Genetic parameters were estimated for frosty pod rot resistance and yield components in cacao using a linear mixed model approach (Restricted Maximum Likelihood and Best Linear Unbiased Prediction). Two progeny trials established in the Rosario Izapa Experimental Field of the INIFAP in Chiapas, Mexico, were evaluated for frosty pod rot incidence and yield components, including the number of pods per tree, bean dry weight per tree, number of beans per pod, bean dry weight per pod, seed index and pod index, for four years. Genetic correlations, age-age correlations and heritabilities were estimated. The narrow-sense heritability for frosty pod rot resistance was 0.46 and for the yield components it ranged from 0.14 to 0.99. Backward selection, using UF-273 and Pa-169 as parents for new combinations, resulted in a reduction from -36.54 to -29.05% in frosty pod rot, respectively. The genetic correlations between bean dry weight per tree and frosty pod rot, number of pods per tree, number of beans per pod and pod index were -0.72, 0.89, 0.25 and 0.40, respectively (all significant at p < 0.01). Age-age correlations and heritabilities from data subsets indicated that the first two years of evaluation and evaluating for only two specific months (August-September) in the year, should result almost in the same genetic estimates as using all four years data, for key traits, including frosty pod rot resistance, reducing the costs and accelerating breeding decisions.

In the two trials, individual trees were assessed weekly for mature pods during eight months from April to November for four consecutive years, beginning in 2009. The measured traits used in the analysis were the: frosty pod rot incidence measured by the total number of diseased pods per tree, bean dry weight per tree (DW, kg/tree), total number of pods per tree (NPd), number of beans per pod measured in all healthy pods harvested (NBPd), bean dry weight per pod (BWPd, g/pod), the seed index (SI) and pod index (PI). The FPR was obtained as the ratio of the diseased pods by the number of all pods harvested in the tree. All yield traits were taken on healthy pods, which were harvested, labeled and transported to the laboratory in polyethylene bags and processed at the same day. Beans were dried at 60 °C for 48 hours in a forced air oven (Binder® Germany). The SI was obtained as the proportion of BWPd and the NBPd; and the PI was obtained as the number of pods required to produce one kg of dry beans.

NA in Phenotypic data. Parental values not observed were predicted based on progeny trial results. Aiming the selection of  potential parent clones for further breeding, backward selection was performed based on the results of the progeny test.