Economic and not ecological variables shape the sparing-sharing trade-off in a mixed cropping landscape

1. The framework of land sparing vs. land sharing provides a useful analytical tool to address the crop-production/biodiversity trade-off. Despite multiple case-studies testing the sparing-sharing trade-off, this framework still lacks the ability to identify the conditions in which sparing, or sharing, would be the preferred strategy for pareto-optimizing both food-production and biodiversity. Under some conditions, ecosystem services may create a positive feedback between biodiversity and crop production, affecting the optimization. 2. This study aims to identify the conditions and the relevant variables that determine the preferred land-use strategy in terms of maximizing both biodiversity and food production, while accounting for positive feedback of ecosystem services in this analysis. We used a simulation model with data from a mixed cropping landscape (100 km2) covering seven crop types, five taxonomic groups, three biodiversity metrics and 23 bioindicators to explore the variables shaping the biodiversity-production trade-off and ecosystem services underlying it. We explored a continuum of sparing large semi-natural patches to sharing by maintaining uncultivated field margins of varying size. 3. Land sparing outperformed land sharing in 62% of the scenarios and it was economically more predictable. The optimization was shaped by costs, associated with crop type, rather than by landscape composition and configuration, biodiversity metric, taxonomic group or bioindicator. 4. Landscape configuration and taxonomic group results corroborate the notion that land sharing benefits mainly small organisms, and that the common width of field-margins in many agri-environmental policies (10 m) is not cost-effective compared to land sparing. 5. Land sharing was the optimal strategy whenever it resulted in minimal costs, despite contributing little to biodiversity. Yet, when field margins were >20 m wide (small-scale sparing), land sharing maintained higher biodiversity and was at least as cost-effective as sparing. 6. Synthesis and applications. Our model highlights the importance of socio-economic variables compared to ecological variables in selecting land-management strategy to pareto-optimize both food production and biodiversity. Considering opportunity costs alongside economic benefits from ecosystem services in various cropping systems may therefore improve the cost-effectiveness of biodiversity conservation policies in agricultural landscapes.This file contains the rawdata analyzed in the statistical models described in the methods section of the paper (excluding personal ownership data): sites - sampling sites metadata including the type of habitat, coordinates, amount of natural habitats in the landscape, slope, altitude. plants...plant assoc. - richness and geometric mean abundance data per site (seasonal total or per visit), used to run the ecological statistical models revenue - revenue, costs and amount of field margins mapped around the plots used to run the economic models. Data about ownership was excluded