Long-Term Shifts in Grasshopper Management and Outbreak-Associated Control Activity in Inner Mongolian Grasslands: Evidence from 12 Years of County-Level Records

Grasshopper outbreaks threaten forage production and rangeland sustainability, yet long-term evidence linking routinely monitored outbreak conditions to pest-management responses remains limited. We analyzed 12 years (2008–2019) of county-level monitoring and management records from 14 purposively selected units in Inner Mongolia, China, spanning five grasshopper-occurrence subregions. Rangeland outbreaks there are dominated by Oedaleus decorus together with Calliptamus italicus, Gomphocerus sibiricus, Myrmeleotettix palpalis, Bryodema spp., and Chorthippus spp. The records included affected and severely affected grassland areas, biological- and chemical-control areas, and operational inputs comprising personnel, equipment, and pesticide quantities. We evaluated long-term change in control composition and associations linking outbreak pressure to treatment scale and resource mobilization. Biological-control share increased markedly over time, with an average linear change of 6.53 percentage points per year, although the trajectory was nonlinear and the increase was concentrated before approximately 2014. Fixed-effects models showed that a 1-SD increase in outbreak extent was associated with a 5.14-percentage-point increase in total control intensity. The corresponding association with biological-control share was positive but too imprecisely estimated to be informative. Conditional severity was positively associated with all six operational resource indicators in the primary models, while only pesticide input remained supported after adjustment for total control intensity. These results indicate that long-term control composition, annual treatment scale, and operational mobilization represent distinct but related dimensions of grassland grasshopper management. With only 14 county-level clusters, the analysis characterizes this purposively selected monitoring sample rather than the whole of Inner Mongolia, and it does not establish treatment efficacy, causal decision rules, or forecasting performance.