Evolutionary Mechanisms of Ecological Agriculture Innovation Systems: Evidence from Chongming Eco-Island, China

Abstract

Constructing an ecological–agricultural innovation system (EAIS) and exploring its evolution patterns are effective ways to overcome the current bottleneck in the world’s agricultural development, which is also an important area of concern for economic geography scholars. Based on innovation system and agricultural innovation system (AIS) theories and relevant research literature, this study constructed an analytical framework for the evolution of the EAIS and analyses the characteristics and mechanisms of the EAIS evolution at three levels. With the help of in-depth research data on the ecological agriculture of Chongming’s eco-island, this study analysed the structural composition, evolutionary paths, and mechanisms of Chongming’s ecological–agricultural innovation system (EAIS). The main findings are that the evolutionary structure of the EAIS can be divided into three levels. At the micro level, ecological–agricultural producers and operators are continuously spun off along with the expansion of industry scale under the role of market orientation, technology promotion, and regulatory management. At the meso level, all actors actively seek specialised associations to obtain innovation resources, forming an innovation network. At the macro level, ecological–agricultural regulatory actors incorporate all actors into the ecological regulatory system so that the evolutionary path of the EAIS can be regulated for cleaner production. EAIS is the evolution of innovation systems in synergy between the three levels of ‘actor’ + ‘network’ + ‘institution’.

1. Introduction

With the growing market demand for quality and diversification of agricultural products, there is an urgent need for innovation in agricultural production models and optimisation of production structures. New types of agriculture, such as organic agriculture, biological agriculture, natural agriculture, ecological agriculture, circular agriculture, sustainable agriculture, and intelligent agriculture, have been proposed one after another [1]. Among them, the ecological–agricultural model, which aims at water and soil conservation, environmental protection, improvement of green vegetation, rational adjustment of economic and crop structures, and protection of ecological balance, has been generally recognised and adopted as an agricultural development strategy in countries or regions around the world [2].

Agricultural innovation systems (AISs) are considered a fundamental approach to assess the mechanisms of agricultural science. The AIS view of agricultural innovation as a process in which technologies, practices, and institutions work together in a coordinated manner with multiple networks provides a perspective on the innovation process in agriculture. It explores the organisational, technological, and institutional innovations of AIS in terms of the evolution of innovation systems, the behaviour of innovation actors, and innovation network linkages [3]. This new perspective is not the result of knowledge transfer, but a continuous process of social, technological, and scientific cooperation of regional and higher-level systems that affect productivity and innovation performance [4]. The AIS theory was originated by P. Cooke of Cardiff University, UK, who proposed the regional innovation system theory in 1992, suggesting that a regional innovation system consists of actors in innovation activities, the linkages and operational mechanisms between actors, and environmental factors affecting innovation [5]. Later, AIS was considered a network of organisations, enterprises, and individuals, focusing on the dissemination of new products with economic use, new processes, and new forms of organisation to be disseminated and diffused within the system, together with the promotion of agricultural development by relevant institutions and policies. The composition of the AIS structural framework is consistent with that of the regional innovation system, which is based on innovation actors, innovation networks, and innovation environment elements. Numerous studies have viewed agricultural innovation as a systems project that includes the synergy of all links in the industrial production chain—all relevant sectors and stakeholders—while AIS is a synergistic system of all these actors and their innovation content, and its synergistic medium is a network organisation based on social relations interacting in a complex dynamic environment [6]. According to the innovation systems methodology research system, technology supply push, participatory development, induced innovation, and innovation systems are the main ways to improve AIS, and these also address the application of agricultural innovation methods at national, regional, and sectoral levels [7].

In summary, innovation has always been the core of agricultural development. While most of the current literature on agricultural innovation has focused on agricultural technology innovation and its diffusion, the measurement of agricultural innovation has mostly focused on investment or input–output black-box approaches, while less attention has been paid to the macro and micro factors involved in the innovation process, such as the evolution of the form and action of actors, the evolution of policy and legal frameworks, the characteristics of knowledge flows, and so on, so that research on the evolutionary mechanism of agricultural actors in the network organisation of AIS been largely neglected. Therefore, this study proposes the concept of the EAIS, analysing ecological–agricultural innovation from the theoretical systems and analytical perspectives of AIS, taking a less-developed region within a developed city—Chongming Eco-Island, in Shanghai—as an example, it empirically analyses the structural characteristics and mechanism of EAIS evolution with the help of a research questionnaire and interview data, with a view to exploring the direction of EAIS development in developing countries under the new situation. The ecological and innovation development of agriculture are both the means and ends of agricultural development, the key is to construct an innovation system that is conducive to agricultural development. This study thus aims to contribute to the further development of research on AISs and enrich the theory of regional innovation systems and explore why an ecological–agricultural innovation system can coordinate resource utilization, stabilize ecosystems, and bring excellent production environments and product quality, fostering the achievement of ecological security and sustainable development in agriculture.

The remainder of this paper is organised as follows. In Section 1 we introduce the significance and main purposes of this research. In Section 2, we review the history and nature of the AIS concept and construct an analytical framework for the evolution of EAIS based on innovation systems. Section 3 details the research methodology, research cases, and data sources of this study. Section 4 analyses the evolutionary features and mechanisms of EAIS at the micro, meso, and macro levels, describing the analysis process and results. Finally, in Section 5, the findings and perspectives of this study are summarised, the potential uses of the framework are presented, and further directions for future research are discussed.

2. Theoretical Framework

The development of agricultural innovation system theory has evolved from the national innovation system (NIS), regional innovation system (RIS), national agricultural research system (NARS), and agricultural knowledge information system (AKIS) to the present AIS. AIS is an agricultural organisational and innovation system consisting of diverse innovation actors, networked innovation processes, and integrated innovation goals, and has been increasingly applied to analyse the combined organisation of technological, social, and regulatory innovation in agriculture [6]. Its ideas focus on understanding the interactive behaviour of various actors in the innovation process, as well as the impact of policies and organisational institutions on these actors, and it and its concepts have been commonly applied to a number of countries, sectors, regions, or specific industries. Whereas early AIS emphasised economic contributions and private sector involvement over sustainability, today AIS conceptualises agricultural innovation as a system of interaction and learning between multiple stakeholders, including research organisations and non-research participants such as farmers, the private sector, civil society organisations, and policy institutions. The evolutionary elements of AIS development are an integrated understanding that includes both social culture and economic foundations, globalisation trends, and regional coordination of development environments, as well as regulatory factors that facilitate the enhancement of innovation capacity, while also being influenced by the relevant role of regional elements. These elements create informal and complex social relations within a limited area by agricultural actors (nodes) through synergies and collective learning processes among themselves, forming a self-organising system with discontinuities and ultimately constituting an integrated innovation system, following a socio-ecological evolutionary mechanism [3].

Constructing an AIS to explore mismatches and systemic failures between organisations, elements, and their linkages with government policies will help improve innovation performance and overall competitiveness, and scientifically guide the development of agricultural innovation policies. A mature innovation system or model has gradually emerged through ‘innovation intermediaries’ or brokers [8]. Technological innovation, including technology development and diffusion, is the most important innovation activity in AIS [9]. Compared with innovation in other fields, agricultural technological innovation has more prominent public goods attributes and is more regional and seasonal, which determines the special and important status of technological innovation in AIS. Organisational innovation in agriculture can make agricultural actors more adaptable to the market. It involves adjustment to original organisational arrangements, and its essence is to produce a more efficient new model or organisational vehicle. Organisational innovation in agriculture has undergone different evolutionary patterns in the course of its development, mainly involving the entrepreneurisation of farmers, the institutionalisation of agricultural organisations, and the spinoff of agricultural institutions. Agricultural regulatory innovation is the key to the evolution of the AIS and the reorganisation and coordination of the various elements of agricultural innovation. Agricultural innovation activities are influenced by regional resources and regulatory environment. Positive incentives for agricultural innovation adopted by government or agricultural management agencies—such as upgrades to the agricultural policy system, the implementation of preferential tax policies for innovation actors such as agricultural enterprises, and the strengthening of patent protection for agricultural innovation—are all regulatory innovations in agricultural development [10]. Innovation actors mainly include agricultural enterprises, university research institutions, farmers, government agencies, and intermediary services. From the perspective of spatial organisation, innovation benefits to a large extent from the combined effect of geographical proximity or clustering of innovation actors and resulting linkages. As shown in Figure 1, this proximity and interconnectedness give rise to a new form of spatial organisation, namely, innovation networks. An innovation network is a relatively stable network of relationships between innovation actors such as enterprises, universities, research institutions, market intermediaries, industry organisations, local governments, and individuals in a specific field, formed through long-term communication and cooperation between them in multiple channels, ways, and levels, and is conceived as consisting of innovation actors, innovation resources, relationship channels, and space. Innovation actors in the AIS gradually form a relatively stable innovation network through formal or informal communication and cooperation in the local environment.

AIS approaches have remained fairly focused on innovation in the agricultural sector and have often concentrated on the agricultural technology innovation paradigm, without yet taking a multifunctional research approach to agricultural systems and organisations or focusing explicitly on eco-efficiency [3]. Mainstream and alternative agriculture are very different in form and concept, making the study of EAIS important for the innovative development and sustainability of ecological agriculture [11]. AIS is the collaboration of multiple elements to achieve innovation; that is, AIS is a ‘conceptual framework’ developed from a systems theory perspective, emphasising the interaction of the various constituent elements involved in the generation and exchange of knowledge, technology, or products in a social, economic, and regulatory context. This means that the interaction of stakeholders in the AIS is a key factor in the evolution of the system [12]. Therefore, the analysis of the AIS should be carried out from a systemic perspective. This paper refers to the analytical frameworks of many scholars on the formation or evolution of the AIS [13], and constructed an analytical framework for the evolution of the EAIS that includes the three levels of ‘actor’ + ‘network’ + ‘institution’, which emphasises that the AIS includes all technological innovation, organisational innovation, and regulatory innovation involved in actor behaviour, actor relations, and the regulatory environment, and also emphasises the evolution of innovation systems in synergy between the three levels (Figure 2).

3. Methodology and Data

3.1. Methodology

In the field of economic geography, social network analysis (SNA) has become a mature set of norms and methods widely used in the study of industrial clusters, enterprise spatial structure, regional innovation systems, urban spatial structure, tourism spatial structure optimisation, and ecosystem management services [14]. In this context, the AIS research has placed great emphasis on the SNA method [15].

Degree centrality is a common metric in SNA. The degree centrality of a node is an indicator that reflects the position of the node in the network and refers to the number of other nodes connected to the node. The greater the number of connected nodes, the higher the degree centrality, the greater the innovation resources and innovation capacity, the greater the ability to obtain information and resources from other nodes, and the greater the network importance and influence (as in the following formula).

$$C_D\left(n_i\right)={{\displaystyle \sum }}_j{X}_{ij}={{\displaystyle \sum }}_i{X}_{ji},$$

where X_ij and X_ji are values of 0 or 1 representing whether node j has a relationship with node i, and mutual out-degree and in-degree relations between node i and j. As degree centrality can be divided into outward and inward centrality, it is generally referred to in terms of out-degree and in-degree centrality. The out-degree represents the outward linkage emanating from that node, which influences other nodes, while the in-degree represents the opposite, influenced by other nodes. In this study, out-degree represents the output of innovation resources from an agricultural actor to other actors, whereas in-degree represents the input of innovation resources to this agricultural actor by other actors.

3.2. Research Case

Chongming Eco-Island is located at the mouth of the Yangtze River; it is the world’s largest estuarine alluvial island and the third-largest island in China, and functions as the ‘ecological barrier’ of the Yangtze River (Figure 3). Chongming was clearly positioned as an ecological island in the Shanghai Urban Master Plan (1999–2020), approved by the State Council in 2001. In 2014, the United Nations Environment Programme (UNEP) released the International Assessment Report on Chongming Eco-Island, which comprehensively assessed the results achieved by Chongming in three major areas of sustainable development: environment, economy, and society. The report concluded that the core values of the ecological construction of Chongming Island reflected the green economy concept of the UNEP, affirming the transformation of Chongming and in particular of family-based, fragmented agricultural operations to scale and the ecological transformation of agricultural production methods, and asserted that Chongming’s eco-agriculture was coming of age. In 2017, Chongming was approved as one of China’s first agro-ecological development pioneer zones, and the ecological agriculture and rural leisure tourism industry ushered in a booming period of development. In 2020, Chongming reached 267 certified green food enterprises in the plantation industry, with a total of 460 products. The green food certification rate in Chongming has exceeded 90%, accounting for 70% of the total number of green food certifications in Shanghai. The China Agricultural Green Development Report 2020, released in Beijing in 2021 by the Chinese Academy of Agricultural Sciences and the China Agricultural Green Development Research Association, shows that Chongming’s agricultural green development index has reached 90.01, ranking first in China. Chongming’s eco-agriculture represents the highest level in China, innovative development and ecological development are the key to its success. Therefore, it is very reliable to take Chongming’s eco-agriculture as the research case of EAIS.

3.3. Data

This study used SNA and its analysis software UCINET 6 to analyse the characteristics and structure of network links between case industry organisations, and focused on the value of information and control advantages represented by the location characteristics of the network structure in which each agricultural actor is located. This method focuses on analysing the relationships between network members and exploring the impact they will have on the actions of network members, focusing more on the relational data of the actors in the network rather than the attribute data. Thus, in the AIS research, we treated actors such as farmers, agro-related enterprises, agricultural research institutions, universities, and agricultural associations as individual units, clarified the overall structure of the network based on different relational attributes, explored the position of each agricultural actor in the network, identified key nodes in the network, and identified specific sub-groups.

With the support of the Chongming Agricultural and Rural Committee and the Development and Reform Commission of Shanghai, we travelled to Chongming for more than 20 days of field research in August and December 2020 and March and April 2021. We approached Chongming’s Eco-Agriculture respondents to conduct semi-structured interviews, instructing them to indicate what organisations they had relationships with for product distribution, market regulation, technology diffusion, market collaboration, innovation collaboration, and organisational affiliation and record the responses on a questionnaire form. In the process of collecting data on the attributes and relationships of the participants in the social network, we first selected the heads of representative agricultural enterprises for interviews and then identified the heads of organisations, such as farmers, cooperatives, government departments, intermediaries, or service organisations with which the enterprise had business or cooperative relationships and asked the heads of the organisations and the farmers to fill in the number of node units with which they had relationship for each relationship question. A total of 282 respondents were interviewed in the main rural areas of Chongming’s eco-agricultural production and operation, covering all sectors related to ecological agriculture in Chongming, and 276 valid questionnaires were obtained, which were collated into the data and materials required for the SNA analysis. Specifically, the respondents were asked to answer the following questions:

• (I) What is your occupation and what is the organisation you work with?
• (II) What does your job entail and what is the business of the organisation to which you belong?
• (III) What are the departments of the organisation you belong to?
• (IV) Which organisations have a subordinate relationship with the organisation you belong to?
• (V) Which organisations have a business or partnership relationship with you or the organisation you belong to?
• (VI) What are your or your organisation’s sources of technology, products, and information?
• (VII) What local policy regimes and regulatory measures affect your or your organisation’s production and business or R & D activities?
• (VIII) What are your or your organisation’s future production, business plan, and development directions?

4. Results

4.1. Micro-Level

4.1.1. Evolutionary Features

At the micro level, ecological–agricultural producers and operators have continued to carry out organisational spinoff and have formed a multi-core-structured organisational spinoff network. Among the 276 research samples obtained, 190, or 68.8%, showed that the actor had a spinoff relationship with at least one other actor. Thus, a matrix of 190 × 190 spinoff relationships exists between ecological–agricultural actors in Chongming. By importing the matrix into UCINET and running it, we can obtain a network of spinoff relationships between ecological–agricultural actors in Chongming, and by processing the nodes with the analysis-centrality measures function, a spinoff network diagram showing the size of the nodes according to their centrality will appear (Figure 4); the multi-core structure is then apparent. Playing an important driving role in the process of enterprise spinoff are samples 6, 25, 62, 89, and 122; these five core actors have a high degree of centrality in the network (24, 20, 16, 18, and 18, respectively), and most of Chongming’s ecological–agricultural enterprises are derived from these five actors, which thus make the greatest contribution to the growth in the number of ecological–agricultural enterprises and industrial expansion in Chongming. There are also eight actors, samples 32, 58, 75, 92, 106, 132, 144, and 162 (degree centrality 8, 7, 9, 7, 9, 5, 6, and 6, respectively), which are significantly less central than samples 6, 25, 62, 89, and 122 but higher than the others; these are secondary contributors to the growth in the number of ecological–agricultural enterprises and to industrial expansion in Chongming.

4.1.2. Mechanism Analysis

Ecological–agricultural producers and operators are continuously spun off along with the expansion of industry scale. Organisational innovation activities at the actor level include business expansion to meet market demand, business outsourcing for agricultural production efficiency, enterprise incubation based on new technologies to capture market share, and the ecological transformation of farmers and business integration to guarantee a high standard of service for ecological tourism projects. Combining the research materials with the results of the data analysis, the spinoff characteristics of Chongming’s ecological–agricultural actors can be divided into three main categories based on the relationship and degree centrality values of the spinoff network: large corporate fission spinoff, enterprise spinoff from division of labour, and family spinoff. The spinoff drivers include industry competition, business outsourcing, technology diffusion, social relations, and market demand.

(1)

Large corporate fission spinoffs (enterprise-to-employee spinoffs)

Employees who have worked in Chongming’s leading ecological–agricultural enterprises have left the parent company and created new enterprises, which is the main way in which Chongming’s ecological–agricultural enterprises spin off. The main paths for fission spinoffs are the management and technology paths, and new firms’ fission spinoffs generally have a competitive relationship with the parent firm [16]. For example, sample 6, which spawned a large number of new enterprises, is the first Happy Farm enterprise approved by the Shanghai Municipal People’s Government, founded in 2012. As a representative high-quality, high-tech, high-value-added agriculture company of Chongming, its business model is well adapted to the market, and many employees who fully understand this business model often start successful business activities in their new company after job-hopping. There are also fission spinoffs through the technology path, where new technological inventions are made in the parent company or where employees jump ship after working in high-tech institutions (including research institutions and companies) and use legal technology brought in from the parent company to spin off new companies. Such spinoff paths are common in the development of technology-focused ecological agriculture. For example, samples 62 and 89, which have many spinoffs, are agricultural technology companies that maintain strategic partnerships with famous seed companies at home and abroad, and often have mature product development and promotion teams, a large number of technology patents, and fission spinoffs through technology paths.

(2)

Enterprise spinoff from the division of labour (enterprise-to-enterprise spinoff)

Spinoff from the division of labour in Chongming’s leading ecological–agricultural enterprises is the main reason for Chongming’s current networked industrial organisation model of symbiosis of large, medium, and small enterprises. The expansion of enterprises in the production and operation of ecological–agricultural products is inevitably accompanied by a specialised division of labour within the enterprise. The deepening of this division of labour will, on the one hand, brings about an obvious efficiency gains, and on the other hand, increases the number of levels and links in the management of the enterprise, thus increasing internal management costs [17]. With the expansion of Chongming’s ecological agriculture market and the improvement of the business system, market transaction costs are decreasing, and the flow of industrial information is becoming more convenient. The division of labour between enterprises linked through market transactions has become an important means of improving the productivity of Chongming’s ecological agriculture enterprises and forming an agglomeration of enterprises. An increasing number of relatively independent, specialised divisions of labour enterprises gradually emerge in the process of deepening the industrial division of labour, for example specialising in the production and operation of intermediate or final products, and/or providing few pre-production, mid-production, and post-production services such as information, technology, and sales. Many enterprises with different types of specialised divisions of labour will form a complete industrial chain through market transactions. For example, samples 25, 122, 75, and 32 have spun off a large number of new enterprises. They have systematically led multiple enterprise spinoffs from division of labour in order to expand their production operations in Chongming and capture a larger market share of ecological–agricultural products, effectively improving the efficiency of production and operations; and in this way many of these enterprises have gradually become the leading ecological–agricultural enterprises in Chongming.

(3)

Family spinoffs (enterprise-to-farmer spinoffs)

Family and relatives of the operators of many ecological agriculture businesses in Chongming actively learn from the technical experience of successful business owners, then create new enterprises of their own; this kind of spinoff of enterprise based on kinship relations is an important source of growth in the number of small businesses in Chongming villages. Relational and geographical proximity facilitate the rapid spillover and diffusion of new knowledge and technology, so that local ties and kinship are key paths for the industry to flourish. Kinship is the most basic type of social relationship, and serves as a credit asset that plays a vital role in the productive life of farmers [18]. Based on successful demonstrations of businesses and trust in relatives, family spinoffs are generally fast, allowing an industry to spawn a large number of related enterprises in a short period of time and thus contributing to the rapid development of that industry. For example, four of the six samples 58, 92, 106, 132, 144, and 162 are plantation cooperatives and two are farm tourism enterprises. What they have in common is small size, simple production and operation models and low technological barriers, lack of a clear strategy for expansion, mainly kin employees, and production and operation models very similar to those of their parent enterprises. The interviews also revealed that new enterprise spinoffs from the family would include upstream and downstream or support service businesses, such as semi-finished products or accessories, in addition to similar businesses, to avoid excessive competition among relatives.

4.2. Meso-Level

4.2.1. Evolutionary Features

At the meso level, the evolution of the Chongming EAIS is characterised by a complex and stable innovation network with a central–peripheral structure formed by all types of agroecological actors and the behavioural relationships between them. The data of the 276 research samples were coded with letters + numbers according to the type of actor: A for Farmer, B for Professional Farmer Cooperatives, C for Agricultural Products Processing Enterprises, D for Agricultural Integrated (production–processing–operation) Enterprises, E for Agricultural Products Sales Enterprises, F for Ecological Farm Tourism Enterprises, G for Seedling Factories, H for Agricultural Service Stations (stations for supplying the means of ecological–agricultural production), I for Agricultural Science and Technology Enterprises, J for Agricultural Associations, K for Seed Associations, L for Agricultural Research Institutions, M for Agricultural Universities, N for Agricultural Technology Extension Stations (stations for popularising ecological–agricultural techniques), O for Ecological Agricultural Means of Production Assessment Centre, P for Ecological Agricultural Products Testing Centre, Q for Ecological Agricultural Management Committee; the numbers following the letters distinguish among different individuals who are the same kind of actor. The actual flow of innovation resources among the actors is assigned a value of 1, whereas those without flow of innovation resources are assigned a value of 0. Accordingly, an innovation relationship matrix of 276 × 276 was constructed among Chongming’s ecological agriculture actors. The matrix was imported into UCINET, and the network nodes were processed with an analysis-centrality measure function to obtain the innovation network diagram among ecological–agricultural actors in Chongming (Figure 5). It can be seen that the innovation behaviour of Chongming’s ecological–agricultural actors is frequently linked to each other. The five actor categories L, M, I, J, and K in the figure have a much higher average degree of centrality in the innovation network than other categories. They are the core actors in the innovation network, participate in more innovation activities, and are more connected to other actors. Three types of actors—E, D, and Q—have the lowest average degree of centrality in the innovation network, are marginal actors in the innovation network, and (as this reflects) are generally not involved in innovation activities.

Agricultural innovation does not depend solely on the individual capabilities and actions of agricultural actors, but is clearly more influenced by innovation networks, where agricultural actors are embedded based on their respective motivations to create or exploit different types of information, technological knowledge, and resources by interacting with various types of agricultural innovation actors, with the result that the innovation-intermediary actor with the most such resources in the network is pushed to a central position; these innovation network evolutionary characteristics are consistent with the findings of numerous resent studies [19].

4.2.2. Mechanism Analysis

The networked interconnection of agricultural actors is the result of an active search for specialised associations driven by the respective interests of network actors, a collection of innovative behaviours, and cooperative links between stakeholders. There are four main innovation paths in Chongming EAIS according to the different goal orientations of innovation behaviour.

(1)

Variety and product innovation led by research institutions

Collaborative innovation between scientific research institutions and farmers, professional farmer cooperatives, and agricultural processing companies to transform cutting-edge scientific research into the latest ecological–agricultural products—this kind of innovation activity led by scientific research institutions is the main reason why ecological agriculture in Chongming is leading the seedling and ecological–agricultural product industries. The leading innovators are strong agricultural technology companies, agricultural research institutes, and agricultural universities that cooperate with foreign researchers and have research teams experienced in ecological agriculture. Such innovative activities often promote the upgrading the industrial structure of Chongming’s ecological agriculture and enhancing their market competitiveness with scientific and technological innovations.

(2)

Renewal of agricultural means of production led by associations and service stations

The dissemination of information from industry associations, the distribution of production materials from Agricultural Service Stations, and technical guidance from Agricultural Technology Extension Stations are all part of the process of precisely matching the latest ecological–agricultural innovation resources optimally to each farmer, resulting in an enhancement of the overall ecological standards and level of innovation in all Chongming’s ecological activities. Information exchange and technology diffusion can influence farmers’ decisions and even change knowledge production and talent mobility in the AIS [20]. Agricultural Service Stations and Agricultural Technology Extension Stations set up by local government in almost every administrative village provide each farmer and cooperative with ecologically sound fertiliser and pesticide as well as ecological technology guidance to ensure the spread of the qualified production materials. Simultaneously, Seed Associations and Agricultural Associations, with their own means of information dissemination, have a huge influence on ecological farming groups. The popularisation of technology and products related to ecological–agricultural production and management and the monthly news reports and policy broadcasts on local agriculture in Chongming are highly effective in promoting timely improvements to the ecological–agricultural production and management methods of Chongming’s ecological–agricultural actors.

(3)

Operational model innovation led by farm tourism enterprises

In the process of expanding the scale of their operations and innovating in their business models to improve service quality, most farm tourism enterprises choose to cooperate with small farmers and farmer cooperatives in the vicinity. As market demand continues to change and local infrastructure becomes more sophisticated, ecological farm tourism enterprises are ushering in larger markets and more visitors with higher service requirements. As a result, experiential farm renovation and B&B-style farmhouse renovation jointly carried out by enterprises and local farmers have become commonplace, with an increasing number of farmers innovating and upgrading the traditional agricultural production model to an integrated production, sales, and service model under the leadership of farm tourism enterprises, transforming the single small farmer production model into an ecological industrial business model, bringing higher economic returns to agricultural actors, and promoting the process of upgrading regional ecological security and industrial ecological transformation.

(4)

Innovation in production models led by integrated (production–processing–operation) enterprises

Integrated (production–processing–operation) enterprises and Chongming’s local farmers or cooperatives carry out joint production, unified planning, division of labour, large-scale intensive operations, sharing technology, information, capital, purchase and marketing, storage and transportation, and other pre-production, production, and post-production services. The research interviews revealed that integrated enterprises have a well-developed system in terms of production base, skilled labour, capital, scientific research, and transport capacity, so they have established a solid market and a good brand reputation. Although this model is very independent and has absolute advantages in production, sales, and scientific research and development, it can also cooperate with local farmers or farmer cooperatives in Chongming to carry out joint production to share the risks of research and operations, as well as benefiting from coordination and promotion by the government and industry associations. This innovative model of joint production involves both cooperation in the production process and production factors such as capital and land, changing the disadvantages faced by individual producers in terms of narrow scale, low technology, backward means of production, closed information, and high market entry costs.

4.3. Macro-Level

4.3.1. Evolutionary Features

At the macro level, ecological–agricultural regulatory actors incorporate all other agricultural actors into the ecological regulatory system so that the evolutionary path of the EAIS is regulated in clean production. The data from the 276 research samples were coded according to the type of actor, as in Section 4.2.1, with the actual existence of market behaviour links between actors being assigned a value of 1 and the absence of these links a value of 0. A market behaviour linkage matrix of 276 × 276 between actors in Chongming’s ecological agriculture industry was constructed. The matrix was imported into UCINET, and the nodes were processed as a function of analysis-centrality measures to obtain a network diagram of market relationships between ecological–agricultural actors in Chongming (Figure 6). It can be seen that the market relationship network among actors in ecological agriculture in Chongming has the obvious characteristic of being centred on the regulatory actors; the market relationship covers all the actors, among which the average degree of centrality of three types of actors, Q, P, and O, was much higher than that of other actors, while the average degree of centrality of the five more types of actors (A, B, C, E, and F) was much lower than that of Q, P, and O, but slightly higher than that of other actors. This shows that the frequency of regulatory activities is much higher than the frequency of production and operation activities in all production and operation activities of ecological agriculture in Chongming and that the ecological–agricultural regulatory actors engaged in regulatory innovation activities and the implementation of ecological management regulations have fully integrated ecological control elements into all links of the industrial chain in ecological agriculture activity in Chongming, guaranteeing the ecological path of the evolution of Chongming EAIS. All stakeholders in the EAIS have their own functional position [21], and Chongming’s ecological–agricultural actors can be divided into three categories according to their main functions: ecological–agricultural producers and operators, ecological–agricultural innovation-intermediary actors, and ecological–agricultural regulatory actors (

Table 1. List of the classifications of ecological agriculture actors and the subtypes of each actor group.

Functional Classifications	Contains Specific Actors (Represented by Letters)
Ecological–agricultural producers and operators	Farmer (A), Professional Farmer Cooperatives (B), Agricultural Products Processing Enterprises (C), Agricultural Integrated (production–processing–operation) Enterprises (D), Agricultural Products Sales Enterprises (E), Ecological Farm Tourism Enterprises (F)
Ecological–agricultural innovation-intermediary actors	Seedling Factories (G), Agricultural Service Stations (H), Agricultural Science and Technology Enterprises (I), Agricultural Associations (J), Seed Associations (K), Agricultural Research Institutions (L), Agricultural Universities (M), Agricultural Technology Extension Stations (N)
Ecological–agricultural regulatory actors	Ecological Agricultural Means of Production Assessment Centre(O), Ecological Agricultural Products Testing Centre (P), Ecological Agricultural Management Committee (Q)

). Ecological–agricultural producers and operators are mainly engaged in agricultural production and processing activities, as well as the sale of agricultural products and farm tourism service business activities. They are the largest group of ecological agriculture actors in Chongming and the main objects regulated by ecological agriculture. The service activities of ecological–agricultural innovation-intermediary actors include research and development of agroecology, updating ecological–agricultural means, and dissemination of information. They are not directly involved in the production and operation of agroecological products, but can contribute to the quality and efficiency of the innovative development of agroecology. Ecological–agricultural regulatory actors carry out policy innovation and market regulation. To summarise the evolutionary features of the Chongming EAIS at the macro-regulation level, the innovative activities and production and operation activities of all agricultural actors are within the ecological regulatory system led by the regulatory actors, and regulatory innovation determines the behaviour of agricultural actors and the evolutionary direction of the EAIS.

4.3.2. Mechanism Analysis

By distilling the characteristics of the inter-actor market relationship network in Figure 6 into a more parsimonious flow chart of actors’ market behaviour linkages (Figure 7), it can be seen that the Ecological Agricultural Means of Production Assessment Centre, Ecological Agricultural Products Testing Centre, and the Agricultural Management Committee have incorporated all aspects of Chongming’s ecological–agricultural industry chain into an ecological system of regulatory management and regulation, with regulatory management and regulatory measures driving the evolution of Chongming EAIS along ecologically sound and innovative paths. Specifically, the main role of the Ecological Agricultural Means of Production Assessment Centre is to assess all pesticides, fertilisers, agricultural machinery, and production technologies on the market. Only those that can meet environmental safety standards are allowed to be distributed and used by the Agricultural Service Stations and Agricultural Technology Extension Stations. It is an ecological regulatory actor for the pre- and mid-production stages of ecological–agricultural production in Chongming. By the end of 2020, the Ecological Agricultural Means of Production Assessment Centre screened and published 67 types of green pesticides and fertilisers suitable for the production of green food in Chongming, the Technical Regulations for the Production of Green Agricultural Products in Chongming County, and the Technical Regulations for the Production of Pollution-Free Agricultural Products in Chongming County, which were compiled and updated to include a total of 75 technical regulations for five major categories of agricultural products: grain, vegetables, fruits, livestock, poultry, and aquatic products. The main role of the Ecological Agricultural Products Testing Centre is to test all agricultural products in Chongming; only those that pass the test can enter the sales chain. Excellent ecological–agricultural products are further subject to China’s national geographical indication product certification, as well as organic, green, and pollution-free agricultural products certification, and are the post-production link management agency for Chongming’s ecological–agricultural production. The quality of Chongming’s ecological–agricultural products is guaranteed on the product side, and ‘eco-certification’ is used to enhance economic benefits and achieve sustainable regional practices [22]. By the end of 2020, Chongming had 187 enterprises with 1263 varieties certified as organic, green, and pollution-free, accounting for 90% of Chongming’s total agricultural production area, which also included two protected geographical indication products and six geographical indication certification trademarks. The main role of the Agricultural Management Committee is to constantly update and improve the relevant regulations and legal frameworks according to Chongming’s ecological–agricultural production difficulties and operation problems, forming the most effective regulatory system for ecological agriculture in Chongming. The expansion of agroecology is inevitably accompanied by an increase in the number of actors. Non-ecological production activities and irregular business practices must be excluded by the ecological regulatory system, for example, the use of substandard pesticides or fertilisers, and the sale of substandard agricultural products, etc. Agricultural Management Committees must keep their regulatory systems and management measures up to date in order to ensure no ecologically unsustainable behaviour takes place here. By the end of 2020, Chongming had a total of 533 registered ecological tourist farms and 1303 ecological professional farmer cooperatives, while the Chongming Ecological Agriculture Standardisation Comprehensive Demonstration Standard System covered 1589 standards and 113 laws and regulations.

Compared with the traditional agricultural model, ecological agriculture, because of its high costs, positive externalities, and public goods characteristics, it cannot easily induce stable development of by market mechanisms alone, and must rely on government regulatory instruments for this purpose; these instruments then play a decisive role in the evolution of EAIS [23]. The traditional monolithic management model and lagging macro-regulation arrangements in rural Chinese areas have suppressed the endogenous growth capacity of ecological–agricultural producers and operators. The speculative mentality and short-sighted management behaviour destroy the social-ecological system of the ecological–agricultural economy and the natural ecological balance of agriculture, and the traditional operation concept rooted in the natural economy fails to achieve an inner fit with the new agricultural operators that need to be armed with modern management knowledge and innovative consciousness. Vague development strategies and backward management concepts often go hand-in-hand. The independent behaviour of stakeholders, lack of coordination, and different visions are the biggest obstacles to innovation in agroecology [24], so the regulatory actors of the Chongming EAIS pay great attention to the innovation behaviour, innovation values, innovation functions, and innovation benefits of other innovation actors. Chongming has built a regulatory system for ecological agriculture consisting of quality standards, monitoring and inspection, product development, trademark management, technical guarantee, and market services. Through the government’s macro regulatory guidance and specific project funding support, the market mechanism is the main means of allocating resources for science and technology innovation. The construction of an innovation network closely integrated with Chongming’s ecological agriculture industry and the efficient coordination and integration of various innovation resources can promote the rapid and effective transformation of high-tech achievements into value added along the science and technology value-chain network system [25].

5. Conclusions

Based on a review of the literature related to the AIS and agricultural innovation, this study constructed a theoretical framework for the evolution of EAIS and analysed the characteristics and mechanisms of EAIS evolution at the micro, meso, and macro levels, taking the ecological agriculture of Chongming Eco-Island in Shanghai, China as an example to provide a theoretical basis for the construction and management of EAIS in developing countries. The main findings of this study are as follows.

First, the evolutionary structure of EAIS includes three levels: the micro level of actors, the meso level of the innovation network, and the macro level of institutions. The evolutionary content of the three levels together forms the evolutionary content of EAIS. Therefore, the evolutionary process of the EAIS is the evolution of innovation and ecological evolution of each ecological–agricultural producer and operator under the regulation of the ecological–agricultural regulatory actors and the innovation-intermediary effect of the ecological–agricultural innovation-intermediary actors, including the organisational spinoff activities at the micro level, the innovation network evolution process at the meso level, and the regulatory innovation activities at the macro level.

Second, the innovation content of the EAIS includes organisational spinoffs, innovation networks, and regulatory innovation. EAIS meets the huge market demand and keeps up with the accelerating pace of technological innovation through organisational spinoffs, and helps agricultural actors to integrate quickly into ecological regulations; adapts to regulations and market changes through innovation networks; efficiently obtains resources such as knowledge overflow, information dissemination, and technological cooperation; and establishes a set of scientific regulatory systems and operational norms to guarantee an innovative and ecologically sound development path for ecological agriculture in Chongming.

Third, the evolutionary mechanism of EAIS is the diffusion and effectiveness of innovation resources and ecological regulation in the inter-actor network, including innovation-intermediary assistance and the ecological regulation of the actor as well as the resource synergy of the innovation network. At the micro level, ecological–agricultural producers and operators are continuously spun off with the expansion of industry scale under market orientation, technology promotion, and regulation. At the meso level, all actors actively seek specialised associations to obtain innovation resources, forming an innovation network. At the macro level, ecological–agricultural regulatory actors incorporate all actors into the ecological regulation system so that the evolutionary path of the EAIS is regulated in cleaner production.

6. Discussion

This study makes a new breakthrough in its analysis of the structure and mechanism of the EAIS evolution, and the theoretical framework constructed can be valuable for guiding the ecologically sound and innovative development of agriculture in developing countries, providing a theoretical basis for government policymakers to construct or manage the EAIS. In terms of theoretical framework construction, innovation system framework construction, innovation actors, innovation networks, and innovation dimensions were the themes of this study. Many scholars have proposed different AIS research frameworks to explore the structural levels and composition of the AIS [26,27], explore its deep-rooted principles and operational mechanisms, and apply case studies to explore the specific behaviours of various sectors. With regard to influencing factors, the government sector plays a decisive role in the direction of the AIS evolution, but sometimes the private sector and intermediary organisations can have an even more important impact on the AIS [28,29,30], and further analysis of non-regulatory factors should be expanded in the future. At the same time, coordination of the claims of various groups and stakeholders in the AIS has also been the focus of scholarly research, with an exploration of how power dynamics among stakeholders affect participatory agricultural innovation initiatives, attempting to find solutions in technology, markets, regulation, and other practices and frameworks [31,32,33,34]. How innovation actors achieve collaborative innovation through innovation networks is key to the sustainable operation of the EAIS, while the creation and adoption of knowledge within and across actors, path dependence and path disruption of technology, and the influence of external market choices and government promotion are the intrinsic mechanisms by which they can evolve collaboratively. Finally, although this study supports the heterogeneous characteristics of the AIS [35,36], because the formation and evolution of the AIS cannot be separated from the influence of the regional socio-cultural environment, different regional contexts, resource endowments, and technological capabilities will affect the evolutionary patterns of the AIS, and thus it is necessary to research the evolution of the EAIS in the economic, demographic, and institutional environments of various regions in the future to enhance the EAIS approach’s universality and expand its theoretical system and scope. However, to do so, the applicability of the AIS methods to specific regions and industries also still needs to be improved, and exploring the regional rootedness of the EAIS will thus be an important aspect of future research.