How many years of evaluation are needed to select new productive cocoa clones?

In perennial plants, the selection of new genotypes requires several years of evaluation to get a reliable estimate of the genetic value and performance potential of the trees. The cocoa tree, which is cultivated for its beans, produces continuously throughout the entire duration of trial plots or field stands, which can last up to 30 years. The genetic value of the trees' productivity over the entire production period is therefore difficult to assess. In particular, how many years are needed to estimate the potential of the trees? To answer this question, the analysis of a trial comparing 42 clones in a statistical design, located in Costa Rica, was proposed. The individual production of each of the 1187 trees in this trial was evaluated over 18 consecutive years. Results indicate that the first year's production is not a reliable predictor of cumulative production over 18 years given its genetic correlation of 0.68. It is necessary to have an accumulation of 4 years for the genetic correlation with the 18-year accumulation to be higher than 0.85, and an accumulation of 7 years for the genetic correlation to be higher than 0.90. Longitudinal data analysis allows for a better understanding of the links between production of the successive years. In this trial, the seventh year was a year of high production followed by a slight decrease in the eighth year. Determining an evaluation period that allows for a reliable estimate of genetic value should therefore be based on the study of production kinetics. Indeed, it seems necessary to continue the evaluation of tree production until a high annual production that reveals the productive capacity of the genotypes. Studies on production kinetics also make it possible to envisage genetic improvement for better sustainability of cocoa trees to be proposed in appropriate cropping systems.

Key-words: cocoa production kinetics, selection strategy, sustainability