Abstract

Landscape transformation due to agriculture affects more than 40% of the planet's land area and is the most important driver of losses of biodiversity and ecosystem services (ES) worldwide. Ecological restoration may significantly reduce these losses, but its effectiveness has not been systematically assessed in agroecosystems at the global level. We quantitatively meta-analyzed the results of 54 studies of how restoration actions reflecting the two contrasting strategies of land sparing and land sharing affect levels of biodiversity and ES in a wide variety of agroecosystems in 20 countries. Restoration increased overall biodiversity of all organism types by an average of 68%. It also increased the supply of many ES, in particular the levels of supporting ES by an average of 42% and levels of regulating ES by an average of 120% relative to levels in the pre-restoration agroecosystem. In fact, restored agroecosystems showed levels of biodiversity and supporting and regulating ES similar to those of reference ecosystems. Recovery levels did not correlate with the time since the last restoration action. Comparison of land sparing and land sharing as restoration strategies showed that while both were associated with similar biodiversity recovery, land sparing led to higher median ES response ratios. Passive and active restoration actions did not differ significantly in the levels of biodiversity or ES recovery. Biodiversity recovery positively correlated with ES recovery. We conclude that ecological restoration of agroecosystems is generally effective and can be recommended as a way to enhance biodiversity and supply of supporting and regulating ES in agricultural landscapes. Whether a land sharing or land sparing strategy is preferable remains an open question, and might be case dependent. Moreover, it is unclear whether crop production on restored land can meet future food production needs. (C) 2015 Elsevier B.V. All rights reserved.