Abstract
Decades of evidence suggests that the relationship between landscape diversity and ecosystem productivity is positive, nonlinear, and saturating—with remarkable consistency across organisms and conditions. Despite this consensus, evidence of the impact of landscape diversity on agricultural productivity remains mixed. We use mixture modeling to systematically examine consistency in the relationship between landscape diversity and crop yields in the United States across heterogeneous model specifications for different major crops and approaches to operationalizing landscape diversity. While the diversity‐productivity relationship varies across crops and landscape diversity metrics, we find a consistent (87%) positive association between landscape diversity and yields. Although not significant at a 95% credibility level, across all conditions examined highly diverse landscapes are associated with yields that are, on average, 2.7% higher than yields in highly simplified landscapes. Moreover, this relationship is not saturating, suggesting that the upper limits of positive landscape diversity effects have not been reached within contemporary U.S. cropping systems.