Driving Mechanisms of Land Degradation Neutrality at the County Level and Zonal Strategies in the Yellow River's Key Ecological Region (Part II : Post‐Assessment Driver Analysis)

Abstract

Achieving land degradation neutrality (LDN) is critical for sustainable socio‐ecological development; however, its practical implementation is often hindered by scale mismatches between assessments and administrative units, coupled with an over‐reliance on static drivers. Focusing on the Key Ecological Region of the Yellow River (KER), this study addresses these challenges by establishing a county‐scale net land improvement rate ( NLIR ) and developing a dynamic–static integrated indicator system aligned with county governance. By combining spatial analysis with Random Forest models, we quantified dominant drivers, spatial heterogeneity, and nonlinear threshold responses of NLIR . Results indicate: (1) From 2000 to 2020, land degradation pressure in the KER remained relatively low, with 79.16% of counties exhibiting NLIR above 20%, although a decline in restoration capacity occurred in human‐intensive regions after 2010. (2) Spatially, counties with low NLIR values exhibited significant clustering in developed regions subjected to intense anthropogenic pressure, whereas high NLIR values were predominantly distributed across climatically favorable regions and areas where ecological conservation is prioritized. (3) The intensity of construction land change emerged as the dominant driver of NLIR , followed by the intensity of cultivated land change and per capita cultivated land area. (4) All driving factors exhibited pronounced nonlinear threshold effects, within which their marginal positive or negative impacts on NLIR were strengthened relative to other value ranges, indicating critical governance tipping points for land management. Based on these quantified driving mechanisms, four differentiated zoning categories are proposed to support territorial space optimization and high‐quality development in the KER.