Urban–Rural Integration and Agricultural Technology Innovation: Evidence from China

Abstract

Urban and rural relations are important for the sustainable development of a nation or region. Urban and rural integration (URI), as a conceptual framework and strategic tool for managing urban–rural relations, is supposed to play a role in linking urban and rural socio-economic systems to promote the two-way flow of resources, facilitate spatial restructuring and functional transformation, enhance resource allocation efficiency, and shorten the gap between urban and rural areas. This article takes agriculture technological innovation in China as an example, and discuss whether and how URI can promote agricultural technological innovation (ATI). Based on a dataset of 288 prefecture-level cities spanning the years from 1999 to 2018, this article analyzes the mechanism and effect of URI on the development of ATI by using the negative binomial regression model and other models based on measuring the level of URI in the municipal area. The main conclusions are as follows. (1) URI positively promotes the development of ATI and exhibits significant heterogeneity across various dimensions, regions, and agricultural production areas. (2) The mechanism analysis shows that governance systems and mature markets positively moderate the impact of URI on promoting ATI, which also exhibits heterogeneous characteristics across various regions. (3) The impact of URI on the development of ATI exhibits a double threshold effect, and its marginal impact on the development of ATI increases with the deepening of URI. Moreover, the threshold effect of URI demonstrates significant heterogeneity. In central and urban areas, after surpassing the threshold value, the influence of URI on agricultural technological innovation becomes greater. Therefore, it is crucial to continuously deepen the development of URI, smooth the channels of factor flow, enhance resource sharing, break down the urban and rural barriers, and create a new engine for ATI development.

1. Introduction

As a fundamental industry, the stable growth of agriculture is strategically significant to a country [1]. However, agriculture has encountered increasing challenges as well as a rising demand for food security and imbalances in nutritional structure because of decreasing labor forces, accelerated urbanization process, and the improvement in high living standards [2,3]; and it also has faced the challenges of climate change and agricultural pollution [4,5]. As a consequence, it is urgent to make scientific and technology innovation as the critical driving force [6], enhance agricultural productivity, and accelerate the transformation to modern agriculture [7]. The European Union has established national agricultural knowledge and innovation systems (AKIS) while advancing the implementation of the Common Agricultural Policy, and other developed countries, such as the United States and Japan, have introduced policies, laws, or plans to promote agricultural technology innovation (ATI) as well [8,9]. Meanwhile, developing countries like China, Brazil, and Zambia have been actively advancing agricultural science and technology, for the purpose of accelerating agricultural development [10,11,12].

ATI generally refers to the invention and creation of new agricultural production technologies (i.e., biotechnology, agricultural environmental pollution control technologies), new equipment (i.e., agricultural machinery), and new management practices (i.e., agricultural production management systems, agricultural e-commerce) [13,14,15]. It spans the pre-production, production, and post-production stages of agriculture; integrates biological, information, and other technologies; involves a diverse group of participants, including farmers, enterprises, governments, universities, research institutions, and intermediary organizations [16]; and covers various activities, such as scientific R & D, experimentation and extension, production application, and innovation diffusion [17] to transform the traditional agricultural production model into a new one. ATI and its application exhibit regional characteristics [18], which vary with changes in regional natural conditions, social and economic foundations, and the distance from the market. It is believed that ATI mainly conglomerates in urban areas [19], rather than in rural areas where the majority of agricultural production is located. Furthermore, the pathways of ATI differ significantly between the two kinds of areas. Universities and research institutes in urban areas are the predominant knowledge-producing organizations of ATI, endowed with abundant innovation resources and talent. Through market linkages and frequent interactions with potential collaborators [20,21], these universities and research institutes enhance regional R & D capabilities in the agricultural industry and make contributions to the application process of ATI. Conversely, farmers and some agriculture-related enterprises in rural areas are constrained by innovation factors and capabilities [21], such as cutting-edge technology, latest information, financial capital, talent and R & D infrastructure, and geographical distance from cities [22,23], which consequently makes their ATI capabilities are relatively low and difficult to significantly improve.

Even though there exists an urban–rural continuum and a close linkage [24] between the two kinds of areas, and relatively low barriers for ATI to diffuse from urban to rural areas [25], a coordinated and unified ATI system has still to be formed. According to the experiences from developed countries, such as the Netherlands [26], the United States [27], and France [28], in cooperation with universities and research institutes and technological companies in urban areas, innovation clusters in agriculture were established in rural communities. As the coordinating relationship between the two kinds of areas has been improved, probably driven by the combination of capital inputs [29], technology advances [30], and market demand [31], the spread of ATI from urban to rural areas has become increasingly significant [31]. As far as developing countries, there exists urban–rural dual system, which impedes the free flow of innovation factors [32]. Therefore, it has been proposed that the linkage between urban and rural areas, or the process by which urban and rural areas facilitate the bidirectional flow and reorganization of resources within subsystems, such as the economy, results in a closely interconnected, functionally complementary, and mutually beneficial urban–rural composite system, namely urban–rural integration (URI), should be established and enhanced so as to improve the circulation channels of innovation factors [33,34]. However, the extant literature pays little attention to the impact of the relationship of URI on ATI.

As a consequence, this article explores whether and how URI can enhance ATI in urban and rural areas, taking China as an example. China, a developing country with a vast population, has been endeavoring to revamp traditional farming through technology-driven agricultural innovation. The central government has successively devised and implemented a series of policies, including “National Agricultural and Rural Science and Technology Development Plan” and “New Urbanization Implementation Plan”. These policies explicitly advocate for the implementation of rural revitalization and innovation-driven strategy, through promoting URI, establishing fair exchange and bidirectional flow systems for economic factors between the two areas [35,36]. In 2021, China’s contribution rate of agricultural scientific and technological progress exceeded 61% [37], highlighting the significant supportive and leading role of technology in the development of agriculture and rural areas. This provides a pertinent case for exploring the relationship between ATI and URI and the urban–rural gap of ATI in the context of URI.

This article has two marginal contributions, as follows. Firstly, this article focuses on the interrelationship between urban and rural areas, about which the extant literature on ATI is scarce, and discusses the impact of URI on ATI. It provides a new perspective for rural/agricultural innovation research and especially for narrowing the urban–rural gap of ATI in developing countries and regions. Secondly, it reveals a nonlinear relationship between URI and ATI, and proves that governance systems and mature markets play a positive moderating role in the process of URI promoting ATI, which provides support for differentiated URI policies and better local institutional strategies to promote regional ATI.

2. Literature Review and Main Hypothesis Development

2.1. Factors Influencing ATI

The process of ATI is complex and diverse, and is affected by various levels of factors. From a macro-perspective, ATI is affected by factors including national or regional economic development levels, industrial structure, human capital, financial support or research investments, institutional environment, agricultural resource endowments, and energy structure, etc. [16,38,39]. It is crucial to enhance the allocation capability of ATI-related resources and factors [40]. From a micro-perspective of a specific locality, factors such as local public facilities, land costs and investment motivations, and geographical proximity (i.e., geographical clusters; universities; research institutions), often become significant metrics for the location choices of ATI actors, including agri-tech enterprises [38,41]. Furthermore, local social capital or networks, values, customs [16], as well as ATI-related actors’ individual and family characteristics, such as cognitive levels, continuously influence their absorption capacity and application level of agricultural science and technology [42,43].

Additionally, Pardey point out that the urban–rural relationship is an essential factor causing regional disparities in ATI [44]. Indeed, urban and rural areas originally have an innate connection in terms of capital, people, and commodities. Much ATI is ultimately applied in rural areas, while cities remain its dominant innovation sites. As capital and personnel in urban areas are increasingly involved into agricultural production activities, a large number of agriculture-related technologies have been spread to rural areas [45]. Furthermore, ATI requires interaction, collaboration, and communication among innovators, users, and other related actors in urban and rural areas, which reflects the importance of urban–rural coordination and cooperation in the ATI process. As the role of innovation networks connecting urban and rural areas strengthens [46], the information gap gradually narrows, agricultural innovation-related technologies and resources are shared among urban and rural innovators [47]. Therefore, García-Cortijo urged that interactions across diverse areas, including urban, suburban, and rural, deserve attention, as far as ATI is concerned [48]. Finally, the EU’s CAP reform proposal also suggests relying on cities to accelerate innovation cooperation between urban and rural areas through knowledge transfer and collaboration, optimizing the utilization of innovative resources, and facilitating the practical transfer of ATI [49]. Consequently, the significance of beneficial interactions between urban and rural areas for the dissemination and service of agricultural technology cannot be overstated, and it is worth further exploration.

2.2. URI and ATI

Unbalanced development between urban and rural areas has become increasingly prominent worldwide as urbanization and industrialization progress [50]. Hence, sustainable URI is viewed as a challenge that nations across the globe must address [51]. Academic attention to urban–rural relationships dates back initially to the utopian socialist notions of urban and rural development [52]. Based on this, Karl Marx first presented the concept of “URI”, which was later developed into the theory of urban–rural linkage. Subsequently, dual-structure theories of urban–rural development, such as the “Lewis–Ranis–Fei” model, and coordinated development theories, exemplified by the “regional network model”, were proposed successively [53]. These theories have provided a theoretical foundation for assessing URI. From a systems theory perspective, URI represents a stage in the evolution of urban–rural relationships [54]. It involves the bidirectional flow and reorganization of urban and rural resources, facilitating the restructuring and transformation of various subsystems, including population, society, and economy, within the urban–rural system. This process ultimately creates a closely interconnected, functionally complementary, and mutually beneficial urban–rural composite system [55]. As a result, resource allocation shifts from inefficiency to efficiency, achieving organic coordination in urban and rural development [56]. Compared to developed countries, China, in its early stages of development, was heavily influenced by an “urban-biased” approach, which led to a city-centric system with nearly unidirectional resource flows. As a result, rural areas and farmers bore much of the costs of industrialization and urbanization, while agricultural modernization lagged significantly behind. This imbalance has persisted due to ongoing urban–rural barriers and the entrenched dual economic structure [57]. Although some major cities and their surrounding areas have begun to show features of an urban–rural continuum, in most regions, the spatial divide between urban and rural areas remains clear, and full URI is yet to be achieved. Therefore, given China’s unique development context, promoting urban–rural linkages to reshape relations between these areas has become a crucial strategic priority [33]. In the study of URI and ATI, proximity theory offers a unique perspective for understanding how urban–rural interactions influence ATI. From the perspective of multidimensional proximity, innovation elements such as talent, knowledge, and information flow between regions through geographic, organizational, social, cognitive, and institutional proximities, thus promoting element sharing and innovation [58]. URI promotes ATI through multidimensional proximity, mainly reflected in the following aspects. Firstly, the sharing of innovation elements. The integration of urban and rural economies and spaces reduces the time–space distance between urban and rural areas by extending infrastructure like transportation and communication to rural regions [59], enhancing geographic proximity. This increased geographic proximity lowers communication costs, prevents information asymmetry due to long distances, and facilitates knowledge spillovers and technology diffusion [60]. Consequently, agricultural technology knowledge and production resources accumulated in urban areas can be quickly transferred to rural areas, effectively enhancing rural agricultural technology and innovation capabilities. Rural areas can also access urban market information and technical support more conveniently, further promoting agricultural technology research and application [61]. Moreover, urban–rural population and social integration, through initiatives such as promoting bidirectional labor flow between urban and rural areas [62] and establishing mechanisms for joint construction, shared governance, and benefit [63], help reduce cognitive gaps between urban and rural populations, thereby facilitating collaboration and exchanges. The increase in cognitive proximity means that innovators share similar knowledge bases and structures, and similar technical and professional competencies, which reduces communication barriers, increases the likelihood of collaboration, and ultimately fosters the spread and cooperation of ATI [64]. Secondly, building cooperative networks. Urban–rural social integration promotes the equalization of public services, aiming to provide equal development opportunities for urban and rural residents, enhancing mutual recognition and social integration between them [65]. This equalization of public services is reflected not only in the expanded coverage of basic services like education and healthcare but also in the gradual balancing of welfare levels between urban and rural areas, helping to establish a foundation of trust between them [63]. Additionally, population integration, through bidirectional labor flow and diverse social interactions, increases contact frequency between urban and rural residents, deepening mutual understanding and trust [62]. Urban–rural ecological integration also promotes shared concepts of environmental protection and green development, fostering a collective awareness and sense of responsibility for environmental conservation. Together, these processes enhance social proximity between urban and rural areas, creating a social network based on trust and cooperation [66]. This increased social proximity facilitates knowledge sharing and technology dissemination. The establishment of trust reduces uncertainties in information exchange, promotes the exchange and transfer of knowledge, and creates conditions for knowledge spillovers and innovation [67], thereby promoting the diffusion and innovation of ATI. Thirdly, improvement in institutional mechanisms. Urban–rural economic integration promotes the coordinated development of urban and rural industries, facilitating organizational cooperation between cities and rural areas [68,69]. For example, extending agricultural supply chains, establishing cooperatives, and fostering joint urban–rural enterprises have strengthened the connections and resource sharing between urban and rural organizations, forming stable communities of interest [70]. This economic cooperation not only increases interaction among different organizations but also significantly improves cooperation efficiency. Additionally, urban–rural social integration promotes communication and cooperation among various organizations through the equalization of public services [63]. With the enhancement in organizational proximity, urban and rural interactions become closer, fostering cooperation among urban research institutions, enterprises, and rural households [1,71], thereby promoting the dissemination and application of ATI in rural areas. Moreover, multidimensional URI has led to policy alignment. By improving policies, laws, and development strategies, institutional barriers between urban and rural areas have gradually been broken down, providing a stable institutional foundation for innovation cooperation, reducing uncertainties in urban–rural interactions, and enhancing institutional coherence [33]. The improvement in institutional proximity creates a favorable policy environment for ATI, promoting knowledge leaps and innovation diffusion, breaking existing path dependencies [72], accelerating the application of agricultural technologies, and contributing to the sustainable development of agricultural innovation. In summary, URI enhances geographic, organizational, social, cognitive, and institutional proximities, forming a multidimensional ecosystem conducive to ATI. These proximities interact to facilitate the flow and sharing of innovation elements, thereby accelerating the diffusion and application of agricultural technologies, providing solid support for agricultural innovation. Additionally, given the discrepancies in regional circumstances, including the availability of resources, the effect of URI on ATI is supposed to vary across regions. Hence, this article put forward a hypothesis as shown below.

Hypothesis 1a: 

URI has a significant positive effect on ATI, and its impact varies across different regions.

Generally, URI typically involves integration across five dimensions: population, economy, society, space, and ecology [33]. Among these, social integration focuses on the interaction of social relations across urban and rural residents, encompassing the sharing and exchange of education, healthcare, culture, and other aspects [73]. It helps to dismantle the urban–rural dual structure, enhance the social capital of rural residents, and increase their ability to participate in social activities. For example, in Italy, local agricultural cooperatives assist farmers in adopting precision agriculture techniques by providing shared equipment, technical training, and knowledge exchange platforms. This support reduces the technological cost for individual farmers and enhances their motivation to adopt new technologies through collective efforts [74]. In Minnesota, USA, farm communities organize regular experience-sharing and agricultural training sessions, improving the rate of technology adoption in farming [75]. Additionally, many smallholder farmers in Uganda rely on “Farmer Field Schools” to acquire new knowledge and techniques. These schools use social capital as a basis for collective learning and field experiments, helping farmers effectively adopt methods such as ecological farming, thereby enhancing agricultural productivity and yields [76]. Population integration refers to the movement and settlement of people across urban–rural regions, including rural-to-urban migration and urban-to-rural return flows. The bidirectional flow of population facilitates the exchange of knowledge, skills, and ideas, which can promote the diffusion of agricultural technologies to some extent [77]. However, its effectiveness is constrained by the scale and frequency of population movement, which are often closely linked to economic factors. Spatial integration involves the effective planning and allocation of construction land in urban–rural regions, including infrastructure development and the sharing of public service facilities [1]. This process can enhance both the living conditions and production environment in rural areas, boosting their attractiveness and providing a favorable physical environment and infrastructure support for ATI. However, its impact is relatively indirect. Ecological integration focuses on the collaborative governance and protection of ecosystems in urban–rural areas, including the sharing of ecological resources and joint efforts to prevent and regulate environmental contamination. A good ecological environment can provide a sustainable foundation for agriculture, thereby enhancing the ecological benefits of agricultural production. However, its impact on ATI is also relatively indirect.

Meanwhile, economic integration denotes the interconnectivity of economic activities across urban–rural areas, encompassing the flow of resources, market alignment, and the extension of industrial chains. This process aids in the collaborative growth of urban-rural economies. Firstly, economic integration improves resource utilization efficiency through optimized allocation and provides financial, technical, and market support for ATI. For instance, the introduction of urban capital can provide the necessary financial support for ATI, enabling new technologies to be swiftly applied to agricultural production. Additionally, advanced urban technology and management expertise can be transferred to rural areas through economic integration, thereby elevating the level of agricultural technology [78,79]. For example, the rapid growth of urban agriculture in the Netherlands has benefited from urban capital supporting greenhouse farming through public investment and agricultural technology incubators, providing smart irrigation systems and vertical farming technologies. The introduction of these advanced technologies has significantly improved agricultural productivity in the Netherlands [26,80]. Secondly, economic integration promotes ATI through market mechanisms. The increased urban demand for agricultural products drives farmers to adopt advanced agricultural technologies to enhance production efficiency and product quality [81]. Changes in market demand not only spur ATI but also facilitate rapid technology diffusion. For example, in Kenya, international market demand for high-quality horticultural products has encouraged farmers to adopt more efficient planting and irrigation techniques, including drip irrigation systems and greenhouse cultivation, to improve product quality and reduce waste. These innovative measures have greatly increased horticultural production efficiency and facilitated the rapid spread of new technologies [82]. Thirdly, economic integration helps extend the agricultural value chain, thereby fostering synergies between agriculture and other industries. The extension of the agricultural value chain encompasses not only the fundamental stages of agricultural production, processing, and sales but also related service and manufacturing industries [83]. Through integration with urban economies, rural areas can attract more businesses and investments, fostering diversified agricultural development. For example, in California, grape cultivation is not limited to producing raw materials but extends the value chain through winemaking, tourism, and brand promotion. By developing a wine culture, California’s wine industry has successfully attracted significant investment and tourists, creating positive synergies between agriculture, tourism, and manufacturing [84]. In comparison, other dimensions have relatively limited impacts on ATI. While social integration can enhance the social capital of rural residents, its direct role in promoting ATI is weak. Population integration facilitates technology diffusion to some extent, but its influence is constrained by the scale and frequency of population movement, making its impact less significant than economic integration. Spatial integration and ecological integration primarily affect ATI indirectly by improving the physical and ecological environments, respectively, and thus have more limited direct effects. In summary, economic integration may have a more significant positive impact on ATI compared to other dimensions. This is achieved through optimizing resource allocation, leveraging market mechanisms, and extending industrial chains. Therefore, this article posits H1b.

Hypothesis 1b: 

Compared to other dimensions, urban–rural economic integration has a more significant positive impact on ATI.

Additionally, the existing research indicates that agricultural technology primarily involves three categories, namely farming and breeding techniques, agricultural mechanical engineering technology, and agricultural service management technology [13,14,15]. Therefore, this study classified ATI patents into three categories based on the International Patent Classification (2024.01 edition) and the existing research on knowledge bases and classification [85], which are science-based analytical knowledge (agricultural farming and breeding patents), engineering-based synthetic knowledge (agricultural mechanical engineering technology patents), and art-based symbolic knowledge (agricultural service management technology patents). Firstly, agricultural service management technology innovation is based on symbolic knowledge, which is typically embedded in individuals and organizations, and exhibits significant spatial stickiness, requiring a high degree of geographic proximity [86]. The innovation process often involves face-to-face communication and social interaction to facilitate knowledge exchange in specific contexts. URI effectively reduces time–space distances, while the integration of urban and rural societies and populations helps attract management and technical personnel to rural areas, improving the specialization and innovation capabilities of service management. It also strengthens trust and social capital between urban and rural residents, promoting effective knowledge transfer, thereby making the marginal effect of URI on agricultural service management technology innovation more significant. Secondly, agricultural mechanical engineering technology innovation belongs to synthetic knowledge, emphasizing the solution of specific problems through interaction and learning in practice. Therefore, geographic proximity is relatively important in the innovation process [87], requiring some level of face-to-face communication and interaction. The improvement in URI levels shortens the urban–rural distance and increases opportunities for face-to-face interactions, thereby facilitating the realization of agricultural mechanical engineering technology innovation to some extent. Lastly, agricultural farming and breeding technology innovation is mainly based on science-based analytical knowledge, which relies more on abstract scientific principles and codified knowledge. Due to its independent knowledge system, it is easier to diffuse and spread, and thus the innovation process has a lower dependency on face-to-face communication [88]. Although URI can support the dissemination and application of analytical knowledge by improving infrastructure and promoting the flow of scientific resources, its impact is relatively indirect. Hence, this article put forward a hypothesis as shown below.

Hypothesis 1c: 

Compared to other types of ATI, URI has a more significant positive impact on agricultural service management technology innovation.

Furthermore, it is noteworthy that URI, while promoting ATI, is constrained by multiple factors, resulting in uncertainties. ATI may exhibit nonlinear characteristics under varying levels of URI. In the initial phase of URI, the persistence of institutional barriers across urban–rural systems, coupled with the significant urban siphoning and agglomeration effects, limits the radiative and trickle-down impacts on rural areas [89]. Consequently, the flow of traditional and innovative production factors remains predominantly unidirectional [90], leading to a weaker growth in ATI during this period. As URI deepens, the barriers hindering the flow of factors across urban–rural regions are gradually dismantled. A substantial influx of traditional and innovative production elements into rural areas [32], adequately allocated under market mechanisms and policy regulations [91], fosters significant advancements in ATI. As a consequence, this article posits H1d.

Hypothesis 1d: 

URI has a threshold effect on ATI, with the marginal impact on the development of ATI increasing as URI deepens.

2.3. The Moderating Effects of Regional Governance Systems and Mature Markets

ATI is intricately linked not only to URI but also to governance systems. Governance systems refer to government, NGO, community groups, and the private sector efforts to promote URI through the formulation and implementation of development strategies and specific policies, creating favorable infrastructural (such as transportation networks, communication networks, and energy supply), capital, and business environments conducive to development. These systems represent an external force pushing URI forward in China [33,92]. Firstly, robust governance systems facilitate dismantling or weakening the barriers across urban–rural areas, and optimize grassroots governance, so as to safeguard URI [93]. For instance, ensuring the uniformity and coherence of urban and rural policies facilitates interaction across these areas, preventing resource wastage and efficiency reduction caused by policy discrepancies. Secondly, effective governance improves the allocation efficiency of innovation resources across urban–rural areas, leading to the increase in ATI [94]. For instance, well-structured governance systems support rural management innovation, attract diverse talents from cities to rural areas, and alleviate the scarcity of skilled personnel in villages, which is crucial for developing agricultural technology services in rural regions [95]. Lastly, through the guidance and support of governance systems, intellectual property protections in agricultural technology have been strengthened, thereby encouraging innovation by enterprises and individuals and safeguarding the legitimate rights of innovators [96]. Simultaneously, the establishment of technology promotion mechanisms has facilitated the widespread application of agricultural technologies, thereby accelerating the transformation and upgrade of traditional agriculture. Therefore, this article posits H2a.

Hypothesis 2a: 

Regional governance systems positively modulate the relationship between URI and ATI.

Furthermore, in a mature market, URI undoubtedly exerts a stronger promotional effect on the development of ATI. The combination of URI and mature markets significantly contributes to the advancement of agricultural technology development. Firstly, efficient interactions among agricultural innovation agents under market mechanisms contribute to the formation of regional ATI ecosystems and establish a virtuous cycle of enhanced innovative capabilities among market participants [91]. Secondly, mature markets facilitate enhanced information transparency, implying that market participants, whether in urban or rural areas, can relatively easily access the latest information and trends regarding agricultural technologies. This high-transparency information environment promotes the rapid dissemination and adoption of such technologies [97]. Thirdly, mature markets are often accompanied by highly competitive environments, which can stimulate the drive for ATI. This competitive pressure compels agricultural participants to continually seek new technologies and methods, such as introducing new techniques, optimizing production processes, and improving product quality, to enhance production efficiency and product quality, thereby maintaining a competitive edge [98]. Lastly, URI provides traditional and innovative production elements necessary for ATI. In this process, if resources are to be allocated efficiently, a market-oriented approach is essential to alleviate market rigidity and optimize resource distribution. Under the influence of market mechanisms, urban and rural elements achieve optimal free flow and equitable exchange [99]. ATI necessitates resources such as capital. Mature markets play a pivotal role in driving rural supply-side reforms, guiding the direction of resource input, optimizing the efficiency of resource allocation, attracting enterprise agglomeration, and facilitating information dissemination, knowledge learning, and technology cooperation among clustered enterprises, thereby reducing the costs of agricultural innovation and enhancing the efficiency of ATI [100]. In summary, at certain levels of URI, different stages of marketization may have differential impacts on the development of ATI. Specifically, a mature market may be a critical factor in enhancing the positive impact of URI on ATI. Hence, this article posits H2b.

Hypothesis 2b: 

Mature markets positively modulate the relationship between URI and ATI.

In summary, the theoretical framework constructed in this article is depicted in Figure 1.

3. Research Design

3.1. Measures

(1) Dependent variable: ATI

According to comprehensive reviews of the existing research, ATI is typically measured by innovation inputs and outputs. However, compared to innovation input indicators, output indicators provide a more objective and realistic reflection of regional ATI achievements. Additionally, because invention patents are closely related to technology progress, they can better characterize the level of regional innovation. Considering the time lag between patent application and approval, this delay may amplify the lagged effects between variables in the analysis model. Therefore, this article utilized the volume of agricultural technology invention patent applications filed from 1999 to 2018 in the CNKI (China National Knowledge Infrastructure) patent database to measure ATI. Moreover, in this article, if the address field contains “county”, “town”, “village”, and “township”, the area is classified as rural; otherwise, it is classified as urban.

(2) Independent variable: URI

Based on the theoretical underpinnings and rich connotations of URI, and referencing prior research findings [32,33,101], this article adhered to the principles of comprehensiveness, representativeness, and data availability, and establishes a comprehensive URI evaluation framework that incorporates demographic, economic, social, spatial, and ecological dimensions. This system included five secondary indicators and sixteen tertiary indicators (

Table 1. URI evaluation index system.

Target Layer	Standardized Layer	Indicator Layer	Calculation or Description of the Indices	Properties	Weights
URI evaluation index system	RUIpop	Urban–rural population structure [32,33]	Permanent urban population/total permanent population	+	0.065
		Urban–rural employment structure [32,33]	Proportion of the workforce employed in secondary and tertiary sectors/Proportion of the workforce employed in primary sectors	+	0.075
	RUIecon	Urban–rural industrial structure [32,101]	Proportion of GDP contributed by secondary and tertiary sectors compared to the primary sector	+	0.131
		Urban–rural income structure [33]	Ratio of urban to rural household disposable income per capita	-	0.041
		Urban–rural consumption structure [32,33]	Ratio of urban to rural household consumption per capita	-	0.031
		Urban–rural industrial technology level [33,101]	Per capita agricultural machinery power	+	0.048
	RUIsoc	Urban–Rural basic education structure [32,33]	Ratio of students to full-time teachers in primary and secondary schools	-	0.089
		Urban–rural cultural, educational, and entertainment structure [33,101]	Urban to rural per capita expenditure ratio on culture, education, and entertainment	-	0.035
		Urban–rural healthcare structure [32,33]	Urban to rural per capita expenditure ratio on healthcare	-	0.049
		Urban–rural social security structure [33,101]	Social security and employment expenditure/Local general public budget expenditure	+	0.026
	RUIspa	Urban spatial agglomeration level [33]	Ratio of built-up area to total land area	+	0.139
		Urban–rural spatial circulation network [32,33]	Ratio of highway and railway mileage to land area	+	0.054
		Urban–rural information connectivity level [32,101]	Urban to rural per capita expenditure ratio on transportation and communication	-	0.04
	RUIecol	Urban–rural greening level [32]	Urban green coverage rate	+	0.024
		Urban–rural health level [32]	Sanitation facilities penetration rate	+	0.053
		Urban–rural environmental protection level [32,33]	Wastewater centralized treatment rate	+	0.099

). The entropy weight method was applied to objectively determine and calculate the weights of these indicators.

(3) Control variables

To mitigate potential biases caused by omitted variables and to comprehensively evaluate the impact of URI on ATI, this article drew on databases from CNKI and Web of Science, filtering the relevant literature published between 1980 and 2023 using keywords such as “agricultural technological innovation”, “agricultural science and technology innovation”, “urban–rural linkages”, “urban–rural interactions”, and “urban–rural development”. Through systematic review and synthesis, the article identified key factors influencing the development of ATI and URI. Based on these identified factors, the article further incorporated control variables across multiple dimensions, including economic development level [16], industrial structure [16], openness [39], environmental regulation [38], climate risk [38], and R & D investment [39], to ensure the comprehensiveness and scientific rigor of the analysis. These variables were: (1) Regional economic level (lnpgdp). As ATI often involves substantial investments with long return periods, the regional economic level significantly influences the capacity for agricultural innovation. This is represented by the log-transformed per capita GDP, which serves as an indicator of economic development; (2) Regional openness degree (lnopen). Regional openness is an important channel for obtaining technology spillover and achieving technology innovation, which not only facilitates the acquisition of agricultural research funding, but also extends the agricultural industry chain and promotes the structural adjustment of the agricultural industry. This article used the logarithm of the ratio of FDI in GDP to characterize the degree of regional openness; (3) Regional industrial structure (lnind). Given that the primary sector forms the practical base for ATIs, this article used the logarithm of the primary sector’s GDP share to characterize the regional industrial structure; (4) Regional environmental level (lnenv). To meet environmental regulations, agricultural stakeholders continuously enhance technology innovation capabilities to reduce production costs and comply with government regulations. This article adopted the logarithm of the environmental regulatory intensity (word frequency of environmental vocabulary of prefecture-level cities/word frequency of prefecture-level city government work report) to characterize the regional level of environmental protection; (5) Regional climate risk (lnrcr): Variability and uncertainty in climate patterns, such as seasonal fluctuations and anomalies, can significantly affect agricultural production and potentially drive advancements in agricultural technology. In this article, regional climate risk is quantified using the climate risk index derived from the China Climate Change Blue Book; (6) agricultural R & D input (lnag-r&d). This variable is a fundamental carrier and support for achieving ATI, effectively reflecting the investment in agricultural R & D funds and personnel by the different prefecture-level cities in China. Therefore, it is characterized by the logarithm of the combined value of full-time equivalent agricultural R & D personnel weighted at 0.41 and internal agricultural R & D expenditure weighted at 0.59.

(4) Moderator variables

Based on the previous analysis, the governance system (MAN) is characterized by the ratio of administrative villages that have formulated village plans to the total number of administrative villages within prefecture-level cities. Mature markets are represented by the marketization index (MAR) of each prefecture-level city, which serves as a proxy variable.

3.2. Model Selection

In this article, the dependent variable (ATI patents) was non-negative integer count-type data that do not conform to normal distribution, typically necessitating a Poisson regression model for analysis. However, an examination of the sample data during the study period revealed that the mean is lower than the variance, indicating overdispersion, which violates the assumptions of the Poisson model. Consequently, a negative binomial regression model was considered as an alternative. The negative binomial model test results rejected the original hypothesis (p < 0.05), justifying its use over the Poisson model. Thus, this article applied a negative binomial model to analyze the effect of URI on ATI [102]. The specific formula as follows:

$$E\left({ATI}_t\right)=\exp \left(\alpha +\beta {RUI}_t+\gamma {Controls}_t+\sum Year+\sum City+ϵ\right)$$

(1)

Subsequently, this article constructed the following model to assess how governance systems and mature markets moderate the relationship between URI and ATI. The specified formula is as follows:

$$\begin{array}{cc}\hfill E\left({ATI}_t\right)=\exp (\alpha & +\beta {RUI}_t+\mu {RUI}_t\times {MAN}_t+\sigma {MAN}_t+\gamma {Controls}_t\hfill \\ & +\sum Year+\sum City+ϵ)\hfill \end{array}$$

(2)

$$\begin{array}{cc}\hfill E\left({ATI}_t\right)=\exp (\alpha & +\beta {RUI}_t+\mu {RUI}_t\times {MAR}_t+\sigma {MAR}_t+\gamma {Controls}_t\hfill \\ & +\sum Year+\sum City+ϵ)\hfill \end{array}$$

(3)

In model (2), the interaction term RUI $\times$ MAN represents the interaction between URI and governance systems. In model (3), RUI $\times$ MAR represents the interaction between URI and mature markets. By examining the coefficients μ of these interaction terms, the article assessed how governance systems and mature markets moderate the mechanism linking URI and ATI. To further investigate whether URI exhibits nonlinear characteristics in promoting ATI, this article adopted threshold model theory to construct the following model [103]:

$$\begin{array}{cc}\hfill E\left({ATI}_t\right)=\exp (\alpha & +{\alpha }_1{RUI}_{t}I\left({RUI}_t\le {\gamma }_1\right)+{\alpha }_2{RUI}_{t}I\left({{\gamma }_1<RUI}_t<{\gamma }_2\right)\hfill \\ & +{\alpha }_3{RUI}_{t}I\left({RUI}_t\ge {\gamma }_2\right)+\gamma {Controls}_t+\sum Year+\sum City\hfill \\ & +ϵ)\hfill \end{array}$$

(4)

Model (4) is designed as a two-threshold test model, where γ are the thresholds to be estimated. The indicator function I (*) is used to designate which regression regime applies based on the thresholds, with other variables retaining the same definitions as previously outlined.

3.3. Data Sources

This article utilized data from multiple sources between 2000 and 2019, including the CNKI Patent Database (https://kns.cnki.net/kns8s/AdvSearch?classid=OORPU5FE (accessed on 24 October 2024)), China Socio-Economic Development Database (https://data.cnki.net/ (accessed on 24 October 2024)), Wind Database (https://www.wind.com.cn/mobile/EDB/zh.html (accessed on 24 October 2024)), and CSMAR (https://data.csmar.com/ (accessed on 24 October 2024)). Given the low proportion of missing values and their uniform distribution in the original dataset, combined with the relatively stable variability of the data, linear interpolation was employed to address the missing entries. Additionally, to account for price fluctuations, the data were adjusted using GDP deflators and commodity price indices, with 1999 as the base year. Meanwhile, in order to reduce the impact of heteroscedasticity, logarithmic transformations were applied to the control variables before their inclusion in the model. The foundational geographic information was sourced from China’s National Geographic Information Center (www.tianditu.gov.cn (accessed on 24 October 2024)). The descriptive statistics and collinearity diagnostics of variables are shown in

Table 2. Results of variable descriptive statistics and collinearity diagnostics.

var	Obs	Mean	Std.Dev	Min	Max	VIF
ATI	5760	49.649	156.057	0	2970	-
RUI	5760	0.539	0.092	0.267	0.899	1.39
GDP	5760	29,240.736	26,936.689	1160	186,125	1.63
OPEN	5760	0.022	0.028	0.006	0.376	1.21
IND	5760	15.824	10.113	0.035	63.822	2.42
ENV	5760	0.003	0.002	0	0.012	1.47
RCR	5760	46.651	16.042	12.88	105.7	1.22
AG-R&D	5760	4260.072	6611.287	0	74,567.005	1.31
MAN	5760	9.273	3.713	1.089	28.009	2.33
MAR	5760	8.373	3.355	1.077	18.257	1.52

.

Table 2 shows there exist significant differences between the maximum and minimum values of ATI and URI, which indicates substantial disparities in the development of ATI and URI among prefecture-level cities in China. Moreover, the VIF values for all variables are below 10, confirming that no multicollinearity exists among the variables.

4. Spatio-Temporal Pattern Analysis

4.1. Spatio-Temporal Dynamics of ATI

The advancement of “agriculture, rural, and farmers” is inextricably linked to ATI within the context of innovation propelling economic and social development. Compared to innovation studies in other industries, research on ATI uniquely requires a deep understanding of its characteristics within rural contexts and the urban–rural divide, in order to fully explore its value to rural development. This article selected 288 prefecture-level cities as research units and divided the study period into four segments: 1999–2003, 2004–2008, 2009–2013, and 2014–2018. It aims to investigate the spatial evolution of China’s ATI patterns, as depicted in Figure 2.

In the early stage of agricultural innovation from 1999 to 2003, ATIs were dispersed across provincial capitals and surrounding cities. Among 263 prefecture-level cities, a total of 5652 patents were recorded. Despite the broad distribution, the number of innovations was limited and not scaled up significantly. However, regions such as Beijing–Tianjin–Hebei (BTH), Jiaodong Peninsula (JDP), and the Yangtze River Delta (YRD) began to show early signs of spatial agglomeration of agricultural innovations. During the rapid growth stage of 2004–2008, the spatial scope of ATI continued to expand from the prefecture-level cities that had shown innovation outputs in the first stage to a broader area. The BTH region, the JDP, and the YRD exhibited an increasingly apparent spatial agglomeration of ATI. At this stage, ATI was observed in 275 prefecture-level cities, generating 15,821 patents, approximately 2.8 times the output of the first stage. Each prefecture-level city was at a relatively low-output stage of innovation, with significant innovations concentrated in metropolitan hubs like Beijing and Shanghai. Additionally, only a few provincial capitals exhibited a comparatively higher output of innovative achievements. During the acceleration growth stage of agricultural innovation from 2009 to 2013, the space scope of ATI expanded further to include all 288 prefecture-level cities, with 62,840 patents recorded, about 11 times that of the first stage. The spatial expansion of innovation was rapid, and the number of innovative outputs increased significantly. Spatially, beyond provincial capitals, high-value agglomeration areas of ATI were found in the core areas of the YRD (Southern Jiangsu and Northern Zhejiang), JDP, and BTH region. In addition, the Pearl River Delta (PRD, around Guangzhou, Dongguan, and Huizhou) and the Chengdu-Chongqing area also emerged as contiguous hotspots of ATI. In the four years from 2014 to 2018, the space of ATI accelerated further, with a total of 201,668 patents. During this stage, the agglomeration characteristics became even more pronounced. Regions that had previously appeared as high-value innovation areas are more intensified, and the space scope of high-value areas is expanded. Especially, the YRD, JDP, PRD, and Chengdu-Chongqing region emerged as major clusters for ATI in the country. Among these, the YRD emerged as the largest high-value cluster area for ATI. The core of high-value innovation shifted from southern Jiangsu and northern Zhejiang to the periphery of the YRD in Anhui Province. After 2014, ATI in Anhui increased rapidly, becoming China’s most concentrated high-value area of ATI. Specifically, Hefei and Wuhu in Anhui are ranked second and ninth nationally, with patent counts of 6098 and 4254, respectively.

A further analysis was conducted using global spatial autocorrelation tests for four time periods, with p-values of 0.000 for each period, passing the significance tests. This indicates varying degrees of positive spatial autocorrelation in ATI at each stage. The Moran’s I index values were 0.030, 0.033, 0.105, and 0.203 for the four respective stages, showing a gradual increase. This suggests that the trend of ATI clustering is strengthening over time, with regional disparities widening. The areas with clusters of ATI are increasingly gaining an advantage, demonstrating a “Matthew effect”.

4.2. Spatio-Temporal Dynamics of URI

The article explored the spatio-temporal dynamics characteristics of URI levels in China across four time periods: 1999–2003, 2004–2008, 2009–2013, and 2014–2018, as depicted in Figure 3.

From 1999 to 2018, the spatial pattern of URI in China exhibited a “multi-core dispersion and clustered aggregation” characteristic, transitioning from high disparity and low quality to low disparity and high quality. The core centers of integration were primarily in the urban agglomerations of the BTH, YRD, and PRD regions, with clustering effects becoming increasingly pronounced and gradually expanding to peripheral cities. During the 1999–2003 period, the urban agglomerations of BTH, YRD, and PRD were the core high-value areas of URI. The Jiaodong and Liaodong Peninsulas were secondary high-value clusters. Most other areas were at a lower or lowest level of URI, forming a “core–periphery” structure, with regions of lower value surrounding the higher value URI areas. In this stage, the economic foundations in both urban and rural regions of China remained relatively weak, the industrial structure required further refinement, and urbanization and industrialization were slowed by the effects of the Asian financial crisis. Consequently, the overall level of URI was characterized by low-quality integration, remaining at a relatively low but balanced level. From 2004 to 2008, the characteristics of high-value URI clusters in the BTH, YRD, and PRD were further enhanced, with secondary high-value clusters forming along the Beijing–Guangzhou railway. In comparison to the preceding period, the number of regions at the lowest level of integration significantly increased, predominantly dispersed in the western areas. During this stage, as economic system reforms deepened and urbanization rapidly progressed, the urban siphon effect intensified, leveraging locational advantages and policy support to concentrate critical resources. This prompted surplus labor from rural and surrounding areas to move to urban for employment opportunities. As a result, a significant dual economic system emerged across urban–rural areas, characterized by the one-way movement of production factors, including rural population and land, toward cities. This process widened the urban–rural development divide, spatially manifesting as an increased concentration of high-value URI regions alongside a growing number of low-value areas. From 2009 to 2018, high-level integration areas, such as the BTH, YRD, and PRD, expanded further, gradually extending from central urban areas to peripheral cities. Medium-level URI areas continued to expand into central China, with low-level areas significantly reduced, mainly located in the southwestern border regions, reflecting a tendency toward low-value lock-in. Additionally, the Chengdu-Chongqing metropolitan cluster in the west also entered the ranks of high-value URI clusters, likely benefiting from the implementation of Chongqing’s national urban–rural integrated reform pilot zone policy in 2007. During this stage, with the rollout of URI, rural revitalization, and similar strategies, the level of urbanization continued to rise, the spillover effects of cities on neighboring rural regions became prominent, and an initial reciprocal mechanism linking urban–rural areas was established. The circulation of production factors transformed from unidirectional to bidirectional, and disparities in education, healthcare, and social security across urban–rural areas gradually narrowed, overall exhibiting the characteristics of high-quality integration.

4.3. Correlation Analysis

We used the Spearman’s correlation analysis method to explore the mechanism by which URI influences ATI. The results indicate a significant correlation between the URI level and ATI (Figure 4). A noteworthy positive correlation was observed between the URI level and ATI, with a correlation coefficient of 0.39. URI facilitates the breakdown of barriers to the flow of elements between urban and rural areas. It promotes comprehensive circulation and rational distribution of essential resources, including labor, capital, and technology over urban–rural areas, thereby injecting new vitality into ATI. However, this analysis does not account for other control factors and the impacts of temporal variations. Therefore, subsequent sections will use econometric analysis methods to explore this further.

5. Mechanism Analysis

5.1. Benchmark Regression

Based on the models and data described above, this section employs a stepwise regression approach for fixed effects estimation, specifically testing the potential of URI to promote ATI. The results of the model are presented in detail in

Table 3. Benchmark regression results of URI’s impact on ATI.

var	ATI
	(1)	(2)	(3)	(4)	(5)	(6)	(7)
RUI	0.53 *** (0.03)	0.29 *** (0.04)	0.3 *** (0.04)	0.38 *** (0.04)	0.38 *** (0.04)	0.4 *** (0.04)	0.43 *** (0.04)
lnpgdp		0.41 *** (0.05)	0.45 *** (0.05)	0.59 *** (0.06)	0.58 *** (0.06)	0.56 *** (0.06)	0.47 *** (0.06)
lnopen			−2.06 *** (0.6)	−2.33 *** (0.61)	−2.29 *** (0.61)	−2.21 *** (0.61)	−2.67 *** (0.61)
lnInd				0.24 *** (0.05)	0.25 *** (0.05)	0.25 *** (0.05)	0.23 *** (0.05)
lnenv					23.98 *** (7.5)	23.49 *** (7.49)	26.68 *** (7.46)
lnrcr						−0.23 *** (0.05)	−0.29 *** (0.06)
lnag-r&d							0.08 *** (0.01)
Cons	−0.55 *** (0.08)	−4.45 *** (0.46)	−4.71 *** (0.47)	−6.52 *** (0.6)	−6.51 *** (0.6)	−5.48 *** (0.65)	−5.01 *** (0.66)
FE(City)	YES
FE(Year)
N	5760	5760	5760	5760	5760	5760	5760

Note: Significant confidence levels of 1% are marked by ***, with standard errors shown in parentheses.

. In column (1), the estimated coefficient of URI on ATI is 0.53 (p < 0.01), excluding the control variables. This indicates that, between 1999 and 2018, URI had a positive effect on promoting ATI without the control variables. There are two potential explanations for this phenomenon. Firstly, the development of URI helps to break down the barriers to the circulation of resources across urban-rural areas, altering the unidirectional “rural-to-urban” movement pattern of production factors, optimizing the efficiency of resource allocation, and fostering new agricultural formats. Consequently, this process promotes the transformation and advancement of the agricultural industry, ultimately enhancing ATI’s capacity. Secondly, URI facilitates the removal of systemic and institutional obstacles, optimizes grassroots governance, and accelerates the process of equalizing public services across urban-rural areas. This provides institutional guarantees and social support for enhancing the capability of ATI. Additionally, columns (2)–column (6) present the fixed effects regression results with gradually added control variables. During this process, the core explanatory variable remains significant (p < 0.01) and the sign does not change, which means that the conclusion that URI can positively promote ATI remains robust. Hence, H1a is supported.

According to the estimation results of the control variables: (1) Regional economic level. The estimated coefficient is significantly positive (p < 0.01), indicating that, as farmers’ income levels continue to increase, a portion of this income can be transformed into agricultural capital and R & D investments, effectively solidifying the material foundation for local technology innovation and thereby promoting progress in agricultural technology. (2) Degree of regional openness. The coefficient is significantly negative (p < 0.01), suggesting that, as foreign capital and multinational corporations gradually engage in the R & D and production segments of China’s agricultural industry chain, domestic companies tend to directly introduce advanced foreign agricultural technologies, reducing the investment in agricultural science and technology funds, thus having a negative spillover effect on Chinese ATI. (3) Regional industrial structure. The coefficient is positive and significant (p < 0.01), indicating that the development of the primary sector in the region effectively promotes technology progress and ATI. (4) Regional environmental level. The estimated coefficient is significantly positive (p < 0.01), meaning that, as the level of urban environmental regulations increases, participants in the agricultural industry chain will continue to increase R & D investment to adapt to environmental policy requirements, promote the application of new production technologies, and pursue innovation compensation benefits, thereby enhancing the level of ATI. (5) Regional climate risk. The coefficient is significantly negative (p < 0.01), suggesting that high climate risk regions may require more resources to mitigate the adverse effects of climate change, thereby constraining the application and innovation of agricultural technologies to a certain extent. (6) Agricultural R & D input. The coefficient is positive and significant (p < 0.01), suggesting that investments in agricultural research funding and human capital in agricultural research effectively enhance the capacity and intensity of ATI, driving progress in agricultural science and technology.

5.2. Heterogeneity Regression

5.2.1. Dimensional Analysis

A further analysis was conducted to assess the impact of the five dimensions of URI on ATI.

Table 4. Heterogeneity regression results of URI’s impact on ATI by dimension.

var	ATI					ATIaf	ATIame	ATIasm
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
RUI						0.42 *** (0.05)	0.45 *** (0.06)	0.46 *** (0.11)
RUIpop	0.18 *** (0.04)
RUIecon		0.43 *** (0.04)
RUIsoc			0.16 *** (0.02)
RUIspa				0.11 *** (0.03)
RUIecol					0.07 *** (0.03)
lnpgdp	0.58 *** (0.06)	0.58 *** (0.06)	0.6 *** (0.06)	0.63 *** (0.06)	0.67 *** (0.06)	0.48 *** (0.07)	0.47 *** (0.08)	0.61 *** (0.14)
lnopen	−2.09 *** (0.6)	−2.6 *** (0.61)	−2.15 *** (0.6)	−1.99 *** (0.61)	−2.14 *** (0.61)	−2.25 *** (0.66)	−3.14 *** (0.79)	1.68 (1.33)
lnInd	0.15 *** (0.05)	0.47 *** (0.06)	−0.01 (0.05)	0.11 ** (0.05)	0.05 (0.05)	0.19 *** (0.06)	0.19 *** (0.06)	0.18 (0.12)
lnenv	24.36 *** (7.48)	23.79 *** (7.46)	24.45 *** (7.45)	27.66 *** (7.49)	26.1 *** (7.48)	37.14 *** (8.49)	7.43 (9.25)	−10.12 (17.24)
lnrcr	−0.23 *** (0.06)	−0.26 *** (0.05)	−0.22 *** (0.05)	−0.27 *** (0.06)	−0.25 *** (0.06)	−0.37 *** (0.06)	0.01 (0.07)	−0.62 *** (0.14)
lnag-r&d	0.08 *** (0.01)	0.09 *** (0.01)	0.09 *** (0.01)	0.07 *** (0.01)	0.07 *** (0.01)	0.08 *** (0.02)	0.1 *** (0.02)	0.16 *** (0.04)
Cons	−6.3 *** (0.65)	−7.07 *** (0.63)	−6.07 *** (0.65)	−6.48 *** (0.65)	−6.7 *** (0.64)	−4.82 *** (0.72)	−6.85 *** (0.85)	−7.17 *** (1.62)
FE(City)	YES
FE(Year)
N	5760	5760	5760	5760	5760	5740	5760	5460

Note: Significant confidence levels of 5%, and 1% are marked by **, and ***, respectively, with standard errors shown in parentheses.

shows that the coefficients for all five dimensions of URI are significant (p < 0.01), indicating that each dimension of URI positively promotes ATI to varying degrees. Notably, the economic dimension of URI exerts the most significant positive effect on ATI. The flow of capital elements enhances the funding for agricultural research, solidifying the material foundation for ATI. The population and social dimensions of URI also significantly promote ATI, ranking just behind the economic dimension in impact. The primary labor force and researchers are critical carriers of ATI. The implementation of policies such as “local talents returning to start businesses” and “talent introduction” has optimized the allocation of human capital across urban–rural areas, thereby enhancing ATI capabilities to a certain extent. The proximity in social customs and living habits further facilitates the bidirectional flow of human capital and other factors across urban–rural areas. The spatial and ecological dimensions of URI have a relatively weaker, yet still positive, effect on ATI. Infrastructure development is essential for strengthening the connectivity of elements across urban–rural areas. This facilitates the spillover of traditional and innovative production factors from cities to rural areas, forming a fundamental base for enhancing ATI. Additionally, as strategies such as carbon neutrality are implemented, they impose higher demands on agricultural development, thereby driving ATI. Moreover, regions with better ecological environments also hold additional advantages in attracting labor, capital, and other factors. Therefore, H1b is supported.

In addition, to explore the impact of URI on different types of ATI, this article con-ducts regression analyses based on three types of agricultural technology, including farming and breeding techniques, agricultural mechanical engineering technology, and agricultural service management technology. The results are shown in columns (6)–(8) of Table 4, indicating that the estimated coefficients of URI for these three types of ATI are all positive (p < 0.01). This suggests that URI significantly promotes the development of various agricultural technologies to different extents. Among these, the promotion effect on agricultural service management technology innovation is the most significant. This may be due to the reliance of such technology innovation on the transfer of symbolic knowledge, which often requires high geographic proximity and frequent face-to-face communication and social interaction during the innovation process. URI effectively re-duces the spatial and temporal distances between urban and rural areas, improving the efficiency of knowledge dissemination for agricultural service management technology. Therefore, hypothesis H1c is confirmed.

5.2.2. Regional Analysis

Given each region’s diverse resource endowments, geographical locations, and development stages, there is also significant heterogeneity in URI and ATI. Therefore, this article categorized the nation into East, Central, and West regions based on the statistical system and classification standards of administrative divisions provided by the National Bureau of Statistics. On this basis, the benchmark model was re-analyzed, with the results displayed in columns (1) to (3) of

Table 5. Heterogeneity regression results by the regional impact of URI on ATI.

var	ATI
	EST	CTR	WST	Urban	Rural
	(1)	(2)	(3)	(4)	(5)
RUI	0.38 *** (0.07)	0.34 *** (0.09)	0.32 *** (0.09)	0.44 *** (0.05)	0.37 *** (0.08)
lnpgdp	0.31 *** (0.1)	0.52 *** (0.12)	0.39 *** (0.12)	0.46 *** (0.06)	0.3 *** (0.1)
lnopen	−4.04 *** (0.82)	−1.45 (1.27)	−5.73 ** (2.54)	−2.83 *** (0.62)	1.34 (1.02)
lnInd	0.09 (0.09)	0.54 *** (0.09)	−0.32 *** (0.1)	0.18 *** (0.05)	0.31 *** (0.08)
lnenv	8.87 (11.27)	29.64 ** (13.18)	36.93 ** (14.86)	20.74 *** (7.71)	40.6 *** (12.99)
lnrcr	0.19 ** (0.09)	−0.73 *** (0.1)	−0.61 *** (0.1)	−0.29 *** (0.06)	0.06 (0.09)
lnag-r&d	0.07 *** (0.02)	0.03 (0.03)	0.11 *** (0.03)	0.09 *** (0.01)	0.09 *** (0.02)
Cons	−4.66 *** (1.07)	−4.32 *** (1.34)	−2 (1.31)	−4.77 *** (0.69)	−6.77 *** (1.08)
FE(City)	YES
FE(Year)
N	2000	2000	1760	5760	5720

Note: Significant confidence levels of 5%, and 1% are marked by **, and ***, respectively, with standard errors shown in parentheses.

. Overall, URI has a significant positive impact on ATI across all regions. However, from a regional perspective, the effect of URI on ATI varies significantly among the regions, presenting a transparent East-to-West gradient, indicating that the facilitating effect of URI on ATI is more pronounced in developed areas. The likely reasons are that Eastern developed regions typically possess better institutional guarantees, higher levels of grassroots governance, and well-developed infrastructure, which can well undertake the promotion effect of URI. Additionally, the mobility and rational allocation of elements across urban–rural areas effectively promote the development of ATI in the Eastern region. Compared to the Western region, the Central area is often the first to receive spillovers of capital and other elements from the more developed Eastern region, and URI further promotes the total flow of the abovementioned elements across urban–rural areas, which provides the necessary conditions for ATI. The Western region, due to its relatively weaker developmental foundation and larger urban–rural disparities, traditionally lags in ATI. However, as URI continues to advance, it offers new perspectives and drives for ATI’s development in the Western region.

According to the standards above, ATI was analyzed separately for urban and rural areas, as shown in columns (4) and (5) of Table 5. URI significantly and positively influences ATI in urban and rural areas. However, its positive effect on ATI is weaker in rural regions compared to urban ones, possibly due to the following reasons: Firstly, although URI promotes the bidirectional flow of elements across urban–rural areas, the scale and efficiency of high-quality production factors, especially high-end talent, advanced technology, and capital, are lower in rural areas compared to urban ones. These factors tend to be concentrated around urban peripheries or economically more developed rural areas, leading to a weaker impact of URI on technology innovation in remote rural areas. Secondly, rural areas lack sufficient economic incentives to attract investment in ATI due to limited agricultural returns in regions where traditional agriculture predominates. Additionally, although URI adds necessary elements for ATI to rural areas, the limited capacity of these areas to absorb, manage, and create an innovative environment means that they cannot fully utilize these innovation factors, thereby somewhat weakening the impact of URI on ATI. Finally, constrained by education levels, acceptance capabilities, and risk preferences, farmers often show little willingness to adopt new technologies, which also affects the spread and application of ATIs in rural areas.

5.2.3. Agricultural Production Area Analysis

According to China’s 12th Five-Year Plan, agricultural production areas are classified into several key regions: the Gansu–Xinjiang production area (GXPA), the Hetao Irrigation District (HIDPA), the Northeast Plain (NPPA), the Fenwei Plain (FWPPA), the Huang –Huai–Hai Plain (HHHPPA), the Southern China production area (SCPA), and the Yangtze River production area (YRPA). The impacts of URI on ATI vary significantly across these regions, as detailed in the regression results in

Table 6. Heterogeneity regression results of URI’s impact on ATI by agricultural production area.

var	ATI
	GXPA	HIDPA	NPPA	FWPPA	HHHPPA	YRPA	SCPA
	(1)	(2)	(3)	(4)	(5)	(6)	(7)
RUI	0.76 *** (0.25)	0.07 (0.33)	−0.03 (0.14)	0.16 (0.23)	0.24 * (0.13)	0.54 *** (0.09)	0.27 *** (0.1)
lnpgdp	−0.05 (0.34)	0.39 (0.32)	0.09 (0.16)	0.7 ** (0.29)	−0.11 (0.18)	−0.15 (0.13)	0.71 *** (0.15)
lnopen	−22.03 (17.73)	−10.35 * (5.57)	−0.12 (1.26)	−1.64 (4.75)	0.1 (1.14)	2.62 ** (1.12)	−1.31 (1.35)
lnInd	−0.63 ** (0.26)	−0.1 (0.31)	0.27 * (0.16)	0.32 (0.2)	−0.05 (0.14)	0.17 * (0.1)	0.13 (0.13)
lnenv	−18.75 (45.39)	6.84 (34.17)	69.57 *** (20.12)	55.47 ** (26.3)	35.89 ** (16.43)	15.97 (12.5)	−4.82 (19.14)
lnrcr	−0.29 (0.24)	−0.3 (0.26)	0.7 *** (0.2)	−0.5 (0.41)	0.12 (0.15)	−0.15 (0.12)	−0.95 *** (0.18)
lnag-r&d	0.15 * (0.09)	0.5 *** (0.13)	−0.03 (0.05)	−0.05 (0.06)	0.1 ** (0.05)	0.08 *** (0.03)	−0.02 (0.04)
Cons	1.65 (3.57)	−5.51 (3.63)	−3.36 * (1.9)	−5.36 (3.32)	−0.36 (2.14)	−0.17 (1.39)	−4.73 *** (1.55)
FE(City)	YES
FE(Year)
N	280	280	680	420	920	1920	1080

Note: Significant confidence levels of 10%, 5%, and 1% are marked by *, **, and ***, respectively, with standard errors shown in parentheses.

. It is shown that URI has an insignificant effect on promoting ATI in the HIDPA, FWPPA, and NPPA. Conversely, URI significantly positively impacts ATI in the GXPA, HHHPPA, YRPA, and SCPA. Moreover, the effect of URI on ATI varies considerably among these agricultural production areas. Among them, the level of URI is notably higher in the HHHPPA, YRPA, and SCPA, characterized by frequent bidirectional flows of capital and technology across urban–rural areas, which effectively foster the development of ATIs. In the GXPA, establishing experimental zones for URI has effectively overcome institutional barriers, promoted the movement of elements, and enhanced ATI capacity.

5.3. Robustness Test

This article carried out five methods of robustness checks, including the panel Pois-son regression model, winsorize, exclusion of municipalities (Beijing, Shanghai, Tianjin, and Chongqing), introduction of interaction fixed effects, and adjustment of the study period. Firstly, winsorize. To eliminate the influence of outliers, all variables were processed at the 1% and 99% percentiles, with the regression results shown in column (1) of

Table 7. Robustness test results of URI’s impact on ATI.

var	ATI
	Winsorize	Change of Method	Deletion of Municipalties	Interaction Fixed Effect	Changing the Time Period of the Research
	(1)	(2)	(3)	(4)	(5)
RUI	0.32 *** (0.04)	1.01 *** (0.02)	0.39 *** (0.04)	0.86 *** (0.04)	0.41 *** (0.07)
lnpgdp	0.76 *** (0.05)	0.92 *** (0.03)	0.52 *** (0.06)	0.89 *** (0.04)	0.38 *** (0.09)
lnopen	−2.84 *** (0.78)	−0.25 * (0.15)	−2.42 *** (0.61)	−6.68 *** (0.61)	1.24 (0.98)
lnInd	0.31 *** (0.05)	−0.01 (0.03)	0.25 *** (0.05)	0.77 *** (0.04)	0.34 *** (0.08)
lnenv	23.42 *** (8.65)	9.17 *** (1.94)	24.91 *** (7.49)	39.34 *** (7.71)	22.27 ** (9.24)
lnrcr	−0.28 *** (0.05)	−0.12 *** (0.02)	−0.27 *** (0.05)	−0.53 *** (0.05)	−0.25 *** (0.07)
lnag-r&d	0.11 *** (0.02)	0.11 *** (0.01)	0.07 *** (0.02)	0.08 *** (0.01)	0.07 *** (0.02)
Cons	−8.15 *** (0.62)	−6.63 *** (0.29)	−5.57 *** (0.65)	−9.23 *** (0.49)	−4.19 *** (1.06)
FE(City)	YES
FE(Year)
N	5760	5760	5680	5760	2592

Note: Significant confidence levels of 10%, 5%, and 1% are marked by *, **, and ***, respectively, with standard errors shown in parentheses.

. Secondly, the estimation method was changed. After switching to the panel Poisson regression model, the results are presented in column (2). Thirdly, given the significant differences in fiscal and industrial structures between municipalities and other cities, four municipalities were excluded from the sample to avoid the effects of anomalous data, with the results shown in column (3). Fourthly, interaction fixed effects between individuals (cities) and time were introduced to eliminate the interference of unobservable factors that vary with both time and individuals, thereby controlling for the heterogeneous impacts of multidimensional shocks across different cities. The results are presented in column (4). Lastly, a robustness check was conducted by selecting a different study period (2010–2018), with the results shown in column (5). Overall, the estimated coefficients of the core explanatory variables remain significantly positive, consistent with the baseline regression results. Therefore, the findings of this study are robust, indicating that URI positively promotes ATI.

5.4. Further Analysis

5.4.1. Moderation Effect Test

According to the theoretical analysis mentioned earlier, governance systems could moderate the relationship between URI and ATI. Based on this premise, this article incorporated an interaction term between URI and governance system indicators into the model to examine the moderating effect of governance systems. Column (1) of

Table 8. Results of the moderation effect test of the governance system.

var	ATI	ATIest	ATImid	ATIwes	ATIurb	ATIrur
	(1)	(2)	(3)	(4)	(5)	(6)
RUI × MAN	0.17 *** (0.03)	0.32 *** (0.06)	0.34 *** (0.07)	0.06 *** (0.05)	0.17 *** (0.03)	0.06 * (0.06)
RUI	0.83 *** (0.04)	0.78 *** (0.07)	0.66 *** (0.08)	1.07 *** (0.09)	0.81 *** (0.04)	0.97 *** (0.08)
MAN	−0.06 (0.05)	0.28 ** (0.13)	0.7 *** (0.12)	−0.53 *** (0.08)	−0.07 (0.06)	0.03 (0.09)
Cons	−9.93 *** (0.51)	−10.65 *** (0.86)	−8.17 *** (0.94)	−6.7 *** (1.02)	−10.04 *** (0.52)	−13.25 *** (0.89)
Ctrls	YES
FE(City)
FE(Year)
Prob > x2	0.00	0.00	0.00	0.00	0.00	0.00
N	5760	2000	2000	1760	5760	5760

Note: Significant confidence levels of 10%, 5%, and 1% are marked by *, **, and ***, respectively, with standard errors shown in parentheses.

presents the regression results. The coefficient of URI is significantly positive (p < 0.01), and the interaction term coefficient between URI and governance systems is 0.17, also positive and significant (p < 0.01). This suggests that governance systems significantly enhance the positive impact of URI on ATI. This is because well-developed governance systems help to break down urban–rural barriers, smooth the channels of element flow, optimize the allocation of resources for innovation entities, and enhance research and development outputs. At the same time, well-developed governance systems support rural management innovation, guiding the flow and aggregation of various talents to rural areas, and providing human and intellectual resources for agricultural technology services. These findings indicate that governance systems have a positive moderating effect between URI and ATI, confirming the validity of H2a.

However, the moderating effects also exhibit significant regional disparities due to varying degrees of governance system development across regions. Given this, the article investigated the impact of URI on ATI from a regional heterogeneity perspective, dividing the regions into different parts (Eastern, Central, Western, urban, rural). The results can be found in columns (2) to (6). From Eastern, Central, and Western viewpoints, the interaction coefficients between URI and governance systems are 0.32, 0.34, and 0.06, respectively, all positive and significant (p < 0.01). This demonstrates that the significant moderating effect of governance systems is widely present across these regions. This clearly shows that research hypothesis H2a is supported across these regions. Among them, the moderating effect of governance systems is particularly significant in the Central region. This could be due to the Eastern region’s higher economic development and smoother URI process, where governance systems are already well-established. Therefore, while the moderating effect of governance systems is present in the Eastern region, its impact is limited. In contrast, due to lower economic development and larger urban–rural disparities, the Western region experiences weaker influence from governance systems because of limitations in infrastructure and resource allocation. Meanwhile, the Central region, serving as a transitional area, benefits from the development advantages of the Eastern region while also confronting the developmental bottlenecks of the West. In this context, the role of governance systems and policy adjustments in the central region becomes more critical, significantly enhancing the potential for ATI through optimized management practices, playing a more pronounced moderating role in URI and ATI. From the urban–rural perspective, the interaction coefficients between URI and governance systems are 0.17 and 0.06, respectively, both positive and significant (p < 0.1). This indicates that the moderating effects of governance systems are significant in both areas. In urban areas, the moderating effect of governance systems is particularly notable. This is likely because urban governance systems are more mature and effectively implemented, guiding resources towards dominant industries, facilitating industrial upgrading and transformation, improving living standards and public services for urban and rural residents alike, optimizing the innovation environment, and stimulating innovative activities. In contrast, rural areas are relatively backward in terms of infrastructure, information transmission systems, and resource allocation capacity, which leads to a weaker implementation and effectiveness of governance systems compared to urban areas. Consequently, the moderating role of governance systems is relatively limited in rural areas.

The moderating role of mature markets in URI to promote ATI needs to be further explored. As displayed in

Table 9. Results of the moderation effect test of mature markets.

var	ATI	ATIest	ATImid	ATIwes	ATIurb	ATIrur
	(1)	(2)	(3)	(4)	(5)	(6)
RUI × MAR	0.39 *** (0.03)	0.52 *** (0.05)	0.85 *** (0.06)	0.23 *** (0.06)	0.37 *** (0.03)	0.31 *** (0.06)
RUI	0.73 *** (0.04)	0.76 *** (0.07)	0.63 *** (0.08)	0.75 *** (0.09)	0.72 *** (0.04)	0.8 *** (0.07)
MAR	0.57 *** (0.06)	0.72 *** (0.1)	0.75 *** (0.1)	0.38 *** (0.13)	0.53 *** (0.06)	0.92 *** (0.11)
Cons	−7.31 *** (0.58)	−7.19 *** (0.94)	−8.26 *** (1.06)	−3.89 *** (1.18)	−7.32 *** (0.6)	−10.14 *** (0.97)
Ctrls	YES
FE(City)
FE(Year)
Prob > x2	0.00	0.00	0.00	0.00	0.00	0.00
N	5760	2000	2000	1760	5760	5760

Note: Significant confidence levels of 1% are marked by ***, with standard errors shown in parentheses.

, the results of the moderation effect test are presented. In the moderation model examining the relationship between mature markets, URI, and ATI, the coefficient for URI is significantly and positive (p < 0.01). The coefficient for the interaction term between URI and mature markets is 0.39, which is also significantly and positive (p < 0.01). The findings indicate that mature markets have a positive moderating effect on the relationship between URI and ATI. As mature markets intensify, it enhances the facilitative effect of URI on ATI, supporting hypothesis H2b. A possible explanation is that regions with more developed markets tend to be more attentive to market information and consumer demands in terms of awareness and capability. Based on these insights, they allocate resources effectively, improving resource use efficiency and advancing ATI to swiftly adapt to market and meet consumer needs.

Regionally, the positive moderating effect of mature markets on URI and ATI is the strongest in the Central region, moderate in the East, and the weakest in the West. A possible explanation is that Central China is at the heart of China’s “Rise of Central China” strategy and is currently at a critical stage of transformation and upgrading, where the demand for ATI is particularly urgent. The strategy of URI facilitates the flow of talents and technology, and the role of mature markets is especially significant in this process, effectively transforming local resource advantages into practical outcomes of ATI, enhancing the practicality and market adaptability of these innovations. The Eastern region, with its mature markets system and highly competitive resource allocation environment, experiences a diminishing marginal effect of innovation activities despite high levels of URI, which limits the impact of mature markets. In the Western region, the effectiveness of mature markets in positively moderating the impact of URI on ATI is limited due to the region’s outdated infrastructure, imperfect market systems, and complex geographical conditions.

From the perspective of urban–rural relationships, mature markets play a more significant positive moderating role in URI and ATI in urban areas compared to rural areas. A likely reason for this is that urban areas possess the abundant resources, well-established infrastructure, and mature markets systems required for ATI activities. Consequently, mature markets can more effectively promote knowledge exchange and technology dissemination, identify market opportunities, and accelerate the conversion and implementation of agricultural science and technology achievements, thereby driving the progress of agricultural innovation and industry upgrading. In contrast, rural areas struggle with inadequate infrastructure, hindering the smooth dissemination of information and technology. Additionally, the outflow of talent reduces the local capacity for innovation and slows the speed of knowledge renewal. As a result, while mature markets do stimulate ATI activities in rural areas to some extent, its motivating effect is not fully achieved due to these structural barriers.

5.4.2. Threshold Effect Test

To further investigate the potential nonlinear relationship between URI and ATI, this article used the URI index as a threshold variable and employs the Bootstrap resampling method, repeating the process 300 times to assess threshold effects. As demonstrated in

Table 10. Results of the threshold effect of URI’s impact on ATI.

Region	var	Threshold	F	p	Boundary Value			BS
					10%	5%	1%
ATIall	RUI	single	641.83	0.000	70.456	91.136	124.487	300
		double	126.75	0.007	56.68	65.077	108.406	300
		triple	63.62	0.82	300.877	356.452	462.242	300
ATIest		single	334.44	0.000	49.254	66.153	98.634	300
		double	117.1	0.003	38.509	54.26	85.368	300
		triple	28.81	0.707	179.793	217.354	262.864	300
ATImid		single	313.85	0.000	55.761	65.649	94.952	300
		double	30.41	0.227	54.567	83.783	162.592	300
ATIwes		single	136.6	0.013	54.097	70.517	147.718	300
		double	46.00	0.06	37.2	47.795	71.124	300
		triple	15.03	0.733	57.778	69.836	98.97	300
ATIurb		single	750.63	0.000	62.614	79.226	103.588	300
		double	198.29	0.000	53.1	73.2	92.871	300
		triple	70.01	0.787	380.681	438.836	520.135	300
ATIrur		single	181.00	0.000	56.336	70.524	92.252	300
		double	40.54	0.16	49.005	57.836	83.97	300

, the results of the threshold effect indicate that both the single and double threshold effects of URI pass the test (p < 0.01), whereas the triple threshold effect does not. This illustrates a meaningful nonlinear connection between URI and ATI, implying that ATI shifts in response to fluctuations in URI intensity. On the other hand, the impact of URI intensity on ATI exhibits a significant spatial heterogeneity. Specifically, double thresholds exist in the Eastern, Western, and urban areas, whereas a single threshold is observed in the Central and rural areas.

Table 11. Results of the panel threshold regression model of URI’s impact on ATI.

ATI (1)		ATIest (2)		ATImid (3)		ATIwes (4)		ATIurb (5)		ATIrur (6)
RUI ≤ 0.63	0.04 ** (0.02)	RUI ≤ 0.55	−0.15 *** (0.03)	RUI ≤ 0.83	0.18 *** (0.02)	RUI ≤ 0.65	0.05 *** (0.03)	RUI ≤ 0.63	0.02 *** (0.02)	RUI ≤ 0.64	0.09 *** (0.02)
0.63 < RUI < 0.71	0.13 *** (0.02)	0.55 < RUI < 0.66	−0.06 ** (0.03)	RUI > 0.83	0.33 *** (0.02)	0.65 < RUI < 0.81	0.09 *** (0.03)	0.63 < RUI < 0.71	0.11 *** (0.02)	RUI > 0.64	0.15 *** (0.02)
RUI ≥ 0.71	0.22 *** (0.02)	RUI ≥ 0.66	0.06 ** (0.03)			RUI ≥ 0.81	0.18 *** (0.03)	RUI ≥ 0.71	0.21 *** (0.02)
Cons	0.05 * (0.03)	−0.22 *** (0.06)		0.3 *** (0.06)		0.17 ** (0.07)		0.03 (0.03)		0.08 *** (0.03)
Ctrls	YES
City-FE
Year-FE
R2	0.26	0.36		0.29		0.21		0.28		0.13
N	5760	2000		2000		1760		5760		5760

Note: Significant confidence levels of 10%, 5%, and 1% are marked by *, **, and ***, respectively, with standard errors shown in parentheses.

shows the estimated parameters for the threshold effects of URI on ATI. The results reveal a positive scale effect, indicating that the impact of URI on ATI intensifies as it develops and becomes more pronounced as the threshold variable exceeds certain threshold values. Specifically, with lower URI levels, a 1% rise in URI results in a 0.04% growth in ATI. When URI is at an intermediate level, each 1% increase in URI results in a 0.13% growth in ATI. When URI reaches a higher level, its impact on ATI is even more significant, with a 0.22% increase for each 1% rise in integration. Overall, these findings indicate varying degrees of threshold effects in URI. As these thresholds are surpassed, the role of URI in advancing ATI is further enhanced, demonstrating a “positive marginal increasing effect”. This supports hypothesis H1d.

Regionally, the impact of URI on the development of ATI shows significant heterogeneity across different areas. In the Eastern region, ATI development is partially restrained until URI development exceeds the second threshold. However, as URI surpasses the first threshold value, the inhibitory effect gradually diminishes. One possible explanation is that, in the initial stages of URI, considering factors such as market and cost, some innovative resources and industries relocate to the Central and Western regions, which offer lower costs and greater potential, thereby imposing certain constraints on ATI’s short-term development in the East. As the Eastern region progresses into advanced stages of URI, it can establish a more refined innovation system and a more efficient resource allocation mechanism, thereby positively promoting the development of ATI. After surpassing the threshold values of URI, the Central–Western regions exhibit a more pronounced promotion impact on ATI. This may be attributed to these regions starting their URI development later than others. Emulating and learning from the advanced experiences of the Eastern region serves as an effective pathway for rapid development, suggesting that less developed areas can leverage strong latecomer advantages.

From an urban–rural perspective, URI significantly boosts ATI once URI development surpasses threshold values. However, the impact is more significant in urban areas than in the rural areas. A likely reason for this disparity is that urban areas benefit from tight industry–academia–research collaboration networks, efficient information dissemination mechanisms, and strong policy support. These advantages enable the swift integration and extensive deployment of new technologies, speeding up the modernization of agricultural production methods. Consequently, the impact of URI on ATI is more pronounced in urban areas.

6. Discussion

Much extant literature focuses separately on the factors influencing ATI in urban–rural areas [43,48], with little attention to the impact of urban–rural relationships. Therefore, this article primarily explored how URI affects ATI and analyzed how governance systems and mature markets moderate the relationship between URI and ATI. This research findings indicate that URI positively influences ATI. This is confirmed by García-Cortijo et al. and Pardey et al., who suggest that strengthening exchanges among urban, suburban, and rural areas, together with enhancing the capacity for allocating agricultural innovation resources, is crucial for promoting agricultural innovation [44,48].

From a dimensional perspective, the positive effect of the economic dimension of URI is most significant, followed by the demographic and social dimensions, while the spatial and ecological dimensions have relatively weaker impacts. This finding corroborates with Mwangi’s conclusion that economic factors are key determinants influencing smallholder farmers’ adoption of new agricultural technologies [104]. However, unlike Wang et al. and Xu et al.’s works, which have overlooked spatial and ecological factors [16,39], this article reveals that these two dimensions of factors actually demonstrate a comparatively weak, but significant, positive influence. Spatial integration provides infrastructural support [105] and optimizes the physical environment, while ecological integration introduces new requirements for agricultural development; both serve as indirect drivers for the advancement of ATI.

As far as regional difference, this article finds that the effect of URI on ATI exhibits a gradient distribution characterized by “East–Central–West”, which is probably because urban–rural segregation is pronounced in the Central and Western regions. However, GXPA in the Western area has a significant effect of URI on ATI. This may be because the policy implementation, such as rural finance or national modern agricultural demonstration zone policies [106], has promoted the flow of production factors [107], thereby enhancing the level of ATI in GXPA. The results further show that the positive effect of URI on ATI in rural areas is weaker than that in urban areas. A possible explanation is that the flow of innovation factors tend to favor urban areas or advanced rural areas, and the limited absorption capacity and management levels of remote rural areas lead to a weaker impact of URI on their technological innovation [19,22,23].

Moreover, this article proved that governance systems significantly enhance the positive effect of URI on ATI. Yin et al. also confirmed that strengthening institutional infrastructure and optimizing rural governance systems facilitate the inflow of technology and talents, consequently creating a favorable environment for rural innovation and entrepreneurship [108]. Regarding the result that mature markets have a positive moderating effect between URI and ATI, Liu et al. confirmed that a high degree of marketization helps to expand fiscal scale and improve fiscal efficiency, thereby increasing investment in agricultural science and technology R & D and effectively promoting the improvement in ATI [106].

Additionally, this article indicated that the impact of URI on ATI exhibits a double-threshold effect and shows a positive marginal increasing trend. As Long-gao and Wigboldus et al. found, in the early stages of URI, the urban–rural dual structure is significant, and the flow of factors is mainly unidirectional [89,90], which restricts the development of ATI in urban and rural areas. However, with the continuous deepening of URI, the urban–rural barriers are broken, and agricultural innovation resources can be effectively allocated, enhancing the driving force for the development of ATI [32].

It is worth noting that this article still has certain limitations. Although the impact of URI on ATI was explored, gaps remain at the policy level. First, the article did not thoroughly examine policies that promote the dissemination of knowledge and technology or specify the areas they support, which are vital for advancing ATI. Second, regional differences in policy effectiveness, particularly the challenges faced in rural areas, were not sufficiently analyzed. Future research should focus on policy development and its regional variations to more comprehensively understand the role of URI in the development of ATI.

7. Conclusions and Policy Implications

7.1. Conclusions

Drawing on panel data spanning from 1999 to 2018 from prefecture-level cities in China, this article explored the impact of URI on ATI and the moderating roles of governance systems and mature markets in their relationship. Additionally, a threshold effect model was formulated to assess the nonlinear relationship between URI and ATI. It drew three conclusions as follows.

(1) URI significantly and positively promotes ATI, displaying heterogeneous characteristics. In terms of different dimensions of URI, economic integration has a particularly noticeable positive effect on ATI, followed by population and social integration, while spatial and ecological integration has a relatively weaker role. From the perspective of the types of ATI, URI had a more significant promotion effect on agricultural service management technology innovation, which is based on symbolic knowledge than on analytical knowledge. Moreover, developed and urban areas get more benefits on the development of ATI from URI than developing and rural areas.

(2) The impact of URI on ATI exhibits dual-threshold characteristics, and its effect on the level of ATI shows a trend of positive marginal increase as URI develops. Specifically, when URI is at a low level, its promotional effect on ATI is limited. However, as URI increases, its positive impact on ATI also rises. The threshold effect also shows spatial heterogeneity. In China, both the Eastern and Western regions exhibit dual-threshold effects, while the Central region exhibits a single-threshold effect and its promotional effect of URI on ATI is most pronounced after surpassing the threshold values. Although both urban and rural areas exhibit dual-threshold effects, the impact of URI on ATI is notably stronger in urban areas than in rural areas after surpassing these thresholds.

(3) Governance systems and mature markets significantly strengthen the positive impact of URI on ATI, also exhibiting spatial heterogeneous. Such moderating effect are the strongest in the Central area of China, the weakest in the West, and is more pronounced in urban settings than rural settings.

7.2. Policy Implications

This research has also some policy implications as follows.

(1) As URI has a notable positive impact on ATI, it is essential to strengthen urban–rural connections, such as through breaking down institutional barriers, and accelerating the equalization of urban–rural public infrastructure and services, so as to promote bi-directional flow of resources between urban and rural areas. Especially, economic URI measures, such as expanding funding channels for technological innovation and establishing a diversified capital investment system, are supposed to be prioritized to enhance economic connections between urban and rural areas. Furthermore, it is considerable that knowledge-sharing platforms and coordination mechanism between urban and rural areas can be established for promoting agricultural service management technologies in rural areas.

(2) The double-threshold effect reveals that the promotion of URI to ATI is relatively limited before a certain critical threshold. It implies that local policymakers should figure out some long-term measures to enhance the level of URI focusing on investments in innovation-related infrastructure connecting urban and rural areas, particularly in underdeveloped areas and rural regions. Differentiated policy interventions should be formulated to address the distinct needs of urban and rural areas, ensuring that all regions benefit from the agricultural innovation fostered by URI under their specific circumstances.

(3) The moderating roles of governance systems and mature markets raise a critical question of how to coordinate between governmental intervention and market mechanism, especially in developing countries with a low level of marketization. It is imperative to improve governance capabilities to establish innovation-oriented institutional environments, enhance administrative efficiency, promote effective allocation and utilization of resources, and cultivate fair competitive markets with a lower-entry barriers for agricultural innovation. More importantly, it is necessary for regional policy making to provide innovation incentives in combination of urban with rural areas, on the one hand, and strengthen intellectual property protection for diffusion and trade of agricultural innovation with a low transaction cost, on the other hand.