Introduction

A river basin constitutes an autonomous and comprehensive area within the realm of physical geography, which is a pivotal origin and central hub for the development of human culture1. The Yellow River Basin (YRB) has been observed to demonstrate social complexity and urbanization coinciding with the Neolithic to Bronze Age transition in northern China2. The agricultural civilization is the foundation of the Yellow River culture3. A unique regional agricultural civilization system with population, production and ecological adaptation has gradually formed in the river basin, and becoming a major birthplace of Chinese civilization. An agricultural civilization is the interaction between man and the land4, creating a large amount of cultural heritage related to agricultural production and living. As a cultural heritage, agricultural civilization heritage exerts a pivotal role in fortifying cultural self-assurance and national solidarity5.

With urbanization and industrialization, as well as excessive development of resources within the YRB6, agricultural civilization heritage faces with multiple external disturbance, as well as crises such as fragmentation, homogenization, and inferiority. It is urgent to conduct integrity research and protection on agricultural civilization heritage in order to inherit the Yellow River agricultural civilization and promote its historical transformation. This has prompted the government to take notice. In October 2020, the Chinese government proposed to construct the Yellow River National Cultural Park and a system of cultural landmarks7. In 2021, Chinese government published an outline document on the ecological protection and high-quality development of the YRB. This document proposes to promote the revitalization, utilization, inheritance and innovation of excellent agricultural civilization heritage in YRB8.

The research on agricultural civilization heritage in China has a long history. As early as the Republic of China period, attention was paid to the protection and research of agricultural heritage system9. The Department of Rural Social Undertakings Promotion, Ministry of Agriculture and Rural Affairs of China, in collaboration with Shanghai Jiao Tong University, conducted a joint research project on the inheritance and preservation of agricultural cultural heritage. Their work promotes practical innovations in exploration and conservation, systematic documentation, and sustainable development within agricultural culture. The research team emphasized that agricultural culture is rooted in human agricultural practices, encompassing both rural material culture (e.g., traditional settlements, agricultural infrastructure, land-use systems, agroforestry products, and farming tools) and rural intangible culture.

Within China’s existing cultural heritage classification system, agricultural civilization heritage is defined as encompassing agricultural production systems, including land-use patterns, agroforestry products, and farming infrastructure. Heritage associated with rural settlements (e.g., traditional villages, historic towns) is categorized separately under "Traditional Chinese Villages" or "Historic and Cultural Towns/Villages", while rural intangible practices fall under Intangible Cultural Heritage (ICH). Collectively, these three domains—agricultural production, settlement life, and intangible traditions—constitute the core components of agricultural culture, representing its material, social, and spiritual dimensions, respectively. In 2002, the Food and Agriculture Organization of the United Nations (FAO) initiated the Globally Important Agricultural Heritage Systems (GIAHS) project, which has designated 95 GIAHS in 28 countries10. As one of the earliest adopters of the project, China boasts the largest number of GIAHS, with 25 items11. Since 2012, China has also recognized seven batches of China’s national-level Important Agricultural Heritage system (China-NIAHS), totaling 188 sites distributed across 31 provincial-level administrative regions. The dynamic conservation of Agricultural Heritage Systems represents a core research focus. Current studies predominantly examine four key aspects: first, the resource valuation12 and developmental potential13 of these systems; second, development models14 with sustainability considerations; third, perceptions among stakeholders including tourists and residents15; and fourth, public engagement mechanisms16. The government-recognized historical and cultural towns and villages, and Traditional Villages represent the most valuable agricultural settlements. So far, a total of 312 historical and cultural towns, 487 historical and cultural villages, and 8155 Chinese traditional villages have been announced17. Research on historical villages and towns encompasses three primary aspects: historical evolution18, conservation and development, and tourism development19. Beyond case-specific investigations, the studies also examine spatial structures and influencing factors at multiple scales20, including national, provincial, and watershed levels. Since the introduction of the Convention for the Safeguarding of the Intangible Cultural Heritage, the safeguarding of the intangible cultural heritage has attracted attention. China leads globally with 44 UNESCO-listed ICH elements, while over 100,000 ICH projects have been registered at national, provincial, municipal, and county levels—including 1,557 at the national level. The spatial dimensions of intangible cultural heritage have become a central focus in geographical research. These studies primarily address two key aspects: first, the examination of spatiotemporal distribution patterns, influencing factors, and geographical response mechanisms across different heritage scales and categories21; and second, the analysis of cultural diffusion processes22 and public perception characteristics associated with intangible cultural heritage23.

The YRB has abundant agricultural civilization heritage. The agricultural civilization heritage in YRB has highly convergent cultural values, close spatial connections and natural contiguous protection conditions, but existing studies tend to focus on a typical heritage type, such as historical and cultural towns and villages24, traditional villages25 and intangible cultural heritage26. There is a paucity of research on the totality of agricultural civilization heritage in YRB, as well as the spatial correlation of different heritage and the driving mechanism.

Based on the agricultural civilization classification framework established by China’s Ministry of Agriculture and Rural Affairs, this study proposes a theoretical model for agricultural civilization heritage conservation. This research was conducted with the agricultural civilization heritage of 69 cities in nine provinces and autonomous regions of the YRB as study object. Using GIS spatial analysis, we employ a multi-method approach to examine the spatial distribution characteristics, interrelationships, and potential influence mechanisms. This study lays an essential basis for preserving the spatial integrity of the heritage of agricultural civilization in YRB, which has the potential to strengthen the inheritance and protection of agricultural civilization in river basins, and provide a theoretical basis and a decision-making reference for the protection of agricultural civilization heritage in YRB and the construction of the National Cultural Park.

materials and methods

Study area

The Yellow River Basin spans longitudes 96° 53′ E to 119° 05′ E and latitudes 32° 10′ N to 42° 50′ N. The YRB traverses the Tibetan Plateau, the Inner Mongolia Plateau, the Loess Plateau, and the Yellow-Huaihai Plain from west to east. It passes through 69 cities in nine provincial-level administrations, including Qinghai, Sichuan, Gansu, Ningxia, Inner Mongolia, Shaanxi, Shanxi, Henan, and Shandon27. The YRB is the cradle of China’s farming civilization, where fertile soil and a suitable climate support crop production. The region’s earliest evidence of primitive agriculture dates back to prehistory. Its unique geography has given rise to many farming cultures. This study employs 69 prefecture-level cities (prefectures) as analytical units instead of provincial divisions or strict hydrological boundaries for three key reasons. First, provincial-scale analysis would incorporate extensive territories beyond the Yellow River Basin’s core zone of ecological-economic influence, introducing extraneous areas that could distort findings. Second, while hydrological boundaries are geographically precise, they frequently fragment administrative units, compromising data continuity; by contrast, cultural heritage practices, historical irrigation systems, and agricultural economies operate within integrated administrative boundaries that enable coherent comparative analysis. Third, the selected cities comprehensively represent the Yellow River’s mainstem core area, its cultural continuity belt comprising historically irrigation-dependent regions, flood-modified agricultural zones, and tributary-based cultural diffusion corridors—a scope demonstrating remarkable congruence with China’s Ecological Protection and High-Quality Development of the Yellow River Basin strategy, thus ensuring policy relevance while maximizing both analytical rigor and data reliability. Therefore, this paper takes 69 cities in 9 provinces in YRB as the study area (Fig. 1), comprising all prefecture-level cities in Shanxi, Qinghai, and Ningxia Hui Autonomous Region, plus selected cities from Shandong (9), Henan (9), Shaanxi (8), Inner Mongolia (7), Gansu (10), and Sichuan (2). The study area covers approximately 2.245 million square kilometers, representing 62.60% of China’s total land area.

Fig. 1
figure 1

Research scope. The map was generated using ArcGIS Pro 3.0.2 (Environmental Systems Research Institute, Redlands, CA, USA; https://www.esri.com).

Full size image

Data

Building on the agricultural civilization classification standards recognized by China’s Ministry of Agriculture and Rural Affairs (MARA), we developed a theoretical framework for agricultural heritage systems through a tripartite mapping of agricultural production, agricultural livelihoods, and agricultural spirituality. We divide agricultural civilization heritage into three categories based on the characteristics of agricultural civilization: Agricultural Production Heritage(APH), Agricultural Life Heritage(ALH), and Agricultural Spiritual Heritage(ASH). Agricultural production heritage refers to the agricultural heritage system, which is a traditional agricultural system created and inherited through the long-term coordinated development of human beings and their environment. Agricultural life heritage refers to settlement heritage, including historical and cultural towns and villages, and Chinese traditional villages recognized by the Chinese government. Agricultural spiritual heritage refers to intangible cultural heritage. Agricultural spiritual heritage refers to intangible cultural heritage18, including global intangible cultural heritage and national-level intangible cultural heritage.

This study determined marked and geocoded the locations the geographical location of the agriculture civilization heritage in Baidu coordinates (Fig. 1), and used ArcGIS software to construct the resource database in YRB. The agricultural production heritage in this study includes four GIAHS and 27 China-NIAHS, with the data sourced from Ministry of Agriculture and Rural Affairs of China. The agricultural life heritage includes 36 historical and cultural towns, 123 historical and cultural villages, 1467 Chinese traditional Villages, with the data sourced from Chinese Traditional Villages Website and State Administration of Culture Heritage. Data on intangible cultural heritage comes from the China Intangible Cultural Heritage Network, which has 734 national intangible cultural heritage items in the study area.

Factors influencing the spatial distribution of agricultural civilization heritage included physical geography factors and socioeconomic factors (Table 1). The DEM (Digital Elevation Model) data are sourced from the Chinese Academy of Sciences data cloud platform, and the elevation and slope data were extracted with the assistance of ArcGIS software. Road and river system data are obtained from the OpenStreetMap platform. Average Annual Temperature and GDP data were obtained from the National Earth System Science Data Center. Population data were sourced from the Oak Ridge National Laboratory. The urbanization level data were derived from the China Statistical Yearbook-2021. Nighttime light (NTL) data were obtained from the Urban Environment Monitoring and Modeling (UEMM) team of the School of Land Science and Technology, China Agricultural University. Obtain geographic spatial data such as administrative divisions of the research area through Baidu Maps and the National Geomatics Center of China, and perform spatial correction and other processing on the data in ArcGIS software to obtain the vector spatial database used for research analysis.

Table 1 Natural and social factors data source summary.
Full size table

In accordance with Geodetector’s requirement for categorical inputs, continuous variables were discretized using optimized classification protocols. Using the GD package in R, the optimal discretization method is automatically selected from among 'equal,' 'natural,' 'quantile,' and ‘geometric’ through an algorithmic approach to discretize continuous variables.

Methods

The nearest neighbor index

Neighbourhood index (R) is the ratio of actual to theoretical neighbourhood distance. It can be used to measure whether the spatial distributions are random, uniform or agglomerative. The method is utilized for the purpose of analysis of the spatial distribution characteristics of diverse types of agricultural civilization heritage in YRB. The calculation formula is as follows29:

$$\begin{array}{*{20}c} {R = \frac{{\overline{r}}}{{\overline{r}_{i} }};\overline{r}_{i} = \frac{1}{{2\sqrt {n/A} }} = \frac{1}{2\sqrt D }} \\ \end{array}$$
(1)

where \(R\) denotes the nearest neighbour index,\(\overline{r}\text{,}{\overline{r}}_{i}\) denote the actual mean distance, theoretical mean distance,\(n\) denotes the number of samples of heritage sites, \(A\) denotes the area of the study area, and \(D\) denotes the density of sites.

Kernel density analysis

Kernel density analysis is a method of calculating the density of points in a neighborhood around those points. This study analyses the three types of agricultural civilization heritage in YRB as a whole and individually in order to summarize the core areas of distribution of each type of heritage. The formula is as follows30:

$$\begin{array}{*{20}c} {F\left( x \right) = \frac{1}{nh}\mathop \sum \limits_{i = 1}^{n} k\left( {\frac{{x - x_{i} }}{h}} \right)} \\ \end{array}$$
(2)

where \(k\left(\frac{x-{x}_{i}}{h}\right)\) denotes the kernel density function, \(n\) denotes the number of heritage site samples, \(h\) denotes bandwidth(> 0).(\(x-{x}_{i}\)) denotes the distance from the \(x\) valuation point to the \({x}_{i}\) heritage point.

Centre of gravity model analysis

The centre of gravity model was initially utilized by American scholars in the realm of socio-economic investigation31, employs the spatial centre of gravity of elements within a region to elucidate the spatial evolution process and spatial distribution between different elements32. The centre of gravity position of the distribution of each type of agricultural civilization heritage in YRB can be obtained. If the distribution of heritage resources is balanced, its center of gravity is the geometric center of the area; If the center of gravity deviates significantly from the geometric center of the region, it indicates that the heritage has an uneven spatial distribution, and the distance of deviation reflects the degree of spatial distribution imbalance. The calculation formula is as follows:

$$\begin{array}{*{20}c} {X = \frac{{\mathop \sum \nolimits_{i = 1}^{n} f_{i} x_{i} }}{{\mathop \sum \nolimits_{i = 1}^{n} f_{i} }};Y = \frac{{\mathop \sum \nolimits_{i = 1}^{n} f_{i} y_{i} }}{{\mathop \sum \nolimits_{i = 1}^{n} f_{i} }}} \\ \end{array}$$
(3)

where \(X\), \(Y\) denote the horizontal and vertical coordinates of the centre of gravity of the heritage point, \(n\) denotes the number of cities and towns, \({x}_{i}\), \({y}_{i}\) denote the horizontal and vertical coordinates of the geometrical centre of the ith city and town, \({f}_{i}\) denotes the value of an attribute of an element in the i-th region. The study involves multiple elements, and the spatial relationship between the two can be examined by the overlap of their centroids. The deviation index can be used to determine the degree of spatial distribution deviation between the elements. The calculation formula is as follows:

$$\begin{array}{*{20}c} {S = \sqrt {\left( {X_{i} - X_{i} } \right)^{2} + \left( {Y_{i} - Y_{i} } \right)^{2} } ;r = 100*S/\sqrt D } \\ \end{array}$$
(4)

where \(S\) denotes the distance between the centres of gravity of the two types of heritages, \(D\) denotes the total area of the 69 cities in YRB, \(r\) denotes the index of deviation from the centres of gravity of the two types of agricultural civilization heritages.

Landscape spatial correlation index

Using ArcGIS for landscape spatial correlation index analysis, we calculate the spatial correlation between two types of heritage relying on 2 × 2 columnar table data. Firstly, create a 20 km × 20 km grid layer using ArcGIS. Then, superimpose the grid layer with the heritage layer to form a composite layer. According to the topology database corresponding to the composite layer, calculate the binary data of landscape elements in each grid. Finally, create a binary contingency table for heritages and further calculate the landscape spatial correlation. The formula is as follows33:

$$\begin{array}{*{20}c} {R = \frac{ad - bc}{{\sqrt {\left( {a + b} \right)\left( {c + d} \right)\left( {a + c} \right)\left( {b + d} \right)} }}} \\ \end{array}$$
(5)

where R represents the landscape spatial correlation index, a denotes the number of grids where both types of agricultural civilization heritage are present, b denotes the number of grids where only the first type of agricultural civilization heritage is present, c denotes the number of grids where only the second type of agricultural civilization heritage appear, and d denotes the number of grids where neither type of agricultural civilization heritage appear. The value of R ranges from -1 to 1. A positive value indicates a positive association between the two types of heritage, while negative values denote a negative correlation. The significance of R value can be tested using the following formula:

$$\begin{array}{*{20}c} {X^{2} = \frac{{n\left( {ad - bc} \right)^{2} }}{{\left( {a + b} \right)\left( {c + d} \right)\left( {a + c} \right)\left( {b + d} \right)}}} \\ \end{array}$$
(6)

where \(n\) denotes the total of samples, when \(a=0.05\), check the distribution table of Xa2(1) , if \({|X}^{2}|>{X}_{a}^{2}(1)\), it means that the quantitative relationship between the two types of heritage is significant, and vice versa is not significant.

The Geodetector

Geodetector is an effective instrument for studying the interactions between different geographic factors34. In this study, with the help of Geodetector’s factor detection module, we analyzed the different effects of the X influence factor on the spatial distribution of Y heritage35. The specific calculation formula is:

$$\begin{array}{*{20}c} {q = 1 - \frac{{\sum_{h = 1}^{L} N_{h} \sigma_{h}^{2} }}{{N\sigma^{2} }}} \\ \end{array}$$
(7)

where \(q\) represents the impact strength of the influencing factor X on the heritage Y, taking values between 0 and 1, with larger values indicating greater influence and vice versa. \(h\) denotes the categorization of the variable \(Y\) or of the influence factor \(X\), \({\sigma }_{h}^{2}\) and \({N}_{h}\) denote the variance and the number of cells of the influence factor of class \(h\). \({\sigma }^{2}\) and \(N\) denote the legacy point sample and the variance.

Results

Spatial distribution

Types of spatial distribution

Using ArcGIS software, the clustering and dispersion types of agricultural civilization heritage resources in YRB were calculated and analyzed (Table 2). The average nearest neighbor index R of agricultural civilization heritage is 0.35, indicating that the agricultural civilization heritages show an agglomeration distribution in YRB. The R of APH is the largest, which is 0.95, indicating that the APH in YRB belongs to the random distribution type. The R of ASH and ALH are 0.45 and 0.47, respectively, which are less than 0.5, showing a clear spatial aggregation distribution characteristic in the, with ALH having a higher degree of aggregation than ASH.

Table 2 Summary Table of average nearest neighbor index of agricultural civilization heritage.
Full size table

Spatial density distribution characteristics

We use Kernel Density in the ArcGIS to estimate the spatial density distribution characteristics of the agricultural civilization heritage in YRB as a whole (Fig. 2). The agricultural civilization heritage form one high-density core area and three sub-high-density cores area in YRB. The high-density core area is centered around the border between Henan Province and Shanxi Province, covering multiple surrounding cities such as Jincheng city. The sub-high-density core areas are located in the eastern part of Qinghai Province, the central part of Shanxi Province, and the central part of Shandong Province.

Fig. 2
figure 2

Map of Kernel Density Distribution of Various Agricultural Civilization Heritage Sites.

Full size image

Figure 2 shows the results of the kernel density. The spatial distribution of APH kernel density is relatively random, with distribution in all nine provinces of the Yellow River (Fig. 2). Henan Province, Shandong Province, Shaanxi Province, and Ningxia Hui Autonomous Region have relatively more distributions, but there is no obvious feature of clustering. The spatial distribution of ASH kernel density is relatively clustered and mainly have two significant density cores, the primary density core area was mainly distributed in the Shanxi Province and Henan Province, Connecting Shandong Province and Shaanxi Province from east to west and showing the “Y” pattern. Sub-density core areas were mainly in areas where ethnic minorities are concentrated, such as in Qinghai and Gansu Provinces. The spatial distribution of ALH kernel density are heavily concentrated in Shaanxi Province and its surrounding areas, in addition to numerous ALH distributed in the ethnic minority gathering areas, especially in the Qinghai Province, Gansu Province and Sichuan Province. As a whole, different types of agricultural cultures have different distribution characteristics and varying degrees of regional overlap, generally shows a distribution of more heritage resources in the Middle and lower reaches of the Yellow River and along its banks than the other places.

Spatial relation

Balance of spatial distribution

By comparing the link between the centre of gravity of the heritage and the geometric centre of the study area, and reflecting the balance of spatial distribution among various agricultural civilization heritages in YRB, the geographic coordinates of the centers of gravity of APH, ASH, and ALH in YRB and each province are first determined. Then, the center model is used to calculate the deviation distance and deviation index (Table 3). From a locational perspective (Table 3), the geometric center coordinates of the YRB are approximately 36° 38′ 28.32″ N latitude and 103° 30′ 15.84″ E longitude. The centroids of APH, ASH, and ALH are all located in Yan’an City, southeast of the geometric center of the YRB (Fig. 3), indicating the spatial distribution of the three heritage types (production, livelihood, and spirituality) in the Yellow River Basin demonstrates significant similarity, revealing their deeply coupled mechanisms within the cultural ecosystem. With regard to the concept of deviation distance, the centers of APH and ALH features the most pronounced displacement distance in YRB are about 128.433 km apart, with a deviation index of 0.086. The centers of APH and ALH are the least significant, with a deviation distance of approximately 44.222 km and a deviation index of only 0.030. As a result, the deviation distance from the centre of gravity of the heritage of each type of agricultural civilization within the YRB is small, indicating a clear spatial correlation.

Table 3 The shift in the center of gravity and the deviation index for agricultural civilization heritage with.
Full size table
Fig. 3
figure 3

Map of the central location of agricultural civilization heritage.

Full size image

The centroids of agricultural civilization heritages in YRB exhibit certain spatial deviations, the extent of deviation varies between different regions. In order to avoid research bias caused by insufficient agricultural heritage system, only the intangible cultural heritage and settlements in each province will be discussed. The ASH and ALH in the Ningxia Hui Autonomous Region features the most pronounced displacement distance, followed by that in the Qinghai province having more deviation distances than other provinces. The Ningxia Hui Autonomous Region is the main settlement of Hui and other ethnic minorities, and the presence of the smallest sub-basin area further results in a larger deviation index. In history, Qinghai Province was the intersection of northwest nomadic tribes and eastern agricultural civilization. The eastern region was mostly a place where ethnic rule and political power changed throughout history, resulting in a significant deviation distance.

The correlation degree of spatial distribution

A 20 km grid system was established across the study area, generating 8842 sample units. Using ArcGIS, we overlaid point layers of agricultural heritage system sites—including 31 APH, 734 ASH, and 1,626 ALH locations across 69 Yellow River Basin cities—onto this grid. This process produced the spatial distribution map shown in Fig. 4. Firstly, the spatial correlation index (R) between APH and ASH is 0.1952 (Table 4) indicating a statistically significant but functionally weak positive association. The grid analysis reveals 19 grids containing exclusively APH, 309 with only ASH, 8 exhibiting both types, and 8506 devoid of either heritage. The spatial correlation index between APH and ALH is 0.1250 (Table 4), significantly weaker than the APH-ASH correlation. Corresponding distribution patterns show: 19 grids containing exclusively APH, 710 containing only ALH, merely 8 featuring both types, and 8105 devoid of either heritage. Finally, ASH and ALH demonstrate a comparatively stronger (R = 0.3952), though still below the threshold for moderate correlation. Distribution analysis identifies 221 grids with exclusively ASH, 622 with solely ALH, as few as 96 containing both heritage types, versus 7903 devoid of either. Collective analysis confirms discernible yet variable positive spatial correlations among the heritage categories (all R > 0, ranging from 0.125 to 0.3952). Furthermore, referring to the research of Chen et al.36, according to the chi square distribution table, when a = 0.05, \({X}_{a}^{2}(1)\) =3.84, and \({|X}^{2}|>{X}_{a}^{2}(1)\), all values passed the significance test at a level of 5%, suggesting a substantial positive correlation in the spatial distribution of the quantities between the three. However, both APH-ASH (R = 0.195) and APH-ALH (R = 0.125) exhibited extremely low grid proportions (≤ 0.1%, 8/8842), suggesting a weak substantive association. In contrast, ASH-ALH (R = 0.395) displayed a substantially higher grid proportion (1.1%, 96/8842), indicating a quasi-moderate association strength.

Fig. 4
figure 4

Spatial Distribution and Sample Plot Decomposition Map of Various Agricultural Civilization Heritages.

Full size image
Table 4 Types of spatial association indices for agricultural civilization heritage sites.
Full size table

Analysis of influencing factors

Drawing on methodologies from Tian et al.37, Li et al.38, and Chen et al.36, this study employs ten key indicators to quantify driving forces behind agricultural civilization heritage distribution (Table 5).

Table 5 Determinants and predictive strength of agricultural civilization heritage.
Full size table

The spatial distribution of agricultural heritage systems in the Yellow River Basin exhibits distinct dual drivers comprising both natural and socio-economic factors. Geodetector analysis indicates that road network density (X7) significantly influences all three heritage types (q APH = 0.464, q ASH = 0.549, q ALH = 0.174), demonstrating particularly strong explanatory power for both APH and ASH. Regions with higher road network density generally display enhanced geographical connectivity, which facilitates the dissemination of agricultural techniques and enables cross-regional exchanges of intangible cultural heritage, including folk arts and oral traditions. Furthermore, these well-connected areas frequently coincide with historically established agricultural economic zones, where prolonged stable agricultural production has yielded substantial agricultural civilization legacies. Urbanization rate (X10) similarly exhibits significant explanatory power across all three heritage types (q APH = 0.369, q ASH = 0.426, q ALH = 0.327), underscoring urbanization’s substantial impact on heritage distribution patterns. Urban expansion induces land-use changes and population mobility that progressively degrade traditional agricultural landscapes, with this effect being particularly pronounced for ALH. In contrast, urbanization simultaneously promotes heritage tourism development and facilitates policy-driven conservation efforts, thereby generating new preservation opportunities for both APH and ASH. Annual precipitation (X5) exerts widespread influence (q APH = 0.312, q ASH = 0.406, q ALH = 0.270), corroborating the essential reliance of traditional agricultural systems on precipitation regimes throughout the Yellow River Basin. Precipitation directly affects surface water and groundwater recharge, which in turn influences soil quality and fertility. On the other hand, early humans had weaker ability to resist natural disasters, and droughts and floods caused by precipitation to some extent affected the living environment of humans, thereby affecting the dissemination of agricultural civilization heritage. These factors, along with agricultural productivity, contribute to population growth and the concentration of settlements, shaping the distribution patterns of agricultural civilization heritage.

For Agricultural Productive Heritage (APH), GDP (X8, q = 0.346) and nighttime light index (X9, q = 0.251) demonstrated strong explanatory power, second only to road network density, reflecting the significant influence of economic development level on productive heritage. The moderate explanatory power of water network density (X3, q = 0.277) indicates the fundamental role of water resources. Regarding Agricultural Spiritual Heritage (ASH), water network density (X3, q = 0.386) also exhibited notable explanatory power, highlighting the particular importance of hydrological conditions for agricultural spiritual-cultural activities. In contrast, economic factors like GDP (X8, q = 0.277) and nighttime lights (X9, q = 0.248) showed relatively weaker influences, while climatic factors such as temperature (X4, q = 0.253) demonstrated moderate explanatory power. For Agricultural Living Heritage (ALH), all driving factors generally displayed lower explanatory power. Although factors including temperature (X4, q = 0.162), population size (X6, q = 0.158), and water network density (X3, q = 0.155) showed relatively weak influences, they still maintained statistical significance.

Overall, the heritage of agricultural civilization is a unique agricultural cultural and geographical landscape product formed under the influence of social and economic factors in the process of symbiosis between human and natural environmental factors. It also breeds related social structures and is an important manifestation of cultural diversity.

Discussion

The agricultural civilization heritage as a crucial carrier of agricultural civilization in YRB, of exceptional cultural value and profound historical significance. From 2020 to 2024, the central government has emphasized rural cultural work in its No. 1 Central Document for five consecutive years, calling for the comprehensive implementation of the Agricultural Culture Inheritance and Protection Project. This initiative has been incorporated into the Key Project Plan for the Inheritance and Development of Excellent Traditional Chinese Culture during the 14th Five-Year Plan period. However, amid rapid urbanization and industrialization, this heritage faces external pressures, along with internal challenges such as fragmentation and homogenization. The establishment of the Yellow River National Cultural Park7, along with efforts to protect, inherit, and innovate upon this heritage, is a key initiative to promote the development and wealth of the Yellow River culture. This study utilizes geospatial data to analyze the spatial patterns and influencing factors of agricultural civilization heritage in YRB, addressing the existing research gap in the holistic study and spatial conservation of this heritage cluster. Overall, the spatial distribution of agricultural civilization heritage in YRB exhibits a distinct clustering pattern, with major concentrations in the middle and lower reaches of the Yellow River, as well as in ethnic minority regions such as Qinghai and Gansu in the upper reaches. Additionally, the three types of heritage demonstrate a high degree of spatial balance, with a significant positive correlation in their quantitative distribution. The findings suggest that these three types of heritage are spatially interconnected and mutually influential, collectively forming the overall spatial structure of agricultural civilization heritage. Therefore, establishing a comprehensive conservation system for this heritage is essential for sustaining the transmission of Yellow River culture and, more broadly, for preserving and promoting the excellence of ancient Chinese culture.

Based on the research findings, this study systematically examines the differential impacts of natural environmental factors and socio-economic factors on the spatial distribution of agricultural civilization heritage. While the distribution patterns of different types of agricultural civilization heritage exhibit varying degrees of disparity, socio-economic factors generally exert a greater influence than natural environmental factors. From the perspective of the natural environment, climate, hydrological systems, and topography all play distinct roles in shaping the spatial distribution of agricultural civilization heritage. The primary reason lies in the favorable conditions of the middle and lower reaches of the Yellow River, which are characterized by vast plains and hills, a temperate climate, abundant precipitation, and convenient transportation. These factors enable human populations to better adapt to and integrate with the natural environment, thereby fostering the emergence and preservation of agricultural civilization heritage36. Given that the YRB spans four geomorphological units, its natural environment exhibits significant variations. Substantial differences in terrain elevation and precipitation within the basin influence regional water availability and climatic conditions, which in turn shape agricultural production methods and local agricultural cultures. On the other hand, mountainous and plateau regions face harsh climatic conditions, frequent geological hazards, and limited land resources, all of which hinder the formation and development of agricultural civilization heritage. However, provinces such as Qinghai, Gansu, and Sichuan have also developed unique agricultural cultures due to complex historical and ethnic influences21,39. Historically, the northwestern region served as a crucial passage of the Silk Road, facilitating frequent cultural exchanges with the Central Plains. Many ethnic minority areas have preserved distinctive traditional agricultural practices, and successive waves of ethnic migration and integration have further fostered mutual influence among different agricultural traditions, ultimately contributing to a diverse agricultural civilization landscape.

In consideration of socio-economic factors, urbanization, GDP, and population dynamics, it is evident that these elements exert a considerable influence on the spatial distribution of agricultural civilization heritage. The underlying reason is that regions with rich human social activities are more likely to experience cultural agglomeration. Areas with higher urbanization rates tend to exhibit a more concentrated distribution of agricultural civilization heritage resources, reflecting the complex interplay between population density, economic development, and land-use transformation40. Conversely, land-use changes and rural population migration during urbanization processes have led to the decline of certain settlements, thereby impacting the preservation and transmission of agricultural civilization heritage41. Additionally, economically developed regions, where people have higher living standards and greater cultural demand, are often the birthplace or key developmental zones of agricultural civilization, which further shapes the spatial distribution of heritage sites20,42. Thus, the spatial differentiation of agricultural civilization heritage is the result of multiple interacting factors, including historical accumulation, population concentration, cultural demand, and policy support.

To further elucidate our findings, we compare the results with previous studies on the distribution and influencing factors of agricultural civilization heritage. For instance, studies by Li et al.43 and Zhang et al.44 examined cultural heritage in the Yangtze River Basin and the Guangdong-Hong Kong-Macao Greater Bay Area, respectively, revealing a similar pattern of heritage clustering in historically rich agricultural regions and economically developed cities. However, our study extends these findings to the YRB and provides a more detailed spatial classification of APH, ASH, and ALH. Unlike previous research, which primarily emphasized historical and policy influences, our study integrates quantitative spatial methods to systematically depict distribution patterns. Moreover, prior studies have often focused on individual cultural heritage types without considering their interconnections45,46. Our research bridges this gap by illustrating the spatial correlations among APH, ASH, and ALH, linking the spatial distribution of agricultural civilization heritage with human activities and cultural continuity. Unlike studies that rely on descriptive statistics, our approach offers a deeper understanding of the interactions between heritage sites and external factors.

Nevertheless, this study has certain limitations. The three categories of agricultural civilization heritage included in our analysis are limited to national-level or higher-designated heritage sites, while the exclusion of provincial-level heritage may reduce the comprehensiveness and diversity of the research sample. It is recommended that future studies augment the sample size, employ stratified sampling and comparative research methods to enhance the reliability and accuracy of findings, and provide more robust guidance for the conservation of heritage in YRB.

Recommendations:

  1. 1.

    Establishing a Digital Information Database for Heritage Resources in YRB

First, a comprehensive digital information database for agricultural civilization heritage across the nine provinces of the Yellow River Basin should be developed. This database should systematically collect fundamental information on agricultural heritage system, intangible cultural heritage, and settlement heritage, integrating these resources into an online digital museum. Advanced technologies such as Geographic Information Systems (GIS) and 3D scanning should be employed to comprehensively document and disseminate heritage culture, effectively distilling the essence of Yellow River culture and providing a robust data foundation for the systematic protection of heritage resources.

  1. 2.

    Developing a Regional Collaborative Protection System for Heritage Resources in YRB

Second, considering the spatial distribution differences of agricultural civilization heritage, a regional collaborative protection strategy should be implemented at both the basin and provincial levels. Priority should be given to the protection of heritage-rich areas, particularly Shanxi Province and its surrounding regions, where the establishment of "Cultural Heritage Protection Zones" is recommended. A heritage cultural corridor should be constructed to facilitate cross-regional cooperation and resource sharing. Additionally, leveraging the corridor system, an integrated "agriculture-culture-commerce-tourism" model should be explored to promote the synergy between heritage conservation and local industries, ensuring the comprehensive and long-term protection of heritage resources. The research findings demonstrate strong consistency with the core mandates of China’s Yellow River Protection Law regarding systematic conservation of Yellow River cultural heritage. Quantitative analysis of heritage sites’ spatial clustering patterns and distribution characteristics enables precise identification of key protected areas with high heritage density, facilitating the establishment of Agricultural Civilization Heritage Ecological Conservation Zones to mitigate industrial encroachment and preserve the authenticity of traditional settlements and folk culture. In alignment with the national rural revitalization strategy, we propose developing integrated cultural-tourism demonstration corridors, with particular emphasis on Intangible Cultural Heritage Rural Revitalization Belts in heritage-rich areas. A Traditional Artisans Repatriation Program offering housing subsidies and tax benefits to heritage practitioners should be implemented, while prioritizing financial investments in high-density heritage clusters to achieve scale economies and optimize conservation outcomes. Priority should be given to establishing a "Yellow River Basin Agricultural Heritage Conservation Alliance" among the nine provincial regions through enhanced cultural-tourism infrastructure and digital preservation initiatives, while coordinating regional regulations and funding mechanisms.

  1. 3.

    Implementing Differentiated Heritage Protection Strategies

To address the varying protection needs of different regions, a zonal analytical approach should be adopted to examine the diverse influencing factors across areas. This approach would enable the formulation of tailored protection strategies, enhancing the precision and effectiveness of agricultural civilization heritage conservation efforts.

  1. 4.

    Promoting the Living Inheritance of Agricultural Civilization Heritage

Lastly, the living transmission of agricultural civilization heritage should be strengthened by integrating it into ecological agriculture and rural tourism, thereby enhancing its economic and social value. Modern agricultural technologies should be utilized to support the sustainable development of agricultural civilization heritage. Through these innovative approaches, local economic vitality can be stimulated, revitalizing heritage culture and ensuring its continued relevance in contemporary society.

Conclusion

This study analyzed the spatial pattern and influencing mechanisms of agricultural civilization heritage in YRB, the following main conclusions are to be drawn:

  1. 1.

    Spatial Distribution Characteristics: Agricultural civilization heritage in YRB exhibits a notable clustering pattern, with one high-density core area and three sub-core areas. ALH and ASH share a grouped clustering distribution pattern, with ALH exhibiting a slightly higher degree of clustering than ASH. These heritage sites are predominantly located in the middle and lower reaches of the YRB, as well as ethnic minority areas in the upper reaches, such as Qinghai and Gansu. In contrast, APH are randomly distributed, mostly situated near the Yellow River without a distinct clustering pattern.

  2. 2.

    Spatial Relationships: The centroids of the three heritage types all fall within Yan’an, Shaanxi Province, indicating a high degree of spatial balance. The deviation indices between any two heritage types are all less than 0.1, further confirming this balance. Among them, ASH and ALH exhibit the smallest centroid deviation, while APH and ALH have the largest deviation, suggesting that APH is more spatially independent, whereas ALH and ASH show a stronger spatial correlation.

  3. 3.

    Influencing Factors: Among socioeconomic factors, road network density has the strongest explanatory power for the spatial distribution of all three heritage types, followed by urbanization rate and GDP. Among natural environmental factors, annual precipitation has the greatest influence, followed by water network density and average temperature. Other factors, such as water network density, road network density, and nighttime light index, have a relatively weaker impact but should not be overlooked.