Genotype x environment interaction studies in rubber (Hevea) clones [Malaysia]

The importance and impact of genotype x environment (GxE) effects on Hevea planting recommendations have been long recognised. However, studies on GxE interaction in rubber were initiated only in the seventies. This paper reviews previous studies on GxE in Hevea and examines the extent, nature and the major factors of GxE interaction. Data on the girth and yield derived from two sets of clonal trials, namely, RRIM 700 series (16 clones tested in 6 locations) and 1954 foreign exchange clone trials (91 clones tested in 5 locations) were used in this study. The data were subject to various statistical analyses. In general, the GxE interaction was present and significant for yield and girth. However, total variance of the GxE interaction was rather small, compared to clonal and environmental effects. The GxE variance components accounted for 12.6 percent and 11.4 percent to 19.3 percent for girth and yeild, respectively. When the results from previous studies were combined, the contribution of GxE variance components over the total variance components for various characters of Hevea was 15.7 percent. The nature of GxE interaction was studied using joint regression analysis. The results showed that differences in linear regression accounted for only a small proportion of the total GxE interaction effects, suggesting the non-linear nature of the GxE interaction for most of the characters in Hevea and the involvement of more than one major environmental factor (probably wind damage and disease, in this study), causing the GxE interaction. Stability estimates of clones were carried out using various statistical techniques such as regression and deviation, coefficient of variation and variance, ecovalence and other stability variances. In general, all the above methods could help classify clones according to their stability of response toward the variation of environment as guideline for choice of clones for planting. Regression technique however needs some caution in application due to the non-linearity of the GxE interaction in Hevea. In contrast, simple statistical parameters such as CV or variance and mean of genotypes over different trial sites can serve as practical and useful guides in clonal selection