The “Ten-Year Fishing Ban” policy in the Yangtze River Basin aims to restore ecological diversity but poses significant challenges for the fishermen in their transition to alternative livelihoods. This study focuses on fishermen who worked on Poyang Lake, using the sustainable livelihood framework and the theory of planned behavior, combined with fuzzy-set qualitative comparative analysis (fsQCA) and descriptive statistics, to explore the interaction effects of livelihood capital, ecological cognition, and compensation policies on fishermen’s behavioral responses. Key findings include the following: natural, financial, and psychological capital are core drivers of enhanced ecological cognition, with combined effects significantly increasing sensitivity to policy and environmental changes through pathways like “ecological transition drive”, “knowledge adaptation support”, and “multi-cooperation synergy”. Attitude, perceived behavioral control, and compensation policy transparency are crucial for positive responses, while social norms and policy fairness can compensate for individual motivation deficits, forming pathways like “ecological drive-policy recognition” and “norm drive-social support”. Current issues such as low compensation standards, insufficient retraining, and gender differences limit policy effectiveness. Optimizing measures like differentiated fishing permits, dynamic compensation mechanisms, and cultural empowerment are needed to balance ecological protection and social equity. The study suggests enhancing financial and psychological capital, improving vocational training systems, and increasing policy transparency to provide theoretical and practical references for sustainable global fisheries management.

Food resources are the fundamental basis for the survival and reproduction of animals. Diet research is the foundation for understanding their ecological habits and is of great significance for evaluating their survival status and carrying out effective protection and management. South China sika deer (Cervus nippon kopschi) is the most endangered subspecies of wild sika deer in China, with a small population and a shrinking distribution area. Here, we used DNA metabarcoding technology to study the diet composition of C. n. kopschi in Taohongling Sika Deer National Nature Reserve in Jiangxi, China. Comparative analysis of diet composition among different areas (A–E) of C. n. kopschi was conducted, as well as C. n. hortulorum raised in the same areas. We found that the dominant families in the diet composition of C. n. kopschi were Rosaceae (46.73% of relative abundance), Anacardiaceae (6.02%), Poaceae (5.54%), and Fabaceae (3.92%), with Rubus (45.43%) being the absolute dominant genus. Thirty-two preferred plant species were identified in the diet of C. n. kopschi. The highest relative abundance (45%) was Rubus reflexus. The dominant family and genus in the diet composition of C. n. hortulorum were Fabaceae (33.89%) and Pueraria (32.87%), respectively. Of the 15 preferred plant species, the highest relative abundance was Pueraria montana (33%). The richness, diversity, and evenness of diet composition in Area B of C. n. kopschi were the highest among all areas, with significant differences compared to Areas C, D, and E. Diet composition of deer in Area A did not differ significantly from those in other areas. Principal coordinate analysis (PCoA) and nonmetric multidimensional scaling (NMDS) both indicated a significant separation in the diet composition of C. n. kopschi and C. n. hortulorum, while Area B showed significant separation from the other areas. This study elaborates on the diet composition information of C. n. kopschi and can provide a reference for the protection and improvement of the habitat of sika deer.

Understanding species–environment relationships and habitat overlap among sympatric ungulates is crucial for biodiversity conservation in mountainous ecosystems. Using a three‐year camera‐trap dataset (2021–2024) from Tongboshan National Nature Reserve, Jiangxi, China, we applied MaxEnt modeling to evaluate habitat suitability and niche overlap of Black muntjac ( Muntiacus crinifrons ), Reeves's muntjac ( Muntiacus reevesi ), and Wild boar ( Sus scrofa ). After spatial filtering, 103 independent occurrence records were analyzed against 12 environmental predictors representing topography, vegetation, soil, water, and anthropogenic disturbance. All models performed well (AUC > 0.90), with vegetation and topography contributing most to species distributions. Black muntjac exhibited a narrow spatial niche breadth, being strongly associated with mid‐ to high‐elevation forests and close proximity to water sources, whereas Reeves's muntjac showed broad ecological flexibility, adapting to coniferous forests and moderately disturbed landscapes. Wild boar demonstrated generalist strategies, favoring forest edges and bamboo stands. Predicted suitable habitats covered 16.75%, 41.54%, and 31.17% of the reserve for Black muntjac, Reeves's muntjac, and Wild boar, respectively. Niche overlap analysis revealed the highest overlap between Reeves's muntjac and Wild boar ( D = 0.822; I = 0.969), while Black muntjac showed limited overlap with either species. Species‐specific habitat preferences and spatial partitioning emerge as key mechanisms enabling coexistence. We recommend prioritizing the maintenance of closed‐canopy forest habitats for Black muntjac as the most critical conservation strategy, alongside enhancing forest heterogeneity in overlap zones and establishing ecological corridors to improve landscape connectivity. Our study provides quantitative evidence to guide adaptive management and support multi‐species conservation in subtropical reserves.

Habitat loss and fragmentation threaten forest-dependent ungulates in subtropical mountain systems, yet integrative assessments linking habitat quality and landscape configuration remain limited. Here, we evaluated habitat suitability and identified core habitat patches for the endangered black muntjac (Muntiacus crinifrons) in Tongboshan National Nature Reserve using an Analytic Hierarchy Process–Habitat Suitability Index (AHP–HSI) framework integrated with camera-trap validation and landscape pattern analysis. Vegetation-related indicators (NDVI and vegetation type) were the dominant suitability drivers, and highly suitable habitats accounted for 62.9% of the reserve (8646.97 ha), forming three major forest blocks with low disturbance levels. Camera-trap detections (n = 58) showed strong concordance with model predictions (98.28% within moderately suitable or higher classes). Landscape metrics revealed contrasting spatial configurations between overall high-suitability habitats and optimal core patches, indicating that demographic source areas are embedded within fragmented peripheral mosaics. Medium patches and forested ridges may function as potential stepping stones and corridors facilitating movement across habitat clusters. These findings highlight the importance of maintaining functional connectivity and mitigating edge disturbances in buffer and experimental zones to ensure long-term population persistence and effective protected-area management for forest ungulates.

Elucidating mechanisms maintaining mountain biodiversity is a central goal in ecology. Integrating phylogenetic and functional dimensions helps understand community assembly. Jiangxi Tongboshan National Nature Reserve, a key subtropical forest, has lacked systematic surveys of birds and mammals. Based on 113 camera-trap sites (2020 - 2024), we recorded 56 species (21 mammals, 35 birds), including 15 nationally protected species. Using taxonomic, functional, and phylogenetic diversity, together with generalized additive models, variation partitioning, and the vertical connectivity hypothesis, we explored community assembly mechanisms. Functional diversity (FDis) exhibited no significant environmental gradient effect in either mammals or birds, suggesting functional redundancy. For mammals, standardized effect size of phylogenetic diversity (SES.PD) exhibited significant nonlinear variation with elevation across all sites (p = 0.0316), and this effect was stronger in evergreen broadleaf forests (p = 0.0116, R² = 0.356). Net relatedness index (NRI) shifted from negative to positive at approximately 494 m (p = 0.00265, R² = 0.426), directly supporting the stress-gradient hypothesis. Birds showed no significant elevational effect. Variation partitioning analysis (VPA) showed very low environmental explanation (<6.2%). The vertical connectivity hypothesis was supported only in mammals (p = 0.001), with β diversity between low and high elevations significantly greater than between low medium and medium high elevations, indicating vertical dispersal limitation. This study reveals a phylogenetic shift from competitive exclusion to environmental filtering along elevation in evergreen broadleaf forests, emphasizes the value of integrating phylogenetic dimensions and habitat-specific analyses, and provides a scientific basis for maintaining elevational gradient integrity in nature reserve.