The Impact of Rural Population Shrinkage on Rural Functions—A Case Study of Northeast China

Abstract

As industrial and urban growth advances, the challenge of rural population shrinkage has grown more pronounced, impacting rural functions. Northeast China is an example in this study, and a rural function evaluation index system is constructed based on four dimensions: agricultural production, economic development, social security, and ecological conservation. The spatio-temporal heterogeneity of the impact of rural population shrinkage on rural functions is quantified in this study using bivariate spatial autocorrelation and geographically and temporally weighted regression (GTWR). The results show that from 2000 to 2020, the rural population in most counties in Northeast China declined, while agricultural production, economic development, social security, and ecological conservation functions generally trended upwards. According to the GTWR model, the positive effect of rural population density on agricultural production weakened over time, slightly promoting social security and continuing to inhibit ecological conservation. In contrast, the supporting effect of average rural population size on economic development strengthened, its inhibitory effect on ecology decreased, and it slightly inhibited social security. While rural population shrinkage generally promoted agricultural development, economic growth, social security, and ecological improvements, its positive impact on agricultural development declined over time, and the promotion effects on social security and ecological conservation partially turned into inhibition after 2020. Policy recommendations are presented in this paper, providing a solid scientific foundation for the sustainable development of rural areas in Northeast China.

1. Introduction

As global urbanization accelerates, the decline in rural populations has emerged as a widespread trend across nations [1]. This phenomenon is not confined to the United States [2] or Europe [3]; it also affects Asian countries, particularly in East Asia that are experiencing severe rural population shrinkage caused by rural-to-urban migration and an imbalanced age structure [4]. It leads to multiple socioeconomic issues. Regional imbalances exacerbate the rural–urban divide [5]. Additionally, it threatens regional food security. Rural areas also encounter challenges such as village depopulation, land abandonment, and population aging [6]. However, rural population shrinkage also brings about specific positive effects, such as the redistribution of agricultural labor. As people migrate away from the countryside, abandoned fields and vacant homes create opportunities for large-scale agribusiness [7]. At the same time, the resulting expansion of available farmland per capita and advances in farming methods go hand in hand with these shifts [8]. Additionally, fewer residents ease the strain on local ecosystems, allowing natural vegetation to rebound more effectively [9]. The decline in population significantly affects rural demographics, environmental management, social frameworks, and community sustainability.

The theory of multifunctional countryside originates from the theory of multifunctional agriculture. The European Union (EU) first formally recognized the multifunctionality of agriculture in 1988, emphasizing that agriculture not only holds economic value but also plays a significant role in environmental governance, social structure, and community viability [10]. Holmes further developed the theory of multifunctional rural transformation, noting that changes in human demands for rural production, living, and ecological functions during the process of social development drive the evolution of rural areas [11]. The connotation of rural multifunctionality refers to the beneficial effects on nature or humanity generated by a specific rural system within a broader regional space, through the exertion of its own attributes and interactions with other systems, at a certain stage of social development [12]. Many researchers believe that pursuing rural multifunctionality effectively means overcoming social regression and maintaining the stability of rural regional systems [13]. Understanding how rural population shrinkage affects rural functions is essential. Currently, studies on rural populations and rural functions primarily concentrate on two aspects: First, in the context of population shrinkage, the changes in individual rural functions or the relationships between multiple rural functions and their influencing factors are studied. For example, variations in rural population greatly influence arable land utilization effectiveness [14], with rural depopulation also facilitating the enhancement of the rural ecological environment [15]; the outmigration of young people from Moldova has reduced the rural labor force and hindered local revitalization efforts [16], and at the same time, remittances have greatly contributed to poverty reduction and improved living conditions, especially in rural areas [17]; in Mediterranean rural areas with population shrinkage, vegetation restoration has been observed in semi-natural highlands, while an increase in food production in agricultural lowlands has been noted [18]; and digitalization impacts the welfare of residents in Finland’s declining rural regions [19]. Li et al. examined the functional trade-offs and synergies within the black soil counties of Northeast China amid population decline [20]. A geographic detector was used to reveal the dominant factors in the decline in rural multifunctionality in areas with population decline in Heilongjiang Province [21]. Second, based on the relationship between people and land [22,23] or the “population–land–industry” framework [24,25], population is incorporated as a factor in the study of rural functions. Qu et al. explored the complex effects of rural population shrinkage and land expansion on the human–land system [26]. Dong et al. demonstrated that rural settlement expansion persisted despite population decline, resulting in significant land-use inefficiency [27]. Based on an empirical analysis of Pinggu District in Beijing, Tian et al. studied the differentiated influence exerted by how rural population, land resources, and industrial elements are distributed on living, production, and ecological functions from an efficiency perspective [13]. To date, most studies have focused on single functions [18] or static analyses, employing methods such as coupling-type measurements [28], spatial autocorrelation [29], and geographically weighted regression [30]. However, a systematic exploration of the dynamic interplay between population decline and rural multifunctionality remains limited. Furthermore, existing research predominantly relies on qualitative discussions [20] or analyses incorporating additional factors [31], with few studies directly quantifying the spatiotemporal relationships between rural population shrinkage and diverse rural functions.

Following the reform and opening-up policies, China experienced rapid growth in industry and urbanization [32]. Policy incentives and natural demographic shifts inevitably contribute to the concentration of resources and population in major urban centers [33]. China’s rural population began to decline in 1995 [34], with a continuous downward trend from 2000 to 2020, accelerating between 2010 and 2020, with Northeast China being the most brutally hit [35]. Northeast China is a resource-rich, old industrial base. With the Chinese economy transitioning into a phase of new normalcy, the region’s industrial competitive advantages have gradually eroded, making contraction an inevitable phase in its socioeconomic development [36]. Northeast China is rich in black soil resources. It is China’s most significant area for commercial grain production and natural forests, which are crucial in ensuring food and ecological security. However, studies indicate that the recent intensification of land degradation in Northeast China has severely disrupted the synergy between agricultural production and environmental protection [37]. As one of the regions experiencing the most severe rural population shrinkage in China, Northeast China faces critical challenges related to rural multifunctionality and sustainable development. To address this challenge, bivariate spatial autocorrelation analysis can be employed to reveal the spatial dependency between rural population and functionality. Building on this, geographically and temporally weighted regression (GTWR) captures the impact of population variables on functional changes and their spatiotemporal heterogeneity. Moreover, by disaggregating rural population data and extracting relative population change values, the influence of varying population baselines is mitigated, thereby addressing the shortcomings of existing studies in analyzing the interaction between rural population dynamics and multifunctionality.

Therefore, this study uses Northeast China as a case study, develops a rural function evaluation index system, applies geographically and temporally weighted regression to examine the impact of population on rural functions, and proposes corresponding policy recommendations, providing a solid scientific basis for the sustainable development of rural areas in Northeast China. Specifically, this study aims to (1) identify the trend of rural population shrinkage in Northeast China and its spatio-temporal changes; (2) develop a rural function evaluation index system and quantify the characteristics of rural function evolution; and (3) examine the impact of rural population changes on rural functions.

2. Theoretical Framework on the Impact of Rural Population Shrinkage on Rural Functions

2.1. The Impact Path of Rural Population Decline on Rural Functions

The persistent decline in rural populations fundamentally reshapes land use patterns, which, in turn, critically influences the evolution of rural functions. This complex process is driven by a confluence of interwoven factors, including industrial restructuring, infrastructure development, natural environmental dynamics, shifts in management systems, technological advancements, and policy initiatives [38]. The impacts of these changes are not uniform; rather, they manifest through diverse pathways, leading to spatially differentiated outcomes across regions, contingent upon their unique natural endowments and prevailing socio-economic conditions.

The interaction between the rural population and agriculture is inherently spatially embedded, meaning that population shrinkage engenders distinct agricultural development trajectories in areas with varying topographical and demographic characteristics. For instance, agricultural plains typically exhibit a linear development model, characterized by “labor reduction–mechanization substitution–efficiency enhancement.” Conversely, mountainous regions show a nonlinear pattern of “labor exodus-land abandonment-productivity decline” [39]. Therefore, in fertile and accessible lowland areas, agriculture has been strengthened, enhancing food production services. In contrast, in mountainous regions have experienced land abandonment and ecological restoration, which have inadvertently enhanced crucial ecosystem services such as water regulation and soil retention capacity [18]. In addition, implementing the Household Responsibility System has undoubtedly increased agricultural productivity. However, over time, the pervasive fragmentation of farmland into smaller units has inadvertently impeded the pace of agricultural modernization. Given China’s constrained land resources, substantial improvements in agricultural labor productivity are unlikely to occur until population out-migration leads to a greater per capita availability of arable land [40]. The “tripartite entitlement system” reform divides land ownership into three types: full land ownership, non-transferable land contract rights, and transferable land management rights [41]. This reform allows farmers to transfer their land management rights to other farmers, agricultural enterprises, and village cooperatives, significantly preventing land abandonment, increasing rural household income, and promoting large-scale agricultural operations [42,43].

The shrinkage of rural populations poses numerous challenges to rural economic development. These include pervasive labor shortages and a reduction in local consumption potential, which collectively diminish the attractiveness of the rural capital environment [44]. The aging and limited education of rural population has exacerbated rural disadvantages [45]. Under the combined influence of these factors, rural areas have inevitably fallen into an economic recession marked by the contraction of local markets and the closure of small businesses [46]. Nevertheless, rural out-migration has also yielded certain positive transformations in rural development. It has fostered the dissemination of agricultural technology, improved labor quality, and increased income opportunities, as departing farmers often gain access to better education and adopt more advanced concepts. Wealthier farmers are more likely to invest in education and intellectual resources, which enhances their quality of life and creates a multiplier effect, thereby further advancing the agricultural economy [47]. However, because labor productivity tends to be high among the out-migrant population, labor mobility may result in a shortage of rural development leaders and exacerbate the urban-rural income gap [8], hindering urban-rural integration.

Rural population shrinkage significantly undermines the sustainability and effectiveness of rural social security functions by exacerbating service accessibility difficulties and weakening economies of scale and governance capabilities [48]. A two-way causal relationship exists between the absence of essential services such as education, finance, healthcare, and broadband coverage and the outflow of the rural population. Inadequate public services are both a driving factor and a long-term consequence of population decline [49]. Poor infrastructure maintenance and lack of community vitality in rural areas have resulted in a downward spiral in rural life quality [46]. While it is generally assumed that areas experiencing significant population shrinkage will also suffer a decline in social security functions, empirical studies suggest that rural infrastructure decline is not always directly correlated with population decline [50]. Merely guaranteeing public services in areas projected to shrink is not, therefore, a sufficient strategy for enhancing rural infrastructure effectiveness [51], as numerous other factors influence the trend.

The reduction in rural population density frequently leads to a decrease in the expansion of construction zones, a cessation of certain agricultural practices, an increase in fallow land, and a slowdown in livestock farming, all contributing to reduced land use intensity. This promotes the self-repair of natural ecosystems and secondary forest succession [52,53], positively influencing vegetation greenness changes [9]. Meanwhile, reduced human activities decrease the consumption of natural resources and result in environmental pollution, contributing to biodiversity conservation [47] and, when combined with ecological conservation projects, enhancing ecological surplus [15]. However, natural recovery resulting from the cessation of agricultural activities may be accompanied by issues such as simplified habitat structures, homogenization of species communities, and loss of ecological connectivity [54], potentially increasing the vulnerability of rural ecosystems.

2.2. Interactions Between Changes in Rural Functions

Against the backdrop of rural population shrinkage, the various rural functions—namely, production, living, and ecological functions—are not evolving in isolation but rather through intricate interactions, interdependencies, and a delicate balance of synergies and trade-offs, as shown in Figure 1. As the foundational function of rural areas, changes in agricultural production inherently drive the development and reconfiguration of other functions. For instance, enhancing livestock and farming yields can bolster economic expansion and boost local service and job opportunities under favorable external conditions like climate and market rates; however, it may also make rural areas more vulnerable to climate or socioeconomic disturbances [55]. Excessive use of chemical pesticides, fertilizers, and plastic mulch in farming, along with improper irrigation, leads to soil erosion and other problems that damage the ecological environment. As non-agricultural industries develop, environmental land is increasingly taken up by non-agricultural construction. Predominantly mountainous areas with significant ecological functions suffer from poor geographical conditions, weak industrial infrastructure, inadequate public services, and limited income sources for farmers, which hinder large-scale population settlement. Regions with high population density, intensive land development, and reduced biodiversity exhibit diminished ecological resilience. This creates a trade-off between ecological and living functions [56]. However, ecological conservation functions do not always directly conflict with other functions. For example, studies have shown that afforestation, as a land use practice, can improve soil fertility, provide wood and non-wood products, increase income and self-sufficiency for small farmers, and offer ecological and socio-economic benefits, provided it is carefully designed and implemented [57]. Ultimately, the changes in rural functions interact in a profoundly complex and intertwined manner. Therefore, when assessing the multifaceted effects of rural population shrinkage on these functions, it is imperative to consider the inherent trade-offs and potential synergies among various functions, which is critical for developing sustainable rural development strategies.

3. Materials and Methods

3.1. Overview of the Study Area and Data Sources

3.1.1. Overview of the Study Area

The geographic extent of Northeast China spans from 115°32′ to 135°09′ E and from 38°42′ to 53°35′ N, and includes the provinces of Jilin, Liaoning, and Heilongjiang, along with four eastern prefecture-level cities in Inner Mongolia—Hulunbeier, Xing’an League, Tongliao, and Chifeng. Based on terrain, vegetation, and climate, Northeast China is divided into six geographical sub-regions: Changbai Mountain Region (CBMR), Lesser Kinggan Mountains Region (LKMR), Greater Kinggan Mountains Region (GKMR), Sanjiang Plain (SJP), Songnen Plain (SNP),and Liao River Plain (LRP) [37,58]. Northeast China was once a resource-rich industrial base that underwent rapid industrialization and urbanization in the 20th century. However, it has experienced a significant population decline in recent years due to aging, out-migration, and economic stagnation [59].

As the fundamental administrative unit in China, counties have become a key focus for promoting rural sustainable development [60]. They encompass many rural areas and populations, serving as the earliest, fastest, most extensive, and largest regional units affected by population shrinkage in China. This shrinkage’s social, economic, and ecological impacts are also more pronounced and profound [61,62]. Therefore, this study uses counties as the evaluation units, excluding regions with clear urbanization based on data availability, as shown in Figure 2, and includes 169 counties as the research subjects.

3.1.2. Data Sources

The main datasets adopted in this research encompass land use data, county-level socioeconomic information, and demographic statistics for 169 counties in Northeast China across the years 2000, 2010, and 2020. Spatial data, such as county-level administrative boundaries and land use, are primarily sourced from the Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences (https://www.resdc.cn, accessed on 24 July 2024). The classification standards from the Chinese Academy of Sciences serve as the foundation for the land use data and have been employed to estimate ecosystem service values. The rural population refers to the permanent residents living in rural areas [63]. County-level rural population data in this study are derived from the “rural population” field in the Fifth, Sixth, and Seventh National Population Censuses of the People’s Republic of China. County-level socioeconomic data are primarily obtained from the China County Statistical Yearbook, with missing data supplemented by statistical yearbooks of the corresponding counties (cities, districts) for the relevant years. The details are as follows: (a) grain production and total power of agricultural machinery; (b) gross domestic product, value added by the primary industry, value added by the secondary industry, and fiscal revenue; (c) number of students enrolled in ordinary secondary schools, primary schools, hospitals, and number of beds in health centers.

3.2. Research Methods

3.2.1. Identification of the Extent and Types of Rural Population Shrinkage

In order to more accurately classify the types of rural population change in Northeast China, along with their spatiotemporal variation, the change rate of the rural population [64] was calculated using the following Formula (1):

$$T_i={\displaystyle \frac{R{P}_{t_2}-R{P}_{t_1}}{R{P}_{t_1}}}\times 100\%$$

(1)

In the formula, $T_i$ represents the rural population change rate of the i-th county. In contrast, $R{P}_{t_1}$ and $R{P}_{t_2}$ denote the rural population of the i-th county at the initial and final states, respectively.

Based on existing research [35], the types of changes in the rural population level are classified into five categories: (I) population growth; (II) accelerated decline; (III) decelerated decline; (IV) overall decline, first increasing then decreasing; (V) overall decline, first decreasing then increasing.

3.2.2. Construction of Rural Function Evaluation Index System and Weight Calculation

The multifunctionality of rural areas arises from their diverse attributes. Rural areas are where rural residents engage in production and other activities, providing both urban and rural residents with agricultural products, social security, local culture, ecological services, and other resources. Drawing on the established framework of rural functions in Northeast China [20], this study classifies explicitly the multifunctionality of rural areas into four dimensions: agricultural production function (APF), economic development function (EDF), social security function (SSF), and ecological conservation function (ECF). As shown in

Table 1. Rural function evaluation index system in Northeast China.

Functions	Indexes/Unit	Effect	Index Connotation and Calculation Method	Weight
Agricultural production function (APF)	Land reclamation rate/%	+	Farmland area/total land area	0.297
	Per unit area grain yield/(t/ha)	+	Total grain output/farmland area	0.254
	Agricultural mechanization degree/(kW·h/ha)	+	Total power of agricultural machinery/farmland area	0.143
	Paddy field ratio/%	+	Paddy field area/total cultivated land area	0.306
Economic development function (EDF)	Grain commodity rate/%	+	(Grain output − 400 kg × permanent population)/total grain output	0.581
	Per unit area agricultural output/(104 yuan RMB/ha)	+	The output value of the primary industry/total land area	0.091
	Economic structure/%	+	The sum of the output value of secondary and tertiary industries/GDP	0.161
	Financial contribution rate/%	+	Financial revenue/GDP	0.167
Social security function (SSF)	Healthcare/per 104 person	+	Number of beds in health institutions owned by 10,000 people	0.137
	Road traffic density/(km/ha)	+	Road mileage/total land area	0.45
	Per capita housing area of rural residents/ha	+	Rural Settlement Area/number of the permanent population	0.166
	Per capita farmland area/ha	+	Farmland area/number of the permanent population	0.247
Ecological Conservation function (ECF)	Climate regulation function/(yuan/ha)	+	Per unit area ecological service, the equivalence factor method of ecological service value	0.228
	Environmental purification function/(yuan/ha)	+	Per unit area ecological service, the equivalence factor method of ecological service value	0.209
	Maintaining nutrient cycling function/(yuan/ha)	+		0.243
	Biodiversity function/(yuan/ha)	+		0.126
	Landscape aesthetics function/(yuan/ha)	+		0.194

, a rural function evaluation index system is subsequently constructed. Although there is some overlap between the agricultural production, economic development, and social security functions in rural areas, they must be assessed separately due to the specific context of rural China. Agricultural production involves not only economic output but also the direct relation to national food security, the livelihood security of rural residents, and the continuation of rural culture. The Rural Comprehensive Revitalization Plan (2024–2027) emphasizes that food security is a core task of rural development. Furthermore, social security assumes particular importance in rural China. Under the collective land ownership system, the right to use house sites and the right to contracted land management are central to the housing, livelihoods, and retirement security of farmers. Therefore, the separation of agricultural production, economic development, and social security functions allows for a more accurate assessment of the multifunctionality of rural areas.

As a central grain-producing region in China, Northeast China possesses abundant black soil resources. Agricultural production functions are key to regional rural development, ensuring food security and agricultural product production. This study uses land reclamation rate, per unit area grain yield, agricultural mechanization degree, and proportion of paddy fields as indicators of agricultural production functions. The paddy field ratio was selected based on the characteristics of Northeast China as the core rice-producing region. The “Conversion of Dryland to Paddy Fields” (CDPF) represents a significant policy-driven initiative in Northeast China aimed at enhancing agricultural production capacity and ensuring national food security [65]. Consequently, this indicator effectively reflects the intensity of regional agricultural production functions and policy orientation. The rural economy relies on the primary industry, influenced by economic structure and fiscal contribution rate. For economic development functions, grain commodity rate, per unit area agricultural output, economic structure, and fiscal contribution rate are selected as indicators. The calculation of the grain commodity rate is based on existing studies [66,67]. This indicator reflects the portion of grain output that is available for market transactions, reflecting the degree of commercialization of agricultural production and the level of marketization of the rural economy. The social security function provides basic needs and services such as healthcare, employment, education, water and electricity facilities, transportation, and other social services. Indicators selected include healthcare, road traffic density, per capita housing area of rural residents, and per capita farmland area. The per capita farmland area in rural China is not only an agricultural production resource but also plays a role in supporting farmers’ livelihood security and social stability. The right to the contracted management of land provides farmers with a safety net, aligning with the essence of the social security function [68,69]. Northeast China’s rich ecosystems provide significant ecological services, including climate regulation, environmental purification, maintaining nutrient cycling, biodiversity, and landscape aesthetics, which form the ecological conservation function.

The Criteria Importance Through Intercriteria Correlation (CRITIC) method is an objective weighting approach based on data volatility [70]. This research used the CRITIC method to calculate indicator weights for all four rural function categories. In order to address the dimensional discrepancies among indicators, the extreme value standardization method was applied to normalize the indicators, and the weights for each rural function indicator were subsequently determined. Using the standardized indicator values and weights, the rural function values for each county can be calculated using Formula (2):

$$R{F}_{ik}={\displaystyle \sum X_{ij}}\times w_j$$

(2)

In the formula, $R{F}_{ik}$ represents the value of the k-th single function for the i-th county; $X_{ij}$ is the range standardized value of the j-th indicator for the i-th county in Table 1, and $w_j$ is the weight of the j-th indicator calculated using the CRITIC weighting method.

3.2.3. Bivariate Spatial Autocorrelation

Spatial autocorrelation serves as an essential statistical tool in geography, used to examine how variables are interconnected across space. It can be broken down into two categories: global and local measures of spatial autocorrelation [71]. The formula for global Moran’s I is as follows:

$$I={\displaystyle \frac{{\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j\ne i}^n{W}_{ij}}}\cdot (Y_i-\overline{Y})\cdot (Y_j-\overline{Y})}{S^2\cdot {\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j\ne i}^n{W}_{ij}}}}}$$

(3)

The calculation for local Moran’s I is as follows:

$$I_i={\displaystyle \frac{Y_i-\overline{Y}}{S_i^2}}\cdot {\displaystyle \sum _{\begin{array}{c}j=1\\ j\ne i\end{array}}^n{W}_{ij}}\cdot (Y_j-\overline{Y})$$

(4)

In the formula, $S^2={\displaystyle \frac{1}{n}}{\displaystyle \sum _{i=1}^n(Y_i-\overline{Y})}$; $\overline{Y}={\displaystyle \frac{1}{n}}{\displaystyle \sum _{i=1}^n{Y}_i}$; $Yi$ and $Yj$ represent the attribute values of unit $i$ and unit $j$, respectively; $n$ is the number of spatial units; and $Wij$ is the weight matrix based on spatial adjacency relationships.

To more comprehensively characterize the spatial correlation between multiple variables, a bivariate global and local spatial autocorrelation method based on Moran’s I index was developed, offering an effective way to analyze the relationship between the spatial distribution of different factors [72]. In this study, GeoDa software was used to calculate the bivariate spatial autocorrelation. This method was applied to examine the spatial interaction between the rural population and various rural functions in Northeast China. The formula is:

$$I_{lm}^p=z_l^p\cdot {\displaystyle \sum _{q=1}^n{W}_{pq}}\cdot z_m^q$$

(5)

In the formula, $z_l^p={\displaystyle \frac{X_I^p-{\overline{X}}_I}{{\sigma }_l}}$; $z_m^q={\displaystyle \frac{X_m^q-{\overline{X}}_m}{{\sigma }_m}}\theta$; $X_l^p$ is the attribute value of spatial unit $p$ for attribute $l$; $X_m^q$ is the attribute value of spatial unit $q$ for attribute $m$; ${\overline{X}}_l$ and ${\overline{X}}_m$ are the mean values of attributes $l$ and $m$, respectively; and ${\sigma }_1$ and ${\sigma }_m$ are the variances of attributes $l$ and $m$, respectively.

3.2.4. Geographically and Temporally Weighted Regression

Using panel data, a geographically and temporally weighted regression (GTWR) model is employed to analyze the spatial relationship between population shrinkage and the evolution of rural functions. To capture the influence of changes in rural population and to reduce the interference caused by inter-county disparities in population scale, the rural population ($\mathrm{RP}$) is divided, selecting three demographic variables—rural population density ($\mathrm{RPD}$), average rural population ($\overline{\mathrm{RP}}$), and relative change in rural population ($\mathrm{RPc}$)—which were incorporated into the GTWR model as explanatory variables. RPc captures how much the rural population in a county has changed relative to its average over time. Their definitions are presented below:

$$RP{D}_i={\displaystyle \frac{R{P}_{it}}{A_i}}$$

(6)

$${\overline{RP}}_i={\displaystyle \frac{1}{n}}{\displaystyle \sum _{t=1}^{n}R}{P}_{it}$$

(7)

$$RP{c}_{it}=R{P}_{it}-{\overline{RP}}_i$$

(8)

In the formula, $RP{D}_i$ represents the rural population density of the $i$-th county, $R{P}_{it}$ represents the rural population of thee $i$-th county in year $t$, $A_i$ represents the rural area of the $i$-th county, ${\overline{RP}}_i$ represents the average rural population of the $i$-th county, and $n$ represents the number of years. In this study, $n$ = 3, and $RP{c}_{it}$ is the rural population of $i$-th county in year $t$.

The GTWR model integrates spatio-temporal characteristics into the basic regression framework, accurately explaining spatial phenomena, reflecting geographical heterogeneity, and accounting for the potential effects of temporal changes, thus making it better suited to research needs [73]. In this study, GTWR was implemented using ArcGIS 10.8. Three population variables served as explanatory variables, and four types of rural functions were used as dependent variables. To address the issues of skewed distribution and heteroscedasticity and improve model interpretability, all variables were log-transformed, and the model’s bandwidth was calculated using the second-order Akaike Information Criterion (AICc) method. The specific formula is as follows:

$$Y_i={\beta }_0(u_i,v_i,t_i)+{\displaystyle \sum _{k=1}^K{\beta }_k}(u_i,v_i,t_i)X_{ik}+{ϵ}_i$$

(9)

$$\widehat{\beta }(u_i,v_i,t_i)={\left[X^{T}W(u_i,v_i,t_i)X\right]}^{-1}{X}^{T}W(u_i,v_i,t_i)Y$$

(10)

$$w_{ij}=\exp \left(-{\left({\displaystyle \frac{d_{ij}}{h}}\right)}^2\right)$$

(11)

In the formula, $Y_i$ represents the dependent variable of the $i$-th observation point, $X_{ik}$ is the explanatory variable of the $i$-th observation point, and ${\beta }_0(u_i,v_i,t_i)$ is the intercept term. ${\beta }_k(u_i,v_i,t_i)$ represents the coefficient of the explanatory variable at the spatial location $(u_i,v_i)$ and time $t_i$, and $W(u_i,v_i,t_i)$ is the spatial weight matrix. The spatial weight $w_{ij}$ indicates the similarity between observation points $i$ and $j$. $d_{ij}$ represents the Euclidean distance between observation points $i$ and $j$, and $h$ refers to the decay parameter that governs the inverse relationship between spatial distance and the spatial weight function.

4. Results

4.1. Spatial and Temporal Patterns of Population Decline and Rural Functions in Northeast China

4.1.1. Spatial and Temporal Patterns of Rural Population Shrinkage

The rural population distribution in Northeast China shows clear temporal and spatial patterns. Figure 3a–c illustrate a stepwise decline in population from the center to the periphery, with lower populations in the GKMR, LKMR, CBMR, and SJP areas compared to the SNP and LRP areas. Between 2000 and 2020, the population distribution further concentrated and centralized, with significant agglomeration near urban areas, likely due to natural environmental factors [63]. Rural development relies on cities, and the concentration of rural populations around urban centers is key to urban-rural integration. The rural population is more concentrated in counties surrounding major cities like Harbin, Shenyang, and Changchun. Rural areas, distant from urban centers and characterized by weak connections to them, are most vulnerable to population decline and marginalization [74]. Plain regions, with higher agricultural productivity and economic development, support larger rural populations, resulting in higher population density than the mountainous areas. The population decline has progressively extended eastward, shifting from western regions to the eastern plains, with increased intensity, although the overall rural population distribution in North-east China remains unchanged.

As illustrated in Figure 3d,e, the rural population in most counties slightly shrank between 2000 and 2010, while some counties continued to experience population growth. However, between 2010 and 2020, the decline in the rural population intensified, with most counties experiencing moderate to severe shrinkage, while only a few counties maintained growth. As illustrated in Figure 3f, from 2000 to 2020, only 1.18% of the rural population grew, while 98.82% shrank—72.79% of which experienced acceleration, 5.92% saw a deceleration in decline, and 18.92% of counties experienced a total decline, initially increasing and then decreasing, while 1.18% of counties saw a total decline, initially decreasing and then increasing.

4.1.2. Spatial and Temporal Patterns of Rural Functions Evolution

Figure 4 illustrates the spatio-temporal evolution of rural functions in Northeast China for 2000, 2010, and 2020, along with functional changes during this period. Spatially, the agricultural production function in rural areas shows clear differentiation, with a high center and low periphery, consistent with the region’s terrain. High-value agrarian production areas are concentrated in SJP, SNP, and LRP, characterized by flat terrain, fertile soil, irrigation, and abundant arable land. These regions are key agricultural hubs in Northeast China. Low-value areas, such as GKMR, LKMR, and CBMR, are predominantly mountainous, with extensive forests and terrain unsuitable for crop cultivation, resulting in poor agricultural conditions. Over time, agricultural production in the plains has intensified and expanded outward.

From 2000 to 2020, the economic development of each county unit changed significantly, with notable increases in SJP, SNP, LRP, and the southeastern region of GKMR. Counties with higher economic development levels are concentrated around major cities in Northeast China, such as Shenyang, Dalian, Harbin, and Changchun. As urbanization progresses, economic activities expand outward, with counties surrounding major cities developing more advanced infrastructure and forming economically more prosperous regions.

The high-value areas of the social security function are mainly concentrated in the southeastern part of the LRP and the SJP. These regions feature flat terrain and well-developed infrastructure, healthcare, social services, and transportation networks. In contrast, mountainous areas such as the GKMR and the CBMR exhibit lower SSF values. In particular, the overall social security function in Inner Mongolia is lower than that in other provinces. Driven by urbanization, which has spurred development in rural areas, the majority of counties experienced an upward trend in SSF from 2000 to 2010, with the exception of a few counties that initially had higher SSF values in 2000 and subsequently saw a decline. From 2010 to 2020, the focus of SSF growth shifted back to the southeastern region, while some counties in the central and northern areas experienced a slight decline in SSF due to the impacts of population changes and terrain constraints.

The ecological conservation function in Northeast China exhibits significant regional differences, contrasting with the spatial distribution of agricultural production functions. Northeast China’s ecological conservation function was predominantly weak in 2000, but it increased significantly from 2000 to 2010 before declining from 2010 to 2020. High-value areas are primarily in mountainous and forest-covered regions, such as GKMR, LKMR, and CBMR. These regions are high in elevation, have extensive forest coverage, and possess strong ecological protection and conservation functions. In contrast, low-value areas are predominantly found in the southwest plains. Ecological conservation function distribution contrasts with agricultural production, reflecting balance and contradictions in resource utilization in Northeast China.

4.2. Spatial Correlation Between Rural Population and Rural Functions

Variable spatial autocorrelation analysis was used to investigate the spatial correlation between the rural population and rural functions in counties from 2000 to 2020 and their changes over time. As shown in

Table 2. Bivariate global spatial autocorrelation of rural population and rural functions.

	APF			EDF			SSF			ECF
Year	2000	2010	2020	2000	2010	2020	2000	2010	2020	2000	2010	2020
Moran’s I	0.495	0.486	0.445	0.122	0.205	0.153	0.122	0.180	0.232	−0.341	−0.221	−0.365
Z score	10.800	10.117	9.858	3.131	4.824	3.844	3.242	4.766	5.931	−7.750	−5.206	−8.589
p-value	0.001	0.001	0.001	0.002	0.001	0.001	0.002	0.001	0.001	0.001	0.001	0.001

, the global Moran’s I index results indicate significant positive spatial autocorrelation between the rural population and APF, with the Moran’s I value slightly decreasing from 0.495 to 0.445 in 2020. This correlation continued to weaken over time, with Z values greater than 9.8. However, its positive spatial correlation with economic development functions increased and decreased. Moran’s I value rose from 0.122 to 0.205 in 2010 before falling to 0.153, accompanied by a significant Z value. For SSF, the positive correlation with rural population is gradually increasing, with Moran’s I rising from 0.122 to 0.232. In contrast to APF, EDF, and SSF, the agricultural population showed a significant negative correlation with ECF, with this correlation first decreasing and then increasing to −0.341, −0.221, and −0.365, respectively.

As shown in Figure 5, for APF, high-high agglomeration was primarily distributed in plain areas, accounting for 17.26% in 2000, rising to 22.62% in 2010, and decreasing to 19.64% in 2020. During this period, its distribution gradually shifted northward, concentrating in SNP and SJP, primarily located in Heilongjiang and Jilin provinces. Low–low agglomeration has consistently concentrated in the northern part of the GKMR, increasing from 9.52% in 2000 to 13.69% in 2010 and decreasing to 11.31% in 2020. The initial increase followed by a decrease in the proportion of high–high and low–low agglomeration indicates that, as time passed, the region’s agricultural production became less reliant on intensive labor. Low–high agglomeration gradually emerged on the edges of SJP, representing a potential area with a small rural population but abundant agricultural resources and efficient agricultural production. The proportion of high–low counties decreased from 5.36% to 3.57%, and the number of counties with dense populations but weak agricultural functions also declined.

For EDF, high–high agglomeration is concentrated in the economically active SNP and SJP regions. In 2000, its proportion was 20.83%, rising significantly to 26.79% in 2010. Simultaneously, the proportion of low–high agglomeration increased. By 2020, some high–high agglomeration transformed into low–high agglomeration and its proportion decreased to 19.05%. Low–low agglomeration remained steadily concentrated in the northwestern GKMR margin and the southeastern CBMR-Liaodong Hills peripheral area, with changes in its proportion consistent with those of the high–high type. The proportion of high–low agglomeration continued to decline, primarily concentrated in the southern marginal area.

The spatial association of SSF with rural populations shows a relatively stable and concentrated trend. High-high agglomeration and high-low agglomeration have a large proportion, and their numbers first decreased and then increased over time. High-high agglomeration gradually concentrated in Liaoning Province, while high-low agglomeration is mainly distributed at the junction of SNP, GKMR, and LKMR. Low-low agglomeration is located in the northern part of the GKMR, with its proportion increasing from 19.35% to 36.17% and then decreasing to 19.17%. The proportion of low-high agglomeration is minimal.

ECF and the rural population exhibited a clear and relatively stable negative correlation pattern. High–low agglomeration accounted for the most significant proportion, first declining and then increasing, reaching 22.02% in 2000, declining to 19.64% in 2010, and rising again to 22.62% in 2020. High–high and low–high agglomeration are significantly concentrated in the ecological barrier areas of the northeast and east, including the northern GKMR, LKMR, and CBMR, which are mountainous and forested areas with relatively low population density. In contrast, low–low and high–low agglomeration are primarily distributed in the central and southern regions, such as the south of GKMR, SJP, and LRP. These areas are predominantly agricultural production zones or regions with high development intensity, and their ecological conservation functions are generally weak.

4.3. GTWR Results Analysis

The applicability of the models was assessed by comparing the adjusted determination coefficient (R²) and AICc values of the ordinary least squares (OLS), geographically weighted regression (GWR), time-weighted regression (TWR), and geographically and temporally weighted regression (GTWR) models. As shown in

Table 3. Comparison of OLS, GWR, TWR, and GTWR model results.

	OLS				TWR
	APF	EDF	SSF	ECF	APF	EDF	SSF	ECF
AICc	−177.805	−61.329	−1081.633	366.149	−252.538	−114.536	−1171.370	9.348
R2	0.507	0.161	0.200	0.241	0.591	0.273	0.356	0.640
R2 Adjusted					0.589	0.268	0.352	0.637
	GWR				GTWR
	APF	EDF	SSF	ECF	APF	EDF	SSF	ECF
AICc	−532.480	−287.452	−1331.130	153.944	−555.980	−336.568	−1381.360	−278.576
R2	0.795	0.520	0.579	0.561	0.808	0.601	0.688	0.854
R2 Adjusted	0.794	0.518	0.576	0.558	0.806	0.598	0.686	0.853

, the GTWR model exhibits a higher R² value than the OLS, GWR, and TWR models, with a lower AICc value, thus validating the rationale and feasibility of selecting the GTWR model.

The regression results of GTWR are shown in

Table 4. Regression results of GTWR model.

Variable	Year	APF			EDF			SSF			ECF
		p ≤ 0.1 (%)	+	−	p ≤ 0.1 (%)	+	−	p ≤ 0.1 (%)	+	−	p ≤ 0.1 (%)	+	−
RPD	2000	100.00%	169	0	66.86%	142	27	80.47%	167	2	75.74%	61	108
	2010	100.00%	169	0	76.92%	72	97	65.09%	159	10	92.31%	41	128
	2020	95.86%	166	3	98.22%	1	168	75.74%	162	7	86.98%	36	133
$\overline{\mathrm{RP}}$	2000	79.29%	57	112	88.17%	141	28	69.23%	18	151	71.60%	9	160
	2010	87.57%	94	75	100.00%	169	0	59.17%	48	121	73.96%	89	80
	2020	81.07%	102	67	100.00%	169	0	64.50%	31	138	43.79%	80	89
RPc	2000	100.00%	0	169	66.86%	86	83	47.34%	18	151	69.23%	42	127
	2010	100.00%	0	169	94.08%	6	163	44.38%	23	146	85.21%	21	148
	2020	100.00%	0	169	85.80%	24	145	29.59%	118	51	47.34%	99	70

and Figure 6. Across Northeast China, RPD significantly impacts APF, but the correlation coefficient decreases over time, and the degree of effect gradually weakens. The positive correlation coefficients between LRP, LKMR, GKMR, and APF are relatively higher. This indicates that higher rural population density corresponds to stronger APF in rural areas, but this effect gradually weakens with technological progress as population-intensive agriculture declines in Northeast China. $\overline{\mathrm{RP}}$ has a relatively small overall impact on APF, with coefficients near 0. However, RPc exerts a strong negative influence on APF, with average correlation coefficients below −0.5 across all sub-regions from 2000 to 2020. This suggests that as the rural population decreases annually, APF has shown an upward trend.

For EDF, the impact of RPD is not significant, but a trend toward a positive influence is observed, initially near zero and gradually shifting to a negative influence over time. Exerts a gradually increasing positive influence on EDF, indicating that regions with higher average rural population levels tend to exhibit stronger EDF in rural areas. RPc also exerts an adverse effect, with EDF increasing as the population decreases. Although the trends differ, this effect is more pronounced in LPR and CBMR. In LPR and LKMR, the negative impact of RPc on EDF continues to weaken, whereas in CBMR and SJP, it shifts from a slight positive impact to a negative one that strengthens over time. In SNP and GKMR, the adverse effect first intensifies and then weakens.

According to the GTWR model results, RPD shows a slight positive impact on SSF, with its influence first decreasing and then increasing, reaching near zero in 2010 before rising again in 2020. Similarly, RP has a slight negative impact on SSF, with the effect also decreasing and then increasing, and the correlation coefficient was close to zero in 2010. This suggests that counties with higher population densities generally exhibit stronger SSF, while those with larger population bases experience some suppression of SSF; however, both effects remain relatively mild. In the central region of Northeast China, particularly in LRP, the southern part of CBMR, SNP, LKMR, and the northern part of GKMR, the influence of RPc on SSF gradually shifted from negative to positive between 2000 and 2020. This change is especially notable in high SSF-value areas such as LRP and the southern part of CBMR. In the remaining areas, including SJP, the northern part of CBMR, and the southern part of GKMR, the negative influence of RPc also diminished over time but did not become positive.

For ECF, RPD exerts an overall negative impact, with areas of high rural population density experiencing a negative effect on ECF. The negative impact of $\overline{\mathrm{RP}}$ gradually weakened or shifted to a slight positive impact over time, indicating that the inhibitory effect of high average population levels on ECF is diminishing. The impact of RPc on ECF exhibited relatively complex temporal and spatial variations. LRP maintained a relatively stable negative impact, and rural population shrinkage enhanced the region’s ecological conservation function. GKMR and CBMR have experienced a weakening of their adverse effects, with their influence turning positive by 2020. The trend in SJP differs slightly, with the average correlation coefficient of RPc changing from −0.38 in 2000 to −0.88 in 2010, significantly strengthening before turning to 0.41 in 2020. SNP and LKMR exhibit similar trends, shifting from a significant positive impact in 2000 to a negative impact in 2010, then weakening to around zero by 2020, with LKMR showing more pronounced changes.

5. Discussion

5.1. Impact of Rural Population Shrinkage on Rural Functions

The findings reveal a marked acceleration in rural population decline across Northeast China from 2000 to 2020. Meanwhile, agricultural production, economic development, social security, and ecological conservation functions in rural areas showed an overall upward trend. The GTWR model results further illustrate the quantitative impact of the rural population on various rural functions: the positive effect of rural population density on agricultural production has weakened over time, and in areas with higher rural population density, the ecological conservation function has been suppressed, while its effect on social security remains slightly positive. The average rural population positively impacts economic development, with counties having larger rural populations corresponding to stronger economic development, while social security functions are slightly suppressed. The negative effect of the average rural population on ecological conservation is diminishing. The impact of relative changes in rural population on various rural functions exhibits a complex and regionally differentiated dynamic pattern. The promoting effect of rural population shrinkage on agricultural production remains relatively consistent across Northeast China but declines over time, while its effect on economic development is more significant in LPR and CBMR. The promotion of social security functions by rural population shrinkage has weakened over time and turned into a suppressive effect in central Northeast China by 2020. Rural population shrinkage generally promotes the enhancement of ecological conservation, but its trend exhibits significant spatio-temporal variation.

Rural population density is positively correlated with APF. In plain areas, high population density and favorable topography promote agricultural development, whereas the opposite occurs in mountainous regions. However, research indicates that this positive influence is weakening, confirming that agricultural production is becoming less reliant on traditional labor-intensive models. Against the backdrop of socio-economic transformation and the diversification of household livelihoods, labor transfer from agriculture to non-agricultural industries is an inevitable process driven by the pursuit of economic efficiency [8]. While this transfer has led to some land abandonment, it has also facilitated the concentration of agricultural land, providing a crucial prerequisite for large-scale, intensive modern agricultural operations [7]. Despite the decline in the rural population, advancements in agricultural technology, increased mechanization, land transfers, and large-scale operations, coupled with policy support under the national food security strategy, may offset the adverse effects of labor force reduction and even enhance output efficiency.

Contrary to the traditional view that “population loss leads to economic decline,” this study found that EDF generally increased despite population shrinkage, with regions having a larger rural population base exhibiting higher EDF. Rural population shrinkage promoted EDF, particularly in LRP and CBMR. Although the decline in the rural population has posed challenges to rural development, the reduction in the rural labor force has not significantly impeded economic growth in rural areas [29]. In fact, development levels continue to improve [14], with some counties shifting to a low-population, high-economic development model. This indicates that factors such as resource endowments, external markets, technological innovation, policy drivers, and the spillover effect of cities on surrounding rural areas are driving economic development in rural areas with shrinking populations. Simultaneously, the positive impact of average rural population levels on EDF has gradually strengthened, with counties possessing large rural populations maintaining advantages in attracting investment and industry. Additionally, as one of China’s major agricultural regions, the rural economy in Northeast China is closely tied to the performance of the primary sector. Population decline, by reshaping agricultural production functions, may also exert an indirect impact on rural economic development.

This study reveals that, in the context of rural population shrinkage, the Social Security Function (SSF) in Northeast China exhibits an overall upward trend. Rural population density exerts a slight positive impact on SSF, whereas the average rural population has a slight negative effect on SSF. Both of these influences initially decrease, then increase over time. The positive influence of rural population shrinkage on SSF has diminished over time, eventually turning into a suppressive effect in the central Northeast region by 2020. This suggests that, while population decline may initially foster the development of SSF by alleviating social pressure and enhancing per capita security resources, over time, labor outflow and social aging may intensify challenges, leading to a structural imbalance in social security demands. Overall, the spatiotemporal differentiation in SSF’s response to rural population shrinkage underscores the need for localized policy formulation, particularly in high-value regions in the central and southern areas, where security strategies during the shrinkage phase should be adjusted to prevent a shift from promotion to suppression. Notably, Li et al. suggested that various forms of population shrinkage in the black soil region of Northeast China contribute to the development of social security functions [20]. This conclusion was mainly attributable to the limitations of Li et al.’s research methodology, which failed to capture the trend of this shift by 2020.

The reduction in the rural population in Northeast China has been associated with enhanced ecological functionality, as evidenced by an inverse relationship between rural population density and ECF. This aligns with the argument that reduced population pressure facilitates ecological restoration [9,15,52] and the strong negative correlation found in bivariate spatial autocorrelation analysis. Ecological pressure resulting from population shrinkage, combined with ecological protection policies in Northeast China, has promoted the development of ecological conservation functions. However, the GTWR results reveal considerable variation over space and time in how RPc affects ECF. The promotional effects of GKMR, CBMR, and SJP gradually weaken over time and even become inhibitory, indicating that the impact of population contraction on ecology is neither linear nor static. Ecological protection policies should be adjusted according to the stage of population contraction and implemented region-specifically based on local conditions.

5.2. Policy Recommendations

This study reveals the complex and significantly heterogeneous spatial and temporal effects of rural population shrinkage on rural functions in Northeast China. Given the marked acceleration of rural population decline in Northeast China from 2000 to 2020 and the difficulty of reversing this trend in the short term, policymakers should focus more on managing the challenges brought by population decline in rural areas rather than simply aiming to restore growth [4]. Counties experiencing sustained population loss can draw on the “smart shrinkage” strategies of developed Western countries [75], which encourage regional planning to accept the trend of population decline. Rather than maintaining sprawling infrastructure and service levels designed for historically larger populations, these strategies call for readjusting public infrastructure and services, reallocating public resources, and optimizing limited fiscal and material resources to precisely meet the needs of the current—often much smaller—rural population. Infrastructure and investments should be aligned with the present population scale to mitigate negative effects such as reduced tax revenues and increased vacant properties. The main objectives are to move away from traditional growth models, optimize limited financial and material resources, meet residents’ living and employment needs, and enhance satisfaction while retaining talent [33]. At the same time, attention should be paid to the endogenous development potential of rural areas, and differentiated policy instruments should be applied based on the spatial and temporal differences in functional responses, in order to facilitate the coordination between demographic transitions and rural functional transformation, thereby enhancing the adaptability of rural systems and contributing to improved resident well-being.

Studies show that the positive impact of rural population density on agricultural production has gradually weakened over time, confirming that agricultural production is progressively reducing its dependence on traditional labor-intensive models. This shift, combined with advancements in agricultural technology, increased mechanization, and land transfer, has allowed large-scale, intensive modern agricultural operations to offset the negative effects of labor reduction, and even improve output efficiency. Therefore, policy focus should shift to vigorously promoting agricultural modernization and transformation, developing capital- and technology-intensive models, strictly implementing the arable land occupation-compensation balance policy, prioritizing the protection of high-quality arable land resources, accelerating the transfer and consolidation of land management rights, and fostering appropriately scaled operations to address labor shortages and improve production efficiency [37]. Although rural population shrinkage may temporarily accelerate large-scale agricultural development, it is not conducive to sustained growth in the long run [7]. Moreover, our research results also indicate that the promoting effect of population shrinkage on agricultural production is gradually weakening over time. Therefore, it is necessary to actively expand the multifunctionality of agriculture by developing deep processing of agricultural products, rural tourism, and recreational farming, among other forms, to achieve the deep integration of agriculture with the secondary and tertiary sectors. This will create more non-agricultural employment opportunities through positive externalities, such as rural landscapes. Additionally, a clear trade-off exists between agricultural production and ecological conservation in Northeast China, and it is essential to further explore ecological agricultural models, reduce the use of chemical fertilizers and pesticides, address agricultural non-point source pollution [76], and promote the synergistic development of agricultural production and ecological conservation functions.

It has been revealed in our study that, despite the decline in the rural population, the economic development function in Northeast China has generally continued to strengthen, and in certain areas, population loss has even stimulated economic growth. In response to this trend, policies should prioritize the cultivation of a diversified rural industrial system adapted to a shrinking population. Building on Northeast China’s comparative advantages—such as its role as a major commodity grain base and its abundant forestry and snow tourism resources—the grain production industry chain should be extended, black soil organic agricultural products developed, and eco-tourism, snow tourism, border tourism, and forest health and wellness industries actively promoted based on unique natural endowments. Social capital should be guided and encouraged to invest in distinctive rural industry projects to invigorate endogenous rural development. Furthermore, the “people-land-industry” framework underscores that land serves as the essential foundation and spatial carrier for both demographic and industrial activities. To support economic transformation, effective mechanisms for revitalizing and utilizing idle residential land and abandoned industrial and mining sites should be explored.

Although the overall trend of rural social security functions has been upward, the positive effects of rural population shrinkage have diminished over time and have even shifted to a constraining influence. This suggests that while initial population decline may alleviate certain social pressures, in the long term, rural depopulation poses challenges to social security, particularly in the central and southern regions where social security functions are relatively strong. In areas where rural social security is weak, the policy focus should shift from merely maintaining the scale of service facilities to improving the efficiency of resource allocation and the accessibility of services, thereby preventing the effects of population shrinkage from shifting from promotion to inhibition. In counties with suitable conditions, medical and educational resources should be concentrated in central towns, and the basic service provision for surrounding rural areas should be ensured through mobile service vehicles and village-level comprehensive service stations. In remote areas without such conditions, the construction and popularization of rural digital infrastructure should be advanced [19], and telemedicine, online education, digital government affairs, online finance, and other measures should be used to effectively compensate for the lack of physical service outlets. In rural areas where social security functions are relatively sound but population loss has resulted in idle infrastructure and long-abandoned homesteads, the actual needs of permanent rural residents should be met, avoiding the creation of new idle rural homesteads. Meanwhile, through the construction of beautiful villages and the preservation of local characteristics, the urban–rural infrastructure gap should be narrowed. Continuous improvement of the road network connecting cities, central towns, key villages, and major transportation arteries should also be pursued to enhance rural accessibility, ensure basic travel and smooth logistics [77], and attract people back to the countryside.

The impact of rural population shrinkage on ecological conservation functions presents complex spatiotemporal dynamics. Therefore, ecological protection policies must be adaptive, adjusting according to the specific stages of population decline and implementing region-specific policies based on local conditions, to prevent potential problems in the natural succession process, such as ecosystem structural simplification or functional degradation, following substantial population loss. In regions like the Lesser Khingan Mountains Region and the northern Songnen Plain, where the effect of population shrinkage on ecology shifts from suppression to promotion but later weakens, ecological advantages should be fully leveraged. This includes developing mountainous industries, establishing and improving ecological compensation mechanisms, advancing eco-tourism and under-forest economies, participating in carbon trading, and exploring effective pathways for realizing the value of ecological products. This approach aims to transform favorable ecological environments into sustainable economic benefits, thereby contributing to local development and promoting a virtuous cycle of ecological protection and economic growth. In areas where population shrinkage gradually suppresses ecological conservation functions, such as the Hulunbeier Grassland, Changbai Mountain Reserve, and the Sanjiang Plain, the protection of ecological sources should be strengthened, ecological corridors constructed, the landscape types of ecological nodes optimized, and an ecological security network for Northeast China should be established [78]. In the Liaohe River Plain, focus should be placed on strengthening the development of ecological agriculture, enhancing farmland ecological protection measures, and achieving synergistic development between agriculture and ecology.

In summary, to effectively respond to the issue of rural population shrinkage in Northeast China, forthcoming strategies for rural development need to follow the trajectory of demographic contraction, prioritize development quality, and support integrated progress across rural functional domains through spatial optimization and efficient resource distribution, ultimately achieving sustainable rural development in the region amid demographic shifts.

5.3. Research Limitations and Future Directions

Although this study has made significant strides in establishing a quantitative relationship between rural population shrinkage and rural functions, limitations remain. Due to the influence of the research scale and scope, data availability has been limited, and some selected indicators may not be precise or directly relevant. For example, using per capita farmland area as an indicator for the social security dimension, although it is somewhat related to agricultural production functions, makes it difficult to fully and accurately represent a single function dimension. Some indicators are not specifically targeted at rural areas but are instead measured at the county scale. Nevertheless, this study has carefully balanced data availability with research objectives in selecting indicators to construct a reasonable indicator system, although room for improvement remains in terms of representational accuracy and scientific rigor. Furthermore, while GTWR can effectively capture the spatio-temporal heterogeneity of population impacts on functions, it fails to capture the interactions and coupling mechanisms between functions.

In future research, datasets such as nighttime light intensity should be used to more precisely delineate rural areas, and indicators directly related to rural functions must be prioritized to enhance the scientific rigor and relevance of the indicator system. Alternatively, qualitative case studies or interviews could be incorporated to enrich and support the research findings, thereby further improving the scientific rigor of rural function evaluation. Emphasizing the interplay between complementary and competing rural functions is expected to redirect attention toward demographic aging and labor force restructuring, while also investigating how technological innovation and policy coordination can jointly address rural population decline.

6. Conclusions

Taking Northeast China as a case study, this research established an index system for evaluating rural functions and identified the types of rural population shrinkage at the county level. A comprehensive study was conducted to explore the quantitative correlation between population dynamics and rural functions, focusing on their spatiotemporal changes. The main findings are as follows:

The rural population in most counties of Northeast China declined rapidly from 2000 to 2020, with population distribution decreasing from the central plains to the surrounding mountains. However, this contraction did not significantly change the spatial distribution pattern of rural populations.

Against the backdrop of population decline, the agricultural production, economic development, social security, and ecological conservation functions in Northeast China exhibited an overall upward trend during the study period. Spatially, areas with high agricultural production and economic development are mainly located in plains and urban fringe areas, while areas with high social security are concentrated in the central and southern regions. High-value ecological conservation areas are primarily concentrated in mountainous and forested regions.

The GTWR model reveals the spatiotemporal heterogeneity in the impact of rural population changes on rural functions. The promoting effect of rural population density on agricultural production weakens over time, while it slightly promotes social security and continues to inhibit ecological conservation. The supporting effect of average population size on economic development strengthens, its inhibitory effect on ecological conservation diminishes, and it weakly inhibits social security. Population shrinkage generally promotes rural agriculture, economic development, and social security and improves ecology, but its impact on agricultural production decreases over time. The promotion effects on social security and ecological conservation partly turn into inhibitions after 2020.