The effectiveness of cocoa agroforests depends on shade-tree canopy height

Abstract

Agroforestry is often proposed as a ‘climate smart’ strategy for allowing agriculture to both adapt to and mitigate climate change and sustainably increase agricultural production. This is because shade trees in agroforests may buffer growing conditions by creating favorable microclimates (climate-change adaptation), and because shade trees can sequester additional carbon from the atmosphere (climate-change mitigation). However, a major challenge for agroforestry is to maximize these potential benefits while minimizing costs to production as a consequence of resource competition between shade trees and the primary crop. While the effects of shade-tree density and canopy cover on the costs and benefits of agroforests are increasingly well understood, the effects of the traits of shade trees on the effectiveness of agroforests have received less attention. Here, we assess how shade trees with different crown architecture influence production, adaptation, and mitigation goals in a major cocoa growing region in Ghana, West Africa. We quantified the effects of shade trees from nine different species across two classes of height-to-crown-base (low vs. elevated canopies) on yield, microclimate, and carbon storage. We show that shade trees with elevated crowns had large positive effects on carbon storage and neutral effects on yield, while shade trees with low crowns had smaller effects on carbon storage and simultaneously caused larger reductions in incoming light, which was associated with lower yield. Trees of both crown classes were equally effective at buffering sub-canopy temperatures and vapor pressure deficit, although trees with low crowns maintained higher relative humidity. Taken together, our results suggest that shade-tree species with elevated crowns improve the effectiveness of cocoa agroforests by providing maximum benefits for climate-change adaptation and mitigation, while minimizing short-term costs to cocoa production.