Abstract
As a part of the genetic improvement programme of Eucalyptus grandis in Kerala, India, plus trees were selected from the existing plantations and these were screened for their genetic superiority. Sixty four families were tested at two sites, one in grass land (Vallakkadavu) and another in area near evergreen forest (Muthanga). The results showed that there was highly significant family-site interaction and growth was better at the site near evergreen forest. It is also seen that performance of families differ from site to site. In general the genotypic coefficient of variation is low, which indicates the narrow genetic base and that the seed origin of all the E. grandis plantations in Kerala must be the same. Both family and single tree heritability as well as the genetic gain were also very low. However, at Muthanga 7 percent gain can be expected for growth by selecting 10% of the total families while at Vallakkadavu it is only 4 per cent. Family selection is recommended in cases of low single tree heritability. Significant amount of non-additive genetic variation was reported in E. grandis for growth characters and this also supports the relevance of family selection as well as clonal propagation.
Keywords