Does Digital Village Construction Affect the Sustainable Intensification of Cultivated Land Use? Evidence from Rural China

Abstract

Digital village construction (DVC) is a crucial pathway for increasing farmland productivity, reducing agricultural waste, and ultimately achieving sustainable development goals (SDGs). However, its effects on the sustainable intensification of cultivated land use (SICLU) remain unclear. To bridge this gap, this study investigated the impact effects and mechanisms of DVC on SICLU across 358 counties in China using ordinary least squares and mediating effect models. The results showed the following: (1) DVC and its four sub-indices had significant and positive impacts on SICLU, which were validated through a series of robustness tests. (2) Heterogeneity analysis showed that DVC significantly improved SICLU in the eastern and central regions, as well as in regions with abundant and relatively scarce resource endowments, whereas no such effect was observed in the western region. (3) The relationship between DVC and SICLU was mediated by farmers’ income, technological innovation, and agricultural informatization. These insights highlight the importance of accelerating DVC to enhance SICLU.

1. Introduction

Cultivated land is the most fundamental resource and material security for human survival and development [1,2,3] and is closely related to global food and ecological security [4,5]. Rising food demand, triggered by global population growth and compounded by decreased cultivated land due to urban expansion, poses significant challenges to global food security [6,7]. Intensifying cultivated land use (CLU) has proven effective for ensuring global food security by increasing the agricultural output per unit area [8,9]. However, the over-intensification of CLU, particularly the excessive input of agricultural chemicals, has triggered a sequence of ecological and environmental problems, such as land degradation, agricultural non-point source pollution, and biodiversity loss [10,11]. This inevitably hinders the sustainability of CLU and threatens food security, thereby impeding sustainable development [12,13,14].

The concept of sustainable intensification (SI) was initially introduced by Pretty (1997) [15] to explain the aim of achieving a considerable increase in agricultural output while protecting the environment and natural resources. SI is widely considered an effective way to balance the contradiction between increasing agricultural output and farmland ecosystem services and protecting the ecological environment [16,17]. SI directly contributes to multiple sustainable development goals (SDGs), particularly SDG 2 (Zero Hunger) and SDG 15 (Life on Land), by enhancing productivity while minimizing environmental degradation, thereby serving as a critical pathway toward achieving sustainable development [18]. As the world’s most populous country, China’s food security and over-intensification of CLU have attracted worldwide attention [19,20]. To address the challenges in utilizing and protecting cultivated land in China, some scholars have conducted targeted SI studies from the perspective of CLU, i.e., the sustainable intensification of cultivated land use (SICLU) [9,21]. In this circumstance, SICLU is deemed to be an important direction for the transition of CLU in China [22].

With the wide application of networking, informatization, and digitization in agricultural and rural development, agricultural and rural areas are moving toward digitalization by using big data, artificial intelligence, 5G, etc. [23,24]. The novel notion of the digital village, referring to the digital transformation of traditional agricultural practices and villages, has emerged and become a global tide [25,26,27]. Digital village construction (DVC) has received widespread attention from the governments of many countries, such as the EU Action for Smart Villages, Digital India, and Land Digital of Germany. In China, DVC is regarded as a necessary condition for achieving agricultural and rural modernization and meeting the increasing challenges of food security and ecological environmental protection [28,29]. The Chinese government has issued a set of policy documents aiming at accelerating DVC’s advancement, such as the Digital Countryside Development Strategy in 2019 and Action Plan for the Development of Digital Village (2022–2025) in 2022. Referring to the Digital Countryside Development Strategy in 2019, this study defines DVC as the endogenous process of modernization and transformation in agricultural and rural areas, which is accompanied by the application of networking, informatization, and digitization in the economic and social development of agricultural and rural areas.

The incorporation of digital technology within agricultural systems and rural development frameworks through DVC can substantially increase the potential of digital technologies [18]. Digital technology has now become deeply integrated into every facet of agricultural development [30]. DVC has enhanced the possibility of increasing agricultural output while reducing negative environmental impacts [31], providing effective solutions to the challenges encountered by agricultural and rural areas [32]. Scholars have explored the impact of DVC on agricultural modernization, county-level economic growth, and farmers’ income growth [27,33,34]. However, the relationship between DVC and SICLU remains unclear. A systematic investigation into their relationship is crucial for enhancing agricultural sustainability and promoting its green transition.

2. Literature Review

Extensive theoretical and empirical studies have been conducted on SICLU, including its conceptual definition [8,35], quantitative measurement [36,37], and driving factors [38,39]. SICLU has been defined from different perspectives. For example, Xie et al. (2021) [9] defined SICLU as enhancing the output efficiency of cultivated land by optimizing the input–output relationship, mitigating ecosystem damage, and augmenting or sustaining the resilience of a CLU system. Lyu et al. (2022) [21] summarized the connotations of SICLU as intensive management, high productivity, resource conservation, maintenance of the ecological environment, and the sustainable development of society. Building on these definitions, this study defines SICLU as optimizing the input–output relationship of a CLU system to increase yield efficiency, reduce environmental load, and promote the ecosystem services of the system. Regarding the evaluation of SICLU, material flow analysis and emergy analysis are the most widely used methods [9,40,41]. Moreover, some scholars have investigated the influencing factors of SICLU from multiple perspectives. Hou et al. (2023) [40] found that economic agglomeration generally exerts negative impacts on the SICLU of the local and surrounding regions. Lyu et al. (2024) [42] pointed out that all cultivated land renting-in, the ratio of family dependency, and the average years of education are conducive to improving SICLU.

To reveal the role of digital village in promoting the collaborative development of food and ecological security, the impact of digital village on the green transformation of agriculture and CLU has been widely investigated [43,44]. Garnett et al. (2013) [45] proposed that modern information and communications technology and suitable financial instruments enable farmers to adopt SI practices. Krintz et al. (2016) [46] noted that modern information technology can help farmers to achieve the dual goals of simultaneously increasing yield and improving the ecological environment. Weersink et al. (2018) [43] stated that digital agricultural transformation allows the environmental management of food systems to be improved. Shen et al. (2022) [47] confirmed that Internet popularization and digital technology are capable of promoting green growth in the agricultural sector. Singh et al. (2022) [48] pointed out that the application of Industry 4.0 Technologies in agriculture could help to map data on the soil and environment, thereby increasing productivity by smart controlling input. Du et al. (2023) [49] found that DVC could significantly improve agricultural green total factor productivity through two potential transmission mechanisms: scaled agribusiness operations and agricultural informatization (AgI). Guo (2024) [50] demonstrated that DVC could enhance agricultural green total factor productivity by accelerating green technological advancements and alleviating land, capital, and labor resource mismatches. Zhang et al. (2024) [30] pointed out that agricultural and rural areas’ digitalization exerted a substantial positive effect on agricultural green total factor productivity. Fu et al. (2024) [44] found that rural digital transformation improves cultivated land green use efficiency through optimizing the allocation of input factors, increasing expected output, and reducing unexpected output. Tan et al. (2024) [51] suggested that digital financial inclusion is conducive to cultivated land green use efficiency by optimizing input and increasing output. Therefore, DVC may affect SICLU by influencing the input–output relationship of the CLU system.

Although existing studies provide important references for the relationship between DVC and SICLU, there is no empirical evidence for this relationship. Therefore, this study investigated how DVC affects SICLU in Chinese counties in order to understand the role of DVC in promoting SICLU. Specifically, we addressed the following three critical questions: (1) Does DVC influence SICLU at the county level? (2) Does the effect that DVC has on SICLU vary across regions? (3) What are the transmission mechanisms underlying the relationship between DVC and SICLU?

The following are the main contributions of this study: First, to our knowledge, this is the first attempt to explore the effect of DVC on SICLU and its heterogeneity. This bridges the gap in the relationship between DVC and SICLU, clarifies how DVC contributes to the achievement of SDGs, and provides policy insights for promoting the transformation of green agriculture through digital technology. Second, the mediating roles of farmers’ income (FI), technological innovation (TI), and AgI were discussed to further reveal the internal mechanism of DVC affecting SICLU. This study could help policymakers to incorporate DVC and these mediating factors into a unified policy system, thus providing more effective guidance for improving SICLU and promoting sustainable agricultural development from the perspective of DVC. Third, different from existing studies that explored the influencing factors of SICLU in provincial administrative regions, the direct and indirect effects of DVC on SICLU in Chinese counties were estimated, respectively, in this study, thereby enriching the research on SICLU.

3. Mechanism Analysis and Research Hypotheses

Figure 1 illustrates the mechanisms of DVC affecting SICLU. It involves both direct and indirect effects. The direct effect reveals the fundamental mechanisms through which DVC affects SICLU. The indirect effect explains how the mediating variables mediate the relationship between DVC and SICLU.

3.1. Direct Effect of DVC on SICLU

DVC stimulates the improvement of cultivated land productivity and the green transition of CLU by integrating digital technology into agricultural production and farmers’ lives [43,52]. The DVC evaluation system used in this study consisted of the following four secondary indicators: digital infrastructure construction in rural areas (DIC), digitalization of the rural economy (DRE), digitalization of rural governance (DRG), and digitalization of rural life (DRL). To more comprehensively reveal the relationship between DVC and SICLU, we analyzed the impact of these four secondary indicators on SICLU, respectively.

(1)

DIC serves as a crucial support and carrier for digital village and can embed digital technology into agricultural production and operation decisions and enhance data acquisition and sharing in rural areas [53]. According to asymmetric information theory, DIC can broaden information channels and coverage and speed up information transmission, which can reduce information asymmetry in agricultural production and management, thereby increasing factor productivity and SICLU [54]. Moreover, DIC, especially the construction of digital infrastructure for collecting and analyzing farmland and environmental monitoring data, can provide information for precise resource inputs, yield monitoring, and environmental management, which helps to boost yields, avoid excessive input of agricultural chemicals, and decrease agricultural pollution emissions, thus promoting the sustainable development of agriculture and improving SICLU [43].

(2)

DRE is the core of DVC and a new driving force for sustainable rural development. DRE improves SICLU mainly by promoting the development of rural digital supply chains and rural digital marketing and the digital transformation of inclusive finance. The advancement of the digital supply chain (the increase in rural logistics networks) and digital marketing (the development of rural e-commerce) could increase the channels for farmers to purchase green agricultural inputs and sell green agricultural products, decrease the transaction costs of agricultural inputs and outputs, and enhance farmers’ green production motivation [55]. This facilitates the implementation of green production technology and the input of green agricultural material, which will reduce negative environmental impacts, thereby promoting sustainability and increasing SICLU [56]. Moreover, digital financial inclusion helps to alleviate the negative impact of capital scarcity in the agricultural sector on the application of new agricultural technology, resulting in an improvement in SICLU [51,57].

(3)

According to the digital governance theory, DRG is capable of changing the information asymmetry between governance entities and enhancing the overall efficiency of rural grassroots work through introducing modern information technology [58]. DRG increases the transparency of various agricultural policies, and thereby promotes their implementation by applying digital tools, such as Alipay, WeChat, and DingTalk, in rural grassroots government services in China. This would help to enhance farmers’ enthusiasm for implementing conservation tillage and increase agricultural production efficiency, thus improving SICLU [59,60]. Strict cultivated land protection is a fundamental requirement for rural governance, and cultivated land governance is an important part of rural governance. The digital transformation of cultivated land governance (e.g., the adoption of information technology in farmland fragmentation governance, the ecological restoration of farmland, and high-standard farmland construction) can improve governance efficiency, boost yields, and improve farmland ecological environment, thereby promoting the sustainability of agriculture and SICLU [61,62,63].

(4)

DRL facilitates access to more entertainment, education, and other resources in rural areas through Internet platforms, thus affecting SICLU. Specifically, increasingly prosperous online information and gradually popularizing smartphones and computers have provided more opportunities to promote agricultural technology extension and enhance farmers’ digital literacy and skills. This can promote the adoption of digital technology in agricultural cultivation and enhance farmers’ farmland management capabilities, thereby increasing SICLU [43,64]. Moreover, DRL enhances farmers’ awareness of green production through expanding their social networks and channels and promoting the sharing and dissemination of information and knowledge on the green transition of CLU [65,66]. This further encourages the adoption of green farming techniques and reduces the dependence on agricultural chemicals, thereby reducing negative environmental impacts and improving SICLU [44,47,67]. In summary, the following hypothesis is formulated:

Hypothesis 1. 

DVC, DIC, DRE, DRG, and DRL contribute to improving SICLU.

3.2. Indirect Effect of DVC on SICLU

3.2.1. The Mediating Role of FI Between DVC and SICLU

DVC increases FI mainly by advancing the progress of digital agricultural and rural e-commerce and popularizing smartphones and broadband. Specifically, the adoption of digital and geospatial information technology in agricultural practices improves agricultural efficiency through facilitating precise resource input and effective farmland management, thereby increasing FI [68,69]. Rural e-commerce can increase the sales volume of agricultural products by broadening sales channels and provide farmers with more opportunities for employment and entrepreneurship, thus driving FI growth [70,71,72]. By using smartphones, farmers can swiftly and easily obtain and share information and learn advanced skills and knowledge, which can enhance their labor skills and decision-making efficiency, thereby increasing FI [69,73]. The increase in FI encourages farmers to invest in high-quality production resources and apply advanced green technologies, which enhance agricultural productivity and decrease agricultural waste emission, thus enhancing SICLU [74,75,76]. Accordingly, we propose the following hypothesis:

Hypothesis 2. 

DVC improves SICLU through increasing FI.

3.2.2. The Mediating Role of TI Between DVC and SICLU

DVC stimulates TI by promoting the adoption of digital technology, accelerating the popularization of mobile phones and broadband, and promoting the development of digital inclusive finance. Specifically, the wide utilization of digital technology provides technical and data support for TI, thereby accelerating it [77,78]. Moreover, the popularization of smartphones and broadband in rural areas widens farmers’ access to information and provides them with more opportunities for agricultural technology training, which increases their willingness and ability to adopt new technology and ultimately fosters new agricultural TI [39,79]. By providing financial support, increasing financial service products, and decreasing financing costs, digital inclusive finance provides financial support for TI, thus accelerating agricultural TI [79,80]. In the theory of endogenous growth, TI is recognized as a determinant of sustained economic growth [81,82]. There is a consensus that TI is a main driver of productivity improvement and the green transformation of agriculture [83,84]. TI, especially the innovation in emergent and green technologies, can increase yields while reducing agricultural inputs, thereby increasing CLU efficiency, reducing environmental load, and ultimately improving SICLU [85,86]. Consequently, the following hypothesis is proposed in this paper:

Hypothesis 3. 

DVC increases SICLU by promoting TI.

3.2.3. The Mediating Role of AgI Between DVC and SICLU

In the theory of production factors, information is considered an important production factor driving long-term economic growth [87]. As the digital information age arrives, digital information has been a crucial production factor in digital agriculture [88]. DVC promotes the wide application of information technology in agricultural production and rural development by accelerating the construction of information infrastructure, thereby promoting full-chain informatization in agriculture and all-round informatization in rural society [89]. Moreover, DVC provides information from various open network platforms, which accelerates information flow and addresses information asymmetry in agricultural production and marketing, thus promoting the development of AgI [49]. AgI stimulates agricultural technical advancement by promoting the dissemination of advanced technology, breaking down information barriers, and penetrating agricultural information into every aspect of agricultural production, thereby improving SICLU [49,90]. Moreover, the advancement of AgI provides more accurate information for agricultural resource input and farm management, which can increase resource use efficiency and reduce waste discharge, thereby improving SICLU [31,91]. Therefore, the following hypothesis is proposed:

Hypothesis 4. 

DVC improves SICLU through promoting AgI development.

4. Data, Method, and Variables

4.1. Study Area and Data Sources

Based on data availability, 358 counties in mainland China were selected as research units (Figure 2). Of these, 43, 147, and 168 counties were located in the eastern, central, and western regions, respectively. DVC data were collected from the County Digital Rural Index in 2020, which was jointly published by the Institute for New Rural Development of Peking University and the Ali Research Institute in May 2022. The other data used in this study included two categories: statistical and raster data. The statistical data included input–output data of the CLU system (e.g., the input of labor, fertilizer, and pesticide) and data on socio-economic development (e.g., the population, patent authorization, and mobile phone owners). The raster data included spatial distribution maps of cultivated land, elevation, solar radiation, and soil.

Table 1. Data source and description.

Data Type	Data Format	Data Resolution	Data Source
County Digital Rural Index	Spreadsheet	–	Institute of New Rural Development, Peking University (http://www.ccap.pku.edu.cn/nrdi/xmycg/yjxm/363361.htm, accessed on 20 April 2024)
Statistical data on the input–output of the cultivated land use system	Spreadsheet	–	The third national land resource survey, the statistical yearbook and water resources bulletin of the prefecture-level city where the county is located
Statistical data on the socio-economic development	Spreadsheet	–	The national economic and social development statistical bulletin of each county; the state statistical bureau (https://www.stats.gov.cn, accessed on 20 April 2024); the national intellectual property (https://www.cnipa.gov.cn, accessed on 20 April 2024)
Cultivated land data	Raster	30 m	Annual China Land Cover Dataset (https://zenodo.org/records/8176941, accessed on 20 April 2024)
Digital elevation model data	Raster	30 m	The Geospatial Data Cloud (https://www.gscloud.cn/sources/index?pid=302, accessed on 20 April 2024)
Solar radiation data	Raster	0.25°	National Ecosystem Science Data Center (http://nesdc.org.cn/sdo/detail?id=62b95e437e281714dccbd1f2, accessed on 20 April 2024)
Soil data	Raster	30 arc-seconds	The National Tibetan Plateau Data Center (https://data.tpdc.ac.cn/zh-hans/data/8ba0a731-5b0b-4e2f-8b95-8b29cc3c0f3a, accessed on 20 April 2024)

lists the data sources and descriptions used in this study.

4.2. Materials and Methods

4.2.1. Calculation Method for SICLU

Emergy analysis can achieve comparability across various input–output elements of the CLU system and indicate the direction of energy flow of this system, thereby facilitating the evaluation of energy flow across various components of the system [9,92,93]. Referring to previous studies [9,40], SICLU was evaluated using emergy analysis in this study. The assessment steps are as follows:

First, this study identified the spatiotemporal boundaries and drew an energy systems language (ESL) diagram of the CLU system [94]. In Figure 3, the large rectangular box is the system boundary; the left part of the system lists the inputs of natural environmental resources, including the sun, rain, and wind; the upper side shows the inputs from human society, such as fuel, fertilizer, and pesticide; and the right part represents the system outputs, including market and non-market outputs (e.g., grain, ecosystem services, and waste).

Second, we identified the major matter and energy flows of the CLU system. Based on this, an emergy indicator table was established (

Table 2. Emergy indicators for evaluating SICLU.

Type	Index	Indicators
Inputs	Local renewable environmental resources	R
	Local non-renewable environmental resources	N
	Renewable purchased inputs	FR
	Non-renewable purchased inputs	FN
	Total emergy input	U = R + N + FR + FN
Outputs	Agricultural product outputs	YA
	Ecosystem services	YE
	Waste outflows	YW
	Total emergy output	Y = YA + YE − YW

). Then, we converted various flows into a unified unit of “solar emergy” (sej, solar energy emjoules) using Equation (1) (the calculation details are listed in Appendix A, Appendix B, Appendix C and Appendix D).

$$E_m={\displaystyle \sum _{i=1}^n{f}_i}\times UE{V}_i$$

(1)

where E_m refers to the total solar emergy of the CLU system; f_i is the ith material and energy flow; and UEV_i represents the unit emergy value of the ith material and energy flow.

Third, we established the assessment indicator system for SICLU. According to emergy theory and SICLU’s connotation, and referring to existing studies [9,93], we established the following four indices: emergy productivity ratio (EPR), emergy yield ratio (EYR), environmental loading ratio (ELR), and environmental economic efficiency (EE). The SICLU was then calculated based on these indicators.

$$EPR={\displaystyle \frac{Y}{U}}$$

(2)

$$EYR={\displaystyle \frac{Y_A}{(F_R+F_N)}}$$

(3)

$$ELR={\displaystyle \frac{(N+F_N)}{(R+F_R)}}$$

(4)

$$EE={\displaystyle \frac{Y_W}{R}}$$

(5)

$$SICLU={\displaystyle \frac{EPR\times EYR}{ELR\times EE}}$$

(6)

4.2.2. Basic Model

According to the mechanism analysis, the classic ordinary least squares (OLS) model was designed to evaluate the direct effect of DVC on SICLU. The formula is as follows:

$$DV{C}_i={\alpha }_0+{\alpha }_{1}SICL{U}_i+{\alpha }_{2}ln{X}_i+{\epsilon }_i$$

(7)

where DVC_i is the explanatory variable, standing for the DVC of county i; SICLU_i is the explained variable, representing the SICLU of county i; X_i indicates a set of control variables; α₀ refers to a constant term; and α₁ and α₂ are the regression coefficients of the corresponding variables.

4.2.3. Mediating Effect Model

To reveal the underlying mechanism of DVC affecting SICLU, we built the following two-step mediating effect model [95]:

$$ln{M}_i={\beta }_0+{\beta }_{1}SICL{U}_i+{\beta }_{2}ln{X}_i+{\epsilon }_i$$

(8)

$$DV{C}_i={\lambda }_0+{\lambda }_{1}SICL{U}_i+{\lambda }_2{M}_i+{\lambda }_{3}ln{X}_i+{\epsilon }_i$$

(9)

where M_i refers to a series of mediating variables; X_i is the control variable; β₀ and λ₀ represent the constant term; and β₁, β₂, and λ₁–λ₃ stand for the regression coefficients of the corresponding variables.

4.3. Variables Selection

4.3.1. Explained Variable

SICLU was selected as the explained variable, representing the level of SICLU in each county.

4.3.2. Explanatory Variable

The DVC was the explanatory variable and was obtained from the County Digital Rural Index (2020). Currently, it is the most authoritative and reasonable indicator of the DVC level in China.

4.3.3. Control Variables

To control the potential influence of the other variables, the following five control variables were incorporated into the econometric models.

(1)

Multiple cropping index (MCI). The improvement of MCI could increase the utilization ratio of cultivated land, thereby increasing agricultural output and SICLU [96]. However, the enhancement of MCI may increase the input of agricultural chemicals and even lead to the excessive exploitation of cultivated land, thereby increasing environmental loading and agricultural waste and even decreasing cultivated land productivity. This will ultimately decrease SICLU. MCI was measured by the ratio of the total planting area of crops to the cultivated land area [96].

(2)

The proportion of the sown area of grain crops (SAGC). Compared with grain crops, cash crops usually require a higher input of agricultural chemicals, thereby producing more agricultural waste and ultimately decreasing SICLU. However, farmers may invest in agrochemicals in grain production to gain more profits, which may threaten SICLU. In this study, SAGC was defined as the proportion of the sown area of grain crops in the total sown area of crops [40].

(3)

Per capita cultivated land (CLA). CLA reflects the cultivated land resource endowments in each county. When CLA is relatively low, other types of production materials will be invested in agricultural production as substitutes for cultivated land. This may further affect SICLU. CLA was calculated by dividing the total cultivated land area by the total resident population [97].

(4)

Ratio of agricultural employees (PAE). To some extent, PAE reflects the abundance of labor force engaged in agricultural production. A sufficient agricultural labor force may promote the intensive cultivation of farmland, reduce the risk of farmland abandonment, and substitute for the use of some agricultural chemicals, which may be conducive to SICLU. This study selected the proportion of agricultural employees in the rural labor force as the proxy for PAE [98].

(5)

Output value per unit area (OVP). OVP reflects the development level of the agricultural economy in a specific region. Agricultural economy development can accelerate the advancement of agricultural technology and scale management of cultivated land, thereby increasing farmland productivity and SICLU. However, as the agricultural economy develops and agricultural production scale expands, agricultural production is confronted with increasing environmental pressure, which is not conducive to improving SICLU. Referring to Cao et al. (2022) [99], the OVP was evaluated as the value added of the primary industry per unit of cultivated land area.

4.3.4. Mediating Variables

According to the mechanism analysis and research hypotheses, the following three mediating variables were selected to explore the underlying transmission mechanism of DVC affecting SICLU: FI, TI, and AgI. FI was represented by the per capita disposable income of farmers [100]. TI was measured using the number of technical patent applications. Specifically, it was calculated as the weighted summation of the number of applications accepted for invention, utility model patents, and design patents, with weights of 0.5, 0.3, and 0.2, respectively [101]. Moreover, following Du et al. (2023) [49], we used the penetration rate of mobile phones as the proxy for AgI.

5. Results

5.1. Spatial Distributions of DVC and SICLU

The spatial distributions of the DVC and SICLU at the county level in 2020 are shown in Figure 4a,b, respectively. We divided the DVC and SICLU values of the sample counties in 2020 into five levels using the natural breakpoint method by ArcGIS 10.2, respectively. Figure 4a,b reveals that both DVC and SICLU were randomly distributed and demonstrated obvious regional heterogeneity. On the whole, DVC and SICLU exhibited relatively similar distribution patterns, with high values mainly concentrated in the eastern and central regions. However, the floating range of SICLU was larger than that of DVC. Figure 4a shows that, among the 358 counties, 102 had DVC values higher than 60.68, whereas 115 had DVC values below 51.41. As shown in Figure 4b, SICLU exceeded 29.32 in only 37 counties; however, 251 counties had SICLU values below 18.13. These results indicate that the overall SICLU in the study area is relatively low and requires further improvement.

5.2. Benchmark Regression Results

The regression results of the direct effect are reported in

Table 3. Regression results of the direct effect of DVC on SICLU.

Variable	(1)	(2)	(3)	(4)	(5)
	SICLU	SICLU	SICLU	SICLU	SICLU
DVC	0.672 ***
	(6.66)
DIC		0.202 ***
		(3.25)
DRE			0.489 ***
			(6.76)
DRG				0.103 **
				(2.41)
DRL					0.402 ***
					(6.72)
MCI	−0.464 ***	−0.481 ***	−0.324 ***	−0.485 ***	−0.533 ***
	(−3.91)	(−3.58)	(−2.85)	(−3.56)	(−4.09)
SAGC	−0.058 **	−0.039	−0.052 **	−0.032	−0.059 **
	(−2.31)	(−1.41)	(−2.25)	(−1.14)	(−2.16)
CLA	−0.516 **	−0.830 ***	−0.490 **	−1.053 ***	−0.735 ***
	(−2.47)	(−3.38)	(−2.19)	(−4.37)	(−3.90)
PAE	0.140 ***	−0.003	0.085 **	−0.044	0.008
	(3.12)	(−0.07)	(1.97)	(−1.01)	(0.20)
OVP	0.124 **	0.114 *	0.034	0.113 *	0.165 **
	(2.28)	(1.78)	(0.50)	(1.73)	(2.48)
Constant	−14.452 **	16.461 **	1.153	30.557 ***	13.558 ***
	(−2.04)	(2.11)	(0.22)	(6.14)	(3.04)
R2	0.3697	0.1279	0.3665	0.1101	0.3332
N	358	358	358	358	358

Notes: The data in parentheses are robust standard error. ***, **, and * represent the significance at 1%, 5%, and 10%, respectively.

. Column (1) presents the direct effect of DVC on SICLU. The estimation results suggest that DVC significantly contributed to improving SICLU. On average, every 1% increase in DVC resulted in a 0.672% advancement in SICLU. For the control variables, the coefficients of MCI, SAGC, and CLA were significantly negative, indicating that the increase in the multiple cropping index, the proportion of the sown area of grain crops, and the per capita cultivated land area restrained the improvement of SICLU. In contrast, the coefficients of PAE and OVP were significantly positive, showing that the ratio of agricultural employees and output value per unit area can stimulate the increase in SICLU. Additionally, the regression results by dimension, which investigate the impact of DIC, DRE, DRG, and DRL on SICLU, are shown in Columns (2)–(5). These results show that a 1% increase in DIC, DRE, DRG, and DRL triggered 0.202%, 0.489%, 0.103%, and 0.402% increases in SICLU, respectively. Therefore, Hypothesis 1 was confirmed.

5.3. Robustness Tests

Four robustness tests were conducted to ensure the credibility of the findings. First, referring to Hou et al. (2023) [40], we used ESI as an alternative indicator of the independent variable SICLU. Second, we applied a two-sided 5% tail reduction for the dependent variable to avoid any bias caused by abnormal values. Third, 15 pilot areas were removed from the sample counties to eliminate the impact of the pilot project for digital village on the regression results. Fourth, we re-estimated Equation (7) by censoring the control variables, PAE and OVP, to check whether the regression results were influenced by the selection of the control variables. As shown in

Table 4. Robustness test results.

Variable	Replace Independent Variable	Process Extreme Values	Remove Pilot Areas for Digital Village	Censor Some Control Variables
	(1)	(2)	(3)	(4)
DVC	0.064 ***	0.565 ***	0.603 ***	0.570 ***
	(15.25)	(5.71)	(6.59)	(5.72)
Controls	Yes	Yes	Yes	Yes
R2	0.2438	0.2816	0.3673	0.3370
N	358	322	343	358

Notes: The data in parentheses are robust standard error. *** represents the significance at 1%.

, the effect of DVC on SICLU remained significantly positive, demonstrating that the basic regression results are robust and that our main findings are credible.

5.4. Heterogeneity Analysis

Given that the impact of DVC on SICLU may be influenced by economic development level and resource endowments, we explored the relationship between DVC and SICLU in different regions. First, the study sample was subdivided into eastern, central, and western regions to conduct regional heterogeneity analysis. Columns (1)–(3) of

Table 5. Heterogeneity analysis results.

Variable	Eastern Region	Central Region	Western Region	Regions with Abundant Resource Endowments	Regions with Relatively Scarce Resource Endowments
	(1)	(2)	(3)	(4)	(5)
DVC	0.854 *	1.109 ***	0.251	0.676 ***	0.755 ***
	(2.07)	(7.41)	(1.36)	(4.16)	(5.10)
Controls	Yes	Yes	Yes	Yes	Yes
R2	0.8251	0.6371	0.1579	0.3555	0.4061
N	20	170	168	179	179

Notes: The data in parentheses are robust standard error. *** and * represent the significance at 1% and 10%, respectively.

report the estimated results. The coefficients of DVC in the eastern and central regions were both significantly positive, suggesting that DVC significantly improved SICLU in these regions. However, this positive effect was not significant in the western region, possibly because the lagging DVC led to its relatively low utility in increasing SICLU.

Second, differences created by the endowment of cultivated land resources may have an additional impact on the relationship between DVC and SICLU. Therefore, we divided 358 counties into two categories, as follows: counties with per capita cultivated land area ranking in the top 50% were classified as regions with abundant resource endowments, while the remaining counties were considered as regions with relatively scarce resource endowments. As shown in Columns (4) and (5), the coefficients of DVC in these two types of regions were significantly positive. These results indicate that DVC was conducive to improving SICLU regardless of whether it is in regions with abundant or relatively scarce CLU resources. However, the promotional effect of DVC on SICLU was greater in the latter regions.

5.5. Mediating Effect of DVC on SICLU

To further explore the mechanism by which DVC affects SICLU, we evaluated the mediating effects of FI, TI, and AgI (

Table 6. Estimation results of the mediating effect of DVC on SICLU.

Variable	(1)	(2)	(3)	(4)	(5)	(6)
	FI	SICLU	TI	SICLU	AgI	SICLU
DVC	0.097 ***	0.534 ***	6.112 ***	0.498 ***	0.652 ***	0.585 ***
	(6.87)	(5.47)	(14.80)	(7.41)	(8.25)	(10.10)
FI		0.378 *
		(1.93)
TI				0.028 ***
				(4.17)
AgI						0.133 ***
						(3.73)
Controls	Yes	Yes	Yes	Yes	Yes	Yes
R2	0.2216	0.3731	0.4915	0.3995	0.2497	0.3937
N	358	358	358	358	358	358

Notes: The data in parentheses are robust standard error. *** and * represent the significance at 1% and 10%, respectively.

). Columns (1) and (2) report the evaluation results of the mediating effect of FI. Column (1) indicates that DVC significantly increased FI. In Column (2), the significantly positive coefficients of DVC and FI jointly show that DVC improved SICLU through increasing FI. Therefore, Hypothesis 2 was verified. The test results of the mediation effect of TI are shown in Columns (3) and (4). All of the coefficients of DVC in Columns (3) and (4) and the coefficient of TI in Column (4) were significantly positive, indicating that TI is an important mediating variable in the relationship between DVC and SICLU. Thus, Hypothesis 3 was confirmed. Columns (5) and (6) present results using AgI as a mediating variable. In Column (5), the significantly positive coefficient of DVC suggests that DVC demonstrated a notable enhancement in AgI. Column (6) shows that DVC and FI jointly improved SICLU. Accordingly, FI mediated the effect of DVC on SICLU, validating Hypothesis 4.

6. Discussion

6.1. DVC Significantly Improves SICLU in Multiple Dimensions

With the advancement of DVC in China, it has become a vital driver for the green transformation of agriculture and the enhancement of agricultural productivity. Previous studies have explored the influence of rural digital transformation on green agricultural growth and cultivated land green use efficiency [44,47]. However, there is little evidence concerning the relationship between DVC and SICLU. This study examined the effects of DVC and its sub-indices (DIC, DRE, DRG, and DRL) on SICLU. The empirical results indicate that DVC, DIC, DRE, DRG, and DRL significantly improved SICLU. These findings are indirectly corroborated by the conclusions of relevant studies [44,47,49], which demonstrated that DVC and rural digital transformation increase agricultural green total factor productivity and cultivated land green use efficiency. Therefore, the positive effect of DVC on improving SICLU should not be ignored. Moreover, the impact of DVC on SICLU presented obvious heterogeneity. Specifically, the effect of DVC on SICLU was significantly positive in the eastern and central regions, as well as regions with abundant and relatively scarce resource endowments in China. However, this effect was not significant in the western region. The reason for this could be that the overall levels of DVC and digital agriculture in the eastern and central regions were comparatively high, which can promote progress in agricultural technology and precise inputs of agricultural production factors, thereby improving SICLU. Nevertheless, constrained by underdeveloped agricultural foundations, low productivity levels, and a shortage of skilled personnel, the adoption of digital technologies in the western region has progressed relatively slowly. This has resulted in an overall underdevelopment of both DVC and digital agriculture, consequently diminishing the promotive effect of DVC on SICLU. Nevertheless, due to the relatively low average level of DVC and digital agriculture in the western region, the promotion effect of DVC on SICLU is limited in this region. These findings are similar to the results of Du et al. (2023) [49], who demonstrated that DVC increases agricultural green total factor productivity in China’s eastern and central regions; however, such effect was not observed in the western region. Overall, digital village can stimulate the green transition of CLU and promote the sustainable development of agriculture, thereby helping to push forward rural vitalization. Our results provide empirical support for policymakers to stimulate DVC.

6.2. FI, TI, and AgI Strongly Mediate the Relationship Between DVC and SICLU

The results of the mediating effect analysis demonstrated that DVC improved SICLU through FI, TI, and AgI. Specifically, DVC remarkably increased FI, as corroborated by the results of Leng (2022) and Liu et al. (2023) [72,102]. Similar to the relevant literature [74,76], FI has been proven to be an important driver in promoting SICLU. Moreover, the estimation results showed that DVC significantly stimulated TI, which can be supported by existing studies [39,78]. We found that TI is conducive to improving SICLU, which is in line with the conclusions of Wang et al. (2021) and Zhang et al. (2023) [103,104]. In terms of AgI, the empirical results were confirmed by Du et al. (2023) [49], who found that DVC can promote its development. We found that AgI improved SICLU, aligning with the findings of Meng et al. (2024) and Ogutu et al. (2014) [31,91]. In summary, DVC not only increased FI, accelerated TI, and promoted AgI, but also indirectly improved SICLU through multiple pathways. These findings enrich our understanding of DVC’s role in achieving the green transformation and sustainable use of cultivated land in China.

6.3. Limitations and Future Research

There are some limitations in this study. First, the sample was limited to 358 counties due to a lack of sufficient data for the other counties in China. The sample size could be increased in future studies when relevant data are available. Second, due to the unavailability of panel data on DVC at the county-level scale, this study employed cross-sectional data for empirical analysis, thus preventing the examination of both short-term and long-term impacts of DVC on SICLU. If possible, future research should employ panel data to distinguish between short-run and long-run effects. Third, owing to data constraints, the overall level of regional TI was used as a proxy for agricultural TI. Subsequent research could consider collecting additional data through other channels, such as field research, to evaluate agricultural TI more accurately and comprehensively.

7. Conclusions and Policy Implications

7.1. Conclusions

This study evaluated the SICLU of 358 Chinese counties in 2020 using emergy analysis and further explored the effect of DVC on SICLU and its internal mechanisms using the OLS and mediating effect models. The following conclusions were obtained:

(1)

DVC had a significant promoting effect on SICLU in Chinese counties. This conclusion was still valid after a sequence of robustness tests. Moreover, all four secondary indicators of DVC (DIC, DRE, DRG, and DRL) can significantly improve SICLU. Therefore, the positive effect of DVC on SICLU should not be overlooked.

(2)

The effect of DVC on SICLU exhibited evident heterogeneity in different regions. This effect was significantly positive in the eastern and central regions but insignificant in the western region. The main reason may be that the western region’s lagging adoption of digital technologies—attributable to agricultural productivity gaps, technological constraints, and talent shortages—has hindered the development of DVC and digital agriculture, thereby limiting the promotional effect of DVC on SICLU. Additionally, the promotional effect of DVC on SICLU was significant in both regions with abundant and relatively scarce resource endowments; however, this effect is greater in the latter regions.

(3)

FI, TI, and AgI were important partial mediating variables through which DVC indirectly improved SICLU. This indicates that the intrinsic mechanisms underlying the relationship between DVC and SICLU are relatively complex. The systematic exploration of the transmission mechanisms of DVC affecting SICLU is conducive to fully leveraging the promotional effect of DVC on SICLU.

7.2. Policy Implications

In light of these discoveries, the following policy implications are proposed:

(1)

Given that DVC and its four sub-indices significantly improved SICLU, policymakers should promote DVC in multiple dimensions. More specifically, the government should increase its financial support to improve the digital infrastructure in rural areas, perfect relevant policies to improve the digital supply chain, and accelerate the development of rural e-commerce. Great efforts should be made to accelerate the empowerment of digital technology to modernize rural governance and raise relevant subsidies and provide more free training to enhance farmers’ digital skill levels. Moreover, scholars should conduct in-depth research to examine the limiting factors of DVC, which could provide empirical evidence for optimizing policy design and thereby enhance the effectiveness of DVC. As a result, DIC, DRE, DRG, DRL, and DVC will be promoted. This will facilitate the increase in agricultural productivity and the reduction in carbon emissions and agricultural non-point source pollution, thereby promoting the achievement of SDG 2 (Zero Hunger), SDG 12 (Responsible Consumption and Production), SDG 13 (Climate Action), and SDG 15 (Life on Land), and accelerating the improvement of SICLU.

(2)

Considering the heterogeneity characteristics of DVC, SICLU, and their relationship, China should establish a regionally differentiated development strategy to promote DVC and give full play to its role in improving SICLU. For eastern and central regions, it is necessary to summarize the successful experiences and deficiencies of DVC and design optimization strategies to promote it, thereby improving SICLU. The central and local government should accelerate the DVC of the western region and leverage its positive role in improving SICLU by improving related policy support system, accelerating the transfer and transformation of digital technology, and strengthening farmers’ digital literacy.

(3)

The mediating effect analysis showed that DVC can improve SICLU through increasing FI, accelerating TI, and promoting AgI. Based on this, we propose further accelerating DVC and tapping into its potential for increasing FI to enhance farmers’ motivation and ability to use green production factors and adopt new technology, thereby helping to improve SICLU. To further enhance the SICLU level, perfecting the agricultural technology innovation system and strengthening agricultural technology training to promote agricultural technology innovation and application should not be overlooked when progressing DVC. Moreover, policymakers should continuously improve digital rural facilities, prioritize supporting the development of AgI, and enrich and improve agricultural information services to accelerate the increase in SICLU.