Competition effects in cocoa (Theobroma cacao L.) hybrid trials

A method developed in coffee tree breeding to assess family competitioneffects (partner effects) in comparative variety trials was applied to thecocoa tree (Theobroma cacao L.). The study was conducted in ahybrid comparative trial planted in French Guiana, involving twelve familiesof 50 trees in a totally randomized single-tree plot design, at a density of1,667 trees per hectare. The trial was thinned at 10 years, at a rate of twoout of four rows. Competition was studied with reference to juvenile andadult vegetative vigour, and to periodic and cumulative yields (number ofpods, potential weight, average weight of one pod, and the production:vigour ratio). At the end of the trial, after thirteen years of monitoring,competition effects were revealed which explained 8 to 10% of theresidual variance after removal of the hybrid and micro-environment effects.These effects, which were detected as early as 18 months, occurred earlierthan generally acknowledged. Under the trial conditions, the families couldbe classed as aggressive, stimulating or passive for their neighbours.Vegetative vigour (trunk cross-section) explained 34% of the competitioneffects (partner effects), which, with hindsight, vindicated the use of theproduction : vigour ratio as the main selection criterion in cocoa breeding.The partner effects noted on the production variables were never explainedby any production variable, hence non-aggressive high-yielding families canbe selected.