Temporal evolution of maize yield spatial heterogeneity in northeast China: shift of dominant factors from human management to climate change

Northeast China (NEC) is a major agricultural region in China but exhibits large spatial heterogeneity in crop yield. To date, it remains unclear how multiple climate and management factors have affected the spatial heterogeneity of crop yields and whether the dominant factors have changed over time. Here, we synthesize multi-source datasets and develop geographic detector and regression models to evaluate the effects of 19 climate and management variables on the spatial heterogeneity of maize yield in NEC during 1984-2013. We found that the spatial heterogeneity of maize yield (measured by the coefficient of variation) has exhibited an upward tendency of 0.74 %center dot y-1 during 1984-2013, with a significant change point detected around the year 2000. Specifically, an increasing trend of 0.47 %center dot y-1 is observed for 1984-2000, while a downward trend of-0.68 %center dot y-1 is revealed for 2001-2013. At the annual scale, fertilization and low temperature are the dominant factors affecting the spatial heterogeneity of maize yield in 1984-2000, whereas precipitation-related factors play leading roles in 2001-2013. At the decadal scale, fertilization and low-temperature control the upward trend in the spatial heterogeneity of maize yield during 1984-2000, while wind is the dominant factor in the downward trend from 2001 to 2013. Our findings highlight that the dominant factors controlling the spatial heterogeneity of maize yield have temporally shifted from management to climate factors, which underscore the importance of prioritizing resources to address rising climate risks by improving irrigation, drainage systems and crop varieties.