Numerous collecting expeditions of Theobroma cacao L. germplasm have been undertaken in Latin-America. However, most of this germplasm has not contributed to cacao improvement because its relationship to cultivated selections was poorly understood. Germplasm labeling errors have impeded breeding and confounded the interpretation of diversity analyses. To improve the understanding of the origin, classification, and population differentiation within the species, 1241 accessions covering a large geographic sampling were genotyped with 106 microsatellite markers. After discarding mislabeled samples, 10 genetic clusters, as opposed to the two genetic groups traditionally recognized within T. cacao, were found by applying Bayesian statistics. This leads us to propose a new classification of the cacao germplasm that will enhance its management. The results also provide new insights into the diversification of Amazon species in general, with the pattern of differentiation of the populations studied supporting the palaeoarches hypothesis of species diversification. The origin of the traditional cacao cultivars is also enlightened in this study.

No abstract

In order to cope with current challenges facing world cocoa production and the obvious lack of sustainability of the intensive model proposed to farmers, more ecologically efficient cocoa cropping systems must be developed, based in particular on a higher cultivated biodiversity level. The performances of cocoa multispecies systems, which involve multiple and hard to quantify interactions, are, however, more complicated to assess than that of monospecies systems. Despite this hurdle, we carried out a study in 48 cocoa agroforests located in three zones in central Cameroon where we conducted an analysis of cocoa yield components and agroforestry system structural characteristics that are likely responsible for observed yield variations. For the first time, we adapted the regional agronomic diagnosis method to demonstrate that the basal area per cocoa tree (mean 61.6 cm 2 ) and the unproductive adult cocoa tree rate (mean 21%) are key factors when assessing the productive performance of the surveyed systems whose average cocoa yield was 737 kg ha −1. From a methodological standpoint, the assessment approach we set up succeeded to overcome the specific obstacles linked with the features of agroforestry systems, especially their complexity (number of species and heterogeneity), by (i) determining relevant indicators and easily measurable variables, (ii) considering the associated tree communities as an environmental component, and (iii) analyzing interactions between cocoa stands and associated tree communities. From an operational standpoint, we showed that farmers can intervene on the structural characteristics of their cocoa agroforests to improve cocoa yields, in particular by eliminating unproductive cocoa trees whose basal area is less than 19 cm 2 to enable the other ones to grow.

No abstract

No abstract