Temporal and Spatial Differentiation and Formation Mechanisms of Island Settlement Landscapes in Response to Rural Livelihood Transformation: A Case Study of the Southeast Coast of China

Abstract

Island settlement landscapes exhibit distinctive characteristics, and investigating their spatio–temporal differentiation features and formation mechanisms is crucial for effective landscape conservation. This study selected Qida Village, Beigang Village, and Jingsha Village in Fuzhou City, Fujian Province, China, as representative cases. It constructed an integrated evaluation framework termed “livelihood transformation–two dimensional expansion–three dimensional form” and systematically analyzed the spatio–temporal differentiation characteristics and driving mechanisms of island settlement landscapes under the context of livelihood transformation by integrating multi-source data. Research findings indicate that livelihood transformation significantly affects both the horizontal expansion and vertical evolution of settlement landscapes. Aquaculture-based villages demonstrate a high expansion rate (15.10%) and pronounced vertical differentiation (building height difference ratio of 13.30) due to industrial agglomeration. Tourism service-oriented villages, influenced by policy regulation, exhibit low architectural style heterogeneity (0.35) and a harmonized skyline. Villages experiencing significant out-migration show a high housing vacancy rate (64.70%) and reduced spatial compactness (0.13) due to population decline. The livelihood model drives landscape differentiation through the “population mobility–economic investment–land use” pathway, where capital accumulation and policy constraints emerge as key determinants of spatial form heterogeneity. This study provides a solid theoretical foundation and methodological support for the differentiated governance of island settlement landscapes.

1. Introduction

Island settlement landscapes serve as spatial manifestations of the long-term interactions between human societies and marine environments, embodying distinctive ecological and cultural values. These landscapes hold critical importance for the preservation of coastal cultural heritage and the maintenance of regional identity. In recent years, rapid urbanization and rural revitalization initiatives have triggered profound transformations in the livelihood strategies of island communities [1], leading to notable spatio–temporal differentiation in settlement landscapes. A systematic analysis of this evolutionary process and its underlying driving mechanisms is urgently needed to enable targeted management and sustainable governance of island landscapes.

Existing research on rural landscapes has developed into a relatively systematic and mature field, primarily centered on three key domains: element identification and classification, analysis of influencing mechanisms, and evaluation methodologies. In the area of element identification, Yang Guiqing et al. introduced a “point–line–surface” classification framework [2]; Liu Xiaoyan et al. categorized rural landscapes into three primary types: natural, artificial, and cultural landscapes [3]; and Li Bohua et al. differentiated between material and non-material elements through the theoretical lens of landscape genes [4,5,6]. Concerning influencing mechanisms, scholarly attention has largely concentrated on the effects of functional transformation and policy intervention. For example, Wang Yong et al. highlighted the pivotal role of functional reconstruction in shaping spatial morphology [7]; Du Lanchun et al. identified economic transformation as a dominant driver of landscape evolution [8]; and Fei Shuyan et al., using the Danjia Fishing Village as a case study, demonstrated how industrial transformation, mediated by spatial commercialization, catalyzes the differentiation of cultural landscapes [9]. Regarding evaluation methods, Liu Binyi et al.’s “five-dimensional” model [10], Huang Yingying et al.’s “triple-composite” framework [11], and Wu Xue et al.’s multi-scale GIS model [12] have collectively advanced the field of quantitative landscape assessment.

Notwithstanding these achievements, existing research has been confronted with significant limitations [5]. Primarily, the majority of studies have concentrated on inland rural areas, leaving island settlements—characterized by geographical isolation and resource constraints—largely understudied. Critical phenomena such as livelihood vulnerability, industrial hollowing-out, and tourism-driven capitalization remain insufficiently examined and systematically understood. Secondly, landscape analysis has largely relied on a two-dimensional static paradigm, with limited integration of three-dimensional morphological indicators (e.g., height-to-area ratio, volume-to-space ratio) and the temporal dynamics of human livelihoods. Although some scholars have proposed dynamic [13,14] and multi-dimensional evaluation frameworks [15,16], the empirical application of three-dimensional spatial data remains underdeveloped. Moreover, current evaluation approaches often depend on single methodological pathways, such as perceptual assessments [17,18], which fail to capture the intrinsic mechanisms driving landscape evolution from an integrated “livelihood–population–capital–space” perspective.

In response to the aforementioned limitations, this study advances innovation across three dimensions: methodology, theoretical framework, and case studies. Methodologically, it integrates three-dimensional metrics, such as building height ratio and coastal interface shading rate, with longitudinal livelihood data, thereby overcoming the constraints of traditional two-dimensional static analysis. Theoretically, it establishes an integrated framework linking “livelihood transformation–two dimensional expansion–three dimensional form,” elucidating the multidimensional differentiation characteristics and underlying mechanisms of landscape evolution driven by livelihood change. In terms of empirical application, the study selects representative island settlements in Fuzhou City, addressing the current research gap from a land–sea interaction perspective. This research focuses on two core questions: (1) What are the differentiation patterns of island settlement landscapes across two-dimensional and three-dimensional dimensions during the process of livelihood transformation? (2) How does livelihood transformation drive landscape spatial reconfiguration through multiple pathways, including demographic restructuring, capital allocation, and land use transition? By conducting a synergistic analysis of multi-source data and multi-dimensional indicators, this study systematically uncovers both the external manifestations and internal mechanisms of island settlement landscape evolution, aiming to provide a robust theoretical foundation for the sustainable governance of island landscapes.

2. Research Area

2.1. Livelihood Transformation Paths of Island Residents

Livelihood refers to the relatively stable and continuous strategies and methods individuals employ to secure the necessities of life [19,20,21]. The transformation of livelihood patterns among island communities in Fuzhou has been influenced by the synergistic interactions of resource availability, institutional frameworks, and market-driven forces. This transformation follows a distinct developmental trajectory: from “traditional fishing”, through “industrial differentiation”, to “diversified symbiosis”. In the initial phase, geographical isolation and limited natural resource endowments restricted livelihood options, with fishing serving as the primary means of subsistence [22]. From the Tang and Song dynasties to the Ming and Qing dynasties, fishing practices gradually shifted from extensive, informal methods to more specialized and organized forms of production. Since the 1980s, the decline in marine resources as well as external market shocks have gradually prompted local residents to diversify their livelihoods into aquaculture, non-local wage labor, and tourism-related services. In recent years, national policies aimed at rural revitalization and the establishment of marine economic demonstration zones have further stimulated the growth of the tourism sector, contributing to its emergence as a key component of local livelihood systems. Currently, the livelihood patterns of island residents in Fuzhou City exhibit a diversified structure, primarily categorized into aquaculture-based, tourism service-oriented, and migrant-worker dependent types. This study selects three representative island communities as case studies to investigate the characteristics of their livelihood transitions and landscape differentiation.

2.2. Three Typical Cases of Livelihood Transformation

In this study, three primary livelihood transformation pathways—aquaculture, tourism services, and migrant labor—were initially identified through preliminary field investigations. Building on this foundation, the maximum difference sampling method was employed to select representative cases that exhibited significant variations in livelihood models among island settlements in Fuzhou City. To ensure the typicality of the selected cases, this study further quantified the differences in livelihood transformation by statistically analyzing the employment structure of local residents. Consequently, three distinct settlement models were identified (Figure 1).

• Aquaculture-based village: Qida Village (Figure 2a), recognized as a “National Demonstration Site for Healthy Aquaculture”, has transitioned from traditional fishing to intensive farming of large yellow croakers and abalones since 2000. Currently, it has a permanent population of 3750, with 91% engaged in aquaculture. The substantial profits from seafood farming have significantly increased residents’ incomes. Driven by capital accumulation, the production area has expanded continuously toward the coastal zone, forming an integrated spatial system combining “fishing port–aquaculture zone–processing facility”.
• Tourism service-based village: Beigang Village (Figure 2b), leveraging the strategic development of “Pingtan Island as an International Tourism Island”, was designated a “National Key Village for Rural Tourism” in 2019 and has successfully developed diverse tourism service formats. It currently has a permanent population of 849, with 81% employed in tourism-related services. The village has established an integrated tourism service chain encompassing “homestay–catering–leisure”.
• Migrant-worker dependent village: Jingsha Village (Figure 2c). Due to the scarcity of offshore fishery resources, since 1980, residents have started to migrate out for work on a large scale. Currently, the proportion of out-migrant workers accounts for up to 90% of the registered population, and only 10% of the population has remained in the village over a long period. The serious population loss and industrial hollowing have led to the vacancy rate of village buildings climbing to 64.7%, and the whole village has fallen into a vicious cycle of “population loss–spatial collapse–functional decline”.

3. Methodology

3.1. Design Logic of Multidimensional Evaluation Indexes

To systematically analyze the spatio–temporal differentiation characteristics of settlement landscapes driven by livelihood transformation, this study establishes a set of multi-dimensional evaluation indicators (

Table 1. Calculation methods of landscape character assessment indices.

Type	Index	Formula	Description
Two-Dimensional Spatial Index	Village Expansion Rate	$E={\displaystyle \frac{A_t-A_a}{A_t}}\times 100\%$	In Equation, “$E$” denotes the village expansion rate, where “$A_a$” and “$A_t$” denote the total building floor areas at the beginning and end of the observation period, respectively.
	Housing Vacancy Rate	$H_v={\displaystyle \frac{A_k}{A_k+A_y}}\times 100\%$	In Equation, “$H_v$” denotes the housing vacancy rate, “$A_k$” denotes the total area of vacant housing, and “$A_y$” denotes the total area of occupied housing.
	Public Space Ratio	$R={\displaystyle \frac{A_p}{S}}\times 100\%$	In Equation, “$R$” represents the ratio of public space, where “$A_p$” denotes the land area allocated to public space, and “$S$” refers to the total construction land area of the village
	Village Compactness	$C={\displaystyle \frac{2\sqrt{\pi Sq}}{P}}$	In Equation, “$C$” denotes the compactness of the village, “$Sq$” represents the area of the region, and “$P$” stands for the perimeter of the region.
Three-Dimensional Space Index	Architectural Style Heterogeneity	$M={\displaystyle \frac{Q_n}{Q_n+Q_c}}$	In Equation, “$M$” denotes architectural style heterogeneity, “$Q_n$” represents the number of non-traditional buildings, and “$Q_c$” refers to the number of traditional buildings
	Building Height Difference Ratio	$P={\displaystyle \frac{H_{max}-H_c}{H_c}}$	In Equation, “$P$” denotes the building height difference ratio, where “$H_{max}$” represents the height of the tallest building and “$H_c$” refers to the height of traditional buildings.
	Building Volume Ratio	$V={\displaystyle \frac{V_{max}-V_c}{V_c}}$	In Equation, “$V$” represents the building volume ratio, where “$V_{max}$” denotes the maximum building volume and “$V_c$” corresponds to the volume of traditional buildings
	Occlusion Rate of the Coastal Interface	$I={\displaystyle \frac{L_o}{L}}\times 100\%$	In Equation, “$I$” denotes the occlusion rate of the coastal buildings, “$L_o$” represents the length of coastline segments visually occluded by building clusters, and “$L$” refers to the total length of the village coastline.
	Color Complexity	$CMI={\displaystyle \frac{C}{B}}$	In Equation, “$CMI$” refers to the color mixing index, where “$C$” denotes the number of color categories and “$B$” represents the number of buildings.

) based on an integrated perspective of “livelihood transformation–two dimensional expansion–three dimensional form”. The proposed index system aims to extend beyond conventional two-dimensional static landscape metrics, such as diversity indices [23,24,25], dominance indices [26,27], and evenness indices [28,29]. By integrating time-series data and three-dimensional spatial parameters, the indexes enable the characterization of multi-scale variations in settlement landscapes under evolving livelihood patterns [30,31,32]. Specifically, two-dimensional spatial indicators, including village expansion rate and housing vacancy rate, are employed to quantify horizontal expansion processes and functional degradation trends of settlements. Three-dimensional morphological indicators, such as building height difference ratio and coastal interface occlusion rate, are introduced to reflect vertical spatial differentiation and visual compatibility within the landscape.

3.2. Construction of Index System and Calculation Methods

3.2.1. Two-Dimensional Spatial Index

• Village Expansion Rate: This indicator reflects the driving effect of livelihood transformation on village land use. While the expansion rate is commonly used to study dynamic changes in urban or rural areas [33], it typically relies on total land area for calculation. To improve accuracy, this study revises the calculation approach by using the “building footprint area” as the reference. Specifically, the village expansion rate refers to the growth rate of a village’s building footprint area over a defined time period. The calculation formula is as follows:

$$E={\displaystyle \frac{A_t-A_a}{A_t}}\times 100\%$$

(1)

In Equation (1), “$E$” denotes the village expansion rate, where “$A_a$” and “$A_t$” denote the total building floor areas at the beginning and end of the observation period, respectively.

• Housing Vacancy Rate: Livelihood transformation determines the scale of rural labor outmigration, and population outmigration, in turn, affects the housing vacancy rate of rural housing. A high housing vacancy rate typically manifests as landscape decline and desolation [34]. Therefore, the housing vacancy rate serves as an indicator of the coupling relationship between livelihood transformation and spatial landscape dynamics. This indicator is defined as the ratio of vacant housing area to the total housing area within a village. The calculation formula is as follows:

$$H_v={\displaystyle \frac{A_k}{A_k+A_y}}\times 100\%$$

(2)

In Equation (2), “$H_v$” denotes the housing vacancy rate, “$A_k$” denotes the total area of vacant housing, and “$A_y$” denotes the total area of occupied housing.

• Public Space Ratio: Public space serves as an indicator of the spatial development quality of a village and reflects the intensity of economic input [35]. The quantitative characteristics of public space can reveal the direction and magnitude of capital flow. Under the influence of economic input gradients, three typical evolutionary paradigms of public space have emerged: the spatial capitalization of tourism areas, the polarization of production areas, and the disappearance of migrant-worker areas. This indicator is defined as the ratio of public space area to the total construction land area within the village. The calculation formula is as follows:

$$R={\displaystyle \frac{A_p}{S}}\times 100\%$$

(3)

In Equation (3), “$R$” represents the ratio of public space, where “$A_p$” denotes the land area allocated to public space, and “$S$” refers to the total construction land area of the village.

• Village Compactness: This metric reflects the intensity of land use within a village. Variations in compactness represent the spatial expression of multiple influencing factors, including industrial capital, policy constraints, and population mobility, during the transformation of rural livelihood systems. The indicator quantifies the geometric cohesion of the village’s outer boundary in two-dimensional space, with values ranging from 0 to 1. A value closer to 0 indicates a more fragmented or dispersed spatial configuration, while a value closer to 1 reflects a more compact and contiguous form [36]. The calculation formula is as follows:

$$C={\displaystyle \frac{2\sqrt{\pi Sq}}{P}}$$

(4)

In Equation (4), “$C$” denotes the compactness of the village, “$Sq$” represents the area of the region, and “$P$” stands for the perimeter of the region.

3.2.2. Three-Dimensional Space Index

• Architectural Style Heterogeneity: This index characterizes the continuity of the village landscape and functions as a key metric for quantifying the diversity of architectural styles within rural settlements [37]. Changes in livelihood strategies are closely associated with shifts in land use functions, which subsequently influence architectural forms and reflect the spatial evolution patterns of human–land interactions. Specifically, this indicator is computed as the ratio of non-traditional buildings to the total number of buildings. The calculation formula is as follows:

$$M={\displaystyle \frac{Q_n}{Q_n+Q_c}}$$

(5)

In Equation (5), “$M$” denotes architectural style heterogeneity, where “$Q_n$” represents the number of non-traditional buildings and “$Q_c$” refers to the number of traditional buildings.

• Building Height Difference Ratio: This indicator quantifies the degree of vertical visual contrast within the village landscape. It reflects the intensity of vertical spatial competition, which is influenced by population dynamics that, in turn, shape spatial demands. These dynamics lead to vertical expansion or the implementation of height restrictions. This metric provides insight into the mechanisms through which interactions between population and space influence landscape evolution. Specifically, the building height difference ratio is defined as the relative difference in height between the tallest building in the village and the average height of traditional buildings. The calculation formula is as follows:

$$P={\displaystyle \frac{H_{max}-H_c}{H_c}}$$

(6)

In Equation (6), “$P$” denotes the building height difference ratio, where “$H_{max}$” represents the height of the tallest building and “$H_c$” refers to the height of traditional buildings.

• Building Volume Ratio: This indicator quantifies the relative volumetric difference between the largest residential building in the village and traditional buildings, reflecting the extent of vertical volumetric contrast within the settlement. The expansion of building volume is driven by economic investment, which is in turn influenced by population agglomeration. This highlights the impact of capital accumulation and spatial demand on the spatial evolution of the settlement. The calculation formula is as follows:

$$V={\displaystyle \frac{V_{max}-V_c}{V_c}}$$

(7)

In Equation (7), “$V$” represents the building volume ratio, where “$V_{max}$” denotes the maximum building volume and “$V_c$” corresponds to the volume of traditional buildings.

• Occlusion Rate of the Coastal Interface: As the core landscape element of island settlements, the coastline and its coastal interface directly influence the level of landscape coordination and serve as a key indicator of settlement landscape quality. This indicator reflects the impact of changes in the livelihood model on rural landscape alienation by quantifying the visual shielding effect caused by the vertical expansion of coastal buildings on the natural coastal landscape. The calculation method is as follows: Measure the length of the coastline sections within the village that are visually obstructed by building clusters of three floors or higher, and compute the ratio of this obstructed length to the total length of the coastline. The calculation formula is as follows:

$$I={\displaystyle \frac{L_o}{L}}\times 100\%$$

(8)

In Equation (8), “$I$” denotes the occlusion rate of coastal buildings, “$L_o$” represents the length of coastline segments visually occluded by building clusters, and “$L$” refers to the total length of the village coastline.

• Color Complexity: The unity and harmony of colors are key indicators of landscape quality [38,39]. An excessive number of disorganized colors can lead to visual conflicts. The transformation of livelihood patterns, driven by increased economic investment, changing spatial demands, and shifts in land use functions, influences the color composition of settlements. To quantitatively assess the degree of color complexity, we use the ratio of the number of distinct color tones observed on building facades from the village’s main viewpoint to the total number of buildings. A higher ratio indicates a greater diversity of colors and a more chaotic visual impression, while a lower ratio suggests fewer color types and greater visual coherence. The calculation formula is as follows:

$$CMI={\displaystyle \frac{C}{B}}$$

(9)

In Equation (9), “$CMI$” refers to the color mixing index, where “$C$” denotes the number of color categories and “$B$” represents the number of buildings.

3.3. Data and Sources

The research employs a multi-source data fusion approach, comprising four key components: (1) Historical literature analysis—A systematic review of “Fuzhou City Annals”, “Fujian Fisheries Annals”, and “Pingtan County Annals” is conducted to extract the historical evolution of livelihood patterns. (2) Remote sensing image interpretation—Google Earth satellite imagery from 2011 to 2024 is analyzed to identify dynamic changes in building base areas and trace spatial expansion trajectories. (3) UAV aerial survey—High-resolution orthophotos (0.05 m) are obtained using the DJI Phantom 4 RTK drone, enabling precise extraction of building height, three-dimensional volume, and coastal interface parameters. (4) Field research and validation—Household surveys in three selected villages (N = 217) are carried out to empirically assess housing vacancy rate and examine characteristics of functional transformation.

4. Analysis of the Results

4.1. The Differentiation Characteristics of Two-Dimensional Planar Landscapes

Villages experiencing livelihood transformation demonstrate distinct spatial expansion patterns (Figure 3). Among the selected case villages, Qida Village (aquaculture type) exhibits the highest spatial expansion rate at 15.10% (

Table 2. Summary of indicator calculation results.

Type	Index	Qida Village	Beigang Village	Jingsha Village
Two-Dimensional Spatial Index	Village Expansion Rate	15.10%	8.00%	5.80%
	Housing Vacancy Rate	13.10%	0.70%	64.70%
	Village Compactness	0.29	0.27	0.13
	Public Space Ratio	5.90%	16.60%	1.50%
Three-Dimensional Space Index	Architectural Style Heterogeneity	0.75	0.35	0.28
	Building Height Difference Ratio	13.30	2.00	2.00
	Building Volume Ratio	24.15	10.10	3.59
	Occlusion Rate of the Coastal Interface	64.70%	12.00%	3.90%
	Color Complexity	0.22	0.06	0.10

), characterized by a radial diffusion pattern. Kernel density estimation reveals a high concentration of newly constructed buildings within the buffer zone surrounding the fishing port, indicating the formation of a dense production cluster in this area. Beigang Village (tourism service type) shows a moderate village expansion rate of 8.00%. Kernel density thermal analysis shows that the new buildings have expanded significantly along the coast and are distributed in a banded pattern. Jingsha Village (out-migrant worker type) presents the lowest village expansion rate at 5.80%. Kernel density analysis indicates a dispersed spatial pattern, with new buildings scattered around the village periphery, exhibiting a gradient of decreasing density from the core to the outer areas.

There are notable variations in the housing vacancy rate across villages with differing livelihood models (Figure 4). The results show that Qida Village, which relies primarily on aquaculture, has a housing vacancy rate of 13.10%, placing it between Beigang Village and Jingsha Village in terms of vacancy levels. In Qida Village, the vacant houses are primarily located in the old residential areas, situated away from water bodies and dispersed toward the periphery of the village, forming scattered yet clustered patterns of vacancy. Beigang Village, characterized by a tourism service-based economy, demonstrates the lowest housing vacancy rate at 0.70%. The vacant houses in this village are scattered as isolated units, mainly situated at the periphery of the influence zone of tourism-related facilities. In contrast, Jingsha Village, where the local economy is heavily dependent on migrant worker remittances, exhibits the highest housing vacancy rate of 64.70%. Vacant houses in Jingsha Village are distributed throughout the entire settlement area, forming a dispersed, “honeycomb-like” spatial pattern.

The livelihood patterns of residents have a notable impact on the spatial compactness of rural settlements (Figure 5). Based on calculated compactness values, Qida Village, where aquaculture constitutes the primary economic activity, exhibits the highest compactness index of 0.29. The spatial configuration of the village deviates from a regular geometric shape, forming an irregular “elliptical” structure. Building outlines collectively form a continuous coastal interface, with the pier extending in a jetty-like manner. Beigang Village, whose economy is primarily oriented toward tourism-related services, has a moderate compactness value of 0.27. The presence of preserved farmland within the central area introduces a “morphological depression”, which interrupts the otherwise compact spatial structure. Homestay clusters are arranged in a necklace-like pattern along the boundary of the controlled farmland area. Jingsha Village, characterized by a livelihood model centered on migrant labor, displays the lowest compactness value of 0.13. This low level of spatial compactness corresponds to a highly fractal contour pattern, indicating a dispersed and fragmented distribution of residential buildings.

The public space ratios, in descending order, are Beigang Village, Qida Village, and Jingsha Village. Among them, Beigang Village exhibits the highest public space ratio of 16.60%. The public space system in Beigang is primarily oriented toward tourism services, forming a “dual-core driven” spatial structure. A coastal landscape corridor integrates a snack street and a recreation park, constituting a linear axis of public activity, which together account for 78.00% of the village’s total public space. Qida Village has a public space ratio of 5.90%. Its spatial configuration prioritizes production-related functions, with the fishing pier and coastal strip space comprising 86.00% of the public space. These areas serve multiple functions, including fish trading and recreational use for residents. Other public spaces, such as temples and ancestral halls, account for only 14.00% of the total public space and are scattered within densely built-up areas. Jingsha Village has the lowest public space ratio at 1.50%. Due to population outflow, traditional public spaces such as ancestral halls and threshing grounds have been converted into storage facilities, leading to a significant reduction in the capacity of public spaces to support communal activities.

4.2. The Morphological Differentiation Characteristics of Three-Dimensional Landscape

Qida Village demonstrates the highest level of architectural style heterogeneity, with a heterogeneity index of 0.75. In the northern part of the village, workshops within the aquaculture operation area, warehousing clusters at the fishing pier, and mid-rise to high-rise residential buildings along the main road are interwoven with traditional structures in the village center. This juxtaposition creates a strong visual contrast and leads to spatial fragmentation between the new developments and the historic buildings. Beigang Village exhibits an architectural style heterogeneity index of 0.35. As a key tourism resource, the village has maintained its traditional architectural style, primarily characterized by stone-built houses. Landscape management policies have ensured that newly constructed buildings conform to traditional architectural forms in critical design parameters, such as eave height and roof pitch. Furthermore, contemporary buildings integrate traditional stone materials in their facades. The visual coherence between old and new structures is enhanced through the harmonization of material textures, reflecting an organic renewal strategy that supports the “coexistence of old and new” within the built environment. Jingsha Village displays the lowest degree of architectural style heterogeneity, with an index of 0.28. Traditional buildings largely retain their original layout, with only a limited number of modern structures scattered along the village periphery. Overall, the village exhibits a static and inert spatial pattern, indicating a notably stagnant developmental trajectory.

The livelihood patterns of residents significantly influence the variation in building heights within rural settlements (Figure 6). Qida Village presents the highest building height difference ratio, leading to a notable visual contrast. In comparison, Beigang Village and Jingsha Village exhibit relatively lower building height difference ratios, indicating a more cohesive vertical visual integration. Specifically, the building height difference ratio in Qida Village reaches 13.30, where significant variations in building height are observed. Newly constructed residential buildings generally exceed the conventional height standards for rural structures, resulting in a fragmented skyline characterized by a spatial pattern of “low-scale traditional settlement–abruptly rising modern residential area”. In Beigang Village, the building height regulation has shown evident effectiveness, with a building height difference ratio of only 2.00 between traditional stone houses and newly added constructions. The tourism service-oriented “height-controlled development” model not only aligns with visual expectations of visitors but also contributes to the preservation of the historical skyline rhythm. In Jingsha Village, the building height difference ratio between traditional stone houses and contemporary buildings is also measured at 2.00. Due to a decline in population, construction activities have largely ceased, resulting in minimal alteration to the vertical structure of the settlement. As a parametric representation of three-dimensional form, the building height difference ratio elucidates the spatial coupling characteristics between changes in livelihood patterns and the evolution of village landscapes by quantifying the intensity of vertical interactions between traditional and modern buildings. An excessively high building height difference ratio—such as 13.30 in Qida Village—can lead to the following: ① ecological consequences, such as obstructing coastal ventilation corridors and intensifying the urban heat island effect [40,41]; ② cultural consequences, including the disruption of the historical rhythm of the traditional skyline [42,43].

The building volume ratio reflects variations in building volumes within a settlement. A higher ratio indicates greater heterogeneity in building volumes and less integration of the built environment with the surrounding landscape, while a lower ratio suggests a more uniform distribution of building volumes and better spatial coherence. The results indicate (Figure 7) that Qida Village has the highest building volume ratio of 24.15. This significant variation between the volumes of newly constructed and traditional buildings leads to abrupt spatial transitions and visual discontinuity. Beigang Village exhibits a moderate building volume ratio of 10.10, reflecting observable but less pronounced differences in building volumes. Jingsha Village has the smallest building volume ratio, at 3.59. The relatively uniform building mass contributes to a more balanced and coherent three-dimensional spatial structure. The harmonious scale relationship between buildings, along with the natural spatial transitions, results in an overall compact and orderly appearance, creating a comfortable and cohesive landscape environment.

The coastal interface represents a significant landscape resource and plays a critical role in shaping the environmental quality of coastal settlements. The permeability of this transitional zone is a key factor in preserving high-quality landscape conditions. Research findings reveal that Qida Village exhibits the highest coastal interface occlusion rate, reaching 64.70%. Over half of its coastline is visually obstructed by the presence of high-rise buildings (Figure 8 and Figure 9a). These tall and densely arranged structures along the shoreline act as a visual barrier, which weakens the spatial continuity between the village and the sea. This condition reduces the openness of the coastal space and diminishes the sense of proximity to the marine environment, resulting in a more enclosed spatial perception. In comparison, Beigang Village demonstrates a significantly lower occlusion rate of 12.00%. The spatial configuration of buildings relative to the coastline in this village reflects a more integrated relationship, with a balanced degree of openness and spatial fluidity (Figure 9b). Jingsha Village has the lowest occlusion rate, at only 3.90%. The occluding buildings are sparsely distributed, forming a well-organized, discontinuous interface that provides a more open view of the sea (Figure 9c).

As a representative indicator of visual order, color complexity quantifies the extent to which villagers’ economic activities affect the traditional color management system under varying livelihood models. Among the three village types, Jingsha Village exhibits a color complexity of 0.10. The scattered presence of newly constructed buildings introduces localized chromatic variations, primarily due to the diverse origins of building materials. Within a dominant context of traditional gray tones, the intermittent appearance of modern colors has resulted in a gradual integration of traditional and contemporary visual elements in the village’s overall appearance. Beigang Village displays the lowest level of color complexity, with only four distinct colors identified in its main visual interface and a complexity value of 0.06. The relatively uniform color palette across building facades contributes to a cohesive and harmonious visual structure, indicating a high degree of visual consistency and order. Qida Village presents the highest color complexity, with 10 distinguishable colors observed in its primary visual interface and a complexity value of 0.22. The newly constructed buildings employ highly saturated, non-traditional color schemes, which contrast sharply with the traditional buildings and give rise to a noticeable sense of visual clutter (Figure 10).

5. Discussion

This study constructed a multidimensional evaluation framework integrating “livelihood transformation–two-dimensional expansion–three-dimensional form”, thereby systematically uncovering the spatio–temporal differentiation characteristics and formation mechanisms of island settlement landscapes in the context of evolving livelihood patterns. The findings reveal that diverse pathways of livelihood transformation exert significant influences on settlement spatial structures and visual landscapes through varying dynamics of population mobility, capital investment, and land use patterns. These insights further substantiate the central role of the “livelihood–space” coupling mechanism in driving landscape evolution.

Livelihood transformation directly drives the expansion and contraction of two-dimensional settlement spaces through the restructuring of population structures and altering the direction of capital flows. Qida Village, as an aquaculture-based settlement, has experienced significant spatial expansion and high compactness due to capital accumulation resulting from industrial agglomeration. However, the absence of effective spatial governance has led to disproportionately high building height and volume ratios, generating pronounced visual conflicts. In contrast, Beigang Village, guided by tourism-oriented policies, has achieved a synergistic relationship between capital investment and spatial development. It maintains low land use complexity and a harmonious skyline, with enhanced public space quality through functional integration. Jingsha Village, however, has entered a negative feedback cycle of “space collapse–functional degradation” due to population outflow and capital withdrawal, characterized by high housing vacancy rate, low spatial compactness, and an overall decline in landscape quality. These findings align with the “functional reconstruction–spatial response” mechanism proposed by Wang Yong et al. [7], further underscoring the critical influence of the interplay between policy regulation and capital flow on the formation and maintenance of landscape order. Notably, this landscape differentiation driven by disparities in capital investment echoes transnational patterns observed in cases such as the Ziz Oasis in Morocco—capital-sensitive Type A settlements (e.g., Qida Village and Beigang Village) undergo significant transformation due to capital injection, whereas Type C settlements lacking industrial support (e.g., Jingsha Village) decline or enter a state of “stagnation” following capital withdrawal [44].

Three-dimensional morphological indicators, such as building height difference ratio, coastal interface shading rate, and color complexity, effectively capture visual coordination issues that remain obscured in traditional two-dimensional analyses. In Qida Village, elevated shading rate and pronounced color complexity reflect the disorderly nature of capital-driven construction activities. These factors not only compromise the aesthetic quality of the landscape but may also intensify ecological challenges—such as ventilation obstruction and heat island effects—as well as to erode cultural identity. These findings corroborate the research of Asadpour et al., which suggests that a lack of “coordination” and “compatibility” in visual organization can directly result in “visual pollution” and a “loss of identity” within rural communities [39]. Beigang Village demonstrates the “coexistence of old and new” through proactive style regulation, illustrating how external policy interventions can effectively mitigate the visual disorder caused by unchecked capital expansion. In contrast, Jingsha Village has experienced minimal changes in its three-dimensional morphology due to stalled construction activities, thereby maintaining a relatively coherent visual appearance. However, this stability is accompanied by a decline in spatial vitality, reflecting the broader issue of functional degradation. These findings corroborate the multidimensional evaluation framework of “visual–ecological–cultural” proposed by Liu Binyi et al. [10], underscoring the added value of three-dimensional indicators in enhancing landscape quality assessment.

Significant variations exist in the configuration of public spaces across villages with distinct livelihood types. Tourist-oriented villages demonstrate enhanced public space quality and higher utilization rates through functional integration. In contrast, villages dependent on traditional production or reliant on migrant labor have witnessed a decline or alienation of public spaces due to functional simplification and population outflow. These patterns indicate that public spaces are not merely outcomes of economic investment, but also manifestations of social capital and the effectiveness of spatial governance.

This study offers critical implications for the sustainable governance of island settlement landscapes by emphasizing the need to develop differentiated spatial control strategies tailored to specific livelihood types. For aquaculture-dependent villages, enhanced regulations on building height and architectural aesthetics are essential to mitigate visual conflicts and ecological degradation. For tourism serviced-oriented villages, functional spatial layouts should be optimized while maintaining visual coherence. For villages experiencing significant out-migration, proactive interventions—such as functional revitalization and targeted social policies—are necessary to counteract spatial decay [31]. Implementing context-specific strategies to achieve a dynamic equilibrium between development and conservation represents an effective approach to preserving island-specific landscape identities and advancing rural revitalization.

Innovation and Limitations: This study makes a theoretical contribution to the understanding of the “livelihood–space” coupling mechanism by incorporating three-dimensional morphological indicators and dynamic time series data. These enhancements effectively address the limitations of traditional landscape analysis, which tends to prioritize two-dimensional and static perspectives over three-dimensional and dynamic dimensions. Methodologically, the developed multidimensional evaluation framework demonstrates strong operational feasibility and broad applicability, making it suitable for comparative analysis and long-term monitoring of diverse rural landscape types. Nevertheless, several limitations remain. First, while the framework encompasses both two- and three-dimensional spatial indicators, it lacks sufficient quantification of social perceptions—such as resident satisfaction [43] and visitor experiences [45]. Second, although the study identifies the influence pathways of “population–capital–policy”, it provides limited modeling of the interactive effects among these factors. Future research could enhance the framework by integrating participatory assessment methods and structural equation modeling (SEM), thereby deepening the understanding of the complex dynamics driving landscape evolution.

6. Conclusions

This study investigated the characteristics and mechanisms of spatio–temporal differentiation in island settlement landscapes, focusing on livelihood-driven transformation through a multi-dimensional evaluation framework. The main findings are summarized as follows:

Different livelihood models lead to gradient-based differentiation in landscape characteristics through processes such as population redistribution, capital allocation preferences, and land use reconfiguration. Villages dominated by aquaculture exhibit vertical spatial expansion and functional polarization. In contrast, tourism serviced-oriented villages display higher levels of landscape coherence. Settlements with a high proportion of migrant labor tend to undergo spatial fragmentation and decline.

Landscape differentiation arises from the multi-dimensional interaction among “livelihood strategies–population dynamics–capital flows–spatial organization”. The combined effects of policy interventions and capital accumulation play a critical role in shaping the trajectory of landscape evolution. Livelihood transformation acts as the primary internal driver of settlement landscape change. When this internal driver is regulated by strong external governance mechanisms, the evolution of the landscape tends to be coordinated and structured. In the absence of effective external regulation, however, landscape evolution may manifest as spatial disorganization and visual incoherence.

Three-dimensional morphological indicators, such as the building height difference ratio and the occlusion rate of the coastal interface, demonstrate effectiveness in quantifying visual conflicts and overcoming the limitations inherent in traditional two-dimensional analytical approaches. The three-dimensional morphological and landscape evaluation indicators developed in this study contribute meaningfully to the characterization of three-dimensional landscape qualities in settlement environments and are capable of distinguishing landscape features across diverse regions.

Research suggestions: The governance of island settlements should integrate considerations of industrial development associated with livelihood transitions with the long-term sustainability of the landscape. A balance between development and conservation can be achieved through dynamic monitoring and categorized management strategies. The comprehensive evaluation framework proposed in this study provides a scientific basis for landscape governance in island settlements and offers actionable insights for the preservation of regional island characteristics within the context of rural revitalization strategies.