Identifying exemplary sustainable cropping systems using a positive deviance approach: Wheat-maize double cropping in the North China Plain

CONTEXT: Sustainable cropping systems need to balance productivity and profitability with resource and environmental conservation. Within a population of cropping system observations, there might be positive deviants that outperform others in terms of sustainability, which could serve as “model systems” for future development. Wheat-maize double cropping is the dominant system in the North China Plain, which is facing multiple economic, societal, and environmental sustainability challenges. Identifying exemplary positive deviants out of a multitude of wheat-maize observations might provide solutions to enhance overall sustainability. OBJECTIVES: We aimed to 1) identify exemplary wheat-maize systems that reached optimal performance across seven sustainability indicators, 2) determine which factors regarding management practices and farming contexts resulted in the sustainability gaps between exemplary and other systems, and 3) propose a sustainable wheat-maize prototype. METHODS: Based on a farmer survey dataset (n = 344), we developed a cropping system-level positive deviance approach, including multi-criteria assessment, positive deviant identification (Pareto ranking) and positive deviant clustering, to identify exemplary wheat-maize systems. We then compared exemplary and other systems to quantify the sustainability gaps and identify the key variables explaining sustainability gaps. RESULTS AND CONCLUSIONS: Sixteen percent of wheat-maize cases were Pareto-optimal and were classified as positive deviants. These were sorted into seven clusters representing contrasting sustainability patterns. Among these clusters, one comprised exemplary systems due to the best compromise over the indicator set. Compared to remaining wheat-maize cases, exemplary systems, on average, resulted in 49% and 17% higher gross margin and dietary energy output, respectively, and 33–51% lower labor use, groundwater depletion, N loss, net greenhouse gas emission, and pesticide use. Key practices conferring exemplary system performance included higher maize seeding density, lower fertilizer N input in wheat, partial substitution of inorganic fertilizer with manure, a smaller number of irrigation events, and a lower frequency of pesticide and herbicide application. No significant difference in farming context was found between exemplary and other systems. SIGNIFICANCE: Since the practices of exemplary systems were already locally adopted and proven, we expect that farmers in the region can increase the sustainability of their wheat-maize production by adjusting their management to resemble the exemplary systems. The positive deviance approach thus provides a pragmatic bottom-up approach to identify practices that can improve the sustainability of cropping systems, and can be used for other cropping systems elsewhere.