Reducing greenhouse gas (GHG) emissions from crop production is essential for mitigating climate change and achieving sustainable agricultural development. Using micro-level data from the Fixed Observation Rural Survey (FORS) spanning 1993-2020 and the China Rural Development Survey (CRDS), we calculate the agricultural total factor productivity (ATFP) and GHG emission intensity from crop production among farmers based on the Cobb-Douglas production function and life cycle assessment, respectively. A two-way fixed effects model was employed to examine their relationship. The results show that ATFP has a rebound effect on GHG emission intensity under current technological and managerial conditions, with the optimal ATFP level below the sample mean. This suggests that continuous improvements in ATFP may initially reduce but eventually increase emission intensity. These findings remain robust across a series of tests. Moderating effect analysis indicated that higher education levels, per capita income, and labor input significantly strengthened the impact of ATFP on emission intensity. Heterogeneity across nine agricultural zones also indicates consistent rebound effects but with varying turning points. The results suggest that policies should focus on farmer education and training, optimize the agricultural planting structure, and adopt region-specific mitigation strategies to enhance ATFP while avoiding rebound effects in GHG emission intensity.

Urban land use change and industrial land space restructuring are critical drivers of carbon emissions, and understanding their intrinsic mechanisms holds significant theoretical value for territorial spatial governance under carbon neutrality goals. Taking Guangzhou as a case study, this research innovatively integrated POI kernel density analysis and random forest algorithms to construct a high-precision industrial land space identification model (overall classification accuracy: 72%), coupled with support vector regression (SVR), random forest regression (RFR), and logarithmic mean Divisia index (LMDI) decomposition, to systematically reveal the correlation and driving factors between industrial land space evolution and carbon emissions from 2012 to 2022. The findings indicate that: (1) The built-up area of Guangzhou exhibited significant outward expansion, increasing from 1091.95 km² to 1370.87 km², with carbon emissions rising from 67.646 million tons to 84.3112 million tons. (2) The correlation between industrial land space and carbon emissions in Guangzhou presented a "mixed-area dominance—high emissions in industrial, commercial and residential zones—low emissions in public service zones". While the total carbon emissions in industrial zones increased, their intensity declined; both total emissions and intensity rose in commercial and residential zones, whereas public service zones maintained low levels. (3) The RFR model, developed under a "natural-economic-spatial" multidimensional framework integrating remote sensing spectral, socioeconomic, and landscape morphological indicators, achieved optimal fitting accuracy for carbon emissions (R² = 0.86). (4) Carbon emission driving mechanisms showed significant heterogeneity: industrial zones were dominated by industrial activity intensity and land scale, commercial and residential zones were driven by socioeconomic activities and population density, and public service zones exhibited a "binding effect" with commercial activities. This study provides scientific support for optimizing polycentric urban structures, promoting green industrial agglomeration, and formulating targeted carbon reduction policies.

The characteristics and future evolution trends of carbon emissions in the multi-scale social water cycle are of great significance for achieving low-carbon sustainability in the water cycle. Firstly, a "top-down" method for estimating carbon emissions in the social water cycle is proposed based on the water-energy-carbon relationship, taking a macro to meso-micro perspective. Secondly, a "bottom-up" method for predicting carbon emissions in the social water cycle is proposed based on a deep learning-nonlinear dynamic system coupled model, which considers the micro to meso-macro perspective. Finally, the carbon emissions in the social water cycle at different scales in Gansu province from 2009 to 2022 were estimated, and the projected trends of carbon emissions in the social water cycle at different scales in this province from 2023 to 2035 were predicted and analyzed. It is found that: (1) From 2009 to 2022, the carbon emissions in the social water cycle in Gansu demonstrated a pattern of stable growth followed by fluctuating decline. Most prefecture-level cities exhibit a trend of fluctuating decrease or a tendency towards a stable state in social cycle carbon emissions, and only Jiuquan and Jiayuguan cities show fluctuating growth. The county level social water cycle carbon emissions exhibited a spatial distribution pattern of "low at both ends and high in the middle". (2) The deep learning-nonlinear dynamic system coupled model demonstrates robust and universal predictive capabilities with respect to carbon emissions in the social water cycle, outperforming traditional dynamic system models and grey prediction models in both temporal and spatial domains. (3) Driven by population outflows and the optimization of water use structures, carbon emissions from the social water cycle in Gansu peaked at an early stage, reaching their maximum level in 2012. Furthermore, the peak timing for most prefecture-level cities was predominantly concentrated within the period of 2011-2013.

An uneven match between inter-city economic trade benefits and carbon emission costs creates inequitable shifts in carbon governance responsibilities. Existing carbon responsibility allocation approaches inadequately address carbon inequity patterns from cross-city trade and carbon transfer. This paper takes the Yangtze River Delta Urban Agglomeration (YRDUA) as an example. Municipal input-output tables and direct carbon emissions data are analyzed using environmental input-output modeling, Tapio decoupling analysis, random forests, and Shapley values. This quantifies income disparities stemming from carbon inequity relationships. Urban interest characteristics are profiled through a multidimensional "social-economic-ecological-equity-efficiency-sustainability" indicator system. A cooperative game-theoretic model achieves carbon quota allocation across the YRDUA. Key results indicate that: (1) Shanghai functions as the core hub for carbon-intensive product flows. Jiangsu and Anhui exhibit substantial internal carbon transfers, with Jiangsu demonstrating stronger external economic linkages than Anhui. Zhejiang maintains closer carbon-economic ties with Shanghai despite lower internal transfers. (2) High-income cities concentrate around provincial capitals, while low-income cities occupy peripheral regions. Shapley value analysis reveals Jiangsu-Zhejiang-Shanghai coalitions securing disproportionate carbon governance benefits. Unregulated competition biases quota allocation toward dominant cities, amplifying regional disparities (the "Matthew effect"). (3) The equity-based carbon quota system balances emission reduction pressures between high- and low-income cities. Complementary cooperation generates enhanced developmental benefits. This approach imposes effective constraints on carbon-intensive cities and advances regional progress toward "dual carbon" goals.

With the growing impacts of human activities and global changes on biodiversity, the concept of biodiversity footprint has rapidly emerged as a key tool for evaluating the processes of biodiversity loss and their influencing factors. However, research across different scales often overlaps, and assessment paradigms remain incompatible, making a systematic review and standardization essential endeavors. This paper distinguishes and compares various eco-environmental footprint concepts, systematically reviews the evolution of the biodiversity footprint concept, identifies the driving factors and key assessment indicators of biodiversity loss, and analyzes and compares the research perspectives, indicator selection, and evaluation methods of biodiversity footprint across three spatial scales (global, national, and local) and their corresponding actors. From a global perspective, biodiversity footprint research primarily uses tools such as country-level species-area relationship models, input-output models, and life cycle assessment (LCA) to analyze biodiversity loss and the transnational transfer of impacts caused by global commodity production-consumption networks. It also employs shared socioeconomic pathways (SSPs) to predict biodiversity loss under different socioeconomic development scenarios. From a national perspective, research mainly adopts input-output models and land-use change analysis to identify leading industries and economic sectors that threaten biodiversity through domestic construction and production activities. At the local scale, biodiversity footprint research often integrates LCA with ecological methods to assess the localized and cross-regional biodiversity impacts of actors such as enterprises, farmers, individuals, and institutions. This includes impacts from investment allocation, biological resources inputs, production disturbances, and pollution emissions. Finally, based on a literature review, this paper suggests that future research should focus on three interconnected areas: (1) developing a comprehensive and standardized indicator system; (2) constructing integrated theoretical models that enable multi-scale integration and dynamic simulations; and (3) translating research findings into policy guidance to advance the theoretical refinement and practical application of biodiversity footprint assessments.

Protected area policies are pivotal not only for conserving and sustainably managing natural resources but also for balancing human-environment interactions to foster sustainable development. Based on the perspective of the geography of public policy, this study develops a three-dimensional analytical framework encompassing policy structure, policy instruments, and policy effectiveness. Employing methods such as social network analysis, content analysis, and the PMC index model, the paper examines the evolutionary patterns and underlying logic behind the transformation of China's protected area policies from 1982 to 2021. The results indicate that: (1) The evolution of these policies can be categorized into three distinct stages (initial exploration from 1982 to 1993, system development from 1994 to 2012, and system refinement and deepening from 2013 to 2021). This evolution illustrates a paradigm shift from "quantitative expansion" to "high-quality development". As the stages have progressed, policy content has gradually become more refined, effectively supporting the standardized development of protected areas. (2) Policy types are predominantly normative texts such as notices and guidelines. Furthermore, the issuance of policies on protected areas is primarily carried out independently by various departments, with the majority of policies concentrated within three key agencies: the Ministry of Ecology and Environment, the National Forestry and Grassland Administration, and the Ministry of Housing and Urban-Rural Development. The level of cooperation among the policy-issuing departments is relatively low, and there is a clear need to enhance interdepartmental collaboration. (3) The policy instruments encompass three major categories, including supply-side, environment-shaping, and demand-side. However, an imbalance exists in the structure of these instruments, with environment-shaping instruments maintaining long-term dominance. At the same time, the comprehensiveness of internal combinations among various policy instruments has continued to strengthen. The joint application of these tools has evolved through distinct phases, progressing from isolated linkages to structural optimization and eventually to the formation of a networked collaborative framework. (4) Protected area policies generally demonstrate high effectiveness, reflecting sound scientific grounding and systematic design. Among them, comprehensive planning-type and guidance-type policies typically achieve higher PMC Index scores than measure-type policies, which tend to focus on specific aspects. In comparison, measure-type policies still have room for improvement in terms of content, functional design, and incentive mechanisms, which is essential for fully enhancing their overall policy effectiveness.

Taking 107 prefecture-level and above cities in the Yangtze River Economic Belt as research samples, this study empirically examines the coordination and interactive effects of their ecological resilience and eco-efficiency during the period of 2009-2023 by using the coupled coordination model, the spatial autocorrelation model, and the panel vector autoregression model. The results show that: (1) between 2009 and 2023, the urban ecological resilience in the study area had been increasing year by year, from 0.1242 in 2009 to 0.1709 in 2023, with an average annual growth rate of 2.31%, and showed a spatial pattern of "high in the south and low in the north". The urban ecological efficiency demostrated a steady upward trend, rising from 0.1960 in 2009 to 0.5475 in 2023, with an average annual growth rate of 7.61%, and a spatial pattern of "downstream region > midstream region > upstream region". (2) The coordination degree between urban ecological resilience and eco-efficiency increased with fluctuation from 0.3818 in 2009 to 0.5329 in 2023, averaging an annual growth rate of 2.41%; and the spatial pattern is characterized by "downstream region > midstream region > upstream region". (3) There is a spatial positive correlation between urban ecological resilience and eco-efficiency coordination degrees, but the intensity of agglomeration presents a decaying trend, which is manifested by the sprouting of emerging agglomerations in the upstream and the contraction of the scope of agglomeration in the downstream, especially in Anhui, where the low-low agglomeration area is drastically shrinking. (4) A dynamic interactive effect exists between urban ecological resilience and eco-efficiency. While the interaction exhibits fluctuations, the net effect is positive. Furthermore, the inertia of each system exerts a stronger influence on its own development than their cross-impact.