Yield estimation in cocoa with partial harvest data

Abstract

Methodology for estimation of crop production varies depending on the nature and duration of the crop. Perennial crops like cocoa, where the harvest is done throughout the year at different intervals, crop cutting experiments are laborious and time consuming. In this study, a methodology for early estimation of cocoa yield based on a few observations is proposed. A model based on observations on number of pods present on the tree during the beginning of major harvest season and after six months have been proposed to estimate the annual yield of cocoa. The accuracy of the methodology was validated using the test data. Subsequently, yield estimation of cocoa in major cocoa growing districts of Kerala and Karnataka were carried out. The effect of irrigation type, major cropping systems and age of the cocoa garden on the yield were also studied using Hierarchical linear modelling approach. The study revealed that cocoa yield in coconut based cropping system is better than any other systems.