Evidence of positive relationships between biodiversity and ecosystem function (BEF) have been reported in numerous experimental studies carried on short spatial and temporal scales. However, it remains uncertain how the BEF relationship varies over larger scales especially given the impact of humans on landscapes through land-use changes and fragmentation. Indeed humans modify functions through nutrient and pollutant deposition, favor specific local species assemblages and alter landscape connectivity, hence changing biodiversity distribution and turnover. Here, we employ a simple theoretical approach to make general predictions about the effect of landscape fragmentation on changes to the BEF relationship across spatial scales. In particular our objective is to determine how the size distribution and configuration of remaining habitat patches in fragmented landscapes influence BEF measured at increasing spatial scales of observation. We generated spatially distributed communities by using point process models with varying degrees of spatial aggregation among individuals. Function was measured for each individual as a function of surrounding biomass and diversity, and strength of complementarity. We then created different fragmentation patterns where fragments had either constant areas or followed a normal, power-law or uniform distribution of sizes.
Results/Conclusions
BEF relationships followed power-laws with exponents varying with spatial scales of observation. The variations in the exponent differed according with the degree of spatial aggregation among individuals in un-fragmented landscapes and show discontinuities for landscape fragmentation patterns presenting a spatial characteristic scale. We discuss the potential implications of these results for management systems, such as agriculture and silviculture, which are characterized by scale-specific landscape transformations.