Introduction

The issue of global climate change is increasingly receiving attention from countries around the world. However, as the main source of greenhouse gas emissions, agriculture plays a crucial role in the process of “dual carbon” (Wolfert and Isakhanyan, 2022). As the acceleration of industrialization and urbanization, the ecological development of agriculture in developing countries face with enormous challenges under the pressure of many factors such as population, resources, and environment. Under the traditional agricultural development model, the pollution of pesticide, fertilizer, livestock breeding waste emissions has exacerbated, which has seriously damaged the surrounding environment and reduced the diversity of agricultural biology, directly threatening human health. Taking China’s experience as an example, the government has successively issued multiple policy documents, striving to promote green agricultural development. Agricultural ecological transformation is an inevitable choice for the development of modern agriculture in China. Ensuring the sustainable development of agricultural ecology not only relates to the process of urbanization, but also to the modernization development of the country. During the period of China’s continuous promotion of rural revitalization strategy, the in-depth analysis of the development level of agricultural ecological development and its obstacles is of great guiding significance for formulating policies and specific measures related to regional rural revitalization, guiding the development of ecological agriculture, and providing important theoretical basis for the sustainable development of agriculture.

In 2022, the scale of China’s digital economy reached 50.2 trillion-yuan, accounting for 41.5% of GDP and provided a crucial support to economic growth. The digital economy is a series of economic activities based on digital technology, with digital platforms as the main medium and digital-enabled infrastructure as an important support (Ma and Zhu, 2022; Pan et al. 2022). Through the overlay, diffusion, and penetration of digital infrastructure, data elements and digital technologies, the digital economy generates superposition, diffusion, and penetration effects on traditional industries, thereby enhancing the resource allocation efficiency and reducing production costs in these industries. The integration and development of digital economy with agriculture and rural economy can effectively alleviate the information asymmetry in traditional agriculture, allowing for the optimal allocation of resources to achieve effects of economic scale and reduce transaction costs (Bansal et al., 2022). As a major agricultural producer and carbon-emitting country, China’s approach towards meeting carbon peak and neutrality targets while embarking on a path of green development with Chinese characteristics is a matter of concern for all sectors of society. Digital economy is a new driving force and engine for the transformation of traditional agriculture into modern agriculture, while the integration and development of digital economy and agricultural ecological transformation not only improves the economic development level of rural areas but also promotes sustainable agricultural development (Şerbu, 2014). Digital economy has become an important focus for promoting green development in agriculture, and is also an urgent requirement for China’s rural areas to cope with climate change in agricultural affairs (Ma and Zhu, 2022). Against this background, exploring the relationship between digital economy and agricultural ecological transformation has become a topic of academic concern in academia.

Under the call for policies to achieve peak carbon dioxide emissions and carbon neutrality, low-carbon technological innovation is the essential path to drive sustainable development of the Chinese economy. Low-carbon technological innovation is the integration and implementation of the two major development concepts of innovation and green. It is based on technological innovation and fully considers green and low-carbon characteristics, while balancing economic development and environmental protection. It is an important approach to achieve agricultural ecological transformation. Haque and Rashid (2023) propose that low-carbon technological innovation is an important factor in promoting green transformation. Through technology diffusion and spatial spillover effects, low-carbon technological innovation can significantly reduce particulate matter and greenhouse gas emissions, thereby improving air quality and promoting ecological transformation. Wang et al. (2022) argue that the sustainable development of agriculture is facing dual pressures of changes in factor endowment structure and resource environment constraints. They selected provincial-level panel data of China’s agricultural industry from 1997 to 2017, combined with DEA-SBM model, Malmquist-Luenberger index decomposition method, and panel data regression analysis method, to empirically explore the changes and internal laws of green technology bias in China’s agricultural industry. Research has found that green technology has a significant enhancing effect on green total factor productivity in China’s agricultural industry. Given the complex and ever-changing international climate situation and the severe environmental protection situation in China, this article will focus on the study of three aspects of issues: Does the digital economy play a role in promoting the agricultural ecological transformation? What is the underlying mechanism of low-carbon technological innovation in improving the agricultural ecological transformation through the digital economy? Is there a difference in the effect of the digital economy on the agricultural ecological transformation at different levels of low-carbon technological innovation?

Accordingly, digital technologies such as big data and artificial intelligence are continuously changing the structure of agricultural industries and promoting agricultural ecological transformation in China. The penetration of low-carbon technological innovation in the agricultural field may also affect agricultural ecological transformation through certain mechanisms. Our research contributions are as follows: First, we have constructed an evaluation index system and measurement framework for agricultural ecological transformation, revealing the real dilemma and matching development pattern from “Input Factor”, “Expected Output”, “Undesirable Output”. Secondly, we incorporated digital economy as an important driving force into the framework of the impact mechanism on agricultural ecological transformation, further considering the differential threshold effect of low-carbon technological innovation in different regions, and answering the question of how to effectively utilize low-carbon technological innovation to enhance agricultural ecological transformation empowered by digital transformation. Based on the nonlinear dynamic panel threshold model, this paper innovatively verifies the heterogeneous “threshold” characteristics caused by regional low-carbon technological innovation differences on the enhancement effect of digital economy, providing new insights for the implementation path of low-carbon agriculture development.

The remaining parts are as follows: Part 2 provides a literature review, Part 3 is theoretical hypothesis analysis, Part 4 estimates agricultural ecological transformation, Part 5 constructs a dynamic threshold model, Part 6 introduces and discusses the regression results, Part 6 is conclusions, and puts forward policy recommendations.

Literature review

The digital economy is not only an important carrier in promoting the revitalization of rural areas in the new era, but also a critical driving force and guarantee in achieving the transformation towards agricultural colonialization and the dual-carbon targets. Digital economy refers to a series of economic activities that are based on digital technologies, use digital platforms as the primary medium, and rely on digital-enabled infrastructure as an important support (Tiwasing et al., 2022). As a new engine for reshaping the development mode of agriculture, the digital economy has significant practical implications for the sustainable development of green agriculture. The integration of digital economy and agriculture can promote green production in agriculture, reduce negative impacts on the environment, and achieve green transformation and upgrading in agriculture. It is a new direction for the modernization of agriculture and rural areas. This will propel China’s agricultural ecological transformation to a new level. Against this background, exploring the relationship between digital economy and agricultural ecological transformation has become a hot topic of concern in the academic community. Currently, the academic community has relatively consistent views on the following aspects: Firstly, the digital economy has become a new productive force for improving agricultural economic benefits. Utilizing digital technology, the digital economy has changed the way in which agricultural economic growth occurs, bringing about a transformation in agricultural production and management modes, improving agricultural operational efficiency, and promoting the economic growth and high-quality development of agriculture. Secondly, the digital economy has achieved a coordinated balance between environmental and economic benefits. By using general information technology, the necessary water and fertilizer demand for agricultural production can be scientifically calculated, thus reducing unnecessary resource depletion and carbon emissions in agriculture. With the optimization and integration of agricultural production factors using digital technology, the digital economy has achieved creative subversion and innovation of traditional agriculture, thereby constructing an efficient and green modern ecological agriculture (Attour and Barbaroux, 2021). Thirdly, digital economy plays a strong empowering role in agricultural ecological transformation through the spillover effect of technology, cultivating and forming a modern agricultural digital economy industry chain, and turning to new ways of developing ecological agriculture. Digital economy can optimize the structure of agricultural industry, promote technological innovation and application of green agricultural production, improve the output rate of agricultural land and resource utilization, thus promoting the coordinated development of regional agricultural digital economy and ecologicalization, and realizing the transformation and upgrading of agricultural ecological environment (Pakseresht et al., 2022).

As the foundation of the national economy, agricultural ecologicalization is an important component of implementing the concept of green development in the economic field. However, agricultural production in developing countries still faces problems such as low level of informatization and limited technological content. It is urgently necessary for the government to vigorously promote technological innovation, actively construct a green, low-carbon circular development agricultural production system, and comprehensively improve the level of traditional agriculture greening. Therefore, in the process of improving agricultural production efficiency and promoting green agricultural development, the role of low-carbon technological innovation cannot be neglected. Some scholars have explored the impact of low-carbon technological innovation on the agricultural ecological transformation from the perspective of enterprises in the digital economy era. Bhandari et al. (2023) found that the application of green nanotechnology plays an important role in agricultural ecosystems. Biogenic nanoparticles obtained from plants, bacteria, fungi, or their metabolites are a sustainable green technology that has a positive impact on crop production, soil health management, and ecological environment. It is beneficial for meeting the world’s food demands and achieving the goal of sustainable agriculture development. Costa et al. (2023) conducted a qualitative research through exploratory single case study, analyzing different business departments of a multinational corporation. They demonstrated that, for developing countries with agriculture as their economic center, digital technology promotes sustainable development of the agriculture industry by optimizing operations and resource utilization. Chen et al. (2023) conducted a study based on panel data from 30 provinces and cities in China between 2013 and 2019. They used the NDDF-LHM model to estimate the forestry green total factor productivity (TFP) and verified that the digital economy can significantly promote the improvement of forestry green TFP. The study also found that green technology innovation is an important mechanism for the digital economy to improve forestry green TFP. However, He et al. (2021) adopted a random effects panel Tobit model to explore the role of green production technology in improving agricultural low-carbon efficiency. The results indicated that the impact of green technology on low-carbon efficiency is heterogeneous, and not all levels of green technology can effectively improve low-carbon efficiency. Similarly, Berkhout and Hertin (2004) argue that the diffusion and use of information and communication technologies are leading to both positive and negative environmental impacts, which are difficult to trace and measure. This also provides a reference for the study of the relationship between low-carbon technological innovation and agricultural ecological transformation.

In summary, the digital economy is of great significance for promoting agricultural ecological transformation and achieving high-quality agricultural development. Previous research has provided theoretical support for the digital economy to empower agricultural ecological transformation, but there are still breakthroughs and improvements to be made. Compared with previous research, the contribution of this article lies in:

Firstly, at the research level, most existing studies have focused on the impact of the digital economy on the green transformation of industry or manufacturing, but have overlooked the challenges faced by agriculture in terms of sustainable development and ecological transformation under the current conditions in China. There is a lack of attention to agricultural ecological transformation under the dual pressure of resources and environment. This article will focus on the agricultural field and explore in depth the impact and intrinsic mechanism of the digital economy on the agricultural ecological transformation, expanding the empirical research on the digital economy in the agricultural field, with the aim of providing theoretical and practical support for achieving green transformation and sustainable development in agriculture.

Secondly, in terms of research methods, most existing studies use traditional econometric models based on economic principles to describe the relationship between digital economy and agricultural ecological transformation. However, this method often fails to reflect the endogenous relationship between variables. In order to avoid or reduce errors in existing econometric methods, this paper adopts an improved dynamic threshold regression method to construct a dynamic threshold model that examines the impact of digital economy on agricultural ecological transformation, considering the endogeneity and dynamic changes of the model. The results are more robust, providing powerful theoretical support and reliable reference for exploring the path selection and policy design of agricultural ecological transformation.

Thirdly, in the expansion of research, this article further constructs a threshold model for the digital economy to drive the agricultural ecological transformation. The low-carbon technological innovation perspective is included in the research framework of the impact mechanism of the digital economy on the agricultural ecological transformation, revealing the nonlinear impact relationship of the digital economy on the agricultural ecological transformation at different levels of low-carbon technological innovation. Then, combined with the differences in the development levels of digital agriculture and regional economies, the heterogeneous mechanism of the interaction between the digital economy and the agricultural ecological transformation is revealed.

Accordingly, we explore a new source of driving force and identifies the heterogeneous driving path of agricultural ecological transformation, achieving a balance between digitization and green development under low-carbon technological innovation, and providing new insights for achieving “carbon reduction and economic promotion” in developing countries.

Theoretical assumptions

Digitalization and green development

The relationship between digitization and green development has always been a focus of international attention. Strengthening digitization construction is an important source of driving green development. The application of digital information technology can break the constraints of time and space, promote the flow of production factors resources, and achieve reasonable docking and matching, thereby improving resource allocation efficiency and productivity, driving green economic development (Popkova et al., 2023). Moreover, the development of digitalization also needs to be controlled within a certain range and either too high or too low may be conducive to achieving the maximum benefits of digitalization. The function of digitalization (DE) and green development level (Y) is set as: Y= ƒ(DE), Then take the derivative of DE, ƒ΄(DE) = 0, ƒ˝(DE) < 0, DE* = T. When DE∈(0, T), ƒ’(DE) > 0, digitalization is positively correlated with green development level; When DE∈(T, +∞), ƒ’(DE) < 0, there is a negative correlation between digitization and green development level; When the digitization level is at T value, the green development level reaches its maximum. In summary, this paper proposes the following hypotheses:

Hypothesis I: Digitalization has a certain impact on green development, and this mechanism of action may have a non-linear relationship.

Digital economy, low-carbon technological innovation, and agricultural ecological transformation

The digital economy empowers high-quality agricultural development through low-carbon technological innovation, accelerates the dissemination and penetration of technology and information elements in the agricultural industry, benefits the reshaping of traditional agricultural production methods, and promotes the transformation of agricultural ecology. At different levels of low-carbon technological innovation, the impact of the digital economy on the ecological transition of agriculture may exhibit certain heterogeneity. For regions with high levels of low-carbon technological innovation, green technology innovation helps farmers achieve technological and clean transformations, actively promotes the upgrading of agricultural industrial structure and optimization of energy structure, thereby improving green all-factor productivity. With the assistance of low-carbon and green technology, the digital economy enhances the efficiency of resource utilization by adjusting the input-output ratio accordingly, injecting new vitality to agricultural production and consumption, and promoting the green development of the agricultural economy (Goel et al., 2021). However, in regions where the level of low-carbon technological innovation is relatively low, the enabling role of the digital economy is limited due to insufficient technological level, which cannot help the digital economy to exert its maximum driving effect, and the transformation of agricultural colonialization is slow (Shen et al., 2022). Based on the logical considerations of change theory, under the condition of low innovation level of low-carbon technology, it is difficult to effectively control the decarbonization cost in the region, provide more effective technical support for the research and development of carbon dioxide utilization, capture and storage technology, and large-scale application, and cannot reduce the regional carbon emission level in a short period of time (Habiba et al., 2022). Additionally, when the green innovation capability is low, local governments will invest some resources into technological research and development. However, in general, technological innovation is difficult and long-lasting, with a large amount of funding being consumed for a long period of time, while the technological output is slow and limited. The diversion of resources prevents the government from focusing all its efforts on agricultural ecological transformation and hinders the modernization process of agriculture. Meanwhile, it is difficult for lower levels of green technology innovation to coordinate the integrated development of green technology among enterprises, universities, and research institutes in digital development, and to attract more high-quality human capital. In addition, the immaturity of technological innovation development and incomplete government regulation make it difficult for innovation entities to achieve data sharing, data openness, and data circulation, resulting in deficiencies in the construction of digital infrastructure (Du et al., 2019). Specifically, in the context of the heterogeneity of low-carbon technological innovation in China, when low-carbon technological innovation is at a low level, the digital economy cannot effectively drive the agricultural ecological transformation. With the improvement of the level of low-carbon technological innovation and its surpassing of the critical threshold, the digital economy begins to positively promote the agricultural ecological transformation. In summary, this paper proposes the following hypotheses:

Hypothesis 2: The impact of the digital economy on the transformation of agricultural ecologicalization is constrained by the threshold effect of low-carbon and green innovation levels.

Finally, we construct a schematic diagram of the mechanism of digital economy on agricultural ecological transformation (Fig. 1):

Fig. 1: Process flow and theory of change.
figure 1

This illustrates the process flow of how digital economy affects agricultural ecological transformation.

Full size image

Agricultural ecological transformation

Super-SBM Model Setting

Based on the literature study of existing documents (Boix-Fayos and de Vente, 2023; Barbier, 2022), we believe that the transformation of agricultural ecologicalization can be summarized into two aspects: on the one hand, in terms of agricultural growth patterns, it emphasizes the process of transforming from extensive agriculture to intensive agriculture. On the other hand, in terms of agricultural pollution control, it emphasizes the process of gradually reducing energy consumption to achieve energy-saving and emissions reduction by reducing high-carbon pollution methods. Therefore, on the basis of the definition of agricultural ecological transformation, the measurement index is calculated using the contribution rate of agricultural green total factor productivity to agricultural economic growth. Among them, the agricultural green total factor productivity can be evaluated using the Super SBM model of unexpected output (Tone, 2001). The Super-SBM model has obvious advantages in effectively dealing with the laxity problem of input-output variables and the fitting problem of unexpected output (Tone, 2002).

Based on the features of the Super-SBM model, this paper constructs a measurement index system for the overall green productivity of agriculture, as shown in Table 1. In terms of input factors, the ratio of the number of employees in the primary industry and the total agricultural output to the total output of agriculture, forestry, animal husbandry, and fishery is selected as the labor input indicator; the planting area of crops is selected as the land input indicator; the effective irrigation area is selected as the irrigation input indicator; the amount of fertilizer used in agriculture is selected as the fertilizer input indicator; the amount of pesticides used is selected as the pesticide input indicator; the amount of agricultural film used is selected as the film input indicator; the amount of diesel fuel used in agriculture is selected as the energy input indicator; and the total power of agricultural machinery is selected as the mechanical input indicator. In terms of output, the agricultural total output value is selected as the expected output indicator, while the agricultural non-point source pollution comprehensive index is chosen as the unexpected output.

Table 1 Evaluation indicator system of agricultural ecological transformation.
Full size table

Calculation results

Based on the constructed the ecological transformation of agriculture indicator system, the Super-SBM method was used to evaluate the level of agricultural ecological transformation in 30 provinces of China from 2013 to 2021. The results are shown in Table 2 and Fig. 2. It can be seen that the overall average level of agricultural ecological transformation China is 0.607 (as shown by the dashed line in Fig. 2), and the general level of agricultural ecological transformation is relatively low. However, it is showing a stable upward trend. From the growth rate perspective, the average annual growth rate from 2013 to 2021 was 6.99%, indicating a relatively fast growth rate. The breadth and depth of regional rural digitization continue to extend mainly due to the continuous expansion of the application scope of digital technology. Although China gradually realizes the importance of agricultural ecological transformation, there is still a gap between China and the advanced international level due to the significant differences in ecologicalization in different regions. Specifically, regarding segmented provinces and regions, Guangdong Province has the highest level of digital rural development, with an average value of 0.866. As a pioneer of China’s agricultural ecological transformation, Guangdong province has vigorously developed a distributed new energy system in rural areas in recent years, improved ecological carbon trading mechanisms, promoted the development of rural carbon economies, accelerated the development of efficient ecological cycle agriculture, and promoted the transformation of agriculture towards ecological sustainability. The province with the lowest level of digital rural development is Jilin province, which reflects the significant gap in the development level of Agricultural ecological transformation in China. With the gradual increase in the construction cost of agricultural ecological transformation in China, coupled with the relatively slow development of rural areas in the region, agriculture ecological transformation is highly challenging. In summary, the path of developing ecological agriculture faces great challenges and requires substantial efforts towards improvement.

Table 2 China’s agricultural ecological transformation (2013–2021).
Full size table
Fig. 2: Average level of China’s agricultural ecological transformation (AEE) (2013–2021).
figure 2

This is an average map of China’s agricultural ecological transformation.

Full size image

Figure 2 is an average map of China’s agricultural ecological transformation. From the figure, it can be seen that the average level of agricultural ecological transformation in China from 2013 to 2021 is 0.607 (shown by the dotted line in Fig. 2). Half of the regions have exceeded the average level, with Guangdong, Shaanxi, Henan and other provinces leading the way, while Qinghai, Shanxi, Jilin and other provinces have a relatively low level of transformation and significant regional differences. Specifically, the average level of agricultural ecological transformation in Guangdong (0.866) is 2.71 times higher than that of Jilin (0.320). How to enhance the level of agricultural ecological transformation in provinces such as Qinghai, Shanxi, and Jilin are a key issue that the government should pay attention to. The provinces of Guangdong, Shaanxi, Henan, and others, where the level of agricultural ecological transformation is relatively high, are mostly located in the central and eastern regions, occupying a clear geographical advantage. The local economic development level is good and the infrastructure construction is relatively complete. There is an abundance of outstanding talent gathered in the area, including many high-tech agricultural skilled personnel who participate in rural and agricultural development. Take Guangdong as an example. Guangdong province has consistently adhered to the principles of quality-driven and green-driven agricultural development. It has adopted a green concept to lead and enhance the growth of its dominant industries, while continuously improving the level of agricultural production through the application of green technology. Guangdong has launched and implemented the largest agricultural non-point source pollution control project in Asia, which is also China’s first project funded by the World Bank. Guangdong has launched and implemented the largest agricultural non-point source pollution control project in Asia, which is also China’s first project funded by the World Bank. The usage of fertilizers and pesticides has seen negative growth for five consecutive years, while the comprehensive utilization rates of livestock and poultry manure and straw have reached 88.4 and 91.1%, respectively. These measures have effectively transformed the development pattern of agriculture, which was previously over-reliant on resource consumption. The pilot projects on green breeding and farming, in which 13 project counties and 194 implementing entities have actively participated, have pushed forward the industrialization of green breeding and farming technology. Shaanxi actively promotes the application of standardized ecological agriculture technology, practices green low-carbon circular production, builds a green ecological brand, advances ecological agriculture construction, and accelerates the development of agricultural green transformation. These cases demonstrate that some central and eastern regions of China, such as Guangdong and Shaanxi are continuously increasing the agricultural ecological transformation, promoting the green development of agriculture and rural areas.

Methodology

Dynamic panel threshold model

Due to the significant differences in the level of low-carbon technological innovation across different regions in China, neglecting the heterogeneity factors of low-carbon technological innovation would affect the accuracy of model results. Hansen (2006) proposed a static panel threshold regression model that estimates the true threshold through a threshold variable. However, this threshold method fails to reflect the dynamic changes or lag effects of sample objects, as well as ignores the treatment of endogenous variables. In order to address this deficiency, this paper adopts an improved dynamic panel threshold regression model. Based on Hansen’s idea, the threshold estimation values are calculated and further dynamic estimation of the partition intervals is conducted using system GMM dynamic methods. By combining Hansen’s ideas and the System GMM dynamic modeling approach, the endogeneity and dynamic characteristics of the model are comprehensively considered. By combining Hansen’s ideas and the System GMM dynamic modeling approach, the endogeneity and dynamic characteristics of the model are comprehensively considered.

In this paper, agricultural ecological transformation (AEE), digital economy (DE), Low-carbon technological innovation (LCI) are explanatory variable, core explanatory variable, and threshold variable, respectively. Control variables include Disasters rate (DR), urbanization rate (URB), agricultural internal structure (AIS), and rural human capital (HC). The paper aims to examine the impact of digital economy on agricultural ecological transformation at different levels of low-carbon technological innovation. Based on this, a panel threshold model (with a single threshold) is constructed as follows:

$$\begin{array}{l}AEE_{it} = \theta + a_1L_1 + a_2L_2 + a_3DR_{it} + a_4URB_{it} + a_5AIS_{it} + a_6HC_{it}\\ \quad \quad \quad \beta _1DE_{it}I\left( {LCI_{it} \le \gamma } \right) + \beta _2DE_{it}I\left( {LCI_{it}\, >\, \gamma } \right) + \mu _i + \nu _t + \varepsilon _{it}\end{array}$$
(1)

Dynamic panel multi-threshold model (double threshold as an example) is:

$$\begin{array}{l}AEE_{it} = \theta + a_1L_1 + a_2L_2 + a_3DR_{it} + a_4URB_{it} + a_5AIS_{it} + a_6HC_{it} + \beta _1DE_{it}I\left( {LCI_{it} \le \gamma _1} \right)\\ \qquad \quad \quad\;\;\, +\, \beta _2DE_{it}I\left( {\gamma _1\, <\, LCI_{it} \le \gamma _2} \right) + \beta _3DE_{it}I\left( {LCI_{it}\, >\, \gamma _2} \right) + \mu _i + \nu _t + \varepsilon _{it}\end{array}$$
(2)

In the model, L1、L2 are lagging items, I(·) represents the indicator function, γ is the threshold value of the variable, μi is the individual-specific effects, vt is the time-specific effects, and εit is the random interference terms.

Variable definitions

Explained variable: agricultural ecological transformation (AEE), we use the results of agricultural ecological transformation as measured above.

Core explanatory variable: digital economy (DE). The digital economy possesses unique technological advantages that enable efficient flow and allocation of resources, rapid sharing and dissemination of knowledge and information, and facilitate the renewal and iteration of ecological innovation. It promotes the agricultural ecological transformation. Building on existing research, this paper constructs an evaluation index system from the characteristics of the digital economy, fully considering the hierarchy of measurement indicators and the availability of data the system includes three subsystems, namely, the construction of digital infrastructure, the development of digital industrialization, and the digitalization of industries, as well as nine criterion levels and a total of 14 measurement indicators. The results are shown in Table 3. According to the specific system in Table 3, we use the entropy method to calculate the development results of the digital economy in various regions. The entropy method is an objective weighting method that is not affected by subjective factors, and can retain all information of the data without special requirements for data distribution. The specific steps are as follows: first, standardize the process, second, calculate the entropy value of the indicators, third, calculate the weights of the indicators, and finally obtain the comprehensive results. Limited to space, the specific calculation process is used as supplementary information for readers.

Table 3 Digital economy measurement system.
Full size table

Threshold variable: Low-carbon technological innovation (LCI). Low-carbon technological innovation is the integration and implementation of the two major development concepts of innovation and green. It is based on technological innovation and fully considers green and low-carbon characteristics, while balancing economic development and environmental protection. It is an important approach to achieve agricultural ecological transformation (Hou et al., 2023). Low-carbon technological innovation emphasizes the environmental and sustainable development attributes of technology, and pursues the maximum utilization of limited resources. Under the guidance of new development concepts, low-carbon technological innovation provides an efficient and scientific driving path for the regional transformation of agricultural ecologicalization. Patents are important indicators for measuring regional technological innovation capacity and research and development levels. However, Green patents, as direct physical evidence supporting innovation-driven sustainable development, can be used to represent the output of green technological innovation. Based on the CNRDS database, this paper selects the number of green patent applications in each region to reflect the level of green technology innovation in that region (Chen et al. 2022; Hou et al. 2023).

Control variables:

  1. (1)

    Urbanization rate (URB). Urbanization is an important influencing factor in the transformation of agricultural ecology. With the migration of the population from rural to urban areas, the living standards of residents have been constantly improving, and there is also a higher demand for agricultural products, which has driven agricultural production to continuously develop towards green and ecological development. We use the proportion of urban population to total population in each province to calculate the urban population proportion (Zhao et al., 2022).

  2. (2)

    Disaster Rate (DR). Compared to other industries, agriculture has a greater dependence on natural conditions and relatively weaker resilience to disasters. Especially in the context of global warming, extreme weather events occur frequently, causing serious impacts on agricultural output. However, agriculture, as the foundation of the country, in order to ensure sufficient production even in circumstances of disaster, the government will continually increase capital and resource inputs to promote sustainable agricultural ecological transformation. We use the ratio of the agricultural disaster area to the crop sowing area to represent the disaster rate (Guan et al., 2021).

  3. (3)

    Agricultural internal structure (AIS). The adjustment of the internal structure of agriculture affects agricultural ecological transformation. Different crops have different growth characteristics and demand for fertilizers and other agrochemicals. Therefore, with different crops, agrochemicals input land changing, agricultural ecological transformation also change. We use the ratio of the sown area of cereal crops to the total sown area of crops to represent the internal structure of agriculture (Huang et al., 2022).

  4. (4)

    Rural human capital (HC). By optimizing the allocation of production factors, rural human capital can assist in promoting clean production and promoting economic growth, which has a certain impact on the transformation of agricultural ecologicalization. We use the rural residents’ per capita spending on education and entertainment to represent rural human capital (Luo et al., 2023).

Data sources

This paper uses panel data of 30 provincial regions in China from 2011 to 2021 (Tibet, Hong Kong, Macau and Taiwan are not included in the research samples due to missing data). The original data are obtained from the National Bureau of Statistics of China, the China Stock Market & Accounting Research Database database (CSMAR), and the Chinese Research Data Services database (CNRDS). This paper uses STATA13.0 to perform the data processing and empirical analysis. Table 4 presents a summary of the descriptive statistics of variables:

Table 4 Descriptive statistics of the variables.
Full size table

The threshold effects of digital economy on agricultural ecological transformation

Estimation results of the dynamic threshold effect

This paper starts with the heterogeneous threshold of low-carbon technological innovation, using a dynamic panel threshold model, focusing on the impact of digital economy on agricultural ecological transformation. To begin with, the results in Table 5 indicate that low-carbon technological innovation has not passed the triple threshold test. However, both single and double thresholds passed the test at a significant level of 5%. According to Hansen’s threshold model, the relationship between digital economy and agricultural ecological transformation has a significant double threshold in terms of low-carbon technological innovation. Based on the results, this paper uses the double threshold model of low-carbon technological innovation to estimate the impact mechanism of digital economy on agricultural ecological transformation.

Table 5 Threshold effect test.
Full size table

Table 6 presents the estimates of the double threshold: Among them, the estimated values for the double threshold are 4.283 and 4.642 respectively, each located within the 95% confidence intervals of [4.234, 4.564] and [2.773, 6.763]. Therefore, we divide the sample into three different regimes as follows: weakly low-carbon technological innovation (LCI ≤ 4.283), moderately low-carbon technological innovation (4.283 < LCI ≤ 4.642), and strongly low-carbon technological innovation (LCI > 4.642). Besides, Fig. 3 represents the estimated threshold values and confidence intervals corresponding to low-carbon technological innovation. It can be observed that the effects of the digital economy on agricultural ecological transformation have a significant double threshold for low-carbon technological innovation.

Table 6 Results of threshold estimator.
Full size table
Fig. 3: Threshold confidence interval of low-carbon technological innovation (LCI).
figure 3

This is the identification of the first threshold value and second threshold value.

Full size image

Furthermore, this paper divides low-carbon technological innovation into different intervals according to the threshold values and discusses the threshold effect and different impacts of digital economy on agricultural ecological transformation under different degrees of low-carbon technological innovation. Table 7 presents the driving mechanisms of the digital economy for agricultural ecological transformation under different levels of low-carbon technological innovation.

Table 7 Results of dynamic threshold regression.
Full size table

Table 7 indicates that when low-carbon technological innovation is relatively weak (LCI ≤ 4.283), the digital economy has a significant negative effect on agricultural ecological transformation. When low-carbon technological innovation is at a moderate level (4.283 < LCI ≤ 4.642), the digital economy is beginning to have a significant and positive effect on agricultural ecological transformation. With the continuous improvement of the level of low-carbon technological innovation (LCI > 4.642), the digital economy has shown a significant promoting effect on agricultural ecological transformation. The above results suggest that the effects of digital economy on agricultural ecological transformation exist a significant threshold for low-carbon technological innovation. In general, digital economy has an inhibitory effect on agricultural ecological transformation when low-carbon technological innovation is at a relatively low threshold. When low-carbon technological innovation increases and surpasses the threshold value, digital economy stimulates low-carbon technological innovation to some extent. Therefore, hypothesis I and Hypothesis II are validated.

Regarding other influencing factors that affect agricultural ecological transformation, the proportion of urban population and the incidence of disasters show a significant positive correlation with agricultural ecological transformation. The reason is that a large number of people are moving from rural to urban areas, and the increase in the urban population size leads to an increase in demand for high-quality agricultural products, which in turn drives the continuous development towards green and ecological agriculture (Zou and Deng, 2022). However, agricultural disaster rates can have an impact on agricultural output. Agriculture is the foundation of a country, and in order to effectively respond to possible natural disasters, the government will continuously increase resource input and continue to promote the agricultural ecological transformation (Guan et al., 2021). The internal structure of agriculture is positively correlated with the agricultural ecological transformation, but the results are not significant. With the continuous optimization and adjustment of the internal structure of agriculture, the development between different departments becomes more balanced, thus having a positive impact on the agricultural ecological transformation. Rural human capital has a certain inhibitory effect on the agricultural ecological transformation. Based on the current situation, this may be because with the popularization of education, the labor force in rural areas has received good quality education and is not willing to engage in agricultural production. The flow of human capital from the agricultural industry to other industries is not conducive to the development of the agricultural ecological transformation (Luo et al., 2023).

Figure 4 indicates the changing trend of the threshold effect of low-carbon technological innovation. In general, in most regions in China, the low-carbon technological innovation is relatively strong, and it is on a trend of strengthening year by year. Under the dual pressures of resource and environmental, low-carbon technological innovation is a new engine for reshaping agricultural development models and holds significant practical implications for the agricultural ecological transformation. Promoting the formation of a new pattern of green and low-carbon development is not only conducive to meeting the growing aspirations of the people for a better life, but also contributes to the strategic goal of achieving peaking carbon emissions and carbon neutrality. Therefore, according to the threshold effects of low-carbon technological innovation in the impact of digital economy on agricultural ecological transformation, the governments can use the mechanism of low-carbon technological innovation to fully exploit the promoting role of digital economy on agricultural ecological transformation. Specifically, in most provinces of China, the level of low-carbon technological innovation is relatively high, while the provinces with low levels of low-carbon technological innovation are showing a downward trend year by year. However, it is still necessary to further improve the level of low-carbon technological innovation in various regions, in order to lead the development of agriculture and rural modernization in a new direction through the digital economy and low-carbon technological innovation. This includes transforming the development mode of agriculture, promoting rural revitalization and high-quality agricultural development, and facilitating the improvement of agricultural quality, efficiency, and ecological transformation.

Fig. 4: Trend in different thresholds of low-carbon technological innovation (LCI).
figure 4

This indicates the changing trend of the threshold effect of low-carbon technological innovation.

Full size image

Discussion

Under the new round of information technology revolution, the digital economy has penetrated into all aspects of society and profoundly changed people’s production and lifestyle, exerting tremendous impact on the development of China’s agricultural economy. The digital economy can be fully and deeply applied to various links in the entire agricultural industry chain. By embedding agricultural factor allocation and production management system, it can promote the greening of the entire process of agricultural production, processing, circulation, and consumption, thus promoting the quality and efficiency of agriculture. Furthermore, the digital economy based on data information as a basic element, its role depends not only on digital infrastructure, but also on the local low-carbon technological innovation level. Only when low-carbon innovation technology reaches a certain level can it meet the usage environment of the digital economy and fully promote the digitization of agricultural ecological transformation (Shen et al., 2022).

The research findings confirm the U-shaped non-linear relationship between digital economy and agricultural ecological transformation. On the one hand, when the level of low-carbon technological innovation is low, the digital economy hinders the agricultural ecological transformation. The development of technology level does not match the development of digital economy, limiting the enhancement effect of digital economy on agricultural ecological transformation. The insufficient input of data elements and low production efficiency resulting from the traditional agricultural development methods have become obstacles to agricultural ecological transformation (Goel et al., 2021). Additionally, when the green innovation capability is low, local governments will invest some resources into technological research and development. However, in general, technological innovation is difficult and long-lasting, with a large amount of funding being consumed for a long period of time, while the technological output is slow and limited. The diversion of resources prevents the government from focusing all its efforts on agricultural ecological transformation and hinders the modernization process of agriculture.

On the other hand, with the further enhancement of low-carbon technological innovation, it can effectively exert the empowering effect of the digital economy, thus promoting the agricultural ecological transformation. As a powerful engine for unleashing economic vitality, the digital economy can break the constraints of information exchange in time and space, effectively promote the transfer and flow of production factors between regions and agricultural industries, which inevitably relies on the promotion of innovative technologies (Pakseresht et al., 2022). The knowledge spillover and technical reward-increasing effect brought by low-carbon technological innovation promote the rational allocation of resources, help to improve traditional agricultural technology level and output efficiency, and promote the transformation of new and old kinetic energy. Then, in the era of the digital economy, low-carbon innovative technology can provide important support for meeting the needs of green production and intensive management, which is conducive to fully unleashing the multifunctionality of modern agriculture product production, economic growth, and ecological protection, extending the industrial chain, and promoting the improvement of the quality and efficiency of agriculture. The continuous iteration and updating of new business formats and models create more employment opportunities, thereby promoting farmers’ income increase and the coordinated development of ecology and economy. In addition, the improvement of low-carbon technological innovation level inevitably accompanies the increase of agricultural science and technology talents, stimulating the development of agricultural electronic commerce, greatly changing the connection mode of agricultural production and consumption, and consumers conveying their demand for greening and health through purchasing preferences and other mechanisms, which in turn promotes farmers to adopt green production technology and forces agriculture to transform towards greening production.

In order to ensure the reliability of the research findings, we conducted a robustness test on the relationship between the digital economy and agricultural ecological transformation. The results are shown in Table 8. The Hansen test shows Prob>chi2 = 0.439, which fails to reject the null hypothesis of instrumental variable validity. The results of Auto-Regressive Moving Average Model (AR) tests also indicate the model is relatively reasonable.

Table 8 AR (1) and AR (2) test.
Full size table

Conclusions, implications and future research directions

We have estimated the level of agricultural ecological transformation based on Super-SBM Model Setting. In addition, from the perspective of low-carbon technological innovation, we have built a non-linear dynamic panel threshold model to systematically explore the heterogeneous effects of the digital economy on the agricultural ecological transformation at different levels of low-carbon technological innovation. The research conclusions are as follows:

  1. (1)

    The current level of agricultural ecological transformation China is relatively low, and there is still significant room for development; In the long-term trend, the overall development in rural areas shows a stable upward trend, with an average annual growth rate of 6.99% from 2013 to 2021, and different regions show local fluctuations, indicating a significant “wealth gap”. Among them, Zhejiang Province has the highest level of digital rural development, while Hainan Province has the relatively lowest level of digital development.

  2. (2)

    Importantly, our new insight is that the role of digital economy in the transformation of agricultural ecology is influenced by the level of low-carbon technological innovation. Lower levels of low-carbon technological innovation will significantly inhibit the impact of digital economy on agricultural ecological transformation. However, as the level of low-carbon technological innovation increases and exceeds the critical value, the driving effect of digital economy is deeply stimulated. Digital economy improves agricultural ecological transformation to a certain extent, indicating a “U” shaped relationship between digital economy and agricultural ecological transformation.

  3. (3)

    For other influencing factors of agricultural ecological transformation, disasters rate, urbanization rate, and agricultural internal structure have a significant positive impact on agricultural carbon productivity, while agricultural industrial structure shows a negative promoting effect on agricultural ecological transformation, but this effect is relatively not significant.

Our research has expanded the growth path of improving agricultural ecological transformation and provided new experience for promoting the development of low-carbon agriculture:

  1. (1)

    How to strengthen, optimize, and stabilize digital economy in developing countries: According to our evaluation index data, firstly, the government should focus on promoting the construction of digital infrastructure, establishing a digital economic platform, and actively promoting the deep integration of traditional industries and digital technology. The government should strengthen policy guidance, increase its financial support for the construction of digital infrastructure to consolidate the foundation of digital economic development. Secondly, the government should increase its efforts in digital technology innovation and talent cultivation through subsidies, rewards, and preferential policies, guiding and supporting enterprises, research institutions, and universities to carry out collaborative innovation in digital technology. A sound system should be established to guarantee the achievements of digital innovation and application, creating a favorable environment for digital innovation and application. Finally, the government should pay attention to the coordinated development of the digital economy among regions. By using the virtual technological characteristics of the digital economy and surpassing the administrative boundaries of regions, digital economy connectivity and sharing should be achieved in larger areas. The eastern region should fully leverage its advantages in digital economic infrastructure and continuously improve the level and quality of digital application, guiding relevant digital technology to spread to the central and western regions. The central and western regions should seize the opportunities for digital economic development, focusing on strengthening the construction of digital infrastructure, bridging the digital divide, increasing investment in digital technology innovation, improving the capacity for digital technology absorption, digestion, and application, and strengthening digital transformation and upgrading.

  2. (2)

    In the era of digital economy, digital transformation is an important path to improve agricultural ecological transformation. In the face of the constraints and challenges of the dual-carbon goal, the digital economy is an important driving force in promoting the agricultural ecological transformation. In the era of rapid changes in information technology, the government should accelerate the construction of digital information infrastructure in rural areas, achieve interconnection of data information among regions, establish a sound digital economic ecosystem, guide the innovation and development of low-carbon technologies driven by the digital economy, and empower the agricultural ecological transformation. Firstly, the digital economy itself should pay attention to digital construction, focusing on the development of core industries and key technologies of the digital economy, emphasizing the research and development of green energy and the application of green technologies, and continuously improving the level of industrial digitization and digital industrialization. Secondly, it is necessary to promote the deep integration of the digital economy and agriculture, expand the scale effect of agricultural production, optimize the flow channels of factors, effectively improve resource utilization efficiency, maximize the promotion of low-carbon and green innovation on the agricultural ecological transformation, strengthen the development of agricultural economy, and promote high-quality development of agriculture. Therefore, regions should formulate suitable policies to incentivize technological innovation and provide a solid information technology infrastructure for the digital economy.

    It should be noted that in the process of promoting the improvement of agricultural ecological transformation, the heterogeneous role of low-carbon technological innovation cannot be ignored. To achieve the regional agricultural ecological transformation, it is necessary to consider the threshold effect of the low-carbon and green innovation level in each region and to plan reasonable paths according to the specific contexts of different regions, while being aware of the negative impact of heterogeneity threshold effects. To address the environmental differences and digital economy’s varying impacts on agricultural ecological transformation in different regions, regional barriers should be broken down, and the integration of digital information technology between regions should be deepened. For regions with relatively low levels of low-carbon technological innovation, such as Qinghai, Ningxia, Inner Mongolia, and other provinces, the local low-carbon technological innovation ability is relatively weak due to geographical location and regional economic development constraints, and its role in promoting China’s agricultural ecological transformation is relatively limited. The government still needs to increase its emphasis on low-carbon technology innovation, actively optimize the types and proportions of factor inputs, enhance investment in technology innovation, cultivate high-level innovative talents, reduce reliance on external clean technologies, and promote the development of innovation cooperation mechanisms among production, academia, and research to improve the region’s green technology innovation ability. Conversely, for regions with relatively high levels of low-carbon and green innovation, digital economic construction should be further strengthened under the existing low-carbon and green innovation level. Moreover, attention should be paid to moderately raising the low-carbon technological innovation level, fully leveraging technology innovation advantages, transforming agricultural development mode, and promoting regional agricultural green development and ecological transformation.

  3. (3)

    (3) Other factors that help improve agricultural ecological transformation: In terms of disaster relief, as agricultural natural disasters are unavoidable, it is necessary to improve measures for defense and control of natural disasters. The standards for agricultural infrastructure construction should be improved, and new materials and technologies should be continuously introduced and promoted to enhance the modern agriculture’s ability to resist natural disasters. Before disaster occurs, local governments should pay attention to climate change and make reasonable arrangements for farming. After the disaster, measures should be taken promptly according to the actual situation to reduce the impact of the disaster within a controllable range. In terms of the urban population ratio, based on the needs of industrial upgrading, population distribution, and other factors in each region, rural residents should be reasonably and orderly transferred to the urban areas to gradually break down the household registration system barrier between urban and rural areas, eliminate obstacles for rural surplus labor to enter the city, and substantially increase the urban population ratio. In terms of agricultural internal structure, the adjustment of the internal structure should be in line with the development law of the market, gradually enhance the market competitiveness, organize special investments, continuously upgrade the advantageous agricultural product industry chain, produce agricultural products with high quality and quantity standards, and continuously cultivate and develop the farm’s superior industries. In terms of rural human capital, we should increase the investment in rural human capital, accelerate the improvement of rural human capital level, formulate detailed policies for the development of rural human resources, establish a sound education and training system for ecological agriculture operators and farmers, strengthen green technology training and practical guidance for farmers, enhance the quality of rural labor force, and continuously improve the skill of green production by farmers.

The article still has some limitations that can be further studied in the future. First, due to the availability of the existing statistical data, the sample selected in this paper mainly focus on the regional macro level. With the continuous updating of the database, future research can consider expanding the data dimension, such as the enterprise or micro level research. Second, the impact of regional heterogeneity factors, in addition to low-carbon technological innovation, may also include environmental regulations, industrial integration, etc, which can be further verified.