Abstract
Highlights: • Crops yield and leftovers were scaled up to the field-scale. • Using spectral images throughout the crop cycle improves productivity predictions. • The distance of influence of trees on crops can be measured by proxydetection and reveals the Land Equivalent Ratio.Abstract: Faidherbia albida is known to affect the yield of various crops, typically in a pattern where the impact decreases with increasing distance from the tree. While several studies have investigated the spatial extent of this effect, limited research has explored how this distance varies across different crops or its relationship with crop yield and the Land Equivalent Ratio. In this study, we used a geostatistical approach combined with multispectral UAV (Unmanned Aerial Vehicle) imagery to address these gaps in understanding. The results showed that, in contrast to its tripling effect on millet yield, F. albida does not have a significant impact on groundnut pod yield, it only improves its haulm yield under its crown by about 50 %. The geostatistical analysis showed that F. albida affects the groundnut crop up to 9.8-m, compared to 18-m for millet. Yield upscaling from subplots to the whole plot was achieved with an error of 8 % for groundnut pod yield and 13 % for haulm yield. Groundnut’s partial Land Equivalent Ratio (LERcp) was 1.02 for pod yield and 1.05 for haulm yield, which was lower than the LERcp for millet. We concluded that the distance at which agroforestry trees influence crops is a reliable predictor of their effect on yield and Land Equivalent Ratio. This approach offers a promising tool for future agroforestry studies, potentially guiding crop management strategies in agroforestry systems.