Differentiated selection of agricultural power development modes from the perspective of livelihood resilience: evidence from 18 prefecture-level cities in Henan Province

Under the context of geopolitical conflicts and economic deglobalization, challenges related to food security, the quality of the agricultural labor force, and innovation in agricultural technology have become increasingly prominent. Advancing the construction of an agricultural powerhouse in China therefore requires a transition from extensive promotion to precision-oriented development. Using 18 prefecture-level cities in Henan Province from 2010 to 2023 as a case study, this research develops an agricultural development mode base and a livelihood resilience indicator system. The entropy weight method is applied to measure livelihood resilience. These results are then integrated with the agricultural development mode base, and comparative model evaluation is conducted. Based on model performance, the Random Forest (RF) and Geographically Weighted Random Forest (GWRF) algorithms are selected for analysis. The findings show the following: 1) At the overall level, agricultural development modes have been shifting toward new types mainly represented by smart agriculture and interactive agriculture. 2) From a temporal perspective, collective agriculture dominated the key agricultural development modes before 2015, while smart agriculture became the dominant mode after 2015. 3) From a spatial perspective, the modes display significant regional differences. For example, agrotourism is concentrated in the central, western, and southern parts of Henan, while smart agriculture is primarily distributed in the central and southern regions. 4) From a combined spatiotemporal perspective, the key agricultural development modes across cities in Henan Province are generally dominated by smart agriculture, with collective agriculture serving as a secondary mode.