Optimum stratification for yield estimation in cocoa (Theobroma Cacao L.)

Abstract

Investigation on "Optimum stratification for yield estimation In cocoa (Theobroma cacao. L)" was carried out in the department of Agricultural Statistics, College of Horticulture, V ellanikkara during 1997-99 using data on 1025 'F orestro' variety cocoa trees from the Cadbury-KAU Co-operative Cocoa Research Project (CCRP), College of Horticulture, Kerala Agricultural University. Four different stratification rules, viz, cumulative root 'f rule, cumulative cube root 'f rule, Ekman's rule and an iterative procedure for function minimisation, were used to arrive at boundary points of strata. For each of these rules stratification was carried out on average yield of yester years, canopy spread, value of HxG2 where 'H' is height and 'G' is the girth of the trees, the first principal component derived from these variates and height and girth of the trees and regression estimate of the study character with the predictor variables mentioned above. Sampling variance of the estimator of the population mean under Neyman allocation for two to five strata situations was estimated in each case, assuming a uniform sample size of 200. Different stratification rules and stratification variables were compared using these estimates. No single rule was found to be appropriate for all the stratification variables and for different number of strata. But in most of the cases the cumulative root 'f rule was found to be good for smaller number of strata followed by the cumulative cube root 'f rule. For large number of strata, the iterative procedure performed consistently well compared to all the other methods. In the case of stratification based on the canopy spread, the Ekman' s method was found to be good for different numbers of strata. Regarding the stratification variables, the average yield of yester years was found to be best followed by the regression estimate in the sense of resulting in smaller sampling variance of the estimator of the population mean. Stratification based on value of HxG2 and the first principal component were found to be inferior.

Average yield of yester years or regression estimate of yield could be used as covariate to perform the analysis of covariance for experiments in cocoa and also blocking of trees could be done based on these for the conduct of planned experiments on cocoa.