Methodological approaches for the selection of genotypes in a progeny trial of Dacryodes edulis (G. Don) H. J. Lam in Cameroon

The World Agroforestry Centre (ICRAF) initiated a program in Cameroon for the selection of improved genotypes of Dacryodes edulis for domestication. There are major challenges in such selection because superior performance of individual trees may be based on site conditions rather than on genetic differences. Open-pollinated seeds were collected from mother trees in four provenances in Cameroon and planted in a progeny trial at Minkoameyos, Cameroon. Progeny from the same mother tree are referred to as family. After one year of growth in the field, growth traits were measured annually for five years and fruit traits were measured in two successive years. Three methodological approaches were investigated: family and within-family selection (combined selection), ranking for phenotypic selection and correlation for multi-trait and early selection. Variation in growth and fruit traits among families was not significant for most provenances, indicating that family selection would not be effective for the selection of genotypes in the trial. Phenotypic selection resulted in the retention of a small number of individuals of all families in the trial. There were strong positive correlations between fruit traits and between fruit and growth traits, while negative correlations were found between fruit traits and yield. The correlations indicate that selecting for one fruit trait may also improve the other fruit traits but reduce fruit yield. Positive and significant juvenile-mature correlations for growth were also found in this trial, indicating that early selection may be possible based ongrowth. Based on these results, phenotypic selection is recommended in the progeny trial for the domestication program in Cameroon and further research on early selection is needed as the trial advances in age