Spatio-Temporal Evolution and Conflict Diagnosis of Territorial Space in Mountainous–Flatland Areas from a Multi-Scale Perspective: A Case Study of the Central Yunnan Urban Agglomeration

Abstract

Investigating spatio-temporal differentiation patterns of land-use conflicts in mountainous and flatland regions provides critical insights for optimizing spatial regulation strategies and advancing sustainable regional development. Using the Urban Agglomeration in Central Yunnan (UACY) as a case study, the production–living–ecological space (PLES) was classified through land-use functional dominance analysis based on 2010–2020 geospatial datasets. Spatio-temporal evolution patterns and mountain–dam differentiation were analyzed using spatial superposition, dynamic degree analysis, transfer matrices, and geospatial TuPu methods. A multi-scale conflict index incorporating landscape metrics was developed to assess PLES conflict intensities across spatial scales, with contribution indices identifying key conflict-prone spatial types. Analysis revealed distinct regional differentiation in PLES distribution and evolutionary trajectories during 2010–2020. Forest Ecological Space (FES) and Agricultural Production Space (APS) dominated both the entire study area and mountainous zones, with APS exhibiting particular dominance in dam regions. Grassland Ecological Space (GES) and Other Ecological Space (OES) experienced rapid conversion rates, contrasting with stable or gradual expansion trends in other space types. Change intensity was significantly greater in mountainous zones compared to flatland area (FA). PLES conflict exhibited marked spatial heterogeneity. FA demonstrated substantially higher conflict levels than mountainous zones, with evident scale-dependent variations. Maximum conflict intensity occurred at the 4000 m scale, with all spatial scales demonstrating consistent escalation trends during the study period. ULS, FES, and WES predominantly occurred in low-conflict zones characterized by stability, whereas APS, Industrial and Mining Production Space (IMPS), RLS, GES, and OES were primarily associated with high-conflict areas, constituting principal conflict sources.

1. Introduction

Rapid urbanization has induced spatial expansion and land fragmentation, resulting in diminished land-use efficiency and exacerbated land-use conflicts [1]. The global scholarly and policy communities have increasingly focused on urbanization-driven land-use conflicts [2], prompting the development of spatial conflict research frameworks that integrate both “land conflict” and “land-use conflict” concepts. Contemporary research predominantly examines stakeholder interest conflicts alongside spatial conflicts stemming from land-use/cover changes or suboptimal spatial allocations [3]. Land-use conflict fundamentally embodies spatio-temporal competition among stakeholders competing for spatially overlapping resources [4], manifested through evolving contradictions between land-use types and functional priorities [5,6]. As a composite system integrating subspaces with land resources at its core [7], land space organization requires strategic planning. China’s recent spatial planning initiatives prioritize rational configuration of “production—living—ecology spaces” (PLES) to harmonize ecological preservation, productive efficiency, and residential livability [8]. Emerging as a comprehensive zoning paradigm [9], the PLES framework aligns with sustainable development objectives that balance economic growth, social progress, and ecological conservation [9]. Scholarly advancements in PLES research span functional identification [10,11], spatial structure evaluation [12], formation mechanism analysis [13,14], spatio-temporal coupling [15,16], evolutionary characterization [11,17], spatial simulation [18], and systemic interrelationship studies [19,20]. Compared to conventional land-use conflicts, PLES conflicts offer a more holistic framework for national strategic zoning [21]. Addressing these conflicts through spatial optimization remains a critical challenge in contemporary planning research [21]. Current spatial conflict studies predominantly focus on human-economic development-induced conflicts, emphasizing typological diagnosis and governance strategies [22], while evolving planning demands necessitate multidimensional analytical frameworks [21,23].

Existing PLES conflict methodologies encompass both qualitative and quantitative approaches. Quantitative methods, termed absolute conflict analysis [24], include landscape pattern-based metrics derived from geographical and landscape ecological perspectives [21]. These incorporate spatial heterogeneity indices such as conflict type–pattern–process indexes [25], patch complexity–vulnerability–stability models [26], and pressure–exposure–stability frameworks [27]. Although objectively applicable across scales, these methods often overlook socioeconomic factors, relying primarily on composite index-map superposition [28]. Complementary approaches employ composite index systems integrating “pressure-state-response” [29], “possibility-exposure-consequence” [30], “source-receptor-effect” [1], and spatial conflict indices [24], utilizing weighted superposition analysis to diagnose conflict intensity through discriminant matrices [21]. Research from a geographical perspective has found that unique spatial characteristics (such as spatial geometry) can exacerbate territorial spatial conflicts by influencing spatial expectations [31]. Qualitative methods adopt sociological techniques like participatory surveys [32] and logical frameworks to analyze stakeholder rights conflicts, yet face limitations in scalability and subjective bias [33].

At the same time, although there is limited research abroad specifically on PLES conflicts, extensive studies have been conducted on land-use conflicts. Research on land-use conflicts has explored various dimensions and contexts globally. In Ethiopia’s agro-pastoral regions, large-scale land investments have been shown to disrupt traditional livelihoods, leading to the reallocation of land resources and exacerbating tensions between local communities and investors [34]. In the Brazilian Amazon, agricultural expansion, particularly soybean cultivation, has triggered significant social and environmental conflicts, including deforestation and the infringement of indigenous land rights [35]. In Switzerland, a typology framework for peri-urban land-use conflicts identified the multifunctionality of land use as a key driver of conflicts during urbanization [4]. In Germany, land-use conflicts at national and regional levels were examined from a stakeholder perspective, emphasizing the critical role of multi-stakeholder participation in policy-making [36]. The relationship between forest multifunctionality and infrastructure has also been studied, revealing that conflicts between ecological conservation and resource development are a primary manifestation of land-use disputes [37]. Compatibility analysis and spatial assessment have been used to identify synergies and conflicts among ecological, socio-cultural, and economic values, providing a scientific basis for land-use planning [38]. In the context of global crises, land-use conflicts have been exacerbated by climate change, resource scarcity, and globalization, particularly in developing countries [39]. Methodologically, participatory mapping has been introduced as a tool for identifying land-use conflicts, highlighting the importance of stakeholder involvement in conflict identification and resolution [40]. Additionally, multi-criteria decision-making (MCDM) methods have been employed to identify areas prone to land-use conflicts, offering quantitative tools for conflict management in land-use planning [31]. These studies collectively underscore the complexity of land-use conflicts and the need for integrated, multi-stakeholder approaches to address them effectively.

Previous studies have achieved substantial progress in land space conflict research [41]. However, two critical limitations remain: (1) Constrained by the natural geographical constraints and developmental disparities inherent to mountain–dam systems, spatial resource allocation exhibits systemic imbalance, with persistent mismatches between ecological capacity and socioeconomic demands. Current methodologies often lack differentiated analytical frameworks for mountain versus FA, leading to imprecise conflict identification and incomplete characterization of regional conflict dynamics. (2) Although landscape ecological indices effectively diagnose ecological risks from spatial development [42], their well-established scale sensitivity induces significant conflict intensity variations across spatial scales. These variations are routinely disregarded in studies employing landscape pattern indices, particularly when quantifying conflict levels through conventional analytical approaches.

China’s mountainous terrain, encompassing plateaus and hills, constitutes over 60% of its terrestrial territory [43]. Exemplifying this landscape, Yunnan Province contains 94% mountainous area, where dam regions sustain dense populations, advanced infrastructure, industrial concentration, and robust economies, contrasting with mountainous zones characterized by sparse demographics and high biodiversity [44]. These spatially constrained FA facilitate intensive urbanization and agricultural production within ecologically sensitive water networks, while mountainous regions function as vital ecological barriers and agricultural bases [45]. The escalating contradiction between resource exploitation and conservation underscores the imperative for mountain–dam conflict analysis and coordinated spatial governance in the UACY.

2. Materials and Methods

2.1. Study Area

The Urban Agglomeration in Central Yunnan (UACY) is located in the central part of Yunnan Province, spanning from 100°43′ to 104°49′ east longitude and 23°01′ to 27°04′ north latitude in southwestern China. Covering an area of approximately 111,400 km², it includes Kunming City, Qujing City, Yuxi City, Chuxiong Yi Autonomous Prefecture, and seven counties in the northern part of Honghe Hani and Yi Autonomous Prefecture, totaling 49 counties (cities and districts), as shown in Figure 1. By the end of 2020, the region had a resident population of 21.96 million and a GDP of CNY 150.74 billion. As one of the 19 major urban agglomerations in China, UACY serves as the economic, political, and cultural hub of Yunnan Province, while also playing a vital role in food production and ecological conservation. The primary land-use types are woodland and arable land, accounting for 32.8% of Yunnan’s total land area, 40.42% of its arable land, and 45% of its construction land. With elevations ranging from 312 to 4313 m and a relative height difference of 4001 m, over 80% of the region consists of mountainous and hilly terrain. This area exemplifies a typical mountain–flatland interlocking zone, marked by significant disparities in natural resource endowment and socio-economic development between flatland and MA. These characteristics pose considerable challenges for land space protection, development, and utilization. Analyzing the spatial patterns and evolutionary trends of the national territory in this region is crucial for addressing the complexities arising from the interplay between mountain and flatland.

2.2. Data Sources

The data used in this study include socioeconomic datasets, land-use data, administrative district boundaries and DEM for 2010 and 2020. Land-use data comes from the Yunnan Provincial Land Use Annual Change Survey database. The administrative divisions and names are based on information available at the end of 2020. A detailed summary of the data sources is presented in

Table 1. Data sources and references.

Data Aspect	Data Content	Time	Data Source	Data Declaration	Reference
Land Use	Arable land, woodland, city, town, village, etc.	2010, 2020	Yunnan Provincial Land Use Survey database	Vector (Shapefile Format)	Yunnan Provincial Department of Natural Resources. (2010, 2020). Land Use Survey Report.
Administrative District	Administrative district boundary	2020	Resource and Environment Science and Data Center	Vector (Shapefile Format)	Resource and Environment Science and Data Center. (2020). Administrative Boundary Data.
Flatland	Flatland boundary	2014	Yunnan Provincial Land Use Survey database	Vector (Shapefile Format)	Yunnan Provincial Department of Natural Resources. (2014). Land Use Survey Report.
Terrain	DEM	2020	Resource and Environment Science and Data Center	Raster (30 m Grid)	Resource and Environment Science and Data Center. (2020). Digital Elevation Model Data.

. All spatial data were unified using the China Geodetic Coordinate System 2000 (CGCS2000) with the Gauss–Krüger projection, adhering to the national standard 3-degree zoning and the 1985 National Elevation Datum, the grid resolution is uniformly 250 m × 250 m.

2.3. Methods

2.3.1. PLES Classification System

This Guided by the established classification principles, this study systematically identifies and classifies PLES functions through spatial functional dominance analysis. To enhance the precision of primary functional identification, a comprehensive PLES classification system (

Table 2. Correspondence between PLES and land-use types in UACY.

PLES		Corresponding Land-Use Type
Primary Classification	Secondary Classification
Production Space (PS)	Agricultural Production Space (APS)	Cultivated Land, garden land, facility agricultural land, ridge
	Industrial and Mining Production Space (IMPS)	Mining land, scenic spots and special land, transportation land, hydraulic construction land
Living Space (LS)	Urban Living Space (ULS)	Cities and towns
	Rural Living Space (RLS)	village
Ecological Space (ES)	Forest Ecological Space (FES)	woodland
	Grassland Ecological Space (GES)	grassland
	Water Ecological Space (WES)	Water area and swamp
	Other Ecological Spaces (OES)	Saline alkali land, sandy land and bare land

) was developed by administering structured expert questionnaires and integrating existing research frameworks [46,47]. The system operates at two hierarchical levels: the primary classification distinguishes production, living, and ecological spaces, while the secondary classification refines these categories to accommodate land multi-functionality and operational practicality. Multi-temporal land survey data (2010 and 2020) were processed into standardized grid datasets for spatial analysis, following coordinate system normalization and topological verification protocols.

2.3.2. Dynamic Index

The dynamic degree metric, conventionally employed to quantify land-use change rates [48,49], is adapted in this study to characterize spatio-temporal structural transformations within land systems. Specifically, we analyze both single and comprehensive dynamic degrees for the three designated space types (production, living, ecological) across the study area during 2010–2020. These metrics are operationalized as follows:

The Single Dynamic Degree quantifies the annualized change rate of specific land spatial types within discrete temporal units. This metric captures both the magnitude and velocity of areal transformations during the study period, formulated as:

$$K=(U_b-U_a)/U_a\times {\displaystyle \frac{1}{T}}\times 100\%$$

(1)

where K is a certain type of dynamic degree during the study period; U_b and U_a are the area of this land type at the beginning and end of the study, respectively; T is the study period, in a.

The Composite Dynamic Index model quantifies the integrated transformation rate of Production–Living–Ecological Spaces (PLES) across temporal units within the study area, formulated as:

$$LU=\left[{\displaystyle \frac{{\displaystyle \sum _{i=1}^n\Delta L{U}_{i-j}}}{2{\displaystyle \sum _{i=1}^{n}L{U}_i}}}\right]\times {\displaystyle \frac{1}{T}}\times 100\%$$

(2)

where LU is the comprehensive dynamic degree; LU_i is the area of type i land space type in the starting year; ΔLU_i−j is the absolute value of the area converted from type i land space type to non type i during the study period, i ≠ j; T is the length of the study period.

2.3.3. Transition Matrix of PLES

The land spatial transfer matrix, structured as a two-dimensional array [50,51], systematically documents land-use transitions through temporal comparisons: with rows denoting origin land-use types at the initial period and columns indicating destination types at the terminal period. Matrix elements quantify either the absolute area or proportional conversion between specific PLES categories during the study interval.

$$T=\left(\begin{array}{cccc}t11& t_{11}& \dots & t_{1n}\\ t_{21}& t_{22}& \dots & t_{2n}\\ \dots & \dots & \dots & \dots \\ t_{n1}& t_{n2}& \dots & t_{nm}\end{array}\right)$$

(3)

where the diagonal element t_nn represents the area or proportion of type n that remains unchanged throughout the study period. The non-diagonal element t_mn, m ≠ n represents the area or proportion from type m to n.

2.3.4. Geo-Information TuPu

The geospatial information TuPu methodology effectively visualizes land-use/cover change dynamics and is employed to quantify structural-quantitative transformations within PLES [52]. Utilizing Formula (4) [53], we generated PLES transition maps spanning 2010–2020, documenting both spatial-type influx and efflux patterns between temporal phases. Spatio-temporal potential analysis of these cartographic outputs revealed expansion–contraction trajectories across PLES categories, thereby systematically characterizing spatio-temporal pattern evolution within the study area [54].

$$T_{2010–2020}=100T_{2010}+T_{2020}$$

(4)

T_2010–2020 refers to the map code, T₂₀₁₀ and T₂₀₁₀ are the code of the PLES map unit in 2010 and 2020, respectively, APS (code 11), IMPS (code 12), ULS (code 21), RLS (code 22), FES (code 31), GES (code 32), WES (code 33), and OES (code 34).

2.3.5. Analysis of the Conflict of PLES

The Pressure–State–Response (PSR) framework is a key model in landscape ecology for assessing human activities, landscape conditions, and societal responses. In constructing a conflict index, PSR helps analyze factors contributing to spatial conflicts:

Pressure: Human activities like urbanization, deforestation, and pollution alter the landscape and create imbalances between land uses, contributing to conflicts. The conflict index accounts for the intensity and distribution of these pressures.

State: The current condition of the landscape (e.g., land cover, biodiversity, fragmentation), revealing how pressures lead to conflicts. The conflict index reflects changes in the landscape’s state, highlighting areas prone to conflict.

Response: Societal measures such as land-use planning and conservation efforts to address pressures. The conflict index evaluates how effective these responses are in mitigating conflicts, with weak responses indicating unresolved issues.

Territorial spatial conflict emerges when anthropogenic interventions and natural variability induce discordance in land spatial structure, triggering significant spatio-temporal pattern transformations [55]. The ecological risk level inversely correlates with land-use conflict intensity: ecosystems with optimized structural configurations exhibit lower conflict magnitudes. Regions exhibiting diminished landscape disturbance intensity correlate with reduced anthropogenic stress and consequently attenuated conflict pressure, while heightened disturbance amplifies these pressures proportionally. Such pressure dynamics directly manifest the anthropogenic impacts on land systems. The conflict evaluation framework aligns with landscape ecological risk assessment through three diagnostic dimensions: Landscape Complexity (Risk Source) reflects the degree of human disturbance to natural landscapes, such as the expansion of construction land and the intensity of agricultural activities. Vulnerability (Risk Receptor) reflects the sensitivity and resilience of ecosystems to external disturbances, such as the distribution of ecologically sensitive areas and levels of biodiversity. Fragmentation (Risk Effect) reflects the degree of landscape fragmentation, such as the dispersion of forest patches and the degradation of wetlands. These metrics, adapted from established methodologies [56,57,58], quantitatively characterize Territorial spatial conflict index (TSCI) intensity in the UACY via Formula (5).

$$TSCI=SCI+SFI-SSI$$

(5)

where TSCI is the land spatial conflict index, and SCI, SFI and SSI are the three factors of spatial complexity, spatial vulnerability and spatial stability, respectively.

(1)

Spatial Complexity Index

By assessing the integrated spatial pressure exerted by target units and their adjacent neighborhoods, this study employs the Area-Weighted Mean Patch Fractal Dimension (AWMPFD) metric to quantify landscape configuration complexity, with fractal geometry characteristics serving as the spatial complexity indicator.

$$AWMPFD={\displaystyle \sum _{i=1}^m{\displaystyle \sum _{j=1}^n\left[\frac{2\ln (0.25P_{ij})}{\ln (a_{ij})}\times \left(\frac{a_{ij}}{A}\right)\right]}}$$

(6)

where P_ij is the perimeter of the patch, a_ij is the patch area, and A is the total landscape area.

(2)

Spatial vulnerability index

The spatial vulnerability index is used to reflect the compressive capacity of patches in land space. The higher the vulnerability index is, the weaker the compressive capacity is, and it is vulnerable to external influence to a large extent. The formula is as follows:

$$SFI={\displaystyle \sum _{i=1}^m{\displaystyle \sum _{s=1}^r{f}_{is}}}\times {\displaystyle \frac{a_{is}}{A}}$$

(7)

where f_is is the landscape vulnerability of s landscape types in i space type; a_is refers to the landscape patch area of s landscape types in class i space type; A represents the area of spatial units, m represents the number of spatial types, n represents the number of patches, and r represents the types of landscape. Vulnerability reflects the exposure of risk receptors. Refer to previous research results [59,60]. The vulnerability index is determined as follows: rural living space, urban living space and industrial and mining production space are 1, FES is 2, grassland ecological space is 3, agricultural production space is 4, water ecological space is 5, and other ecological space is 6. Normalize the calculation results of each spatial element to between 0 and 1.

(3)

Spatial Stability Index

Land-use stability can be measured using the landscape fragmentation index, and the formula is as follows:

$$SSI=1-PD=1-{\displaystyle \frac{n_i}{A}}$$

(8)

where PD is plaque density; n_i is the number of patches of type i space type in each space unit; A is the area of each space unit. The greater the PD value, the higher the degree of spatial fragmentation, the lower the stability of its spatial landscape unit, and the lower the stability of the corresponding regional ecosystem. The calculation results of each spatial unit are standardized between 0 and 1.

(4)

Multi-scale method

Given the well-documented scale sensitivity of landscape patterns [61,62], this study implements a bisectional scaling protocol to systematically evaluate multi-scale PLES conflict dynamics. Scale gradations were determined through systematic consideration of: (1) the spatial domain extent of the UACY, (2) PLES typological complexity, and (3) operational constraints in data acquisition and computational processing. The hierarchical grid system maintains embedded spatial relationships across scales, with 250 m × 250 m PLES classification data (2010–2020) serving as the baseline spatio-temporal dataset.

$$Sn=2^n\times S$$

(9)

where S_n is the scale size, n is the number of levels, S is the basic grid scale of 250 m × 250 m, and n = 1, 2, 3, 4, 5, and 6 are selected to obtain six scales of 500 m × 500 m, 1 km × 1 km, 2 km × 2 km, 4 km × 4 km, 8 km × 8 km, 16 km × 16 km as the moving window for calculating the conflict index of land spatial use. A hexagonal tessellation framework was generated using ArcGIS 10.8 Data Management tools, with landscape units partitioned according to these six hierarchical resolutions [56]. Subsequent landscape pattern indices were computed through integrated analysis using ArcGIS 10.8 and FRAGSTATS, followed by spatial conflict index derivation through our established quantification model. All resultant values were normalized to the [0,1] interval using min–max scaling.

(5)

Conflict Revealed Comparative Advantage (CRCA)

To systematically examine the correlation between spatial conflict patterns and PLES structural configurations, and to delineate the differential impacts of PLES typologies on conflict manifestation, this investigation quantifies the spatial-type distributions across four conflict categories. The Revealed Comparative Advantage (RCA) methodology [63,64] is employed to evaluate categorical contributions of distinct spatial types, culminating in the development of a PLES’ CRCA index model:

$$CRCA=(Cij/Cit)/(Ckj/Ckt)$$

(10)

where CRCA is the conflict contribution of i space types at the time of j conflict, C_ij is the area of i space types at the time of j conflict, C_ij is the area of i space types at the time of j conflict, C_ki is the area of I space types in the region, and C_kt is the area of a region. CRCA greater than 1 indicates that the spatial type has a comparative advantage in the conflict, which is a strong contribution. If it is less than 1, it does not have a comparative advantage, which is a weak contribution [11].

3. Results

3.1. Evolution of PLES Structure

Analysis of PLES structural composition reveals sustained dominance of Ecological Spaces FES and APS across the entire study region and mountainous zones, whereas dam systems exhibit APS as the predominant spatial type. The spatial distribution demonstrates marked mountain area (MA) and flatland area (FA) differentiation, particularly evident in APS and FES allocations. As quantified in 2010, APS occupied 27.99% (UACY), 23.02% (MA), and 64.71% (FA) of spatial coverage, contrasting sharply with FES distributions at 54.74%, 60.81%, and 9.81%, respectively (Figure 2). Secondary spatial categories, including ULS, RLS, GES, and WES, further exhibited measurable regional disparities.

Temporal Dynamics (2010–2020): APS exhibited divergent trajectories, with expansion in mountainous zones (2.3% increase) contrasting contraction in dam systems (4.1% reduction), maintaining regional equilibrium at 27.9%. FES demonstrated progressive growth across all units: regional (54.74% → 59.34%), mountainous (60.81% → 65.89%), and dam systems (9.81% → 10.89%). IMPS, ULS, RLS, and WES showed universal expansion, notably IMPS with 165% regional growth (0.91% → 2.41%) and an 87.4% surge in FA (2.69% → 5.04%). Conversely, GES experienced halved regional coverage (9.38% → 4.30%), while OES plummeted 89.9% regionally (2.68% → 0.27%), with mountainous systems exceeding a 90% reduction (2.82% → 0.29%).

Table 3. Dynamic degree of different space types.

Type	Dynamic Degree
	UACY	MA	FA
APS	0.00%	0.34%	−1.01%
IMPS	6.21%	6.73%	4.66%
ULS	3.10%	4.39%	2.79%
RLS	2.98%	3.25%	2.57%
FES	0.77%	0.77%	0.99%
GES	−11.85%	−12.07%	−5.99%
WES	1.38%	2.06%	0.55%
OES	−89.12%	−87.11%	−124.20%

delineates the single dynamic degree of each spatial type during 2010–2020. APS exhibits structural equilibrium at the regional scale (dynamic degree = 0%). However, subregional variations emerge between MA and FA, with values of 0.34% (MA) and −1.01% (FA), demonstrating low-intensity incremental growth and controlled contraction, respectively. IMPS display accelerated transformation rates across all units: 6.21% (UACY), 6.73% (MA), and 4.66% (FA) reflecting high-magnitude spatial restructuring.

ULS and RLS exhibit moderate transition velocities (3–5% dynamic degree range). FES maintains minimal variation (<1% across UACY, MA, and FA), while WES shows intermediate dynamics (1.38% UACY, 2.06% MA, 0.55% FA). GES experience rapid depletion, evidenced by dynamic degrees of −11.85% (UACY), −12.07% (MA), and −5.99% (FA). The most drastic reduction occurs in OES, with values plummeting to −89.12% (UACY), −87.11% (MA), and −124.20% (FA), indicative of a catastrophic decline.

3.2. Characteristics of PLES Transfer

Analysis of PLES transformations from 2010 to 2020 reveals comprehensive dynamic degrees of 2.36% (UACY), 2.37% (MA), and 2.33% (FA), indicating intensive spatial restructuring. To elucidate these changes, transition matrix analysis was conducted using the PLES transfer area matrices (

Table 4. UACY global PLES transfer matrix from 2010 to 2020 (ha).

Year 2010	Year 2020
	APS	IMPS	ULS	RLS	FES	GES	WES	OES	Outflow
APS	2,387,184	97,625	22,209	66,057	500,092	10,758	32,274	2069	731,084
IMPS	10,068	40,646	13,182	11,339	17,562	6920	1326	498	60,895
ULS	2341	2912	73,383	3589	2808	1312	442	17	13,421
RLS	3200	6280	9103	183,351	1948	204	210	6	20,951
FES	470,583	88,725	4025	20,050	5,379,816	111,352	18,418	6018	719,171
GES	185,531	20,492	1674	3908	542,786	276,406	8456	6197	769,044
WES	15,534	3255	1534	1113	8181	3225	154,041	142	32,984
OES	45,231	8293	635	1429	157,952	68,389	1920	15,227	283,848
Inflow	732,488	227,583	52,362	107,484	1,231,329	202,161	63,046	14,946	2,631,399

,

Table 5. Spatial transfer matrix of PLES in UACY MA from 2010 to 2020 (ha).

Year 2010	Year 2020
	APS	IMPS	ULS	RLS	FES	GES	WES	OES	Outflow
APS	1,681,187	59,536	5872	36,132	451,704	8447	15,078	1876	578,644
IMPS	6224	27,113	4273	6424	15,011	5552	785	443	38,712
ULS	724	744	7877	1429	1841	634	124	12	5509
RLS	2237	3278	1618	108,071	1685	154	98	5	9075
FES	431,302	83,175	2808	16,700	5,304,521	107,535	16,925	5837	664,281
GES	173,524	18,923	993	3277	535,843	270,756	7222	6045	745,828
WES	6995	1373	171	323	6357	1795	78,925	121	17,136
OES	37,487	7184	263	1107	149,676	65,635	1624	14,202	262,976
Inflow	658,492	174,213	15,999	65,392	1,162,118	189,752	41,855	14,339	2,322,160

and

Table 6. Transfer matrix of PLES in UACY FA from 2010 to 2020 (ha).

Year 2010	Year 2020
	APS	IMPS	ULS	RLS	FES	GES	WES	OES	Outflow
APS	705,997	38,089	16,337	29,925	48,388	2311	17,196	193	152,440
IMPS	3844	13,534	8909	4915	2550	1369	541	55	22,184
ULS	1617	2168	65,507	2159	967	678	318	5	7912
RLS	963	3002	7485	75,279	264	50	111	0	11,876
FES	39,282	5550	1217	3350	75,295	3817	1494	181	54,890
GES	12,007	1569	681	631	6942	5649	1234	152	23,217
WES	8539	1882	1363	790	1824	1430	75,116	21	15,849
OES	7743	1109	371	322	8276	2755	296	1025	20,872
Inflow	73,996	53,370	36,364	42,092	69,211	12,409	21,191	607	309,239

), complemented by spatial geoscience information TuPu mapping (Figure 3). Key findings demonstrate the following:

APS demonstrates considerable inflow and outflow magnitudes, with dynamic patterns that elude comprehensive capture through quantitative analysis alone. Transition matrix analysis delineates substantial divergence in internal transition patterns: inflows predominantly derive from FES, followed sequentially by GES and OES, while outflows predominantly transition to FES, followed by IMPS and RLS. Geospatial clustering reveals inflows concentrated in eastern Qujing and western Chuxiong, contrasting with outflows dispersed along urban peripheries. MA exhibits net inflow dominance, whereas FA demonstrates net outflow prevalence. Regional comparisons identify elevated CES intensity in FA relative to mountainous zones, with inverse patterns observed for FES coverage. Mechanistically, APS diminution in FA was associated with urbanization processes, rural revitalization initiatives, and infrastructure expansion, while mountainous reductions correlate with ecological conservation policies such as farmland-to-forest conversion.

IMPS demonstrates net inflow predominance, with inflows predominantly originating from APS and FES, concentrated within the urban agglomeration’s core zones. Outflows primarily transition to FES, APS, IMPS, and RLS, distributed across the central-southern sectors of the urban agglomeration. Both MA and FA exhibit inflow surpluses, though mountainous regions show significantly higher incremental gains. Spatial redistribution patterns diverge regionally: MA prioritizes conversions to FES, RLS, and APS, while FA predominantly shifts IMPS to ULS, RLS, and APS, with PES and APS serving as principal sources. These dynamics indicate that IMPS expansion relies critically on transformations of agricultural and forest ecological spaces, likely associated with industrialization and resource exploitation processes. The larger IMPS transition magnitudes observed in FA further substantiate their role as primary industrialization hubs.

ULS exhibited net inflow dominance, with primary inflows originating from APS, IMPS, and RLS. These inflows concentrated in Kunming’s urban core and county-level residential zones, while outflows predominantly transitioned to RLS, FES, IMPS, and APS, spatially dispersed around Kunming’s periphery. Regional conversion patterns diverged significantly: MA primarily converted ULS to RLS and FES, whereas FA shifted ULS to RLS, IMPS, and APS. Spatial source analysis revealed MA contributed higher FES proportions to ULS conversions compared to FA, while FA exhibited markedly greater RLS and APS contributions. ULS expansion primarily relied on the transformation of agricultural production and rural living spaces, reflecting intensified urbanization processes. The larger-scale ULS transitions observed in FA further confirmed their status as primary urbanization zones.

RLS exhibited pronounced net inflow dominance with rapid spatial expansion. Inflows primarily originated from APS, FES, and IMPS, displaying spatially dispersed sourcing patterns across all subregions. Outflows predominantly transitioned to ULS, IMPS, and APS, concentrated in rapid urbanization zones surrounding Kunming’s metropolitan core. Regional disparities emerged in conversion patterns: FA demonstrated markedly larger RLS-to-ULS conversion areas compared to MA, while MA showed significantly higher retention of APS in spatial transitions. Source analysis revealed mountainous RLS conversions were predominantly fueled by APS and FES, contrasting with FA where APS and IMPS served as primary contributors. These dynamics reflect intensive rural spatial restructuring through bidirectional exchanges with agricultural and ecological spaces, likely associated with rural land-use optimization and ecological conservation measures.

FES exhibited the most substantial transitions within UACY, demonstrating net inflow dominance. Inflows primarily derived from GES, APS, and OES, concentrated in northern Chuxiong, southern Yuxi, southern Qujing, and Yuanmou-Dongchuan areas. Outflows dispersed across the entire region, with APS, GES, and IMPS dominating total and MA transitions, while FA showed APS and IMPS as primary outflow targets. Source analysis revealed MA predominantly sourced transitions from GES and APS, contrasting with FA where APS constituted the principal origin. These dynamics reflect the effective implementation of forest conservation policies in MA, coupled with interzonal cultivated land exchanges between mountain and dam systems.

GES exhibited net outflow dominance across the entire region, MA, and FA. Inflows primarily originated from FES and OES, spatially concentrated in the southern Honghe River, eastern Luoping, and northern Dongchuan. Outflows predominantly transitioned to FES and APS, with northern regions serving as the core transition zone. Mountain–dam divergence emerged in transition dynamics: MA received inflows mainly from FES and OES, while FA incorporated additional PES contributions. Outflow patterns further differentiated these zones—MA prioritized FES followed by APS, whereas FA exhibited reversed dominance (APS > FES). These spatial disparities suggest mountainous GES transitions align with ecological restoration objectives, contrasting with FA where agricultural expansion drives grassland conversion.

WES exhibits net inflow dominance, with primary contributions originating from APS and FES. These inflow sources demonstrate spatially dispersed patterns across the study area, paralleled by analogous conversion dynamics between APS and FES. Comparative analysis reveals distinct spatial sourcing: MA exhibits significantly higher APS contributions than FA, while FES inputs demonstrate inverse spatial patterns. These transitional characteristics likely correlate with water resource scarcity in FA, compounded by regional water management strategies and ecological conservation efforts.

OES exhibited pronounced transformation intensity across the entire region, demonstrating net outflow dominance. Inflows predominantly originated from GES, FES, and APS, spatially concentrated in eastern areas with severe rocky desertification (e.g., Luoping and Shizong). Outflows primarily transitioned to FES and GES, displaying analogous spatial clustering. These dynamics reflect bidirectional exchanges between OES and adjacent ecological spaces (FES/GES), likely associated with ecological restoration initiatives and land-use optimization strategies targeting degraded landscapes.

From 2010 to 2020, the PLES transformations in UACY exhibited marked spatio-temporal heterogeneity. Transition matrix analysis delineates distinct spatial flow patterns driven by urbanization, industrialization, ecological conservation, and land-use optimization. The FA demonstrated stronger responsiveness to urbanization and industrialization pressures, whereas MA exhibited pronounced sensitivity to ecological conservation mandates. These findings establish critical empirical support for optimizing territorial spatial planning and ecosystem governance frameworks in ecologically fragile mountainous regions.

3.3. PLES Conflict Diagnosis

3.3.1. Overall Characteristics of Multi-Scale Conflicts

The comprehensive conflict indices of PLES in the UACY were systematically quantified across six spatial scales (500 m, 1 km, 2 km, 4 km, 8 km, 16 km) for three temporal intervals (Figure 4). At the regional scale, the mean conflict values demonstrated an ascending trajectory from 2010 to 2020, particularly within the 500–8000 m scale range. For instance, conflict intensity escalated from 0.469 to 0.529 (+12.8%) at 500 m resolution and from 0.618 to 0.681 (+10.2%) at 4000 m resolution. Notably, the 16,000 m scale exhibited stability, while the 1000 m scale uniquely displayed a 6.4% decline (0.551 → 0.516). MA mirrored regional trends but with marginally lower conflict magnitudes. The 500 m and 8000 m scales recorded 12.2% and 11.2% increases, respectively, while stability persisted at 16,000 m and a minor reduction occurred at 1000 m. In contrast, FA exhibited significantly elevated conflict levels, surpassing both regional and MA values across all scales. All dam-area scales demonstrated marked intensification (2010–2020), exemplified by 500 m scale escalation from 0.535 to 0.617 (+15.3%) and 8000 m scale growth from 0.636 to 0.723 (+13.7%). Spatio-temporal analysis identified a threshold effect at 4000 m resolution, where mountain–dam differentiation peaked across triennial data, indicating maximum PLES conflict intensity. Ascendant regional conflict values reflect intensifying land-use tensions, particularly pronounced at smaller or mesoscales (500–8000 m). This scale-dependent divergence manifests in FA consistently outperforming regional and MA conflict levels, signifying heightened spatial friction. Conversely, MA maintained comparatively subdued conflict indices, suggesting relative spatial stability.

3.3.2. The Spatio-Temporal Evolution Characteristics of Multi-Scale Conflicts

The spatio-temporal evolution of PLES conflicts across mountain–dam regions was investigated using an explicit land-use conflict evaluation model. Spatial conflict indices in UACY (2010–2020) were assessed through cumulative frequency curve analysis, revealing an inverted “U” shaped evolutionary pattern [65]. Building upon established methodological frameworks [55,58], the conflict intensity values within the study area were classified into four discrete tiers using an equal-interval approach: Stable and Controllable (UC) [0, 0.25), Basically Controllable (BC) [0.25, 0.5), Moderate Conflict (MC) [0.5, 0.75), and Severe Conflict (SC) [0.75, 1.0].

(1)

General Characteristics

MC predominated across all study periods, though the proportion of SC increased substantially, indicating intensifying spatial tensions. Conflict intensity followed a distinct spatial hierarchy: FA exhibited the highest levels, followed by the entire region, with MA demonstrating the lowest intensity. Notably, SC proportions in FA significantly exceeded those in both the regional aggregate and MA, confirming heightened conflict severity in lowland urbanization cores.

In 2010, MC constituted the primary conflict type in UACY (Figure 5). The proportions of BC, UC, and SC displayed substantial scale-dependent variability, with significant structural divergences across spatial resolutions. Geospatial analysis revealed strong multi-scale consistency in conflict distribution: UC clustered predominantly in southwestern sectors, BC and UC exhibited synergistic spatial coupling within these areas, MC concentrated heavily in northern regions, and SC localized primarily in eastern zones. By 2020, conflict-level distributions diverged markedly from 2010 patterns. UC and BC proportions decreased significantly, while MC and SC escalated substantially. Despite MC maintaining dominance, SC, BC, and UC exhibited pronounced scale-specific fluctuations. Comparative spatial analysis demonstrated shifting conflict distributions relative to 2010 baselines, though multi-scale spatial patterns maintained consistency (Figure 6).

(2)

Global Evolution Characteristics

At the 500 m scale, MC and BC dominated across the entire region, exhibiting extensive spatial distribution. UC clustered predominantly in southwestern sectors, while SC accounted for merely 1.88%, localized primarily in eastern zones. Scale-dependent divergence emerged in conflict proportions with increasing spatial resolution: UC and BC followed “U”-shaped trajectories, declining from 500 m to 4000 m (reaching 1.15% and 16.21%, respectively) in western and northwestern regions, then rebounding to 4.96% and 26.00% beyond the 4000 m threshold.MC demonstrated biphasic dynamics—escalating from 500 m to 2000 m (peaking at 72.22%) through BC/UC conversions in western/northwestern areas, then progressively declining to 51.54% at 16,000 m via transitions to other conflict types in eastern/southwestern sectors. SC exhibited monotonic growth, increasing to 17.49% at 16,000 m, driven predominantly by eastern regional contributions. Comparative analysis revealed marked temporal shifts: by 2020, MC at 500 m increased to 60.72% (concentrated in northern/northeastern areas), while BC and UC declined to 29.78% and 6.43%, respectively. SC concurrently rose to 6.43%, primarily aggregating in eastern/southeastern zones (Figure 7 and Figure 8).

(3)

Evolution Characteristics in MA

MA exhibited broad consistency with regional conflict evolution patterns, though demonstrating regional divergences during scale transitions. At the 500 m scale, conflict characteristics mirrored regional dynamics: Moderate Conflict (MC: 46.02%) and Basic Control (BC: 37.54%) predominated, with Severe Conflict (SC: 1.78%) remaining minimal. Scale-dependent variations emerged as follows: UC and BC followed U-shaped trajectories, decreasing from 500 m to 4000 m (UC: 1.23%; BC: 17.60%) before rebounding to 5.31% and 27.06% at 16,000 m. MC displayed biphasic dynamics, rising from 500 m to 2000 m (71.01%) through BC/UC conversions, then declining to 51.72% via transitions to other conflict types. SC exhibited monotonic growth, progressively increasing to 15.92% at 16,000 m. From 2010 to 2020, UC and BC decreased substantially, while MC increased significantly and SC rose moderately. These trends mirrored regional patterns, with all conflict types showing pronounced scale-dependent divergence as spatial resolution increased.

(4)

Evolution Characteristics in FA

The FA exhibited distinct conflict evolution patterns compared to regional and MA trends, particularly in SC dynamics. At the 500 m scale, MC (63.51%) dominated, followed by BC (30.92%), with UC: 2.93% and SC (2.64%) being marginal. Scale-dependent transitions revealed divergent trajectories: UC and BC followed U-shaped trajectories, decreasing from 500 m to 4000 m (UC: 2.93% → 0.61%; BC: 30.92% → 6.36%) before rebounding to 2.17% and 17.39% at 16,000 m, with an inflection point at 4000 m. MC displayed biphasic dynamics, peaking at 80.86% (500–2000 m) before declining to 50.00% (inflection at 2000 m). SC demonstrated M-shaped fluctuation, initially rising to 23.84% at 4000 m, dipping at 8000 m, then surging to 30.43% at 16,000 m. Comparative analysis (2010 vs. 2020) showed substantial increases in MC and SC alongside marked decreases in UC and BC. These shifts underscored the FA’s intensified conflict restructuring, diverging sharply from regional and MA evolutionary patterns.

3.3.3. Conflict Analysis of Different Space Types

To elucidate conflict dynamics across spatial types and their geographic contributions, the 4000 m scale was identified as the peak conflict intensity threshold for UACY based on prior analytical findings. This scale was selected to investigate spatial-type conflict configurations, complemented by 250 m resolution landscape pattern analysis(

Table 7. Conflict level of different space types.

Type	Conflict Average Value
	2010	2020
APS	0.6065	0.6642
IMPS	0.6396	0.6692
ULS	0.6536	0.7176
RLS	0.6226	0.6777
FES	0.6044	0.6617
GES	0.6128	0.6771
WES	0.6208	0.6750
OES	0.6473	0.7146

).

From 2010 to 2020, the APS conflict index increased significantly to 0.6642, yet remained below the annual regional average, indicating localized intensification within an overall low-conflict context. IMPS conflict escalated to 0.6692, similarly staying subregional to global averages. ULS demonstrated pronounced conflict amplification from 0.6536 to 0.7176, persistently exceeding regional averages and maintaining high-intensity status. RLS, FES, and GES conflicts showed marked increases while remaining below regional means, suggesting system-wide low-intensity persistence. WES conflict rose from 0.6208 to 0.6750, exceeding the regional average in 2010 but falling below it by 2020, reflecting moderated intensity relative to broader trends. OES exhibited the most acute volatility, with indices persistently exceeding regional averages and demonstrating significant escalation, confirming sustained high-intensity conflict.

3.3.4. Analysis of Conflict Space Contribution

In 2010, FES constituted 88.43% of the primary spatial composition within UC zones, with other space types each accounting for less than 5%, reflecting FES dominance in low-conflict regions characterized by expansive, stable configurations. Conflict contribution analysis identified ULS, WES, and FES as dominant contributors, with ULS exhibiting the highest contribution index (6.17). This prominence stems from ULS’s role as urban built-up areas shaped by population and economic agglomeration, which foster stable spatial morphologies. WES, represented by six major lakes (Dianchi, Fuxian, Yangzonghai, Xingyun, Yilong, Qilu), maintained stable distributions due to minimal anthropogenic disturbances. FES further demonstrated strong contributions owing to its ecological stability, with most forested areas retaining original integrity. In contrast, RLS and GES exhibited negligible contribution indices (0.13 and 0.09, respectively), significantly below the threshold of 1 (Figure 9).

Under BC conditions, FES accounted for 77% of the primary spatial composition, showing a decrease compared to UC, while APS (11.39%) and GES (5.45%) exhibited proportional increases. This shift reflects enhanced spatial aggregation and structural stability among these three types. Conflict contribution analysis revealed ULS, WES, and FES remained dominant contributors, consistent with UC patterns, though their indices declined to 1.4–1.7. ULS persisted as the primary contributor despite reduced morphological integrity compared to UC urban cores. FES maintained partial occupancy with structural continuity, sustaining its contribution role. Secondary WES features (e.g., reservoirs) emerged as key contributors in BC. Peripheral contributors (RLS, IMPS, GES) showed rising indices but remained subdominant.

Under MC conditions, FES decreased to 54.05%, while APS and DES increased to 29.33% and 9.70%, respectively, with minimal contributions from other spaces. Spatial conflict contribution indices exhibited balanced distribution, predominantly ranging between 0.9 and 1.1, contrasting with the skewed patterns observed in UC and BC scenarios. APS, IMPS, RLS, and GES showed marginally elevated indices (slightly >1), reflecting interspersed spatial configurations. This fragmented distribution correlated with reduced structural stability compared to UC/BC regimes.

Under SC conditions, APS constituted 43.89% of spatial composition, followed by FES (30.13%), with GES and OES each approximating 10%. ULS represented merely 0.42%. Conflict contribution analysis identified OES (index: 3.38), RLS, and APS (both >1.5) as primary contributors, while GES and IMPS exhibited moderate contributions (>1). WES, FES, and ULS demonstrated negligible contributions (≈0.5), highlighting the oppositional complementarity between SC and UC regimes. APS and RLS fragmentation constituted principal conflict drivers, linked to urbanization-driven spatial conversions that preserve macrostructural stability. FES maintained concentrated distributions but faced localized encroachments in human-adjacent zones, correlating with low conflict contributions. RLS development displayed unregulated expansion due to minimal planning constraints, exacerbating APS occupation conflicts. OES emerged as the strongest contributor, associated with land degradation and dispersed spatial patterns.

Under MC conditions, FES exhibited increased proportions while APS decreased compared to 2010 baselines, maintaining their status as dominant spatial types. GES demonstrated significant reduction, with other spaces collectively constituting minor proportions. Contribution analysis revealed relative equilibrium across spatial types in 2020, with indices predominantly ranging from 0.86 to 1.10 (except OES: 0.58), lacking pronounced dominance hierarchies. Notably, dominant contributors diverged from 2010 patterns, as FES and WES emerged as primary contributors. This shift likely reflects intensified conflict dynamics, wherein FES and WES transitioned from high-contribution roles under UC/BC regimes to MC-dominated contributions, thereby diminishing other spatial contributions.

Under SC conditions, APS maintained stability at 43.67%, while FES increased significantly to 40.03% compared to 2010 levels. IMPS and RLS exhibited marked upward trends, contrasting with declines in OES and GES proportions. Conflict contribution patterns remained temporally consistent, with APS, IMPS, RLS, GES, and OES persisting as dominant contributors. IMPS and GES demonstrated elevated contribution indices, whereas OES and RLS showed reductions, reflecting complementary interactions among SC, UC, and BC regimes. The 2020 conflict escalation primarily stemmed from spatial encroachment by APS, IMPS, and RLS into FES and GES territories, compounded by inter-sector competition (e.g., RLS/IMPS occupation of APS). FES maintained spatial aggregation but experienced localized fragmentation near settlements, while dispersed OES patterns exacerbated conflict susceptibility due to land degradation (Figure 10).

4. Discussion

As a representative mountainous province, Yunnan exhibits distinctive geomorphological characteristics dominated by limited FA. FA constitutes merely 6% of the province’s total land area, contrasting with MA occupying 94%. However, these spatially constrained FA paradoxically host intensive human production and living activities while maintaining critical ecological functions, exemplified by dense hydrological networks including nine plateau lakes. MA serves dual roles as ecological barriers and human activity carriers, preserving core ecosystem services while supporting traditional agricultural practices. Significant disparities exist between mountain and dam systems in topography, climatic conditions, and resource distribution. Investigating these spatial contrasts enhances understanding of regional differentiation mechanisms and enriches geographical theories. The functional differences in spatial organization and ecological sensitivities between these zones provide critical insights for optimizing territorial planning frameworks and formulating elevation-dependent development strategies.

4.1. Evolution Characteristics of PLES

The structural configuration of PLES in the UACY exhibits pronounced spatial differentiation across UACY, MA, and FA scales. FES and APS dominate the entire region and MA, whereas APS constitutes the primary spatial category in FA, reflecting the deterministic influence of mountain–dam physiographic conditions on PLES distribution patterns. Dynamically, APS manifests compensatory trends with increases in MA and decreases in FA, achieving regional equilibrium, while IMPS demonstrates robust expansion. Contrasting with previous findings emphasizing production space growth near urban peripheries [66], this study identifies significant IMPS expansion in both MA and FA. The dam-area growth aligns with accelerated industrialization, while MA increases correlate with mining sector development and Yunnan Province’s “industrial uphill” policy, which strategically relocates industrial land to uplands for farmland conservation. Ecological spaces exhibit divergent trajectories: FES shows moderate growth, contrasting sharply with GES, which undergoes rapid decline. This dichotomy reflects intensified ecological protection pressures in FA, where urbanization-driven encroachment severely impacts GES. Spatial transition analysis reveals that MA experiences large-scale bidirectional transfers in Ecological Spaces (FES/GES/WES/OES), confirming their ecological conservation priority status [67,68]. APS predominantly converts to FES, evidencing effective implementation of farmland-to-forest policies, while limited ULS/RLS transitions indicate restrained urbanization. In FA, substantial IMPs and ULS inflows mark them as industrialization–urbanization cores. Despite minor APS outflows, sustained agricultural significance is evidenced by notable APS inflows, whereas minimal Ecological Space transitions suggest inadequate ecological governance. Mechanistically, APS reduction links to ecological restoration policies and agricultural restructuring, particularly in MA. Regional development strategies diverge distinctly, with ecological preservation prioritized in mountains versus industrialization–urbanization dominance in FA [69].

4.2. PLES Conflicts

Landscape ecological risk assessment has become a widely adopted method for quantifying land-use conflicts, yet existing studies frequently overlook the scale sensitivity of landscape indices—particularly the variations in risk values across different analytical scales. This study addresses this gap by employing six moving window scales (500 m, 1 km, 2 km, 4 km, 8 km, 16 km) to systematically analyze spatial conflicts in UACY. The results demonstrate distinct hierarchical characteristics and evolutionary patterns of conflicts across scales.

The landscape risk model elucidates scale-dependent effects and regional differentiation characteristics of PLES spatial conflicts. Analysis reveals a unimodal trend in conflict intensity across scales ranging from 500 m to 16,000 m, peaking at 4000 m resolution. This finding methodologically diverges from prior research [24], which identified 2000 m as a stabilization threshold but lacked multi-scale validation. The critical role of scale selection in conflict identification is evident, with the bisectional scaling protocol employed in this study ensuring spatial correspondence of grid units across resolutions, thereby enhancing methodological rigor. Geospatial differentiation analysis demonstrates significantly elevated conflict levels in FA compared to regional and MA, attributable to compounded physiographic constraints and intensive socioeconomic activities. Temporal analysis further reveals conflict intensification across all scales from 2010 to 2020, particularly in eastern urban clusters where severe conflict spatial agglomeration emerges. This trajectory aligns with urbanization-driven land-use intensification patterns documented in Bon [60]. Conflict contribution analysis identifies distinct spatial-type impacts: ULS, FES, and WES predominantly associate with low-conflict zones, whereas APS, IMPS, RLS, GES, and OES concentrate in high-conflict areas. Notably, urban cores maintain structural integrity without direct conflict generation, while peripheries exhibit acute escalation. This spatial paradox stems from two mechanisms—urban expansion fragments APS through edge encroachment, increasing landscape complexity, and IMPS preferentially occupies ecologically sensitive zones (FES/APS) due to siting requirements, exacerbating spatial friction. These observations corroborate peripheral conflict intensification patterns reported in Yang [70]. In FA, accelerated urbanization drives severe ecological space erosion, potentially generating compound conflict patterns.

This study conducts a multi-scale analysis of “Production-Living-Ecological” (PLE) space conflicts. Compared to previous research, it considers the sensitivity of landscape indices, and the results show that PLE space conflicts do not exhibit a single trend with scale changes but rather demonstrate certain threshold effects. In multi-scale conflict research, Ai et al. [71] analyzed scales ranging from 15 m to 240 m to determine the optimal scale, but they only considered granularity and did not account for the impact of extent changes, which also significantly influence landscape indices. Wang et al. [72] conducted a multi-scale analysis of conflict evolution in Chongqing at the county, town, and grid levels, achieving notable results in simulating multi-scenario land-use conflicts. However, differences in conflict patterns across grid scales require further investigation. Additionally, this study focuses on typical mountainous and flatland areas, where unique geographical conditions may lead to richer landscape types. Using the Urban Agglomeration in Central Yunnan as the study area provides both typicality and representativeness. The landscape conflict index demonstrates strong applicability across diverse geographical environments, such as flatland and mountainous areas. In flatlands, it effectively captures conflicts from agricultural intensification and urban expansion, while in mountainous regions, it identifies ecological conservation issues like deforestation and habitat fragmentation. By adjusting parameter weights to reflect environmental specifics (e.g., slope sensitivity in mountains), the index proves versatile and adaptable, making it a robust tool for spatial conflict analysis in varied contexts.

In addition, research on spatial conflicts abroad primarily focuses on land-use conflicts [34,36]. Both “land-use-conflicts” and “PLES conflicts” fall within the research scope of “spatial conflicts”, with the distinction lying in the classification of spatial types [21]. Compared to land-use spaces, PLE spaces represent a more comprehensive spatial classification model, derived from national strategic needs. Spatial conflict risk, as a subset of spatial conflict research, assesses the likelihood of future conflicts. Unlike current research on “spatial conflict risk based on land use change prediction”, it provides a more direct prediction of future conflict areas. Therefore, conducting conflict research based on PLE spaces better aligns with the needs of China’s territorial spatial planning and governance in its developmental context.

4.3. Limitations

Although this study has achieved certain results in multi-scale PLE space conflict analysis, some limitations remain. First, data limitations are a concern. The study primarily relies on publicly available statistical yearbooks and local survey data. While these data sources are authoritative and reliable, their coverage and precision have certain limitations. For example, variations in data quality across different years and sources may lead to inconsistencies in the analysis. Early land-use data, with lower resolution and accuracy, may affect the reliability of long-term time series analysis. Furthermore, the study’s time span from 2010 to 2020, while reflecting recent spatial utilization and conflict characteristics, may not fully capture long-term trends and cyclical patterns, potentially limiting the long-term applicability of the optimization project. Future research could improve methodologies by diversifying data sources, establishing long-term monitoring, and enhancing data quality, thereby strengthening the scientific and practical value of the findings.

Although there is limited research abroad specifically on PLES conflicts, extensive studies have been conducted on land-use conflicts globally. Research has explored various dimensions, such as large-scale land investments disrupting traditional livelihoods in Ethiopia’s agro-pastoral regions [34], soybean expansion triggering deforestation and indigenous land rights conflicts in the Brazilian Amazon [35], and peri-urban land-use conflicts driven by multifunctionality in Switzerland [4]. In Germany, stakeholder perspectives have emphasized multi-stakeholder participation in policy-making to address national and regional land-use conflicts [36]. Studies also highlight conflicts between ecological conservation and resource development, with compatibility analysis and spatial assessment identifying synergies and conflicts among ecological, socio-cultural, and economic values. Methodologically, participatory mapping and multi-criteria decision-making (MCDM) have been employed to identify and manage land-use conflicts [31], underscoring the need for integrated, multi-stakeholder approaches to address their complexity effectively.

The global body of research on land-use conflicts provides a strong foundation for advancing the study of PLES conflicts. By integrating international methodologies and insights, PLES conflict research can address the unique challenges of balancing production, living, and ecological spaces in China. This synthesis not only enriches the theoretical understanding of PLES conflicts but also offers practical tools and strategies for sustainable land-use planning and governance, contributing to the broader goal of achieving ecological civilization and sustainable development.

4.4. Future Research Directions

While this study reveals scale-dependent characteristics of spatial conflicts, future research should prioritize elucidating their multi-scale mechanisms through integrating higher-resolution remote sensing data to uncover finer-grained conflict patterns. The identified differentiation between MA and FA warrants deeper exploration of conflict formation processes, particularly through synthesizing socioeconomic datasets and ecological modeling to quantify policy impacts, such as urbanization dynamics and ecological conservation measures. These findings provide foundational insights for land-use planning, yet advancing conflict mitigation requires developing scenario simulation frameworks to assess how diverse land management policies influence conflict evolution. Such analyses could evaluate trade-offs between developmental priorities and ecological sustainability, ultimately informing regionally differentiated governance strategies. At the same time, future research should prioritize the generalizability of existing methods.

5. Conclusions

This study investigated the spatio-temporal evolution and conflict dynamics of PLES in the UACY, a representative in MA. Utilizing 2010 and 2020 land-use data, we established a PLES classification framework and systematically analyzed spatial differentiation patterns across UACY, MA, and FA contexts. A multi-scale landscape risk model (500–16,000 m) quantified conflict intensity variations, complemented by spatial typology contribution analysis.

(1)

The spatial structure of the UACY exhibits distinct mountain–dam differentiation. FES and APS dominate regional and MA, while APS prevails in FA. Spatially, the entire region demonstrates relatively intense transitions, with varying dynamics across space types: APS exhibits compensatory expansion in MA alongside contraction in FA, maintaining equilibrium. IMPS display aggressive growth, contrasting with moderate increases in ULS and RLS, while GES degrade rapidly. Other OES undergo the most dramatic decline, particularly in ecologically sensitive intermountain basins. Transition intensities in MA systematically exceed those in dam zones, reflecting amplified responsiveness to ecological pressures and human activities in topographically complex regions.

(2)

The spatial conflict of PLES exhibits distinct scale effects and regional differentiation characteristics. From a scale of 500 m to 1600 m, the spatial conflict generally demonstrates an initial upward trend followed by a downward trend, exhibiting a significant threshold effect. The scale of 4000 m represents the maximum conflict intensity, with the conflict level in the FA at each scale significantly higher than that in the entire region and the MA. Severe conflict is concentrated in the eastern part of the urban agglomeration, with the scale of 4000 m again being the maximum intensity. From 2010 to 2020, conflicts at all scales have shown an intensifying trend.

(3)

Overall, when the conflict type is stable and basically controllable, the strong contribution areas are primarily ULS, FES, and WES. When the conflict intensity is moderate, the strong contribution areas include APS, IMPS, and RLSGES. When the conflict intensity is severe, the strong contribution areas are RLS, IMPS, RSL, GES, and OES. ULS, FES, and WES are primarily involved in low-conflict areas such as those that are stable and basically controllable, while APS, IMPS, RLS, GES, and OES are predominantly found in high-conflict areas, serving as the primary sources of conflict.

From the findings above, it is evident that the spatial conflicts within the PLES in the UACY are intensifying. Given the significant differences in the distribution of PLE spaces and spatial conflicts between mountainous and flatland areas, it is crucial to fully consider the conflicts and contradictions among construction land, cultivated land, and ecological land during territorial spatial planning. Differentiated policies should be formulated to allocate spatial resources accordingly.