Under the dual pressures of increasing fragmentation of farmland and profound changes in the agricultural labor force, the problem of abandoned farmland in rural China has become increasingly prominent, posing a serious threat to food security. Previous studies have mainly focused on the impact of socioeconomic factors on farmland abandonment behavior; however, there remains a lack of systematic investigations into the underlying mechanisms through which farmland fragmentation influences farmers’ land-use decisions. Therefore, this study applies behavioral theory to construct a “fragmentation–mechanization–abandonment” analytical framework to analyze the potential mechanisms of land-use decision-making. Using micro-level household survey data from multiple provinces across China, we conduct empirical tests through Probit models and Karlson Holm and Breen (KHB) mediation effect analysis. Our findings include: (1) Farmland fragmentation significantly increases the probability of farmland abandonment, which remains robust across multiple tests. (2) Farmland fragmentation indirectly influences abandonment behavior through the mediating variable of agricultural mechanization level, but this mechanism plays only a partial mediating role. (3) Heterogeneity tests indicate that the positive effect of farmland fragmentation is significantly stronger among households with low access to finance than among those with high access to finance. This influence is also stronger among households with high farmland accessibility. This study indicates that abandonment governance requires moving beyond a singular focus on land consolidation and shifting toward a synergistic approach that combines differentiated regional strategies with targeted support for smallholders. From a micro-mechanism perspective, this study provides empirical evidence for targeted implementation of land consolidation and livelihood support for smallholders, offering important policy implications for achieving agricultural modernization.

As China's urbanization accelerates, the ongoing expansion of cities has exacerbated tight land resource constraints and the reduction of ecological space. Under these conditions, conventional extensive modes of development are unable to simultaneously accommodate economic growth, resource efficiency, and environmental carrying capacity, thus impeding the further development of urban sustainable productivity (USP). Drawing on resource scarcity theory and spatial equilibrium theory, this paper develops a theoretical analytical framework of “urban land use conflict (ULUC)–green innovation (GI)–USP”. Taking the urban agglomeration in the Yangtze River Basin (YRB) as a case study, this paper employs fixed effects models and mediation effect tests, and other empirical techniques, to investigate the mechanisms through which ULUC influences USP. These results show that: (1) ULUC exerts a significant positive impact on USP, and this effect exhibits marked variation across different sections of the YRB. (2) GI plays a mediating role in the relationship between ULUC and USP, while fiscal environmental expenditure (FEE) and green finance support (GFS) enhance this influence through moderating mechanisms. (3) Industrial structure upgrading displays a clear threshold effect, whereby the impact of ULUC on USP is significantly reinforced only when the urban industrial structure has reached a certain level of development. Against the backdrop of increasingly binding land resource constraints, cities are suggested to enhance their GI capacity, increase investment in environmental policy implementation, and promote industrial structure upgrading so as to advance USP.

Pollution control and carbon mitigation often involve the same energy inputs and production processes, yet the role of finance in shaping their combined urban outcomes remains insufficiently understood. Using panel data for 110 Chinese cities in the Yangtze River Economic Belt (YREB) from 2010 to 2023, this study examines the relationship between green finance (GF) and joint pollution and carbon reduction performance (JPCRP), while JPCRP is measured using an outcome-based composite index covering carbon emissions, energy use, and conventional pollution. The empirical analysis combines city and year fixed effects with two stage least squares, indirect path, interaction, and panel threshold models. The results show that GF is positively associated with JPCRP, while exploratory subgroup estimates vary in magnitude across basin locations and development levels. Specifically, the fiscal expenditure analysis identifies a countervailing pathway in which GF is positively associated with fiscal environmental expenditure (FEE), while FEE is negatively associated with JPCRP after controlling for GF and the other covariates. Green technological innovation (GTI) conditions the relationship between GF and JPCRP, with the negative GF × GTI interaction indicating that the marginal association between GF and JPCRP becomes weaker as GTI increases. The estimated threshold for urban spatial development efficiency (USDE) is 0.66, above which the GF coefficient remains positive but declines from 0.1412 to 0.0845. These findings indicate that the environmental contribution of GF depends not only on the supply of green capital but also on fiscal allocation, technological capacity, and urban spatial conditions.