Does Regional Digital-Intelligent Transformation Mitigate the Quantity-Quality Dilemma in Agricultural Production? —Empirical Evidence from China's Tea Industry

Does Regional Digital-Intelligent Transformation Mitigate the Quantity-Quality Dilemma in Agricultural Production? —Empirical Evidence from China's Tea Industry | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Does Regional Digital-Intelligent Transformation Mitigate the Quantity-Quality Dilemma in Agricultural Production? —Empirical Evidence from China's Tea Industry Lu Jing, Xu Yao, Lin Dongkai, Wu Wulin This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9428414/v2 This work is licensed under a CC BY 4.0 License Status: Posted Version 2 posted You are reading this latest preprint version Show more versions Abstract Agricultural production constitutes the foundation of national economies and livelihoods, and countries worldwide attach considerable importance to the efficiency, quality, and scale of agricultural operations. Nevertheless, an inherent dilemma has long persisted between high-quality agricultural output relying on intensive inputs and high-yield production driven by cost reduction and scale expansion. This trade-off has become increasingly difficult to reconcile amid the continuous outflow of agricultural labor globally. As a core transformative force advancing Agriculture 4.0, the regional diffusion and application of digital-intelligent technologies raise a critical research question: Can Regional Digital-Intelligent Transformation (RDIT) alleviate this long-standing dilemma? This study uses the value of regional public tea brands to measure the quantity-quality equilibrium level of agricultural production. Following the principle of organic integration grounded in a holistic indicator system rather than simple numerical aggregation, we construct the Regional Digital-Intelligent Transformation Index. Using panel data on China's tea industry from 2012 to 2023, this paper empirically examines the impact of RDIT on the Quantity-Quality Equilibrium Level of Tea Production (QQEL_TP). The main findings are as follows. First, RDIT significantly promotes QQEL_TP. After controlling for individual fixed effects, time fixed effects, and other confounding factors, a 1% increase in RDIT is associated with a statistically significant 0.432% rise in QQEL_TP. Second, the core conclusion remains robust after addressing endogeneity via the instrumental variable method, excluding observations from special years, and adjusting sample specifications. Third, RDIT enhances QQEL_TP by fostering the development of Taobao villages (TBV). Fourth, the positive impact of RDIT on QQEL_TP is negatively moderated by government investment in science and technology (GI_ST). Agricultural Economics & Policy Quantity-quality equilibrium Agricultural production Value of the regional public tea brand Digital-intelligent transformation Tea industry Taobao villages Government expenditure on science and technology Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction During periods of material scarcity and poverty, people yearn for adequate food and clothing, and expanding production volume has consistently been the primary objective of agricultural endeavors. As human society progresses, numerous countries have transcended poverty, and there is now a heightened emphasis on enhancing the quality of life. Consequently, improving the quality of agricultural products has emerged as a pivotal production goal for many agricultural powerhouses. Augmenting quantity is tantamount to boosting yield, whereas enhancing quality encompasses not only the refinement of sensory attributes such as color, aroma, taste, and form but also entails stricter food safety regulations and more meticulous initial processing standards. Given constant mechanical efficiency, escalating the application of fertilizers and pesticides, curtailing processing and management durations, and compromising on the quality standards of finished products can substantially elevate yields, albeit at the expense of quality. When rural labor resources are abundant or even in surplus, labor costs remain low and may not even be factored into the cost calculations by smallholder farmers. By deploying substantial labor in precision farming, these farmers can not only diminish the reliance on fertilizers and pesticides but also uphold stringent operational standards across all stages. Under such circumstances, the dichotomy between quality and quantity in agricultural production is not pronounced. However, with the advent of industrialization, a growing number of rural laborers migrate to urban areas, driving up labor costs. In the absence of sustained precision farming practices and adequate capital to invest in advanced machinery, smallholder farmers are more inclined to adopt production strategies that prioritize quantity over quality, such as indiscriminate use of fertilizers and pesticides, streamlining processing and management procedures, and relaxing quality standards for finished products. Diminished quality leads to lower market acceptance and reduced profitability for producers, perpetuating a vicious cycle wherein producers increasingly favor quantity at the expense of quality. This cycle persists until profits turn negative, prompting an exodus of agricultural labor. Nevertheless, with technological advancements, mechanical substitution can partially offset labor shortages, enabling large-scale agricultural producers to reconcile the conflicting objectives of quality and quantity. In the early 21st century, the improvement in mechanization-driven labor substitution has been insufficient to offset the severe shortages of agricultural labor, leading to increasingly extensive agricultural production practices (Li et al. 2002 ) Consequently, quality-related incidents such as excessive pesticide residue limits and the deterioration of flavor characteristics of agricultural products have been frequently documented in public reports (Wei 2003 ; Wang. 2005; Li 2007 ; Fang and Zhang 2008 ). Brand-oriented development facilitates differentiated competition for agricultural entities that adopt quality-first or balanced quantity-quality production strategies, drawing a clear demarcation from opportunistic producers obsessed with pure output expansion. Nevertheless, most major agricultural economies worldwide belong to developing regions, where scattered smallholder households dominate agricultural production (Taherzadeh et al.2026). These small-scale producers lack sufficient intellectual capacity and financial resources to support systematic brand development. To break the vicious cycle of pursuing high yields at the expense of quality, the Chinese government has vigorously promoted the development of regional public agricultural product brands. It guides local agricultural operators to coordinate via technology sharing, unified joint quality control, and industrial collaborative governance, thereby collectively upholding the stable quality of agricultural products (Yang et al. 2020 ). The establishment of regional public brands has transformed the decision-making dynamics concerning agricultural production quantity and quality, transitioning from a zero-sum game constrained by individual fixed resources to a positive-sum game that seeks collective gains through coordination. The value of regional public brands for agricultural products, a common metric for assessing the effectiveness of such brand construction, represents a comprehensive value proposition centered on quality, augmented by considerations of effective scale and market realization. This value effectively captures the overall outcome of the strategic interplay between quantity expansion and quality enhancement in local agricultural production, reflecting the Quantity-Quality Equilibrium Level of agricultural production. The mathematical essence of resolving the dilemma in agricultural production quantity and quality decisions lies in enhancing this equilibrium level. Based on this premise, this paper proposes utilizing the value of regional public brands for agricultural products as a proxy for the Quantity-Quality Equilibrium Level of agricultural production, thereby indicating the extent to which the quantity-quality dilemma in agricultural production has been mitigated in the region. As a pivotal transformative force in Agriculture 4.0, digital and intelligent technologies have been empirically validated by numerous natural science experiments to enhance agricultural production efficiency. Nevertheless, within the realm of economics, the effectiveness of their dissemination and application in alleviating the quantity-quality dilemma in agricultural production remains to be substantiated. The tea industry belongs to labor-intensive and high-value agriculture. Compared with conventional agricultural commodities such as corn, wheat, and rice, tea production entails more intricate operational links, stringent standardized protocols, and higher thresholds for practical experience and professional skills. Even sensory quality assessment faces evident technical barriers, leading to pronounced information asymmetry and a more severe quantity-quality predicament (Dong et al. 2010 ; Yang and Hu 2011 ; Dong et al. 2014 ). Adopting panel data of China's tea industry spanning 2012 to 2023, this paper empirically investigates how Regional Digital-Intelligent Transformation (RDIT) affects the Quantity-Quality Equilibrium Level (QQEL_TP) in agricultural production. This study delivers four major marginal contributions to the literature. First, targeting the tea industry as the specific research context, it clarifies the theoretical logic whereby RDIT improves QQEL_TP, and further unpacks the internal driving mechanism mediated by Taobao villages development (TBV), as well as the external contextual influence moderated by government investment in science and technology (GI_ST). Second, the value of the regional public tea brand is introduced as a rigorous proxy indicator to measure QQEL_TP in agricultural production. Third, defining Digital-Intelligent Transformation as an organic integration of digitalization and intellectualization rather than simplistic numerical aggregation, we develop the Regional Digital and Intelligent Transformation Index grounded in holistic systematic logic for indicator construction. Fourth, at the macro level, this research quantitatively estimates the baseline influence magnitude of RDIT on QQEL_TP of tea production, while systematically verifying its underlying mediating pathways and moderating effects. Literature review Due to difficulties in acquiring data on measurable proxy variables, empirical economic research directly targeting the Quantity-Quality Equilibrium Level in agricultural production is relatively limited. Early scholarly attempts to mitigate the inherent quantity-quality predicament in agriculture generally adopted a fragmented analytical perspective, independently investigating the driving mechanisms for improving either agricultural output volume or product quality separately. By contrast, recent literature has mainly focused on industrial efficiency optimization that simultaneously boosts both quantity and quality performance. Research concerning Digital-Intelligent Transformation belongs to this latest research stream. Although ample economic empirical analyses have been conducted for agriculture as a whole, relevant studies specific to the tea industry are predominantly confined to experimental research in natural science and engineering fields. From a methodological perspective, evaluation frameworks for quantifying Digital-Intelligent Transformation have matured progressively as academics deepen their comprehension of digital-intelligent connotations. Even so, few construction paradigms effectively embody the organic integration of digitalization and intelligentization grounded in holistic logical reasoning, instead resorting merely to simplistic numerical combination strategies. Owing to the scarcity of studies that directly concentrate on agricultural quantity-quality equilibrium issues, the academic community still lacks adequate systematic assessments and measurement investigations of the Quantity-Quality Equilibrium Level in agriculture. Digital-Intelligent Transformation and agricultural production Extant studies generally acknowledge that Digital-Intelligent Transformation is far more than a mere upgrade of technical tools; it serves as the logical starting point and inevitable path to cultivate new quality productive forces in agriculture. Through systematically reshaping laborers, means of labor, and objects of labor, it drives agricultural production to shift from traditional experience-dominated patterns to modern data-driven operations (Sun et al. 2024). At the micro level, Digital-Intelligent Transformation significantly elevates the productivity of listed agricultural enterprises by improving innovation efficiency, diversifying supply chain structures, easing financing constraints, and accelerating human capital accumulation, while broadening their innovation boundaries simultaneously (Su et al.2026; Dong et al.2025). At the macro level, the Digital-Intelligent Transformation of agricultural producer services covering financial support, scientific and information services, and commercial circulation services can prominently boost agricultural green total factor productivity (Liu 2024 ). It also enhances the resilience of agricultural industrial supply chains by facilitating the integrated development of primary, secondary, and tertiary industries in rural areas (Liu 2025 ). In specific agricultural practices, embedding digital and intelligent technologies into pre-production and in-production segments allows real-time collection of crop growth and environmental information, alongside unmanned farming machinery executing operational decisions released by data centers. This achieves scientific regulation and standardized documentation of water, fertilizer, and pesticide application as well as farming management practices. It effectively cuts the input costs of general production factors whilst ensuring high quality, high yield, efficiency, and safety of agricultural products. Large-scale field trials verify that digitally and intelligently upgraded farmlands reduce labor costs by around 50%, pesticide usage by 30%, and fertilizer consumption by 15%-25%. Relative to conventional irrigation, overall water conservation reaches 50%, and drip irrigation delivers an additional water-saving effect of 20%-30%. Correspondingly, the per-unit yield rises by 15%-25% and comprehensive economic efficiency increases by approximately 25% (Yu et al. 2024 ). In addition, the embedding of digital-intelligent technologies into post-production links greatly improves the supply capacity and operational efficiency of high-quality agricultural goods. The integration of deep learning algorithms and hyperspectral detection not only enhances the accuracy and efficiency of pesticide residue identification but also reduces inspection expenditures and sample attrition losses (Augustin and Kiliroor 2025 ). The collaborative application of blockchain, artificial intelligence, and the Internet of Things within agricultural supply chain management systems significantly lowers coordination and verification costs between producers, retailers, and wholesalers enabled by full lifecycle production data traceability (Paul et al. 2025 ). Meanwhile, reinforcement learning algorithms dynamically optimize temperature and humidity control throughout cold chain logistics, curbing the decay and damage ratio of agricultural commodities during transportation and warehousing (Dhal et al.2025). However, equipment-only upgrades of Digital-Intelligent Transformation cannot generate substantial improvements in agricultural economic performance (Nie et al. 2022 ). The investment and deployment of digital-intelligent facilities face prominent entry barriers. Sound institutional arrangements are essential to align the interests of multiple stakeholders, including governments, agricultural producers, investors, and specialized agricultural service firms. Positive institutional incentives combined with external competitive pressures can motivate all participants to mobilize resources synergistically and promote the practical and effective implementation of Digital-Intelligent Transformation across the whole agricultural production process (Xiao and Fu 2024 ). Digital-Intelligent Transformation and tea production Experimental research in natural sciences and engineering has fully confirmed that embedding digital and intelligent technologies into every segment of the tea industrial chain can significantly enhance operational accuracy and efficiency at each stage (Wei et al. 2024 ). In essence, achieving simultaneous improvements in quality and efficiency for the whole tea industrial chain through Digital-Intelligent Transformation is technically viable. In tea cultivation and field management, the combination of deep learning and remote sensing can assess the suitability of tea-growing regions with an accuracy of 94% (Wei and Zhou 2023 ). When integrated with hyperspectral imaging technology, it supports non-destructive monitoring of nitrogen accumulation in tea leaves (Hiroto et al. 2020 ) and the growth conditions of tea buds and roots (Li et al. 2024 ). Coupled with cloud computing services, smartphone-captured imagery enables real-time identification and early prediction of tea plant diseases (Lanjewar and Panchbhai 2023 ). If further connected to Internet of Things facilities, these technologies enable scientific planning of tea garden layouts, precise water and fertilizer management, and targeted pest and disease control, thereby enhancing operational efficiency while reducing labor costs associated with manual patrols. During the harvesting stage, advanced mechanical harvesting systems powered by deep learning have been developed. These systems integrate tea bud detection, picking coordinate positioning, and motion trajectory planning to support targeted harvesting of tender high-grade fresh tea shoots (Li et al. 2023 ). Robotic arms embedded with deep learning algorithms can rapidly classify fresh tea leaves at a maximum accuracy of 92%, resolving the uneven maturity problem commonly found in mechanically harvested raw tea materials (Zhang et al. 2023 ). Integrated with the Internet of Things framework, intelligent automated tea harvesting with real-time status transmission and dynamic quality assessment has been realized (Zhang and Li 2021 ), which helps alleviate the acute labor shortage during peak harvesting periods. Regarding processing and quality detection, the synergistic application of deep learning, spectral analysis, computer vision, and mobile intelligent terminals achieves non-destructive monitoring of in-process tea characteristics across four critical procedures: withering, fixation, fermentation, and drying (Mao et al. 2022 ; Yu et al. 2020 ; Sheng et al. 2023 ). This avoids inconsistent finished tea quality stemming from subjective empirical judgment bias in manual operations. Moreover, these integrated technologies realize rapid and accurate quantification of pesticide residue levels (Sun et al. 2022 ), tea adulteration rates (Zou et al. 2023 ), physicochemical attributes (Ren et al. 2024 ), and flavor compound concentrations in finished tea products (Luo et al. 2022 ), substantially elevating overall quality inspection efficiency. For finished tea quality grading and supply chain traceability, the fusion of machine learning, radio frequency identification, the Internet of Things, and blockchain enables full lifecycle traceability of tea production and circulation data (Paul et al. 2021 ). It also empowers quality supervisors and ordinary consumers without professional tea evaluation expertise to accurately distinguish tea varieties (Jiang et al. 2023 ), flavor grades (Hu et al. 2023 ), and geographical provenances (Peng et al. 2023 ). The large-scale popularization of such technological systems will greatly reduce transaction costs and optimize the overall operational efficiency of the tea supply chain. Measurement of macroeconomic Digital-Intelligent Transformation level When Digital-Intelligent Transformation was still an emerging academic concept, Chinese researchers initially treated it as a simplistic juxtaposition of digitalization and intelligentization. Correspondingly, early evaluation indicator systems adopted a dual-dimensional structure that assessed digitalization and intelligentization independently. Luo and Chen ( 2022 ) first construct a digitalization index by selecting indicators from four dimensions: digital infrastructure development, digital industrial development, digital technological innovation, and corporate digital transformation. They then build an intelligentization index using indicators across three aspects: intelligent technologies, intelligent outcomes, and competitiveness and benefits. Finally, they combine the two indices into a single digital-intelligent index using principal component analysis (PCA). Zhang and Bai ( 2023 ) adopt a similar calculation logic but differ in the construction of digitalization and intelligentization evaluation systems and their final aggregation method. They measure digitalization from three dimensions: digital infrastructure, digital platforms, and digital users. For intelligentization, they use the logarithmic number of patents granted in the artificial intelligence, 5G, and blockchain industries. They ultimately measure the level of Digital-Intelligent Transformation as the product of the two indices. Han et al. ( 2025 ) employ a similar framework, with two key distinctions: they replace "digital platforms" with "digital technologies" in the digitalization evaluation system, and adopt the entropy weight method for the final index synthesis. Most subsequent studies follow Zhang and Bai's framework, and only revised partial sub-indicators and final aggregation approaches (Tang et al. 2025 ; Zhou et al. 2025 ). Notably, Liu et al. ( 2024 ) incorporate more sophisticated yet economically more convincing indicators—including the digital financial inclusion index and industrial robot penetration (Acemoglu and Restrepo 2020 )—into their evaluation framework. Building on this work, Liu et al.(2025) further clarify the framework’s logic and expand it into a well-structured three-level indicator system. Furthermore, Lyu et al. ( 2024 ) draw on the strengths of prior studies by introducing Python web scraping and word frequency analysis, widely used in micro-level research, to develop a more comprehensive and precise regional digital-intelligent assessment system. As new-generation information technologies have become increasingly pervasive, researchers have gained a more sophisticated understanding of digital-intelligent transformation. Instead of treating it as a mere combination of digitization and intelligence, they now conceptualize it as a novel system arising from their organic integration. Recently, several scholars have transcended the two-dimensional framework by reconstructing the evaluation system with four first-level indicators: digital-intelligent infrastructure, digital-intelligent technology, digital-intelligent application, and digital-intelligent benefits. Using the entropy method, they measured the level of digital-intelligent transformation across prefecture-level cities (Yuan et al. 2025 ). Measurement of the Quantity-Quality Equilibrium Level in agricultural production In empirical economic studies, the quantity of agricultural products is generally measured by output and planting area, whereas the measurement of quality varies considerably depending on research perspectives and conceptual understandings of quality. At the micro level, researchers either directly measure product quality based on the sensory attributes of agricultural products within specific sectors (Nie et al. 2021 ), or indirectly infer quality through survey -based consumer satisfaction statistics (Zhang and Wang 2025 ). From a macro-level perspective, when quality is defined as product safety and compliance with pesticide residue limits, it is commonly proxied by indicators such as the pass rate of random quality inspections, the quantity or share of organic agricultural products, and the number or proportion of registered geographical indication products (You and Wan 2022 ; George et al. 2019 ). When quality is interpreted as systematic safety, nutrition, and premium characteristics guaranteed throughout all supply-side links, scholars establish standardized evaluation indices covering cultivation, processing, circulation, and marketing to indirectly characterize overall agricultural quality (Jing 2021 ). When quality is treated as a latent variable reflecting demand shifts, it is estimated using the residual term derived from inverse demand functions (Wei et al. 2024 ; Wang 2024 ). For studies that encompass all the above dimensions of quality, the total number of regional public agricultural brands is adopted as the proxy measure (Wang et al. 2024 ). Recently, one study has directly focused on the balance between agricultural product quantity and quality, using the coupling degree, coordination degree, and comprehensive evaluation index of the two as proxy variables (Shao et al. 2024 ). Based on production-side data, this method effectively captures the quantity-quality equilibrium of agricultural products. However, production decisions typically fluctuate between surplus and shortage over time. Such measures only reflect producers' autonomous choices and fail to adequately capture the realized quantity-quality outcomes shaped by market competition. Using circulation-side data that connects production and consumption, this paper proposes regional agricultural product brand value as a measure of the equilibrium level of quantity-quality decisions. Although extant literature has thoroughly explored the roles of Digital-Intelligent Transformation in boosting agricultural output, efficiency, and income, theoretical analyses and empirical studies on how such transformation mitigates the conflict between quantitative growth and quality improvement of agricultural products remain absent. Taking China's tea industry—a typical sector plagued by long-standing quantity-quality coordination dilemmas at the production stage—as an example, this paper investigates the impacts of Regional Digital-Intelligent Transformation (RDIT) on the Quantity-Quality Equilibrium Level of Tea Production (QQEL_TP) from an economic perspective. It aims to offer empirical evidence for policy design addressing the quantity-quality dilemma in agricultural production, while contributing to theoretical development and methodological innovation. Theoretical analysis and research hypotheses The mathematical essence of the quality-quantity dilemma in agricultural production lies in how to maximize the total income of products by balancing the quantity and quality of production process inputs. Over a certain period, after undergoing sufficient market competition and receiving consumer feedback, agricultural producers will ultimately determine the most appropriate proportion of production factor inputs between quantity expansion and quality improvement to maximize total income. For a region, at this juncture, all agricultural producers' inputs in terms of quantity and quality factors reach a Nash equilibrium, which can be termed the Quantity-Quality Equilibrium in agricultural production. The better the decision-making coordination between agricultural producers regarding quantity expansion and quality improvement, the higher the level of Quantity-Quality Equilibrium in agricultural production. For a given technological level and a predominance of small-scale, decentralized farming, labor becomes the decisive factor shaping the Quantity-Quality Equilibrium Level in agricultural production. The more abundant the labor supply, the more effectively it can offset shortages in other production factors, thereby allowing coordinated expansion of output and improvement of quality even under the dual constraints of land endowment and capital investment. However, when labor outmigration results in scarcity, coordinated expansion of output and improvement of quality become unachievable. The relationship between quantity and quality in agricultural production thus shifts to a zero-sum game, which may in the long run degenerate into a negative-sum game featuring sacrificing quality for quantity. Against the overriding trend of advancing industrialization and urbanization, agricultural labor outmigration has become irreversible, rendering technological progress the pivotal approach to mitigating the quantity-quality dilemma in agricultural production. Advancements in traditional agricultural technology are mainly characterized by the independent adoption and direct application of individual production factors or pieces of technical equipment, with their systematic integration effect remaining relatively weak. Limited combinations of such factors are only observed among large-scale producers. By contrast, progress in digital-intelligent technology is distinguished by its move beyond reliance on a single device. It instead relies on public carriers such as 5G networks, cloud computing platforms, and intelligent control centers to build an integrated technical system. This model uses data to penetrate all links of the production process, enables cross-entity collaboration through platformization, and achieves the systematic integration of factors and procedures before terminal application. As a result, it can first release the systematic integration effect, promote the efficient use of technical equipment, and thereby reduce the application threshold for smallholder farmers. The large fixed investments required to develop public digital-intelligent infrastructure are mainly borne by governments, platform enterprises, and social capital. Smallholders can access digital-intelligent systems through lightweight terminals such as smartphones, thus sharing the benefits generated by integration effects. RDIT has therefore shown considerable potential to foster high-quality agricultural development. As a labor-intensive and high-value-added agricultural sector, the tea industry imposes strict demands on both the quantity and quality of labor input for premium tea production, resulting in a particularly acute quantity-quality dilemma. This study employs the tea industry as a case to analyze the specific logic and mechanisms through which RDIT contributes to an improved Quantity-Quality Equilibrium Level in agricultural production. High-quality tea production relies on considerable labor input and sophisticated craftsmanship, as tea leaves must be harvested and processed in a timely manner. Moreover, the cultivation cycle for skilled tea processors is lengthy, and the current stock of such talent is severely limited. During the peak tea-picking and processing season, shortages exist not only in ordinary tea pickers but also in skilled tea artisans, representing a structural labor imbalance. Although mixed cultivation of tea varieties with different maturity stages can alleviate the peak labor demand to some degree, the challenges of labor scarcity and rising labor costs in high-quality tea production remain pronounced nationwide. On the one hand, labor shortages directly push up production costs in the tea sector, forcing growers to loosen picking standards and deviate from optimal harvesting schedules. In addition, the absence of uniform post-harvest processing norms among smallholder farmers significantly impairs the consistency of the sensory quality of final tea products. On the other hand, labor shortages lead to insufficient routine management and maintenance of tea gardens, with some growers resorting to unregulated chemical applications, thereby creating potential food safety and quality risks. Existing mechanical harvesting and processing equipment cannot satisfy the strict quality requirements for premium and famous teas, pushing myopic smallholders toward quantity-oriented production strategies at the expense of quality. Given that the Chinese tea industry is dominated by smallholder operations, the persistent outflow of rural labor further intensifies the quantity-quality imbalance in tea production. Evidently, labor shortage represents the core constraint behind the quantity-quality dilemma in China's tea production system. RDIT not only alleviates labor shortages but also directly improves the rationality of tea farmers'production decisions, thereby enhancing the Quantity-Quality Equilibrium Level of Tea Production (QQEL_TP). First, RDIT promotes the upgrading and openness of public service platforms, lowers the costs and spatiotemporal constraints of skill training, strengthens the knowledge spillover effect, and improves the quality of the existing labor supply, ultimately raising product quality without compromising quantity. The refinement and accessibility of expert knowledge systems and disaster early-warning systems provide tea producers with access to advanced processing technologies and pest management strategies, thereby improving the quality of tea garden management and processing operations and consequently enhancing the quality of final tea products. Second, RDIT can expand the effective labor supply during the peak harvesting season by encouraging labor return and optimizing labor allocation efficiency, thus safeguarding both the quantity and quality of tea products. This transformation is often accompanied by the rise of the platform economy and the expansion of cross-regional and cross-industry cooperation networks. The platform economy creates additional employment opportunities that attract labor back or support part-time participation. The expansion of cooperation networks helps break down information barriers in the labor market, improves the efficiency of labor matching, and enables famous tea-producing areas to quickly connect with idle labor resources in surrounding regions, thereby easing labor shortages during the peak picking period. Both factors increase the labor available to the tea sector, ensure compliance with harvesting standards and timely processing, and thus stabilize the quality and supply scale of finished tea products. Third, RDIT can correct distortions in the market mechanism and reverse small-scale producers' quantity-over-quality bias through the accurate transmission of price signals. The construction of a comprehensive quality traceability system is a core part of RDIT. It digitizes and transmits information to consumers, including tea origin, producer, batch, and grade, allowing high-quality tea to be recognized and priced at a premium. This eliminates the market distortion caused by information asymmetry and prevents the phenomenon of bad money driving out good. The accurate transmission of price signals enhances the rationality of tea farmers'production decisions, encouraging them to shift from short-term opportunistic behavior that sacrifices quality for quantity to long-term production strategies that balance quantity and quality. Meanwhile, regional data integration enables the government to implement targeted brand certification and differentiated subsidies, strengthening market incentive and restraint mechanisms. This further improves the rationality of tea farmers' production decisions and ultimately elevates QQEL_TP. Synthesizing the foregoing discussions, this paper puts forward the core theoretical proposition as Hypothesis 1. H1 Regional Digital-Intelligent Transformation (RDIT) can significantly enhance the Quantity-Quality Equilibrium Level of Tea Production (QQEL_TP). The effective realization of the above three pathways requires tea producers to actively access platforms or systems fostered by RDIT. Taobao Villages (TBV), as rural e-commerce industrial clusters developed on the basis of the platform economy, are themselves typical outcomes of RDIT permeating rural areas. Their emergence and development can enhance tea producers' willingness to participate in such transformation from two perspectives, thereby improving the QQEL_TP. First, TBV provide tea producers with sales channels directly targeting national consumers, reducing supply chain links and retaining more industrial chain profits. This in turn increases rural household income (Zhang et al. 2024 ), expands agricultural capital accumulation (Zhou et al. 2021 ), and equips local farmers with greater financial capacity and motivation to adopt digital-intelligent platforms. Second, to cope with market competition, TBV must proactively integrate into RDIT and connect with platforms or systems conducive to improving production and marketing performance. In doing so, they effectively guide and drive tea producers to understand, familiarize themselves with, and utilize public service platforms and collaborative networks, thereby improving production conditions. Existing research indicates that the development of TBV can significantly influence the quality and variety of products in the village's dominant industries (Zhang et al. 2023 ). Furthermore, the development of TBV can be promoted by RDIT. During this transformation, the popularization of broadband networks, the construction of data centers, and the expansion of mobile payment services have benefited rural areas, providing digital-intelligent infrastructure support for the formation of TBV and enabling remote rural regions to access national and even global markets at low cost. Meanwhile, the diffusion and application of digital-intelligent technologies have reduced the costs of information searching and skills training, helping to improve farmers'digital literacy and stimulate their entrepreneurial vitality (Mei et al. 2020 ). This makes small-scale and decentralized rural e-commerce operations economically viable, thereby fostering the emergence of more Taobao Villages (Wu et al. 2020 ). Based on the above logical deduction, Hypothesis 2 is formally proposed. H2 Regional Digital-Intelligent Transformation (RDIT) improves the Quantity-Quality Equilibrium Level of Tea Production (QQEL_TP) by promoting Taobao Villages development (TBV). The promotional effect of RDIT on QQEL_TP is not independent, and its effectiveness hinges on regional capacities for local technology absorption and conversion. To some extent, Government investment in science and technology (GI_ST) shapes local capabilities for technology absorption and translation, thereby moderating the degree to which RDIT elevates QQEL_TP. On the one hand, by providing financial support for scientific research projects and targeted subsidies for equipment purchases, GI_ST directly lowers the upfront cost barriers for tea researchers, farmers, and enterprises to adopt cutting-edge facilities such as the Internet of Things, intelligent sensors, and smart agricultural machinery. This accelerates the widespread deployment of digital-intelligent equipment across tea production and strengthens stakeholders' capabilities in the application, secondary development, and scenario-based utilization of digital-intelligent technologies. Furthermore, increased GI_ST focused on digital-intelligent R&D and talent cultivation sends positive policy signals, boosts market confidence, and attracts social capital, high-skilled labor, and computing infrastructure to support RDIT. This, in turn, amplifies the positive effect of RDIT on enhancing QQEL_TP. On the other hand, the moderating effect of GI_ST is not uniformly positive. Excessive public financial investment may give rise to a typical "resource curse," turning the moderating effect negative. Some subsidy programs lack rigorous evaluation systems for actual technology application performance. Poorly designed project acceptance criteria or excessive administrative intervention in R&D activities tend to erode tea producers' incentives for independent exploration of efficient digital-intelligent solutions and autonomous innovation, slowing the diffusion of digital-intelligent technologies in tea production. Moreover, administration-led capital allocation is prone to information lags and resource misallocation. When fiscal subsidy priorities diverge from the real operational needs of the tea industry, formalism emerges. Resources flow into inefficient projects or obsolete traditional capacity, leaving digital-intelligent equipment underutilized. Such distortions crowd out funds that market forces would otherwise channel to promising digital-intelligent innovators, hindering the deep integration of digital-intelligent technologies with core production links and ultimately dampening the marginal contribution of RDIT to improving QQEL_TP. Overall, the sign and magnitude of GI_ST's moderating role in the RDIT–QQEL_TP nexus remain theoretically ambiguous and await rigorous empirical validation. Accordingly, Hypothesis 3 is proposed. H3 Government investment in science and technology (GI_ST) plays a significant moderating role in the influence of Regional Digital-Intelligent Transformation (RDIT) on the Quantity-Quality Equilibrium Level of Tea Production (QQEL_TP). Research design Model specification Following the approach of Li et al. ( 2024 ), this study constructs a two-way fixed effects model to test H1. This model can effectively control for omitted variable bias, individual heterogeneity, and time heterogeneity, thereby mitigating endogeneity concerns in panel data to a certain extent. To alleviate heteroskedasticity and facilitate the elastic interpretation of regression coefficients, the natural logarithm is applied to all continuous variables. The specific specification is as follows: ln ln ln (1) ln QQE_TP it = α 0 + α 1 ln RDIT it + α j ln Z it + δ i + µ t + ε it (1) Where QQE_TP it refers to the Quantity-Quality Equilibrium Level of Tea Production of province i in year t ; RDIT it denotes the Regional Digital-Intelligent Transformation Index for province i in year t ; Z it is a vector of control variables. δ i , µ t , and ε it represent individual fixed effects, time fixed effects, and the random error term, respectively. Based on Eq. (1), we further develop a mediating effect model to test H2, as below: ln TBV it = β 0 + β 1 ln RDIT it + β j ln Z it + δ i + µ t + ε it (2) ln QQE_TP it = γ 0 + γ 1 ln RDIT it + γ 2 ln TBV it + γ j Z it + δ i + µ t + ε it (3) Where TBV it represents the level of Taobao Villages development in province i in year t . Eq. (2) is first adopted to estimate the impact of RDIT on TBV. Subsequently, Eq. (3) is used to assess how RDIT influences QQEL_TP after incorporating TBV as a controlled variable. The total mediating effect is expressed as β 1 × γ 2 . Considering that the sampling distribution of the mediating effect estimator may fail to follow a normal distribution, this paper applies the Bootstrap approach to conduct repeated resampling trials and determines statistical significance according to bias-corrected confidence intervals, thereby improving the overall robustness of the mechanism examination results. To test H3, a moderating effect model is further formulated on the basis of Eq. (1), as below: ln ln (ln×ln) (4) ln QQE_TP it = η 0 + η 1 RDIT it + η 2 ln GI_ST it + η 3 (ln RDIT it ×ln GI_ST it ) +η j Z it + δ i + µ t + ε it (4) Where GI_ST it represents government investment in science and technology of province i in year t , and the coefficient η 3 of the interaction term characterizes the direction and intensity of the moderating effect. Given that the meaningful variation in both mediating and moderating variables primarily arises from cross-regional disparities and gradual temporal evolution, controlling for both individual and time fixed effects would absorb much of their core variation, leading to insufficient identifying information for mechanism analysis and overly conservative estimates. Accordingly, four model specifications are adopted to test the mediating and moderating effects: no fixed effects, time fixed effects only, provincial fixed effects only, and both time and provincial fixed effects. The results are mainly used for exploratory analysis and preliminary validation of the proposed mechanism channels. Variable selection and data sources Extant studies identify 2012 as the starting point of China's digital agriculture development (Zhu 2023 ). In April 2012, the China Tea E-commerce Alliance was established in Xinchang, Zhejiang, marking a new era of self-regulated and standardized development for China's tea e-commerce industry (Yuan 2012 ). In June 2012, the Chinese government rolled out the "Broadband China" strategy to boost digital development and prioritize next-generation information infrastructure (The State Council 2012). Taken together, the profound influence of Digital-Intelligent Transformation on agricultural production, the standardized operational maturity of Taobao Villages in major tea-growing regions, and the targeted tilt of government technological funding toward digital-intelligent sectors all converged around the year 2012. Accordingly, this study sets 2012 as the starting point of its research timeline. Owing to divergent release frequencies and update cycles across datasets, the latest consistent and complete observations are available up to 2023 after rigorous data integration. The empirical analysis thus employs a panel dataset covering the period 2012–2023. Dependent variable QQEL_TP is measured by the value of regional public tea brands. As theorized earlier, better alignment between production scale and quality enhances industrial reputation and is ultimately reflected in brand value, which therefore serves as a valid proxy for QQEL_TP. Brand value data are retrieved from the annual Evaluation Report on the Value of China's Regional Public Tea Brands, published systematically by the China Agricultural Brand Research Center. This evaluation represents the only authoritative third-party assessment widely recognized in China's tea industry (Huang and Sun 2023 ). The original dataset covers major tea-producing provinces in China, including 215 brands and 1,456 raw observations. Brand-level data are aggregated to the provincial level by value-weighted averaging, resulting in a balanced panel of 15 provinces and 180 valid observations. Core explanatory variable RDIT is measured by the regional digital-intelligent transformation index constructed by the authors. Guided by the connotation of Digital-Intelligent Transformation, we build a comprehensive evaluation system consisting of 3 primary indicators and 11 secondary indicators. From a technological perspective, Digital-Intelligent Transformation is not merely a simple combination of digitization and intelligence. Rather, it constitutes an evolving state of information technology that progresses from basic digitization toward full-fledged intelligence, wherein digitization and intelligence become organically integrated and mutually reinforcing. From an economic standpoint, Digital-Intelligent Transformation encompasses the full spectrum of advances in and applications of next-generation information technologies. Building on digitization, it leverages increasingly sophisticated computing power and algorithms to collect and integrate data from across all dimensions of socioeconomic activities. By fully unlocking data value, it continuously enhances productivity and reshapes production modes accordingly, ultimately propelling human society toward an intelligent state in which all economic activities enable autonomous decision-making, automatic execution, and self-optimization. Accordingly, the regional digital-intelligent transformation index is constructed from three dimensions: public foundation, application level, and economic benefits. First, the public foundation dimension incorporates intelligent elements beyond conventional digitalization, covering basic infrastructure, investment, human capital, as well as advanced computing facilities and R&D. Second, consistent with Marx's social reproduction theory, the application level is assessed across production, circulation, distribution, and consumption. Third, it captures the economic benefits generated by RDIT through both supply-side gains and demand satisfaction. The evaluation indicator system for RDIT is presented in Table 1. In selecting proxy variables, we follow standard literature practices but introduce two tailored adjustments for our research objectives. First, we use patent applications for new-generation information technology to measure technological R&D, rather than aggregating patents across digital technology, 5G, AI, and blockchain. With the rapid diffusion and innovation of information technology, these traditional indicators fail to adequately capture emerging digital-intelligent technologies. New-generation information technology more accurately represents core enabling technologies such as cloud computing, IoT, VR, AR, and the BeiDou Navigation System, which are often not covered in prior studies. Second, we measure supply-side gains using gross profit in the manufacturing of computers, communications, and electronic devices, instead of revenue, operating profit, or total profit. While revenue reflects supply scale, it is highly volatile to market prices and demand shocks and cannot identify technological progress or cost efficiency. The digital industry features high R&D intensity, fast fixed-asset depreciation, and strong subsidy dependence, making operating and total profits prone to policy and accounting distortions. By contrast, gross profit better captures cost reductions and efficiency gains driven by technological innovation, providing a more reasonable measure of supply-side gains. Control variables Regional consumption level (RCL), agricultural mechanization level (AML), and logistics accessibility (LA) are selected as control variables. Based on Keynes' effective demand theory, a higher regional consumption level expands market demand for tea products, encouraging tea enterprises to expand production and optimize operations, thus positively promoting improvements in QQEL_TP. From the perspective of induced technological change theory, agricultural mechanization substitutes labor inputs, improves operational accuracy and efficiency, and supports coordinated growth in tea yield and quality. Grounded in new economic geography theory, efficient logistics networks lower transport costs, shorten circulation cycles, enhance flexible production scheduling, and further optimize quantity-quality decision-making performance. Specifically, RCL is measured as the share of total retail sales of consumer goods in regional GDP; AML is proxied by total agricultural machinery power; and LA is represented by total freight volume. All control variable data are obtained from provincial statistical yearbooks. Instrumental variable The number of new-generation information technology patent applications filed between 1985 and 1996 is chosen as the instrumental variable (IV). On the one hand, the historical patent stock reflects the evolution and penetration of foundational digital-intelligent technologies, which directly determines the future direction and scale of RDIT, thus satisfying the relevance condition. On the other hand, given the lengthy time span of more than two decades, early information technology patents exert negligible exogenous effects on current tea sector quantity-quality production decisions, strictly meeting the exogeneity requirement for valid instrumental variables. Mechanism variables Consistent with Hypothesis 2, Taobao Village development (TBV) is treated as the mediating variable. Limited by data availability, the number of Taobao Villages is used to measure its development, with raw data obtained from the Annual Research Report on China's Taobao Villages issued by the Alibaba Research Institute. In line with Hypothesis 3, government investment in science and technology (GI_ST) serves as the moderating variable, proxied by provincial government expenditure on science and technology, with relevant data collected from the National Bureau of Statistics of China. Descriptive statistics and data processing of variables Descriptive statistics of variables Descriptive statistics of all variables are shown in Table 2. Data processing First, all nominal variables are deflated to constant 2012 prices. Second, sporadic missing values are imputed using interpolation. Finally, all processed variables are log-transformed prior to regression analysis. Excel is used for preliminary data cleaning and preprocessing, while Stata is employed for baseline regressions, robustness checks, and endogeneity correction. The KMO statistic for the RDIT indicator system exceeds 0.8, indicating strong correlations among dimensions. PCA is thus used to construct the aggregate Regional Digital and Intelligent Transformation Index. In addition, the measurement of industrial robot penetration involves complex calculations, which are detailed below. First, using the ISIC Rev.4 manufacturing classification as a bridge, this study matches industrial robot categories from the International Federation of Robotics (IFR) (Müller and Christopher, 2025) to China's Industrial Classification for National Economic Activities (GB/T 4754 − 2017) ① . A concordance table is constructed covering 29 Chinese manufacturing sub-industries (C13–C41) ② and 13 corresponding IFR sectors. Second, we aggregate the cumulative stock of industrial robots at the industry level in China for 2012–2023 according to this concordance. Third, taking 2013 as the base year ③ and drawing on the frameworks of Acemoglu and Restrepo ( 2020 ), Wang and Dong ( 2023 ), and Zhang et al. ( 2025 ), we construct a province-year measure of manufacturing robot penetration. The specific formula is given below: $$\:\text{E}\text{Exposure}\text{\:−}\text{CH}\text{}\text{=}\sum\:_{\text{i}\in\text{I}}\frac{\text{em}{\text{p}}_{\text{ji}\text{,}\text{t}\text{=201}\text{3}}}{\text{em}{\text{p}}_{\text{j}\text{,}\text{t}\text{=201}\text{3}}}\frac{\text{M}{\text{R}}_{\text{it}}^{\text{CH}}}{{\text{L}}_{\text{i}\text{,}\text{t}\text{=201}\text{3}}^{\text{CH}}}\text{}\text{}\text{}\text{(5)}$$ Where \(\:{\text{MR}}_{\text{it}}^{\text{CH}}\) stands for the stock of industrial robots in China's industrial sector i in year t ; \(\:{\text{L}}_{\text{i}\text{,}\text{t}\text{=201}\text{3}}^{\text{CH}}\) ​denotes the national baseline employment of sector i in 2013. \(\:\frac{\text{em}{\text{p}}_{\text{ji}\text{,}\text{t}\text{=201}\text{3}}}{\text{em}{\text{p}}_{\text{j}\text{,}\text{t}\text{=201}\text{3}}}\) represents the share of sector i's employment in total employment of province j in the same base year, which captures the regional manufacturing employment structure. Data on China's baseline employment are obtained from the 2013 China Economic Census Yearbook . Results and discussion Spatiotemporal characteristics of RDIT and QQEL_TP To clearly depict the spatiotemporal evolutionary features of the RDIT and the QQEL_TP across major tea-producing provinces in China, this paper conducts analyses from both static and dynamic dimensions. MATLAB is employed to plot kernel density distribution curves for the period 2012–2023, and Stata is utilized to generate spatiotemporal evolution trend maps for the representative years of 2012, 2017, and 2023. A geographical location map (Fig. 1) is provided below to identify the actual spatial distribution of each sample province across China's territory. Temporal and spatial evolution of RDIT From a temporal perspective (Fig. 2), RDIT across China's major tea-producing provinces shows steady improvement and divergent development. Before 2014, the main peak was narrow and high, the secondary peak was indistinct, and both were concentrated in the low-level interval, indicating that the overall RDIT of the sample provinces was generally low. This is consistent with weak infrastructure in the early stage of Digital-Intelligent Transformation. Over time, both the main and secondary peaks shifted continuously to the right, with the height of the main peak declining and the secondary peak rising, suggesting that the clustering of samples at low digital and intelligent transformation levels gradually weakened, while the proportion of high-level regions increased steadily. After 2020, a third peak emerged, and all peaks continued to shift rightward, indicating that the digital and intelligent transformation level of the samples kept improving and regional divergence became more pronounced. Distinct high, medium, and low tiers have formed, reflecting that regional digital and intelligent transformation has entered a stage of normalized development, with the effects of policy promotion and technology diffusion gradually becoming evident. From a spatial perspective (Fig. 3), the regional digital and intelligent transformation (RDIT) of China's major tea-producing provinces exhibits a clear gradient distribution. In all sample years, Zhejiang, Jiangsu, and Guangdong consistently rank at the forefront, while Shandong and Fujian remain in the medium-high group. Although Guangxi and Yunnan have improved gradually over the sample period, they still fall into the low-RDIT category, reflecting the advantage of coastal locations in advancing Digital-Intelligent Transformation. Sichuan and Anhui show considerable temporal shifts in their gradient positions. Benefiting from technological spillovers from neighboring high-RDIT provinces, Anhui has steadily risen from the medium-low group in 2012 to the medium-high group. Sichuan has advanced remarkably from a low level in 2012 to the medium‑high group, and given the low RDIT performance of most surrounding provinces, this improvement can be largely attributed to targeted policy support. By contrast, Hubei and Hunan have experienced continuous declines, falling from medium-high to medium-low levels. Jiangxi, despite its proximity to high-performing provinces, has long remained trapped in the low-RDIT group. Overall, these patterns indicate that sustained and targeted regional policies are essential for achieving inclusive Digital-Intelligent Transformation. Spatiotemporal evolution of QQEL_TP From a temporal perspective (Fig. 4), the distribution of QQEL_TP across China's major tea-producing provinces shows a pattern of gradual improvement and increasing concentration. During 2012–2016, the main peak was narrow and high, concentrated in the low-value range, with two distinct secondary peaks in the medium- and high-value ranges. This suggests a highly dispersed tripartite pattern in quantity-quality equilibrium with an overall low level in the early period. Over time, the main peak declined, and the two secondary peaks converged and merged by 2017. Although the secondary peak rose moderately afterward, it shifted left and right intermittently, indicating that QQEL_TP gradually converged but improved slowly and unstably. This evolution reflects the complexity of provincial quantity-quality decisions. Initially, most provinces prioritized quantity, a few balanced quantity and quality, and only a small number emphasized quality primacy. With rising consumer quality awareness and stricter agricultural product quality policies, most provinces shifted toward balancing quantity and quality. Amid fiercer market competition, even quality-oriented provinces began to adopt a balanced strategy and develop mid-to-low-end products to sustain profits. The diffusion of digital-intelligent technologies has to some extent accelerated this transition. As illustrated in the spatial evolution map (Fig. 5), the distribution of QQEL_TP has transitioned from regional polarization to convergent development, reflecting strengthened spatial convergence and spillover effects among major tea-producing provinces. At the provincial level, Zhejiang and Fujian have consistently taken the lead. Notably, Guangdong, a neighboring province with China's highest GDP, has remained at a relatively low level. This suggests that unobservable factors beyond provincial economic development exert a significant influence on QQEL_TP, which may be closely related to the reputation of historical famous teas: 60% of China's top ten famous teas originate from Zhejiang and Fujian. Therefore, to rigorously examine whether RDIT can improve QQEL_TP, it is essential to control for provincial fixed effects to mitigate unobservable regional heterogeneity. Baseline regression results Baseline regression results for Hypothesis 1 are presented in Table 3. Columns (1)–(5) sequentially add controls and various fixed effects, with column (5) representing our preferred two-way fixed effects specification. Baseline results confirm that RDIT exerts a significant positive effect on QQEL_TP, with coefficient robustness strengthening as model specification improves. Even in the parsimonious specification without controls, the core coefficient remains positive and significant, indicating that RDIT inherently boosts QQEL_TP. Adding control variables further increases the magnitude of the core coefficient, implying that omitting factors such as regional consumption, agricultural mechanization, and logistics accessibility leads to downward bias in estimating RDIT's true impact. Among controls, RCL is significantly positive, consistent with demand-pull theory. Greater domestic demand improves supply–demand matching through Taobao Village–driven platforms, promoting quality upgrading and production efficiency. By contrast, AML and LA are statistically insignificant, likely reflecting tea's unique production characteristics: tea cultivation relies little on standard farm machinery and general logistics, so their marginal effects are overshadowed by digital-intelligent gains. Further controlling for provincial and time fixed effects substantially increases the core coefficient, which remains significant at the 5% level. This suggests that unobserved regional heterogeneity and aggregate time shocks distort baseline estimates, and removing such confounding factors enables cleaner identification of RDIT's net effect. Overall, these findings strongly support that RDIT significantly improves QQEL_TP. In our preferred two-way fixed-effects specification, a 1% increase in RDIT raises QQEL_TP by 0.432%. Endogeneity Test Limited by access to industry-specific panel data for tea, our empirical model may suffer from endogeneity concerns, including omitted unobservables and measurement errors arising from inconsistent statistical standards across databases. In addition, significant bidirectional causality exists between RDIT and QQEL_TP. On the one hand, higher RDIT maturity improves QQEL_TP; on the other, regions with better QQEL_TP tend to have greater incentives and capacity to invest in digital-intelligent infrastructure, further promoting RDIT. To address these endogeneity issues, we adopt an instrumental variable (IV) approach, with results reported in Table 4. In the first-stage regression, the coefficient of the instrumental variable is significant at the 1% level, with an F-statistic of 50.05, well above conventional critical values, thus ruling out weak instrument concerns. In the second-stage estimation, the coefficient of RDIT is 1.189 and significant at the 5% level, representing a substantially larger marginal effect relative to baseline estimates. The LM underidentification test rejects the null hypothesis at the 1% level, further confirming the statistical adequacy of the instrument. Two-stage estimation results validate the effectiveness of the instrumental variable design. The instrument is strongly correlated with RDIT and exogenous to the disturbance term, effectively mitigating estimation bias driven by reverse causality and omitted variables. After accounting for endogeneity, RDIT still exerts a significant positive influence on QQEL_TP, with an increased coefficient magnitude compared with baseline findings. This reinforces the causal inference that RDIT improves QQEL_TP, and indicates that baseline results tend to understate the actual promoting effect. Overall, although neglecting endogeneity leads to conservative coefficient estimates, the core conclusion that RDIT enhances QQEL_TP remains robust under a more rigorous causal identification framework. Robustness checks This study employs three systematic robustness strategies, with results presented in Table 5. First, considering that the COVID‑19 pandemic substantially disrupted macroeconomic conditions from 2020 to 2022, observations in these abnormal years are excluded. The coefficient of the core independent variable remains positive and significant, with magnitude and sign highly consistent with baseline estimates. The positive effect of RDIT on QQEL_TP remains robust even after removing approximately one-quarter of the sample, confirming that the baseline results are not driven by temporary macroeconomic shocks or extreme public health events. Second, to mitigate bias caused by extreme outliers in continuous variables, all continuous indicators are winsorized at the 5th and 95th percentiles. Re-estimation yields a significantly positive and stable coefficient for the core explanatory variable, indicating that the main findings are not affected by extreme observations. Third, the logarithmic transformation of the normalized core explanatory variable generates missing values where original entries are zero, reducing the effective sample to 179. To rule out estimation distortion caused by such data truncation, zero values are replaced with 0.001 before regression. The sign and significance of the RDIT coefficient remain unchanged, further verifying that the conclusions are robust to minor data adjustments and not dependent on specific data-processing rules. Mediating effect analysis To explore the mechanism through which RDIT affects QQEL_TP, the bootstrap method is employed to test the mediating role of TBV. As shown in Table 6, the indirect effect is significantly positive in the specification without fixed effects, supporting the mediating channel: RDIT improves QQEL_TP by fostering TBV development. This positive indirect effect remains robust after controlling for provincial fixed effects, indicating that the transmission path RDIT → TBV → QQEL_TP is stable even after accounting for unobserved regional heterogeneity. By contrast, the mediating effect becomes statistically insignificant when only time fixed effects or two-way fixed effects are included. This implies that the mediating role of TBV is sensitive to time-varying shocks arising from macro policy adjustments and market fluctuations. Overall, RDIT enhances QQEL_TP partially through the promotion of TBV. Further analysis of direct effects across specifications yields additional implications. In the baseline model without fixed effects, the direct effect of RDIT on QQEL_TP is significantly negative, whereas the total effect is insignificant. This suggests the existence of other unobserved mediating channels that may exert adverse influences on QQEL_TP. After controlling for time, provincial, or two-way fixed effects, both direct and total effects become insignificant. These results indicate that the relationship between RDIT and QQEL_TP is characterized by complex mechanisms and high sensitivity to time-varying heterogeneous shocks. Moderating effect analysis To examine the contextual boundary conditions under which RDIT affects QQEL_TP, a moderating effect model is estimated using GI_ST as the moderating variable. As shown in Table 7, the interaction term is significantly negative when both fixed effects are omitted or only time fixed effects are included (Columns 1 and 2). When provincial fixed effects or two-way fixed effects are introduced (Columns 3 and 4), the interaction term remains negative but loses statistical significance. These results indicate that GI_ST exerts a weakly negative moderating effect on the relationship between RDIT and QQEL_TP. The moderating effect diminishes substantially after accounting for unobserved provincial heterogeneity, suggesting that GI_ST is characterized by notable regional heterogeneity. In general, the positive impact of RDIT on QQEL_TP is weakened under high levels of GI_ST. This pattern can be explained by the insufficiently mature digital-intelligent application scenarios in the tea industry. Excessive fiscal investment and resource misalignment prevent public funding from being effectively converted into industry-specific technologies. Instead, the mismatch between advanced digital technologies and on-farm production needs creates an inhibitory effect. These findings also point to widespread resource redundancy and low technology conversion efficiency in the current allocation of public fiscal resources for science and technology. Conclusions, policy implications, and research limitations Conclusions Digital-Intelligent technologies have been extensively integrated into contemporary agricultural production systems. A large body of natural science experiments has confirmed that such technologies can simultaneously enhance both the output quantity and product quality in agricultural practices. However, constrained by difficulties in obtaining qualified measurable proxy data in economic empirical research, existing studies mostly examine the agricultural quantity-quality dilemma from fragmented and isolated perspectives. Few scholars have directly focused on the integrated quantity-quality equilibrium level, and whether RDIT can effectively alleviate this long-standing agricultural dilemma still lacks solid empirical validation. This study proposes using the value of regional public tea brands as a proxy indicator to reflect the degree of relief from the agricultural quantity-quality dilemma, thereby accurately characterizing QQEL_TP. Taking the tea industry—where the production-side quantity-quality contradiction is particularly prominent—as the research context, this paper clarifies the theoretical mechanisms and quantifies the empirical effects of how RDIT promotes the coordination of agricultural quantity and quality. The core conclusions are as follows. (1) From 2012 to 2023, the overall RDIT level across major tea-producing provinces in China maintained steady growth, while QQEL_TP remained generally low with moderate improvement momentum. Spatially, significant inter-provincial disparities persisted for both variables; RDIT exhibited a clear gradient-driven development pattern, and the spatial distribution of QQEL_TP evolved from scattered imbalance in the early stage to a relatively concentrated layout characterized by coordinated regional development. (2) RDIT exerts a significantly positive impact on QQEL_TP. After controlling for individual heterogeneity, temporal shocks, and other confounding factors, a 1% increase in RDIT is associated with a notable 0.432% rise in QQEL_TP. (3) The core finding that RDIT significantly improves QQEL_TP remains highly robust after addressing endogeneity through the instrumental variable approach, excluding samples from abnormal pandemic years, and adjusting sample distributions via standardized statistical treatments. (4) RDIT indirectly elevates QQEL_TP by promoting the development of TBV, and this mediating pathway is sensitive to time-dependent macroeconomic fluctuations. (5) The enhancing effect of RDIT on QQEL_TP is negatively moderated by GI_ST, and such moderating characteristics display sensitivity to cross-regional heterogeneous variations. Policy implications The empirical findings based on the tea industry provide targeted policy implications for addressing the widespread agricultural quantity-quality dilemma. First, the agricultural quantity-quality equilibrium should be embedded into the core objective system of high-quality agricultural development. Policymakers should optimize evaluation and incentive mechanisms that prioritize quality improvement, operational efficiency, and long-term sustainability. They should guide smallholder farmers to actively participate in the construction of regional public brand systems and promote agricultural development strategies that balance output expansion and quality enhancement. Second, formulate categorized and differentiated policies to promote the implementation of RDIT, and leverage the integrated empowerment dividends of digital intelligence as the core driver to enhance the synergy between agricultural quantity and quality. For provinces with mature digital foundations such as Zhejiang, Jiangsu, and Guangdong, priorities should be placed on the deep integration of digital-intelligent technologies into the entire agricultural industrial chain, as well as the large-scale deployment of intelligent sensing equipment and precision farming systems across all production links. For less developed regions, including Guangxi and Yunnan, where digital infrastructure lags behind, policymakers should first address shortages in basic facilities, popularize low-cost and replicable technology application models, and establish cross-regional cooperation mechanisms to accelerate technology spillovers and practical experience sharing, so as to gradually narrow inter-regional development gaps. For provinces with moderate RDIT maturity, emphasis should be placed on building high-end digital-intelligent supporting facilities, breaking institutional barriers that restrict the free flow of data, and giving full play to the pivotal role of data factors in forecasting external changes in agricultural markets and improving the efficiency of cross-regional supply-demand matching. Third, continuously upgrade the comprehensive ecosystem of rural e-commerce. By improving logistics service networks for agricultural products and enhancing the digital marketing capabilities of rural practitioners, traditional e-commerce platforms can be upgraded from simple transaction terminals to multi-functional comprehensive service carriers, further strengthening the mediating role of TBV in boosting the positive nexus between RDIT progress and the improvement of agricultural quantity-quality equilibrium. Fourth, restructure the allocation mechanism of public fiscal resources for science and technology and alleviate the crowding-out effect caused by inappropriate government intervention on the autonomous digital investment of market participants. Given that excessive GI_ST has exerted certain inhibitory distortions on the effectiveness of RDIT, it is advisable to formulate practical project evaluation criteria based on actual agricultural production needs, adopt third-party independent performance evaluation and competitive funding allocation mechanisms, curb resource misallocation and redundant infrastructure investment, and ultimately unlock the positive moderating potential of rational government S&T input in fostering RDIT-driven optimization of QQEL_TP. Research limitations This study provides meaningful theoretical references and practical guidance for alleviating the agricultural quantity-quality dilemma, yet several inevitable limitations remain. Firstly, the analytical framework is rooted in the reality of China's smallholder-dominated decentralized farming, which limits the generalizability of the core conclusions to agricultural economies dominated by large-scale commercial farms. Secondly, inconsistent data release timelines across multi-dimensional evaluation indicators restrict the estimation of the RDIT Index to 2023 only, failing to cover the period 2024–2025—a phase characterized by further upgrades in digital intelligence penetration. Such data limitations may underestimate the true positive promotional effect of RDIT on the improvement of QQEL_TP. Thirdly, constrained by the unavailability of operational scale statistics, this paper only uses the number of Taobao Villages to characterize the development of TBV, which inevitably weakens the statistical robustness of the mediating effect tests to a certain extent. Finally, this research mainly focuses on verifying whether RDIT can effectively address the agricultural quantity-quality dilemma through theoretical deduction and empirical modeling, while in-depth exploration of the complex underlying influencing mechanisms remains insufficient. Both mediating and moderating regression results confirm the existence of complex heterogeneous transmission pathways between RDIT evolution and QQEL_TP improvement. Accordingly, future research can conduct more in-depth theoretical elaboration and refined empirical tests to systematically unpack the sophisticated internal driving logic. References Augustin A, Kiliroor CC (2025) EXPRESS: Advanced Hyperspectral Signature Processing for Chemical Stress Detection in Vegetable Leaves Using Hierarchical Feature Extraction and Enhanced Ensemble Model. App spec 37028251411953. https://doi.org/10.1177/00037028251411953 Acemoglu D, Restrepo P (2020) Robots and Jobs: Evidencefrom US Labor Markets. J Poll Eco 128(6):2188–2244. https://doi.org/10.1086/705716 Dong YH, CHENG WJ, ZHANG XY (2010) Analysis on the causes and Countermeasures of quality problems of famous green tea. Chinese Tea, 32(01):27–28. https://doi.org/CNKI:SUN:CAYA.0.2010-01-012 (in Chinese) Dong YH, Yang CX, Lv JD (2014) Reasons for the decline of the quality of Yongxi Huoqing tea and its improvement measures. Chinese Tea 36(03):25–26. https://doi.org/CNKI:SUN:CAYA.0.2014-03-012 (in Chinese) Dong Y, Guo HW, Zhang JH (2025) The impact of digital and intelligent transformation on the innovation The impact of digital and intelligent transformation on the innovation. Res Agri Moder 46(03):495–505. https://doi.org/10.13872/j.1000-0275.2025.0174 (in Chinese) Dhal SB, Kar D (2025) Leveraging artificial intelligence and advanced food processing techniques for enhanced food safety, quality, and security: a comprehensive review. Discov Appl Sci 7:75. https://doi.org/10.1007/s42452-025-06472-w Fang CM, Zhou ZT (2008) Yantai, Shandong Province, issued measures to deal with severely exceeding pesticide residues. China Qual News 2008/12/30(006). (in Chinese) George RV, Harsh HO, Ray P (2019) Food quality traceability prototype for restaurants using blockchain and food quality data index. J Clea Pro 240:118021. https://doi.org/10.1016/j.jclepro.2019.118021 The State Council of the People’s Republic of China (2012) Several opinions on vigorously promoting the development of informatization and effectively safeguarding information security (Guo Fa [2012] 23). 28 June. (in Chinese) Hiroto Y, Rei S, Yuhei H (2020) Dissection of hyperspectral reflectance to estimate nitrogen and chlorophyll contents in tea leaves based on machine learning algorithms. Sci Rep 10:16654. https://doi.org/10.1038/s41598-020-73745-2 Hu Y, Huang P, Wang Y (2023) Determination of Tibetan tea quality by hyperspectral imaging technology and multivariate analysis. J Food Com& Ana 117:105136. https://doi.org/10.1016/j.jfca.2023.105136 Han XF, Xiao J, Song WF (2025) Digital intelligence and urban rural common prosperity: evidence from village level micro space. Fin Eco (05):123–136. (in Chinese) Huang BQ, Sun YL (2023) Research on the Construction of Tea Regional Public Brand under the Background of Rural Revitalization. J Macro-quality Res 11(06):1–13. https://doi.org/10.13948/j.cnki.hgzlyj.2023.06.001 (in Chinese) Jiang J, Tan X, Zhang L (2023) Hybrid N-way partial least squares and random forest model for brick tea identification based on excitation–emission matrix fluorescence spectroscopy. Food Bio Tech 16(6):1335–1342. https://doi.org/10.1007/s11947-023-03006-3 Jing ER (2021) Study on the Construction of Evaluation Index System of Supply Quality about Agricultural Product in Northwest China. J North Minzu Uni 0265–72. https://doi.org/10.3969/j.issn.1674-6627.2021.02.008 (in Chinese) Li WH, Ren WY, Xie WP (2002) The quality of agricultural products in China is facing challenges. Jilin Agri 08:13 (in Chinese) Li ZY, WANG M (2007) Amperometry for pesticide residues. Prog Chem 19(10):1585–1591 Liu YT (2024) Research on the the Influence of Digital andIntelligent Transformation of AgriculturalProductive Services on Agricultural GreenTotal Factor Production. Zhengzhou Institute of Aeronautical Industry Management. https://doi.org/10.27898/d.cnki.gzhgl.2024.000086 . (in Chinese) Liu H (2025) The digital intelligence of agricultural productive services, the integration of Rural three industries and the toughness of agricultural industrial chain supply chain. Stati Deci 41(21):110–115. https://doi.org/10.13546/j.cnki.tjyjc.2025.21.018 (in Chinese) Li H, Mao YL, Shi HT (2024) Establishment of deep learning model for the growth of tea cutting seedlings based on hyperspectral imaging technique. Sci Horti 320:113106. https://doi.org/10.1016/j.scienta.2024.113106 Lanjewar MG, Panchbhai KG (2023) Convolutional neural network based tea leaf disease prediction system on smart phone using paas cloud. NEU COMP APP 35(3):2755–2771. https://doi.org/10.1007/s00521-022-07743-y Li YT, Wu SK, He LY (2023) Development and field evaluation of a robotic harvesting system for plucking high-quality tea. Com Elec Agri 206:107659. https://doi.org/10.1016/j.compag.2023.107659 Luo N, Li YL, Yang BH (2022) Prediction Model for Tea Polyphenol Content with Deep Features Extracted Using 1D and 2D Convolutional Neural Network. AGRICULTURE-BASEL 12(9):1299. https://doi.org/10.3390/agriculture12091299 Luo BY, Chen YX (2022) How Can Digital Intelligence Enable High-quality Economic Development:A Study on the Regulatory Role of Business Environment. Sci Tech Pro& Poli 39(05):61–71. https://doi.org/10.6049/kjjbydc.2021040437 (in Chinese) Liu L, Ruan JJ, Zhuang HT (2024) Impact of Digital Intelligence on Green Development in the Yangtze River Delta Urban Agglomeration. Eco Geo 44(09):123–132. https://doi.org/10.15957/j.cnki.jjdl.2024.09.013 (in Chinese) Liu RZ, Li X, He C (2025) Mechanism and Verification of Empowering High-Quality Economic Development through Urban Digitization Transformation. Ur Dev Stu 31(01):6–11 (in Chinese) Lyu K, Pan JB, Lin ZS (2024) Innovative Province Construction, Regional Smartization, and Competitiveness of Regional Innovation Ecosystems——Mechanism of Influence and Quasi-Natural Experiments. R&D Manage 36(06):137–150. https://doi.org/10.13581/j.cnki.rdm.20231424 (in Chinese) Li DH, Xiong Y, Chen JH (2024) The Impact of Digital Village Construction on the Regional Public Brand Value of Agricultural Products——Empirical analysis based on the regional public brand of tea. J Macro-quality Res 12(04):101–114. https://doi.org/10.13948/j.cnki.hgzlyj.2024.04.008 (in Chinese) Mao YL, Li H, Wang Y (2022) Prediction of Tea Polyphenols, Free Amino Acids and Caffeine Content in Tea Leaves during Wilting and Fermentation Using Hyperspectral Imaging. Foods 11(16):2537. https://doi.org/10.3390/foods11162537 Mei Y, Mao DL, Lu YH (2020) Effects and mechanisms of rural E-commerce clusters on households' entrepreneurship behavior in China. Growth Change 51(4):1588–1610. https://doi.org/10.1111/grow.12424 Müller C (2025) World robotics 2025 – industrial robots. IFR Statistical Department, VDMA Services GmbH, Frankfurt am Main, Germany. Available at: https://ifr.org/img/worldrobotics/Sources_Methods_WR_2025_Industrial_Robots.pdf?utm_source=chatgpt.com (Accessed: 1 November 2025) Nie J, Wang Y, Li Y (2022) Artificial intelligence and digital twins in sustainable agriculture and forestry: a survey. Turkish J Agri Fore 46(5):642–661. https://doi.org/10.55730/1300-011X.3033 Nie WJ, Fu HM, Li TP (2021) A Theoretical Basis for and Path Towards Quality Grading System of Agricultural Products in China——Welfare Evaluation Based on Choice Experiment Survey. J Agrotech Eco( 0765–78. https://doi.org./10.13246/j.cnki.jae.2021.07.005 (in Chinese) Paul T, Islam N, Rakshit S (2025) The impact of artificial intelligence (AI)-enabled collaborative approach: achieving sustainable supply chain performance. J Busi Log 46(4):e70030. https://doi.org/10.1111/JBL.70030 Paul T, Mondal S, Islam N (2021) The impact of blockchain technology on the tea supply chain and its sustainable performance. Tech Fore Soc Chan 173:121163. https://doi.org/10.1016/j.techfore.2021.121163 Peng Y, Zheng C, Guo S (2023) Metabolomics integrated with machine learning to discriminate the geographic origin of Rougui Wuyi rock tea. Npj Sci Food 7:7. https://doi.org/10.1038/s41538-023-00187-1 Ren G, Wu R, Yin L (2024) Intelligent Evaluation of the Appearance Modality of Black Tea Based on Machine Vision and Hyperspectral Imaging. Ana Lett 57(2):176–189. https://doi.org/10.1080/00032719.2023.2201706 Sun SY, Li S (2024) The Purpose and Driving Direction of the Digitization Logic for Cultivating and Developing New Quality Productive Forces. China Rev Poli Eco 15(06):101–114 (in Chinese) Su YM, FANG F, WANG YH (2026) The impact of digital and intelligent transformation on the new quality productivity of agricultural enterprises. J Henan Agri Uni 1–16. https://doi.org/10.16445/j.cnki.1000-2340.20251104.002 . (in Chinese) Sheng XF, Zan JZ, Jiang YW (2023) Data fusion strategy for rapid prediction of moisture content during drying of black tea based on micro-NIR spectroscopy and machine vision. Optik 276:170645. https://doi.org/10.1016/j.ijleo.2023.170645 Sun J, Hu Y, Zou YL (2022) Identification of pesticide residues on black tea by fluorescence hyperspectral technology combined with machine learning. Food Sci Tech 42:e55822. https://doi.org/10.1590/fst.55822 Shao YTY, Ni JW, Zhou SJ (2024) Safety assessment of agricultural products and the pesticide regulation trend in China. Agr Food Eco 12:29. https://doi.org/10.1186/s40100-024-00322-w Taherzadeh O, Cai H, Mogollón JM (2026) Small farms contribute a third of the food consumed in high-income nations. Nat Food 7:66–73. https://doi.org/10.1038/s43016-025-01276-y Tang YX, Cai XJ, Zhuang HT (2025) The Mechanism and Test of Digital Intelligence Drives of the Chinese-Style Modern Rural Revitalization——Based on the Marxist Theory of Productive Forces. Pra Fin Eco 46(02):114–122. https://doi.org/10.16339/j.cnki.hdxbcjb.2025.02.014 (in Chinese) Wei WC (2003) Guangxi Bureau strictly controls excessive pesticide residues. China Guomen Times 2003/7/15. (in Chinese) Wang W (2005) The phenomenon of excessive pesticide residues in vegetables sampled in Changsha is rising. China Quality News 2005/08/27(006). (in Chinese) Wei Y, Wen YQ, Huang XL (2024) The dawn of intelligent technologies in tea industry. Trends Food Sci& Tech 145:104337. https://doi.org/10.1016/j.tifs.2024.104337 Wei G, Zhou R (2023) Comparison of machine learning and deep learning models for evaluating suitable areas for premium teas in Yunnan, China. PLoS ONE 18(2):e0282105. https://doi.org/10.1371/journal.pone.0282105 Wei P, Liu HM, Xu CK (2024) Does Green Food Certification promoting agri-food export quality? Evidence from China. J Integra Agri 23(3):1061–1074. https://doi.org/10.1016/j.jia.2023.11.033 Wang YQ (2024) Quality measurement and influencing factors analysis of export agricultural products in Gansu Province -- Based on RCEP. North Eco Tra (06):115–119. (in Chinese) Wang GY, Wang M, Ding S (2024) Differentiation and Realization of Millet’s Production Capacity and Value from the Perspective of Technical Efficiency. 10.13246/j.cnki.jae.2024.05.002 . (in Chinese) J Agrotechn Econ https://doi.org/ Wu W, Zhang Y, Fan Y (2020) ICT Empowers the Formation and Development of Rural E-Commerce in China. IEEE Access 8:135264–135283. https://doi.org/10.1109/ACCESS.2020.3011593 Wang YQ, Dong W (2023) Man machine interaction: the impact of the rise of robots on the income of Chinese workers. J World Eco 46(07):88–115. https://doi.org./10.19985/j.cnki.cassjwe.2023.07.007 (in Chinese) Xiao H, Fu YL (2024) Practical Logic and Optimization Strategy of Smart Agriculture Development–Based on A Case Study of Digitally Intelligent Unmanned Orchards on Hilly Mountainous Terrain in Q Village. J Northwest A&F University (Social. Sci Edition) 24(03):58–65. https:.//doi.org/10.13968/j.cnki.1009-9107.2024.03.07 (in Chinese) Yu ZY, Wang J, Zheng T (2020) An Online Tea Fixation State Monitoring Algorithm Based on Image Energy Attention Mechanism and Supervised Clustering (IEAMSC). Sensors 20(15):4312. https://doi.org/10.3390/s20154312 Yang XL, Fu L, MOU ED (2020) Cultivation of regional public brand of agricultural products: operation mechanism and realization path. Agri Eco (01):125–127. https://doi.org/CNKI:SUN:NYJJ.0.2020-01-047 (in Chinese) Yang WS, Hu SD (2011) Only urea application is the main reason for the deterioration of the quality of improved varieties of tea. J Tea Busi 33(02):67–68. https://doi.org/10.16015/j.cnki.jteabusiness.2011.02.014 (in Chinese) Yu ZY, Quan CX, Fang Z (2024) Key technology, equipment and prospect of digital and intelligent farmland construction. Tran Chin Soc Agri Eng 40(23):1–14. https://doi.org/10.11975/j.issn.1002-6819.202407107 (in Chinese) Yuan L, Zhang SQ, He WJ (2025) Impact of Digital Intelligence Level on Green Total Factor Productivity in the Yangtze River Economic Belt. Res &Env Yangtze Basin 34(05):923–936 (in Chinese) You XW, Wan JY (2022) Research on the Spatial Spillover Effect of Local Officials' Agriculture-related Experience on the Quality and Safety of Agricultural Products:Evidence from Chinese Cities. South China J Eco 0337–52. https://doi.org/10.19592/j.cnki.scje.391317 (in Chinese) Yuan HM (2012) ‘Tea e-commerce alliance calls for honest operation’. Market Herald, 16 Apr. Available at: https://www.zjscdb.com/detail.php?newsid=69168 (Accessed: 12 Mar. 2026).(in Chinese) Zhang XC, Wu ZM, Cao CM (2023) Design and operation of a deep-learning-based fresh tea-leaf sorting robot. Com &Elec Agri 206:107664. https://doi.org/10.1016/j.compag.2023.107664 Zhang J, Li ZC (2021) Intelligent Tea-Picking System Based on Active Computer Vision and Internet of Things. Secu Comm Net 5302783. https://doi.org/10.1155/2021/5302783 Zou ZY, Wu QS, Long T (2023) Classification and adulteration of Mengding Mountain green tea varieties based on fluorescence hyperspectral image method. J Food Com Ana 117:105141. https://doi.org/10.1016/j.jfca.2023.105141 Zhang Y, Bai PW (2023) How digital intelligence affects double cycle participation and income gap -- Based on provincial industry level data. J Manage World 39(10):58–83. https://doi.org/10.19744/j.cnki.11-1235/f.2023.0126 (in Chinese) Zhou JW, Guo W, Cai XJ (2025) How Digital Intelligence Promotes Education Equity. J Edu Sci Hunan Normal Uni 24(03):12–21. https://doi.org/10.19503/j.cnki.1671-6124.2025.05.002 (in Chinese) Zhang J, Wang HY (2025) Quality Traceability Subsidies and Cotton Quality:Mechanism and Empirical Test. J Macro-qua Res 13(02):1–14. https://doi.org/10.13948/j.cnki.hgzlyj.2025.02.001 (in Chinese) Zhang N, Yang WT, Ke HQ (2024) Does rural e-commerce drive up incomes for rural residents? Evidence from Taobao villages in China. Eco Ana& Pol 82:976–998. https://doi.org/10.1016/j.eap.2024.04.23 Zhou J, Yu L, Choguill CL (2021) Co-evolution of technology and rural society: The blossoming of taobao villages in the information era, China. J Rural Stu 83:81–87. https://doi.org/10.1016/j.jrurstud.2021.02.022 Zhang J, Wang CC, Phelps NA (2023) Rural E-Commerce and Emerging Paths Toward Product Renewal: Evidence from Taobao Villages in Zhejiang Province, China. Profe Geo 75(3):521–535. https://doi.org/10.1080/00330124.2022.2111689 Zhu J (2023) Research on digital economy enabling high quality development of China's agriculture. Southwest University of Finance and Economics. https://doi.org/10.27412/d.cnki.gxncu.2023.000037 . (in Chinese) Zhang YZ, Zhu JN, Zhang YL (2025) Development of artificial intelligence and labor mobility. Eco Pers (01):128–145. https://link.cnki.net/urlid/11.1057.F.20250307.1006.016 . (in Chinese) Appendix Appendix C of the Industrial Classification for National Economic Activities (GB/T 4754–2017), jointly issued by the National Bureau of Statistics of China and the China National Institute of Standardization, provides a concordance table between China's industrial classification and ISIC Rev. 4. Although the 2013 China Economic Census Yearbook adopts the 2011 version of the classification (GB/T 4754–2011), the overall framework of China's manufacturing sectors in the 2011 system is highly consistent with that in the 2017 version. Hence, this concordance table remains applicable. The categories coded C42 (comprehensive utilization of waste resources) and C43 (repair of metal products, machinery, and equipment) under China's industrial classification are not classified as manufacturing in international standards. As such, they are not included in IFR statistics and cannot be matched, so these two sectors are excluded from the analysis. This study starts from 2012. If 2012 were set as the base year, data for only 21 disaggregated manufacturing sectors would be available from the China Industry Statistical Yearbook. By contrast, using 2013 as the base year allows us to obtain data for 29 detailed sectors from the 2013 China Economic Census Yearbook. Given that no major policy shifts or industrial adjustments occurred between 2012 and 2013, regional industrial structures remained nearly unchanged. To improve measurement accuracy, 2013 is chosen as the base year. Additional Declarations The authors declare no competing interests. Tables 1 to 6 are available in the Supplementary Files section. Supplementary Files tabel1andtable2.png Table 1 Evaluation System of Regional Digital-Intelligent Transformation Index (RDIT)；Table 2 Descriptive Statistics of Variables tabel35.png Table 3 Baseline Regression Results；Table 4 Endogeneity Test；Table 5 Robustness Tests tabel6andtable7.png Table 6 Results of Mediating Effect Analysis；Table 7 Results of Moderating Effect Analysis V1EthicalStatement.docx Ethical Statement Cite Share Download PDF Status: Posted Version 2 posted You are reading this latest preprint version Show more versions Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9428414","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":624730980,"identity":"a88b9767-43d9-4087-b387-4ebfc937b335","order_by":0,"name":"Lu Jing","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAvklEQVRIie2RPQrCQBBGJwhjs8Z2QIlX2BBI7VE2janSSsrYbBrP4DFST4ik2hvYBLxEClHTCYIwdoL7qim+x/wBeDw/SAgdd7cSI5zXLFMQ+mxQLkxC5YxUUYle2Cg70VYLlcClRJjmlsDAWDaiwfak1a6wqwMHR3eRKF1DhvrCrtnMAitSOCXWjxzJaKlyTuLKoPlG6bMrMMZ2OnIr2mVJ0yuDO242dd0OYylQgPhV86fQW5tKlvN4PJ4/5gmSVzrQM5HJ5AAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0002-8722-1572","institution":"School of Economics, Fujian Normal University","correspondingAuthor":true,"prefix":"","firstName":"Lu","middleName":"","lastName":"Jing","suffix":""},{"id":625938537,"identity":"1db12188-c0a2-4dd5-a76d-f65e44663e91","order_by":1,"name":"Xu Yao","email":"","orcid":"","institution":"School of Economics, Fujian Normal University","correspondingAuthor":false,"prefix":"","firstName":"Xu","middleName":"","lastName":"Yao","suffix":""},{"id":625938538,"identity":"b00b6b16-5dea-4fae-b620-a9982a10421c","order_by":2,"name":"Lin Dongkai","email":"","orcid":"","institution":"Shortage Talent Reserve Center, Committee Organization Department of Guigang City","correspondingAuthor":false,"prefix":"","firstName":"Lin","middleName":"","lastName":"Dongkai","suffix":""},{"id":625938539,"identity":"89164459-fd2d-4709-a72e-97a899b54702","order_by":3,"name":"Wu Wulin","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAsklEQVRIiWNgGAWjYFAC5sYHCQb/5NjY2w8Qq4Wx2eBDxQFjPp4zCURraZOcceZA4jwJBwPiNBjcSGyT5m27k94mwZDA8KNiG1Famq15257ltkk3HmDsOXObsBaz24mNt3nbmHPbZA4kMDO2EaelAegw5nQ2iQQDorU0Ab1/OIF4Lfb3H4ICOc2wDRjIB4nyi2TP4YPAqLSRl29vP/jgRwURWlDAARLVj4JRMApGwSjABQBvRkIYZ9bb8wAAAABJRU5ErkJggg==","orcid":"","institution":"School of Economics, Fujian Normal University","correspondingAuthor":true,"prefix":"","firstName":"Wu","middleName":"","lastName":"Wulin","suffix":""}],"badges":[],"createdAt":"2026-04-15 14:55:52","currentVersionCode":2,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-9428414/v2","doiUrl":"https://doi.org/10.21203/rs.3.rs-9428414/v2","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107665774,"identity":"1aca51ba-ef0a-41e9-94af-a7f616b0232a","added_by":"auto","created_at":"2026-04-23 18:48:47","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":867822,"visible":true,"origin":"","legend":"\u003cp\u003eSee image above for figure legend.\u003c/p\u003e","description":"","filename":"figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-9428414/v2/5ca4f1ade6a131c0607a2099.png"},{"id":107707917,"identity":"10c014b2-90ab-42d4-954b-6cffb8bee3eb","added_by":"auto","created_at":"2026-04-24 09:21:25","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1219607,"visible":true,"origin":"","legend":"\u003cp\u003eSee image above for figure legend.\u003c/p\u003e","description":"","filename":"figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-9428414/v2/4c26499b809beed3d8e80e07.png"},{"id":107706383,"identity":"f0694a29-9c6b-4060-9acf-025e576c0c67","added_by":"auto","created_at":"2026-04-24 09:17:58","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1544194,"visible":true,"origin":"","legend":"\u003cp\u003eSee image above for figure legend.\u003c/p\u003e","description":"","filename":"figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-9428414/v2/7bc76c7ad21cec10e8abc793.png"},{"id":107708053,"identity":"5c5ef692-7d06-442c-9d2f-eee6f3e8b79f","added_by":"auto","created_at":"2026-04-24 09:21:46","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1079361,"visible":true,"origin":"","legend":"\u003cp\u003eSee image above for figure legend.\u003c/p\u003e","description":"","filename":"figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-9428414/v2/99995b669badc9bc5a761c98.png"},{"id":107707923,"identity":"80e97d2f-3376-4816-a841-bfba532eb15c","added_by":"auto","created_at":"2026-04-24 09:21:25","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1747781,"visible":true,"origin":"","legend":"\u003cp\u003eSee image above for figure legend.\u003c/p\u003e","description":"","filename":"figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-9428414/v2/86c043f2a27451ec8312b8f7.png"},{"id":107710440,"identity":"030a6077-efba-4603-ad77-69400812e040","added_by":"auto","created_at":"2026-04-24 09:40:52","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":5157592,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9428414/v2/dfd7ca47-7d93-4bed-bc1e-c2f80761f708.pdf"},{"id":107707802,"identity":"8abff2fb-7c72-4905-9782-551c197d24fd","added_by":"auto","created_at":"2026-04-24 09:21:09","extension":"png","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":3124052,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTable 1 \u0026nbsp;\u003c/strong\u003eEvaluation System of Regional Digital-Intelligent Transformation Index (RDIT)；\u003cstrong\u003eTable 2 \u0026nbsp;\u003c/strong\u003eDescriptive Statistics of Variables\u003c/p\u003e","description":"","filename":"tabel1andtable2.png","url":"https://assets-eu.researchsquare.com/files/rs-9428414/v2/f967e0906d0ff2e7e3d9510b.png"},{"id":107665776,"identity":"9835e9e1-2ed0-4d8a-9bc4-45897b19f2a9","added_by":"auto","created_at":"2026-04-23 18:48:47","extension":"png","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":2028682,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTable 3 \u003c/strong\u003eBaseline Regression Results；\u003cstrong\u003eTable 4 \u0026nbsp;\u003c/strong\u003eEndogeneity Test；\u003cstrong\u003eTable 5 \u0026nbsp;\u003c/strong\u003eRobustness Tests\u003c/p\u003e","description":"","filename":"tabel35.png","url":"https://assets-eu.researchsquare.com/files/rs-9428414/v2/1f8e44db300fe38ae76e9a5f.png"},{"id":107707978,"identity":"300588d1-f3c8-4518-afa0-ca01b2c2da22","added_by":"auto","created_at":"2026-04-24 09:21:33","extension":"png","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":1496522,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTable 6 \u0026nbsp;\u003c/strong\u003eResults of Mediating Effect Analysis；\u003cstrong\u003eTable 7 \u0026nbsp;\u003c/strong\u003eResults of Moderating Effect Analysis\u003c/p\u003e","description":"","filename":"tabel6andtable7.png","url":"https://assets-eu.researchsquare.com/files/rs-9428414/v2/816b3fcce1618093a6bf5347.png"},{"id":107665780,"identity":"90795e35-0136-4fb8-a228-9876e9edff36","added_by":"auto","created_at":"2026-04-23 18:48:47","extension":"docx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":15841,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEthical Statement\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"V1EthicalStatement.docx","url":"https://assets-eu.researchsquare.com/files/rs-9428414/v2/f830192af3424c4903ce75d6.docx"}],"financialInterests":"\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003cp\u003eTables 1 to 6 are available in the Supplementary Files section.\u003c/p\u003e","formattedTitle":"Does Regional Digital-Intelligent Transformation Mitigate the Quantity-Quality Dilemma in Agricultural Production?\n—Empirical Evidence from China's Tea Industry","fulltext":[{"header":"Introduction","content":"\u003cp\u003eDuring periods of material scarcity and poverty, people yearn for adequate food and clothing, and expanding production volume has consistently been the primary objective of agricultural endeavors. As \u003cem class=\"Highlight ht71194251-f7a6-4c2d-a145-3d9f25b46662\" style=\"font-style: inherit;\"\u003ehuman\u003c/em\u003e society progresses, numerous countries have transcended poverty, and there is now a heightened emphasis on enhancing the quality of life. Consequently, improving the quality of agricultural products has emerged as a pivotal production goal for many agricultural powerhouses. Augmenting quantity is tantamount to boosting yield, whereas enhancing quality encompasses not only the refinement of sensory attributes such as color, aroma, taste, and form but also entails stricter food safety regulations and more meticulous initial processing standards. Given constant mechanical efficiency, escalating the application of fertilizers and pesticides, curtailing processing and management durations, and compromising on the quality standards of finished products can substantially elevate yields, albeit at the expense of quality. When rural labor resources are abundant or even in surplus, labor costs remain low and may not even be factored into the cost calculations by smallholder farmers. By deploying substantial labor in precision farming, these farmers can not only diminish the reliance on fertilizers and pesticides but also uphold stringent operational standards across all stages. Under such circumstances, the dichotomy between quality and quantity in agricultural production is not pronounced. However, with the advent of industrialization, a growing number of rural laborers migrate to urban areas, driving up labor costs. In the absence of sustained precision farming practices and adequate capital to invest in advanced machinery, smallholder farmers are more inclined to adopt production strategies that prioritize quantity over quality, such as indiscriminate use of fertilizers and pesticides, streamlining processing and management procedures, and relaxing quality standards for finished products. Diminished quality leads to lower market acceptance and reduced profitability for producers, perpetuating a vicious cycle wherein producers increasingly favor quantity at the expense of quality. This cycle persists until profits turn negative, prompting an exodus of agricultural labor. Nevertheless, with technological advancements, mechanical substitution can partially offset labor shortages, enabling large-scale agricultural producers to reconcile the conflicting objectives of quality and quantity. In the early 21st century, the improvement in mechanization-driven labor substitution has been insufficient to offset the severe shortages of agricultural labor, leading to increasingly extensive agricultural production practices (Li et al. \u003cspan class=\"CitationRef\"\u003e2002\u003c/span\u003e) Consequently, quality-related incidents such as excessive pesticide residue limits and the deterioration of flavor characteristics of agricultural products have been frequently documented in public reports (Wei \u003cspan class=\"CitationRef\"\u003e2003\u003c/span\u003e; Wang. 2005; Li \u003cspan class=\"CitationRef\"\u003e2007\u003c/span\u003e; Fang and Zhang 2008 ).\u003c/p\u003e \u003cp\u003eBrand-oriented development facilitates differentiated competition for agricultural entities that adopt quality-first or balanced quantity-quality production strategies, drawing a clear demarcation from opportunistic producers obsessed with pure output expansion. Nevertheless, most major agricultural economies worldwide belong to developing regions, where scattered smallholder households dominate agricultural production (Taherzadeh et al.2026). These small-scale producers lack sufficient intellectual capacity and financial resources to support systematic brand development. To break the vicious cycle of pursuing high yields at the expense of quality, the Chinese government has vigorously promoted the development of regional public agricultural product brands. It guides local agricultural operators to coordinate via technology sharing, unified joint quality control, and industrial collaborative governance, thereby collectively upholding the stable quality of agricultural products (Yang et al. \u003cspan class=\"CitationRef\"\u003e2020\u003c/span\u003e). The establishment of regional public brands has transformed the decision-making dynamics concerning agricultural production quantity and quality, transitioning from a zero-sum game constrained by individual fixed resources to a positive-sum game that seeks collective gains through coordination. The value of regional public brands for agricultural products, a common metric for assessing the effectiveness of such brand construction, represents a comprehensive value proposition centered on quality, augmented by considerations of effective scale and market realization. This value effectively captures the overall outcome of the strategic interplay between quantity expansion and quality enhancement in local agricultural production, reflecting the Quantity-Quality Equilibrium Level of agricultural production. The mathematical essence of resolving the dilemma in agricultural production quantity and quality decisions lies in enhancing this equilibrium level. Based on this premise, this paper proposes utilizing the value of regional public brands for agricultural products as a proxy for the Quantity-Quality Equilibrium Level of agricultural production, thereby indicating the extent to which the quantity-quality dilemma in agricultural production has been mitigated in the region. As a pivotal transformative force in Agriculture 4.0, digital and intelligent technologies have been empirically \u003cem class=\"Highlight ht4fc55b9d-f515-4fa8-9e5d-6731d62f45ba\" style=\"font-style: inherit;\"\u003evalidated\u003c/em\u003e by numerous natural science \u003cem class=\"Highlight ht4fc55b9d-f515-4fa8-9e5d-6731d62f45ba\" style=\"font-style: inherit;\"\u003eexperiments\u003c/em\u003e to enhance agricultural production efficiency. Nevertheless, within the realm of economics, the effectiveness of their dissemination and application in alleviating the quantity-quality dilemma in agricultural production remains to be substantiated.\u003c/p\u003e \u003cp\u003eThe tea industry belongs to labor-intensive and high-value agriculture. Compared with conventional agricultural commodities such as corn, wheat, and rice, tea production entails more intricate operational links, stringent standardized protocols, and higher thresholds for practical experience and professional skills. Even sensory quality assessment faces evident technical barriers, leading to pronounced information asymmetry and a more severe quantity-quality predicament (Dong et al. \u003cspan class=\"CitationRef\"\u003e2010\u003c/span\u003e; Yang and Hu \u003cspan class=\"CitationRef\"\u003e2011\u003c/span\u003e; Dong et al. \u003cspan class=\"CitationRef\"\u003e2014\u003c/span\u003e). Adopting panel data of China's tea industry spanning 2012 to 2023, this paper empirically investigates how Regional Digital-Intelligent Transformation (RDIT) affects the Quantity-Quality Equilibrium Level (QQEL_TP) in agricultural production. This study delivers four major marginal contributions to the literature. First, targeting the tea industry as the specific research context, it clarifies the \u003cem class=\"Highlight ht29216696-c42e-4f00-932a-aea34347df6a\" style=\"font-style: inherit;\"\u003etheoretical\u003c/em\u003e logic whereby RDIT improves QQEL_TP, and further unpacks the internal driving mechanism mediated by Taobao villages development (TBV), as well as the external contextual influence moderated by government investment in science and technology (GI_ST). Second, the value of the regional public tea brand is introduced as a rigorous proxy indicator to measure QQEL_TP in agricultural production. Third, defining Digital-Intelligent Transformation as an organic integration of digitalization and intellectualization rather than simplistic numerical aggregation, we develop the Regional Digital and Intelligent Transformation Index grounded in holistic systematic logic for indicator construction. Fourth, at the macro level, this research quantitatively estimates the baseline influence magnitude of RDIT on QQEL_TP of tea production, while systematically verifying its underlying mediating pathways and moderating effects.\u003c/p\u003e \u003cp\u003eLiterature review\u003c/p\u003e \u003cp\u003eDue to difficulties in acquiring data on measurable proxy variables, empirical economic research directly targeting the Quantity-Quality Equilibrium Level in agricultural production is relatively limited. Early scholarly attempts to mitigate the inherent quantity-quality predicament in agriculture generally adopted a fragmented analytical perspective, independently investigating the driving mechanisms for improving either agricultural output volume or product quality separately. By contrast, recent literature has mainly focused on industrial efficiency optimization that simultaneously boosts both quantity and quality performance. Research concerning Digital-Intelligent Transformation belongs to this latest research stream. Although ample economic empirical analyses have been conducted for agriculture as a whole, relevant studies specific to the tea industry are predominantly confined to \u003cem class=\"Highlight ht4fc55b9d-f515-4fa8-9e5d-6731d62f45ba\" style=\"font-style: inherit;\"\u003eexperimental\u003c/em\u003e research in natural science and engineering fields. From a methodological perspective, evaluation frameworks for quantifying Digital-Intelligent Transformation have matured progressively as academics deepen their comprehension of digital-intelligent connotations. Even so, few construction paradigms effectively embody the organic integration of digitalization and intelligentization grounded in holistic logical reasoning, instead resorting merely to simplistic numerical combination strategies. Owing to the scarcity of studies that directly concentrate on agricultural quantity-quality equilibrium issues, the academic community still lacks adequate systematic assessments and measurement investigations of the Quantity-Quality Equilibrium Level in agriculture.\u003c/p\u003e \u003cp\u003eDigital-Intelligent Transformation and agricultural production\u003c/p\u003e \u003cp\u003eExtant studies generally acknowledge that Digital-Intelligent Transformation is far more than a mere upgrade of technical tools; it serves as the logical starting point and inevitable path to cultivate new quality productive forces in agriculture. Through systematically reshaping laborers, means of labor, and objects of labor, it drives agricultural production to shift from traditional experience-dominated patterns to modern data-driven operations (Sun et al. 2024). At the micro level, Digital-Intelligent Transformation significantly elevates the productivity of listed agricultural enterprises by improving innovation efficiency, diversifying supply chain structures, easing financing constraints, and accelerating \u003cem class=\"Highlight ht71194251-f7a6-4c2d-a145-3d9f25b46662\" style=\"font-style: inherit;\"\u003ehuman\u003c/em\u003e capital accumulation, while broadening their innovation boundaries simultaneously (Su et al.2026; Dong et al.2025). At the macro level, the Digital-Intelligent Transformation of agricultural producer services covering financial support, scientific and information services, and commercial circulation services can prominently boost agricultural green total factor productivity (Liu \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e). It also enhances the resilience of agricultural industrial supply chains by facilitating the integrated development of primary, secondary, and tertiary industries in rural areas (Liu \u003cspan class=\"CitationRef\"\u003e2025\u003c/span\u003e). In specific agricultural practices, embedding digital and intelligent technologies into pre-production and in-production segments allows real-time collection of crop growth and environmental information, alongside unmanned farming machinery executing operational decisions released by data centers. This achieves scientific regulation and standardized documentation of water, fertilizer, and pesticide application as well as farming management practices. It effectively cuts the input costs of general production factors whilst ensuring high quality, high yield, efficiency, and safety of agricultural products. Large-scale field trials verify that digitally and intelligently upgraded farmlands reduce labor costs by around 50%, pesticide usage by 30%, and fertilizer consumption by 15%-25%. Relative to conventional irrigation, overall water conservation reaches 50%, and drip irrigation delivers an additional water-saving effect of 20%-30%. Correspondingly, the per-unit yield rises by 15%-25% and comprehensive economic efficiency increases by approximately 25% (Yu et al. \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e). In addition, the embedding of digital-intelligent technologies into post-production links greatly improves the supply capacity and operational efficiency of high-quality agricultural goods. The integration of deep learning algorithms and hyperspectral detection not only enhances the accuracy and efficiency of pesticide residue identification but also reduces inspection expenditures and \u003cem class=\"Highlight ht29216696-c42e-4f00-932a-aea34347df6a\" style=\"font-style: inherit;\"\u003esample\u003c/em\u003e attrition losses (Augustin and Kiliroor \u003cspan class=\"CitationRef\"\u003e2025\u003c/span\u003e). The collaborative application of blockchain, artificial intelligence, and the Internet of Things within agricultural supply chain management systems significantly lowers coordination and verification costs between producers, retailers, and wholesalers enabled by full lifecycle production data traceability (Paul et al. \u003cspan class=\"CitationRef\"\u003e2025\u003c/span\u003e). Meanwhile, reinforcement learning algorithms dynamically optimize temperature and humidity control throughout cold chain logistics, curbing the decay and damage ratio of agricultural commodities during transportation and warehousing (Dhal et al.2025). However, equipment-only upgrades of Digital-Intelligent Transformation cannot generate substantial improvements in agricultural economic performance (Nie et al. \u003cspan class=\"CitationRef\"\u003e2022\u003c/span\u003e). The investment and deployment of digital-intelligent facilities face prominent entry barriers. Sound institutional arrangements are essential to align the interests of multiple stakeholders, including governments, agricultural producers, investors, and specialized agricultural service firms. Positive institutional incentives combined with external competitive pressures can motivate all participants to mobilize resources synergistically and promote the practical and effective implementation of Digital-Intelligent Transformation across the whole agricultural production process (Xiao and Fu \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eDigital-Intelligent Transformation and tea production\u003c/p\u003e \u003cp\u003e\u003cem class=\"Highlight ht4fc55b9d-f515-4fa8-9e5d-6731d62f45ba\" style=\"font-style: inherit;\"\u003eExperimental\u003c/em\u003e research in natural sciences and engineering has fully confirmed that embedding digital and intelligent technologies into every segment of the tea industrial chain can significantly enhance operational accuracy and efficiency at each stage (Wei et al. \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e). In essence, achieving simultaneous improvements in quality and efficiency for the whole tea industrial chain through Digital-Intelligent Transformation is technically viable. In tea cultivation and field management, the combination of deep learning and remote sensing can assess the suitability of tea-growing regions with an accuracy of 94% (Wei and Zhou \u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e). When integrated with hyperspectral imaging technology, it supports non-destructive monitoring of nitrogen accumulation in tea leaves (Hiroto et al. \u003cspan class=\"CitationRef\"\u003e2020\u003c/span\u003e) and the growth conditions of tea buds and roots (Li et al. \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e). Coupled with cloud computing services, smartphone-captured imagery enables real-time identification and early prediction of tea plant diseases (Lanjewar and Panchbhai \u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e). If further connected to Internet of Things facilities, these technologies enable scientific planning of tea garden layouts, precise water and fertilizer management, and targeted pest and disease control, thereby enhancing operational efficiency while reducing labor costs associated with manual patrols. During the harvesting stage, advanced mechanical harvesting systems powered by deep learning have been developed. These systems integrate tea bud detection, picking coordinate positioning, and motion trajectory planning to support targeted harvesting of tender high-grade fresh tea shoots (Li et al. \u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e). Robotic arms embedded with deep learning algorithms can rapidly classify fresh tea leaves at a maximum accuracy of 92%, resolving the uneven maturity problem commonly found in mechanically harvested raw tea materials (Zhang et al. \u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e). Integrated with the Internet of Things framework, intelligent automated tea harvesting with real-time status transmission and dynamic quality assessment has been realized (Zhang and Li \u003cspan class=\"CitationRef\"\u003e2021\u003c/span\u003e), which helps alleviate the acute labor shortage during peak harvesting periods. Regarding processing and quality detection, the synergistic application of deep learning, spectral analysis, computer vision, and mobile intelligent terminals achieves non-destructive monitoring of in-process tea characteristics across four critical procedures: withering, fixation, fermentation, and drying (Mao et al. \u003cspan class=\"CitationRef\"\u003e2022\u003c/span\u003e; Yu et al. \u003cspan class=\"CitationRef\"\u003e2020\u003c/span\u003e; Sheng et al. \u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e). This avoids inconsistent finished tea quality stemming from subjective empirical judgment bias in manual operations. Moreover, these integrated technologies realize rapid and accurate quantification of pesticide residue levels (Sun et al. \u003cspan class=\"CitationRef\"\u003e2022\u003c/span\u003e), tea \u003cem class=\"Highlight ht29216696-c42e-4f00-932a-aea34347df6a\" style=\"font-style: inherit;\"\u003eadulteration\u003c/em\u003e rates (Zou et al. \u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e), physicochemical attributes (Ren et al. \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e), and flavor compound concentrations in finished tea products (Luo et al. \u003cspan class=\"CitationRef\"\u003e2022\u003c/span\u003e), substantially elevating overall quality inspection efficiency. For finished tea quality grading and supply chain traceability, the fusion of machine learning, radio frequency identification, the Internet of Things, and blockchain enables full lifecycle traceability of tea production and circulation data (Paul et al. \u003cspan class=\"CitationRef\"\u003e2021\u003c/span\u003e). It also empowers quality supervisors and ordinary consumers without professional tea evaluation expertise to accurately distinguish tea varieties (Jiang et al. \u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e), flavor grades (Hu et al. \u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e), and geographical provenances (Peng et al. \u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e). The large-scale popularization of such technological systems will greatly reduce transaction costs and optimize the overall operational efficiency of the tea supply chain.\u003c/p\u003e \u003cp\u003eMeasurement of macroeconomic Digital-Intelligent Transformation level\u003c/p\u003e \u003cp\u003eWhen Digital-Intelligent Transformation was still an emerging academic concept, Chinese researchers initially treated it as a simplistic juxtaposition of digitalization and intelligentization. Correspondingly, early evaluation indicator systems adopted a dual-dimensional structure that assessed digitalization and intelligentization independently. Luo and Chen (\u003cspan class=\"CitationRef\"\u003e2022\u003c/span\u003e) first construct a digitalization index by selecting indicators from four dimensions: digital infrastructure development, digital industrial development, digital technological innovation, and corporate digital transformation. They then build an intelligentization index using indicators across three aspects: intelligent technologies, intelligent outcomes, and competitiveness and benefits. Finally, they combine the two indices into a single digital-intelligent index using principal component analysis (PCA). Zhang and Bai (\u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e) adopt a similar calculation logic but differ in the construction of digitalization and intelligentization evaluation systems and their final aggregation method. They measure digitalization from three dimensions: digital infrastructure, digital platforms, and digital users. For intelligentization, they use the logarithmic number of patents granted in the artificial intelligence, 5G, and blockchain industries. They ultimately measure the level of Digital-Intelligent Transformation as the product of the two indices. Han et al. (\u003cspan class=\"CitationRef\"\u003e2025\u003c/span\u003e) employ a similar framework, with two key distinctions: they replace \"digital platforms\" with \"digital technologies\" in the digitalization evaluation system, and adopt the entropy weight method for the final index synthesis. Most subsequent studies follow Zhang and Bai's framework, and only revised partial sub-indicators and final aggregation approaches (Tang et al. \u003cspan class=\"CitationRef\"\u003e2025\u003c/span\u003e; Zhou et al. \u003cspan class=\"CitationRef\"\u003e2025\u003c/span\u003e). Notably, Liu et al. (\u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e) incorporate more sophisticated yet economically more convincing indicators—including the digital financial inclusion index and industrial robot penetration (Acemoglu and Restrepo \u003cspan class=\"CitationRef\"\u003e2020\u003c/span\u003e)—into their evaluation framework. Building on this work, Liu et al.(2025) further clarify the framework’s logic and expand it into a well-structured three-level indicator system. Furthermore, Lyu et al. (\u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e) draw on the strengths of prior studies by introducing Python web scraping and word frequency analysis, widely used in micro-level research, to develop a more comprehensive and precise regional digital-intelligent assessment system. As new-generation information technologies have become increasingly pervasive, researchers have gained a more sophisticated understanding of digital-intelligent transformation. Instead of treating it as a mere combination of digitization and intelligence, they now conceptualize it as a novel system arising from their organic integration. Recently, several scholars have transcended the two-dimensional framework by reconstructing the evaluation system with four first-level indicators: digital-intelligent infrastructure, digital-intelligent technology, digital-intelligent application, and digital-intelligent benefits. Using the entropy method, they measured the level of digital-intelligent transformation across prefecture-level cities (Yuan et al. \u003cspan class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eMeasurement of the Quantity-Quality Equilibrium Level in agricultural production\u003c/p\u003e \u003cp\u003eIn empirical economic studies, the quantity of agricultural products is generally measured by output and planting area, whereas the measurement of quality varies considerably depending on research perspectives and conceptual understandings of quality. At the micro level, researchers either directly measure product quality based on the sensory attributes of agricultural products within specific sectors (Nie et al. \u003cspan class=\"CitationRef\"\u003e2021\u003c/span\u003e), or indirectly infer quality through \u003cem class=\"Highlight ht71194251-f7a6-4c2d-a145-3d9f25b46662\" style=\"font-style: inherit;\"\u003esurvey\u003c/em\u003e-based consumer satisfaction statistics (Zhang and Wang \u003cspan class=\"CitationRef\"\u003e2025\u003c/span\u003e). From a macro-level perspective, when quality is defined as product safety and compliance with pesticide residue limits, it is commonly proxied by indicators such as the pass rate of random quality inspections, the quantity or share of organic agricultural products, and the number or proportion of registered geographical indication products (You and Wan \u003cspan class=\"CitationRef\"\u003e2022\u003c/span\u003e; George et al. \u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e). When quality is interpreted as systematic safety, nutrition, and premium characteristics guaranteed throughout all supply-side links, scholars establish standardized evaluation indices covering cultivation, processing, circulation, and marketing to indirectly characterize overall agricultural quality (Jing \u003cspan class=\"CitationRef\"\u003e2021\u003c/span\u003e). When quality is treated as a latent variable reflecting demand shifts, it is estimated using the residual term derived from inverse demand functions (Wei et al. \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e; Wang \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e). For studies that encompass all the above dimensions of quality, the total number of regional public agricultural brands is adopted as the proxy measure (Wang et al. \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e). Recently, one study has directly focused on the balance between agricultural product quantity and quality, using the coupling degree, coordination degree, and comprehensive evaluation index of the two as proxy variables (Shao et al. \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e). Based on production-side data, this method effectively captures the quantity-quality equilibrium of agricultural products. However, production decisions typically fluctuate between surplus and shortage over time. Such measures only reflect producers' autonomous choices and fail to adequately capture the realized quantity-quality outcomes shaped by market competition. Using circulation-side data that connects production and consumption, this paper proposes regional agricultural product brand value as a measure of the equilibrium level of quantity-quality decisions.\u003c/p\u003e \u003cp\u003eAlthough extant literature has thoroughly explored the roles of Digital-Intelligent Transformation in boosting agricultural output, efficiency, and income, \u003cem class=\"Highlight ht29216696-c42e-4f00-932a-aea34347df6a\" style=\"font-style: inherit;\"\u003etheoretical\u003c/em\u003e analyses and empirical studies on how such transformation mitigates the conflict between quantitative growth and quality improvement of agricultural products remain absent. Taking China's tea industry—a typical sector plagued by long-standing quantity-quality coordination dilemmas at the production stage—as an example, this paper investigates the impacts of Regional Digital-Intelligent Transformation (RDIT) on the Quantity-Quality Equilibrium Level of Tea Production (QQEL_TP) from an economic perspective. It aims to offer empirical evidence for policy design addressing the quantity-quality dilemma in agricultural production, while contributing to \u003cem class=\"Highlight ht29216696-c42e-4f00-932a-aea34347df6a\" style=\"font-style: inherit;\"\u003etheoretical\u003c/em\u003e development and methodological innovation.\u003c/p\u003e \u003cp\u003e\u003cem class=\"Highlight ht29216696-c42e-4f00-932a-aea34347df6a\" style=\"font-style: inherit;\"\u003eTheoretical\u003c/em\u003e analysis and research hypotheses\u003c/p\u003e \u003cp\u003eThe mathematical essence of the quality-quantity dilemma in agricultural production lies in how to maximize the total income of products by balancing the quantity and quality of production process inputs. Over a certain period, after undergoing sufficient market competition and receiving consumer feedback, agricultural producers will ultimately determine the most appropriate proportion of production factor inputs between quantity expansion and quality improvement to maximize total income. For a region, at this juncture, all agricultural producers' inputs in terms of quantity and quality factors reach a Nash equilibrium, which can be termed the Quantity-Quality Equilibrium in agricultural production. The better the decision-making coordination between agricultural producers regarding quantity expansion and quality improvement, the higher the level of Quantity-Quality Equilibrium in agricultural production. For a given technological level and a predominance of small-scale, decentralized farming, labor becomes the decisive factor shaping the Quantity-Quality Equilibrium Level in agricultural production. The more abundant the labor supply, the more effectively it can offset shortages in other production factors, thereby allowing coordinated expansion of output and improvement of quality even under the dual constraints of land endowment and capital investment. However, when labor outmigration results in scarcity, coordinated expansion of output and improvement of quality become unachievable. The relationship between quantity and quality in agricultural production thus shifts to a zero-sum game, which may in the long run degenerate into a negative-sum game featuring sacrificing quality for quantity. Against the overriding trend of advancing industrialization and urbanization, agricultural labor outmigration has become irreversible, rendering technological progress the pivotal approach to mitigating the quantity-quality dilemma in agricultural production.\u003c/p\u003e \u003cp\u003eAdvancements in traditional agricultural technology are mainly characterized by the independent adoption and direct application of individual production factors or pieces of technical equipment, with their systematic integration effect remaining relatively weak. Limited combinations of such factors are only observed among large-scale producers. By contrast, progress in digital-intelligent technology is distinguished by its move beyond reliance on a single device. It instead relies on public carriers such as 5G networks, cloud computing platforms, and intelligent control centers to build an integrated technical system. This model uses data to penetrate all links of the production process, enables cross-entity collaboration through platformization, and achieves the systematic integration of factors and procedures before terminal application. As a result, it can first release the systematic integration effect, promote the efficient use of technical equipment, and thereby reduce the application threshold for smallholder farmers. The large fixed investments required to develop public digital-intelligent infrastructure are mainly borne by governments, platform enterprises, and social capital. Smallholders can access digital-intelligent systems through lightweight terminals such as smartphones, thus sharing the benefits generated by integration effects. RDIT has therefore shown considerable potential to foster high-quality agricultural development. As a labor-intensive and high-value-added agricultural sector, the tea industry imposes strict demands on both the quantity and quality of labor input for premium tea production, resulting in a particularly acute quantity-quality dilemma. This study employs the tea industry as a \u003cem class=\"Highlight ht71194251-f7a6-4c2d-a145-3d9f25b46662\" style=\"font-style: inherit;\"\u003ecase\u003c/em\u003e to analyze the specific logic and mechanisms through which RDIT contributes to an improved Quantity-Quality Equilibrium Level in agricultural production.\u003c/p\u003e \u003cp\u003eHigh-quality tea production relies on considerable labor input and sophisticated craftsmanship, as tea leaves must be harvested and processed in a timely manner. Moreover, the cultivation cycle for skilled tea processors is lengthy, and the current stock of such talent is severely limited. During the peak tea-picking and processing season, shortages exist not only in ordinary tea pickers but also in skilled tea artisans, representing a structural labor imbalance. Although mixed cultivation of tea varieties with different maturity stages can alleviate the peak labor demand to some degree, the challenges of labor scarcity and rising labor costs in high-quality tea production remain pronounced nationwide. On the one hand, labor shortages directly push up production costs in the tea sector, forcing growers to loosen picking standards and deviate from optimal harvesting schedules. In addition, the absence of uniform post-harvest processing norms among smallholder farmers significantly impairs the consistency of the sensory quality of final tea products. On the other hand, labor shortages lead to insufficient routine management and maintenance of tea gardens, with some growers resorting to unregulated chemical applications, thereby creating potential food safety and quality risks. Existing mechanical harvesting and processing equipment cannot satisfy the strict quality requirements for premium and famous teas, pushing myopic smallholders toward quantity-oriented production strategies at the expense of quality. Given that the Chinese tea industry is dominated by smallholder operations, the persistent outflow of rural labor further intensifies the quantity-quality imbalance in tea production. Evidently, labor shortage represents the core constraint behind the quantity-quality dilemma in China's tea production system.\u003c/p\u003e \u003cp\u003eRDIT not only alleviates labor shortages but also directly improves the rationality of tea farmers'production decisions, thereby enhancing the Quantity-Quality Equilibrium Level of Tea Production (QQEL_TP). First, RDIT promotes the upgrading and openness of public service platforms, lowers the costs and spatiotemporal constraints of skill training, strengthens the knowledge spillover effect, and improves the quality of the existing labor supply, ultimately raising product quality without compromising quantity. The refinement and accessibility of expert knowledge systems and \u003cem class=\"Highlight ht29216696-c42e-4f00-932a-aea34347df6a\" style=\"font-style: inherit;\"\u003edisaster\u003c/em\u003e early-warning systems provide tea producers with access to advanced processing technologies and pest management strategies, thereby improving the quality of tea garden management and processing operations and consequently enhancing the quality of final tea products. Second, RDIT can expand the effective labor supply during the peak harvesting season by encouraging labor return and optimizing labor \u003cem class=\"Highlight ht29216696-c42e-4f00-932a-aea34347df6a\" style=\"font-style: inherit;\"\u003eallocation\u003c/em\u003e efficiency, thus safeguarding both the quantity and quality of tea products. This transformation is often accompanied by the rise of the platform economy and the expansion of cross-regional and cross-industry cooperation networks. The platform economy creates additional employment opportunities that attract labor back or support part-time participation. The expansion of cooperation networks helps break down information barriers in the labor market, improves the efficiency of labor matching, and enables famous tea-producing areas to quickly connect with idle labor resources in surrounding regions, thereby easing labor shortages during the peak picking period. Both factors increase the labor available to the tea sector, ensure compliance with harvesting standards and timely processing, and thus stabilize the quality and supply scale of finished tea products. Third, RDIT can correct distortions in the market mechanism and reverse small-scale producers' quantity-over-quality bias through the accurate transmission of price signals. The construction of a comprehensive quality traceability system is a core part of RDIT. It digitizes and transmits information to consumers, including tea origin, producer, batch, and grade, allowing high-quality tea to be recognized and priced at a premium. This eliminates the market distortion caused by information asymmetry and prevents the phenomenon of bad money driving out good. The accurate transmission of price signals enhances the rationality of tea farmers'production decisions, encouraging them to shift from short-term opportunistic behavior that sacrifices quality for quantity to long-term production strategies that balance quantity and quality. Meanwhile, regional data integration enables the government to implement targeted brand certification and differentiated subsidies, strengthening market incentive and restraint mechanisms. This further improves the rationality of tea farmers' production decisions and ultimately elevates QQEL_TP. Synthesizing the foregoing discussions, this paper puts forward the core \u003cem class=\"Highlight ht29216696-c42e-4f00-932a-aea34347df6a\" style=\"font-style: inherit;\"\u003etheoretical\u003c/em\u003e proposition as Hypothesis 1.\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eH1\u003c/strong\u003e \u003c/p\u003e\u003cp\u003eRegional Digital-Intelligent Transformation (RDIT) can significantly enhance the Quantity-Quality Equilibrium Level of Tea Production (QQEL_TP).\u003c/p\u003e \u003cp\u003e\u003c/p\u003e \u003cp\u003eThe effective realization of the above three pathways requires tea producers to actively access platforms or systems fostered by RDIT. Taobao Villages (TBV), as rural e-commerce industrial clusters developed on the basis of the platform economy, are themselves typical outcomes of RDIT permeating rural areas. Their emergence and development can enhance tea producers' willingness to participate in such transformation from two perspectives, thereby improving the QQEL_TP. First, TBV provide tea producers with sales channels directly targeting national consumers, reducing supply chain links and retaining more industrial chain profits. This in turn increases rural household income (Zhang et al. \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e), expands agricultural capital accumulation (Zhou et al. \u003cspan class=\"CitationRef\"\u003e2021\u003c/span\u003e), and equips local farmers with greater financial capacity and motivation to adopt digital-intelligent platforms. Second, to cope with market competition, TBV must proactively integrate into RDIT and connect with platforms or systems conducive to improving production and marketing performance. In doing so, they effectively guide and drive tea producers to understand, familiarize themselves with, and utilize public service platforms and collaborative networks, thereby improving production conditions. Existing research indicates that the development of TBV can significantly influence the quality and variety of products in the village's dominant industries (Zhang et al. \u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e). Furthermore, the development of TBV can be promoted by RDIT. During this transformation, the popularization of broadband networks, the construction of data centers, and the expansion of mobile payment services have benefited rural areas, providing digital-intelligent infrastructure support for the formation of TBV and enabling remote rural regions to access national and even global markets at low cost. Meanwhile, the diffusion and application of digital-intelligent technologies have reduced the costs of information searching and skills training, helping to improve farmers'digital literacy and stimulate their entrepreneurial vitality (Mei et al. \u003cspan class=\"CitationRef\"\u003e2020\u003c/span\u003e). This makes small-scale and decentralized rural e-commerce operations economically viable, thereby fostering the emergence of more Taobao Villages (Wu et al. \u003cspan class=\"CitationRef\"\u003e2020\u003c/span\u003e). Based on the above logical deduction, Hypothesis 2 is formally proposed.\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eH2\u003c/strong\u003e \u003c/p\u003e\u003cp\u003eRegional Digital-Intelligent Transformation (RDIT) improves the Quantity-Quality Equilibrium Level of Tea Production (QQEL_TP) by promoting Taobao Villages development (TBV).\u003c/p\u003e \u003cp\u003e\u003c/p\u003e \u003cp\u003eThe promotional effect of RDIT on QQEL_TP is not independent, and its effectiveness hinges on regional capacities for local technology absorption and conversion. To some extent, Government investment in science and technology (GI_ST) shapes local capabilities for technology absorption and translation, thereby moderating the degree to which RDIT elevates QQEL_TP. On the one hand, by providing financial support for scientific research projects and targeted subsidies for equipment purchases, GI_ST directly lowers the upfront cost barriers for tea researchers, farmers, and enterprises to adopt cutting-edge facilities such as the Internet of Things, intelligent sensors, and smart agricultural machinery. This accelerates the widespread deployment of digital-intelligent equipment across tea production and strengthens stakeholders' capabilities in the application, secondary development, and scenario-based utilization of digital-intelligent technologies. Furthermore, increased GI_ST focused on digital-intelligent R\u0026amp;D and talent cultivation sends positive policy signals, boosts market confidence, and attracts social capital, high-skilled labor, and computing infrastructure to support RDIT. This, in turn, amplifies the positive effect of RDIT on enhancing QQEL_TP. On the other hand, the moderating effect of GI_ST is not uniformly positive. Excessive public financial investment may give rise to a typical \"resource curse,\" turning the moderating effect negative. Some subsidy programs lack rigorous evaluation systems for actual technology application performance. Poorly designed project acceptance criteria or excessive administrative intervention in R\u0026amp;D activities tend to erode tea producers' incentives for independent exploration of efficient digital-intelligent solutions and autonomous innovation, slowing the diffusion of digital-intelligent technologies in tea production. Moreover, administration-led capital \u003cem class=\"Highlight ht29216696-c42e-4f00-932a-aea34347df6a\" style=\"font-style: inherit;\"\u003eallocation\u003c/em\u003e is prone to information lags and resource misallocation. When fiscal subsidy priorities diverge from the real operational needs of the tea industry, formalism emerges. Resources flow into inefficient projects or obsolete traditional capacity, leaving digital-intelligent equipment underutilized. Such distortions crowd out funds that market forces would otherwise channel to promising digital-intelligent innovators, hindering the deep integration of digital-intelligent technologies with core production links and ultimately dampening the marginal contribution of RDIT to improving QQEL_TP. Overall, the sign and magnitude of GI_ST's moderating role in the RDIT–QQEL_TP nexus remain theoretically ambiguous and await rigorous empirical \u003cem class=\"Highlight ht4fc55b9d-f515-4fa8-9e5d-6731d62f45ba\" style=\"font-style: inherit;\"\u003evalidation.\u003c/em\u003e Accordingly, Hypothesis 3 is proposed.\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eH3\u003c/strong\u003e \u003c/p\u003e\u003cp\u003eGovernment investment in science and technology (GI_ST) plays a significant moderating role in the influence of Regional Digital-Intelligent Transformation (RDIT) on the Quantity-Quality Equilibrium Level of Tea Production (QQEL_TP).\u003c/p\u003e \u003cp\u003e\u003c/p\u003e "},{"header":"Research design","content":"\u003cp\u003eModel specification\u003c/p\u003e\u003cp\u003eFollowing the approach of Li et al. (\u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e), this study constructs a two-way fixed effects model to test H1. This model can effectively control for omitted variable bias, individual heterogeneity, and time heterogeneity, thereby mitigating endogeneity concerns in panel data to a certain extent. To alleviate heteroskedasticity and facilitate the elastic interpretation of regression coefficients, the natural logarithm is applied to all continuous variables. The specific specification is as follows:\u003c/p\u003e\n\u003ch3\u003eln ln ln (1)\u003c/h3\u003e\n\u003cdiv class=\"Heading\"\u003eln\u003cem\u003eQQE_TP\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e=\u0026thinsp;α\u003c/em\u003e\u003csub\u003e\u003cem\u003e0\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;α\u003c/em\u003e\u003csub\u003e\u003cem\u003e1\u003c/em\u003e\u003c/sub\u003eln\u003cem\u003eRDIT\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;α\u003c/em\u003e\u003csub\u003e\u003cem\u003ej\u003c/em\u003e\u003c/sub\u003eln\u003cem\u003eZ\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;δ\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;\u0026micro;\u003c/em\u003e\u003csub\u003e\u003cem\u003et\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;ε\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e (1)\u003c/div\u003e \u003cp\u003eWhere \u003cem\u003eQQE_TP\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e refers to the Quantity-Quality Equilibrium Level of Tea Production of province \u003cem\u003ei\u003c/em\u003e in year \u003cem\u003et\u003c/em\u003e; \u003cem\u003eRDIT\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e denotes the Regional Digital-Intelligent Transformation Index for province \u003cem\u003ei\u003c/em\u003e in year \u003cem\u003et\u003c/em\u003e; \u003cem\u003eZ\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e is a vector of control variables. \u003cem\u003eδ\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e, \u003cem\u003e\u0026micro;\u003c/em\u003e\u003csub\u003e\u003cem\u003et\u003c/em\u003e\u003c/sub\u003e, and \u003cem\u003eε\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e represent individual fixed effects, time fixed effects, and the random error term, respectively.\u003c/p\u003e \u003cp\u003eBased on Eq.\u0026nbsp;(1), we further develop a mediating effect model to test H2, as below:\u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eln\u003cem\u003eTBV\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e=\u0026thinsp;β\u003c/em\u003e\u003csub\u003e\u003cem\u003e0\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;β\u003c/em\u003e\u003csub\u003e\u003cem\u003e1\u003c/em\u003e\u003c/sub\u003eln\u003cem\u003eRDIT\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;β\u003c/em\u003e\u003csub\u003e\u003cem\u003ej\u003c/em\u003e\u003c/sub\u003eln\u003cem\u003eZ\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;δ\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;\u0026micro;\u003c/em\u003e\u003csub\u003e\u003cem\u003et\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;ε\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e (2)\u003c/h2\u003e \u003cdiv id=\"Sec4\" class=\"Section3\"\u003e \u003ch2\u003eln\u003cem\u003eQQE_TP\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e=\u0026thinsp;γ\u003c/em\u003e\u003csub\u003e\u003cem\u003e0\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;γ\u003c/em\u003e\u003csub\u003e\u003cem\u003e1\u003c/em\u003e\u003c/sub\u003eln\u003cem\u003eRDIT\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;γ\u003c/em\u003e\u003csub\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sub\u003eln\u003cem\u003eTBV\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;γ\u003c/em\u003e\u003csub\u003e\u003cem\u003ej\u003c/em\u003e\u003c/sub\u003e\u003cem\u003eZ\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;δ\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;\u0026micro;\u003c/em\u003e\u003csub\u003e\u003cem\u003et\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;ε\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e (3)\u003c/h2\u003e \u003cp\u003eWhere \u003cem\u003eTBV\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e represents the level of Taobao Villages development in province \u003cem\u003ei\u003c/em\u003e in year \u003cem\u003et\u003c/em\u003e. Eq.\u0026nbsp;(2) is first adopted to estimate the impact of RDIT on TBV. Subsequently, Eq.\u0026nbsp;(3) is used to assess how RDIT influences QQEL_TP after incorporating TBV as a controlled variable. The total mediating effect is expressed as \u003cem\u003eβ\u003c/em\u003e\u003csub\u003e\u003cem\u003e1\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e\u0026times;\u0026thinsp;γ\u003c/em\u003e\u003csub\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sub\u003e. Considering that the sampling distribution of the mediating effect estimator may fail to follow a normal distribution, this paper applies the Bootstrap approach to conduct repeated resampling trials and determines statistical significance according to bias-corrected confidence intervals, thereby improving the overall robustness of the mechanism examination results.\u003c/p\u003e \u003cp\u003eTo test H3, a moderating effect model is further formulated on the basis of Eq.\u0026nbsp;(1), as below:\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e\n\u003ch3\u003eln ln (ln×ln) (4)\u003c/h3\u003e\n\u003cdiv class=\"Heading\"\u003eln\u003cem\u003eQQE_TP\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e=\u0026thinsp;η\u003c/em\u003e\u003csub\u003e\u003cem\u003e0\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;η\u003c/em\u003e\u003csub\u003e\u003cem\u003e1\u003c/em\u003e\u003c/sub\u003e\u003cem\u003eRDIT\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;η\u003c/em\u003e\u003csub\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sub\u003eln\u003cem\u003eGI_ST\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;η\u003c/em\u003e\u003csub\u003e\u003cem\u003e3\u003c/em\u003e\u003c/sub\u003e(ln\u003cem\u003eRDIT\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e\u0026times;ln\u003cem\u003eGI_ST\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e)\u003cem\u003e+η\u003c/em\u003e\u003csub\u003e\u003cem\u003ej\u003c/em\u003e\u003c/sub\u003e\u003cem\u003eZ\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;δ\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;\u0026micro;\u003c/em\u003e\u003csub\u003e\u003cem\u003et\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;ε\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e (4)\u003c/div\u003e \u003cp\u003eWhere \u003cem\u003eGI_ST\u003c/em\u003e\u003csub\u003e\u003cem\u003eit\u003c/em\u003e\u003c/sub\u003e represents government investment in science and technology of province \u003cem\u003ei\u003c/em\u003e in year \u003cem\u003et\u003c/em\u003e, and the coefficient \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003e3\u003c/em\u003e\u003c/sub\u003e of the interaction term characterizes the direction and intensity of the moderating effect.\u003c/p\u003e \u003cp\u003eGiven that the meaningful variation in both mediating and moderating variables primarily arises from cross-regional disparities and gradual temporal evolution, controlling for both individual and time fixed effects would absorb much of their core variation, leading to insufficient identifying information for mechanism analysis and overly conservative estimates. Accordingly, four model specifications are adopted to test the mediating and moderating effects: no fixed effects, time fixed effects only, provincial fixed effects only, and both time and provincial fixed effects. The results are mainly used for exploratory analysis and preliminary validation of the proposed mechanism channels.\u003c/p\u003e \u003cp\u003eVariable selection and data sources\u003c/p\u003e \u003cp\u003eExtant studies identify 2012 as the starting point of China's digital agriculture development (Zhu \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). In April 2012, the China Tea E-commerce Alliance was established in Xinchang, Zhejiang, marking a new era of self-regulated and standardized development for China's tea e-commerce industry (Yuan \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). In June 2012, the Chinese government rolled out the \"Broadband China\" strategy to boost digital development and prioritize next-generation information infrastructure (The State Council 2012). Taken together, the profound influence of Digital-Intelligent Transformation on agricultural production, the standardized operational maturity of Taobao Villages in major tea-growing regions, and the targeted tilt of government technological funding toward digital-intelligent sectors all converged around the year 2012. Accordingly, this study sets 2012 as the starting point of its research timeline. Owing to divergent release frequencies and update cycles across datasets, the latest consistent and complete observations are available up to 2023 after rigorous data integration. The empirical analysis thus employs a panel dataset covering the period 2012\u0026ndash;2023.\u003c/p\u003e\n\u003ch3\u003eDependent variable\u003c/h3\u003e\n\u003cp\u003eQQEL_TP is measured by the value of regional public tea brands. As theorized earlier, better alignment between production scale and quality enhances industrial reputation and is ultimately reflected in brand value, which therefore serves as a valid proxy for QQEL_TP. Brand value data are retrieved from the annual Evaluation Report on the Value of China's Regional Public Tea Brands, published systematically by the China Agricultural Brand Research Center. This evaluation represents the only authoritative third-party assessment widely recognized in China's tea industry (Huang and Sun \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The original dataset covers major tea-producing provinces in China, including 215 brands and 1,456 raw observations. Brand-level data are aggregated to the provincial level by value-weighted averaging, resulting in a balanced panel of 15 provinces and 180 valid observations.\u003c/p\u003e\n\u003ch3\u003eCore explanatory variable\u003c/h3\u003e\n\u003cp\u003eRDIT is measured by the regional digital-intelligent transformation index constructed by the authors. Guided by the connotation of Digital-Intelligent Transformation, we build a comprehensive evaluation system consisting of 3 primary indicators and 11 secondary indicators. From a technological perspective, Digital-Intelligent Transformation is not merely a simple combination of digitization and intelligence. Rather, it constitutes an evolving state of information technology that progresses from basic digitization toward full-fledged intelligence, wherein digitization and intelligence become organically integrated and mutually reinforcing. From an economic standpoint, Digital-Intelligent Transformation encompasses the full spectrum of advances in and applications of next-generation information technologies. Building on digitization, it leverages increasingly sophisticated computing power and algorithms to collect and integrate data from across all dimensions of socioeconomic activities. By fully unlocking data value, it continuously enhances productivity and reshapes production modes accordingly, ultimately propelling human society toward an intelligent state in which all economic activities enable autonomous decision-making, automatic execution, and self-optimization. Accordingly, the regional digital-intelligent transformation index is constructed from three dimensions: public foundation, application level, and economic benefits. First, the public foundation dimension incorporates intelligent elements beyond conventional digitalization, covering basic infrastructure, investment, human capital, as well as advanced computing facilities and R\u0026amp;D. Second, consistent with Marx's social reproduction theory, the application level is assessed across production, circulation, distribution, and consumption. Third, it captures the economic benefits generated by RDIT through both supply-side gains and demand satisfaction. The evaluation indicator system for RDIT is presented in Table\u0026nbsp;1.\u003c/p\u003e \u003cp\u003eIn selecting proxy variables, we follow standard literature practices but introduce two tailored adjustments for our research objectives. First, we use patent applications for new-generation information technology to measure technological R\u0026amp;D, rather than aggregating patents across digital technology, 5G, AI, and blockchain. With the rapid diffusion and innovation of information technology, these traditional indicators fail to adequately capture emerging digital-intelligent technologies. New-generation information technology more accurately represents core enabling technologies such as cloud computing, IoT, VR, AR, and the BeiDou Navigation System, which are often not covered in prior studies. Second, we measure supply-side gains using gross profit in the manufacturing of computers, communications, and electronic devices, instead of revenue, operating profit, or total profit. While revenue reflects supply scale, it is highly volatile to market prices and demand shocks and cannot identify technological progress or cost efficiency. The digital industry features high R\u0026amp;D intensity, fast fixed-asset depreciation, and strong subsidy dependence, making operating and total profits prone to policy and accounting distortions. By contrast, gross profit better captures cost reductions and efficiency gains driven by technological innovation, providing a more reasonable measure of supply-side gains.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eControl variables\u003c/h2\u003e \u003cp\u003eRegional consumption level (RCL), agricultural mechanization level (AML), and logistics accessibility (LA) are selected as control variables. Based on Keynes' effective demand theory, a higher regional consumption level expands market demand for tea products, encouraging tea enterprises to expand production and optimize operations, thus positively promoting improvements in QQEL_TP. From the perspective of induced technological change theory, agricultural mechanization substitutes labor inputs, improves operational accuracy and efficiency, and supports coordinated growth in tea yield and quality. Grounded in new economic geography theory, efficient logistics networks lower transport costs, shorten circulation cycles, enhance flexible production scheduling, and further optimize quantity-quality decision-making performance. Specifically, RCL is measured as the share of total retail sales of consumer goods in regional GDP; AML is proxied by total agricultural machinery power; and LA is represented by total freight volume. All control variable data are obtained from provincial statistical yearbooks.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eInstrumental variable\u003c/h3\u003e\n\u003cp\u003eThe number of new-generation information technology patent applications filed between 1985 and 1996 is chosen as the instrumental variable (IV). On the one hand, the historical patent stock reflects the evolution and penetration of foundational digital-intelligent technologies, which directly determines the future direction and scale of RDIT, thus satisfying the relevance condition. On the other hand, given the lengthy time span of more than two decades, early information technology patents exert negligible exogenous effects on current tea sector quantity-quality production decisions, strictly meeting the exogeneity requirement for valid instrumental variables.\u003c/p\u003e\n\u003ch3\u003eMechanism variables\u003c/h3\u003e\n\u003cp\u003eConsistent with Hypothesis 2, Taobao Village development (TBV) is treated as the mediating variable. Limited by data availability, the number of Taobao Villages is used to measure its development, with raw data obtained from the \u003cem\u003eAnnual Research Report on China's Taobao Villages\u003c/em\u003e issued by the Alibaba Research Institute. In line with Hypothesis 3, government investment in science and technology (GI_ST) serves as the moderating variable, proxied by provincial government expenditure on science and technology, with relevant data collected from the National Bureau of Statistics of China.\u003c/p\u003e \u003cp\u003eDescriptive statistics and data processing of variables\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eDescriptive statistics of variables\u003c/h2\u003e \u003cp\u003eDescriptive statistics of all variables are shown in Table\u0026nbsp;2.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eData processing\u003c/h2\u003e \u003cp\u003eFirst, all nominal variables are deflated to constant 2012 prices. Second, sporadic missing values are imputed using interpolation. Finally, all processed variables are log-transformed prior to regression analysis. Excel is used for preliminary data cleaning and preprocessing, while Stata is employed for baseline regressions, robustness checks, and endogeneity correction. The KMO statistic for the RDIT indicator system exceeds 0.8, indicating strong correlations among dimensions. PCA is thus used to construct the aggregate Regional Digital and Intelligent Transformation Index. In addition, the measurement of industrial robot penetration involves complex calculations, which are detailed below.\u003c/p\u003e \u003cp\u003eFirst, using the ISIC Rev.4 manufacturing classification as a bridge, this study matches industrial robot categories from the International Federation of Robotics (IFR) (M\u0026uuml;ller and Christopher, 2025) to China's Industrial Classification for National Economic Activities (GB/T 4754\u0026thinsp;\u0026minus;\u0026thinsp;2017) \u003csup\u003e①\u003c/sup\u003e. A concordance table is constructed covering 29 Chinese manufacturing sub-industries (C13\u0026ndash;C41)\u003csup\u003e②\u003c/sup\u003e and 13 corresponding IFR sectors. Second, we aggregate the cumulative stock of industrial robots at the industry level in China for 2012\u0026ndash;2023 according to this concordance. Third, taking 2013 as the base year\u003csup\u003e③\u003c/sup\u003e and drawing on the frameworks of Acemoglu and Restrepo (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), Wang and Dong (\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), and Zhang et al. (\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), we construct a province-year measure of manufacturing robot penetration. The specific formula is given below:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:\\text{E}\\text{Exposure}\\text{\\:\u0026minus;}\\text{CH}\\text{}\\text{=}\\sum\\:_{\\text{i}\\in\\text{I}}\\frac{\\text{em}{\\text{p}}_{\\text{ji}\\text{,}\\text{t}\\text{=201}\\text{3}}}{\\text{em}{\\text{p}}_{\\text{j}\\text{,}\\text{t}\\text{=201}\\text{3}}}\\frac{\\text{M}{\\text{R}}_{\\text{it}}^{\\text{CH}}}{{\\text{L}}_{\\text{i}\\text{,}\\text{t}\\text{=201}\\text{3}}^{\\text{CH}}}\\text{}\\text{}\\text{}\\text{(5)}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eWhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{MR}}_{\\text{it}}^{\\text{CH}}\\)\u003c/span\u003e\u003c/span\u003e stands for the stock of industrial robots in China's industrial sector \u003cem\u003ei\u003c/em\u003e in year \u003cem\u003et\u003c/em\u003e; \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{L}}_{\\text{i}\\text{,}\\text{t}\\text{=201}\\text{3}}^{\\text{CH}}\\)\u003c/span\u003e\u003c/span\u003e​denotes the national baseline employment of sector \u003cem\u003ei\u003c/em\u003e in 2013. \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\frac{\\text{em}{\\text{p}}_{\\text{ji}\\text{,}\\text{t}\\text{=201}\\text{3}}}{\\text{em}{\\text{p}}_{\\text{j}\\text{,}\\text{t}\\text{=201}\\text{3}}}\\)\u003c/span\u003e\u003c/span\u003e represents the share of sector i's employment in total employment of province \u003cem\u003ej\u003c/em\u003e in the same base year, which captures the regional manufacturing employment structure. Data on China's baseline employment are obtained from the \u003cem\u003e2013 China Economic Census Yearbook\u003c/em\u003e.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results and discussion","content":"\u003cp\u003eSpatiotemporal characteristics of RDIT and QQEL_TP\u003c/p\u003e \u003cp\u003eTo clearly depict the spatiotemporal evolutionary features of the RDIT and the QQEL_TP across major tea-producing provinces in China, this paper conducts analyses from both static and dynamic dimensions. MATLAB is employed to plot kernel density distribution curves for the period 2012\u0026ndash;2023, and Stata is utilized to generate spatiotemporal evolution trend maps for the representative years of 2012, 2017, and 2023. A geographical location map (Fig.\u0026nbsp;1) is provided below to identify the actual spatial distribution of each sample province across China's territory.\u003c/p\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eTemporal and spatial evolution of RDIT\u003c/h2\u003e \u003cp\u003eFrom a temporal perspective (Fig.\u0026nbsp;2), RDIT across China's major tea-producing provinces shows steady improvement and divergent development. Before 2014, the main peak was narrow and high, the secondary peak was indistinct, and both were concentrated in the low-level interval, indicating that the overall RDIT of the sample provinces was generally low. This is consistent with weak infrastructure in the early stage of Digital-Intelligent Transformation. Over time, both the main and secondary peaks shifted continuously to the right, with the height of the main peak declining and the secondary peak rising, suggesting that the clustering of samples at low digital and intelligent transformation levels gradually weakened, while the proportion of high-level regions increased steadily. After 2020, a third peak emerged, and all peaks continued to shift rightward, indicating that the digital and intelligent transformation level of the samples kept improving and regional divergence became more pronounced. Distinct high, medium, and low tiers have formed, reflecting that regional digital and intelligent transformation has entered a stage of normalized development, with the effects of policy promotion and technology diffusion gradually becoming evident.\u003c/p\u003e \u003cp\u003eFrom a spatial perspective (Fig.\u0026nbsp;3), the regional digital and intelligent transformation (RDIT) of China's major tea-producing provinces exhibits a clear gradient distribution. In all sample years, Zhejiang, Jiangsu, and Guangdong consistently rank at the forefront, while Shandong and Fujian remain in the medium-high group. Although Guangxi and Yunnan have improved gradually over the sample period, they still fall into the low-RDIT category, reflecting the advantage of coastal locations in advancing Digital-Intelligent Transformation. Sichuan and Anhui show considerable temporal shifts in their gradient positions. Benefiting from technological spillovers from neighboring high-RDIT provinces, Anhui has steadily risen from the medium-low group in 2012 to the medium-high group. Sichuan has advanced remarkably from a low level in 2012 to the medium‑high group, and given the low RDIT performance of most surrounding provinces, this improvement can be largely attributed to targeted policy support. By contrast, Hubei and Hunan have experienced continuous declines, falling from medium-high to medium-low levels. Jiangxi, despite its proximity to high-performing provinces, has long remained trapped in the low-RDIT group. Overall, these patterns indicate that sustained and targeted regional policies are essential for achieving inclusive Digital-Intelligent Transformation.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eSpatiotemporal evolution of QQEL_TP\u003c/h2\u003e \u003cp\u003eFrom a temporal perspective (Fig.\u0026nbsp;4), the distribution of QQEL_TP across China's major tea-producing provinces shows a pattern of gradual improvement and increasing concentration. During 2012\u0026ndash;2016, the main peak was narrow and high, concentrated in the low-value range, with two distinct secondary peaks in the medium- and high-value ranges. This suggests a highly dispersed tripartite pattern in quantity-quality equilibrium with an overall low level in the early period. Over time, the main peak declined, and the two secondary peaks converged and merged by 2017. Although the secondary peak rose moderately afterward, it shifted left and right intermittently, indicating that QQEL_TP gradually converged but improved slowly and unstably. This evolution reflects the complexity of provincial quantity-quality decisions. Initially, most provinces prioritized quantity, a few balanced quantity and quality, and only a small number emphasized quality primacy. With rising consumer quality awareness and stricter agricultural product quality policies, most provinces shifted toward balancing quantity and quality. Amid fiercer market competition, even quality-oriented provinces began to adopt a balanced strategy and develop mid-to-low-end products to sustain profits. The diffusion of digital-intelligent technologies has to some extent accelerated this transition.\u003c/p\u003e \u003cp\u003eAs illustrated in the spatial evolution map (Fig.\u0026nbsp;5), the distribution of QQEL_TP has transitioned from regional polarization to convergent development, reflecting strengthened spatial convergence and spillover effects among major tea-producing provinces. At the provincial level, Zhejiang and Fujian have consistently taken the lead. Notably, Guangdong, a neighboring province with China's highest GDP, has remained at a relatively low level. This suggests that unobservable factors beyond provincial economic development exert a significant influence on QQEL_TP, which may be closely related to the reputation of historical famous teas: 60% of China's top ten famous teas originate from Zhejiang and Fujian. Therefore, to rigorously examine whether RDIT can improve QQEL_TP, it is essential to control for provincial fixed effects to mitigate unobservable regional heterogeneity.\u003c/p\u003e \u003cp\u003eBaseline regression results\u003c/p\u003e \u003cp\u003eBaseline regression results for Hypothesis 1 are presented in Table\u0026nbsp;3. Columns (1)\u0026ndash;(5) sequentially add controls and various fixed effects, with column (5) representing our preferred two-way fixed effects specification.\u003c/p\u003e \u003cp\u003eBaseline results confirm that RDIT exerts a significant positive effect on QQEL_TP, with coefficient robustness strengthening as model specification improves. Even in the parsimonious specification without controls, the core coefficient remains positive and significant, indicating that RDIT inherently boosts QQEL_TP. Adding control variables further increases the magnitude of the core coefficient, implying that omitting factors such as regional consumption, agricultural mechanization, and logistics accessibility leads to downward bias in estimating RDIT's true impact. Among controls, RCL is significantly positive, consistent with demand-pull theory. Greater domestic demand improves supply\u0026ndash;demand matching through Taobao Village\u0026ndash;driven platforms, promoting quality upgrading and production efficiency. By contrast, AML and LA are statistically insignificant, likely reflecting tea's unique production characteristics: tea cultivation relies little on standard farm machinery and general logistics, so their marginal effects are overshadowed by digital-intelligent gains. Further controlling for provincial and time fixed effects substantially increases the core coefficient, which remains significant at the 5% level. This suggests that unobserved regional heterogeneity and aggregate time shocks distort baseline estimates, and removing such confounding factors enables cleaner identification of RDIT's net effect. Overall, these findings strongly support that RDIT significantly improves QQEL_TP. In our preferred two-way fixed-effects specification, a 1% increase in RDIT raises QQEL_TP by 0.432%.\u003c/p\u003e \u003cp\u003eEndogeneity Test\u003c/p\u003e \u003cp\u003eLimited by access to industry-specific panel data for tea, our empirical model may suffer from endogeneity concerns, including omitted unobservables and measurement errors arising from inconsistent statistical standards across databases. In addition, significant bidirectional causality exists between RDIT and QQEL_TP. On the one hand, higher RDIT maturity improves QQEL_TP; on the other, regions with better QQEL_TP tend to have greater incentives and capacity to invest in digital-intelligent infrastructure, further promoting RDIT. To address these endogeneity issues, we adopt an instrumental variable (IV) approach, with results reported in Table\u0026nbsp;4.\u003c/p\u003e \u003cp\u003eIn the first-stage regression, the coefficient of the instrumental variable is significant at the 1% level, with an F-statistic of 50.05, well above conventional critical values, thus ruling out weak instrument concerns. In the second-stage estimation, the coefficient of RDIT is 1.189 and significant at the 5% level, representing a substantially larger marginal effect relative to baseline estimates. The LM underidentification test rejects the null hypothesis at the 1% level, further confirming the statistical adequacy of the instrument. Two-stage estimation results validate the effectiveness of the instrumental variable design. The instrument is strongly correlated with RDIT and exogenous to the disturbance term, effectively mitigating estimation bias driven by reverse causality and omitted variables. After accounting for endogeneity, RDIT still exerts a significant positive influence on QQEL_TP, with an increased coefficient magnitude compared with baseline findings. This reinforces the causal inference that RDIT improves QQEL_TP, and indicates that baseline results tend to understate the actual promoting effect. Overall, although neglecting endogeneity leads to conservative coefficient estimates, the core conclusion that RDIT enhances QQEL_TP remains robust under a more rigorous causal identification framework.\u003c/p\u003e \u003cp\u003eRobustness checks\u003c/p\u003e \u003cp\u003eThis study employs three systematic robustness strategies, with results presented in Table\u0026nbsp;5. First, considering that the COVID‑19 pandemic substantially disrupted macroeconomic conditions from 2020 to 2022, observations in these abnormal years are excluded. The coefficient of the core independent variable remains positive and significant, with magnitude and sign highly consistent with baseline estimates. The positive effect of RDIT on QQEL_TP remains robust even after removing approximately one-quarter of the sample, confirming that the baseline results are not driven by temporary macroeconomic shocks or extreme public health events. Second, to mitigate bias caused by extreme outliers in continuous variables, all continuous indicators are winsorized at the 5th and 95th percentiles. Re-estimation yields a significantly positive and stable coefficient for the core explanatory variable, indicating that the main findings are not affected by extreme observations. Third, the logarithmic transformation of the normalized core explanatory variable generates missing values where original entries are zero, reducing the effective sample to 179. To rule out estimation distortion caused by such data truncation, zero values are replaced with 0.001 before regression. The sign and significance of the RDIT coefficient remain unchanged, further verifying that the conclusions are robust to minor data adjustments and not dependent on specific data-processing rules.\u003c/p\u003e \u003cp\u003eMediating effect analysis\u003c/p\u003e \u003cp\u003eTo explore the mechanism through which RDIT affects QQEL_TP, the bootstrap method is employed to test the mediating role of TBV. As shown in Table\u0026nbsp;6, the indirect effect is significantly positive in the specification without fixed effects, supporting the mediating channel: RDIT improves QQEL_TP by fostering TBV development. This positive indirect effect remains robust after controlling for provincial fixed effects, indicating that the transmission path RDIT \u0026rarr; TBV \u0026rarr; QQEL_TP is stable even after accounting for unobserved regional heterogeneity. By contrast, the mediating effect becomes statistically insignificant when only time fixed effects or two-way fixed effects are included. This implies that the mediating role of TBV is sensitive to time-varying shocks arising from macro policy adjustments and market fluctuations. Overall, RDIT enhances QQEL_TP partially through the promotion of TBV. Further analysis of direct effects across specifications yields additional implications. In the baseline model without fixed effects, the direct effect of RDIT on QQEL_TP is significantly negative, whereas the total effect is insignificant. This suggests the existence of other unobserved mediating channels that may exert adverse influences on QQEL_TP. After controlling for time, provincial, or two-way fixed effects, both direct and total effects become insignificant. These results indicate that the relationship between RDIT and QQEL_TP is characterized by complex mechanisms and high sensitivity to time-varying heterogeneous shocks.\u003c/p\u003e \u003cp\u003eModerating effect analysis\u003c/p\u003e \u003cp\u003eTo examine the contextual boundary conditions under which RDIT affects QQEL_TP, a moderating effect model is estimated using GI_ST as the moderating variable. As shown in Table\u0026nbsp;7, the interaction term is significantly negative when both fixed effects are omitted or only time fixed effects are included (Columns 1 and 2). When provincial fixed effects or two-way fixed effects are introduced (Columns 3 and 4), the interaction term remains negative but loses statistical significance. These results indicate that GI_ST exerts a weakly negative moderating effect on the relationship between RDIT and QQEL_TP. The moderating effect diminishes substantially after accounting for unobserved provincial heterogeneity, suggesting that GI_ST is characterized by notable regional heterogeneity. In general, the positive impact of RDIT on QQEL_TP is weakened under high levels of GI_ST. This pattern can be explained by the insufficiently mature digital-intelligent application scenarios in the tea industry. Excessive fiscal investment and resource misalignment prevent public funding from being effectively converted into industry-specific technologies. Instead, the mismatch between advanced digital technologies and on-farm production needs creates an inhibitory effect. These findings also point to widespread resource redundancy and low technology conversion efficiency in the current allocation of public fiscal resources for science and technology.\u003c/p\u003e \u003cp\u003eConclusions, policy implications, and research limitations\u003c/p\u003e \u003c/div\u003e"},{"header":"Conclusions","content":"\u003cp\u003eDigital-Intelligent technologies have been extensively integrated into contemporary agricultural production systems. A large body of natural science experiments has confirmed that such technologies can simultaneously enhance both the output quantity and product quality in agricultural practices. However, constrained by difficulties in obtaining qualified measurable proxy data in economic empirical research, existing studies mostly examine the agricultural quantity-quality dilemma from fragmented and isolated perspectives. Few scholars have directly focused on the integrated quantity-quality equilibrium level, and whether RDIT can effectively alleviate this long-standing agricultural dilemma still lacks solid empirical validation. This study proposes using the value of regional public tea brands as a proxy indicator to reflect the degree of relief from the agricultural quantity-quality dilemma, thereby accurately characterizing QQEL_TP. Taking the tea industry\u0026mdash;where the production-side quantity-quality contradiction is particularly prominent\u0026mdash;as the research context, this paper clarifies the theoretical mechanisms and quantifies the empirical effects of how RDIT promotes the coordination of agricultural quantity and quality.\u003c/p\u003e \u003cp\u003eThe core conclusions are as follows. (1) From 2012 to 2023, the overall RDIT level across major tea-producing provinces in China maintained steady growth, while QQEL_TP remained generally low with moderate improvement momentum. Spatially, significant inter-provincial disparities persisted for both variables; RDIT exhibited a clear gradient-driven development pattern, and the spatial distribution of QQEL_TP evolved from scattered imbalance in the early stage to a relatively concentrated layout characterized by coordinated regional development. (2) RDIT exerts a significantly positive impact on QQEL_TP. After controlling for individual heterogeneity, temporal shocks, and other confounding factors, a 1% increase in RDIT is associated with a notable 0.432% rise in QQEL_TP. (3) The core finding that RDIT significantly improves QQEL_TP remains highly robust after addressing endogeneity through the instrumental variable approach, excluding samples from abnormal pandemic years, and adjusting sample distributions via standardized statistical treatments. (4) RDIT indirectly elevates QQEL_TP by promoting the development of TBV, and this mediating pathway is sensitive to time-dependent macroeconomic fluctuations. (5) The enhancing effect of RDIT on QQEL_TP is negatively moderated by GI_ST, and such moderating characteristics display sensitivity to cross-regional heterogeneous variations.\u003c/p\u003e \u003cp\u003ePolicy implications\u003c/p\u003e \u003cp\u003eThe empirical findings based on the tea industry provide targeted policy implications for addressing the widespread agricultural quantity-quality dilemma.\u003c/p\u003e \u003cp\u003eFirst, the agricultural quantity-quality equilibrium should be embedded into the core objective system of high-quality agricultural development. Policymakers should optimize evaluation and incentive mechanisms that prioritize quality improvement, operational efficiency, and long-term sustainability. They should guide smallholder farmers to actively participate in the construction of regional public brand systems and promote agricultural development strategies that balance output expansion and quality enhancement.\u003c/p\u003e \u003cp\u003eSecond, formulate categorized and differentiated policies to promote the implementation of RDIT, and leverage the integrated empowerment dividends of digital intelligence as the core driver to enhance the synergy between agricultural quantity and quality. For provinces with mature digital foundations such as Zhejiang, Jiangsu, and Guangdong, priorities should be placed on the deep integration of digital-intelligent technologies into the entire agricultural industrial chain, as well as the large-scale deployment of intelligent sensing equipment and precision farming systems across all production links. For less developed regions, including Guangxi and Yunnan, where digital infrastructure lags behind, policymakers should first address shortages in basic facilities, popularize low-cost and replicable technology application models, and establish cross-regional cooperation mechanisms to accelerate technology spillovers and practical experience sharing, so as to gradually narrow inter-regional development gaps. For provinces with moderate RDIT maturity, emphasis should be placed on building high-end digital-intelligent supporting facilities, breaking institutional barriers that restrict the free flow of data, and giving full play to the pivotal role of data factors in forecasting external changes in agricultural markets and improving the efficiency of cross-regional supply-demand matching.\u003c/p\u003e \u003cp\u003eThird, continuously upgrade the comprehensive ecosystem of rural e-commerce. By improving logistics service networks for agricultural products and enhancing the digital marketing capabilities of rural practitioners, traditional e-commerce platforms can be upgraded from simple transaction terminals to multi-functional comprehensive service carriers, further strengthening the mediating role of TBV in boosting the positive nexus between RDIT progress and the improvement of agricultural quantity-quality equilibrium.\u003c/p\u003e \u003cp\u003eFourth, restructure the allocation mechanism of public fiscal resources for science and technology and alleviate the crowding-out effect caused by inappropriate government intervention on the autonomous digital investment of market participants. Given that excessive GI_ST has exerted certain inhibitory distortions on the effectiveness of RDIT, it is advisable to formulate practical project evaluation criteria based on actual agricultural production needs, adopt third-party independent performance evaluation and competitive funding allocation mechanisms, curb resource misallocation and redundant infrastructure investment, and ultimately unlock the positive moderating potential of rational government S\u0026amp;T input in fostering RDIT-driven optimization of QQEL_TP.\u003c/p\u003e \u003cp\u003eResearch limitations\u003c/p\u003e \u003cp\u003eThis study provides meaningful theoretical references and practical guidance for alleviating the agricultural quantity-quality dilemma, yet several inevitable limitations remain. Firstly, the analytical framework is rooted in the reality of China's smallholder-dominated decentralized farming, which limits the generalizability of the core conclusions to agricultural economies dominated by large-scale commercial farms. Secondly, inconsistent data release timelines across multi-dimensional evaluation indicators restrict the estimation of the RDIT Index to 2023 only, failing to cover the period 2024\u0026ndash;2025\u0026mdash;a phase characterized by further upgrades in digital intelligence penetration. Such data limitations may underestimate the true positive promotional effect of RDIT on the improvement of QQEL_TP. Thirdly, constrained by the unavailability of operational scale statistics, this paper only uses the number of Taobao Villages to characterize the development of TBV, which inevitably weakens the statistical robustness of the mediating effect tests to a certain extent. Finally, this research mainly focuses on verifying whether RDIT can effectively address the agricultural quantity-quality dilemma through theoretical deduction and empirical modeling, while in-depth exploration of the complex underlying influencing mechanisms remains insufficient. Both mediating and moderating regression results confirm the existence of complex heterogeneous transmission pathways between RDIT evolution and QQEL_TP improvement. Accordingly, future research can conduct more in-depth theoretical elaboration and refined empirical tests to systematically unpack the sophisticated internal driving logic.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAugustin A, Kiliroor CC (2025) EXPRESS: Advanced Hyperspectral Signature Processing for Chemical Stress Detection in Vegetable Leaves Using Hierarchical Feature Extraction and Enhanced Ensemble Model. App spec 37028251411953. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/00037028251411953\u003c/span\u003e\u003cspan address=\"10.1177/00037028251411953\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAcemoglu D, Restrepo P (2020) Robots and Jobs: Evidencefrom US Labor Markets. J Poll Eco 128(6):2188\u0026ndash;2244. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1086/705716\u003c/span\u003e\u003cspan address=\"10.1086/705716\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDong YH, CHENG WJ, ZHANG XY (2010) Analysis on the causes and Countermeasures of quality problems of famous green tea. Chinese Tea, 32(01):27\u0026ndash;28. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/CNKI:SUN:CAYA.0.2010-01-012\u003c/span\u003e\u003cspan address=\"https://doi.org/CNKI:SUN:CAYA.0.2010-01-012\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDong YH, Yang CX, Lv JD (2014) Reasons for the decline of the quality of Yongxi Huoqing tea and its improvement measures. Chinese Tea 36(03):25\u0026ndash;26. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/CNKI:SUN:CAYA.0.2014-03-012\u003c/span\u003e\u003cspan address=\"https://doi.org/CNKI:SUN:CAYA.0.2014-03-012\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDong Y, Guo HW, Zhang JH (2025) The impact of digital and intelligent transformation on the innovation The impact of digital and intelligent transformation on the innovation. Res Agri Moder 46(03):495\u0026ndash;505. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.13872/j.1000-0275.2025.0174\u003c/span\u003e\u003cspan address=\"10.13872/j.1000-0275.2025.0174\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDhal SB, Kar D (2025) Leveraging artificial intelligence and advanced food processing techniques for enhanced food safety, quality, and security: a comprehensive review. Discov Appl Sci 7:75. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s42452-025-06472-w\u003c/span\u003e\u003cspan address=\"10.1007/s42452-025-06472-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFang CM, Zhou ZT (2008) Yantai, Shandong Province, issued measures to deal with severely exceeding pesticide residues. China Qual News 2008/12/30(006). (in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGeorge RV, Harsh HO, Ray P (2019) Food quality traceability prototype for restaurants using blockchain and food quality data index. J Clea Pro 240:118021. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jclepro.2019.118021\u003c/span\u003e\u003cspan address=\"10.1016/j.jclepro.2019.118021\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThe State Council of the People\u0026rsquo;s Republic of China (2012) Several opinions on vigorously promoting the development of informatization and effectively safeguarding information security (Guo Fa [2012] 23). 28 June. (in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHiroto Y, Rei S, Yuhei H (2020) Dissection of hyperspectral reflectance to estimate nitrogen and chlorophyll contents in tea leaves based on machine learning algorithms. Sci Rep 10:16654. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41598-020-73745-2\u003c/span\u003e\u003cspan address=\"10.1038/s41598-020-73745-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHu Y, Huang P, Wang Y (2023) Determination of Tibetan tea quality by hyperspectral imaging technology and multivariate analysis. J Food Com\u0026amp; Ana 117:105136. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jfca.2023.105136\u003c/span\u003e\u003cspan address=\"10.1016/j.jfca.2023.105136\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHan XF, Xiao J, Song WF (2025) Digital intelligence and urban rural common prosperity: evidence from village level micro space. Fin Eco (05):123\u0026ndash;136. (in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang BQ, Sun YL (2023) Research on the Construction of Tea Regional Public Brand under the Background of Rural Revitalization. J Macro-quality Res 11(06):1\u0026ndash;13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.13948/j.cnki.hgzlyj.2023.06.001\u003c/span\u003e\u003cspan address=\"10.13948/j.cnki.hgzlyj.2023.06.001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang J, Tan X, Zhang L (2023) Hybrid N-way partial least squares and random forest model for brick tea identification based on excitation\u0026ndash;emission matrix fluorescence spectroscopy. Food Bio Tech 16(6):1335\u0026ndash;1342. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11947-023-03006-3\u003c/span\u003e\u003cspan address=\"10.1007/s11947-023-03006-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJing ER (2021) Study on the Construction of Evaluation Index System of Supply Quality about Agricultural Product in Northwest China. J North Minzu Uni 0265\u0026ndash;72. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3969/j.issn.1674-6627.2021.02.008\u003c/span\u003e\u003cspan address=\"10.3969/j.issn.1674-6627.2021.02.008\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi WH, Ren WY, Xie WP (2002) The quality of agricultural products in China is facing challenges. Jilin Agri 08:13 (in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi ZY, WANG M (2007) Amperometry for pesticide residues. Prog Chem 19(10):1585\u0026ndash;1591\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu YT (2024) Research on the the Influence of Digital andIntelligent Transformation of AgriculturalProductive Services on Agricultural GreenTotal Factor Production. Zhengzhou Institute of Aeronautical Industry Management. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.27898/d.cnki.gzhgl.2024.000086\u003c/span\u003e\u003cspan address=\"10.27898/d.cnki.gzhgl.2024.000086\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu H (2025) The digital intelligence of agricultural productive services, the integration of Rural three industries and the toughness of agricultural industrial chain supply chain. Stati Deci 41(21):110\u0026ndash;115. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.13546/j.cnki.tjyjc.2025.21.018\u003c/span\u003e\u003cspan address=\"10.13546/j.cnki.tjyjc.2025.21.018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi H, Mao YL, Shi HT (2024) Establishment of deep learning model for the growth of tea cutting seedlings based on hyperspectral imaging technique. Sci Horti 320:113106. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.scienta.2024.113106\u003c/span\u003e\u003cspan address=\"10.1016/j.scienta.2024.113106\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLanjewar MG, Panchbhai KG (2023) Convolutional neural network based tea leaf disease prediction system on smart phone using paas cloud. NEU COMP APP 35(3):2755\u0026ndash;2771. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00521-022-07743-y\u003c/span\u003e\u003cspan address=\"10.1007/s00521-022-07743-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi YT, Wu SK, He LY (2023) Development and field evaluation of a robotic harvesting system for plucking high-quality tea. Com Elec Agri 206:107659. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.compag.2023.107659\u003c/span\u003e\u003cspan address=\"10.1016/j.compag.2023.107659\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuo N, Li YL, Yang BH (2022) Prediction Model for Tea Polyphenol Content with Deep Features Extracted Using 1D and 2D Convolutional Neural Network. AGRICULTURE-BASEL 12(9):1299. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/agriculture12091299\u003c/span\u003e\u003cspan address=\"10.3390/agriculture12091299\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuo BY, Chen YX (2022) How Can Digital Intelligence Enable High-quality Economic Development:A Study on the Regulatory Role of Business Environment. Sci Tech Pro\u0026amp; Poli 39(05):61\u0026ndash;71. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.6049/kjjbydc.2021040437\u003c/span\u003e\u003cspan address=\"10.6049/kjjbydc.2021040437\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu L, Ruan JJ, Zhuang HT (2024) Impact of Digital Intelligence on Green Development in the Yangtze River Delta Urban Agglomeration. Eco Geo 44(09):123\u0026ndash;132. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.15957/j.cnki.jjdl.2024.09.013\u003c/span\u003e\u003cspan address=\"10.15957/j.cnki.jjdl.2024.09.013\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu RZ, Li X, He C (2025) Mechanism and Verification of Empowering High-Quality Economic Development through Urban Digitization Transformation. Ur Dev Stu 31(01):6\u0026ndash;11 (in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLyu K, Pan JB, Lin ZS (2024) Innovative Province Construction, Regional Smartization, and Competitiveness of Regional Innovation Ecosystems\u0026mdash;\u0026mdash;Mechanism of Influence and Quasi-Natural Experiments. R\u0026amp;D Manage 36(06):137\u0026ndash;150. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.13581/j.cnki.rdm.20231424\u003c/span\u003e\u003cspan address=\"10.13581/j.cnki.rdm.20231424\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi DH, Xiong Y, Chen JH (2024) The Impact of Digital Village Construction on the Regional Public Brand Value of Agricultural Products\u0026mdash;\u0026mdash;Empirical analysis based on the regional public brand of tea. J Macro-quality Res 12(04):101\u0026ndash;114. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.13948/j.cnki.hgzlyj.2024.04.008\u003c/span\u003e\u003cspan address=\"10.13948/j.cnki.hgzlyj.2024.04.008\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMao YL, Li H, Wang Y (2022) Prediction of Tea Polyphenols, Free Amino Acids and Caffeine Content in Tea Leaves during Wilting and Fermentation Using Hyperspectral Imaging. Foods 11(16):2537. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/foods11162537\u003c/span\u003e\u003cspan address=\"10.3390/foods11162537\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMei Y, Mao DL, Lu YH (2020) Effects and mechanisms of rural E-commerce clusters on households' entrepreneurship behavior in China. Growth Change 51(4):1588\u0026ndash;1610. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/grow.12424\u003c/span\u003e\u003cspan address=\"10.1111/grow.12424\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eM\u0026uuml;ller C (2025) World robotics 2025 \u0026ndash; industrial robots. IFR Statistical Department, VDMA Services GmbH, Frankfurt am Main, Germany. Available at: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://ifr.org/img/worldrobotics/Sources_Methods_WR_2025_Industrial_Robots.pdf?utm_source=chatgpt.com\u003c/span\u003e\u003cspan address=\"https://ifr.org/img/worldrobotics/Sources_Methods_WR_2025_Industrial_Robots.pdf?utm_source=chatgpt.com\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (Accessed: 1 November 2025)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNie J, Wang Y, Li Y (2022) Artificial intelligence and digital twins in sustainable agriculture and forestry: a survey. Turkish J Agri Fore 46(5):642\u0026ndash;661. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.55730/1300-011X.3033\u003c/span\u003e\u003cspan address=\"10.55730/1300-011X.3033\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNie WJ, Fu HM, Li TP (2021) A Theoretical Basis for and Path Towards Quality Grading System of Agricultural Products in China\u0026mdash;\u0026mdash;Welfare Evaluation Based on Choice Experiment Survey. J Agrotech Eco( 0765\u0026ndash;78. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org./10.13246/j.cnki.jae.2021.07.005\u003c/span\u003e\u003cspan address=\"https://doi.org./10.13246/j.cnki.jae.2021.07.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePaul T, Islam N, Rakshit S (2025) The impact of artificial intelligence (AI)-enabled collaborative approach: achieving sustainable supply chain performance. J Busi Log 46(4):e70030. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/JBL.70030\u003c/span\u003e\u003cspan address=\"10.1111/JBL.70030\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePaul T, Mondal S, Islam N (2021) The impact of blockchain technology on the tea supply chain and its sustainable performance. Tech Fore Soc Chan 173:121163. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.techfore.2021.121163\u003c/span\u003e\u003cspan address=\"10.1016/j.techfore.2021.121163\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePeng Y, Zheng C, Guo S (2023) Metabolomics integrated with machine learning to discriminate the geographic origin of Rougui Wuyi rock tea. Npj Sci Food 7:7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41538-023-00187-1\u003c/span\u003e\u003cspan address=\"10.1038/s41538-023-00187-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRen G, Wu R, Yin L (2024) Intelligent Evaluation of the Appearance Modality of Black Tea Based on Machine Vision and Hyperspectral Imaging. Ana Lett 57(2):176\u0026ndash;189. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/00032719.2023.2201706\u003c/span\u003e\u003cspan address=\"10.1080/00032719.2023.2201706\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun SY, Li S (2024) The Purpose and Driving Direction of the Digitization Logic for Cultivating and Developing New Quality Productive Forces. China Rev Poli Eco 15(06):101\u0026ndash;114 (in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSu YM, FANG F, WANG YH (2026) The impact of digital and intelligent transformation on the new quality productivity of agricultural enterprises. J Henan Agri Uni 1\u0026ndash;16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.16445/j.cnki.1000-2340.20251104.002\u003c/span\u003e\u003cspan address=\"10.16445/j.cnki.1000-2340.20251104.002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSheng XF, Zan JZ, Jiang YW (2023) Data fusion strategy for rapid prediction of moisture content during drying of black tea based on micro-NIR spectroscopy and machine vision. Optik 276:170645. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ijleo.2023.170645\u003c/span\u003e\u003cspan address=\"10.1016/j.ijleo.2023.170645\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun J, Hu Y, Zou YL (2022) Identification of pesticide residues on black tea by fluorescence hyperspectral technology combined with machine learning. Food Sci Tech 42:e55822. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1590/fst.55822\u003c/span\u003e\u003cspan address=\"10.1590/fst.55822\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShao YTY, Ni JW, Zhou SJ (2024) Safety assessment of agricultural products and the pesticide regulation trend in China. Agr Food Eco 12:29. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40100-024-00322-w\u003c/span\u003e\u003cspan address=\"10.1186/s40100-024-00322-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTaherzadeh O, Cai H, Mogoll\u0026oacute;n JM (2026) Small farms contribute a third of the food consumed in high-income nations. Nat Food 7:66\u0026ndash;73. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s43016-025-01276-y\u003c/span\u003e\u003cspan address=\"10.1038/s43016-025-01276-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang YX, Cai XJ, Zhuang HT (2025) The Mechanism and Test of Digital Intelligence Drives of the Chinese-Style Modern Rural Revitalization\u0026mdash;\u0026mdash;Based on the Marxist Theory of Productive Forces. Pra Fin Eco 46(02):114\u0026ndash;122. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.16339/j.cnki.hdxbcjb.2025.02.014\u003c/span\u003e\u003cspan address=\"10.16339/j.cnki.hdxbcjb.2025.02.014\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWei WC (2003) Guangxi Bureau strictly controls excessive pesticide residues. China Guomen Times 2003/7/15. (in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang W (2005) The phenomenon of excessive pesticide residues in vegetables sampled in Changsha is rising. China Quality News 2005/08/27(006). (in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWei Y, Wen YQ, Huang XL (2024) The dawn of intelligent technologies in tea industry. Trends Food Sci\u0026amp; Tech 145:104337. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.tifs.2024.104337\u003c/span\u003e\u003cspan address=\"10.1016/j.tifs.2024.104337\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWei G, Zhou R (2023) Comparison of machine learning and deep learning models for evaluating suitable areas for premium teas in Yunnan, China. PLoS ONE 18(2):e0282105. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0282105\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0282105\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWei P, Liu HM, Xu CK (2024) Does Green Food Certification promoting agri-food export quality? Evidence from China. J Integra Agri 23(3):1061\u0026ndash;1074. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jia.2023.11.033\u003c/span\u003e\u003cspan address=\"10.1016/j.jia.2023.11.033\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang YQ (2024) Quality measurement and influencing factors analysis of export agricultural products in Gansu Province -- Based on RCEP. North Eco Tra (06):115\u0026ndash;119. (in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang GY, Wang M, Ding S (2024) Differentiation and Realization of Millet\u0026rsquo;s Production Capacity and Value from the Perspective of Technical Efficiency. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.13246/j.cnki.jae.2024.05.002\u003c/span\u003e\u003cspan address=\"10.13246/j.cnki.jae.2024.05.002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (in Chinese) J Agrotechn Econ https://doi.org/\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu W, Zhang Y, Fan Y (2020) ICT Empowers the Formation and Development of Rural E-Commerce in China. IEEE Access 8:135264\u0026ndash;135283. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/ACCESS.2020.3011593\u003c/span\u003e\u003cspan address=\"10.1109/ACCESS.2020.3011593\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang YQ, Dong W (2023) Man machine interaction: the impact of the rise of robots on the income of Chinese workers. J World Eco 46(07):88\u0026ndash;115. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org./10.19985/j.cnki.cassjwe.2023.07.007\u003c/span\u003e\u003cspan address=\"https://doi.org./10.19985/j.cnki.cassjwe.2023.07.007\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXiao H, Fu YL (2024) Practical Logic and Optimization Strategy of Smart Agriculture Development\u0026ndash;Based on A Case Study of Digitally Intelligent Unmanned Orchards on Hilly Mountainous Terrain in Q Village. J Northwest A\u0026amp;F University (Social. Sci Edition) 24(03):58\u0026ndash;65. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps:.//doi.org/10.13968/j.cnki.1009-9107.2024.03.07\u003c/span\u003e\u003cspan address=\"https:.//10.13968/j.cnki.1009-9107.2024.03.07\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu ZY, Wang J, Zheng T (2020) An Online Tea Fixation State Monitoring Algorithm Based on Image Energy Attention Mechanism and Supervised Clustering (IEAMSC). Sensors 20(15):4312. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/s20154312\u003c/span\u003e\u003cspan address=\"10.3390/s20154312\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang XL, Fu L, MOU ED (2020) Cultivation of regional public brand of agricultural products: operation mechanism and realization path. Agri Eco (01):125\u0026ndash;127. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/CNKI:SUN:NYJJ.0.2020-01-047\u003c/span\u003e\u003cspan address=\"https://doi.org/CNKI:SUN:NYJJ.0.2020-01-047\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang WS, Hu SD (2011) Only urea application is the main reason for the deterioration of the quality of improved varieties of tea. J Tea Busi 33(02):67\u0026ndash;68. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.16015/j.cnki.jteabusiness.2011.02.014\u003c/span\u003e\u003cspan address=\"10.16015/j.cnki.jteabusiness.2011.02.014\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu ZY, Quan CX, Fang Z (2024) Key technology, equipment and prospect of digital and intelligent farmland construction. Tran Chin Soc Agri Eng 40(23):1\u0026ndash;14. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.11975/j.issn.1002-6819.202407107\u003c/span\u003e\u003cspan address=\"10.11975/j.issn.1002-6819.202407107\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYuan L, Zhang SQ, He WJ (2025) Impact of Digital Intelligence Level on Green Total Factor Productivity in the Yangtze River Economic Belt. Res \u0026amp;Env Yangtze Basin 34(05):923\u0026ndash;936 (in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYou XW, Wan JY (2022) Research on the Spatial Spillover Effect of Local Officials' Agriculture-related Experience on the Quality and Safety of Agricultural Products:Evidence from Chinese Cities. South China J Eco 0337\u0026ndash;52. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.19592/j.cnki.scje.391317\u003c/span\u003e\u003cspan address=\"10.19592/j.cnki.scje.391317\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYuan HM (2012) \u0026lsquo;Tea e-commerce alliance calls for honest operation\u0026rsquo;. Market Herald, 16 Apr. Available at: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.zjscdb.com/detail.php?newsid=69168\u003c/span\u003e\u003cspan address=\"https://www.zjscdb.com/detail.php?newsid=69168\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (Accessed: 12 Mar. 2026).(in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang XC, Wu ZM, Cao CM (2023) Design and operation of a deep-learning-based fresh tea-leaf sorting robot. Com \u0026amp;Elec Agri 206:107664. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.compag.2023.107664\u003c/span\u003e\u003cspan address=\"10.1016/j.compag.2023.107664\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang J, Li ZC (2021) Intelligent Tea-Picking System Based on Active Computer Vision and Internet of Things. Secu Comm Net 5302783. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1155/2021/5302783\u003c/span\u003e\u003cspan address=\"10.1155/2021/5302783\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZou ZY, Wu QS, Long T (2023) Classification and adulteration of Mengding Mountain green tea varieties based on fluorescence hyperspectral image method. J Food Com Ana 117:105141. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jfca.2023.105141\u003c/span\u003e\u003cspan address=\"10.1016/j.jfca.2023.105141\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang Y, Bai PW (2023) How digital intelligence affects double cycle participation and income gap -- Based on provincial industry level data. J Manage World 39(10):58\u0026ndash;83. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.19744/j.cnki.11-1235/f.2023.0126\u003c/span\u003e\u003cspan address=\"10.19744/j.cnki.11-1235/f.2023.0126\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou JW, Guo W, Cai XJ (2025) How Digital Intelligence Promotes Education Equity. J Edu Sci Hunan Normal Uni 24(03):12\u0026ndash;21. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.19503/j.cnki.1671-6124.2025.05.002\u003c/span\u003e\u003cspan address=\"10.19503/j.cnki.1671-6124.2025.05.002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang J, Wang HY (2025) Quality Traceability Subsidies and Cotton Quality:Mechanism and Empirical Test. J Macro-qua Res 13(02):1\u0026ndash;14. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.13948/j.cnki.hgzlyj.2025.02.001\u003c/span\u003e\u003cspan address=\"10.13948/j.cnki.hgzlyj.2025.02.001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang N, Yang WT, Ke HQ (2024) Does rural e-commerce drive up incomes for rural residents? Evidence from Taobao villages in China. Eco Ana\u0026amp; Pol 82:976\u0026ndash;998. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.eap.2024.04.23\u003c/span\u003e\u003cspan address=\"10.1016/j.eap.2024.04.23\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou J, Yu L, Choguill CL (2021) Co-evolution of technology and rural society: The blossoming of taobao villages in the information era, China. J Rural Stu 83:81\u0026ndash;87. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jrurstud.2021.02.022\u003c/span\u003e\u003cspan address=\"10.1016/j.jrurstud.2021.02.022\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang J, Wang CC, Phelps NA (2023) Rural E-Commerce and Emerging Paths Toward Product Renewal: Evidence from Taobao Villages in Zhejiang Province, China. Profe Geo 75(3):521\u0026ndash;535. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/00330124.2022.2111689\u003c/span\u003e\u003cspan address=\"10.1080/00330124.2022.2111689\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu J (2023) Research on digital economy enabling high quality development of China's agriculture. Southwest University of Finance and Economics. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.27412/d.cnki.gxncu.2023.000037\u003c/span\u003e\u003cspan address=\"10.27412/d.cnki.gxncu.2023.000037\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (in Chinese)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang YZ, Zhu JN, Zhang YL (2025) Development of artificial intelligence and labor mobility. Eco Pers (01):128\u0026ndash;145. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://link.cnki.net/urlid/11.1057.F.20250307.1006.016\u003c/span\u003e\u003cspan address=\"https://link.cnki.net/urlid/11.1057.F.20250307.1006.016\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (in Chinese)\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Appendix","content":"\u003cspan\u003e \u003cli\u003e \u003cp\u003eAppendix C of the Industrial Classification for National Economic Activities (GB/T 4754\u0026ndash;2017), jointly issued by the National Bureau of Statistics of China and the China National Institute of Standardization, provides a concordance table between China's industrial classification and ISIC Rev. 4. Although the 2013 China Economic Census Yearbook adopts the 2011 version of the classification (GB/T 4754\u0026ndash;2011), the overall framework of China's manufacturing sectors in the 2011 system is highly consistent with that in the 2017 version. Hence, this concordance table remains applicable.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eThe categories coded C42 (comprehensive utilization of waste resources) and C43 (repair of metal products, machinery, and equipment) under China's industrial classification are not classified as manufacturing in international standards. As such, they are not included in IFR statistics and cannot be matched, so these two sectors are excluded from the analysis.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eThis study starts from 2012. If 2012 were set as the base year, data for only 21 disaggregated manufacturing sectors would be available from the China Industry Statistical Yearbook. By contrast, using 2013 as the base year allows us to obtain data for 29 detailed sectors from the 2013 China Economic Census Yearbook. Given that no major policy shifts or industrial adjustments occurred between 2012 and 2013, regional industrial structures remained nearly unchanged. To improve measurement accuracy, 2013 is chosen as the base year.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[{"identity":"93194529-09bf-4684-8c56-6c59904d8bbd","identifier":"10.13039/501100002338","name":"Ministry of Education of the People's Republic of China","awardNumber":"25YJC790119","order_by":0},{"identity":"c321b815-c108-47e8-936b-606bf8a75d69","identifier":"10.13039/501100013259","name":"Policy Research Center, National Graduate Institute for Policy Studies","awardNumber":"25MZKA12","order_by":1}],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"Fujian Normal University","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Quantity-quality equilibrium, Agricultural production, Value of the regional public tea brand, Digital-intelligent transformation, Tea industry, Taobao villages, Government expenditure on science and technology","lastPublishedDoi":"10.21203/rs.3.rs-9428414/v2","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9428414/v2","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eAgricultural production constitutes the foundation of national economies and livelihoods, and countries worldwide attach considerable importance to the efficiency, quality, and scale of agricultural operations. Nevertheless, an inherent dilemma has long persisted between high-quality agricultural output relying on intensive inputs and high-yield production driven by cost reduction and scale expansion. This trade-off has become increasingly difficult to reconcile amid the continuous outflow of agricultural labor globally. As a core transformative force advancing Agriculture 4.0, the regional diffusion and application of digital-intelligent technologies raise a critical research question: Can Regional Digital-Intelligent Transformation (RDIT) alleviate this long-standing dilemma? This study uses the value of regional public tea brands to measure the quantity-quality equilibrium level of agricultural production. Following the principle of organic integration grounded in a holistic indicator system rather than simple numerical aggregation, we construct the Regional Digital-Intelligent Transformation Index. Using panel data on China's tea industry from 2012 to 2023, this paper empirically examines the impact of RDIT on the Quantity-Quality Equilibrium Level of Tea Production (QQEL_TP). The main findings are as follows. First, RDIT significantly promotes QQEL_TP. After controlling for individual fixed effects, time fixed effects, and other confounding factors, a 1% increase in RDIT is associated with a statistically significant 0.432% rise in QQEL_TP. Second, the core conclusion remains robust after addressing endogeneity via the instrumental variable method, excluding observations from special years, and adjusting sample specifications. Third, RDIT enhances QQEL_TP by fostering the development of Taobao villages (TBV). Fourth, the positive impact of RDIT on QQEL_TP is negatively moderated by government investment in science and technology (GI_ST).\u003c/p\u003e","manuscriptTitle":"Does Regional Digital-Intelligent Transformation Mitigate the Quantity-Quality Dilemma in Agricultural Production?\n—Empirical Evidence from China's Tea Industry","msid":"","msnumber":"","nonDraftVersions":[{"code":2,"date":"2026-04-23 18:48:42","doi":"10.21203/rs.3.rs-9428414/v2","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}},{"code":1,"date":"2026-04-17 06:41:36","doi":"10.21203/rs.3.rs-9428414/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"8d823cc0-4a1d-4b97-ac9b-d95e5520faf5","owner":[],"postedDate":"April 23rd, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":66496353,"name":"Agricultural Economics \u0026 Policy"}],"tags":[],"updatedAt":"2026-04-17T06:41:36+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-23 18:48:42","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v2","identity":"rs-9428414","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9428414","identity":"rs-9428414","version":["v2"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}