studies [90] also reported significant G × E interaction effect on total sugars among six tomato varieties grown under field and screenhouse conditions. This problem implies that tomato varieties that were developed and selected under field conditions may not perform to its full potential when farmers grow them under controlled environments. Therefore, the extent of G × E interactions effect for most traits of economic importance needs to be taken into account during the selection process in order to obtain crop varieties that will give consistent performance across environments and seasons.

#### 14.3. Elimination of genotype × environment interaction

Breeding of crops involves different attributes of the genetic materials that are subject to variation in environmental conditions [91]. In some cases, direct selection is slow due to low heritability, polygenic control, epistasis, and significant G × E interaction on the trait of interest [92]. To mitigate the confounding effect of G × E interaction on selection efficiency, plant breeders have devised strategies to ensure progress in selection efficacy. For this reason, genotypes are tested in diverse environments to assess their adaptability and stability [85]. After this sound, analyses are carried out using the appropriate software to assess the extent of G × E interaction effect. Genotypes whose G × E effects are not significant are considered to be stable and therefore selected [62].

problem of GEI is that its effect thwarts the selection of suitable varieties by breeders because elite varieties developed for one location may not perform the same in different locations. In some cases, the quality of fruits of tomatoes is significantly influenced by genotype by environment interaction. Such interactions confuse the selection of the superior cultivars by altering their relative productiveness in different environments. Though tomatoes do well in both tropical and temperate climates, its performance can vary with respect to the environments.

, Joseph Adjebeng-Danquah<sup>2</sup>

Genotype × Environment Interaction: A Prerequisite for Tomato Variety Development

, Essie Blay3 and Hans Adu-Dapaah<sup>1</sup>

,

http://dx.doi.org/10.5772/intechopen.76011

85

Author details

Agyemang Danquah<sup>3</sup>

References

1958;26:349-358

2010. pp. 41-60

DOI: 10.1017/s0954579407000478

Plant Molecular Biology Reporter. 1991;9:210-220

sequences. Molecular and General Genetics. 1988;213:254-261

Michael Kwabena Osei1,3\*, Benjamin Annor1

, Eric Danquah<sup>3</sup>

2 CSIR-Savannah Agricultural Research Institute, Tamale, Ghana

3 West Africa Centre for Crop Improvement, University of Ghana, Legon, Ghana

[1] Haldane JBS. The interaction of nature and nurture. Annals of Eugenics. 1946;13:197-202

[2] McBride G. The environment and animal breeding problems. Animal Breeding Abstracts.

[3] Tabery J. Biometric and developmental gene-environment interactions: Looking back, moving forward. Development and Psychopathology. 2007;19:961-976. PMID: 17931428.

[4] Sesardic N. Making Sense of Heritability. Cambridge: Cambridge University Press; 2005. p. 48

[5] Tabery J, Griffiths PE. Historical and philosophical perspectives on behavioural genetics and developmental science. In: Hood KE, Halpern CT, Greenberg G, Lerner RM, editors. Handbook of Developmental Science, Behavior, and Genetics. Malden: Wiley-Blackwell;

[6] Arumuganathan K, Earle ED. Nuclear DNA content of some important plant species.

[7] Zamir D, Tanksley SD. Tomato genome is comparised mainly of fast evolving single copy

[8] Vallejos CE, Tanksley SD, Bernatzky R. Localization in the tomato genome of DNA restriction fragments containing sequences coding for Rdna (45s), ribulose bisphosphate

carboxylase and chlorophyll a/b binding protein. Genetics. 1986;112:93-105

Address all correspondence to: oranigh@hotmail.com

1 CSIR-Crops Research Institute, Kumasi, Ghana

Stability analysis is performed to estimate the performance of genotypes as linear function of the level of productivity in each environment [93]. Eberhart and Russell suggested joint regression analysis to estimate the average performance of a genotype in different environments relative to the mean performance of all genotypes in the same environment [68]. The use of multiplicative models which include the additive main effect and multiplicative interaction (AMMI) model has also been used to assess the stability of other crops [94, 95]. The AMMI model allows fitting of the sum of several multiplicative terms rather than only one multiplicative term in dissecting the performance of genotypes in different environments [93]. Yan also suggested the use of the genotype and genotype × environment interaction (GGE) biplot to graphically visualize genotypic performance across several environments [96]. The use of these strategies will enable the breeder to make informed decisions in where to place which variety based on their adaptability for optimum performance.

#### 15. Conclusion

The pounding prominence of tomato as a vegetable is reflected by large volume of research on almost all aspects of the crop. In every crop improvement program, promising genotypes are tested for their performance for some years at a number of sites, to identify genotypes which possess the dual qualities of high-yield sustainability to adverse changes in environment condition. This interplay refers to genotype by environment interaction. A genotype × environment interaction is a change in the relative performance of a character of two or more genotypes measured in two or more environments. Its origin is linked to two concepts: biometric and developmental interaction. Interactions may therefore involve changes in order for genotypes between environments and changes in the absolute and relative magnitude of the genetic, environmental, and phenotypic variances between environments. These can further be classified as no GEI, non-crossover interaction, and crossover interaction. Complex quantitative traits, such as yield, with multiple contributing traits are highly influenced by environment interaction effects. Tomato production, though weather dependent and highly seasonal, can be grown under both field and greenhouse conditions (controlled environment). Researchers perform multilocational trials to evaluate new or improved genotypes across multiple environments (locations and years), before they are promoted for release and commercialization. This organized approach helps increase yield stability of new crop varieties in stress-prone environments. To obtain information on the performance of the genotypes in terms of adaptability and stability, an analysis of the GEI is paramount. Even though several statistical methods have been proposed for analysis and interpretation of GEI, the joint regression analysis method has been widely used; nonetheless, it has numerous limitations. Many other researchers have also found AMMI and GGE biplot efficient for analyzing GEI. A major problem of GEI is that its effect thwarts the selection of suitable varieties by breeders because elite varieties developed for one location may not perform the same in different locations. In some cases, the quality of fruits of tomatoes is significantly influenced by genotype by environment interaction. Such interactions confuse the selection of the superior cultivars by altering their relative productiveness in different environments. Though tomatoes do well in both tropical and temperate climates, its performance can vary with respect to the environments.
