Evaluating China’s National Park Pilots: Constructing an Indicator System for Performance Assessment

Abstract

With the designation of the first cohort of national parks and the continued operation of remaining pilots, China’s national park reform has entered a critical stage requiring consolidation and adaptive improvement. A key challenge lies in the ambiguous status of five pilot zones, which lack a standardized evaluation mechanism to guide decisions on future inclusion or exit. This study develops a comprehensive indicator system specifically tailored to assess the construction and development of national park pilots, thereby supporting evidence-based governance beyond initial entry criteria. Drawing on relevant theories and China’s institutional context, the framework employs Analytic Hierarchy Process, expert consultation, and fuzzy scoring to determine indicator weights and evaluation standards. The resulting system integrates three dimensions—ecological protection system, management system, and public service system. Nanshan National Park was selected as a case study, scoring 87.77 in 2024 (Class II, “Proficient”), with strong overall performance but notable weaknesses in landscape connectivity, recreational product diversity, and regional integration. These findings suggest the need for targeted improvements in ecological corridors, service enrichment, and community benefit-sharing. Overall, the proposed framework provides a replicable tool for evaluating pilot zones, offering practical insights for refining China’s national park development and enhancing governance effectiveness.

1. Introduction

The establishment of protected area systems is a cornerstone of global biodiversity conservation strategies [1]. International consensus underscores the role of effectively governed protected areas in safeguarding ecological integrity, preserving key habitats, and maintaining essential ecosystem services. In response to these global imperatives, China has embarked on a sweeping reform of its nature protection regime: the development of a formal National Park System [2]. This initiative, formally launched with pilot programs beginning in 2015, represents a pivotal shift towards a more unified and ecologically focused approach to protected area governance in the country [3]. With the official announcement of the first batch of national parks in 2021, the system has transitioned from a trial phase to an institutionalized framework with more rigorous national regulation and strategic ecological zoning.

However, this milestone has also raised a practical governance question: how to manage the remaining pilot sites that were not selected for formal designation in the first batch. At the time of writing, there are still five “transitional pilot zones” (such as Nanshan and Shennongjia), which have undergone years of pilot management and financial investment, but their future development direction is still uncertain. Simply extending the pilot phase indefinitely is unsustainable, yet automatic promotion without rigorous evaluation risks undermining the integrity of the nascent National Park System. This situation underscores a critical gap in the current institutional architecture: the lack of a standardized, science-based mechanism to comprehensively assess the performance of these pilot zones and provide an evidence-based basis for subsequent policy choices, whether concerning formal incorporation or alternative arrangements.

While international frameworks—such as the IUCN’s Green List, the Management Effectiveness Tracking Tool (METT), and the World Commission on Protected Areas (WCPA) standards—offer valuable references [4,5,6], they are typically designed for mature protected areas and emphasize ongoing management effectiveness rather than construction progress or system-readiness [7]. Similarly, domestic evaluation systems, including China’s own Assessment and Evaluation Standard for National Parks (GB/T 39739-2020) [8], focus heavily on annual operational metrics and ecological outcomes [9,10]. They are less equipped to assess the multi-dimensional performance of pilot areas transitioning toward national designation. Furthermore, most existing policy documents focus heavily on access criteria—that is, what qualifies an area for designation as a national park [11]. By contrast, very limited attention has been paid to how pilot areas are systematically evaluated after the trial phase. Yet, in a large-scale national system such as China’s, where territorial coverage is vast and resources are limited, developing a transparent evaluation framework is crucial to determine whether pilot areas meet the standards for formal designation or require phased adjustment before further advancement [12].

Therefore, there is a pressing need to develop a dedicated evaluation indicator system specifically designed for China’s national park pilot zones. Such a system should provide a transparent and objective benchmark for assessing the current construction status, progress, and relative performance of pilot zones against clearly defined goals, while also supporting longer-term evaluations of sustained performance beyond the pilot phase. This would facilitate meaningful comparison between designated and non-designated pilots, clarify whether these areas meet the composite expectations of ecological protection, institutional effectiveness, and public service delivery, and inform future policy adjustments.

2. Materials and Methods

2.1. Study Area: Nanshan National Park Pilot Zone

The Nanshan National Park Pilot Zone (NNP) is situated in the southern region of Hunan Province, China, encompassing parts of Chengbu Miao Autonomous County, Xinning County, and Suining County within Shaoyang City, as well as the southwestern portion of Dong’an County in Yongzhou City (Figure 1). Strategically positioned within China’s ecological security framework, Nanshan serves as a vital ecological barrier. Following boundary optimization and consolidation efforts by the Chinese government in 2021, the pilot zone now covers a total area of 1303.8 km² [13].

NNP plays a critical role in regional hydrology as the source and watershed for three major river systems: the Yuanjiang and Zishui (Yangtze Basin) and the Xijiang (Pearl River Basin), underpinning crucial water conservation functions. Water quality monitoring indicates high standards, with key areas like the former Nanshan Pasture and Jintongshan National Nature Reserve consistently meeting Class I standards (GB 3838-2002) [14], and others like Liangjiang Gorge and Baiyun Lake meeting Class II standards. The park’s complex terrain and transitional location foster exceptional biodiversity across ecosystems, species, and genetic levels. It harbors over 40% of its ecogeographic region’s species, including 2835 vascular plant species (477 pteridophytes, 23 gymnosperms, 2333 angiosperms) with 3 National Grade I and 73 Grade II protected species, and 359 wild vertebrate species (33 fish, 31 amphibians, 49 reptiles, 197 birds, 49 mammals), including 9 National Grade I and 68 Grade II protected animals [15]. Endemism is high, with 786 plant and 153 vertebrate species endemic to China.

Socio-economically, the park area supports a dispersed population of 12,140 permanent residents (14,611 registered) across 4135 households, primarily engaged in agriculture, forestry, animal husbandry, and limited small businesses. A declining population trend is noted, linked to limited arable land and agricultural income [16]. The region is culturally rich, home to significant populations of Miao, Dong, and Yao ethnic groups who maintain unique traditions such as Taolin Nuo performances, Miao embroidery, leaf-blowing music, and the Chengbu Hanging Dragon art [17]. Additionally, NNP preserves important historical relics reflecting its role in China’s revolutionary history, including the “High Mountain Red Sentinel” and sections of the Long March route like Laoshanjie.

At present, Nanshan National Park is divided into two areas: the general control area and the core protection area (Figure 2). The core protection area, which means Strictly Protected, serves as the ecological cornerstone of Nanshan National Park, enforcing non-intervention conservation to preserve primeval forest dynamics, critical habitats for endemic species (e.g., Abies ziyuanensis), and migratory bird corridors. Human access is prohibited except for sanctioned scientific monitoring, ensuring minimal anthropogenic disturbance [13]. General Control Zones adopt a multi-functional zoning regime: (1) Ecological Conservation Zones facilitate long-term research (e.g., 24 ha forest dynamics plots); (2) Recreation and Display Zones offer regulated ecotourism under environmental carrying capacity constraints; (3) Traditional Use Zones enable sustainable livelihoods via Authorized Operations (e.g., tea cultivation) and pastoralism quotas. This framework exemplifies China’s integrated approach to reconciling biodiversity conservation with community well-being.

2.2. Research Method

This study adopted a three-phase sequential framework to evaluate Nanshan National Park’s (NNP) management efficacy. Phase I established a multidimensional evaluation system through structured Delphi expert consultations (n = 16), integrating ecological integrity, socio-economic sustainability, and institutional governance indicators. Subsequently, Phase II employed the Analytic Hierarchy Process (AHP) coupled with weighted expert judgment to assign priority-based weights to all system metrics. Finally, Phase III executed field-based quantification of 64 evaluation items across NNP’s functional zones, deriving a composite conservation performance score (range: 0–100) through linear aggregation. All statistical analyses were conducted using SPSSAU (Version 24.0; SPSSAU Software, Beijing, China). Evidence-based policy recommendations were then formulated by the above contents.

2.2.1. Expert Survey Method

Following Von Soest (2023) [18], we implemented a structured expert survey methodology comprising Delphi-style consultations and semi-structured interviews. Sixteen specialists were recruited from four key disciplines essential to national park management: forestry planning (n = 3), regional planning (n = 3), resource/environmental economics (n = 4), ecology (n = 2), and geography (n = 4). Full expert credentials are documented in

Table 1. The expertise and sphere of selected 16 experts.

Respondent	Expertise	Sphere
R1	Resources and Environmental Economics	School of Economics, Nankai University
R2/R3/R4	Resources and Environmental Economics	Business School, Hunan Normal University
R5	Forestry Planning and Design	College of Forestry, Central South University of Forestry and Technology
R6	Forestry Planning and Design	Hunan Provincial Forest Exploration Institute
R7	Forestry Planning and Design	Forestry Bureau of Hunan Province
R8	Regional Planning	Yunnan Provincial Institute of Geological Environment Monitoring
R9/R10	Regional Planning	Institute of Geology, Chinese Academy of Geological Sciences
R11	Ecological governance	Institute of Subtropical Agricultural Ecology, Chinese Academy of Sciences
R12	Ecological governance	Hunan Provincial Forest Exploration Institute
R13	Geology and Geography	Institute of Geology, Chinese Academy of Geological Sciences
R14/R15	Geology and Geography	Institute of Geography and Oceanography, Nanning Normal University
R16	Geology and Geography	College of Earth Sciences, Guilin University of Technology

.

This study employed a hybrid Analytic Hierarchy Process (AHP)-Delphi framework to construct and operationalize the evaluation system. Phase 1 engaged 16 domain experts through three-round Delphi consultations to refine preliminary indicators, achieving content validity ratios (CVR) > 0.75 and inter-rater reliability (ICC) of 0.89. Phase 2 utilized expert-informed pairwise comparisons to generate AHP judgment matrices for weight assignment, with all consistency ratios (CR) verified below 0.08—satisfying Saaty’s acceptability threshold [19]. Phase 3 applied the validated indicator system to assess Nanshan National Park (NNP).

2.2.2. Analytic Hierarchy Process (AHP) Method

Evaluating China’s national park development constitutes a complex multidimensional challenge. The Analytic Hierarchy Process (AHP) addresses this by deconstructing the system into hierarchical layers, enabling integrated analysis of quantitative and qualitative criteria while mitigating methodological arbitrariness through pairwise comparison-based weighting [20]. Consequently, AHP proves superior to alternative approaches—such as simple weighting or Principal Component Analysis (PCA)—for determining indicator weights in this study. We therefore employ AHP to establish multilevel weights throughout our evaluation framework, establishing a rigorous quantitative foundation for subsequent comprehensive assessment.

I.

Building a hierarchical model.

The hierarchical model is the basis of hierarchical analysis. Combined with the first step of the expert investigation method, this paper constructs the hierarchical analysis model. Based on the complexity of the evaluation object, this paper divides the overall goal into three sub-goals: the ecological protection system, the management system, and the public service system. There is a logical connection between the sub-objective layer and the criterion layer. Each criterion in the criterion layer is directly related to a certain aspect of the sub-objective layer, ensuring that the combined effect of all criteria can effectively promote the realization of the objective. The role of the element layer in the indicator system is to connect the upper and lower levels. It transforms the abstract standards of the criterion layer into specific and operational elements, providing a detailed framework for the setting and evaluation process of the indicator layer. The bottom layer is the indicator layer, which is directly related to data collection and analysis. It transforms the relatively abstract concepts of the element layer into specific indicators that can be measured and counted, providing support for decision-making in understanding the current situation and identifying problems.

Overall, based on expert judgment and the final evaluation system, this study divides the hierarchical model into five levels, namely the objective level, the sub-objective level, the criterion level, the element level and the indicator level.

II.

Construct the judgment matrix.

Starting from the second level of the hierarchical model, a judgment matrix is constructed for each element belonging to the upper level, and the 1–9 scale method is adopted to quantify the relative importance degree among the indicators. The Analytic Hierarchy Process (AHP) determines the relative importance among indicators by constructing a judgment matrix. Suppose there are n elements P1, P2, P3, P4... Pn. To understand the relative importance among these elements, it is necessary to compare the elements pairwise, using 1 (p1/p1), p1/p2, p1/p3,... p1/pn represents the relative importance of each element, and the resulting judgment matrix table can be obtained (

Table 2. Matrix table.

	P1	P2	P3	…	Pn
P1	1	p1/p2	p1/p3	…	p1/pn
P2	p2/p1	1	p2/p3	…	p2/pn
P3	p3/p1	p3/p2	1	…	p3/pn
…	…	…	…	…	…
Pn	pn/p1	pn/p2	pn/p3	…	1

).

The purpose of constructing the judgment matrix is to determine the relative importance of the indicators. In this paper, a scale from 1 to 9 is used to quantify the relative importance among the indicators [10]. To ensure objectivity when judging the index weights, in the process of constructing the judgment matrix, this paper collected the judgments on the index weights from a total of 16 experts in the research fields related to nature reserves and national parks. The expert information is shown in Table 1. The constructed judgment matrix can be found in the Appendix A section.

III.

Consistency test

Consistency verification is essential for ensuring the reliability of AHP results, serving to detect logical contradictions or subjective biases in expert pairwise comparisons. This procedure evaluates the consistency ratio (CR) against the random index (RI), where CR < 0.1 indicates acceptable matrix consistency [19]. Matrices failing this threshold require iterative revision until compliance is achieved. Notably, second-order matrices inherently satisfy consistency requirements and thus bypass validation. Failure to attain acceptable CR values after adjustment may signal fundamental issues in indicator selection, hierarchy construction, or expert competence, necessitating methodological refinement.

The Consistency test requires the use of two Index values, namely the Consistency Index (CI) and the random consistency index (RI). The former can be calculated from the maximum eigenvalue, while the latter has a fixed reference value table (

Table 3. RI hierarchical fixed value table.

Order	1	2	3	4	5	6	7
RI	0.00	0.00	0.52	0.89	1.12	1.26	1.36

) according to the order of the judgment matrix.

$$\mathrm{C}\mathrm{I}={\displaystyle \frac{{\lambda }_{\mathrm{m}\mathrm{a}\mathrm{x}}-\mathrm{n}}{\mathrm{n}-1}}$$

The consistency ratio CR is

$$\mathrm{C}\mathrm{R}={\displaystyle \frac{\mathrm{C}\mathrm{I}}{\mathrm{R}\mathrm{I}}}$$

When CR is less than 0.1, the judgment matrix satisfies the consistency test; otherwise, the judgment matrix should be appropriately adjusted and analyzed again.

IV.

Evaluation Grading System

Given the limited quantifiability of indicators in China’s national park assessments, this study approximates true performance through triangulation of literature analysis, policy reviews, and expert elicitation. Qualitative factors were converted to quantitative scores, weighted, and aggregated into a unified metric. Final composite scores (range: 0–100) determine four grading tiers, computed as

$$\mathrm{S}=\sum _{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{w}}_{\mathrm{i}}{\mathrm{s}}_{\mathrm{i}}$$

where S = total score, ${\mathrm{w}}_{\mathrm{i}}$ = weight of indicator i, and ${\mathrm{s}}_{\mathrm{i}}$ = score on indicator i.

Grading Criteria employ dual constraints: total score thresholds + minimum dimension averages (≥60/100 per criterion layer) to ensure balanced performance (

Table 4. Scale and meaning of judgment matrix.

Tier	Score Range	Performance Level	Key Characteristics
Ⅰ	90–100	Exemplary	Exceptional ecosystem representativeness and integrity Optimal protection-management synergy Maximized educational/recreational services with stakeholder harmony
Ⅱ	75–89	Proficient	Minor improvements needed in ≤2 dimensions Correct developmental trajectory Moderate enhancement potential
Ⅲ	60–74	Adequate	Meets baseline standards Critical improvements required but non-fatal Aligns with ecological civilization principles
Ⅳ	0–59	Deficient	Fundamental flaws (e.g., unrepresentative ecosystems) Critical human-nature conflicts Management system failures

).

3. Results and Discussion

3.1. Finalized Indicator System for National Park Performance Evaluation

Through the initial screening of the comprehensive evaluation index for the national park system pilot areas, indicators and factors that meet China’s basic requirements for the construction of national parks and international evaluation criteria of national parks were screened [4,5,6,9,10,16,17,21,22,23,24,25,26,27,28,29,30,31,32] (

Table A34. Preliminary screening evaluation indicators.

Composite Index	Single Indicator	Indicator Type
Representativeness	Location representativeness, importance, species representativeness, landscape carriers and elements, landscape aggregation index	Ecological protection system
Authenticity	Natural authenticity (proportion of natural habitats/wilderness degree, frequency of extreme weather, variation in forest line height), cultural authenticity (preservation degree of traditional culture and its carriers)	Ecological protection system
Integrity	Composition (species richness, species diversity index), structure (degree of landscape fragmentation, number of trophic levels, integrity index of major food chains), and function (ecosystem service function)	Ecological protection system
Management Foundation	Management entity, infrastructure (distance of maintenance stations), management guarantee (financial guarantee, institutional guarantee, land ownership, boundary planning)	Management system
Management process	Monitoring capacity, daily management, and social participation (number and proportion of volunteers, proportion of public welfare positions for ecological management and protection, number of business projects, and other channels for social participation)	Management system
Scientific research and education	The number of cooperations with scientific research institutions, scientific research projects, international cooperation and exchanges, scientific research facilities, the degree of interpretation and facility completeness, the number of nature education activities, and the annual number of nature education and ecological experience tourists received	Public service system
Recreational experience	Accessibility of transportation, completeness of infrastructure, richness of ecological and cultural products, and the number of tourists involved in safety accidents	Public service system
Regional development	The number of permanent residents, Engel’s coefficient, annual household income, industrial structure, and residents’ awareness	Public service system

in Appendix A). The screening process is based on the importance and relevance of these indicators and factors.

The preliminary indicator framework (see

Table 5. Indicator system for national park performance evaluation.

Objective Layer	Sub-Objective Layer	Criterion Layer	Element Layer	Indicator Layer
National park evaluation index system S	Ecological protection system A1	National representativeness B1	Spatial representativeness C1	Natural area representation D1
				Ecological niches’ importance D2
			Species representativeness C2	Proportion of endemic species D3
				Proportion of State Key Protected Wild Species D4
			Landscape value C3	Integrity of landscape value elements D5
				The uniqueness of landscape value D6
		Ecosystem authenticity B2	Natural authenticity C4	Proportion of natural habitats D7
				Proportion of alien species D8
				Human activity intervention D9
			Cultural authenticity C5	Authenticity of folk culture D10
				Authenticity of humanistic facilities D11
		Ecosystem integrity B3	Composition integrity C6	Species richness D12
				Uniformity of species distribution D13
			Structural integrity C7	Landscape patch fragmentation D14
				Major food chain integrity D15
			Ecosystem service integrity C8	Forest stock volume D16
				Air quality D17
				Carbon sequestration and oxygen release D18
				Soil conservation function D19
				Water conservation function D20
				Humus thickness D21
				Water quality status D22
	Management system A2	Management Foundation B4	Management entity C9	Man-post matching degree D23
				Adequacy of grassroots staff D24
				Capacity development of managers D25
			Management facilities C10	Adequacy of maintenance stations D26
				Distribution of maintenance stations D27
				Level of management informatization D28
				Completeness of emergency facilities D29
		Management Support B5	Institutional support C11	Clarity of functional zoning D30
				Completeness of the overall norms D31
			Funds safeguard C12	Adequacy of funds D32
				Transparency of fund allocation D33
				Proportion of fixed expenditures D34
			Technical support C13	Construction of research institutions D35
				Research achievements D36
				Cooperation and communication situation D37
			Security of ownership C14	Clarity of land ownership D38
				Clarity of natural resources ownership D39
		Management process B6	Monitoring operation C15	Coverage of environmental monitoring D40
				Application of monitoring data D41
			Inspection operation C16	Daily patrol situation D42
				Effectiveness of ecological restoration D43
				Ecological compensation situation D44
				Emergency drill situation D45
				Species rescue situation D46
			Social participation C17	Volunteer participation situation D47
				Participation of NGOs D48
				Ecological positions D49
	Public service system A3	Recreational education B7	Recreational experience C18	Transportation accessibility D50
				Completeness of recreational facilities D51
				Richness of recreational products D52
				Number of tourists involved in safety D53
				Completeness of entrance community infrastructure D54
			Nature education C19	Completeness of explanation system D55
				Training for tourists before entering D56
				Implementation of nature education activities D57
				Number of tourists receiving nature education D58
		Regional development B8	Economic development C20	Revenue from tourism business projects D59
				Income from land management projects D60
				Realization of ecological products D61
			Social development C21	Engel’s coefficient D62
				The community employment D63
				Residents’ awareness of national parks D64

) underwent a rigorous three-round Delphi process to enhance validity and operational feasibility. Sixteen domain experts (demographics in Table 1) independently rated each indicator on dual criteria: ecological representativeness (1–5 Likert scale) and policy relevance (1–5 Likert scale). Indicators scoring below 3.5 on either dimension were iteratively refined or eliminated, achieving final consensus.

The optimized system comprises 3 sub-objectives, 8 criteria, 21 elements and 64 indicators (Table 5), structured as

The evaluation system employs a five-tiered hierarchical structure, with each tier sequentially cascading to operationalize the assessment framework. The target layer represents the overarching goal: “Comprehensive Development Status of China’s National Parks”. This tier defines the evaluation’s strategic orientation, serving as both the conceptual foundation and ultimate endpoint of the system. Sub-target layer decomposes the primary objective into three actionable domains: Ecological Integrity\Management Efficacy and Public Service Delivery. This decomposition enables targeted governance while maintaining holistic coherence. The Criterion Layer establishes measurable standards directly linked to sub-objectives, and the element layer translates abstract criteria into operational components across 21 elements. At last, the indicator layer constitutes 64 quantifiable metrics that transform elements into measurable variables\Enable empirical data collection and scoring, then supporting evidence-based decision-making through statistical diagnostics.

3.2. Weights of the Items in the Evaluation Index System

Expert-informed pairwise comparison matrices (

Table A1. S-A Judgment matrix.

S	A1	A2	A3	$\mathsf{\omega }$
A1	1	5.459	2.997	0.662
A2	0.183	1	1.179	0.162
A3	0.334	0.849	1	0.176

,

Table A2. A1-B Judgment matrix.

A1	B1	B2	B3	$\mathsf{\omega }$
B1	1	2.611	1.934	0.524
B2	0.383	1	1.497	0.258
B3	0.517	0.668	1	0.218

,

Table A3. A2-B Judgment matrix.

A2	B4	B5	B6	$\mathsf{\omega }$
B4	1	1.742	1.471	0.437
B5	0.574	1	1.976	0.336
B6	0.680	0.506	1	0.228

,

Table A4. A3-B Judgment matrix.

A3	B7	B8	$\mathsf{\omega }$
B7	1	0.913	0.478
B8	1.095	1	0.523

,

Table A5. B1-C Judgment matrix.

B1	C1	C2	C3	$\mathsf{\omega }$
C1	1	1.499	1.362	0.414
C2	0.667	1	1.497	0.327
C3	0.734	0.668	1	0.259

,

Table A6. B2-C Judgment matrix.

B2	C4	C5	$\mathsf{\omega }$
C4	1	5.319	0.842
C5	0.188	1	0.158

,

Table A7. B3-C Judgment matrix.

B3	C6	C7	C8	$\mathsf{\omega }$
C6	1	2.457	0.897	0.419
C7	0.407	1	0.924	0.236
C8	1.115	1.082	1	0.345

,

Table A8. B4-C Judgment matrix.

B4	C9	C10	$\mathsf{\omega }$
C9	1	2.190	0.687
C10	0.457	1	0.313

,

Table A9. B5-C Judgment matrix.

B5	C11	C12	C13	C14	$\mathsf{\omega }$
C11	1	1.913	3.053	1.007	0.369
C12	0.523	1	1.566	0.955	0.221
C13	0.328	0.638	1	0.709	0.147
C14	0.993	1.047	1.410	1	0.263

,

Table A10. B6-C Judgment matrix.

B6	C15	C16	C17	$\mathsf{\omega }$
C15	1	1.769	1.449	0.444
C16	0.565	1	1.179	0.284
C17	0.690	0.848	1	0.272

,

Table A11. B7-C Judgment matrix.

B7	C18	C19	$\mathsf{\omega }$
C18	1	1.845	0.649
C19	0.542	1	0.351

,

Table A12. B8-C Judgment matrix.

B8	C20	C21	$\mathsf{\omega }$
C20	1	1.398	0.583
C21	0.715	1	0.417

,

Table A13. C1-D Judgment matrix.

C1	D1	D2	$\mathsf{\omega }$
D1	1	1.184	0.542
D2	0.845	1	0.458

,

Table A14. C2-D Judgment matrix.

C2	D3	D4	$\mathsf{\omega }$
D3	1	1.208	0.547
D4	0.828	1	0.453

,

Table A15. C3-D Judgment matrix.

C3	D5	D6	$\mathsf{\omega }$
D5	1	1.147	0.534
D6	0.872	1	0.466

,

Table A16. C4-D Judgment matrix.

C4	D7	D8	D9	$\mathsf{\omega }$
D7	1	3.587	3.536	0.64
D8	0.279	1	1.105	0.186
D9	0.283	0.905	1	0.174

,

Table A17. C5-D Judgment matrix.

C5	D10	D11	$\mathsf{\omega }$
D10	1	1.180	0.541
D11	0.848	1	0.459

,

Table A18. C6-D Judgment matrix.

C6	D12	D13	$\mathsf{\omega }$
D12	1	0.989	0.497
D13	1.011	1	0.503

,

Table A19. C7-D Judgment matrix.

C7	D14	D15	$\mathsf{\omega }$
D14	1	0.759	0.432
D15	1.317	1	0.569

,

Table A20. C8-D Judgment matrix.

C8	D16	D17	D18	D19	D20	D21	D22	$\mathsf{\omega }$
D16	1	1.679	1.112	0.779	1.518	0.904	0.833	0.151
D17	0.596	1	1.017	1.951	1.104	0.748	2.085	0.161
D18	0.899	0.983	1	0.784	0.603	0.628	1.163	0.115
D19	1.284	0.512	1.275	1	1.629	1.477	2.524	0.178
D20	0.659	0.905	1.659	0.614	1	0.837	1.655	0.134
D21	1.106	1.337	1.592	0.677	1.194	1	0.563	0.143
D22	1.201	0.480	0.860	0.396	0.604	1.777	1	0.119

,

Table A21. C9-D Judgment matrix.

C9	D23	D24	D25	$\mathsf{\omega }$
D23	1	2.666	1.181	0.455
D24	0.375	1	0.530	0.182
D25	0.847	1.886	1	0.363

,

Table A22. C10-D Judgment matrix.

C10	D26	D27	D28	D29	$\mathsf{\omega }$
D26	1	1.109	0.803	0.771	0.221
D27	0.902	1	0.520	0.750	0.190
D28	1.245	1.923	1	0.643	0.275
D29	1.297	1.333	1.555	1	0.314

,

Table A23. C11-D Judgment matrix.

C11	D30	D31	$\mathsf{\omega }$
D30	1	1.589	0.614
D31	0.629	1	0.386

,

Table A24. C12-D Judgment matrix.

C12	D32	D33	D34	$\mathsf{\omega }$
D32	1	3.512	3.204	0.625
D33	0.285	1	1.245	0.198
D34	0.312	0.803	1	0.177

,

Table A25. C13-D Judgment matrix.

C13	D35	D36	D37	$\mathsf{\omega }$
D35	1	0.805	1.392	0.325
D36	1.242	1	2.949	0.479
D37	0.719	0.339	1	0.196

,

Table A26. C14-D Judgment matrix.

C14	D38	D39	$\mathsf{\omega }$
D38	1	1.514	0.602
D39	0.661	1	0.398

,

Table A27. C15-D Judgment matrix.

C15	D40	D41	$\mathsf{\omega }$
D40	1	1.614	0.617
D41	0.620	1	0.383

,

Table A28. C16-D Judgment matrix.

C16	D42	D43	D44	D45	D46	$\mathsf{\omega }$
D42	1	0.850	0.706	1.384	3.413	0.217
D43	1.176	1	2.759	1.796	3.455	0.325
D44	1.416	0.362	1	3.143	2.992	0.247
D45	0.723	0.557	0.318	1	0.628	0.114
D46	0.293	0.289	0.334	1.593	1	0.098

,

Table A29. C17-D Judgment matrix.

C17	D47	D48	D49	$\mathsf{\omega }$
D47	1	0.732	0.510	0.235
D48	1.365	1	1.190	0.382
D49	1.962	0.841	1	0.384

,

Table A30. C18-D Judgment matrix.

C18	D50	D51	D52	D53	D54	$\mathsf{\omega }$
D50	1	1.510	1.436	1.365	2.068	0.276
D51	0.662	1	0.742	1.718	0.851	0.188
D52	0.696	1.348	1	0.827	1.293	0.194
D53	0.732	0.582	1.210	1	1.558	0.191
D54	0.484	1.175	0.774	0.642	1	0.152

,

Table A31. C19-D Judgment matrix.

C19	D55	D56	D57	D58	$\mathsf{\omega }$
D55	1	2.036	1.221	0.583	0.263
D56	0.491	1	1.099	0.436	0.170
D57	0.819	0.910	1	1.110	0.235
D58	1.715	2.293	0.901	1	0.332

,

Table A32. C20-D Judgment matrix.

C20	D59	D60	D61	$\mathsf{\omega }$
D59	1	0.579	0.439	0.201
D60	1.728	1	0.950	0.374
D61	2.276	1.053	1	0.425

and

Table A33. C21-D Judgment matrix.

C21	D62	D63	D64	$\mathsf{\omega }$
D62	1	1.056	1.116	0.350
D63	0.947	1	1.382	0.363
D64	0.896	0.723	1	0.287

in Appendix A) generated through the Analytic Hierarchy Process yielded all consistency ratios (CR) below the 0.10 acceptability threshold [19], with maximum CR = 0.09 (

Table 6. Judgment matrix consistency test.

Judgment Matrix	${\mathbf{\lambda }}_{\mathbf{m}\mathbf{a}\mathbf{x}}$	CI	RI	CR	Feature Vector
A1.A2.A3	3.066	0.033	0.520	0.063	${\left(\mathrm{1.985,0.485,0.529}\right)}^{\mathrm{T}}$
B1.B2.B3	3.055	0.028	0.520	0.053	${\left(\mathrm{1.573,0.773,0.654}\right)}^{\mathrm{T}}$
B4.B5.B6	3.081	0.041	0.520	0.078	${\left(\mathrm{1.311,1.007,0.682}\right)}^{\mathrm{T}}$
B7.B8	2.000	0.000	0.000	\	${\left(\mathrm{0.955,1.045}\right)}^{\mathrm{T}}$
C1.C2.C3	3.028	0.014	0.520	0.027	${\left(\mathrm{1.243,0.981,0.776}\right)}^{\mathrm{T}}$
C4.C5	2.000	0.000	0.000	\	${\left(\mathrm{1.684,0.316}\right)}^{\mathrm{T}}$
C6.C7.C8	3.097	0.048	0.520	0.093	${\left(\mathrm{1.256,0.709,1.035}\right)}^{\mathrm{T}}$
C9.C10	2.000	0.000	0.000	\	${\left(\mathrm{1.373,0.627}\right)}^{\mathrm{T}}$
C11.C12.C13.C14	4.061	0.020	0.890	0.023	${\left(\mathrm{1.476,0.884,0.589,1.050}\right)}^{\mathrm{T}}$
C15.C16.C17	3.015	0.007	0.520	0.014	${\left(\mathrm{1.332,0.852,0.816}\right)}^{\mathrm{T}}$
C18.C19	2.000	0.000	0.000	\	${\left(\mathrm{1.297,0.703}\right)}^{\mathrm{T}}$
C20.C21	2.000	0.000	0.000	\	${\left(\mathrm{1.166,0.834}\right)}^{\mathrm{T}}$
D1.D2	2.000	0.000	0.000	\	${\left(\mathrm{1.084,0.916}\right)}^{\mathrm{T}}$
D3.D4	2.000	0.000	0.000	\	${\left(\mathrm{1.094,0.906}\right)}^{\mathrm{T}}$
D5.D6	2.000	0.000	0.000	\	${\left(\mathrm{1.069,0.931}\right)}^{\mathrm{T}}$
D7.D8.D9	3.001	0.001	0.520	0.001	${\left(\mathrm{1.920,0.557,0.523}\right)}^{\mathrm{T}}$
D10.D11	2.000	0.000	0.000	\	${\left(\mathrm{1.082,0.918}\right)}^{\mathrm{T}}$
D12.D13	2.000	0.000	0.000	\	${\left(\mathrm{0.995,1.005}\right)}^{\mathrm{T}}$
D14.D15	2.000	0.000	0.000	\	${\left(\mathrm{0.863,1.137}\right)}^{\mathrm{T}}$
D16.D17.D18.D19.D20.D21.D22	7.450	0.075	1.360	0.055	${\left(\mathrm{1.054,1.125,0.802,1.246,0.936,1.003,0.834}\right)}^{\mathrm{T}}$
D23.D24.D25	3.004	0.002	0.520	0.003	${\left(\mathrm{1.366,0.545,1.090}\right)}^{\mathrm{T}}$
D26.D27.D28.D29	4.059	0.020	0.890	0.022	${\left(\mathrm{0.882,0.760,1.100,1.257}\right)}^{\mathrm{T}}$
D30.D31	2.000	0.000	0.000	\	${\left(\mathrm{1.227,0.773}\right)}^{\mathrm{T}}$
D32.D33.D34	3.011	0.005	0.520	0.010	${\left(\mathrm{1.875,0.595,0.530}\right)}^{\mathrm{T}}$
D35.D36.D37	3.032	0.016	0.520	0.031	${\left(\mathrm{0.974,1.438,0.588}\right)}^{\mathrm{T}}$
D38.D39	2.000	0.000	0.000	\	${\left(\mathrm{1.204,0.796}\right)}^{\mathrm{T}}$
D40.D41	2.000	0.000	0.000	\	${\left(\mathrm{1.235,0.765}\right)}^{\mathrm{T}}$
D42.D43.D44.D45.D46	5.351	0.088	1.120	0.078	${\left(\mathrm{1.085,1.624,1.234,0.568,0.489}\right)}^{\mathrm{T}}$
D47.D48.D49	3.032	0.016	0.520	0.031	${\left(\mathrm{0.705,1.145,1.151}\right)}^{\mathrm{T}}$
D50.D51.D52.D53.D54	5.128	0.032	1.120	0.029	${\left(\mathrm{1.378,0.942,0.968,0.953,0.758}\right)}^{\mathrm{T}}$
D55.D56.D57.D58	4.147	0.049	0.890	0.055	${\left(\mathrm{1.051,0.682,0.941,1.327}\right)}^{\mathrm{T}}$
D59.D60.D61	3.006	0.003	0.520	0.005	${\left(\mathrm{0.604,1.123,1.274}\right)}^{\mathrm{T}}$
D62.D63.D64	3.008	0.004	0.520	0.008	${\left(\mathrm{1.051,1.088,0.861}\right)}^{\mathrm{T}}$

). This statistically robust consistency validation finalized the indicator weights underpinning the evaluation framework. The “Feature vector” in Table 6 denotes the principal eigenvector corresponding to ${\lambda }_{\mathrm{m}\mathrm{a}\mathrm{x}}$ of each judgment matrix, representing the raw relative importance of indicators before normalization. When normalized, this eigenvector gives the relative weights (denoted as ω in Appendix A tables), which are then applied in the hierarchical aggregation of indicators.

To further test the robustness of the weighting results, a one-at-a-time sensitivity analysis was conducted. Each expert was removed in turn, and the AHP repeated to recalculate weights at the objective level. As shown in

Table 7. Weights of primary dimensions under expert removal scenarios.

Scenario	Ecological Protection System A1	Management System A2	Public Service System A3
Baseline (all experts)	0.662	0.162	0.176
Remove Expert 1	0.656	0.161	0.183
Remove Expert 2	0.658	0.165	0.176
Remove Expert 3	0.664	0.158	0.178
Remove Expert 4	0.654	0.161	0.185
Remove Expert 5	0.656	0.159	0.185
Remove Expert 6	0.670	0.166	0.164
Remove Expert 7	0.685	0.162	0.153
Remove Expert 8	0.658	0.167	0.175
Remove Expert 9	0.663	0.164	0.173
Remove Expert 10	0.653	0.168	0.178
Remove Expert 11	0.654	0.159	0.187
Remove Expert 12	0.653	0.165	0.181
Remove Expert 13	0.668	0.153	0.179
Remove Expert 14	0.656	0.169	0.176
Remove Expert 15	0.664	0.168	0.167
Remove Expert 16	0.661	0.166	0.173

, the weight of the ecological protection system (A1) fluctuates only within the range of 0.653–0.685, consistently remaining the highest among the three primary dimensions, while the management system (A2: 0.153–0.169) and public service system (A3: 0.153–0.187) show only minor variations. In addition, a perturbation test was carried out by adjusting selected pairwise comparison values in the final judgment matrix by ±10%. The results (

Table 8. Weights of primary dimensions under ±10% perturbation scenarios.

Scenario	Ecological Protection System A1	Management System A2	Public Service System A3
Baseline (all experts)	0.662	0.162	0.176
+10% perturbation	0.681	0.158	0.161
–10% perturbation	0.640	0.166	0.194

) indicate that even with such changes, the overall ranking of the three primary dimensions is preserved, with A1 still dominant and only limited shifts observed for A2 and A3.

3.2.1. The Weights of Items in the Primary Dimensions

As shown in Figure 3, the Analytic Hierarchy Process assigned prioritization weights across three objective layers, with ecological integrity receiving the highest weighting (66.18%)—reflecting China’s “ecological primacy” doctrine in national park governance. Public service delivery (17.64%) and management efficacy (16.18%) followed, demonstrating the former’s critical bridging role between conservation imperatives and societal needs. This weighting structure ensures the evaluation system holistically captures long-term conservation outcomes while maintaining policy coherence with national strategic directives.

3.2.2. The Weights of Items in the Criterion Layer

Figure 4 provides a clear indication that within the national park ecosystem hierarchy, National Representativeness (34.70%) constitutes the paramount indicator, followed by Ecosystem Authenticity and Ecosystem Integrity. For management systems, Management Foundations and Safeguarding Mechanisms collectively outweigh Management Processes. Conversely, in public services, Regional Development (9.22%) marginally exceeds Recreation and Education in weighting.

This distribution reflects core evaluation priorities: National representativeness—fundamental to a park’s legitimacy—receives the highest weighting (34.70%), while Regional Development emerges as the top non-ecological indicator (9.22%) given its critical role in sustaining ecological-economic-social synergies. Although management components remain essential, their lower relative weights prove justifiable since they primarily serve protective functions rather than direct conservation outcomes. Ultimately, the predominance of ecological metrics (particularly representativeness) aligns with national parks’ intrinsic value proposition, whereas service-oriented weights prioritize developmental sustainability.

3.2.3. The Weights of Items in the Element Layer

Factor-level weighting clarifies critical priorities for national park development, with Figure 5 revealing that Spatial Representativeness (14.38%) and Natural Authenticity (14.36%) constitute the highest-weighted factors—demonstrating their primacy in conservation objectives. Within management systems, Management Entities emerge as the dominant component (4.85%). For public services, Recreational Experience (5.46%) and Economic Development (5.38%) form closely aligned core priorities, collectively highlighting their operational significance within the service delivery framework.

3.2.4. The Weights of Items in the Indicator Layer

The final weighting of each indicator is derived through multiplicative aggregation of its hierarchical weights across all superior levels up to the evaluation target layer (Figure 6). Critical focal points emerge within subsystems: Natural Habitat Coverage constitutes the highest-weighted ecological indicator, affirming its primacy in national park assessment; Person-Position Matching dominates management criteria; Transportation Accessibility is paramount for recreation, while Eco-Product Value Realization leads economic development metrics. Among all 64 indicators, ecological metrics occupy nine of the top ten weights—spearheaded by Natural Habitat Coverage (9.19%) alongside Regional Representativeness and Landscape Value Integrity. This distribution underscores ecosystem integrity as the cornerstone of China’s national park framework, deliberately reflecting the foundational principle that any park compromising natural habitat representativeness or succumbing to excessive anthropogenic interference inherently fails as a legitimate Chinese national park.

Notably, other high-impact indicators include Staff Competency Development, Functional Zoning Clarity, Environmental Monitoring Coverage, and Commercial Project Implementation Effectiveness—collectively highlighting operational and governance dimensions critical to park functionality beyond core ecological priorities. And the detailed scoring standards and case study results are provided in Appendix B.

3.3. Comprehensive Evaluation of the Nanshan National Park

Combining the weight distribution of the indicator system and the specific indicator scores, this section starts from the synergy of the ecosystem, management system and public service system, and in combination with the key contradictions in the construction of national parks mentioned earlier (the contradiction between public power and private power, and the problem of the imperfect governance decision-making system), deeply explores the main factors affecting the evaluation results.

3.3.1. The Analysis of the Scores of Items in Criterion Layer

The scoring rates of the evaluation indexes in the criterion layer are all higher than 80% (

Table 9. Scores of items in the criterion layer.

Criterion Layer	Score	Full Score	Scoring Rate
National representativeness B1	31.44	34.69	90.64%
Ecosystem authenticity B2	15.31	17.05	89.79%
Ecosystem integrity B3	12.13	14.42	84.12%
Management Foundation B4	6.13	7.07	86.70%
Management Support B5	5.04	5.43	92.82%
Management process B6	3.32	3.68	90.22%
Recreational education B7	6.79	8.43	80.55%
Regional development B8	7.61	9.21	82.63%

), which indicates that the overall situation is good. Among them, the representation of national parks exceeds 90%, indicating that Nanshan National Park currently stands out in terms of ecological location, species representation and distinctive landscapes. Meanwhile, the score rate of management guarantee and management process also exceeds 90%, suggesting that Nanshan National Park is complete and efficient in ecological protection and management, and the direction of management guarantee and management process is correct. However, it should also be noted that Nanshan National Park scored less than 85% in terms of ecosystem integrity, recreational education and regional development indicators. This means that compared with other criterion-level indicators, the construction in recreational education and regional development is insufficient.

Based on the above statistical results, the bar chart of the score of the indicator layer of the evaluation system for the construction and development of Nanshan National Park is sorted out as shown in Figure 7. The values circled in red in the figure are the indicators with scores lower than 80, and they are also the indicators that Nanshan National Park needs to pay the most attention to and improve at present.

3.3.2. The Analysis of the Scores of Items in the Indicator Level

Spatial–statistical analysis (Figure 7) reveals 7 critical bottlenecks (<80 points) among NNP’s 64 evaluation metrics. The severe landscape fragmentation (D14 = 50) reflects historical land-use conflicts in buffer zones. Humus thickness (D21 = 78) indicates a relatively thin humus layer, reflecting lower soil organic matter and weaker nutrient cycling, which may be associated with historical human disturbance or land-use changes. Clarity of land ownership (D38 = 78) points to tenure ambiguities that complicate unified management and hamper effective compensation. Transportation accessibility (D50 = 75) remains limited, constraining visitor access and regional service integration. Richness of recreational products (D52 = 65) is low, indicating insufficient diversity of cultural and educational offerings. Training for tourists before entering (D56 = 70) is inadequate, exposing the park to inappropriate visitor behavior and weakening environmental education outcomes. Suboptimal Engel coefficients (D62 = 76) indicate persistent livelihood precarity despite tourism growth—contradicting “ecological civilization” co-benefit objectives.

To further illustrate how the weighted aggregation was derived, Figure 8 presents the product of each indicator’s weight and score, forming the basis of the total score.

3.3.3. Comprehensive Evaluation of Nanshan National Park

It is clear from the result of the calculation that the comprehensive score of Nanshan National Park is 87.77 points (out of 100), among which the management system (89.56) performs best, followed by the ecological dimension (88.97), while the public service dimension (81.63) lags behind. This suggests that governance and institutional safeguards in Nanshan are comparatively more developed, whereas service delivery to communities and visitors remains an area needing improvement. Taken together, these results highlight the differentiated performance of the three dimensions, providing a useful basis for understanding their interactions. Nanshan National Park demonstrates a dynamic interplay among its three core components: the ecosystem serving as the foundational asset with significant ecological and economic value; the management system enabling sustainable resource utilization through effective governance; and the public service system converting ecological value into socio-economic benefits while funding conservation. Synergistic development across these dimensions is essential for maximizing ecological, economic, and social outcomes.

(1)

Ecological protection system: Conservation Success vs. Landscape Fragmentation. The ecosystem sub-goal achieved relatively high performance (accounting for 66.18% of total weight with 58.88% score contribution), evidenced by strong metrics including Natural Habitat Coverage (D7), Species Richness (D12: 90/100), and Biodiversity Index (D13: 88/100). However, Landscape Fragmentation (D14: 50/100) emerged as a critical weakness. Remote sensing analysis confirms widespread habitat dissection by roads and scattered settlements, indicating insufficient ecological corridor connectivity between core protected zones. This reflects structural tensions between rigid conservation policies and traditional land-use practices.

(2)

Management system: Institutional Strength Constrained by Limited Stakeholder Engagement. Management systems scored robustly overall, particularly in Managerial Expertise (C9: 88/100), with staff competency (D25: 90/100) being exceptional. Yet Stakeholder Participation (C17) underperformed significantly, characterized by low Volunteer Involvement (D47) and minimal NGO Project Engagement (D48). Field studies reveal restricted community access to decision-making processes and failure to integrate traditional ecological knowledge, resulting in disjointed conservation implementation.

(3)

Public service system: Modest Internal Disparities Hindering Regional Synergy. Overall, the public service system attained an 81.63% score rate, with modest internal disparities: Recreation and Education Services (B7: 80.55/100) showed uneven outcomes (e.g., Richness of Recreational Products D52: 65/100 vs. Completeness of Recreational Facilities D51: 90/100), whereas Regional Development (B8: 82.63/100) performed slightly better overall. Notably, Community Economic Benefits (D60) from park-related industries remained limited despite high Eco-Product Value Realization (D61: 85/100). Fixed operational expenditures (D34) dominate funding allocations, crowding out community development resources—highlighting systemic flaws in market-based mechanisms and sustainable financing.

(4)

Balancing Conservation and Community Welfare. Although ecological compensation (D44) partially mitigates conflicts, persistent challenges remain, as evidenced by community Engel’s Coefficients exceeding 30% (D62). This indicates only preliminary success in reconciling protection and development. Compensation standards fall below land opportunity cost thresholds, necessitating strategies like flexible zoning management (e.g., seasonal grazing rotations) and concession revenue-sharing to internalize externalities.

Nanshan’s evaluation reflects both achievements in China’s national park development and transitional challenges: rigid ecological policies vs. insufficient community adaptation flexibility; bureaucratic management vs. demands for polycentric governance; and untapped economic potential despite rich ecological capital.

3.3.4. Recommendations for the NNP

Based on the comprehensive evaluation results, Nanshan National Park demonstrates notable achievements in ecosystem integrity, management capacity, and public service delivery. However, persistent challenges such as landscape fragmentation, insufficient stakeholder participation, regional development gaps, and inadequate ecological compensation remain. To address these issues and promote sustainable progress, the following recommendations are proposed.

(1)

Enhancing ecological connectivity and mitigating landscape fragmentation.

The evaluation highlighted significant weaknesses in habitat connectivity. Roads, scattered settlements, and historical land-use practices have fragmented the park’s ecosystems, restricting wildlife migration, reducing habitat size, and weakening genetic exchange among species. This fragmentation also intensifies edge effects and increases the vulnerability of threatened species, undermining long-term conservation goals.

To mitigate these challenges, priority should be given to the restoration of ecological corridors and degraded landscapes. Low-efficiency farmland and degraded plantations can be targeted for ecological restoration, including reforestation with native species and vegetation buffers. Abandoned plots may be converted into natural regeneration zones to serve as stepping stones between isolated habitats. In addition, low-traffic roads in sensitive areas should be considered for closure or seasonal restrictions to reduce ecological disturbance. Land-use zoning should strictly enforce ecological redlines, limiting construction and settlement expansion in critical habitats. These measures would gradually restore landscape connectivity and strengthen the overall ecological integrity of the park.

(2)

Building participatory governance mechanisms and expanding stakeholder involvement.

Although Nanshan National Park exhibits strong professional management capacity, the level of community and NGO participation remains weak. Volunteer engagement is limited, and local residents have minimal influence in decision-making processes. This lack of inclusion restricts the integration of traditional ecological knowledge, reduces community ownership of conservation outcomes, and risks policy-practice misalignment.

To address these governance gaps, formal participatory mechanisms should be established. Community councils or co-management platforms can be created to ensure that local residents, NGOs, and academic institutions are systematically engaged in planning, monitoring, and enforcement. Regular stakeholder meetings would provide opportunities for dialogue and trust-building, helping to reduce conflicts between conservation and livelihood needs. Moreover, incorporating traditional ecological knowledge into park management—such as community expertise in resource use, species behavior, or ecological cycles—would complement scientific methods and increase compliance. Incentives such as ecological protection funds, volunteer credit systems, or recognition programs could further encourage wider public participation in conservation.

(3)

Optimizing public service delivery and promoting regional synergy.

The evaluation indicated that while the public service system scored well overall, there are imbalances between infrastructure development and community benefits. Tourism accessibility and recreational services are strong, but regional development indicators are weaker. Much of the operational budget is tied to fixed expenditures, leaving insufficient resources for supporting community development. As a result, tourism growth has not translated into proportional economic gains for local residents, creating tensions between conservation and livelihood needs.

To overcome these shortcomings, a more equitable benefit-sharing model should be developed. Community-based tourism initiatives, such as eco-lodges, cultural interpretation programs, and educational tourism, should be supported through partnerships with local households. Branding and marketing of local eco-products—such as specialty tea, medicinal herbs, and organic produce—would increase income and diversify livelihood opportunities. At the same time, a restructuring of financial allocations is necessary. Reducing rigid operational spending and reallocating resources toward community services, education, and livelihood programs can improve the park’s long-term sustainability and strengthen regional support for conservation.

(4)

Improving ecological compensation schemes and balancing conservation with livelihoods.

Despite existing compensation programs, the evaluation revealed that current standards remain below the opportunity cost of land use, resulting in inadequate livelihood support for local communities. Engel’s coefficient analysis shows that household living burdens remain relatively high, suggesting that ecological protection policies have not yet been fully reconciled with community welfare. If unaddressed, this misalignment may undermine community support for conservation in the long run.

To address this challenge, ecological compensation levels should be recalibrated to more accurately reflect actual livelihood losses. A dynamic adjustment mechanism should also be established to ensure compensation remains fair under changing economic conditions. In addition, diverse compensation strategies should be introduced. These could include direct revenue-sharing from tourism projects, profit returns from eco-product sales, and the implementation of flexible land-use arrangements such as rotational grazing, seasonal harvesting, or community-managed resource use. These measures would enhance local acceptance of conservation policies, promote livelihood security, and improve the balance between ecological protection and community development.

4. Conclusions

This study has developed China’s first comprehensive evaluation framework for national park development, integrating ecosystem integrity, managerial efficacy, and public service delivery through a four-tiered hierarchical structure (3 objective layers → 8 criterion layers → 21 factor layers → 64 indicator layers). Validated through rigorous application in Nanshan National Park—which achieved an overall score of 87.77 (Grade II: “Proficient”)—the system demonstrates robust diagnostic capabilities. Its core innovation lies in embedding conflict-sensitive metrics addressing land-right tensions and institutional immaturity within China’s emerging conservation paradigm.

The framework functions as a dynamic polycentric governance tool, balancing standardized oversight with contextual adaptation. In Nanshan’s case, it precisely identified critical challenges: severe landscape fragmentation (D14: 50/100), inadequate transportation accessibility, and underdeveloped recreation offerings. These findings empirically validate three priority interventions: establishing ecological corridors to enhance habitat connectivity, upgrading transportation infrastructure, and diversifying cultural-educational products (e.g., revolutionary heritage experiences, ethnic minority cultural programs).

As China’s national park system expands, this framework offers dual scalability: Cross-park comparability through standardized core metrics (e.g., 34.70% weighting for National Representativeness); Contextual adaptability via modular add-ons for regional attributes (e.g., indigenous knowledge indicators). Such flexibility enables knowledge transfer from pioneer parks like Nanshan to newer reserves while maintaining ecosystem primacy—exemplified by Regional Development’s status as the top non-ecological indicator (9.22%). Ultimately, this system redefines conservation success through synergistic resilience, where ecological legitimacy (≥34.70% representativeness), community prosperity, and adaptive governance co-evolve. It establishes an actionable pathway for China to achieve its dual mandate of ecological preservation and sustainable development within protected areas.

At the system level, the findings from Nanshan are not isolated but rather symptomatic of broader challenges across China’s national park pilots and newly designated parks. Despite significant progress since the official launch of the national park system, issues such as fragmented landscapes, insufficient stakeholder participation, imbalanced regional development, and underdeveloped compensation schemes remain widely observed. These structural tensions reflect the inherent complexity of reconciling rigid ecological protection mandates with the socio-economic realities of densely populated regions.

Nevertheless, important reforms are already underway. Ecological compensation mechanisms are being piloted in multiple parks, seeking to recalibrate standards to better reflect opportunity costs borne by local communities. Parallel to this, “ecological mobilization” initiatives—such as community-based conservation cooperatives, volunteer ranger programs, and participatory planning platforms—are gradually expanding, signaling a shift toward more inclusive governance. At the same time, national-level policy guidance has begun to promote adaptive zoning, flexible land-use arrangements, and cross-regional coordination, aimed at mitigating fragmentation and enhancing ecological connectivity.

These ongoing efforts suggest that China’s national park system is in a dynamic stage of institutional maturation. While the evaluation framework developed in this study requires further refinement and validation across multiple cases, its application to Nanshan demonstrates the feasibility of diagnosing systemic weaknesses and guiding policy responses. As ecological compensation, community mobilization, and adaptive management practices continue to evolve, the framework can serve as a critical monitoring and decision-support tool to align conservation imperatives with social welfare objectives. In this sense, the evaluation of Nanshan not only provides insights for a single pilot zone but also contributes to shaping the trajectory of national park governance in China as a whole.

Methodologically, while the Analytic Hierarchy Process (AHP) and fuzzy scoring provided the most viable weighting and evaluation approaches available, two limitations warrant acknowledgment. First, expert-dependent weight determination—though systematically implemented—introduces subjectivity; future iterations should integrate objective methods (e.g., entropy weighting) as multi-park datasets expand. Moreover, while most indicators are based on official statistics or field survey data, a few (e.g., residents’ awareness of national parks) rely more heavily on expert judgment. This reliance introduces subjectivity, yet it does not necessarily undermine scientific validity. Expert evaluation, when grounded in long-term experience and interdisciplinary knowledge, often captures nuanced socio-ecological realities that are difficult to quantify through purely objective metrics. Thus, the combination of expert judgment and data-based evaluation represents a balanced and pragmatic approach under current data constraints, ensuring both rigor and contextual relevance in the interpretation of early-warning signals. Second, rapid institutional evolution (e.g., park boundary adjustments, tenure reforms) challenges real-time assessment due to insufficient dynamic monitoring platforms. This may have partially underestimated Nanshan’s recent progress in management safeguards and regional development. And this framework is terrestrial-only.

It is worth noting that while the present framework applies only to terrestrial national parks, marine conservation areas constitute an essential and emerging component of China’s future national park system. At this stage, the lack of consistent ecological baselines, governance data, and long-term monitoring mechanisms makes it premature to directly apply the terrestrial framework to marine contexts. A unified assessment tool across both systems would risk oversimplifying their fundamental ecological and managerial differences. Accordingly, the current study explicitly focuses on terrestrial systems, and we acknowledge this as a methodological limitation rather than an omission. Future research could focus on developing a dedicated, scientifically grounded framework for marine protected areas—one that develops its own indicators, weighting methods, and evaluation models tailored to the dynamic and transboundary nature of marine ecosystems. Such advancement would complement the present work and contribute to building a truly comprehensive evaluation system encompassing both terrestrial and marine national parks within China’s evolving protected area network.

Finally, it is also important to consider how under-performing pilot zones might be addressed within the evolving framework. A phased adjustment pathway has been raised as a sensitive yet necessary consideration. Its purpose is corrective, ensuring that the highest category of protected areas continues to represent the most effective models of conservation. In this study, the indicator system can serve as a practical foundation for such a process by identifying early warning signals across ecological protection, management, and public service systems. Applied to the five transitional pilot zones that were not formally designated, the framework provides a transparent scoring basis to judge whether they meet the expected standards. Low scores would trigger early warnings and indicate the need for targeted rectification before any decision on formal designation is made. Only if serious deficiencies remain unresolved, or core conservation values are irreversibly lost, should withdrawal be considered. In this sense, adjustment is better understood as a dynamic process rather than simple abandonment, strengthening the credibility and resilience of China’s national park reform.