Answers are grounded in this paper and cite their sources.
West African cocoa farmers combine diverse practices for managing the trees associated with their plantations, but whether these practices organize into coherent profiles or vary idiosyncratically remains unknown. We test this question on a national sample of 452 farms covering, for the first time, the three major cocoa-growing « loops » of Côte d'Ivoire (109 East/South-East, 175 Centre/Centre-West, 168 South/South-West), using Multiple Correspondence Analysis ( MCA ) on 15 active management variables, followed by hierarchical agglomerative clustering (Ward's method) on the first two factorial axes (64.0% of Benzécri-corrected inertia). The number of profiles is selected by silhouette score ( k = 3, silhouette = 0.353), confirmed by K -means (Adjusted Rand Index = 0.689) and stable under resampling (mean Adjusted Rand Index = 0.526). Three profiles emerge: a « diversified » profile ( n = 241, 53.3%), combining all three tree origins and the most widespread belief in their usefulness; a « planted-dominant and specialized » profile ( n = 160, 35.4%), almost exclusively of planted origin; and a « passive and poorly-advised » profile ( n = 51, 11.3%), overwhelmingly spontaneous/residual, with technical extension nearly absent (4%) and the only profile with predominantly negative perceived effects. The profiles differ significantly in yield (Kruskal-Wallis, H = 9.23, p = 0.010; passive profile 141 kg/ha/yr versus 238–242 for the other two) and are not independent of the production loop (Chi², p < 0.0001). A joint model of determinants (multinomial logistic regression, cross-checked with random forest) confirms that the production loop ( χ² =26.40, p < 0.0001) and education level ( χ² =11.88, p = 0.018) are the factors most robustly associated with profile membership, ahead of ethnicity and plantation age; cooperative membership is not significant. These results indicate that cocoa agroforestry management in Côte d'Ivoire organizes into recognizable profiles, one of which, a minority, combines a yield deficit with very limited access to technical extension.
No takes yet. Share an insight, caveat, or question.
Adji et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: