How far does the tree affect the crop in agroforestry? New spatial analysis methods in a Faidherbia parkland

摘要: Abstract The trees in agroforestry plots create spatial heterogeneity of high interest for adaptation, mitigation, and the provision ecosystem services. But to what distance, exactly, from tree? We tested a novel approach, based upon geostatistics Unmanned Aerial Vehicle (UAV) sensing, infer distance at which single tree affects surrounding under-crop, map yield, litter (i.e. stover) compute crop-partial Land Equivalent Ratio (LERcp) whole-plot level. In an agro-silvo-pastoral parkland semi-arid western Africa dominated by multi-purpose Faidherbia albida, we harvested pearl-millet under-crop scale (ca. 1 ha) also subplot transects, three distances trunks. observed that yield was times higher below crown (135.6 g m−2) than five tree-crown radii trunk (47.7 g m−2). Through geostatistical analysis multi-spectral, centimetric-resolution images obtained UAV overflight entire plot, determined ‘Range’ parameter semi-variogram (assumed be influence on Normalized difference vegetation index (NDVI)) 17 m. correlated (r2 = 0.41; RRMSE = 48 %) production (r2 = 0.46; RRMSE = 35 subplots with NDVI, generated maps scale. measured (0.73 t ha-1) differed one estimated via mapping only 20 %, thereby validating overall approach. similarly 1.05 tC ha-1 yr-1 mapped. Using proxy sole crop, LERcp 1.16, despite low density. This new method handle systems is first application. propose strategies extension landscape