Global Bias-Aware Synthesis of Meta-Analyses Reveals Agroforestry's Potential for Improving Soil Health.

Agroforestry systems (AFS), which integrate trees into agricultural landscapes, offer a promising strategy for climate change mitigation and biodiversity restoration. The trees can access additional resources, improving soil functions and contributing to soil regeneration through increased organic matter inputs. However, evidence is scattered and thus hindering unified global quantification of impacts. Here, we present the first bias-adjusted global quantitative synthesis of agroforestry effects on distinct soil and water attributes, drawing on 1590 primary studies summarized in 26 meta-analyses. We evaluated each meta-analysis against 16 standardized quality criteria. Less than half of the meta-analyses met > 75% of these criteria, while most did not weigh effect-sizes or assess heterogeneity and publication bias. We quantified primary studies overlap, finding an overall redundancy of 18% across all soil parameters, most of which arise from studies reporting on soil carbon. Accounting for meta-analysis quality and bias, AFS tend to increase soil organic carbon by 20%, improve chemical soil quality by 59%, physical soil quality by 22%, and enhance water regulation by 71%. Biological soil quality (86%), nutrient leaching reduction (67%), and erosion control (6%) also benefit from AFS, though with higher heterogeneity. Evidence is extensive for soil organic carbon and chemical properties (73% of studies), whereas it remains limited for AFS effects on biological (< 12%), physical (< 19%), and water-regulation traits (< 27%). 58% of meta-analyses report no details at all on system characteristics, such as stand age, species diversity, or tree density, limiting the identification of AFS management practices that maximize soil health attributes and highlight the need for standardized definitions of agroforestry and consistent sampling protocols. By systematically evaluating meta-analysis quality and study overlap, this synthesis provides a robust framework to distinguish reliable from uncertain outcomes and demonstrates the capacity of agroforestry to deliver multifunctional benefits, particularly regarding soil health.