Issue Ebookcover storyIssue CoverIssue Contents
List of IssuesDocument Types
- Abstract：Focusing on the development needs of China’s animal husbandry industry， this paper systematically reviews the research and application status of artificial intelligence technology in intelligent monitoring and regulation of breeding environments， intelligent diagnosis of livestock health and diseases， livestock genetic breeding， and precision feeding. Combined with the cases of Shandong Shanxian Country Sheep Industry Technology Backyard and the Otog Banner Animal Husbandry Technology Backyard， the operational logic and implementation effects of each model are analyzed. The study shows that： 1） AI combined with IoT， machine vision， and deep learning technologies can effectively improve breeding standards. However， the promotion of ‘AI + Science and Technology Backyard’ still faces many obstacles， such as weak localization of AI models， high construction and operation costs， shortage of multidisciplinary talents， and incomplete supporting policies and multi-party coordination mechanisms. 2） This study proposes three types of application models for science and technology small institutes： research-driven， industry service-oriented， and full-chain empowerment. 3） The paper proposes methods such as developing localized AI models， improving the coordination mechanism among government， industry， academia， research， and application， optimizing policy subsidies， and cultivating multidisciplinary talents. In summary， this article can provide theoretical basis and practical reference for upgrading technology institutes， digital transformation of animal husbandry， and comprehensive rural revitalization.Keywords：artificial intelligence;smart livestock husbandry;science and technology backyard;digital technology91|27|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：High-quality fish breeding is the foundation and seed-source guarantee for high-quality， efficient and sustainable aquaculture. With the increasing application of artificial intelligence in fisheries， smart breeding technologies characterized by multidimensional environmental sensing， precise phenotypic monitoring and multi-omics integration have developed by combining knowledge from the environment， phenome， metabolome and genome. Focusing on how field environmental and phenotypic data from aquaculture can be transformed into evidence for germplasm evaluation， this review summarizes the application of artificial intelligence in stress tolerance evaluation， disease resistance evaluation， growth and feed efficiency analysis， quality detection， germplasm resource digitization， gene mining， multi-omics modeling and digital twin platforms. It also discusses the applicability and limitations of these technologies in different farming scenarios. Overall， artificial intelligence provides new data sources and analytical tools for fish breeding， but its application is still limited by data accumulation costs， collection standards， model generalization， cross-species transferability and collaboration among research institutions， enterprises and farms. Future work should rely on improved seed farms， facility-based aquaculture units and regional demonstration bases for long-term validation， establish traceable data standards and sharing mechanisms， and promote the use of smart aquaculture data in aquatic germplasm innovation and green， efficient aquaculture.Keywords：artificial intelligence;fish breeding;germplasm resources;phenotypic data acquisition;smart breeding;digital twin67|54|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：To solve the problem of difficulty in fine classification of ornamental fish， this study conducted a PETFish-MViT model based on MobileViT. By introducing a mutual labeling enhancement module to optimize attention allocation， reconstruct the feature fusion structure， and adopt a label smoothing loss function， the model’s classification performance for ornamental fish is improved， alleviating the problem of uneven data distribution. The results showed that：1） The classification of the model on the self-built ornament fish data reached 91.4%， with an F1 value of 90.7%. Compared to the original model， the accuracy increased by 2.8%， and the F1-score increased by 2.8%. 2） On the public dataset WildFish test， the model achieves accuracy rate of 77.0%， indicating that the PETFish-MViT model exhibits certain generalization performance. In summary， this model can provide a classification method and tool for ornamental fish enthusiasts， playing a significant role in the ornamental fish industry.Keywords：ornamental fish classification;deep learning;Transformer;MobileViT59|27|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：To systematically review the current state， research hotspots， and evolutionary trends in the field of genomics driven by artificial intelligence， this study takes the relevant literature obtained from the China National Knowledge Infrastructure （CNKI） and Web of Science （WOS） databases from 2015 to 2025 as the research object. CiteSpace software combined with bibliometric methods was used to analyze the four aspects： research status， citation patterns， hotspots and trends， and applications in agriculture. The results indicated that： 1） The number of publications in the field of genomics driven by artificial intelligence has grown rapidly and is highly concentrated in international academic journals， with China and the United States as the leading contributors. However， the centrality of China in the international collaboration networks remains relatively low. 2） Journals publishing papers in this field are widely distributed， and no dominant core journals have emerged. At the same time， the integration of method innovation and tool development （e.g.， TensorFlow/Pse-in-One） is the key to driving the sustainable development of genomics. 3） Research hotspots of this field have undergone significant dual iterations of technologies and applications， shifting toward multi-omics integration and cross-modal fusion， and shifting from basic tasks including sequence classification to the elucidation of complex molecular mechanisms. It is worth noting that cutting-edge artificial intelligence technologies represented by Transformer are still in the diffusion stage in the field of genomics， yet they are highly likely to become the core driving force in the future. 4） In agriculture， publications on AI-driven genomics account for merely 3.3% of the entire literature dataset of this study. Nevertheless， they have already demonstrated enormous application potential in precision breeding and the decipherment of complex agronomic traits， and will serve as a vital new quality productive force for modern agriculture. In summary， this study thoroughly reviews the current research landscape and frontier trends in the interdisciplinary field of artificial intelligence and genomics， offering valuable references and implications for future methodological innovations within the field and their application in agriculture.Keywords：genomics;artificial intelligence;agriculture;bibliometrics;CiteSpace;research hotspots51|37|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：To address the challenges of diverse inputs， long growth cycle， complex field conditions， and poor traceability of prescription execution in cotton production， this review synthesizes research progress， engineering constraints， and deployment conditions of Artificial Intelligence （AI） for whole-growth-cycle cotton management under the Perception-Decision-Execution （PDE） closed-loop framework. Publications on cotton precision agriculture and AI from 2015 to 2025 were retrieved from Web of Science， Scopus， Google Scholar， CNKI， and IEEE Xplore， with earlier foundational methods and representative equipment studies included as appropriate. A total of 81 publications were included， of which 41 were published in 2023 or later. The literature was organized into five layers： data and infrastructure， perception， prediction and inference， decision and prescription， and execution and machinery； a closed-loop evaluation framework covering model performance， agronomic benefit， resource-environment impact， and economic scalability was used throughout the analysis. The review reveals that： 1） The perception layer contains the highest concentration of studies and task types， covering at least five tasks， including emergence， growth status， pest and weed detection， boll opening， and boll detection. UAV remote sensing， deep semantic segmentation， and object detection are the dominant methods， though domain shift across years， fields， and complex backgrounds still causes performance degradation. 2） The prediction layer has expanded from yield， water stress， and nutrient diagnosis to interval-based forecasting， with recent studies focusing on UAV multi-source imagery， thermal infrared sensing， salinity retrieval， nitrogen diagnosis， and multimodal modeling； however， uncertainty expression and the alignment between lead time and management action remain insufficient. 3） Decision layer studies are still dominated by single-objective or local optimization for irrigation， fertilization， crop protection， and growth regulation， while constraint-aware decisions integrating yield， quality， cost， operation windows， and machinery capacity remain underdeveloped. 4） The execution layer is constrained by prescription formats， equipment interfaces， positioning accuracy， operation logs， and quality verification， and cotton-specific adaptation of standards such as ISO-XML/ISOBUS remains limited； the evaluation layer can be supported by empirical studies on variable-rate nitrogen management， water stress diagnosis， soil salinity monitoring， UAV spraying， and drift control. The bottleneck of digital and intelligent cotton is shifting from single-algorithm accuracy to data governance， interface standards， error attribution， and cross-actor collaboration. Traceable data infrastructure， low-annotation learning paradigms， and end-to-end closed-loop benchmarks should be prioritized to support scalable， auditable， and iterative deployment.Keywords：cotton;precision agriculture;artificial intelligence;PDE closed loop;variable rate application44|48|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：Existing large language model （LLM） agent frameworks lack semantic reliability verification for multi-source heterogeneous agricultural perception data， leaving them vulnerable to the “garbage in， garbage out”（GIGO） effect and catastrophic hallucinated decisions. This study proposes a four-layer functionally decoupled safety architecture grounded in an “untrusted-by-default” design principle and evaluates its decision-making safety in Xinjiang cotton production scenarios. The architecture reduces the catastrophic hallucination rate from 45% （tightly coupled baseline） to 13%， a markedly greater reduction than that achieved by the soft-validation reflective baseline （29%）. Ablation experiments confirm the independent contributions of the perception-layer hard validation and symbolic audit modules， which together constitute a defense-in-depth mechanism. On the Xinjiang Cotton Benchmark， a domain-adapted 8B model attains 72.1% accuracy-surpassing Llama3-70B （61.3%） with approximately one-ninth of the parameters. The architecture provides a trustworthy， efficient， and transferable solution for real-time multi-source heterogeneous data fusion and decision-making under computational constraints in smart cotton production.Keywords：cotton production;large language model;agent;hallucination mitigation;trusted decision-making35|27|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：Unmanned farms， by integrating biotechnology and engineering technology and through big data analysis， precise decision-making and intelligent management， have achieved unmanned agricultural production in all-weather， the entire process and spaces， significantly enhancing the comprehensive benefits of agriculture. They represent a typical form of agriculture 4.0. This article， in combination with the production process and specialized scenarios f open-field vegetables， analyzed the core technologies of unmanned farms for open-field vegetables， including information perception， data transmission， decision-making， execution and control. It also explains the basic path and current situation of building unmanned farms for open-field vegetables， and points out the constraints faced by the development of unmanned farms for open-field vegetables in China， such as open-field vegetable production models， artificial intelligence technology， intelligent equipment and inter-disciplinary talent. It further indicates the future development directions of unmanned farms for open-field vegetables， including agricultural standardization， precise decision-making and intelligent equipment.The development of open-field vegetable unmanned farms will promote the deep integration of artificial intelligence with the vegetable industry， leading to its transformation and upgrading.Keywords：open-field vegetable;unmanned farm;current status;challenges37|40|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：To clarify how artificial intelligence （AI） adoption shapes the green transition in farmland utilization， based on the 2023 China Household Finance Survey （CHFS） conducted by Southwestern University of Finance and Economics， this study employs a Probit model to empirically investigate the impact and transmission mechanisms of AI on farmer organic fertilizer adoption. The findings are as follows： 1） AI significantly promotes the green transformation of farmland utilization. Baseline regression results show that： Farmers using AI-assisted decision-making exhibit a 14.5% higher probability of adopting organic fertilizer. This conclusion remains robust after replacing the dependent variable， altering clustered standard errors， and employing instrumental variable estimation. 2） Mechanism analysis reveals that AI operates through two pathways： first， by mitigating farmers’ time preferences through the visualization of long-term benefits， thereby reducing discount rates； second， by enhancing the accessibility of agricultural credit， alleviating financial constraints associated with green production. 3） Heterogeneity analysis indicates that AI exhibits stronger marginal effects among farmers in regions with weak technical training infrastructure and those located far from agricultural service centers， suggesting its potential to compensate for deficiencies in traditional agricultural extension services. This study provides micro-level evidence for understanding how digital technologies empower agricultural green transformation. From a policy perspective， priority should be given to promoting AI applications in regions with weak agricultural extension services， with emphasis on leveraging its mechanisms to improve farmers’ time preferences and credit constraints.Keywords：artificial intelligence;green transformation of farmland utilization;organic fertilizer adoption;time preference;agricultural credit;China Household Finance Survey （CHFS）49|33|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：To scientifically and accurately measure the development level of smart agriculture across Chinese provinces and clarify the mechanisms influencing its development， this study constructed a comprehensive evaluation index system based on panel data from 30 Chinese provinces from 2018 to 2024. The entropy weight method was used to measure the development level of smart agriculture in each province. Furthermore， based on the technology-organization-environment （TOE） theoretical framework， the fuzzy-set qualitative comparative analysis （fsQCA） method was applied to analyze the configurational pathways leading to high-level and non-high-level smart agriculture. The results indicate that： 1） The development level of smart agriculture in eastern and central China is higher than that in western and northeastern regions. The application of information technology is the key to smart agriculture， while agricultural R&D investment plays an important driving role in its development. 2） The combination of multiple factors such as agricultural R&D investment， policy attention level， and agricultural production level can effectively improve the development level of smart agriculture. 3） China’s high-level smart agriculture mainly follows two configurational pathways： the “technology R&D-digital foundation driven” pathway and the “technology R&D-human capital-fiscal support” pathway. In contrast， non-high-level smart agriculture can be summarized into three constraint pathways： insufficient R&D investment， insufficient transformation of policy support， and insufficient fiscal support with constrained transformation of agricultural foundations. The key difference lies in whether agricultural technological innovation capacity and agricultural R&D investment can provide stable support. Accordingly， it is recommended that all regions implement targeted policies， strengthen government-led overall planning and categorized support， emphasize the cultivation of interdisciplinary talents for smart agriculture， deeply empower agricultural production with digital technologies， and comprehensively improve the development level of smart agriculture in China.Keywords：smart agriculture;configuration analysis;TOE framework;fsQCA method;influencing factors analysis64|27|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：This study aims to explore the influence of AI TCM digital human image recommendations on agricultural live-streaming platforms on consumers’ willingness to purchase functional agricultural products. Based on 1 090 valid samples， four experiments were conducted to investigate the interactive effect of AI TCM digital human types and functional agricultural product types on consumers’ purchase intention， as well as the underlying influencing mechanism. The research results indicate that： 1） A significant interaction effect exists between the type of AI TCM digital human and the type of functional agricultural products on consumers’ purchase willingness. 2） Perceived value plays a mediating role in the interactive impact of AI TCM digital human recommendation and functional agricultural product attributes on consumers’ purchase willingness. 3） In terms of moderating effects， when consumers have individual perceived anxiety， traditional AI TCM digital human recommendation exerts a significantly stronger positive effect on purchase willingness than modern AI TCM digital human recommendation for health-preserving functional agricultural products. By contrast， for special conditioning and improvement-oriented functional agricultural products， modern AI TCM digital humans show a significantly stronger promotional effect on consumer purchase willingness than traditional ones when consumers perceive negative outcome anxiety. Accordingly， this study puts forward corresponding management implications and optimization strategies from the perspectives of live-streaming platform operation， TCM digital human image recommendation， functional agricultural product classification， and enterprise live-streaming marketing.Keywords：AI TCM digital human image;functional agricultural products;live streaming platform;perceived value;health anxiety52|34|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：In order to explore how the Science and Technology Backyard， as a new organizational carrier rooted in the countryside， can empower the integrated development of rural cultural tourism industry through the “Science and Technology Backyard +” model.In the study， its internal mechanisms and practical paths were analyzed.An embeddedness-collaboration-integration analytical framework was constructed based on the theory of embeddedness and collaborative governance. Taking the Science and Technology Courtyard in Gusheng Village， Yunnan as an example， the practical experience of the Science and Technology Backyard in empowering the development of rural cultural tourism through the mechanism path of “embedding-collaboration-integration” was analyzed. The results show that Gusheng Village Science and Technology Backyard first takes root in the local society through social embedding， institutional embedding and cultural embedding then， through a multi-dimensional collaborative mechanism of subject collaboration， resource collaboration， and functional collaboration， resources can be effectively integrated. Finally，it achieves deep integration of the cultural tourism industry through various forms of integration of industry integration， value integration and space integration. “Embedding-collaboration-integration” constitutes the core logical closed loop of empowerment in science and technology courtyards. Embedding is the basic premise of empowerment， while collaboration is the practical path of empowerment， and integration is the final effect of empowerment. Gusheng Village Science and Technology Backyard can provide reference for the functional expansion of science and technology backyards and the integrated development of cultural and tourism industries in other areas， and provide practical reference for optimizing and promoting the “Science and Technology Backyard +” model.Keywords：science and technology backyard;integration of culture and tourism;embeddedness theory;empowerment mechanism59|27|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：To explore the core issue of the integration of agriculture， culture， sports and tourism in the county driving urban-rural integration， based on the analysis framework of conjugate driving theory， this paper systematically analyzes its internal mechanism and explores the realization path that adapts to the actual development of the county. The results showed that： 1） The integration development of county-level agriculture， culture， sports and tourism mainly drives the integration development of urban and rural areas through multi-dimensional conjugation mechanism， forming a multi-dimensional action mechanism of factor conjugation， space conjugation， function conjugation， value conjugation and governance conjugation. The five conjugation dimensions are mutually linked and empowered， which promotes the mutual communication of urban and rural resources， complementary functions and mutual development. 2） The landing of the five conjugate dimensions needs the support of a systematic system， and a two-way flowing and dynamically balanced factor conjugate system should be constructed to solve the problem of mismatch between urban and rural factors. Create a spatial conjugate pattern with global coordination and distinctive characteristics， and optimize the spatial layout structure of urban and rural areas. Consolidate the functional conjugate closed loop of agricultural， cultural and sports tourism linkage empowerment， and strengthen the complementary functions of urban and rural industries. Build a value conjugate system of creating， sharing and balancing interests， and coordinate the value demands of urban and rural development. Improve the governance conjugation mechanism of multi-synergy and digital empowerment， and improve the governance efficiency of urban-rural integration.Keywords：county territory;urban-rural integration;integration of agricultural， cultural and sports tourism;conjugate drive theory43|27|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：To clarify the coupling mechanism between spatial production and the new productive forces in rural areas， based on Henri Lefebvre’s tripartite dialectics of space， this study takes tourism-oriented traditional villages as the research object， systematically expounding the theoretical logic and practical approaches driving the iterative update of spatial production and rural productivity. The research findings are as follows： 1） The tourism industry creates an industrial foundation for the productivity renewal of traditional villages. Spatial practice expands labor materials， representation of space creates production space， and spaces of representation revitalize labor objects. 2） Unordered tourism involvement is likely to lead to development dilemmas in traditional villages， such as spatial commodification， cultural abstraction， land non-agriculturalizing， and alienation of production nature. 3） In the construction of new productive forces in tourism-oriented traditional villages， attention should be paid to the innovative utilization of old- quality tourism production factors in the tripartite space through connection， with additional emphasis to the optimal allocation of the relationship between “people” and “things” in spatial production. From the perspectives of rural social space， material space and governance systems， it is essential to construct the value co-creation paths of self-organization and other-organization， explore the transformation of tourism production mode from “resource-consuming type” to “industry-supporting type”， promote the deep integration of village culture and modern tourism， and construct a new quality productive force oriented from resource consumption to talent cultivation.Keywords：traditional village;new quality productive forces;production of space;rural revitalization;tourism industry34|25|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：To consolidate the foundation of food security and promote high-quality development of the grain industry， this study utilizes the balanced panel data from 30 Chinese provinces from 2011 to 2022 （Statistics exclude the Xizang Autonomous Region and the Hong Kong， Macao， and Taiwan regions. The same applies below）. An evaluation index system is constructed for new-quality productive forces in the grain sector. Methods such as the Dagum Gini coefficient， kernel density estimation， Markov chain transition matrices， and spatial Durbin models are employed to investigate regional disparities， dynamic evolution， and influencing factors. The results show that： 1） The level of the new quality productivity of grain has been increasing year by year， showing a zonal pattern of “main production areas>main sales areas>production and sales balanced areas”. High-level areas are concentrated in the eastern coastal areas and central main production areas， while low-value areas are concentrated in the western regions and production-sales balanced areas. 2） From the perspective of dimensions， the new quality labor means of grain have the fastest growth rate， while the growth rates of the new quality laborers and the optimization of factor combinations are relatively slow. 3） The overall difference of the new quality productivity of grain has fluctuated and risen， and the regional differences are the main source of the overall differences. The polarization trend within each region has gradually emerged over time. 4） The development of the new quality productivity of grain is relatively stable， and there is a “club convergence”phenomenon. At the same time， the level of the new quality productivity of grain has a significant synergy with the type of neighboring areas， and high-level areas can drive the development of neighboring areas. 5） The level of economic development can significantly promote the development of the new quality productivity of grain in the local and surrounding areas. The transformation of consumption structure and the urbanization rate have a positive effect on the development of the new quality productivity of grain in the local area， but have a negative effect on the development of neighboring areas. Extreme climate disasters severely restrict the development level of the new quality productivity of grain. In view of this， this study proposes policy recommendations from four aspects： focusing on the core development shortfalls， improving the regional coordination mechanism， establishing a dynamic monitoring system， and optimizing the combination of policy and influencing factors.Keywords：new quality productivity in grain production;intrinsic characteristics;regional variations;spatio-temporal evolution;influencing factors41|32|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：To investigate the “overshooting” phenomenon in international grain price transmission and address the dual challenges of structural supply gaps and the lack of international pricing power in China’s grain market， this study constructs static and dynamic spillover index models based on futures and spot price data of major domestic and international grain varieties from 2018 to 2024. The price spillover effects， elasticity ranges， and asymmetric characteristics of international grain futures on the domestic market under exogenous shocks are systemically examined. The findings are as follows： 1） The spillover intensity of international grain futures is significantly higher than that of domestic markets， with Chicago Board of Trade （CBOT） soybeans exhibiting the strongest spillover effects， and a significant “upside sensitivity” phenomenon is observed in both domestic and international markets； 2） Price spillovers exhibit significant heterogeneity across varieties and asymmetry， with soybeans and corn being the most affected by exogenous shocks. Furthermore， the spillover effects are significantly stronger during price increase phases than during decrease phases； 3） The market adjustment mechanism demonstrates a U-shaped dynamic pattern of “short-term suppression and long-term enhancement”under exogenous shocks， where policy intervention plays a crucial role in stabilizing the market. Based on results， this study proposes the following policy recommendations： constructing variety-specific risk management tools to enhance the market resilience of grain futures； implementing categorized regulation strategies based on supply-demand elasticity and trade dependence； and establishing a coordinated regulatory mechanism that integrates “short-term emergency response” and “long-term fundamental solutions”.Keywords：exogenous shocks;grain futures;dynamic spillover index;asymmetry33|35|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：To explore the evolution characteristics and driving mechanisms of China’s grain supply chain resilience， based on panel data from 31 provinces in China （The data do not include those of Hong Kong， Macao， and Taiwan regions，the same below） from 2014 to 2023， this study constructed an evaluation index system and employed the entropy weight method to measure the resilience level of the grain supply chain. The Theil index was used to explore the sources of regional differences， the spatial Markov chain was utilized to characterize state transitions and spatial linkage features， and the Spatial Durbin Model （SDM） was constructed to identify driving factors and spatial spillover effects. The results showed that： 1） The resilience of China’s grain supply chain generally exhibited a fluctuating upward trend， presenting an unbalanced spatial pattern of “high in the east， low in the west and stable in the northeast”. 2） The decomposition of the Theil index indicated that the overall difference， inter-group difference， and intra-group difference of grain supply chain resilience were all narrowing， with intra-group differences being the primary source of the overall difference. 3） Spatial Markov chain analysis demonstrated that grain supply chain resilience possesses “path dependence” characteristics and was influenced by the spillover from the status of neighboring provinces. 4） Driving force analysis revealed that basic transportation infrastructure， scientific and technological investment， and digital foundation had significant positive direct effects on local grain supply chain resilience. Among them， the digital foundation served as the core driving force with both local efficiency enhancement and positive spatial spillover effects， while scientific and technological investment showed negative spatial spillovers due to barriers in achievement transformation. Additionally， financial support for agriculture supported local resilience but exhibited negative spatial spillover. In conclusion， it is suggested to strengthen digital empowerment to consolidate the foundation of full-industry chain synergy， optimize regional linkage mechanisms to balance policy spillover effects， and formulate differentiated resilience enhancement paths for the eastern， central， and western regions to fortify the national grain security defense line.Keywords：grain supply chain resilience;spatiotemporal differences;spatial Markov chain;spatial Durbin model;spillover effects37|24|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：To investigate the impact of policy-oriented agricultural insurance on grain production resilience， this study utilizes panel data from 31 Chinese provinces from 2005 to 2023 （The data do not include those of Xizang， Hong Kong， Macao， and Taiwan regions. The same below）. Employing a quasi-natural experiment based on the pilot policies of full-cost insurance and crop revenue insurance， a multi-period Difference-in-Differences model is used to empirically examine the policy’s effect on grain production resilience and its underlying mechanisms. The findings reveal that： 1） The pilot policy significantly enhances the level of grain production resilience， a result that remains valid after a series of robustness tests. 2） The positive impact of the pilot policy on grain production resilience is primarily achieved through three channels： the income guarantees effect， the mitigation of risk shocks， and the incentive to expand the scale of operations. 3） The effect of the pilot policy on grain production resilience exhibits significant heterogeneity across different regions， with more pronounced effects observed in western regions and areas with advanced large-scale agricultural management. These findings provide important policy implications for further promoting the nationwide expansion of the pilot policy and for fully leveraging the role of agricultural insurance in enhancing the resilience of grain production.Keywords：policy-based agricultural insurance;grain production resilience;income guarantee effect;mitigation of risk shocks;incentive for scale expansion36|31|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：To explore the impact of digital rural construction on grain production resilience of western county-level regions， this research employs panel data from 1 041 counties in western China from 2001 to 2023. Using the national digital rural pilot policy as a quasi-natural experiment， a difference-in-differences （DID） model is applied to empirically examine the effects and underlying mechanisms of digital rural construction on grain production resilience. The findings are as follows： 1） Digital rural construction notably improves grain production resilience in western counties， and this conclusion remains valid after robustness tests. 2） Heterogeneity analysis indicates that the impact differs across counties. Improvements are particularly pronounced in mountainous or hilly counties， non-major grain-producing areas， northwest regions， and counties with higher initial resilience levels. 3） Mechanism test shows that digital rural construction enhances grain production resilience by increasing local fiscal support， raising the utilization rate of facility agriculture， and promoting rural labor transfer. It is suggested that western county-level regions promote digital rural construction tailored to local conditions to empower grain production， facilitate the synergy of fiscal revenue and expenditure with technology， break through the application bottleneck of facility agriculture， and promote the non-agricultural employment of labor force.Keywords：digital rural construction;national digital rural pilot policy;grain production resilience;difference-in-differences model;effect evaluation28|23|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：To clarify how the high-standard farmland construction policy affects the consistency between food security and farmers’ income increase， this study comprehensively employs the Regression Discontinuity Design （RDD） and Difference-in-Differences （DID） models to examine the impact and mechanisms of the policy on the consistency between food security and farmers’ income increase. The results show that： 1） The high-standard farmland construction policy significantly improves the consistency between food security and farmers’ income increase. 2） The implementation effects of the policy vary across stages and regions， with relatively more pronounced effects during the comprehensive deepening stage， in low-relief areas， and in major grain-producing regions. 3） Promoting rural industrial integration， developing agricultural socialized services， and facilitating rural labor transfer are three key mechanisms through which the policy enhances the consistency between food security and farmers’ income increase. 4） The consistency effect of the policy on food security and farmers’ income increase exhibits long-term sustainability. Based on these findings， it is recommended to strengthen the full-cycle management of the high-standard farmland construction policy， promote targeted measures tailored to different regions and categories， and continuously optimize policy transmission pathways to fully unleash the policy’s effectiveness.Keywords：high-standard farmland construction policy;food security;farmers’ income increase;regression discontinuity design;difference-in-differences38|32|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：To systematically evaluate the drought resistance of winter wheat germplasms at the seedling stage and identify excellent drought-resistant accessions， 197 winter wheat germplasms were used as materials. In the seedling-stage experiment， two treatments were set up in a controlled growth chamber： Drought stress （DR1， with stress applied on day 7 and day 12 after sowing） and continuous watering treatment （CK1）. At the two-leaf-one-heart stage， 12 traits were measured， including the length of the first leaf， width of the first leaf， first leaf area， straw dry mass， fresh weight of straw， seedling water loss， total root system length， root system surface area， root system volume， root dry mass， root fresh mass， and root water loss. Drought resistance coefficients， the membership-function method， principal-component analysis， and cluster analysis were employed to comprehensively evaluate the seedling-stage drought resistance of the 197 germplasm resources. In the field experiment， two treatments were implemented： Natural drought （DR2， relying solely on natural precipitation） and conventional irrigation treatment （CK2）. At the maturity stage， the plant height， spike length， thousand-grain weight， and other traits were examined to determine the early-prediction ability of seedling-stage indicators for wheat yield and to screen drought-resistant germplasm. The results showed that： 1） In the seedling-stage experiment， the germplasm under drought stress （DR1） exhibited rich genetic variation. The total root length， root dry mass， and seedling water loss showed the largest coefficients of variation （CV>0.75）， while length of the first leaf and width of the first leaf were relatively stable （CV<0.40）. Under the continuous-watering treatment （CK1）， the mean values of total root length， root system surface area， and root system volume were significantly higher than those in the DR1 treatment， increasing by 29.85%， 29.24%， and 37.50%， respectively. 2） Two principal components extracted by principal-component analysis cumulatively explained 84.71% of the total variance. Based on the comprehensive drought-resistance coefficient （D）， cluster analysis classified the germplasm into five drought-resistance groups： 11 highly drought-resistant， 26 drought-resistant， 78 moderately drought-resistant， 63 drought-sensitive， and 19 highly drought-sensitive accessions. 3） Correlation analysis between the drought-resistance coefficients of each index and the indicators related to the maturity period. Under drought stress， the dry weight of the underground part of the tested germplasm at the DR1 seedling stage was significantly negatively correlated with the yield indicators such as grain length， grain width， and thousand-grain weight at the DR2 maturity period （P<0.05）. The fresh weight of the underground part of DR1 seedlings was significantly positively correlated with the ear length， grain circumference and thousand-grain weight of DR2 at maturity （P<0.05）. The dry weight and fresh weight of the underground part of CK1 seedlings were significantly positively correlated with the grain length， grain width and thousand-grain weight of CK2 at the mature stage （P<0.01）. In conclusion， the dry weight and fresh weight of the underground part of seedlings can be used as effective comprehensive indicators for the early identification of drought resistance of winter wheat at the seedling stage. A total of 11 excellent highly drought-resistant winter wheat accessions were identified based on the D-value， namely ‘24jx-102’， ‘24jx-113’， ‘24jx-182’， ‘24jx-195’， ‘24jx-138’， ‘24jx-8’， ‘24jx-150’， ‘24jx-83’， ‘24jx-88’， ‘24jx-68’， and ‘24jx-171’， which can be used in wheat drought-resistant breeding.Keywords：winter wheat;germplasm resources;drought resistance coefficient;cluster analysis;comprehensive drought resistance evaluation49|29|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：To screen for superior japonica rice varieties （lines） with combined tolerance to salt， drought， and low temperature during germination stage， 295 rice varieties （lines） were used as materials. Under simulated abiotic stress conditions of drought （DS）， salt-alkali （SS）， and low temperature （LT） inside an artificial climate chamber during germination stage， indicators such as germination potential， germination rate， root length， shoot length， and root number of the tested materials were measured with normal germination conditions as the control （CK）. The relative values and coefficients of variation （CV） of these indicators were determined. A comprehensive multi-tolerance evaluation system was constructed using the weighted membership function method， principal component analysis， and cluster analysis， and stress-tolerant germplasm resources were screened. The results showed that the CV of each relative trait varied with stress type. Under salt stress and drought stress， relative germination potential exhibited the highest at 4.63 and 1.44， respectively， whereas under low-temperature stress， it exhibited the highest， at 1.55. Based on the comprehensive evaluation value （D） and correlation analysis of relative traits， suitable identification indicators were determined for different stresses. Under salt stress， the identification indicators were relative germination potential， relative germination rate， relative shoot length， relative root number， and relative root length. Under drought stress， the identification indicators were relative germination rate， relative root length， relative germination potential， and relative root number. Under low-temperature stress， the identification indicators were relative germination rate， relative root number， relative shoot length， and relative root length. Through cluster analysis， the rice varieties under each stress were divided into three major groups： Highly tolerant， moderately tolerant， and sensitive. The D values ranged from 0.150 to 0.780 under salt stress， from 0.261 to 0.720 under PEG-simulated drought stress， and from 0.153 to 0.736 under low-temperature stress. Based on the combination of the mean membership function value （
- Abstract：To investigate the optimal water-saving and nitrogen reduction strategy for maize cultivation in mildly saline-alkali soils of Ningxia Yellow River irrigation area， a split-plot experiment was conducted in this region. The main plots were assigned to two irrigation amounts： 4 500 （conventional irrigation， W1） and 3 600 m³/hm2 （20% water saving， W2）. The sub-plots were assigned to four nitrogen application rates： 300 （conventional nitrogen， N1， calculated as pure N， same below）， 225 （25% nitrogen reduction， N2）， 150 （50% nitrogen reduction， N3）， and 0 kg/hm2 （N0）. Soil moisture content， salt content， nitrate-nitrogen distribution， maize root morphology， bleeding intensity， aboveground dry matter accumulation， root-shoot ratio， yield， and water-and nitrogen-use efficiencies were measured. Correlation analysis was performed to explore the relationships among these indicators. The results indicated that： 1） Compared with W1， W2 treatment significantly increased soil moisture content in the >40-60 cm soil layer at the silking stage （P<0.05）. Under the W2 condition， the soil moisture content in the >20-60 cm layer at the silking stage first increased and then decreased with decreasing nitrogen application rate， with the highest value in N2， which was 5.42%， 4.50% and 4.50% higher than that in N0， N1 and N3， respectively. During the silking stage and maturity stage， the soil salt content in the 0-20 cm layer both showed a continuous decreasing trend with the decreasing nitrogen application rate； N0 had the lowest value， which was 9.19%， 4.46%， and 0.98% lower than the two-stage average values of in N1， N2， and N3， respectively. The soil nitrate nitrogen content in the 0-80 cm layer also showed a continuous decreasing trend during the silking stage and maturity stage， and the values in N1， N2， and N3 were significantly higher than those in N0 （P<0.05）. 2） In the 0-20 and >40-80 cm soil layers， the root length density and root weight density of maize under the W2 treatment increased by 23.35% and 13.00%， respectively， compared with those under W1 （P < 0.05）. In the >40 - 80 cm layer， the root surface area density under W2 was 19.45% higher than under W1 （P < 0.05）. At the silking stage， the bleeding intensity under W2 increased by 17.81% compared with that under W1 （P < 0.05）.Under W2 condition， the root length density in >40-80 cm layer， as well as the root surface area density and root weight density in >20-80 cm all first increased and then decreased with decreasing nitrogen application rate. These three parameters all reached their maximum values under N2.Compared with N0， N2 increased these three parameters by 59.75%， 53.08%， and 81.81% （P<0.05）， respectively； compared with N1， by 56.54%， 27.46%， and 55.28% （P<0.05）； and compared with N3， by 72.62%， 18.51%， and 49.60% （P<0.05）， respectively. At 50 d after silking， the bleeding intensity in N2 was significantly higher than that in N0， N1， and N3（P<0.05）. 3） The water use efficiency under W2 was 9.38% higher than that under W1 （P<0.05）. Under W2 condition， the root-shoot ratio， kernel number per ear， 1 000-kernel weight， yield， and water use efficiency all first increased and then decreased with decreasing nitrogen application rate， with N2 showing the highest values. Compared with N0， N2 increased these five parameters by 50.36%， 7.06%， 15.04%， 19.00%， and 17.55%， respectively； compared with N1， the increases were 11.01%， 5.23%， 2.85%， 1.73%， and 2.01%， respectively； and compared with N3， the increases were 47.42%， 6.97%， 11.22%， 16.64%， and 15.53%， respectively. The agronomic efficiency of nitrogen fertilizer also first increased and then decreased with decreasing nitrogen application rate， reaching its highest value under N2， which was 49.64% and 659.40% higher than that in N1 and N3， respectively. 4） Soil salt content was significantly negatively correlated with yield （P<0.05）. In contrast， nitrate-nitrogen distribution， root length density， root weight density， bleeding intensity， aboveground dry matter accumulation， root-shoot ratio， kernels per ear， and 1 000-kernel weight were all highly significantly positively correlated with yield （P<0.01）. In summary， for spring maize production in mildly saline-alkali soils of the Ningxia Yellow River irrigation area， the combination of a 20% water-saving irrigation amount （3 600 m3/hm2） with a 25% nitrogen reduction rate （225 kg/hm2）， designated as W2N2， represents the optimal water-and nitrogen-saving strategy.Keywords：water and nitrogen regulation;spring maize;root system;dry matter accumulation;root-shoot growth;yield43|35|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：In the arid and semi-arid Yellow River irrigation districts of northwestern China， maize yield is jointly influenced by multiple factors including remote sensing， meteorological， soil， and irrigation conditions. The widespread presence of multicollinearity， information redundancy， and potential spurious correlations among these variables limits the predictive stability and interpretability of conventional correlation-driven models. To address this issue， this study developed a framework integrating Causal forest， Temporal convolutional network， and Temporal attention mechanism （CF-TCN-TA） for county-level maize yield prediction. The framework first employs Causal forest to estimate the potential effects of multi-source variables on maize yield and to identify key driving factors with strong explanatory significance， thereby mitigating the impact of redundant variables and potential spurious correlations on model training. Subsequently， the selected key variables are fed into a Temporal convolutional network to extract temporal variation characteristics during the maize growing season， and a Temporal attention mechanism is incorporated to adaptively enhance the information contribution of critical months. Validation was conducted using county-level maize yield data and multi-source remote sensing， meteorological， and soil data from the Ningxia Hui Autonomous Region from 2019 to 2023. The results indicated that the CF-TCN-TA model achieved the optimal predictive performance， with a coefficient of determination （R2） of 0.830 3， a root mean square error （RMSE） of 882.54 kg/hm²， a mean absolute error （MAE） of 706.60 kg/hm²， and a mean absolute percentage error （MAPE） of 10.80%， consistently outperforming conventional machine learning models and standard deep learning models. Causal effect evaluation further revealed that soil pH， deep-layer field water-holding capacity， air temperature， and remote sensing indicators such as the fraction of photosynthetically active radiation （FPAR） and the enhanced vegetation index （EVI） were among the important driving factors influencing spatial variations in county-level maize yield within the study area. The integration of causal-driven feature selection and temporal attention modeling can enhance model interpretability while improving yield estimation accuracy， offering a methodological reference for county-level maize yield monitoring and agricultural production management in the arid and semi-arid Yellow River irrigation districts of northwestern China.Keywords：maize yield prediction;county-level;Yellow River irrigation district;causal forest;temporal convolutional network;temporal attention mechanism;interpretable modeling31|20|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：To improve the model capability of remote sensing retrieval of Leaf Area Index （LAI） under genetically complex backgrounds and to dissect the growth stage-specific regulation of canopy dynamics by genetic variation， this study selected 800 genetically diverse maize accessions as study objects. UAV multispectral and RGB imagery were synchronously acquired at the jointing， large bell， and tasseling-silking stages， and LAI and plant height were measured in 72 field plots. Based on the imagery， nineteen canopy spectral vegetation indices and five canopy morphological-structural parameters were extracted to construct a high-dimensional feature set. Five feature selection strategies， namely Pearson correlation， Variance Inflation Factor， Mutual Information （MI）， Recursive Feature Elimination， and model-embedded importance assessment， were compared. Combined with Bayesian optimization algorithm for hyperparameter determination， XGBoost regression models were constructed and optimized. Based on the optimal model， spatiotemporal distribution maps of LAI for all plots were generated through inversion. SHAP （SHapley Additive exPlanations） analysis dissected feature contributions and mechanistic influences on model predictions. Finally， Kruskal-Wallis test and effect sizes （ε2） quantified the significance of LAI differences among genetic background groups. The results show that： 1） MI could effectively capture nonlinear dependencies and eliminate redundancy， with the five critical features identified constituting a robust feature combination adaptive to genetic variation. The resultant XGBoost model achieves superior predictive accuracy （RMSE＝0.302， R²＝0.722）； 2） Spatiotemporal mapping reveals a dynamic convergence in LAI variation throughout ontogeny. The mean LAI continued to increase from jointing to tasseling and silking， while the coefficient of variation declined markedly from 57.13% to 15.90%， indicating enhanced canopy uniformity with crop development. 3） Moreover， the influence of genetic background on LAI intensified significantly across developmental stages. The effect was negligible at the jointing stage （ε²＝0.008， P＝0.095）， it was weak at the pre-tasseling stage （ε²＝0.020， P<0.001）， and it reached moderate magnitude at the tasseling-silking stage （ε²＝0.138， P<0.001）. In summary， the phenotyping framework proposed in this study， which integrates multi-source remote sensing features with interpretable machine learning， successfully identified the synergistic response patterns of canopy spectral and morphological traits， providing a reliable technical pathway for precise LAI retrieval in genetically diverse populations.Keywords：UAV;maize;LAI;XGBoost;genetic background44|23|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：To clarify the important indicators affecting oat lodging and identify superior oat germplasm resources with lodging resistance， 1 018 oat germplasm resources were used as experimental materials. A field experiment was conducted to measure 19 indicators including panicle traits （main panicle length， panicle height， panicle height coefficient， panicle fresh weight， number of spikelets per main panicle， panicle branch number， grain number per main panicle， grain weight per main panicle， and 1 000-grain weight）， stem morphology traits （plant height， single-stem fresh weight， center-of-gravity height， first and second internode lengths， stem diameter， and stem wall thickness）， and mechanical traits （stem bending resistance， puncture strength， and lodging index）. Principal component analysis， cluster analysis， and comprehensive evaluation were adopted to evaluate lodging resistance. Random forest and structural equation models were further applied to analyze the lodging resistance and key indicator pathways of the tested oat germplasm resources. The results showed that the 1 018 oat germplasm resources exhibited abundant phenotypic variation. Plant height was extremely significantly positively correlated with panicle height， single-stem fresh weight， and center-of-gravity height （P<0.01）， while lodging index was extremely significantly negatively correlated with stem bending resistance and stem puncture strength （P<0.01）. Principal component analysis showed that the eigenvalues of the first seven principal components were all greater than 1.045， with a cumulative contribution rate of 76.128%. The first principal component was associated with plant morphology， the second principal component was associated with panicle traits， and the third， fourth， and seventh principal components were associated with stem mechanical properties. Cluster analysis divided the 1 018 oat germplasm resources into five groups. Group Ⅳ was characterized by low center-of-gravity height， short internodes， and high stem mechanical strength， and six lodging-resistant germplasm resources were identified from this group， namely ‘ND2116’ ‘ND2633’ ‘ND2208’ ‘ND139’ ‘ND2581’ ‘ND1595’. In the random forest model， stem bending resistance showed the highest relative importance， approximately 77%， and the model explained 92% of the variation. The structural equation model showed that plant height and center-of-gravity height had positive effects on lodging index， with standardized path coefficients of 0.368 and 0.981， respectively， whereas stem mechanical indicators had a negative effect on lodging index， with a standardized path coefficient of -0.73. The model explained 78.6% of the variation in lodging index. In conclusion， plant height， center-of-gravity height， and stem bending resistance were the key indicators affecting oat lodging. Lower plant height and center-of-gravity height， together with higher stem bending resistance， can serve as important criteria for screening lodging-resistant oat germplasm. The 1 018 oat germplasm resources were divided into five groups by cluster analysis， and the six lodging-resistant germplasm resources identified from group Ⅳ can be used for improving lodging resistance in oat.Keywords：oat;germplasm resources;lodging resistance;comprehensive evaluation48|24|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：The unclearly understood mechanisms of drop breakage and collision acceleration response in millet root stubble-membrane-soil composites have led to a lack of theoretical basis for the structural design and operational parameter regulation of residual film recycling equipment. To address this， a systematic indoor bench test was conducted. Using field-collected composites as test specimens， moisture content， hardness， mass， drop height， and screen inclination were selected as experimental factors， with soil removal rate as the response variable. An orthogonal experimental design was first employed to identify the dominant influencing factors， followed by single-factor experiments to investigate the variations in soil removal rate and peak value of collision acceleration with each significant factor， and to elucidate the underlying drop breakage mechanism. The orthogonal results show that the significance of the factors affecting soil removal rate， in descending order， is： composite moisture content， drop height， composite hardness， interaction between moisture content and mass， screen sieve inclination， interaction between moisture content and hardness， and composite mass. All factors except composite mass were found to be highly significant （P<0.01）. The results of single-factor experiments further indicate that the soil removal rate decreases with increasing moisture content and hardness of the composite， first increases and then decreases with increasing composite mass， and increases with increasing drop height and sieve inclination； the peak collision acceleration exhibits a monotonically increasing trend with the increase of all five factors. This study fills a research gap in the drop breakage and kinematic characteristics of millet root stubble-membrane-soil composites， and provides both theoretical and data support for the structural optimization and reasonable operational parameter matching of millet residual film recycling machines.Keywords：composite;breaking characteristics;peak value of collision acceleration;soil removal rate;residual film recycling38|29|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：Vitis vinifera L. is the most economically valuable and widely cultivated grape species globally. However， biotic and abiotic stresses seriously restrict its growth， development， yield， and fruit quality. Breeding new varieties with high quality and stress resistance is an effective approach to fundamentally address these cultivation constraints. The study reviews intraspecific recurrent selection methods of Vitis vinifera L. based on a review of recent domestic and international literature on high-quality and stress-resistant grape breeding， combined with the relevant research findings of our research team. The results show that： 1） The stress resistance of V. vinifera is controlled by minor-polygenes， which can be stably inherited to hybrid offspring.2） Recurrent selection can pyramid minor resistance polygenes from multiple parents into a single genotype through gene substitution， thereby significantly enhancing the stress resistance of new germplasm. 3） Intraspecific recurrent selection within V. vinifera does not rely on wild grape species as parents.On the premise of ensuring the fruit quality of new varieties， it can achieve effective breeding of grapes with high quality and comprehensive stress resistance through the substitution and accumulation of minor resistance polygenes. In summary， intraspecific recurrent selection within V. vinifera can effectively realize the breeding of grapes with high quality and comprehensive stress resistance， and this study opens up a new pathway for high-quality comprehensive resistance breeding of grapes.Keywords：Vitis vinifera L.;recurrent selection;high-quality and stress-resistant breeding;breeding method;minor-polygenes27|16|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：Aiming at the problems of poor road transport clearance， difficult headland turning， and low adaptability to complex terrain of wide-width cotton film-laying seed drills， a horizontally folding wide-width film-laying seed drill was designed in this study. The overall structure and working principle were described， and theoretical analyses were carried out on key components such as the folding frame， lifting mechanism， parallelogram profiling mechanism， and hydraulic system. The main dimensions of each component and the force variation pattern of the connecting rods during lifting were determined. A rigid-flexible coupled model of the whole machine was established using ANSYS-ADAMS， and dynamic simulation of the lifting process was conducted to obtain the location of the maximum stress point on the folding frame and the deformation curves of the lifting device. Stress tests on key positions during the dynamic folding process were performed with a multi-channel signal conditioning system. The correlation coefficients between the fitted simulation curves and the measured curves at all test points were above 0.88， and both the residual standard deviations and the amplitude errors remained within 15%. Field trial results showed： a void rate of 2.4%， a single-seed rate of 96%， a seed spacing qualification rate of 97.1%， a seed damage rate of 0.57%， a qualified rate of sowing depth under film of 95.4%， a qualified rate of soil cover thickness at film edge of 96.22%， a qualified rate of soil cover width at film edge of 97.32%， a longitudinal elongation of the plastic film of 5.11%， a qualified rate of width of the daylighting surface of 91.8%， a flatness of the daylighting surface of 98.8%， and a longitudinal elongation of the drip tape of 0.76%. All operation indicators met the requirements of national and professional standards.Keywords：cotton;mulch-sowing machine;wide;horizontal folding;coupling of stiffness and flexibility50|31|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
- Abstract：To continuously advance the high-quality development of high-standard farmland construction， this study takes the strategic interactions among multiple stakeholders involved in China’s high-standard farmland projects as its analytical focus. Drawing on evolutionary game theory， a tripartite dynamic model is constructed encompassing the government， construction contractors， and smallholder farmers. Through numerical simulations， the evolutionary trajectories and equilibrium states of strategy choices under varying conditions are comparatively analyzed， and elucidating the evolutionary logic of multi-agent strategic interactions and the sensitivity of key variables， thereby providing a theoretical basis for optimizing multi-agent coordination strategies in high-quality farmland development. The results show that： 1） Government behavior tends to converge toward stringent regulation， with penalty mechanisms exhibiting threshold effects that effectively curb contractors’ short-term profit-seeking incentives. 2） Smallholder participation exerts an endogenous disciplining effect on construction quality， moderate increases in subsidies and reductions in participation costs enhance farmers’ engagement， excessive incentives may impose fiscal burdens and diminish allocative efficiency. 3） Compared with single policy instruments， a composite strategy is more conducive to the rapid convergence of the evolutionary game system toward a high-quality steady state in high-standard farmland construction. In conclusion， the effectiveness of high-standard farmland construction hinges upon the strategic coordination among governments， construction contractors， and smallholder farmers. Featuring “policy incentives， dynamic regulation， and multi-actor collaborative governance”， this three-dimensional optimization framework promotes the alignment of diverse stakeholder interests and enhances the efficiency of resource allocation， thereby providing a robust institutional underpinning for the sustainable operation of high-standard farmland systems.Keywords：high-standard farmland construction;evolutionary game;multi-actor collaboration;numerical simulation;path construction66|34|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-07-07
