Construction of a GeogDetector-based model system to indicate the potential occurrence of grasshoppers in Inner Mongolia steppe habitats

Grasshopper plagues have seriously disturbed grassland ecosystems in Inner Mongolia, China. The accurate prediction of grasshopper infestations and control of grasshopper plagues have become urgent needs. We sampled 234, 342, 335, and 369 plots in Xianghuangqi County of Xilingol League in 2010, 2011, 2012, and 2013, respectively, and measured the density of the most dominant grasshopper species, Oedaleus decorus asiaticus, and the latitude, longitude, and associated relatively stable habitat factors at each plot. We used Excel-GeogDetector software to explore the effects of individual habitat factors and the two-factor interactions on grasshopper density. We estimated the membership of each grasshopper density rank and determined the weights of each habitat category. These results were used to construct a model system evaluating grasshopper habitat suitability. The results showed that our evaluation system was reliable and the fuzzy evaluation scores of grasshopper habitat suitability were good indicators of potential occurrence of grasshoppers. The effects of the two-factor interactions on grasshopper density were greater than the effects of any individual factors. O. d. asiaticus was most likely to be found at elevations of 1300-1400 m, flat terrain or slopes of 4-6 degrees, typical chestnut soil with 70-80% sand content in the top 5 cm of soil, and medium-coverage grassland. The species preferred temperate bunchgrass steppe dominated by Stipa krylovii and Cleistogenes squarrosa. These findings may be used to improve models to predict grasshopper occurrence and to develop management guidelines to control grasshopper plagues by changing habitats.