[Objective] This study addresses regional disparities and imbalanced drivers in developing smart agriculture, a core approach to fostering new quality agricultural productivity. It aims to: (1) construct a comprehensive evaluation system incorporating geographical and industrial heterogeneity; (2) empirically analyze the synergistic drive between data elements and agricultural physical capital; and (3) reveal their role in inter-provincial spatial linkages. The significance lies in its potential to inform strategic planning and policy-making, thereby contributing to the sustainable transformation of agriculture and balanced regional development within the context of rural revitalization. [Methods] A quantitative spatial econometric approach was employed using panel data from 30 Chinese provinces spanning from 2015 to 2023. The research was executed in three key stages. First, a comprehensive provincial-level smart agriculture development index was constructed. This index integrated multiple dimensions and was weighted by combining the Analytic Hierarchy Process and the Entropy method, with adjustments made for terrain and leading industry heterogeneity. Second, a series of econometric models were specified. Baseline fixed-effects and generalized method of moments models were used to examine the driving role of data elements, while mediating and moderating effect models were employed to systematically verify the synergistic mechanism between data and physical capital and its pathways. Third, spatial autocorrelation tests and spatial Durbin models were employed with three spatial weight matrices—geographical contiguity, agricultural resource zoning, and agricultural economic structure similarity—to identify spatial correlation characteristics and spillover patterns. Direct and indirect effects were decomposed to precisely quantify local impacts and spatial spillovers. [Results and Discussions] The analysis yielded four clusters of key findings that confirmed and refined the proposed core hypotheses. 1) Gradient development and regional mismatch: Smart agriculture development exhibited a pronounced "ladder-like" spatial pattern. The Huang-Huai-Hai region remained the persistent leader in absolute development level. The southwestern region also maintained a relatively high level, which appeared partly "forced" by its challenging terrain, necessitated efficiency-seeking technology. In contrast, the Northeast and Northwest regions lagged. A critical systemic mismatch was revealed: Regions with the highest growth momentum were not necessarily those with the highest current smartization levels, indicating divergent developmental pathways. 2) The core synergistic mechanism: A significant positive interaction was found between data inputs and agricultural physical capital. Crucially, the independent coefficients of each were often found to be insignificant or even negative in spatial models, which underscored that limited or even negative marginal returns were yielded by isolated, uncoordinated investment in either domain. Significant systemic gains were unlocked precisely by their structural complementarity. Furthermore, this synergy was found to operate partially through the channel of technological capital accumulation. The mediating effect of technological capital was confirmed to be significant, and its own impact on smart agriculture output was exhibited as a nonlinear threshold characteristic. This confirmed that a critical mass of technological capital had to be accumulated before its benefits could be fully realized. 3) Competition-dominated spatial interactions: The spatial analysis revealed that inter-provincial dynamics were characterized primarily by competition rather than cooperation. A significant negative spatial spillover was detected specifically under the economic structure similarity matrix. This indicated that resources, talent, and investment were competed for by provinces with similar agricultural economic profiles, potentially hindering each other's growth. However, a nuanced finding was observed: while raw data or capital might be siphoned away, significant positive spatial spillovers were generated by successful provincial models of "data-capital" synergy. This suggested that best practices and development paradigms could be diffused, offering a pathway to transcend pure competition. 4) The effectiveness of the core drivers was found to be highly context-dependent. From a zoning perspective, the Huang-Huai-Hai region was characterized by digital-drive but was found to lack deep synergy; the Northeast was constrained by traditional path dependency; the Southwest was shown to exhibit a paradox of high knowledge spillover but low local application; and the northwest was characterized by singular, weak drivers. From a developmental stage perspective, the role of RD investment was assessed as stable, while the payoff from digital infrastructure was seen to be contingent on an "efficiency threshold", and its spatial spillover effect was observed to diminish as regional total factor productivity increased. [Conclusions] It is demonstrated that development is not merely a function of increased inputs, but is critically determined by the structural coupling of data and physical capital. This coupling is facilitated by technological capital, which acts as a nonlinear mediator. At a practical level, a one-size-fits-all approach to investment policy is argued against. For leading regions such as the Huang-Huai-Hai, policy focus should be placed on deepening existing synergies. For regions like the northeast, breaking path dependence is seen to require policies that forcefully couple new digital tools with legacy physical assets. The pervasive "siphon effect" is identified as necessitating national-level coordination mechanisms among structurally similar provinces to mitigate destructive competition. Ultimately, the promotion of smart agriculture is concluded to require spatially differentiated policies that strategically foster local factor synergy while managing the competitive externalities inherent in regional linkages.

[Significance] The adoption of smart agricultural technologies has significantly accelerated the development of modern agriculture, emerging as a pivotal driver for green transition within the agricultural sector. By enabling a fundamental shift in agricultural production paradigms, these innovations not only enhance desirable agricultural outputs but also effectively reduce undesirable environmental outputs. This dual effect substantially contributes to the improvement of agricultural green total factor productivity (AGTFP). [Progress] A systematic investigation was undertaken into the pivotal role of smart agriculture technology innovation in driving the green transition within the agricultural sector. The analysis was meticulously structured across three sequential stages: Firstly, it elucidated the intrinsic mechanisms through which technological advancements serve as enablers for this green transition, detailing how innovations in agriculture technology translate into environmental sustainability gains. Secondly, it identified and categorized the existing implementation barriers that hinder the widespread adoption of smart agriculture solutions, providing insights into the challenges faced by stakeholders at various levels. The advancement of smart agricultural technologies facilitated a paradigm shift, moving from isolated, fragmented data systems to cohesive, interconnected ecological systems that support sustainable growth. The core technological logic underpinning this transition operates through a synergistic "data-algorithm-equipment-ecology" evolutionary trajectory: (1) Data infrastructure construction: Establishing comprehensive, multi-source heterogeneous agricultural data acquisition systems that collect, process, and analyze data from diverse sources, forming the foundational data assets necessary for the development and deployment of smart agriculture solutions. (2) Algorithmic decision-making: Developing sophisticated AI-driven resource allocation models that leverage big data analytics to enable precision management, optimizing resource use, and enhancing productivity through intelligent decision-making systems. (3) Equipment modernization: Creating technical closed loops characterized by a virtuous cycle of "equipment upgrading-efficiency enhancement-cost reduction-environmental benefits" through the implementation of automated production management and intelligent harvesting systems that minimize human intervention and maximize output. (4) Ecological optimization: Building circular systems for waste valorization, carbon footprint tracking, and ecological monitoring that facilitate the recycling of agricultural by-products, reduce greenhouse gas emissions, and promote the restoration and preservation of agricultural ecosystems, ultimately achieving self-regulating, resilient agricultural environments. However, the implementation process faces three categories of barriers: (1) Technological barriers: High infrastructure costs, inefficient technology commercialization, insufficient digital literacy among farmers, and inadequate full-process automation/intelligence levels constrain transformation depth. (2) Market barriers: Significant mismatches between supply-side capabilities and demand-side requirements in technology applicability. (3) Institutional barriers: Firstly, the lack of policy coordination is evident in the "prioritizing construction over operation" approach, leading to many agricultural big data platforms failing to effectively implement green technologies post-completion due to inadequate policies supporting sustained operations and practical applications. Secondly, structural deficiencies in the standard system, particularly the absence of unified classification standards for agricultural information resources, create technical barriers to cross-departmental and cross-regional data sharing. [Conclusions and Prospects] To effectively advance smart agriculture technology in driving the agricultural green transition and systematically address identified barriers, the following policy recommendations are proposed: (1) To effectively overcome the persistent "last mile" barrier that impedes the widespread adoption of technological innovations among smallholder farmers, it is imperative to actively promote multi-stakeholder participation across the agricultural sector. This involves fostering collaboration among governments, private enterprises, non-profit organizations, and local communities. Additionally, establishing a comprehensive technical service network is crucial, one that seamlessly integrates government guidance to ensure policy alignment, market operations to drive efficiency and sustainability, and social engagement to enhance community buy-in and knowledge dissemination. (2) To significantly boost the efficiency of green resource allocation within the agricultural sector, it is imperative to harness the power of market-oriented reforms. These reforms should serve as the cornerstone for forging a novel supply-demand-driven model that is both dynamic and responsive. By integrating intelligent algorithms and advanced data analytics, this coordinated approach will facilitate the precise matching of essential resources such as land, capital, and cutting-edge green technologies. Such precision will enable a paradigm shift from traditional, scale-oriented production methods to more flexible, demand-driven customization tailored to specific market needs. Ultimately, this transformation will cultivate sustainable endogenous growth momentum, providing a robust foundation for the agricultural sector's green transition and ensuring its long-term viability and competitiveness. (3) Governments ought to take the lead in driving systemic and comprehensive reform of existing policy formulation frameworks, ensuring that agricultural development strategies are aligned with long-term sustainability goals. Additionally, they should strengthen technological support systems for agricultural resource zoning, leveraging advanced tools such as big data analytics and geographic information systems to optimize land use and resource allocation. By doing so, governments can foster a new universal smart agriculture development paradigm that integrates digital innovation with ecological principles. These targeted measures will not only enhance agricultural productivity but also accumulate valuable experience for comprehensively advancing the agricultural green transition and promoting rural ecological revitalization in a systematic manner.

[Objective] In recent years, China has issued a series of policy documents, which emphasize digital transformation as a key driver for deepening the integration of digital and intelligent technologies (DITs) with the agricultural and rural economy, promoting high-quality agricultural development, and advancing the goal of building a strong agricultural sector. However, the ultimate impact of these technologies depends not only on their potential effectiveness but also on the extent to which they are adopted by farmers. Against this backdrop, the aim of this research is to assess the current adoption of agricultural DITs among agricultural producers in China and to identify the main challenges facing their application in the future development of agriculture and rural areas. [Methods] Using two representative household survey datasets—the China Agriculture and Rural Development Survey and a 2023 survey of farmers in major grain-producing regions—this study employs descriptive statistics to document the adoption of various DITs. Then they were combined with a review of the literature to examine the potential impacts and underlying mechanisms of technology adoption, and to analyze the challenges and their causes in the diffusion of these technologies by linking theory with practice. [Results and Discussions] The results showed that, for non-embedded digital technologies, smartphone use among primary agricultural decision-makers was widespread, with an average penetration rate of 76%. The most frequently used applications were WeChat, TikTok, and Kuaishou. Despite the high usage of WeChat, its application in agricultural production and marketing remains limited: about two-thirds of respondents report rarely or never following agriculture-related official accounts. Similarly, access to agricultural policy information via smartphones was limited, with approximately 78% of respondents indicating that they seldom or only occasionally followed such information. In terms of e-commerce, about 52% of rural households engaged in online shopping in 2023, but very few used online channels for sales: only 0.32% sold agricultural products and 0.15% sold other goods online. In digital finance, only 27% of surveyed households were aware that loans could be applied for online, and the share that had actually applied for online loans in the past two years was even lower, at around 2%. By contrast, embedded digital technologies, such as drones and Beidou-based navigation systems, were more widely used in agricultural production. Data from nine provinces show that about 8% of farmers used drones for crop protection in 2023, reflecting a clear upward trend. In major grain-producing regions, drones covered 39.3% of cultivated land for plant protection, largely through service outsourcing. In addition, the share of farmers using machinery equipped with Beidou navigation systems, weighted by the number of households and cultivated land area, reaches 8% and 12.5%, respectively. Building on these findings, the study analyzes the relatively low adoption of non-embedded technologies in agricultural settings from the perspectives of farmer awareness, digital literacy, infrastructure, and policy support. It also explores the diffusion of embedded technologies from the angles of cost, agricultural organization (particularly the development of service providers), and industry standards. [Conclusions] Overall, China's agricultural and rural digital transformation is still at an early stage. The adoption of non-embedded applications in agricultural production and marketing remains very limited, while the use of embedded technologies exhibits substantial heterogeneity across farm sizes and application contexts. In addition, although smallholders rarely invest in digital machinery themselves, they can still benefit from digital transformation through outsourced services and socialized service provision. Further analysis suggests that advancing agricultural digitalization requires improving farmers' digital literacy, fostering new agricultural business entities with higher human capital, promoting the development of appropriately scaled farming, and strengthening the capacity of service providers. These measures are essential to enable broader and more effective diffusion of DITs and to support agricultural modernization.

[Significance] The escalating complexity of genotype-phenotype-environment interactions and the explosive growth of multi-omics big data have necessitated a paradigm shift in crop breeding from empirical selection to intelligent design (Breeding 5.0). The aim of this paper is to systematically explore the underlying logic of the AI-driven crop breeding paradigm shift, comprehensively outline its generational evolution, analyze its core technical implementations in phenomics, genomics, multi-omics integration, and molecular design, and dissect how large agricultural foundation models reconstruct the entire seed industry workflow. [Progress] The historical evolution of crop breeding was first traced from 1.0 empirical domestication to 5.0 smart Breeding characterized by the deep integration of biotechnology (BT) and information technology (IT). In germplasm resource evaluation, deep learning algorithms enabled high-dimensional pattern recognition and unsupervised feature compression to unlock rare alleles from unannotated sequences, large language models (LLMs) like PlantConnectome and wheat germplasm information extraction (WGIE) leverage retrieval-augmented generation (RAG) to automatically construct structural knowledge graphs from unstructured historical literature, achieving predictive and dynamic germplasm evaluations. In high-throughput phenotyping, industrial platforms capture 3D point cloud and multi-spectral data at 0.1 mm resolution, while convolutional neural networks couple with the integrated genomic-enviromic prediction (iGEP) framework to build full-lifecycle digital twin crop models in virtual space. Regarding genomic prediction, the limitations of linear paradigms were dissected and cutting-edge deep learning architectures were highlighted: SoyDNGP applied a 3D-CNN to map chromosomal topology for complex soybean traits; DPCformer employed self-attention mechanisms to dynamically calculate environmental weights under multi-adversarial constraints; and HyenaDNA utilized long-convolution filters to bypass the O(N²) computational complexity limitation of standard Transformers, reducing it to O(N·log N) for chromosome-scale modeling. For multi-omics integration, intermediate fusion strategies were elucidated for their superior capacity to capture cross-layer biological compensatory pathways. In molecular design breeding, foundational plant language models were highlighted, such as AgroNT for zero-shot expression prediction and OpenCRISPR-1, the world's first de novo AI-generated genome editor built to capture underlying physico-chemical syntax. Furthermore, the emergence logic of major domestic and international agricultural foundation models was analyzed through the mathematical lenses of scaling laws, parameter-efficient fine-tuning, and multi-task evaluation benchmarks. Finally, empirical effectiveness was comprehensively evaluated through multinational success stories, including Bayer's Climate FieldView, IRRI's night-temperature thermal models for "Green Super Rice", and China's state-led breeding platforms for stress-resistant maize and high-yield soybean. [Conclusions and Prospects] Key structural challenges that smart breeding faces are thoroughly dissected: data silos and standardization dilemmas, extreme computational resource asymmetry and high training costs, the lack of biological causal logic in deep learning "black boxes", and the structural scarcity of interdisciplinary BT-IT talent. To overcome these bottlenecks, future research and policy efforts should focus on four pillars: 1) Establishing standardized open data ecosystems following FAIR (Findable, Accessible, Interoperable, Reusable) principles and promoting federated learning under a national AI data copyright integration platform to safeguard digital borders; 2) Exploring cloud-to-edge lightweight model deployment pathways through parameter-efficient fine-tuning, quantization, and knowledge distillation to empower real-time field-side decision-making and realize compute equity; 3) Deepening the mechanistic integration of AI and synthetic biology by absorbing breakthroughs from multi-scale modeling frameworks into the virtual "design-build-test-learn" cycle to break natural evolutionary thresholds; and 4) improving regulatory approval frameworks for AI-designed crops and modernizing agricultural education to cultivate a new generation of digital agronomists.

[Objective] As agricultural modernization accelerates, intelligent agricultural machinery is playing an increasingly important role in enhancing productivity and ensuring food security. Among these technologies, BeiDou navigation tractors have shown considerable potential in improving operational precision and efficiency. However, their adoption among farmers remains limited, and large-scale diffusion has yet to be achieved. Understanding what drives or constrains farmers' adoption decisions is therefore critical for promoting the effective use of intelligent agricultural machinery. The aim of this research is to identify the key determinants of farmers' adoption behavior and to uncover the underlying mechanisms shaping these decisions. [Methods] Micro-survey data collected from 1 242 maize farmers across six provinces in China were used as the basis. An analytical framework was constructed from five dimensions, including household head characteristics, family endowments, operational conditions, regional development, and the external environment. Explainable machine learning methods were employed to identify the key driving factors influencing farmers' adoption of BeiDou navigation tractors. [Results and Discussions] The results showed that farmers' adoption decisions regarding BeiDou navigation tractors were characterized by the dominance of a small number of core factors. Among these, policy support for intelligent agricultural machinery, total household income, farm size, age, and participation in agricultural training ranked as the top five determinants. This indicated that policy incentives, economic capacity, land endowments, and human capital were the primary factors influencing farmers' adoption decisions. In terms of the direction of influence, stronger human, physical, and social capital effectively reduced the barriers to technology adoption. The consolidation and contiguity of farmland, together with the application of complementary technologies, significantly improved the suitability of BeiDou navigation tractors. In addition, a well-developed policy support system and agricultural socialized service network created favourable conditions for adoption. The heterogeneity analysis further revealed that smallholder farmers relied more on demonstration effects and external service support, whereas large-scale farmers' adoption decisions were more strongly driven by policy incentives and economic returns. Younger farmers exhibited adoption behaviour that was more evidently driven by economic and policy factors, while older farmers showed greater dependence on services and a stronger tendency towards risk aversion. Farmers operating under less favourable terrain conditions were more influenced by resource endowments and policy support, whereas those in more favourable conditions were more driven by access to services and human capital. [Conclusions] Based on these findings, it is suggested that promoting the large-scale adoption of intelligent agricultural machinery requires improvements in policy support. First, the focus should shift from hardware subsidies to performance enhancement. On the one hand, differentiated subsidy policies should be implemented. On the other hand, subsidy mechanisms should be extended from equipment purchase to operational use. Second, efforts should be made to promote moderate-scale farmland management in order to improve the conditions for applying intelligent agricultural machinery. This includes facilitating land transfer and service arrangements such as land trusteeship to encourage land consolidation and contiguity, thereby enabling the precision operation advantages of BeiDou navigation tractors to be fully realised. At the same time, the coordinated application of agricultural machinery and agronomic practices should be promoted. Under suitable conditions, integrating BeiDou navigation tractors with high-density planting techniques can further enhance their cost-saving and efficiency-improving potential. Finally, it is necessary to improve the agricultural socialized service system and training framework, and to implement differentiated promotion strategies for different groups. On the one hand, greater support should be given to agricultural machinery cooperatives and agricultural service organisations to provide outsourced services such as ploughing, planting, and land management for smallholders. On the other hand, more digital skills and practical training resources should be directed towards large-scale farmers, alongside technical guidance on equipment maintenance and data management, so as to better utilise their leading role in the adoption and diffusion of intelligent agricultural machinery.

[Objective] Plant factories with artificial lighting (PFALs) provide year-round production, controllable environments, consistent product quality, and high space-use efficiency, positioning them as an important form of controlled-environment agriculture (CEA) moving toward intensification and digitalization. However, their commercialization has been constrained by high capital investment, electricity dependence, labor and operation-and-maintenance costs, and insufficient realization of market value. Existing studies have typically examined crop yield, light-environment control, energy use, capital cost, or market price in isolation, with limited integration of crop production, control maturity, energy conditions, and market realization into a computable framework. The aim is to identify the profitability boundary of PFAL lettuce, clarify how control maturity affects marketable yield, unit electricity use, labor substitution, annualized capital cost, and unit cost, and identify feasible transition pathways. [Methods] A production-side accounting boundary was adopted. Annualized capital cost, maintenance cost, electricity cost, labor cost, nutrient solution and seed costs, and other operating costs were included, whereas cold-chain logistics, retail terminal costs, brand advertising costs, financing costs, and complete channel-organization costs were excluded. Three levels of control maturity, three energy scenarios, and three market scenarios were specified, resulting in 27 deterministic scenarios. The model was constructed following the logic of "scenario input – control mapping – cost – benefit calculation – profitability boundary identification – vulnerability diagnosis". Control maturity was incorporated into the profit function through gross yield, marketable rate, unit electricity consumption per unit of marketable product, labor-substitution coefficient, and unit capital expenditure (CAPEX). An economic vulnerability index (EVI), consisting of profit gap, energy exposure, and carbon-constraint exposure, was further constructed. Local elasticity analysis, weight-robustness tests, and extended scenarios involving policy support and channel costs were used to examine the explanatory boundary of the results. [Results and Discussions] The economic feasibility of PFAL lettuce exhibited a distinct "narrow-window" characteristic. Among the 27 deterministic scenarios, only 5 achieved positive profit, accounting for 18.5%, and all were concentrated in the high-value direct-supply market. Under the benchmark scenario of "conventional grid electricity + high-value direct-supply market", upgrading control maturity from basic control to closed-loop intelligent control increased marketable yield from 70.40 to 109.25 kg/(m²·year), reduced unit electricity consumption from 12.0 to 8.4 kWh/kg, decreased unit cost from 27.67 to 19.26 CNY/kg, and increased profit from -258.36 to 518.04 CNY/(m²·year). Cost decomposition showed that although control upgrading increased annualized capital cost per unit area, higher output diluted capital cost per unit product. Meanwhile, improved labor substitution and reduced unit electricity consumption lowered labor cost and electricity cost, respectively. Break-even analysis indicated that higher control maturity flattened the break-even boundary between selling price and electricity price, reflecting lower sensitivity to electricity price fluctuations. The EVI results further showed that profitability ranking and vulnerability ranking were not fully consistent. The best scenario was "closed-loop intelligent control + energy-abundant condition + high-value direct-supply market", with an EVI of 0.088, whereas the worst scenario was "basic control + high-price and high-carbon electricity condition + conventional fresh-food market", with an EVI of 0.727. Local elasticity analysis showed that marketable yield had the largest effect on unit cost, with an elasticity of approximately -0.57, followed by unit CAPEX at approximately 0.49. The elasticities of electricity price and unit electricity consumption were both approximately 0.33. Sensitivity analysis of policy support and channel costs showed that investment subsidies and preferential electricity prices improved the financial performance of some boundary scenarios, whereas additional costs associated with packaging, fulfillment, channel maintenance, and sales organization compressed profit margins in high-value markets. [Conclusions] The feasibility of PFAL lettuce production is not determined by single-factor cost reduction, but by the joint effects of control maturity, energy conditions, market value realization, and channel costs. Conventional fresh-food markets and general premium-brand markets are unlikely to support profitable PFAL lettuce production. Only in high-value direct-supply markets, and when the control level reaches at least the enhanced-control stage, can the system cross the break-even line. The value of control upgrading should not be understood merely as electricity saving, but as a comprehensive mechanism that simultaneously increases marketable yield, improves the marketable rate, reduces unit electricity consumption, enhances labor substitution, and dilutes capital cost. A more robust transition pathway should therefore be built on the synergy among control upgrading, favorable electricity conditions, value-chain upgrading, and policy support. The conclusions of this study are applicable to production-side boundary identification under publicly available data conditions, but should not be directly interpreted as evidence of stable profitability for specific commercial projects.

[Significance] Since the establishment of the "dual carbon" goals, promoting the low-carbon transformation of lifestyles has become a core issue of green and high-quality development. However, most existing studies selected urban residents as the research subject for investigating low-carbon behaviors, systematic exploration from the perspective of rural residents needs to be supplemented. The implementation of the Digital Rural Development Strategy creates favorable conditions for digital empowerment in the low-carbon transformation of rural residents' lifestyles. Based on the perspective of rural residents' behavior, this paper systematically sorts out relevant literature, aiming to explore the core difficulties of rural residents' lifestyle low-carbon transformation and the feasible paths of digital empowerment, so as to provide theoretical reference and practical guidance for the deepening of research in this field and the optimization of relevant policies. [Progress] Firstly, the mainstream theories and research findings were reviewed on resident behavior, concluding that existing studies mostly focus on urban residents, with insufficient relevance analysis and in-depth discussion in rural contexts. Secondly, the core dilemmas in the low-carbon lifestyle transformation of rural residents were analyzed from the two dimensions of behavior and cognition. At the behavioral level, rural residents have low acceptance of new low-carbon practices; their behavioral decisions are strongly influenced by social relations such as regional and kinship ties, and external strong stimuli are needed to drive behavioral change. At the cognitive level, rural residents have limited exposure to low-carbon knowledge, which hinders the formation of low-carbon awareness. Meanwhile, there is a significant deviation between low-carbon cognition and actual behavior, and more external incentives are required for rural residents to translate awareness into action. The proposal of the Digital Countryside has provided hardware infrastructure and policy support for digitally driving the low-carbon transformation of rural residents' behavior. Accordingly, the common enabling scenarios of digitalization were further sorted out in rural life and explored feasible paths for the low-carbon transformation of rural residents' behavior in the digital era. From one aspect, embedding digital technologies and their derivative products is a measure to reshape residents' living habits. Specifically, digital technologies can optimize residents' daily life experience and improve efficiency, while digital financial products and services can lower the costs of low-carbon practices for rural residents. From another aspect, leveraging digital platforms as communication channels helps gradually cultivate residents' low-carbon awareness. This is reflected in the efficient dissemination of information that expands rural residents' scope of low-carbon knowledge, and digital social interactions that transform their traditional mindsets. Finally, China's current policy system related to guiding residents' low-carbon behavior was reviewed and evaluated. It is characterized by national macro-plans that set the overall development direction, and local policies that focus on strengthening digital infrastructure construction. [Conclusions and Prospects] Digitalization can accurately address the behavioral and cognition difficulties in the low-carbon transformation of rural residents' lifestyles through four paths: technology embedding, financial empowerment, information dissemination, and social guidance. Future research needs to further expand the applicability of low-carbon behavior theories in rural scenarios, verify the carbon reduction effect and heterogeneous impact of digitalization in rural living scenarios, and construct a two-way interactive digital carbon reduction policy system featuring government guidance and resident participation, so as to provide more targeted theoretical support and policy ideas for the coordinated development of rural residents' lifestyle low-carbon transformation and rural revitalization.

[Significance] Digital literacy is essential for farming and herding households to integrate into digital financial markets and access credit services. Against the backdrop of the rapid development of inclusive digital finance, a systematic review of existing research on the relationship between digital literacy and the borrowing behaviour of farming and herding households is significant for deepening rural financial theory, revealing the mechanisms underlying the digital divide and formulating targeted digital financial policies for rural and pastoral areas. This paper focuses on farming and herding households, emphasising that, in a fintech-driven environment, digital literacy influences access to financing, economic resilience, and sustainable development capabilities, and is not merely a matter of technological adoption. [Progress] The concept of digital literacy had undergone a progressive evolution from information literacy, internet literacy, and media literacy to digital literacy, gradually forming a comprehensive competency framework that encompassed multiple dimensions, including information retrieval, communication and collaboration, content creation, and security. About measurement, the EU's DigComp 2.2 and UNESCO's DLGF had been adopted extensively. Research had constructed indicator systems across multiple dimensions, including general, social, creative, security, and financial aspects. However, the adaptability of existing measurement tools in pastoral contexts remained insufficient. Regarding the loan behaviour of farming and herding households, extant research was found to focus primarily on financing accessibility, the "last-mile" bottlenecks in rural finance, and the role of digital literacy in promoting digital financial services. The accessibility of financing was influenced by a multitude of factors, including individual and household characteristics, social capital, the policy environment, and technological development. Rural financial bottlenecks were identified as the result of several challenges, such as poor institutional transmission, insufficient technological adaptation, lack of resource guarantees, and mismatched service design. Digital literacy was demonstrated to enhance the availability of both formal and informal credit for farming and herding households by means of improving information access efficiency, expanding social networks, increasing credit visibility, and optimising risk identification. It was also shown to play a positive role in promoting income growth, technology adoption, and economic resilience. [Conclusions and Prospects] The extant research was found to confirm the positive impact of digital literacy on the loan-taking behaviour of farming and herding households. However, significant limitations remained to be addressed. Firstly, the research perspective had been primarily focused on agricultural areas and farming households, with a severe lack of specialised studies on pastoral areas and herding households, thereby overlooking the unique circumstances of pastoral areas, such as the seasonal nature of production, the biological nature of assets, and the dispersed nature of settlements. Secondly, digital literacy measurement tools were inadequately adapted to pastoral contexts, failing to cover core pastoral production scenarios such as pasture monitoring, remote livestock management, and e-commerce sales. Thirdly, research on the underlying mechanisms remained superficial; specifically, how digital literacy influenced borrowing behaviour through intermediary pathways such as social capital, financial knowledge, and risk preference, particularly the differentiated mechanisms between formal and informal channels, had yet to be clarified. Fourthly, difficulties in data acquisition and limitations in research methods were found to have constrained the reliability of causal inferences. Fifthly, there was a mismatch between digital finance policies and the actual needs of pastoral areas. It was recommended that future research efforts concentrate on the development of a theoretical analytical framework and measurement system that was specifically tailored to the pastoral context. This was to be accompanied by the advancement of large-scale, long-term micro-surveys and the development of panel data. Researchers were also advised to actively incorporate methods such as experimental economics and machine learning into their studies. Furthermore, there was a need to deepen the analysis of stratified studies on the internal heterogeneity of herding households. The implementation outcomes of existing digital inclusive finance policies and training programmes were to be systematically evaluated. Additionally, it was considered crucial to proactively address emerging topics such as green credit, insurtech and supply chain finance. Finally, critical issues including digital risk governance and algorithmic fairness were noted as requiring due attention. This provides both theoretical support and practical guidance for ensuring that digital finance benefits a wide range of farmers and herders in a more precise and equitable manner.

[Objective] Digital rural construction has become an important strategy for promoting rural revitalization and breaking the long-standing urban-rural dual structure in China. Under the goals of common prosperity and agricultural modernization, sustained income growth for rural residents is a key policy concern. Existing studies have explored the economic effects of digital rural development, but several gaps remain. Most research focuses on a single outcome rather than jointly considering income growth and broader inclusive development. Many studies also rely on composite indices that may be endogenous to local conditions. In addition, the channels through which digital rural policies affect rural income have not been fully clarified. To address these issues, the National Digital Rural Pilot Policy was used as a quasi-natural experiment in this research to examine whether digital rural construction raises rural residents' income, the transmission mechanisms of this effect and its effect on the urban-rural income gap. [Methods] Using the county-level panel data for 727 counties in China from 2014 to 2023, the 2020 launch of the National Digital Rural Pilot Policy was treated as a quasi-natural experiment, and a difference-in-differences model was applied to identify the policy's effect on rural residents' per capita disposable income.The analysis controlled for economic foundation, fiscal expenditure, financial development, population density, savings level, industrialization, service-sector development, and education. To test robustness, parallel trend tests, placebo tests, the exclusion of special samples, controls for other concurrent policies, and propensity score matching combined with DID estimation were conducted. Mechanism tests were used to examine whether the policy works through labor allocation optimization and enhanced entrepreneurial activity, while moderating effect models assessed whether county economic foundation, digital financial inclusion, and industrialization strengthen the income-enhancing effect. [Results and Discussions] The results showed that the National Digital Rural Pilot Policy significantly increased rural residents' income. After county and year fixed effects as well as other relevant factors were controlled for, the policy led to a significant increase in rural per capita disposable income, and this finding remained robust across a series of tests. Dynamic analysis showed no significant difference in pre-policy trends between pilot and non-pilot counties, while the positive effect emerged after policy implementation and strengthened over time, indicating a sustained and cumulative policy impact. Mechanism analysis identified two main channels through which the policy promoted income growth. First, digital rural construction improved labor allocation by fostering new forms of rural economic activity, such as digital agriculture and rural e-commerce, reducing labor market frictions, and improving the matching efficiency between labor and employment opportunities. Second, it enhanced entrepreneurial activity by lowering market entry and operating costs, increasing the vitality of agriculture-related business entities, and creating more opportunities for local business development and income generation. The income-enhancing effect was more pronounced in counties with a stronger economic foundation, higher levels of digital financial inclusion, and greater industrialization. However, although the policy significantly increased rural residents' income, it did not significantly reduce the urban-rural income gap, suggesting that absolute income growth did not necessarily lead to relative distributional convergence. [Conclusions] Digital rural construction is an effective pathway for increasing rural residents' income in China. By exploiting the National Digital Rural Pilot Policy as a quasi-natural experiment, this study provides more credible evidence on the income effects of digital rural development and clarifies its main transmission mechanisms. Future policy efforts should continue to advance digital rural construction while focusing on improving labor allocation, enhancing entrepreneurial activity, and strengthening county-level supporting conditions. Greater attention should also be paid to digitally disadvantaged areas and vulnerable rural groups in order to promote the inclusive sharing of digital dividends and foster more balanced urban-rural development.

[Objective] As a key policy tool for empowering the development of the rural digital economy and broadening farmers' income channels, the rural e-commerce policy holds significant practical importance for strengthening farmers' risk resistance capabilities and improving sustainable livelihood levels. The purpose of this research is to systematically investigate the effects, mechanisms, and heterogeneous characteristics of the rural e-commerce policy on farmers' livelihood resilience, scientifically evaluate the policy implementation outcomes, and provide theoretical references and decision-making bases for optimizing the arrangement of rural e-commerce policies, improving the digital policy system benefiting farmers, and enhancing farmers' sustainable development capabilities. [Methods] Based on data from the China Family Panel Studies (CFPS) from 2012 to 2022, the National Rural E-commerce Comprehensive Demonstration Policy was treated as a "quasi-natural experiment" and a staggered difference-in-differences (DID) model was employed to empirically assess the effects, mechanisms, and heterogeneity of rural e-commerce access on farmers' livelihood resilience. To solidify causal identification, a series of robustness tests and endogeneity treatments were conducted. Addressing potential biases in staggered DID estimations, a Bacon decomposition was introduced to refine the policy effect estimates. Placebo tests were utilized to rule out interference from unobservable factors and random shocks. Simultaneously, the analysis controlled for the interference of concurrent policies such as the Information Entering Villages and Households Project, the Broadband China Strategy, and the National E-commerce Demonstration Cities, and removed special samples to purify the data. To address sample selection bias, a PSM-DID model was employed for re-estimation, matching characteristics to enhance inter-group comparability. Lagged-term regressions were introduced to mitigate reverse causality issues, clarifying the temporal relationship between the policy and farmers' internet use, thereby further strengthening the rigor of the conclusions. [Results and Discussions] It was confirmed through parallel trend tests that there were no significant differences in trends between the treatment and control groups before policy implementation, satisfying the prerequisite for quasi-natural experiment identification. Placebo tests excluded interference from random factors and omitted variables, and the Bacon decomposition refined policy effect estimates, avoiding biases from heterogeneous treatment effects. Furthermore, the analysis ruled out the confounding effects of concurrent policies and excluded specific special samples to purify the data. Addressing endogeneity concerns, the PSM-DID model mitigated sample selection bias, and lagged-term regressions dismissed potential reverse causality. Following these multi-dimensional robustness checks and endogeneity treatments, the core conclusion that the rural e-commerce policy significantly enhanced farmers' livelihood resilience remained robustly valid. Inspired by the Theory of Planned Behavior, the rural e-commerce policy influenced farmers' livelihood resilience through the synergistic effect of three core dimensions: cognitive, social, and skills. Cognitively, the policy improved rural network and logistics infrastructure, highlighted the value of the internet in production and sales, shifted farmers' traditional conservative mindsets, and strengthened positive expectations regarding e-commerce for income growth. This stimulated their intrinsic motivation to actively participate in e-commerce and diversify livelihood channels, thereby solidifying the psychological and cognitive basis for resilience enhancement. Socially, leveraging demonstration county platforms, the policy connected with industry associations and e-commerce cooperatives to conduct training and resource matching, integrating farmers into organized systems. Utilizing kinship and geographic networks, it facilitated information sharing and peer demonstration effects, building a supportive social environment and consolidating external support for risk resistance. In terms of skills, the policy provided supporting practical e-commerce training, addressing farmers' deficiencies in digital skills. It enhanced their practical abilities in online operations, customer maintenance, and logistics coordination, overcoming psychological resistance related to e-commerce operations. This empowered farmers to optimize resource allocation and flexibly respond to market risks, providing solid skill support for enhancing livelihood resilience. Heterogeneity Analysis: From an economic regional perspective, the policy's effect was significant in central and western regions but insignificant in eastern regions. Regarding policy support background, the policy effect was stronger in poverty-stricken counties and non-old revolutionary base areas. Concerning county-level logistics infrastructure, areas with better logistics conditions exhibited a more pronounced policy effect. [Conclusions] The rural e-commerce policy significantly enhanced farmers' livelihood resilience, a conclusion that holds after multi-dimensional robustness checks. Mechanism analysis verified the crucial roles of cognitive transformation, organizational embedding, and skill improvement in the policy transmission process. Heterogeneity manifests as significant regional and conditional differences in the policy effect. In the future, policy design should shift from infrastructure investment towards systemic capacity building, continuously prioritizing resources towards central and western regions and poverty-stricken counties, while simultaneously improving the county-level circulation system.