Patterns of shade plant diversity in four agroforestry systems across Central America: a meta-analysis.

Agroforestry systems can potentially increase tree diversity within agricultural landscapes, but to date, there is little understanding of the patterns of shade plant diversity within different agroforestry systems (AFS) at large spatial scales. Using compiled plant inventory data (from 23 sources, 2517 plots, and 148,255 individuals) encompassing four AFS (shaded coffee; shaded cocoa; dispersed trees on pastures; and live fences) across six countries in Central America we estimated different metrics of diversity to assess the conservation value of different AFS for shade plants. 458 shade plant species were recorded across the four agroforestry systems. Primary forest species accounted for 28% of the shade species recorded, but only 6% of the recorded individuals. No single AFS was consistently the most diverse across countries when considering rarefied species richness. Trees on pastures can potentially reach a similar species richness as cocoa and coffee systems but require sampled areas 7-30 times larger. In terms of composition, 29 species were shared across the agroforestry systems in different countries, illustrating the strong selection pressure of farmers for species that provide timber, firewood, and fruit. Our study highlights the potential contribution and limitations of different AFS for tree diversity conservation within agricultural landscapes.

Shade tree spatial structure and pod production explain frosty pod rot intensity in cacao agroforests, Costa Rica.

Vegetation composition and plant spatial structure affect disease intensity through resource and microclimatic variation effects. The aim of this study was to evaluate the independent effect and relative importance of host composition and plant spatial structure variables in explaining disease intensity at the plot scale. For that purpose, frosty pod rot intensity, a disease caused by Moniliophthora roreri on cacao pods, was monitored in 36 cacao agroforests in Costa Rica in order to assess the vegetation composition and spatial structure variables conducive to the disease. Hierarchical partitioning was used to identify the most causal factors. Firstly, pod production, cacao tree density and shade tree spatial structure had significant independent effects on disease intensity. In our case study, the amount of susceptible tissue was the most relevant host composition variable for explaining disease intensity by resource dilution. Indeed, cacao tree density probably affected disease intensity more by the creation of self-shading rather than by host dilution. Lastly, only regularly distributed forest trees, and not aggregated or randomly distributed forest trees, reduced disease intensity in comparison to plots with a low forest tree density. A regular spatial structure is probably crucial to the creation of moderate and uniform shade as recommended for frosty pod rot management. As pod production is an important service expected from these agroforests, shade tree spatial structure may be a lever for integrated management of frosty pod rot in cacao agroforests.