The Construction and Analysis of a Spatial Gene Map of Marginal Villages in Southern Sichuan

Abstract

With the acceleration of modernization, villages in Southwest China are experiencing spatial fragmentation and homogenization, leading to the loss of traditional identity. Addressing how to balance scientific planning with cultural and spatial continuity has become a key challenge in rural governance. This study takes Xuyong County in Luzhou City as a case and develops a three-tier analytical framework—“genome–spatial factors–specific indicators”—based on the space gene theory to identify, classify, and map spatial patterns in marginal villages of southern Sichuan. Through cluster analysis, common and distinctive spatial genes are extracted. Common genes—such as medium surface roughness (GeneN-2-b), medium building dispersion (GeneA-3-b), and low intelligibility (GeneT-2-b)—are prevalent across multiple village types, reflecting shared adaptive strategies to complex terrains, ecological constraints, and historical development. In contrast, distinctive genes—such as high building dispersion (GeneA-3-a) and linear boundaries (GeneB-1-c)—highlight unique spatial responses that are shaped by local cultural and environmental conditions. The results contribute to a deeper understanding of spatial morphology and adaptive mechanisms in rural settlements. This research offers a theoretical and methodological basis for village classification, conservation zoning, and spatial optimization, providing practical guidance for rural revitalization efforts focusing on both development and heritage protection.

1. Introduction

The unique topography and transportation conditions of the Sichuan Basin have shaped the local people’s production and lifestyle, resulting in distinctive regional characteristics and complex spatial configurations within its villages. With the deepening of globalization, the traditional agriculture-based production and lifestyle are undergoing significant changes [1,2]. During the processes of urbanization and modernization [3,4,5,6], rural cultures in the peripheral areas of the basin are disappearing or being assimilated, with numerous natural villages facing issues such as destructive construction [7,8], aging traditional architecture [9,10], and functional degradation [11,12]. Particularly severe damage has been inflicted upon the overall village landscape and natural environmental features [13,14,15]. Consequently, inheriting and preserving the regional spatial characteristics, cultural identity, and architectural styles of natural villages have become key objectives for rural development in the new era.

Early studies on the spatial morphology of villages primarily focused on descriptive scenarios, emphasizing the classification of visual features [16,17,18]. With advancements in computer and remote sensing technologies, quantitative analysis methods have been gradually integrated into spatial research on villages, facilitating a shift from purely descriptive analyses to a combination of quantitative and qualitative approaches. The research focus has also expanded from morphological descriptions to diverse aspects such as geographical environment [19,20,21], transportation networks [22,23,24], and human decision-making behaviors [25,26,27]. These developments highlight the complexity of village spatial systems while simultaneously revealing the limitations of single-discipline or single-perspective approaches in comprehensively analyzing village’s spatial characteristics. Therefore, establishing an interdisciplinary research framework is essential to systematically understand village spatial structures and their interrelationships with regional culture and the natural environment.

In 2003, Chinese scholar Liu Peilin [28] first proposed the concept of “Cultural Landscape Genes of Traditional Settlements (CLCTS)”, identifying key elements of village spaces from six dimensions: psychological, ecological, esthetic, environmental, cultural, and temporal [29,30,31,32,33,34,35]. This laid the foundation for a comprehensive research system in village spatial studies. Shortly thereafter, Duan Jin [36] advanced this framework by introducing the “space gene” theory, which further focuses on the integrated relationships among humans, environment, and space. Adopting a holistic perspective and dynamic developmental approach, this theory analyzes village spatial structures [37] and examines the complex interactions among space, humans, and the environment [38] through multidisciplinary lenses, thereby more accurately identifying the core contradictions in village spatial transformations.

Currently, the research and practice related to space gene theory are predominantly led by Chinese scholars. Within the hierarchy of village spatial research, the identification and extraction of spatial genes help reveal the unique spatial forms of regional villages and thoroughly analyze their spatial layouts and evolutionary logic [39,40,41,42]. Based on biodiversity theory, Zhou Hui et al. [43] conducted detailed analyses of traditional villages’ dwellings, public buildings, nodal spaces, street networks, and overall spatial features, proposing a series of region-specific spatial patterns. Zhang Zhenlong et al. [44] combined spatial imagery and space syntax to summarize regionally characteristic spatial combination patterns and key spatial elements from the perspectives of overall layout, street networks, and nodal spaces. Xie Mingjing et al. [45] utilized diversity indices to explore the impact of external spatial genes on traditional village spatial features, focusing on railway and road networks surrounding villages, thereby enriching the scope of spatial gene research. Wang Kai et al. [46] proposed a “six-step” theoretical framework—”cognition–scenario–analysis–refinement–evaluation–translation”—systematizing the methodological approach to spatial gene identification and promoting its practical application in urban and rural spatial studies. Cheng Junjie et al. [47] abstracted village spatial forms and spatial gene characteristics to create schematic diagrams, providing robust practical support for the visualization and mapping of village spatial genes.

Although the proposition of space gene theory has laid an important theoretical foundation for systematic research on village spaces, current studies exhibit two notable limitations: First, the research samples predominantly focus on typical traditional villages with well-preserved architectural features, while paying insufficient attention to transitional zones like the periphery of the Sichuan Basin that possess unique geo-cultural characteristics. In addition, methodologically speaking, standardized paradigms for spatial gene identification have yet to be established, particularly regarding multidimensional quantitative analysis and systematic schematic visualization, which hinders accurate characterization of region-specific spatial genes.

Building on previous work, this study takes Xuyong County, Luzhou City, Sichuan Province, as a case study to further investigate the spatial gene characteristics of villages in the southern periphery of the Sichuan Basin. Through a quantitative analysis of five dimensions—natural space, architectural space, transportation space, village space, and topological space—in typical villages of this region, the study aims to identify and refine representative spatial genes and construct a systematic schematic model. This endeavor seeks to elucidate the spatial structural logic and generative mechanisms of villages in the marginal areas of the Sichuan Basin.

2. Research Methodology and Research Subjects

2.1. Space Gene Theory

For the marginal villages of southern Sichuan, the encoding, replication, and expression processes of spatial genes correspond to three fundamental characteristics of rural spatial morphology.

Firstly, spatial genes are encoded through specific spatial elements, arranged according to certain rules and carrying stable information about spatial patterns. In the context of rural spatial research, it manifests as the identification of morphological elements such as proportional relationships and sequential structures. In addition, spatial genes often serve as templates, replicating themselves across different temporal and spatial contexts to produce one or more copies with identical information. This corresponds to the hierarchical differentiation of spatial indicators in village layouts. Finally, the expression of spatial genes dynamically adapts to environmental changes by regulating the combination of different spatial indicators, thereby shaping the village’s responsive spatial evolution.

2.2. Overview of the Study Area

Xuyong County is situated in the transitional mid-low mountainous zone between the southern Sichuan Basin and the northern Yunnan–Guizhou Plateau, featuring diverse landforms such as mountains, hills, and river valleys (Figure 1). The southern region exhibits a well-developed karst topography, while Danxia landforms are common in the north. The terrain slopes from the southeast to the north, with multiple vertical natural zones and an elevation drop of approximately 1650 m. Mountain ranges extend from the Yunnan–Guizhou border into Sichuan through six branches, forming 148 peaks and comprising 212 natural villages. The intricate geographical environment and multi-ethnic settlement patterns have resulted in distinct spatial characteristics among the natural villages that are located at the tri-junction of Sichuan, Yunnan, and Guizhou provinces.

2.3. Characteristics of Research Subjects

After acquiring village location data, this study incorporated river hydrological data, elevation data, and 10 m resolution land-use raster data to analyze the spatial distribution patterns of the villages.

Hydrological Analysis (Figure 2): The majority of natural villages are uniformly distributed along rivers, with over 95% being located within a 2 km buffer zone of river systems. This distribution reflects the traditional Chinese settlement principle of “settling near water and resting by water.”

Land-Use Composition (Figure 3): Forest land dominates the study area, accounting for 80.9% of the total area, followed by cultivated land (14.4%) and grassland (3.4%). Water bodies, wasteland, and wetlands collectively constitute less than 1%. Consequently, subsequent analyses of village land-use patterns will emphasize the differentiation of village resources.

Elevation Profile Analysis (Figure 4): A north–south cross-sectional analysis reveals that the northern part of the study area features a relatively flat terrain, with villages being primarily distributed at elevations between 200 m and 600 m. The central transitional zone exhibits mixed landforms, with villages being concentrated in lower-lying areas. In contrast, the southern region has higher elevations and horizontally distributed mountain ranges, with most villages being situated in valleys and a few scattered along mid-slopes.

Selection of Typical Villages: Based on the principles of representativeness (covering ancient, modern, and contemporary construction periods), completeness (encompassing all spatial elements within administrative village boundaries), and diversity (reflecting varying cultural backgrounds and natural landform characteristics), 18 villages were selected for study (Figure 5).

2.4. Data Source and Processing

This study employed a multi-source data collection approach, primarily involving four types of fundamental data. Satellite imagery from August 2022 was acquired through the Google Earth platform, with subsequent extraction of buildings and roads being based on this foundation. The Digital Elevation Model (DEM) data was obtained from the PALSAR sensor of the ALOS satellite, with a precision of 12.5 m. The research integrated the 2021 global land cover data (featuring 11 land categories) released by ESA with a 10 m resolution (published in October 2022). The administrative boundary data of Xuyong County and residential settlement data were sourced from the “Xuyong County Map”, edited by Chengdu Cartographic Publishing House under the supervision of Sichuan Civil Affairs Department. Village-level administrative boundary data was derived from the China Administrative Village Boundary Dataset, published by the Scientific Data Registration and Publishing System of the Geographic Remote Sensing Ecological Network. All data underwent uniform preprocessing procedures, including georeferencing and projection transformation on the ArcGIS platform, culminating in the acquisition of a comprehensive vector dataset for the entire Xuyong County through spatial clipping.

3. Extraction of Spatial Gene Elements

This study adopts the space gene theory proposed by Chinese Professor Duan Jin from Southeast University, establishing a three-level analytical framework encompassing the “micro-, meso-, and macro-”scales. Five dimensional elements—natural space, built environment, border space, mobility space, and network space—are identified as genomes. Through quantitative and qualitative analyses of 13 specific indicators across 13 associated spatial factors, this study constructs a “genome–spatial factor–specific indicator” research system to identify and extract representative spatial gene elements. This approach systematically maps the spatial gene patterns of marginal villages in southern Sichuan.

3.1. Natural Space

3.1.1. Terrain Morphology

We used the terrain morphology as the foundational framework for identifying village spatial patterns and developmental characteristics. Based on field surveys and satellite imagery analysis, the geographical landforms of villages can be categorized into four types (Figure 6).

Canyon-Type Villages: These villages are predominantly located between mountain ranges and characterized by deeply incised valleys with significant vertical elevation differences. The height variation between mountain peaks and valley floors typically ranges from 400 m to 600 m, resulting in pronounced topographic relief and visual enclosure.

Flat-Basin-Type Villages: Situated in relatively flat terrain with minimal slopes, these villages exhibit internal elevation differences that are generally below 200 m. A representative example is Sankuaishi Village, where the gentle topography provides favorable land resources and convenient transportation, making it suitable for agricultural cultivation and village expansion.

Gentle-Slope-Type Villages: Distributed across moderately steep foothills, these villages feature noticeable terrain slopes with internal elevation differences of approximately 400 m. Yunshanba Village exemplifies this type, where buildings are typically arranged in a terraced spatial pattern along the mountainside due to topographic constraints.

Mid-Slope-Type Villages: Primarily concentrated at the base and mid-slopes of mountain ranges, these villages occupy steeper terrain, with overall elevation differences ranging from 800 m to 1000 m. Shuinaopu Village is a typical case, demonstrating a dispersed spatial layout that extends along the mountain contours.

3.1.2. Surface Roughness

The surface roughness serves as a critical metric for evaluating terrain undulation and irregularity, effectively characterizing geomorphological features and the topographic complexity of village areas.

Villages exhibiting high surface roughness values demonstrate pronounced topographic variations, as exemplified by Hongyan Village, where the mountainous terrain partitions the settlement into distinct sectors. Consequently, residential clusters are predominantly concentrated in relatively flat, low-elevation zones. Conversely, villages with low values display minimal internal elevation differences, featuring planar topographies that permit relatively unconstrained settlement development.

Through standardized deviation classification of elevation data, the case study villages were systematically categorized into three distinct types (Figure 7).

3.1.3. Feature Typology

The feature typology reveals the spatial interdependence between natural environments and village settlements. Among all natural elements, forest land constitutes the dominant category (55–90%). To mitigate the overwhelming influence of forest coverage and highlight inter-village variations, this study focuses specifically on the distances between settlements and other natural elements.

The distribution patterns of key elements reveal distinct variations: cultivated land serves as the secondary element, accounting for 8.4% to 38.9% of the composition. Grassland additionally appears only in limited instances, with a maximum representation of 22.2%, while the presence of other elements remains negligible across the sample.

Through correlating natural elements with built structures, this study identifies five distinct spatial patterns based on predominant element adjacencies (Figure 8), which are described below.

Forest-Encircled Type: Surrounded by extensive forests and constrained by rugged terrain, these villages have limited arable land and rely primarily on forestry. In Xiangba Village, settlements follow mountain contours, with scattered farmland and a forestry-based economy.

Mixed Forest–Cultivation Type: Situated on gentler terrain, these villages feature a “cropland core, forest periphery” layout. In Liangsanpo Village, farmlands surround the central settlement, transitioning into forests and supporting a mix of agriculture and sustainable forestry.

River-Dependent Type: These villages align with rivers, integrating hydrology into daily life. Xindian Village focuses on irrigated crops like rice and corn, while Sankuaishi Village emphasizes aquaculture. The settlement patterns adapt to river flow, reflecting aquatic cultural traits.

Composite Grassland–Cultivation Type: Found in mountainous areas, land-use is vertically zoned. In Shuinaopu Village, farmland lies on lower slopes, pastures on mid-elevations, and woodlands higher up—supporting integrated agro-pastoral–forestry systems.

Cultivation-Concentrated Type: Defined by expansive farmland, these villages are structured around intensive agriculture. In Hongyan Village, dwellings align along farmland edges, forming a compact settlement focused on specialized crop production.

3.1.4. Feature Diversity

The feature diversity is a key indicator for measuring the complexity of land-use structures and ecosystem heterogeneity in villages, primarily reflected by the Shannon Diversity Index (SHDI). A higher SHDI value indicates richer land-use types and more complex landscape structures in the village, while a lower value suggests a more homogenized landscape pattern.

The studied villages are largely covered by extensive forest resources, so their diversity is mainly reflected in elements such as cropland and grassland. A higher proportion of cropland signifies stronger agricultural dependence, with land-use tending toward intensification and efficiency. Villages that are dominated by a mix of grassland and forest are often more suitable for grazing, foraging, or understory economy.

Using the standard deviation classification method, the studied villages were categorized into three types (Figure 9).

3.2. Built Environment

3.2.1. Construction Intensity

The construction intensity is primarily measured by the built-up area of villages, reflecting the development intensity of construction land and spatial utilization efficiency. It reveals the degree of land-use intensification and spatial expansion characteristics within villages. Most villages exhibit relatively balanced construction scales, without showing excessive clustering or extreme dispersion in their spatial distribution.

Villages with a high construction intensity, such as Yunshanba Village, leveraged the advantages of their flat terrain in the early stages to develop agriculture as a pillar industry, thereby expanding their built-up areas. In contrast, some villages are constrained by their topography, significantly limiting the potential for land development and resulting in a low construction intensity.

Calculations show that the built-up area of villages ranges from 6.02 hm² to 24.64 hm². Using the mean–standard deviation method, these villages can be classified into three levels (Figure 10).

3.2.2. Building Dispersion

Building dispersion refers to the spatial distribution density of structures within a village, reflecting its spatial layout and land-use patterns. An analysis of the building distribution reveals that the degree of dispersion is closely correlated with surrounding mountains, valleys, and other topographical features.

In some villages, such as Shanguan Village and Lianmeng Village, buildings are separated by mountain ranges and constructed along slopes, resulting in linear or clustered settlement patterns. Conversely, villages that are located in relatively flat terrains without intersecting rivers or mountains—like Yueming Village—exhibit more uniform building distributions. Additionally, certain villages, such as Xindian Village and Lianmeng Village, are segmented by water systems, with structures built along rivers or lakes.

To quantify the building dispersion, the mean distance from each building to the village center was calculated. Based on the mean–standard deviation method, dispersion levels were classified into high, medium, and low categories (Figure 11).

3.2.3. Architectural Style

The architectural style refers to the distinctive design features—including form, typology, and ornamentation—of landmark structures and traditional buildings that embody cultural identity and landscape significance.

In terms of roof types, buildings are primarily classified into flat roofs and pitched roofs, with the latter exhibiting greater diversity. Besides the predominant overhanging gable roof, other variants include the saddle roof and suspension roof. Structurally, buildings fall into two categories: timber-framed structures and brick–concrete hybrids.

By crossreferencing roof and structural types, this study determined whether a building qualifies as traditional (Figure 12). Furthermore, based on the Traditional Architecture Proportion (TAP), the studied villages were categorized into three architectural typologies (Figure 13).

3.3. Border Space

3.3.1. Aspect Ratio

This study introduces the Minimum Bounding Rectangle (MBR) concept to more precisely characterize the length–width ratio (α) in village planar morphology. The MBR delineates village boundaries through an intuitive geometric framework, effectively revealing the relationship between regularity and complexity (Figure 14). Village forms are classified into three distinct patterns based on their α values:

Linear (α ≥ 2): Villages exhibit elongated ribbon-like configurations, typically distributed along rivers, roads, or valleys. Their administrative boundaries appear narrow and elongated, demonstrating strong coupling with terrain features and transportation corridors.

Semi-linear (1.5 ≤ α < 2): Villages display transitional belt cluster forms, maintaining a general linear distribution trend while preserving relatively compact core areas. Boundary extensions may occur due to topographic constraints or land-use characteristics.

Compact (α < 1.5): Villages assume clustered block-like patterns with regular administrative boundaries and concentrated forms. These are typically found in relatively flat terrains, showing minimal influence from external geographic conditions.

3.3.2. Fractal Dimension

The fractal dimension (D) serves as a quantitative measure for characterizing the morphological features of village boundaries. The principle states that for a given area, a longer boundary corresponds to a higher fractal dimension value, indicating greater boundary complexity and highly fragmented morphological characteristics. Conversely, lower fractal dimension values reflect more regular, smooth boundaries and a more stable overall spatial structure.

Given the non-normal distribution of village fractal dimension data, the natural breaks (Jenks) classification method was employed to categorize villages into two distinct groups (Figure 15).

3.4. Mobility Space

3.4.1. Network Circularity

Network circularity (θ) serves as a key metric for evaluating the closure degree of road networks, characterizing the presence of cyclic paths within the transportation system and reflecting its connectivity and accessibility.

A lower value indicates that the road network tends to exhibit a tree-like or radial structure, with numerous branches but fewer internal circuits, resulting in limited traffic options and stronger path dependency. Conversely, a higher value suggests that the road network features a grid-like or multi-loop closed structure with high internal permeability, enabling pedestrians and vehicles to flexibly switch between multiple routes, thereby enhancing the richness of traffic connections and the convenience of travel (Figure 16).

3.4.2. External Node Count

The external node count (k) reflects the openness of village transportation networks. A higher value indicates a more open road network with stronger connections to surrounding townships or major traffic arteries, thereby exhibiting greater economic vitality and regional synergy. In contrast, villages with fewer external nodes often demonstrate higher internal self-sufficiency, where traffic organization primarily serves local needs, resulting in relatively weaker external accessibility (Figure 17).

3.5. Network Space

3.5.1. Integration

Integration refers to the degree of aggregation or dispersion between a given spatial element and others within a system, measuring its capacity to attract traffic as a destination and reflecting its centrality in the overall network.

A higher global integration value indicates that the village occupies a core position in the transportation network, with high accessibility that facilitates resource flow and population concentration, thereby driving economic development. Conversely, a lower value suggests a more enclosed road structure, limiting significant population convergence and resulting in slower resource circulation (Figure 18).

3.5.2. Intelligibility

Intelligibility refers to the degree of consistency between the local spatial structure and the global spatial structure within a village’s spatial system. A higher intelligibility value indicates a more legible road network and smoother traffic organization, suggesting greater developmental potential for the village. Conversely, lower intelligibility reflects a more convoluted road network, restricted movement for residents, and relatively limited growth prospects.

The intelligibility data of villages also did not follow a normal distribution; thus, they were classified into two categories using the natural breaks method (Jenks) (Figure 19).

4. Spatial Gene Map

4.1. Construction of the Map

The spatial gene map adopts a three-tiered structure of “genome–spatial factor–specific indicator” to comprehensively encompass genetic types, graphic representation, and coding systems across all levels. It employs a combination of color schemes, alphabetic codes, and numeric identifiers to ensure scientific rigor, systematic organization, and readability (Figure 20).

At the core of the map lies the classification of five major genomes: natural space (GeneN), built environment (GeneA), border space (GeneB), mobility space (GeneM), and network space (GeneT). They collectively shape the rural settlement’s spatial resilience and sustainability potential, thereby constituting critical dimensions that require systematic integration in rural revitalization strategies.

Each genome is subdivided into several spatial factors (14 in total), such as the terrain morphology (GeneN-1), architectural style (GeneA-1), and aspect ratio (GeneB-1), which serve as intermediate analytical categories. Furthermore, each factor can be disaggregated into 2 to 5 specific indicators, ultimately forming a typology of 40 indicators. These indicators are encoded using a consistent alphanumeric format—e.g., GeneA-3-b denotes the second subtype under building dispersion (GeneA-3)—ensuring systematic readability and precise referencing.

To more intuitively visualize the expression patterns of spatial genes, this study further normalized the quantitative data and analyzed it using Origin 2021 software, generating a chord diagram to represent the abundance distribution of village spatial genes (Figure 21). The chord diagram employs multicolored arcs for visual representation, where the left semicircle (from GeneN-2 to GeneT-2) corresponds to various factors of the spatial gene evaluation system, while the right semicircle represents 18 typical case study villages. The connecting chords elucidate co-occurrence relationships between spatial gene factors and villages, with their widths being proportional to the strength of correlation.

Among the specific factors, GeneA-1 (architectural style) exhibited the longest arc segment, followed by construction dispersion GeneA-3 (building dispersion)—both classified as high-abundance spatial genes. This indicates their prevalent expression across multiple village spatial morphologies and their role as key contributors to inter-village differentiation. In contrast, GeneT-2 (intelligibility) showed the shortest arc segment, suggesting relatively uniform levels of road network legibility and circulation efficiency among the studied villages.

At the village level, Yunshanba, Shuinaopu, and Shanguan villages demonstrated overall high expression levels, with multiple indicators significantly outperforming others after normalization. This reflects their complex spatial organization and strong regional adaptability. Conversely, Xiangba Village exhibited limited spatial gene expression, with most indicators falling within low-value ranges, indicating simplified and homogeneous spatial structures.

4.2. Typology of Case Study Villages

In the formation of village spatial genes, both natural, environmental, and human factors have exerted varying degrees of influence on southern Sichuan villages, shaping their spatial patterns through long-term evolution and adaptation. Therefore, this study further analyzes different village types to explore the shared and distinctive spatial characteristics that have developed during their evolution.

Quantitative data were imported into SPSS 27.0.1 for cluster analysis, classifying the case study villages into three types. The spatial gene indicators for each village type were normalized to determine their typological characteristics (Figure 22).

Different village types are color-coded in the radar chart, where each data point represents the overall strength of a specific indicator for that village type. Collectively, all points of the same color reflect the spatial gene characteristics of the corresponding village type. For instance, Type 3 exhibits a notably higher value in GeneB-2-a (High-D), whereas Type 2 shows prominence in GeneB-2-b (Low-D). This distinction suggests that GeneB-2-a serves as a key discriminant indicator for differentiating Type 3 from other village types.

Additionally, the specific village types and corresponding spatial gene indicators were visualized using Origin to generate a Sankey diagram (Figure 23), revealing the commonalities and distinctions in spatial genes and illustrating the distribution of different genetic factors across village types.

The nodes represent spatial gene indicators, with their widths being proportional to their gene expression intensity. High-flow nodes correspond to common genes, indicating their widespread presence and substantial influence across multiple village types. In contrast, low-flow nodes represent specific genes, demonstrating strong regional exclusivity or typological particularity. The connecting edges illustrate gene-sharing patterns, with their topological relationships revealing transmission mechanisms of genes among different village types.

The findings from this analysis are presented below.

Common Spatial Genes: Most spatial gene indicators exhibited connections across two or more village types, forming a shared genetic set that reflects the spatial morphology of southern Sichuan’s marginal villages. These genes embody collective adaptive strategies that have been developed in response to the natural topography, ecological constraints, and historical evolution, while also providing a structural framework and path dependence for spatial formation. The most prominent shared genes include the following:

Medium surface roughness (GeneN-2-b);

Medium building dispersion (GeneA-3-b);

Low intelligibility (GeneT-2-b).

These indicators were widely distributed across all three village types, demonstrating strong universality and stability. Their consistent spatial expression reflects the common challenges that are faced by marginal villages in southern Sichuan, such as a complex terrain, limited construction resources, and irregular traffic networks. As core spatial genes, they shape the macro-level spatial layout and micro-level organization of village elements, holding pivotal roles in the spatial gene map.

Distinctive Spatial Genes: In contrast, distinctive genes are those with significant expression in only one village type or low-frequency expression in others. These genes reflect unique spatial organizational patterns, architectural layouts, or traffic modes, representing village-specific adaptations to terrain, resource endowments, cultural traditions, or social structures. Examples include the following:

High building dispersion (GeneA-3-a);

Linear boundary (GeneB-1-c);

Inward-focused (GeneM-2-c).

Such genes serve as key identifiers for specific village types, highlighting the localized mechanisms of spatial adaptation formed through “site-specific and context-responsive” strategies. They contribute to the diversity and complexity of village spaces, enriching typological expressions and providing targeted indicators for future village classification and conservation efforts.

5. Discussion

This study contributes to the growing body of research on rural spatial morphology by introducing a gene-based analytical framework for villages in the marginal mountain regions of southern Sichuan. The construction of a “genome–spatial factor–specific indicator” system allows for both a macro-level typological classification and a micro-level diagnosis of spatial structure, enabling a more systematic understanding of spatial heterogeneity in traditional rural settlements. The identification of both common and distinctive spatial genes provides valuable insights into how geography, land-use constraints, and cultural patterns co-shape a village’s form, offering an applicable methodological reference for similar mountainous or fragmented contexts.

Nevertheless, the space gene theory, while advantageous in offering structural clarity, is inherently limited in its capacity to represent temporal dynamics and the socio-political processes that drive spatial transformation. Compared to more actor-centered frameworks such as Actor–Network Theory (ANT) or Social–Ecological Systems (SESs), which emphasize governance, resilience, and the interplay of human and non-human agents, the spatial gene model tends to fixate on formal abstraction and spatial typification. This methodological reduction may overlook important factors such as local agency, institutional evolution, and community-led change.

Moreover, the reliance on cross-sectional spatial data constrains the capacity to understand spatial evolution over time. The incorporation of diachronic spatial datasets—such as multi-temporal satellite imagery, historical cadastral maps, and archival records—could help reveal how spatial genes adapt, fragment, or persist in the face of socio-economic transitions, migration, and policy shifts. These longitudinal perspectives would significantly enhance the interpretative capacity of space gene theory, contributing to more resilient and adaptive rural planning strategies.

Future research should explore the integration of spatial gene mapping with participatory ethnographic methods and socio-spatial interaction models. By aligning formal spatial analysis with lived experiences and cultural memory, researchers can better account for the mutual constitution of space, identity, and everyday practice. Such interdisciplinary fusion—drawing from cultural geography, vernacular studies, and rural sociology—could bridge the current gap between form and meaning and provide a more comprehensive framework for heritage-informed rural transformation.

6. Conclusions

This study pioneers a spatial gene model for natural villages in the marginal zones of southern Sichuan, employing a three-tier analytical framework (“genome–spatial element–characteristic indicator”) to systematically extract and classify core spatial components. Through cluster analysis, the research delineates a regionally representative spatial gene typology that elucidates both shared spatial characteristics and distinctive morphological adaptations that are shaped by topographic, ecological, and socio-cultural determinants.

The scholarly contributions are threefold. Primarily, this study establishes an analytical methodology for diagnosing and mapping village spatial structures in complex mountainous environments. Secondly, it develops a spatial coding system that facilitates village classification and conservation zoning, providing scientific support for planning decisions within rural revitalization initiatives. Thirdly, it constructs a theoretical framework for spatial structure analysis through a “gene logic” lens, thereby advancing academic discourse on the interplay between spatial form and environmental–cultural systems.

Nevertheless, the inherent limitations of spatial pattern typification must be acknowledged. While methodologically sound, the spatial gene model risks oversimplifying the region’s profound cultural heterogeneity and intricate architectural traditions. Overly rigid classification may lead to preservation approaches that privilege form over meaning, potentially compromising the living cultural continuity of villages. Preservation strategies should therefore maintain flexibility, demonstrate place sensitivity, and incorporate local knowledge systems.

Future research should prioritize three directions:

Integrating multi-temporal spatial data to trace spatial gene evolution and assess the resilience of spatial configurations.

Adopting participatory ethnography to decode the localized knowledge systems and cultural logics that are embedded in spatial structures.

Synthesizing interdisciplinary perspectives—particularly cultural landscape theory, political ecology, and vernacular heritage studies—to develop more robust analytical frameworks for rural transformation.

In conclusion, this study establishes a more nuanced, evidence-based, and culturally attuned paradigm for spatial planning and conservation in mountainous rural areas, with its methodological innovations carrying implications for sustainable rural development research across broader geographical contexts.