Environmental DNA reveals coral-driven shifts in reef fish networks: Implications for conservation

Environmental DNA reveals coral-driven shifts in reef fish networks: Implications for conservation | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Environmental DNA reveals coral-driven shifts in reef fish networks: Implications for conservation Cong Zeng, Zheying Lin, Zhiyi Su, Yuanbin Zhao, Yue Liu, Mingjie Li, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9465484/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 6 You are reading this latest preprint version Abstract Coral reef fish communities are highly sensitive to changes in coral habitat condition, yet the ecological trajectories and non-linear responses of these communities remain poorly understood. In the biodiverse South China Sea, we used environmental DNA metabarcoding to profile reef-fish assemblages across sites categorized as Excellent, Moderate, or Poor in coral condition. Integrating alpha and beta diversity metrics with ecological network analysis and piecewise structural equation modeling (SEM), we quantified how fish communities shift along this degradation gradient. Moderate coral decline was associated with (i) a rise in species richness, (ii) higher network modularity, and (iii) an increase in core molecular operational taxonomic units (MOTUs). In contrast, severely degraded reefs exhibited abrupt collapses in species richness and network stability, compositional homogenization, and loss of core taxa, revealing a hysteretic, non-linear response. SEM identified coral condition as the dominant direct driver of fish community structure, mediated by habitat complexity, and as an indirect driver via fishing pressure. Intensified fishing simultaneously degraded coral (r = -0.21) and reduced species richness (r = -0.25), amplifying biodiversity loss. Our results reveal stage-dependent resilience limits beyond which reef-fish communities shift abruptly, underscoring the need for integrated management that addresses both global climate stressors and local anthropogenic impacts. Protecting coral integrity while curbing fishing pressure is essential to maintain ecosystem function and biodiversity, especially in geopolitically contested marine regions. Coral reef ecosystem Fish community structure Environmental DNA Ecological network South China Sea Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 1 Introduction Climate change and human disturbances are accelerating marine biodiversity loss, with well-documented impacts on ecosystem functioning and resilience (Cardinale et al. 2012 ; IPCC 2022 ). Coral reefs, often described as the “rainforests of the sea”, have experienced dramatic declines, with over half of live coral cover lost since the 1950s due to ocean warming-induced bleaching, disease and predator outbreaks, and direct human impacts including overfishing, sedimentation, and pollution (Dunne et al. 2004 ; Mora et al. 2016 ; Hughes et al. 2017 ; Gove et al. 2023 ). These declines compromise the structural complexity and three-dimensional habitats that reef-building corals provide, altering shelter availability, trophic resources, and overall habitat quality. As a consequence, reef fishes, which rely on these habitats for foraging, reproduction, and refuge from predators, are directly affected. Reef fish assemblages play key roles in maintaining trophic balance and regulating algal growth (Zhang et al. 2014 ; Graham et al. 2017 ; Fontoura et al. 2020 ), and are widely used as bio-indicators of reef health, offering early-warning signals for ecosystem disruption (Darling et al. 2017 ; Stier et al. 2025 ). However, traditional measures of species richness or composition often fail to capture the broader ecological consequences of habitat degradation, particularly the disruption of species interactions, functional roles, and emergent properties that govern community resilience. Beyond alpha and beta diversity, coral habitat degradation can reshape the interactions and organization of reef fish communities. Obligate coral specialists, such as Dascyllus aruanus and corallivorous butterflyfishes ( Chaetodon spp.), typically decline sharply following coral loss, whereas generalist species may persist or temporarily expand, altering trophic interactions, competitive dynamics, and network modularity (Pratchett et al. 2008 , 2012 , 2015 ; Alvarez-Filip et al. 2011 ; Olán-González et al. 2020 ). Habitat attributes such as structural complexity, coral morphology, bleaching condition, and benthic heterogeneity mediate these interactions by providing refugia, facilitating niche complementarity, and buffering predation or competition (Kerry and Bellwood 2015 ; Darling et al. 2017 ; Clements and Choat 2018 ; Fisher 2023 ). The simplification or loss of these microhabitats can reduce functional redundancy, weaken trophic links, and decrease network robustness, potentially triggering community-level tipping points, hysteresis, or delayed recovery even after stressors are alleviated (Graham et al. 2006 ; Darling et al. 2017 ; McWilliam et al. 2020 ; Santoso et al. 2022 ). Examining how coral habitat condition influences species composition and ecological network structure can help identify key species and interactions that support community stability (Pozas-Schacre et al. 2021 ; Wolfe et al. 2021 ; Hill and Hoogenboom 2022 ; Young et al. 2024 ). Despite extensive research on richness and beta diversity, the effects of coral habitat condition on fish ecological networks remain poorly understood. Traditional biodiversity metrics, such as species richness or diversity indices, often fail to capture the underlying species interactions, functional redundancy, or modular organization that govern ecosystem stability (Montoya et al. 2006 ; Tylianakis et al. 2010 ). Network-based approaches provide a powerful framework to quantify these properties, including connectivity, centrality of keystone or hub species, modularity, and redundancy of functional roles, all of which are critical for understanding how communities maintain resilience under environmental and anthropogenic stress (Pozas-Schacre et al. 2021 ; Wolfe et al. 2021 ; Hill and Hoogenboom 2022 ; Young et al. 2024 ). Highly connected nodes or module hubs may buffer against species loss by sustaining key trophic interactions, whereas the breakdown of such nodes can trigger cascading effects and community destabilization (McCann 2000 ; Dunne et al. 2004 ). Functional redundancy among species performing similar ecological roles can also mitigate the impacts of habitat degradation, allowing ecosystems to retain functionality even as some species decline (Bellwood et al. 2004 ; Mouillot et al. 2014 ). Changes in coral habitat condition, including reduced structural complexity, shifts in coral morphology, and habitat fragmentation, tend to affect module hubs and high-trophic-level species more strongly. As a result, network connectivity declines, energy flow is disrupted, and community assembly processes are altered (Santoso et al. 2022 ; Young et al. 2024 ). These network-level effects often precede observable declines in richness or diversity, highlighting the importance of considering species interactions and network architecture when assessing the ecological consequences of coral degradation. Network-based perspectives also provide practical guidance for coral reef management and restoration. Metrics derived from ecological networks can serve as early-warning indicators of functional collapse, helping managers detect ecosystem destabilization before species losses become irreversible (Emslie et al. 2020 ; McWilliam et al. 2020 ). They can inform the selection of sites for coral transplantation or habitat enhancement by identifying reefs with sufficient structural complexity to support key functional interactions, including predator-prey relationships and trophic regulation. Post-restoration monitoring can then evaluate whether functional redundancy is re-established, module hubs recover, and energy and interaction flows within the community are restored, thereby supporting adaptive management and enhancing long-term ecosystem resilience (Luza et al. 2024 ). By linking habitat condition, species interactions, and network structure, this approach helps clarify how coral degradation reshapes fish communities and can inform reef restoration and management. However, incorporating network metrics into restoration and monitoring is often constrained by the need for high-resolution data on species presence and interactions. Recent advances in molecular ecological network (MEN) methods, such as RMT-based network construction (Zhou et al. 2010 ; Deng et al. 2012 ), provide robust frameworks for defining ecological networks from high-throughput data. Together with environmental DNA (eDNA) metabarcoding, which is defined here as genetic material shed by macroorganisms and suspended in environmental samples such as water (Miya 2022 ) and allows efficient detection of fish assemblages, including rare or cryptic taxa, these methods enable the estimation of ecological networks in reef settings in which traditional surveys would be impractical(Meyer et al. 2020 ). The South China Sea (SCS) is a global hotspot of coral and fish biodiversity but is increasingly threatened by anthropogenic activities and climactic stress(Huang et al. 2015 ; Pang et al. 2021 ; Shi et al. 2022 ). Its offshore archipelagos exhibit pronounced spatial gradients in coral habitat condition, offering a natural system to investigate how fish communities respond to habitat transformation. In this study, we applied eDNA metabarcoding to characterize reef fish assemblages across three archipelagos, integrating alpha and beta diversity analyses with ecological network metrics. Specifically, we aimed to (1) quantify differences in fish community composition and network structure among coral habitat conditions, (2) explore whether these differences exhibit stage-mediated or hysteretic patterns, and (3) evaluate how local anthropogenic pressures, particularly fishing intensity, influence these responses. By combining eDNA metabarcoding with network analysis, this study examines how coral habitat condition influences fish community structure and interaction patterns, with implications for reef monitoring and management. (Emslie et al. 2020 ; McWilliam et al. 2020 ). 2 Material and methods 2.1 Habitat condition assessment During June 2023 and June 2024, a total of 141 samples were collected from 62 reef sites across the three principal archipelagos of the South China Sea (Xisha Islands, 12 sites; Zhongsha Islands, 7 sites; and Nansha Islands, 43 sites; Fig. 1 ). These sites encompassed coral reef systems representing a range of geomorphological types, including submerged reefs. Coral habitat condition was evaluated using two complementary metrics: (i) live scleractinian coral cover, and (ii) the incidence of visible bleaching. Coral cover was quantified from two 50 m photo-transects per site (~ 0.4 m above the reef), while bleaching incidence was determined from diver observations along the transects. Reefs were then categorized into three condition classes: Excellent (> 30% live coral cover and no bleaching), Moderate (5–30% live coral cover and/or localized bleaching), and Poor (< 5% live coral cover or extensive bleaching regardless of cover). These thresholds were defined based on expert recommendations for the long-term ecological status of coral reefs in the South China Sea. Environmental parameters, including seawater temperature, salinity, dissolved oxygen, and pH, were measured at discrete depth layers using a pump-CTD system. In addition, 1 L water samples were collected from each site to quantify NO₃⁻ and NO₂⁻ concentrations using an AutoAnalyzer 3 (AA3). At each site, water samples for eDNA analysis were collected following depth-specific strata. At the single shallow site (< 10 m depth), two 5 L seawater samples were collected from 3–5 m depth. At sites deeper than 10 m, two 5 L samples were collected from each of two depth layers: subsurface (3–5 m depth) and near-bottom (~ 2 m above the seabed), resulting in 20 L per site. Samples were vacuum-filtered through sterile 0.45 µm polyethersulfone (PES) membranes within 10 min of collection. Filters were fixed in absolute ethanol and stored at -20°C until DNA extraction. 2.2 Environmental DNA extraction and sequencing DNA extraction was performed using a modified CTAB protocol based on the method of (Porebski et al. 1997 ). For PCR amplification, the MiFish-U primer was employed (Forward: GTCGGTAAAACTCGTGCCAGC; Reverse: CATAGTGGGGTATCTAATCCCAGTTTG), targeting an approximately 170 bp fragment of the 12S rRNA gene specific to fish (Miya et al. 2015 ). The PCR thermal cycling conditions included an initial denaturation at 98°C for 3 min, followed by 34 cycles of denaturation at 98°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 45 s, with a final extension step at 72°C for 5 min. Library preparation and high-throughput sequencing were conducted using the Illumina NovaSeq 6000 platform (PE250) by Shanghai Personal Biotechnology Co., Ltd. To monitor contamination, multiple negative controls were included: (i) a field blank consisting of sterile distilled water exposed during sampling and subsequently filtered, (ii) a filtration blank (sterile water passed through PES filters), and (iii) PCR negative controls. All blanks were carried through DNA extraction and amplification, and none yielded detectable amplicons. 2.3 Bioinformatics analysis Illumina sequencing reads were initially processed using Cutadapt (v2.3) to identify and remove primer sequences and low-quality reads. Subsequent processing of the 12S rRNA dataset and generation of Molecular Operational Taxonomic Units (MOTUs) was performed using the VSEARCH pipeline. Paired-end reads were first merged with the fastq_mergepairs function, followed by quality filtering with a maximum expected error threshold of 0.5. Sequences were then pre-clustered at 98% similarity, and chimeric sequences were subsequently removed. The remaining unique, non-chimeric sequences were clustered at 97% similarity to define MOTUs. MOTUs with fewer than one read or a total relative abundance of less than 0.1% across all samples were excluded to avoid the sequencing errors. To improve annotation accuracy, a regional reference database compatible with the MiFish primer was constructed using the rCRUX R package, which queried the NCBI nucleotide database to retrieve relevant sequence records (Curd et al. 2024 ). This database was further curated by cross-referencing with historical records of fish species reported in the South China Sea, sourced from published literatures (Fang and Lyu 2019 ; Qiu 2021 ) and Fishbase ( https://www.fishbase.se , accessed: 26/02/2024), to exclude taxa unlikely to occur in the region. Taxonomic annotation of MOTUs was performed using MEGABLAST in Geneious Prime. MOTU sequences with ≥ 97% identity and ≥ 90% query coverage were assigned to the species level, while those with ≥ 95% identity were annotated at the genus level. Sequences with identity ≥ 85% were classified at higher taxonomic ranks (Lamy et al. 2021 ). All species names and taxonomic assignments were subsequently verified and standardized using the World Register of Marine Species (WoRMS). 2.4 Fish community comparison Alpha diversity of all samples was quantified using Hill numbers, specifically the number of observed MOTUs and Shannon diversity (q = 1). Species richness was reflected by the number of observed MOTUs. Additionally, Faith's Phylogenetic Diversity (q = 0) derived from MOTU sequences was calculated within the Hill number framework utilizing the hilldiv package (Alberdi and Gilbert 2019 ). Statistical comparisons of α-diversity indices across coral health conditions were performed using the Kruskal-Wallis test. Pairwise comparisons among groups were subsequently conducted using the Wilcoxon test with p-values adjusted for multiple comparisons via the Benjamini-Hochberg method. All visualizations were generated using the ggplot2 package (Wickham 2016 ). Beta diversity was assessed to evaluate differences in fish community composition among sampling sites and coral habitat condition. Prior to analysis, the data were rarefied to the library with the lowest read count to minimize biases arising from unequal sequencing depths across libraries, followed by normalization and Hellinger transformation. Taxonomic β-diversity was quantified using Jaccard and Bray-Curtis dissimilarity indices, while weighted UniFrac distance was employed to assess phylogenetic β-diversity. To test for significant differences in community composition across coral health condition categories, PERMANOVA was performed using the adonis function from the vegan package (Oksanen J et al. 2025 ). 2.5 Community network comparison To evaluate the influence of coral habitat condition on the stability of associated fish communities, molecular ecological networks were constructed using the ggClusterNet package (Wen et al. 2022 ). Correlation matrices were generated based on Spearman’s rank correlation coefficients (|r| ≥ 0.8, p < 0.05), and topological properties (e.g., number of nodes, edges, connectivity, clustering coefficient, and centralization degree) were calculated. Network complexity was quantified as the normalized average of these properties, following established molecular ecological network frameworks (Zhou et al. 2010 ; Zhang et al. 2024 ). Network stability was quantified using robustness and invulnerability, which provides a comprehensive assessment of network stability under both random and targeted perturbation scenarios. Network robustness was assessed through the Robustness.Random.removal function, which evaluates network integrity following the random removal of 50% nodes. Network invulnerability was evaluated through the natural connectivity decay curve generated by sequentially removing nodes in descending order of their combined betweenness and degree centrality scores, with slower decay rates indicating higher invulnerability. 2.5 Analysis of the driving factors Structural Equation Modeling (SEM) was employed to examine the mediating role of coral reef health in the relationships between human activities, climate change, and fish communities. Two primary hypotheses were formulated based on existing literature: (i) environmental factors (temperature, pH, DO, NO 3 and NO 2 ), human activities (fishing, construction), and thermal stress were hypothesized to directly influence fish community diversity and network stability; and (ii) these factors were posited to exert indirect effects on fish communities through their impact on coral. Given the relatively small sample size and the need for mediation analysis, segmented structural equation models were constructed using the piecewiseSEM package (Lefcheck 2016 ). Accumulated thermal stress data were obtained from the NOAA Coral Reef Watch (CRW) version 3.1 daily global 5km satellite coral bleaching heat stress database ( https://coralreefwatch.noaa.gov/satellite/index.php ), with the cumulative maximum annual Degree Heating Weeks (DHW) from 2018 to 2022 being utilized as an indicator of long-term thermal stress. Fishing intensity data were derived from the Global Fishing Watch dataset ( https://globalfishingwatch.org/ ), where the sum of apparent fishing effort from 2018 to 2022 was calculated. Construction data were collected through field surveys and subsequently verified using Google Earth imagery. Ecological network stability was quantified through network robustness under different coral habitat condition. 3 Results Among the 141 samples analyzed, 32 were classified as having excellent coral habitat condition, 32 as moderate, and 77 as poor. The highest level of coral reef degradation was observed in the Xisha Islands, where 95.24% of samples were classified as poor. This was followed by the Zhongsha Islands, with 68.42% of sites exhibiting degraded conditions, and the Nansha Islands, where 43.56% of stations were in poor condition. From these samples, a total of 8,116 molecular operational taxonomic units (MOTUs) were recovered, with an average of 122,629 sequencing reads per sample. These MOTUs correspond to 367 fish species, spanning 33 orders and 73 families. 3.1 Fish community difference among coral habitat conditions The total number of MOTUs exhibited an inverse relationship with habitat condition, declining progressively from Poor to Excellent reefs. A total of 1,600 MOTUs were shared across all three habitat condition categories, representing 19.71% of the total MOTUs detected. The greatest overlap in MOTU composition was observed between the Moderate and Poor categories, while the Excellent and Poor categories shared the fewest MOTUs. This pattern suggests increasing community dissimilarity with greater differences in habitat condition, and indicates substantial compositional shifts occurring at more advanced stages of reef degradation. Across all habitat condition categories, the most dominant fish orders were Perciformes, Eupercaria incertae sedis , and Acanthuriformes. Comparative analysis of α-diversity indices across coral habitat condition categories revealed distinct patterns in biodiversity distribution (Table S1 ). The number of observed MOTUs differed significantly between reefs classified as Moderate and Poor (p < 0.05), but no significant difference was observed between Excellent and Moderate reefs (Fig. 2 a). In contrast, neither Shannon diversity nor Faith’s phylogenetic diversity (PD) index showed significant variation across habitat condition categories (p > 0.05), indicating that these metrics may not adequately capture biodiversity changes associated with habitat degradation (Fig. 2 b-c). This suggests that richness is more sensitive than diversity or phylogenetic metrics to coral degradation, likely due to the persistence of phylogenetically redundant taxa in early stages of reef decline. PERMANOVA analyses revealed that coral habitat condition significantly influenced fish community composition, with consistent effects across Jaccard, Bray-Curtis, and weighted UniFrac distances (p < 0.05; Table 1 ). Although α-diversity metrics showed significant changes only at later stages of habitat decline, β-diversity patterns revealed that compositional shifts in fish communities began earlier. Specifically, pairwise comparisons showed significant differences in taxonomic β-diversity between Excellent reefs and those classified as Moderate or Poor (p < 0.05; Table S2), suggesting that community restructuring precedes measurable losses in richness. Together, these findings highlight the ecological sensitivity of fish assemblages to coral habitat degradation and emphasize the importance of maintaining reef integrity to preserve biodiversity. Table 1 PERMANOVA results based on Jaccard, Bray-Curtis, and Weighted Unifrac distances Df SS MS F R2 p Jaccard Conditions 2 1.155 0.578 1.212 0.017 0.001 Residuals 138 65.769 0.477 0.983 Total 140 66.924 1 Bray-Curtis Conditions 2 1.265 0.632 1.388 0.020 0.001 Residuals 138 62.862 0.456 0.980 Total 140 64.126 1 Weighted-Unifrac Conditions 2 0.030 0.015 1.287 0.018 0.141 Residuals 138 1.633 0.012 0.982 Total 140 1.663 1 3.2 Network structural changes among coral habitat conditions Analysis of multidimensional topological metrics revealed a pronounced nonlinear response of fish community network structure to coral habitat condition (Fig. 4 ). Reefs in Moderate condition exhibited the highest overall network complexity (Table S3), characterized by increased node connectivity and a more compact, integrated network architecture. However, as coral condition deteriorated further, network fragmentation became increasingly pronounced. In Moderate-condition reefs, only 13.53% of nodes were isolated, whereas in Poor-condition reefs, this proportion nearly doubled to 25%, indicating a substantial breakdown of species interactions. Additional network properties corroborated this threshold-like transition: Moderate-condition networks had the highest edge connectivity (34.45) and maintained average path lengths (2.29) comparable to those in Excellent reefs, reflecting efficient information flow and robust taxonomic associations. By contrast, Poor-condition networks exhibited significantly longer average path lengths (2.46) and reduced clustering coefficients, key indicators of weakened community cohesion and impaired ecological connectivity. These patterns indicate that Moderate-condition reefs maintain higher network connectivity compared to both Excellent and Poor reefs, highlighting a nonlinear response to habitat degradation. 3.3 Role of module hubs and network robustness in fish communities In fish community networks associated with coral reefs in Excellent and Moderate condition, module hubs, nodes with disproportionately high within-module connectivity, were consistently identified, whereas no such hubs were detected in Poor-condition reefs (Fig. 4 ). These hubs indicate species that play central roles in maintaining intra-module cohesion and stabilizing interaction structures. The taxonomically annotated module hub MOTUs ranked within the top 30% in relative abundance for their respective reef conditions (Table S4), suggesting a strong correlation between numerical dominance and network centrality. These hub MOTUs were primarily assigned to 25 genera, including Epinephelus, Carcharhinus, and Cirrhitichthys. Among them, 14 MOTUs were identified as carnivores and 2 as herbivores, with trophic levels exceeding 3.8 for the carnivorous taxa, indicating their role as high-level predators. In contrast, their decline or disappearance in Poor-condition reefs likely disrupts trophic structuring and erodes the regulatory capacity of the community. Most module hubs were carnivorous taxa suggesting that higher trophic-level species play a key role in maintaining network integrity in healthier reefs. Network robustness analysis further revealed significantly higher structural stability in fish communities associated with healthy reefs compared to degraded ones (p < 0.05; Fig. 5 a). In particular, Poor-condition reefs exhibited the lowest robustness values across all categories (p < 0.05). Complementary natural connectivity analysis (Fig. 5 b) showed that Excellent-condition networks maintained the highest overall resilience under simulated node removal, although Moderate-condition reefs exhibited superior resistance to intermediate disturbance levels (20–40% node loss). In contrast, Poor-condition networks suffered rapid breakdown after just 20% of nodes were removed, despite initially appearing relatively connected. 3.4 Stressors related with richness and network stability Structural equation modeling (SEM) revealed distinct pathways through which anthropogenic and environmental stressors influence coral reef fish communities (Fig. 6 a). Among these, fishing pressure exerted the most pronounced direct negative effect on fish species richness (β = -0.25, p = 0.022), while ocean acidification, indicated by pH, showed a marginally significant direct negative impact (β = -0.24, p = 0.053). Other environmental variables, including accumulated thermal stress and coastal construction, did not demonstrate significant direct effects on species richness. In contrast, coral habitat condition emerged as a critical positive driver of fish community network stability (β = 0.96, p < 0.001), underscoring the foundational role of reef health in maintaining ecological connectivity. Coastal construction also exhibited a modest but significant positive influence on network stability (β = 0.09, p = 0.003), potentially reflecting localized management or habitat modifications. Other factors, thermal stress, fishing pressure, and environmental variables aside from temperature, showed negligible direct effects on network stability. Notably, fishing pressure significantly undermined coral habitat condition (β = -0.21, p = 0.048), with accumulated thermal stress also approaching a marginally significant negative impact (β = -0.21, p = 0.054). Coral habitat condition acted as a key mediator, channeling indirect effects from fishing pressure to fish biodiversity and community structure. For species richness, the indirect effect of fishing pressure via coral habitat condition was similar in magnitude to its direct effect. More strikingly, for network stability, the indirect influence of fishing pressure mediated through coral habitat condition surpassed both its direct and total effects, highlighting the central role of reef condition in buffering or amplifying anthropogenic impacts. Together, these findings indicate that fishing pressure primarily undermines reef fish networks indirectly by degrading coral habitat, reinforcing the central role of reef condition as a mediator of anthropogenic stress. 4 Discussion 4.1 Harnessing eDNA for Coral Reef Fish Monitoring: Opportunities and Limitations for Management Coral reefs on remote oceanic islands are ecologically isolated and highly sensitive to climate-induced degradation (Strona et al. 2021 ). Monitoring reef biodiversity in remote islands remains challenging due to logistical constraints, highlighting the need for alternative survey methods (Samoilys et al. 2022 ). Traditional underwater visual census (UVC) methods face limitations in these habitats due to strong currents, turbidity, and logistical challenges (Polanco Fernández et al. 2021 ). As a diver-independent alternative, environmental DNA (eDNA) metabarcoding has gained traction. Despite its strengths, including cost-efficiency, taxonomic breadth, and the ability to detect rare or elusive species, the approach is constrained by incomplete reference databases, particularly for cryptobenthic fishes that dominate coral reef ecosystems (Brandl et al. 2018 ; Gómez-Buckley et al. 2023 ). In our study, only 45.6% of the 2,669 regionally documented species had matching barcodes in public repositories, underscoring this limitation. Nonetheless, eDNA metabarcoding detected 265 species from 21 water samples across seven reef sites, surpassing prior UVC records in the region (Shuting et al. 2022 ). Taxonomic coverage was extensive, spanning 33 orders, and included ecologically and functionally important coral reef taxa such as Acanthuridae and Serranidae, aligning with previous research ((Zhao et al. 2023 , 2024 ). Although some cryptobenthic taxa remained unidentified, MOTUs provided reliable biodiversity proxies (Juhel et al. 2020 ), allowing their ecological roles to be integrated into community-level and network-based analyses (Mathon et al. 2022 ; Malik et al. 2025 ). From a conservation and management perspective, these findings emphasize that eDNA is not simply a substitute for traditional approaches but a complementary and often more practical tool for large-scale biodiversity assessment. Especially in remote, politically sensitive, or physically challenging regions, eDNA provides a scalable method to track biodiversity dynamics where conventional surveys are infeasible (Bohmann et al. 2014 ; Stat et al. 2017 ). The integration of eDNA with UVC or baited remote underwater video (BRUV) surveys could enhance both the breadth and depth of monitoring, creating hybrid approaches that are directly applicable to management programs (Evans et al. 2017 ; Muenzel et al. 2024 ). Despite its limitations, eDNA provides a practical approach for monitoring reef fish assemblages, particularly where early detection of biodiversity change is needed (Rees et al. 2014 ). 4.2 Nonlinear and Stage Responses of Fish Networks to Coral Habitat Loss: Early-Warning Signals for Management Our findings reveal a nonlinear response of coral reef fish diversity and ecological stability to habitat degradation, aligning with prior studies that suggest lagged responses of ichthyofaunal assemblages to Scleractinian coral loss (Wilson et al. 2006 ; Adam et al. 2014 ). This pattern may be driven by two key mechanisms. First, the three-dimensional structural complexity of live coral provides a buffering effect in the early stages of degradation. Even during initial bleaching events, reefs may retain sufficient architectural integrity to support diverse microhabitats, modulate predator-prey interactions, and sustain interspecific competition (Darling et al. 2017 ). Transitional states, such as the coexistence of bleached coral and algal cover, can temporarily support herbivorous and omnivorous fishes (Koester et al. 2023 ). This is consistent with our ZIPI network analysis, which identified herbivorous species as module hub nodes in moderately degraded reef systems. Second, fish recruitment dynamics exhibit pronounced time-lag effects. Species that depend heavily on live coral during juvenile stages may persist into adulthood even after coral decline, delaying the onset of biodiversity loss (Wilson et al. 2008 ). This may contribute to the temporary resilience of fish communities during the early stages of coral degradation observed in our study. Indeed, our results indicate that coral reef fish communities exhibit ecological network robustness during moderate degradation, supporting trophic models that predict sustained fish productivity under initial stress (Rogers et al. 2018a , 2018b ). The temporary increase in prey availability may even benefit large-bodied predators (Graham et al. 2007 ), which aligns with our finding that key MOTUs in moderately degraded reefs are primarily high-trophic-level species. However, this resilience is fragile. Our analysis revealed that moderate degradation coincides with reduced energy flow velocity and lower species turnover (Morais et al. 2020 ), suggesting that apparent stability masks underlying vulnerability. Once coral condition declines past a certain “refuge threshold,” resilience mechanisms break down (Morais et al. 2020 ). Severely degraded reefs exhibited markedly lower ecological network robustness, accompanied by the loss of keystone species and functional redundancy (Kerry and Bellwood 2012 ). The disappearance of critical MOTUs likely reflects compounding stressors, including habitat collapse, delayed predator recruitment failure (Pratchett et al. 2014 ), and intensified overfishing pressure (Sherman et al. 2023 ; Zhao et al. 2024 ). Interestingly, our robustness analysis indicates that some degree of natural connectivity persists even after critical MOTUs are lost, suggesting that reef fish communities may transition into new, alternative stable states (Fung et al. 2011 ; Luza et al. 2022 ). However, these alternative states are more vulnerable to disturbance and support fewer ecological functions than those on healthier reefs. This insight has direct management implications: monitoring programs should move beyond coral cover alone and incorporate network-based indicators such as redundancy, hub persistence, or trophic flow velocity as early-warning signals of ecosystem collapse. These results suggest that early changes in network structure should be monitored to detect declines before major losses in community function occur (Bellwood et al. 2012 ). 4.3 Linking Anthropogenic and Climatic Pressures to Reef Fish Decline: Implications for Multilateral Governance Our study highlights that coral reef condition is the most influential factor affecting the stability of reef fish ecological networks. However, this condition is itself shaped by a suite of anthropogenic and climatic stressors that exert both direct and indirect impacts. Among these, escalating fishing pressure in the South China Sea (SCS) has emerged as a dominant driver. Despite China’s seasonal fishing moratoriums covering over 40% of regional yields(Wu et al. 2024 ), the absence of coordinated governance among the ten littoral states (Yu and Chang 2023 ; Lai and Yu 2025 ) has permitted unsustainable exploitation of remote reefs (Li et al. 2023 ; Hong et al. 2024 ). Our findings align with previous reports documenting trophic downgrading and biodiversity collapse in the Xisha (Zhao et al. 2023 ; Qiu et al. 2024 )and Nansha Islands (Dai et al. 2022 ; Kang et al. 2024 ), where non-selective fishing methods such as trawling and gillnetting (Guan et al. 2021 )disproportionately depleted carnivorous key MOTUs at degraded sites. Beyond biomass removal, these extractive practices exacerbate habitat degradation. Anchoring, bottom trawling, and destructive methods (e.g., blast or cyanide fishing) inflict physical damage on coral structures, reducing reef complexity and fragmenting habitat mosaics (Pet-Soede et al. 1999 ; Arai 2015 ; Williams et al. 2020 ; Borland et al. 2021 ). This fragmentation is evident in coral cover gradients, from 35% in relatively intact Nansha systems(Tkachenko and Hoang 2022 ), reflecting shifts in fishing effort (Wu et al. 2024 ). Such fragmentation interacts synergistically with thermal stress, as cumulative warming accelerates bleaching events and reef collapse (Gordon et al. 2018 ; Asbury et al. 2024 ). The resultant decline in structural complexity diminishes fish recruitment, dispersal, and connectivity, reinforcing negative feedback loops that erode resilience (Kubicek and Reuter 2016 ). These findings underscore that reef condition is not an isolated ecological variable, but one embedded within a complex network of interacting anthropogenic and climatic drivers. Effective management will require coordinated actions, such as regulating fishing practices, protecting structurally complex reefs, and improving cooperation among neighboring countries. Island construction and restoration projects further complicate this picture. While often assumed to cause permanent coral loss (Tkachenko and Hoang 2022 ), our model detected no significant negative effects on fish diversity (p > 0.05) and, unexpectedly, a positive effect on ecological network stability (p < 0.05), likely reflecting post-construction recovery periods that allowed partial ecological reorganization. Nevertheless, these results reinforce a critical conservation principle: the long-term resilience of reef fish communities hinges on preserving coral habitat integrity. Without coordinated management that addresses both local pressures and warming events, reefs that appear stable today may still shift toward degraded states with reduced ecological function. 5 Conclusion and Management Implications Our findings demonstrate that reef fish assemblages do not decline smoothly with coral degradation. Instead, modest early losses of live coral can trigger short-lived increases in biodiversity and connectivity, but once a critical threshold is surpassed, species richness, network stability, and key ecological functions collapse abruptly. By coupling high-resolution eDNA metabarcoding with ecological network and structural-equation analyses, we reveal that top-down control by apex predators and other stabilizing interactions erode progressively along degradation gradients. Fishing pressure and cumulative thermal stress emerged as major drivers, acting directly by removing fish biomass and bleaching corals, and indirectly by simplifying habitat architecture and food-web structure. While fish communities may temporarily reorganize into stable but functionally compromised states, such resilience is fragile and potentially irreversible if stressors persist. Effective conservation therefore requires proactive, multilateral strategies that curb local anthropogenic impacts, mitigate global climate stress, and preserve structural complexity before reefs are pushed beyond ecological tipping points. Our results highlight three key applications: Incorporating network metrics into monitoring frameworks as early-warning indicators of collapse. Integrating eDNA into regional biodiversity monitoring programs, particularly for remote and politically sensitive reefs. Designing multi-scalar governance frameworks, combining local enforcement with transboundary MPAs, to address both anthropogenic and climatic stressors. Such coordinated action is especially urgent in the South China Sea and other geopolitically complex regions where management jurisdictions overlap and climate extremes are intensifying. By linking eDNA-based biodiversity data with network-level ecological insights, our study provides a mechanistically grounded framework for anticipating coral reef collapse and guiding interventions that sustain resilience in the Anthropocene. Declarations Conflict of Interest Statement The authors declare that there are no financial or non-financial conflicts of interest that could have influenced the work reported in this paper. Author Contribution Cong Zeng: Conceptualization, Methodology, Supervision, Writing - review & editing. Zheying Lin: Investigation, Formal analysis, Writing - original draft. Zhiyi Su, Yuanbin Zhao: Writing - review & editing. Yue Liu: Data curation, Formal analysis. Mingjie Li, Chengxuan Zou, Yue Zheng, Hongqiang Yang: Field investigation. Qiang Lin: Funding acquisition, Field investigation. Ling Cao: Conceptualization, Supervision, Writing - review & editing, Funding acquisition. Acknowledgement We gratefully acknowledge the financial support provided by the Ministry of Science and Technology of China (2022-24), the National Natural Science Foundation of China (42425603) and National Natural Science Foundation of China (42206082). The views, findings, and conclusions presented within this material are solely those of the authors and do not necessarily represent the perspectives of the funding organizations. Data Availability All environmental DNA (eDNA) amplicon sequencing data generated in this study have been deposited in the NCBI BioSample database under accession numbers SAMN50205399–SAMN50205543 as part of submission SUB15490360. References Adam TC, Brooks AJ, Holbrook SJ, Schmitt RJ, Washburn L, Bernardi G (2014) How will coral reef fish communities respond to climate-driven disturbances? Insight from landscape-scale perturbations. Oecologia 176:285–296 Alberdi A, Gilbert MTP (2019) hilldiv: an R package for the integral analysis of diversity based on Hill numbers. 545665 Alvarez-Filip L, Gill JA, Dulvy NK (2011) Complex reef architecture supports more small-bodied fishes and longer food chains on Caribbean reefs. Ecosphere 2:art118 Arai T (2015) Diversity and conservation of coral reef fishes in the Malaysian South China Sea. Rev Fish Biol Fisheries 25:85–101 Asbury M, Innes-Gold AA, Wulstein DM, Madin EMP, Madin JS, McManus LC (2024) Recovery potential of fish and coral populations following ecological disturbance. Ecosphere 15:e4915 Bellwood DR, Baird AH, Depczynski M, González-Cabello A, Hoey AS, Lefèvre CD, Tanner JK (2012) Coral recovery may not herald the return of fishes on damaged coral reefs. Oecologia 170:567–573 Bellwood DR, Hughes TP, Folke C, Nyström M (2004) Confronting the coral reef crisis. Nature 429:827–833 Bohmann K, Evans A, Gilbert MTP, Carvalho GR, Creer S, Knapp M, Yu DW, de Bruyn M (2014) Environmental DNA for wildlife biology and biodiversity monitoring. Trends in Ecology & Evolution 29:358–367 Borland HP, Gilby BL, Henderson CJ, Leon JX, Schlacher TA, Connolly RM, Pittman SJ, Sheaves M, Olds AD (2021) The influence of seafloor terrain on fish and fisheries: A global synthesis. Fish and Fisheries 22:707–734 Brandl SJ, Goatley CHR, Bellwood DR, Tornabene L (2018) The hidden half: ecology and evolution of cryptobenthic fishes on coral reefs. Biological Reviews 93:1846–1873 Cardinale BJ, Duffy JE, Gonzalez A, Hooper DU, Perrings C, Venail P, Narwani A, Mace GM, Tilman D, Wardle DA, Kinzig AP, Daily GC, Loreau M, Grace JB, Larigauderie A, Srivastava DS, Naeem S (2012) Biodiversity loss and its impact on humanity. Nature 486:59–67 Chen G, Li Y, Chen X (2007) Species diversity of fishes in the coral reefs of South China Sea. Biodiversity Science 15:373 Clements KD, Choat JH (2018) Nutritional Ecology of Parrotfishes (Scarinae, Labridae). Biology of Parrotfishes. CRC Press, Curd EE, Gal L, Gallego R, Silliman K, Nielsen S, Gold Z (2024) rCRUX: A rapid and versatile tool for generating metabarcoding reference libraries in R. Environmental DNA 6:e489 Dai X, Li Y, Cai Y, Gong Y, Zhang J, Chen Z (2022) Variations in Fish Community Structure at the Lagoon of Yongshu Reef, South China Sea between 1999 and 2016–2019. Diversity 14:763 Darling ES, Graham NAJ, Januchowski-Hartley FA, Nash KL, Pratchett MS, Wilson SK (2017) Relationships between structural complexity, coral traits, and reef fish assemblages. Coral Reefs 36:561–575 Deng Y, Jiang Y-H, Yang Y, He Z, Luo F, Zhou J (2012) Molecular ecological network analyses. BMC Bioinformatics 13:113 Dunne JA, Williams RJ, Martinez ND (2004) Network structure and robustness of marine food webs. Marine Ecology Progress Series 273:291–302 Emslie MJ, Bray P, Cheal AJ, Johns KA, Osborne K, Sinclair-Taylor T, Thompson CA (2020) Decades of monitoring have informed the stewardship and ecological understanding of Australia’s Great Barrier Reef. Biological Conservation 252:108854 Evans NT, Shirey PD, Wieringa JG, Mahon AR, Lamberti GA (2017) Comparative Cost and Effort of Fish Distribution Detection via Environmental DNA Analysis and Electrofishing. Fisheries 42:90–99 Fang H, Lyu X (2019) Reef Fish Identification of Nansha Islands. China Ocean University Press. https://www.hceis.com/home/book_view.aspx?id=11498 Fisher WS (2023) Relating fish populations to coral colony size and complexity. Ecological Indicators 148:110117 Fontoura L, Zawada KJA, D’agata S, Álvarez-Noriega M, Baird AH, Boutros N, Dornelas M, Luiz OJ, Madin JS, Maina JM, Pizarro O, Torres-Pulliza D, Woods RM, Madin EMP (2020) Climate-driven shift in coral morphological structure predicts decline of juvenile reef fishes. Global Change Biology 26:557–567 Fung T, Seymour RM, Johnson CR (2011) Alternative stable states and phase shifts in coral reefs under anthropogenic stress. Ecology 92:967–982 Gómez-Buckley MC, Gallego R, Arranz V, Halafihi T, Stone K, Erdmann M, Tornabene LM (2023) Comparing anesthetic stations and environmental DNA sampling to determine community composition of cryptobenthic coral reef fishes of Vava’u, Kingdom of Tonga. Coral Reefs 42:785–797 Gordon TAC, Harding HR, Wong KE, Merchant ND, Meekan MG, McCormick MI, Radford AN, Simpson SD (2018) Habitat degradation negatively affects auditory settlement behavior of coral reef fishes. Proceedings of the National Academy of Sciences 115:5193–5198 Gove JM, Williams GJ, Lecky J, Brown E, Conklin E, Counsell C, Davis G, Donovan MK, Falinski K, Kramer L, Kozar K, Li N, Maynard JA, McCutcheon A, McKenna SA, Neilson BJ, Safaie A, Teague C, Whittier R, Asner GP (2023) Coral reefs benefit from reduced land–sea impacts under ocean warming. Nature 621:536–542 Graham N a. J, Wilson SK, Jennings S, Polunin NVC, Robinson J, Bijoux JP, Daw TM (2007) Lag Effects in the Impacts of Mass Coral Bleaching on Coral Reef Fish, Fisheries, and Ecosystems. Conservation Biology 21:1291–1300 Graham NAJ, McClanahan TR, MacNeil MA, Wilson SK, Cinner JE, Huchery C, Holmes TH (2017) Human Disruption of Coral Reef Trophic Structure. Current Biology 27:231–236 Graham NAJ, Wilson SK, Jennings S, Polunin NVC, Bijoux JP, Robinson J (2006) Dynamic fragility of oceanic coral reef ecosystems. Proceedings of the National Academy of Sciences 103:8425–8429 Guan Y, Zhang J, Zhang X, Li Z, Meng J, Liu G, Bao M, Cao C (2021) Identification of Fishing Vessel Types and Analysis of Seasonal Activities in the Northern South China Sea Based on AIS Data: A Case Study of 2018. Remote Sensing 13:1952 Hill TS, Hoogenboom MO (2022) The indirect effects of ocean acidification on corals and coral communities. Coral Reefs 41:1557–1583 Hong X, Zhang K, Li J, Xu Y, Sun M, Xu S, Cai Y, Qiu Y, Chen Z (2024) Stock Assessment of the Commercial Small Pelagic Fishes in the Beibu Gulf, the South China Sea, 2006–2020. Biology 13:226 Huang D, Licuanan WY, Hoeksema BW, Chen CA, Ang PO, Huang H, Lane DJW, Vo ST, Waheed Z, Affendi YA, Yeemin T, Chou LM (2015) Extraordinary diversity of reef corals in the South China Sea. Mar Biodiv 45:157–168 Hughes TP, Barnes ML, Bellwood DR, Cinner JE, Cumming GS, Jackson JBC, Kleypas J, van de Leemput IA, Lough JM, Morrison TH, Palumbi SR, van Nes EH, Scheffer M (2017) Coral reefs in the Anthropocene. Nature 546:82–90 IPCC (2022) Climate Change 2022: Impacts, Adaptation, and Vulnerability. Contribution of Working Group II to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change.Cambridge University Press. Juhel J-B, Utama RS, Marques V, Vimono IB, Sugeha HY, Kadarusman, Pouyaud L, Dejean T, Mouillot D, Hocdé R (2020) Accumulation curves of environmental DNA sequences predict coastal fish diversity in the coral triangle. Proc Biol Sci 287:20200248 Kang Z, Wang T, Li C, Zhao J, Shi J, Xie H, Liu Y (2024) Species composition and succession of coral reef fishes in Huaguang Reef, Xisha Islands. Water Biology and Security 3:100273 Kerry JT, Bellwood DR (2012) The effect of coral morphology on shelter selection by coral reef fishes. Coral Reefs 31:415–424 Kerry JT, Bellwood DR (2015) Do tabular corals constitute keystone structures for fishes on coral reefs? Coral Reefs 34:41–50 Koester A, Gordó–Vilaseca C, Bunbury N, Ferse SCA, Ford A, Haupt P, A’Bear L, Bielsa M, Burt AJ, Letori J, Mederic E, Nancy E, Sanchez C, Waller M, Wild C (2023) Impacts of coral bleaching on reef fish abundance, biomass and assemblage structure at remote Aldabra Atoll, Seychelles: insights from two survey methods. Front Mar Sci 10:1230717 Kubicek A, Reuter H (2016) Mechanics of multiple feedbacks in benthic coral reef communities. Ecological Modelling 329:29–40 Lai M, Yu W (2025) Analysis of fishery ecosystem structure in the South China Sea. JFSC 31:1336–1350 Lamy T, Pitz KJ, Chavez FP, Yorke CE, Miller RJ (2021) Environmental DNA reveals the fine-grained and hierarchical spatial structure of kelp forest fish communities. Sci Rep 11:14439 Lefcheck JS (2016) piecewiseSEM: Piecewise structural equation modelling in r for ecology, evolution, and systematics. Methods in Ecology and Evolution 7:573–579 Li M, Xu Y, Sun M, Li J, Zhou X, Chen Z, Zhang K (2023) Impacts of Strong ENSO Events on Fish Communities in an Overexploited Ecosystem in the South China Sea. Biology 12:946 Luza AL, Bender MG, Ferreira CEL, Floeter SR, Francini-Filho RB, Longo GO, Pinheiro HT, Quimbayo JP, Bastazini VAG (2024) Coping with collapse: Functional robustness of coral-reef fish network to simulated cascade extinction. Global Change Biology 30:e17513 Luza AL, Quimbayo JP, Ferreira CEL, Floeter SR, Francini-Filho RB, Bender MG, Longo GO (2022) Low functional vulnerability of fish assemblages to coral loss in Southwestern Atlantic marginal reefs. Sci Rep 12:17164 Malik MDA, Ambariyanto A, Hartati R, Nursalim N, Kholilah N, Kurniasih EM, Anggoro AW, Prasetia R, Syamsyuni Y, Muh F, Cahyani NKD (2025) eDNA uncovers hidden fish diversity in the coral reef ecosystems of Karimunjawa National Park, Indonesia. Regional Studies in Marine Science 81:103945 Mathon L, Marques V, Mouillot D, Albouy C, Andrello M, Baletaud F, Borrero-Pérez GH, Dejean T, Edgar GJ, Grondin J, Guerin P-E, Hocdé R, Juhel J-B, Kadarusman, Maire E, Mariani G, McLean M, Polanco F. A, Pouyaud L, Stuart-Smith RD, Sugeha HY, Valentini A, Vigliola L, Vimono IB, Pellissier L, Manel S (2022) Cross-ocean patterns and processes in fish biodiversity on coral reefs through the lens of eDNA metabarcoding. Proc Biol Sci 289:20220162 McCann KS (2000) The diversity–stability debate. Nature 405:228–233 McWilliam M, Pratchett MS, Hoogenboom MO, Hughes TP (2020) Deficits in functional trait diversity following recovery on coral reefs. Proceedings: Biological Sciences 287:1–9 Meyer JM, Leempoel K, Losapio G, Hadly EA (2020) Molecular Ecological Network Analyses: An Effective Conservation Tool for the Assessment of Biodiversity, Trophic Interactions, and Community Structure. Front Ecol Evol 8:588430 Miya M (2022) Environmental DNA Metabarcoding: A Novel Method for Biodiversity Monitoring of Marine Fish Communities. Annual Review of Marine Science 14:161–185 Miya M, Sato Y, Fukunaga T, Sado T, Poulsen JY, Sato K, Minamoto T, Yamamoto S, Yamanaka H, Araki H, Kondoh M, Iwasaki W (2015) MiFish, a set of universal PCR primers for metabarcoding environmental DNA from fishes: detection of more than 230 subtropical marine species. R Soc Open Sci 2:150088 Montoya JM, Pimm SL, Solé RV (2006) Ecological networks and their fragility. Nature 442:259–264 Mora C, Caldwell IR, Birkeland C, McManus JW (2016) Dredging in the Spratly Islands: Gaining Land but Losing Reefs. PLOS Biology 14:e1002422 Morais RA, Depczynski M, Fulton C, Marnane M, Narvaez P, Huertas V, Brandl SJ, Bellwood DR (2020) Severe coral loss shifts energetic dynamics on a coral reef. Functional Ecology 34:1507–1518 Mouillot D, Villéger S, Parravicini V, Kulbicki M, Arias-González JE, Bender M, Chabanet P, Floeter SR, Friedlander A, Vigliola L, Bellwood DR (2014) Functional over-redundancy and high functional vulnerability in global fish faunas on tropical reefs. Proceedings of the National Academy of Sciences 111:13757–13762 Muenzel D, Bani A, De Brauwer M, Stewart E, Djakiman C, Halwi, Purnama R, Yusuf S, Santoso P, Hukom FD, Struebig M, Jompa J, Limmon G, Dumbrell A, Beger M (2024) Combining environmental DNA and visual surveys can inform conservation planning for coral reefs. Proceedings of the National Academy of Sciences 121:e2307214121 Oksanen J, Simpson G L, Blanchet F G, Kindt R, Legendre P, O’Hara R B, Solymos P, Stevens M H H, Szoecs E, Wagner H, Barbour M, Bedward M, Bolker B, Borcard D, Borman T, Carvalho G, Chirico M, De Caceres M, Durand S, Evangelista H B A, FitzJohn R, Friendly M, Furneaux B, Hannigan G, Hill M O, Lahti L, Martino C, McGlinn D, Ouellette M H, Ribeiro Cunha E, Smith T, Stier A, Ter Braak C J F, Weedon J (2025) vegan: Community Ecology Package. R package version 2.8-0, https://vegandevs.github.io/vegan/ . Olán-González M, Reyes-Bonilla H, Álvarez-Filip L, Pérez-España H, Olivier D (2020) Fish diversity divergence between tropical eastern pacific and tropical western Atlantic coral reefs. Environ Biol Fish 103:1323–1341 Pang J, Ren G, Shi Q, Zhu H, Hu Y, Dong J, Ma Y (2021) Analysis of coral reef bleaching in Yongle Islands of Xisha from 2005 to 2018 based on sediment types change monitoring. MS 45:92–106 Pet-Soede C, Cesar HSJ, Pet JS (1999) An economic analysis of blast fishing on Indonesian coral reefs. Environmental Conservation 26:83–93 Polanco Fernández A, Marques V, Fopp F, Juhel J-B, Borrero-Pérez GH, Cheutin M-C, Dejean T, González Corredor JD, Acosta-Chaparro A, Hocdé R, Eme D, Maire E, Spescha M, Valentini A, Manel S, Mouillot D, Albouy C, Pellissier L (2021) Comparing environmental DNA metabarcoding and underwater visual census to monitor tropical reef fishes. Environmental DNA 3:142–156 Porebski S, Bailey LG, Baum BR (1997) Modification of a CTAB DNA extraction protocol for plants containing high polysaccharide and polyphenol components. Plant Mol Biol Rep 15:8–15 Pozas-Schacre C, Casey JM, Brandl SJ, Kulbicki M, Harmelin-Vivien M, Strona G, Parravicini V (2021) Congruent trophic pathways underpin global coral reef food webs. Proceedings of the National Academy of Sciences 118:e2100966118 Pratchett MS, Berumen ML, Marnane MJ, Eagle JV, Pratchett DJ (2008) Habitat associations of juvenile versus adult butterflyfishes. Coral Reefs 27:541–551 Pratchett MS, Blowes SA, Coker D, Kubacki E, Nowicki J, Hoey AS (2015) Indirect benefits of high coral cover for non-corallivorous butterflyfishes. Coral Reefs 34:665–672 Pratchett MS, Coker DJ, Jones GP, Munday PL (2012) Specialization in habitat use by coral reef damselfishes and their susceptibility to habitat loss. Ecology and Evolution 2:2168–2180 Pratchett MS, Hoey AS, Wilson SK (2014) Reef degradation and the loss of critical ecosystem goods and services provided by coral reef fishes. Current Opinion in Environmental Sustainability 7:37–43 Qiu (2021) Checklist of the coral fish fauna of Xisha Islands, China. Biodiversity Data Journal 9:e63945 Qiu S, Liu X, Chen B, Wang Y, Liao J, Du J (2024) Reef fish diversity and community pattern of the Xisha Islands, South China Sea. MES 41:395–401 Rees HC, Maddison BC, Middleditch DJ, Patmore JRM, Gough KC (2014) REVIEW: The detection of aquatic animal species using environmental DNA – a review of eDNA as a survey tool in ecology. Journal of Applied Ecology 51:1450–1459 Rogers A, Blanchard JL, Mumby PJ (2018a) Fisheries productivity under progressive coral reef degradation. Journal of Applied Ecology 55:1041–1049 Rogers A, Blanchard JL, Newman SP, Dryden CS, Mumby PJ (2018b) High refuge availability on coral reefs increases the vulnerability of reef-associated predators to overexploitation. Ecology 99:450–463 Samoilys M, Alvarez-Filip L, Myers R, Chabanet P (2022) Diversity of Coral Reef Fishes in the Western Indian Ocean: Implications for Conservation. Diversity 14:102 Santoso P, Setiawan F, Subhan B, Arafat D, Bengen DG, Iqbal Sani LM, Humphries AT, Madduppa H (2022) Influence of Coral Reef Rugosity on Fish Communities in Marine Reserves Around Lombok Island, Indonesia. Environ Biol Fish 105:105–117 Sherman CS, Simpfendorfer CA, Pacoureau N, Matsushiba JH, Yan HF, Walls RHL, Rigby CL, VanderWright WJ, Jabado RW, Pollom RA, Carlson JK, Charvet P, Bin Ali A, Fahmi, Cheok J, Derrick DH, Herman KB, Finucci B, Eddy TD, Palomares MLD, Avalos-Castillo CG, Kinattumkara B, Blanco-Parra M-P, Dharmadi, Espinoza M, Fernando D, Haque AB, Mejía-Falla PA, Navia AF, Pérez-Jiménez JC, Utzurrum J, Yuneni RR, Dulvy NK (2023) Half a century of rising extinction risk of coral reef sharks and rays. Nat Commun 14:15 Shi J, Li C, Wang T, Zhao J, Liu Y, Xiao Y (2022) Distribution Pattern of Coral Reef Fishes in China. Sustainability 14:15107 Shuting Q, Xinming L, Bin C, Yanguo W, Jianji L, Jianguo D (2022) Reef fish diversity and community pattern of the Xisha Islands,South China Sea. Haiyang Huanjing Kexue 41:395–401 Stat M, Huggett MJ, Bernasconi R, DiBattista JD, Berry TE, Newman SJ, Harvey ES, Bunce M (2017) Ecosystem biomonitoring with eDNA: metabarcoding across the tree of life in a tropical marine environment. Sci Rep 7:12240 Stier AC, Chase TJ, Osenberg CW (2025) Fish services to corals: a review of how coral-associated fishes benefit corals. Coral Reefs 44:825–834 Strona G, Beck PSA, Cabeza M, Fattorini S, Guilhaumon F, Micheli F, Montano S, Ovaskainen O, Planes S, Veech JA, Parravicini V (2021) Ecological dependencies make remote reef fish communities most vulnerable to coral loss. Nat Commun 12:7282 Tkachenko KS, Hoang DT (2022) Concurrent effect of crown-of-thorns starfish outbreak and thermal anomaly of 2020 on coral reef communities of the Spratly Islands (South China Sea). Marine Ecology 43:e12717 Tylianakis JM, Laliberté E, Nielsen A, Bascompte J (2010) Conservation of species interaction networks. Biological Conservation 143:2270–2279 Wen T, Xie P, Yang S, Niu G, Liu X, Ding Z, Xue C, Liu Y-X, Shen Q, Yuan J (2022) ggClusterNet: An R package for microbiome network analysis and modularity-based multiple network layouts. iMeta 1:e32 Wickham H (2016) ggplot2 Elegant Graphics for Data Analysis. Springer International Publishing, Cham Williams A, Althaus F, Maguire K, Green M, Untiedt C, Alderslade P, Clark MR, Bax N, Schlacher TA (2020) The Fate of Deep-Sea Coral Reefs on Seamounts in a Fishery-Seascape: What Are the Impacts, What Remains, and What Is Protected? Front Mar Sci 7:567002 Wilson SK, Burgess SC, Cheal AJ, Emslie M, Fisher R, Miller I, Polunin NVC, Sweatman HPA (2008) Habitat Utilization by Coral Reef Fish: Implications for Specialists vs. Generalists in a Changing Environment. Journal of Animal Ecology 77:220–228 Wilson SK, Graham N a. J, Pratchett MS, Jones GP, Polunin NVC (2006) Multiple disturbances and the global degradation of coral reefs: are reef fishes at risk or resilient? Global Change Biology 12:2220–2234 Wolfe K, Kenyon TM, Mumby PJ (2021) The biology and ecology of coral rubble and implications for the future of coral reefs. Coral Reefs 40:1769–1806 Wu W, Zhang P, Wang Q, Kang L, Su F (2024) Analysis of fishing intensity in the South China Sea based on automatic identification system data: A comparison between China and Vietnam. Marine and Coastal Fisheries 16:e10309 Young HS, McCauley FO, Micheli F, Dunbar RB, McCauley DJ (2024) Shortened food chain length in a fished versus unfished coral reef. Ecological Applications 34:e3002 Yu C, Chang Y-C (2023) China’s Incentives and Efforts against IUU Fishing in the South China Sea. Sustainability 15:7255 Zhang C, Lei S, Wu H, Liao L, Wang X, Zhang L, Liu G, Wang G, Fang L, Song Z (2024) Simplified microbial network reduced microbial structure stability and soil functionality in alpine grassland along a natural aridity gradient. Soil Biology and Biochemistry 191:109366 Zhang SY, Speare KE, Long ZT, McKeever KA, Gyoerkoe M, Ramus AP, Mohorn Z, Akins KL, Hambridge SM, Graham NAJ, Nash KL, Selig ER, Bruno JF (2014) Is coral richness related to community resistance to and recovery from disturbance? PeerJ 2:e308 Zhao J, Li C, Wang T, Shi J, Song X, Liu Y (2023) Composition and Long-Term Variation Characteristics of Coral Reef Fish Species in Yongle Atoll, Xisha Islands, China. Biology 12:1062 Zhao J, Wang T, Li C, Shi J, Xie H, Luo L, Xiao Y, Liu Y (2024) Seven decades of transformation: evaluating the dynamics of coral reef fish communities in the Xisha Islands, South China Sea. Rev Fish Biol Fisheries 34:1261–1281 Zhou J, Deng Y, Luo F, He Z, Tu Q, Zhi X (2010) Functional Molecular Ecological Networks. mBio 1: 10.1128/mbio.00169 – 10 Additional Declarations No competing interests reported. Supplementary Files SupportingInformation.docx Cite Share Download PDF Status: Under Review Version 1 posted Reviewers agreed at journal 16 May, 2026 Reviewers agreed at journal 07 May, 2026 Reviewers invited by journal 05 May, 2026 Editor assigned by journal 24 Apr, 2026 Submission checks completed at journal 20 Apr, 2026 First submitted to journal 19 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9465484","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":637812334,"identity":"2d9797db-7afb-43ac-b551-d1d12769b5db","order_by":0,"name":"Cong Zeng","email":"","orcid":"","institution":"Shanghai Jiao Tong University","correspondingAuthor":false,"prefix":"","firstName":"Cong","middleName":"","lastName":"Zeng","suffix":""},{"id":637812335,"identity":"c9fac3e5-7624-4500-966d-aa26cd7a9102","order_by":1,"name":"Zheying Lin","email":"","orcid":"","institution":"Shanghai Jiao Tong University","correspondingAuthor":false,"prefix":"","firstName":"Zheying","middleName":"","lastName":"Lin","suffix":""},{"id":637812336,"identity":"0104ad64-6720-4cc6-9271-d19b5f4b02a4","order_by":2,"name":"Zhiyi Su","email":"","orcid":"","institution":"Shanghai Jiao Tong University","correspondingAuthor":false,"prefix":"","firstName":"Zhiyi","middleName":"","lastName":"Su","suffix":""},{"id":637812337,"identity":"6a226643-e5ce-4c6f-9791-f9011e1a5a74","order_by":3,"name":"Yuanbin Zhao","email":"","orcid":"","institution":"Shanghai Jiao Tong University","correspondingAuthor":false,"prefix":"","firstName":"Yuanbin","middleName":"","lastName":"Zhao","suffix":""},{"id":637812340,"identity":"0b3101f4-3493-4cfc-8200-4a0ba7c8047b","order_by":4,"name":"Yue Liu","email":"","orcid":"","institution":"Xiamen University","correspondingAuthor":false,"prefix":"","firstName":"Yue","middleName":"","lastName":"Liu","suffix":""},{"id":637812343,"identity":"d6753e3a-d719-4784-bc8d-d6828ec563fe","order_by":5,"name":"Mingjie Li","email":"","orcid":"","institution":"Xiamen University","correspondingAuthor":false,"prefix":"","firstName":"Mingjie","middleName":"","lastName":"Li","suffix":""},{"id":637812345,"identity":"b9eae548-eafe-4af8-9199-d62d8b9a4b2d","order_by":6,"name":"Chengxuan Zou","email":"","orcid":"","institution":"Shanghai Jiao Tong University","correspondingAuthor":false,"prefix":"","firstName":"Chengxuan","middleName":"","lastName":"Zou","suffix":""},{"id":637812347,"identity":"ab6abfcd-71fc-45ab-85ed-54e3026d0fde","order_by":7,"name":"Qiang Lin","email":"","orcid":"","institution":"South China Sea Institute of Oceanology, Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Qiang","middleName":"","lastName":"Lin","suffix":""},{"id":637812348,"identity":"ce319c7f-900e-487a-aa3a-708c77cab4bc","order_by":8,"name":"Yue Zheng","email":"","orcid":"","institution":"Xiamen University","correspondingAuthor":false,"prefix":"","firstName":"Yue","middleName":"","lastName":"Zheng","suffix":""},{"id":637812351,"identity":"4a5a53d7-7f51-457b-bed8-b9c9011f0f52","order_by":9,"name":"Hongqiang Yang","email":"","orcid":"","institution":"South China Sea Institute of Oceanology, Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Hongqiang","middleName":"","lastName":"Yang","suffix":""},{"id":637812352,"identity":"bf032b3f-afe4-44f5-8b14-b2ee863d15d5","order_by":10,"name":"Ling Cao","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAy0lEQVRIiWNgGAWjYHACxgMJDAxyEDYbkXpAWoxJ1ALEiQ1Ea+FvP3zgwMMdtekbrp0xYPhQdpiBf3YDfi0SZ9ISDiSeOZ47c3aOAeOMc4cZJO4cwK/FgCHH4EBi27HcfukcA2betsMMBhIJBLTwvwFrSWcDaflLlBYJsC01CfwgLYzEaJG48Qzol7YDhjNnpxUc7DmXziNxg4AW/v7kgw9/ttXJG9xO3vjgR5m1HP8MAlqg4DCYPADEPESpB4I6YhWOglEwCkbBSAQAG4JGJYHozw0AAAAASUVORK5CYII=","orcid":"","institution":"Xiamen University","correspondingAuthor":true,"prefix":"","firstName":"Ling","middleName":"","lastName":"Cao","suffix":""}],"badges":[],"createdAt":"2026-04-20 01:38:39","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9465484/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9465484/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":109246835,"identity":"64884183-d350-4c8f-9579-9a2d1d785ba5","added_by":"auto","created_at":"2026-05-14 08:11:01","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1411962,"visible":true,"origin":"","legend":"\u003cp\u003eSampling sites in the South China Sea from 2023 to 2024, indicating sites sampled in both years, and those unique to either 2023 or 2024.\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9465484/v1/ed32be5128188737003d63a2.jpeg"},{"id":109246832,"identity":"d4d3198b-5ff4-4e1b-8484-af37a154d735","added_by":"auto","created_at":"2026-05-14 08:11:00","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":156140,"visible":true,"origin":"","legend":"\u003cp\u003eDistribution of MOTUs across coral‑habitat conditions. The number of MOTUs detected in each habitat condition is shown in (a). The stacked chart (b) displays the number of shared MOTUs and its annotation of TOP10 orders across different combinations of condition. The matrix of connected dots (c) indicates the specific combinations of habitat conditions represented by each bar in (b).\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-9465484/v1/d9f471917651f656cab0cb1d.png"},{"id":109246893,"identity":"6d2774ed-a88c-4773-8643-ffa108be33de","added_by":"auto","created_at":"2026-05-14 08:11:17","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":382654,"visible":true,"origin":"","legend":"\u003cp\u003eComparisons of α-diversity based on MOTU between different coral habitat conditions (Note: * indicated the significant level of p \u0026lt; 0.05).\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-9465484/v1/2daa1ad76562ed954bb3b07c.png"},{"id":109246857,"identity":"4a3640eb-b6e9-4070-9686-eb8619a5feb6","added_by":"auto","created_at":"2026-05-14 08:11:07","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":28292150,"visible":true,"origin":"","legend":"\u003cp\u003eNetwork structure and modular analysis of reef fish communities across coral health conditions\u003c/p\u003e","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9465484/v1/251573e0614eb00ac36b0f7b.jpeg"},{"id":109246833,"identity":"68fd70ce-a5a8-4643-acc8-3854a6ce4a9e","added_by":"auto","created_at":"2026-05-14 08:11:00","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1158007,"visible":true,"origin":"","legend":"\u003cp\u003eNetwork robustness (a) and invulnerability (b) of reef fish community networks along the coral health gradient\u003c/p\u003e","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9465484/v1/2ac7f6e2fcec8a3b017bd614.jpeg"},{"id":109246851,"identity":"c0a7d672-4d58-4582-b743-e0b7d227ae4b","added_by":"auto","created_at":"2026-05-14 08:11:06","extension":"jpeg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":3699018,"visible":true,"origin":"","legend":"\u003cp\u003eStructural equation modeling (SEM) shows how human activities (fishing, construction), accumulated thermal stress, and environmental factors (temperature, DO, pH, NO\u003csub\u003e3\u003c/sub\u003e⁻ and NO\u003csub\u003e2\u003c/sub\u003e⁻) affect coral reef health, which in turn influences fish species richness and network stability. Arrows indicate standardized path coefficients (β), with thickness and style reflecting effect size and significance (\u003csup\u003e#\u003c/sup\u003ep\u0026lt;0.1; *p \u0026lt; 0.05; **p \u0026lt; 0.01; ***p \u0026lt; 0.001). R² values represent variance explained. (b) SEM-derived direct and indirect effects of each driver on species richness and network stability. Coral = coral reef health\u0026nbsp; conditions; Ats = accumulated thermal stress; Temp = temperature; Cons = construction; Sr = species richness; NN = NO\u003csub\u003e3\u003c/sub\u003e⁻ and NO\u003csub\u003e2\u003c/sub\u003e⁻.\u003c/p\u003e","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9465484/v1/c5bf6d6169205557bfed455e.jpeg"},{"id":109252489,"identity":"09549360-48d0-43e9-aa62-01cd90369283","added_by":"auto","created_at":"2026-05-14 09:27:11","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":35421659,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9465484/v1/98021a9e-9ebb-4ceb-a8cf-359804cfcb41.pdf"},{"id":109246845,"identity":"837c841e-5c80-4346-b685-129c233629bd","added_by":"auto","created_at":"2026-05-14 08:11:04","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":25006,"visible":true,"origin":"","legend":"","description":"","filename":"SupportingInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-9465484/v1/72551ed4537bb6f40e8d9814.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Environmental DNA reveals coral-driven shifts in reef fish networks: Implications for conservation","fulltext":[{"header":"1 Introduction","content":"\u003cp\u003eClimate change and human disturbances are accelerating marine biodiversity loss, with well-documented impacts on ecosystem functioning and resilience (Cardinale et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; IPCC \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Coral reefs, often described as the \u0026ldquo;rainforests of the sea\u0026rdquo;, have experienced dramatic declines, with over half of live coral cover lost since the 1950s due to ocean warming-induced bleaching, disease and predator outbreaks, and direct human impacts including overfishing, sedimentation, and pollution (Dunne et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Mora et al. \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Hughes et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Gove et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). These declines compromise the structural complexity and three-dimensional habitats that reef-building corals provide, altering shelter availability, trophic resources, and overall habitat quality. As a consequence, reef fishes, which rely on these habitats for foraging, reproduction, and refuge from predators, are directly affected. Reef fish assemblages play key roles in maintaining trophic balance and regulating algal growth (Zhang et al. \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Graham et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Fontoura et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), and are widely used as bio-indicators of reef health, offering early-warning signals for ecosystem disruption (Darling et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Stier et al. \u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). However, traditional measures of species richness or composition often fail to capture the broader ecological consequences of habitat degradation, particularly the disruption of species interactions, functional roles, and emergent properties that govern community resilience.\u003c/p\u003e \u003cp\u003eBeyond alpha and beta diversity, coral habitat degradation can reshape the interactions and organization of reef fish communities. Obligate coral specialists, such as \u003cem\u003eDascyllus aruanus\u003c/em\u003e and corallivorous butterflyfishes (\u003cem\u003eChaetodon\u003c/em\u003e spp.), typically decline sharply following coral loss, whereas generalist species may persist or temporarily expand, altering trophic interactions, competitive dynamics, and network modularity (Pratchett et al. \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2008\u003c/span\u003e, \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2012\u003c/span\u003e, \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Alvarez-Filip et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Ol\u0026aacute;n-Gonz\u0026aacute;lez et al. \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Habitat attributes such as structural complexity, coral morphology, bleaching condition, and benthic heterogeneity mediate these interactions by providing refugia, facilitating niche complementarity, and buffering predation or competition (Kerry and Bellwood \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Darling et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Clements and Choat \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Fisher \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The simplification or loss of these microhabitats can reduce functional redundancy, weaken trophic links, and decrease network robustness, potentially triggering community-level tipping points, hysteresis, or delayed recovery even after stressors are alleviated (Graham et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Darling et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; McWilliam et al. \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Santoso et al. \u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Examining how coral habitat condition influences species composition and ecological network structure can help identify key species and interactions that support community stability (Pozas-Schacre et al. \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Wolfe et al. \u003cspan citationid=\"CR92\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Hill and Hoogenboom \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Young et al. \u003cspan citationid=\"CR94\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eDespite extensive research on richness and beta diversity, the effects of coral habitat condition on fish ecological networks remain poorly understood. Traditional biodiversity metrics, such as species richness or diversity indices, often fail to capture the underlying species interactions, functional redundancy, or modular organization that govern ecosystem stability (Montoya et al. \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Tylianakis et al. \u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). Network-based approaches provide a powerful framework to quantify these properties, including connectivity, centrality of keystone or hub species, modularity, and redundancy of functional roles, all of which are critical for understanding how communities maintain resilience under environmental and anthropogenic stress (Pozas-Schacre et al. \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Wolfe et al. \u003cspan citationid=\"CR92\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Hill and Hoogenboom \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Young et al. \u003cspan citationid=\"CR94\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Highly connected nodes or module hubs may buffer against species loss by sustaining key trophic interactions, whereas the breakdown of such nodes can trigger cascading effects and community destabilization (McCann \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Dunne et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). Functional redundancy among species performing similar ecological roles can also mitigate the impacts of habitat degradation, allowing ecosystems to retain functionality even as some species decline (Bellwood et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Mouillot et al. \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). Changes in coral habitat condition, including reduced structural complexity, shifts in coral morphology, and habitat fragmentation, tend to affect module hubs and high-trophic-level species more strongly. As a result, network connectivity declines, energy flow is disrupted, and community assembly processes are altered (Santoso et al. \u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Young et al. \u003cspan citationid=\"CR94\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). These network-level effects often precede observable declines in richness or diversity, highlighting the importance of considering species interactions and network architecture when assessing the ecological consequences of coral degradation.\u003c/p\u003e \u003cp\u003eNetwork-based perspectives also provide practical guidance for coral reef management and restoration. Metrics derived from ecological networks can serve as early-warning indicators of functional collapse, helping managers detect ecosystem destabilization before species losses become irreversible (Emslie et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; McWilliam et al. \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). They can inform the selection of sites for coral transplantation or habitat enhancement by identifying reefs with sufficient structural complexity to support key functional interactions, including predator-prey relationships and trophic regulation. Post-restoration monitoring can then evaluate whether functional redundancy is re-established, module hubs recover, and energy and interaction flows within the community are restored, thereby supporting adaptive management and enhancing long-term ecosystem resilience (Luza et al. \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). By linking habitat condition, species interactions, and network structure, this approach helps clarify how coral degradation reshapes fish communities and can inform reef restoration and management. However, incorporating network metrics into restoration and monitoring is often constrained by the need for high-resolution data on species presence and interactions. Recent advances in molecular ecological network (MEN) methods, such as RMT-based network construction (Zhou et al. \u003cspan citationid=\"CR100\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Deng et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2012\u003c/span\u003e), provide robust frameworks for defining ecological networks from high-throughput data. Together with environmental DNA (eDNA) metabarcoding, which is defined here as genetic material shed by macroorganisms and suspended in environmental samples such as water (Miya \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) and allows efficient detection of fish assemblages, including rare or cryptic taxa, these methods enable the estimation of ecological networks in reef settings in which traditional surveys would be impractical(Meyer et al. \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe South China Sea (SCS) is a global hotspot of coral and fish biodiversity but is increasingly threatened by anthropogenic activities and climactic stress(Huang et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Pang et al. \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Shi et al. \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Its offshore archipelagos exhibit pronounced spatial gradients in coral habitat condition, offering a natural system to investigate how fish communities respond to habitat transformation. In this study, we applied eDNA metabarcoding to characterize reef fish assemblages across three archipelagos, integrating alpha and beta diversity analyses with ecological network metrics. Specifically, we aimed to (1) quantify differences in fish community composition and network structure among coral habitat conditions, (2) explore whether these differences exhibit stage-mediated or hysteretic patterns, and (3) evaluate how local anthropogenic pressures, particularly fishing intensity, influence these responses. By combining eDNA metabarcoding with network analysis, this study examines how coral habitat condition influences fish community structure and interaction patterns, with implications for reef monitoring and management. (Emslie et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; McWilliam et al. \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e"},{"header":"2 Material and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Habitat condition assessment\u003c/h2\u003e \u003cp\u003eDuring June 2023 and June 2024, a total of 141 samples were collected from 62 reef sites across the three principal archipelagos of the South China Sea (Xisha Islands, 12 sites; Zhongsha Islands, 7 sites; and Nansha Islands, 43 sites; Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). These sites encompassed coral reef systems representing a range of geomorphological types, including submerged reefs. Coral habitat condition was evaluated using two complementary metrics: (i) live scleractinian coral cover, and (ii) the incidence of visible bleaching. Coral cover was quantified from two 50 m photo-transects per site (~\u0026thinsp;0.4 m above the reef), while bleaching incidence was determined from diver observations along the transects. Reefs were then categorized into three condition classes: Excellent (\u0026gt;\u0026thinsp;30% live coral cover and no bleaching), Moderate (5\u0026ndash;30% live coral cover and/or localized bleaching), and Poor (\u0026lt;\u0026thinsp;5% live coral cover or extensive bleaching regardless of cover). These thresholds were defined based on expert recommendations for the long-term ecological status of coral reefs in the South China Sea. Environmental parameters, including seawater temperature, salinity, dissolved oxygen, and pH, were measured at discrete depth layers using a pump-CTD system. In addition, 1 L water samples were collected from each site to quantify NO₃⁻ and NO₂⁻ concentrations using an AutoAnalyzer 3 (AA3).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAt each site, water samples for eDNA analysis were collected following depth-specific strata. At the single shallow site (\u0026lt;\u0026thinsp;10 m depth), two 5 L seawater samples were collected from 3\u0026ndash;5 m depth. At sites deeper than 10 m, two 5 L samples were collected from each of two depth layers: subsurface (3\u0026ndash;5 m depth) and near-bottom (~\u0026thinsp;2 m above the seabed), resulting in 20 L per site. Samples were vacuum-filtered through sterile 0.45 \u0026micro;m polyethersulfone (PES) membranes within 10 min of collection. Filters were fixed in absolute ethanol and stored at -20\u0026deg;C until DNA extraction.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Environmental DNA extraction and sequencing\u003c/h2\u003e \u003cp\u003eDNA extraction was performed using a modified CTAB protocol based on the method of (Porebski et al. \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e1997\u003c/span\u003e). For PCR amplification, the MiFish-U primer was employed (Forward: GTCGGTAAAACTCGTGCCAGC; Reverse: CATAGTGGGGTATCTAATCCCAGTTTG), targeting an approximately 170 bp fragment of the 12S rRNA gene specific to fish (Miya et al. \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). The PCR thermal cycling conditions included an initial denaturation at 98\u0026deg;C for 3 min, followed by 34 cycles of denaturation at 98\u0026deg;C for 30 s, annealing at 55\u0026deg;C for 30 s, and extension at 72\u0026deg;C for 45 s, with a final extension step at 72\u0026deg;C for 5 min. Library preparation and high-throughput sequencing were conducted using the Illumina NovaSeq 6000 platform (PE250) by Shanghai Personal Biotechnology Co., Ltd.\u003c/p\u003e \u003cp\u003eTo monitor contamination, multiple negative controls were included: (i) a field blank consisting of sterile distilled water exposed during sampling and subsequently filtered, (ii) a filtration blank (sterile water passed through PES filters), and (iii) PCR negative controls. All blanks were carried through DNA extraction and amplification, and none yielded detectable amplicons.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Bioinformatics analysis\u003c/h2\u003e \u003cp\u003eIllumina sequencing reads were initially processed using Cutadapt (v2.3) to identify and remove primer sequences and low-quality reads. Subsequent processing of the 12S rRNA dataset and generation of Molecular Operational Taxonomic Units (MOTUs) was performed using the VSEARCH pipeline. Paired-end reads were first merged with the fastq_mergepairs function, followed by quality filtering with a maximum expected error threshold of 0.5. Sequences were then pre-clustered at 98% similarity, and chimeric sequences were subsequently removed. The remaining unique, non-chimeric sequences were clustered at 97% similarity to define MOTUs. MOTUs with fewer than one read or a total relative abundance of less than 0.1% across all samples were excluded to avoid the sequencing errors.\u003c/p\u003e \u003cp\u003eTo improve annotation accuracy, a regional reference database compatible with the MiFish primer was constructed using the rCRUX R package, which queried the NCBI nucleotide database to retrieve relevant sequence records (Curd et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). This database was further curated by cross-referencing with historical records of fish species reported in the South China Sea, sourced from published literatures (Fang and Lyu \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Qiu \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) and Fishbase (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.fishbase.se\u003c/span\u003e\u003cspan address=\"https://www.fishbase.se\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e, accessed: 26/02/2024), to exclude taxa unlikely to occur in the region. Taxonomic annotation of MOTUs was performed using MEGABLAST in Geneious Prime. MOTU sequences with \u0026ge;\u0026thinsp;97% identity and \u0026ge;\u0026thinsp;90% query coverage were assigned to the species level, while those with \u0026ge;\u0026thinsp;95% identity were annotated at the genus level. Sequences with identity\u0026thinsp;\u0026ge;\u0026thinsp;85% were classified at higher taxonomic ranks (Lamy et al. \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). All species names and taxonomic assignments were subsequently verified and standardized using the World Register of Marine Species (WoRMS).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Fish community comparison\u003c/h2\u003e \u003cp\u003eAlpha diversity of all samples was quantified using Hill numbers, specifically the number of observed MOTUs and Shannon diversity (q\u0026thinsp;=\u0026thinsp;1). Species richness was reflected by the number of observed MOTUs. Additionally, Faith's Phylogenetic Diversity (q\u0026thinsp;=\u0026thinsp;0) derived from MOTU sequences was calculated within the Hill number framework utilizing the hilldiv package (Alberdi and Gilbert \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Statistical comparisons of α-diversity indices across coral health conditions were performed using the Kruskal-Wallis test. Pairwise comparisons among groups were subsequently conducted using the Wilcoxon test with p-values adjusted for multiple comparisons via the Benjamini-Hochberg method. All visualizations were generated using the ggplot2 package (Wickham \u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e2016\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eBeta diversity was assessed to evaluate differences in fish community composition among sampling sites and coral habitat condition. Prior to analysis, the data were rarefied to the library with the lowest read count to minimize biases arising from unequal sequencing depths across libraries, followed by normalization and Hellinger transformation. Taxonomic β-diversity was quantified using Jaccard and Bray-Curtis dissimilarity indices, while weighted UniFrac distance was employed to assess phylogenetic β-diversity. To test for significant differences in community composition across coral health condition categories, PERMANOVA was performed using the adonis function from the vegan package (Oksanen J et al. \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Community network comparison\u003c/h2\u003e \u003cp\u003eTo evaluate the influence of coral habitat condition on the stability of associated fish communities, molecular ecological networks were constructed using the ggClusterNet package (Wen et al. \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Correlation matrices were generated based on Spearman\u0026rsquo;s rank correlation coefficients (|r| \u0026ge; 0.8, p\u0026thinsp;\u0026lt;\u0026thinsp;0.05), and topological properties (e.g., number of nodes, edges, connectivity, clustering coefficient, and centralization degree) were calculated. Network complexity was quantified as the normalized average of these properties, following established molecular ecological network frameworks (Zhou et al. \u003cspan citationid=\"CR100\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Zhang et al. \u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eNetwork stability was quantified using robustness and invulnerability, which provides a comprehensive assessment of network stability under both random and targeted perturbation scenarios. Network robustness was assessed through the Robustness.Random.removal function, which evaluates network integrity following the random removal of 50% nodes. Network invulnerability was evaluated through the natural connectivity decay curve generated by sequentially removing nodes in descending order of their combined betweenness and degree centrality scores, with slower decay rates indicating higher invulnerability.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Analysis of the driving factors\u003c/h2\u003e \u003cp\u003eStructural Equation Modeling (SEM) was employed to examine the mediating role of coral reef health in the relationships between human activities, climate change, and fish communities. Two primary hypotheses were formulated based on existing literature: (i) environmental factors (temperature, pH, DO, NO\u003csub\u003e3\u003c/sub\u003e and NO\u003csub\u003e2\u003c/sub\u003e), human activities (fishing, construction), and thermal stress were hypothesized to directly influence fish community diversity and network stability; and (ii) these factors were posited to exert indirect effects on fish communities through their impact on coral. Given the relatively small sample size and the need for mediation analysis, segmented structural equation models were constructed using the piecewiseSEM package (Lefcheck \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2016\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAccumulated thermal stress data were obtained from the NOAA Coral Reef Watch (CRW) version 3.1 daily global 5km satellite coral bleaching heat stress database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://coralreefwatch.noaa.gov/satellite/index.php\u003c/span\u003e\u003cspan address=\"https://coralreefwatch.noaa.gov/satellite/index.php\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), with the cumulative maximum annual Degree Heating Weeks (DHW) from 2018 to 2022 being utilized as an indicator of long-term thermal stress. Fishing intensity data were derived from the Global Fishing Watch dataset (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://globalfishingwatch.org/\u003c/span\u003e\u003cspan address=\"https://globalfishingwatch.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), where the sum of apparent fishing effort from 2018 to 2022 was calculated. Construction data were collected through field surveys and subsequently verified using Google Earth imagery. Ecological network stability was quantified through network robustness under different coral habitat condition.\u003c/p\u003e \u003c/div\u003e"},{"header":"3 Results","content":"\u003cp\u003eAmong the 141 samples analyzed, 32 were classified as having excellent coral habitat condition, 32 as moderate, and 77 as poor. The highest level of coral reef degradation was observed in the Xisha Islands, where 95.24% of samples were classified as poor. This was followed by the Zhongsha Islands, with 68.42% of sites exhibiting degraded conditions, and the Nansha Islands, where 43.56% of stations were in poor condition. From these samples, a total of 8,116 molecular operational taxonomic units (MOTUs) were recovered, with an average of 122,629 sequencing reads per sample. These MOTUs correspond to 367 fish species, spanning 33 orders and 73 families.\u003c/p\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Fish community difference among coral habitat conditions\u003c/h2\u003e \u003cp\u003eThe total number of MOTUs exhibited an inverse relationship with habitat condition, declining progressively from Poor to Excellent reefs. A total of 1,600 MOTUs were shared across all three habitat condition categories, representing 19.71% of the total MOTUs detected. The greatest overlap in MOTU composition was observed between the Moderate and Poor categories, while the Excellent and Poor categories shared the fewest MOTUs. This pattern suggests increasing community dissimilarity with greater differences in habitat condition, and indicates substantial compositional shifts occurring at more advanced stages of reef degradation. Across all habitat condition categories, the most dominant fish orders were Perciformes, Eupercaria \u003cem\u003eincertae sedis\u003c/em\u003e, and Acanthuriformes.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eComparative analysis of α-diversity indices across coral habitat condition categories revealed distinct patterns in biodiversity distribution (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). The number of observed MOTUs differed significantly between reefs classified as Moderate and Poor (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05), but no significant difference was observed between Excellent and Moderate reefs (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). In contrast, neither Shannon diversity nor Faith\u0026rsquo;s phylogenetic diversity (PD) index showed significant variation across habitat condition categories (p\u0026thinsp;\u0026gt;\u0026thinsp;0.05), indicating that these metrics may not adequately capture biodiversity changes associated with habitat degradation (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb-c). This suggests that richness is more sensitive than diversity or phylogenetic metrics to coral degradation, likely due to the persistence of phylogenetically redundant taxa in early stages of reef decline.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003ePERMANOVA analyses revealed that coral habitat condition significantly influenced fish community composition, with consistent effects across Jaccard, Bray-Curtis, and weighted UniFrac distances (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05; Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Although α-diversity metrics showed significant changes only at later stages of habitat decline, β-diversity patterns revealed that compositional shifts in fish communities began earlier. Specifically, pairwise comparisons showed significant differences in taxonomic β-diversity between Excellent reefs and those classified as Moderate or Poor (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05; Table S2), suggesting that community restructuring precedes measurable losses in richness. Together, these findings highlight the ecological sensitivity of fish assemblages to coral habitat degradation and emphasize the importance of maintaining reef integrity to preserve biodiversity.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ePERMANOVA results based on Jaccard, Bray-Curtis, and Weighted Unifrac distances\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDf\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSS\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMS\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eF\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eR2\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003ep\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"7\" nameend=\"c7\" namest=\"c1\"\u003e \u003cp\u003eJaccard\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eConditions\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.155\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.578\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.212\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.017\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eResiduals\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e138\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e65.769\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.477\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.983\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e140\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e66.924\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"7\" nameend=\"c7\" namest=\"c1\"\u003e \u003cp\u003eBray-Curtis\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eConditions\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.265\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.632\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.388\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.020\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eResiduals\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e138\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e62.862\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.456\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.980\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e140\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e64.126\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"7\" nameend=\"c7\" namest=\"c1\"\u003e \u003cp\u003eWeighted-Unifrac\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eConditions\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.030\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.015\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.287\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.018\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.141\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eResiduals\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e138\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.633\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.012\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.982\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e140\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.663\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Network structural changes among coral habitat conditions\u003c/h2\u003e \u003cp\u003eAnalysis of multidimensional topological metrics revealed a pronounced nonlinear response of fish community network structure to coral habitat condition (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Reefs in Moderate condition exhibited the highest overall network complexity (Table S3), characterized by increased node connectivity and a more compact, integrated network architecture. However, as coral condition deteriorated further, network fragmentation became increasingly pronounced.\u003c/p\u003e \u003cp\u003eIn Moderate-condition reefs, only 13.53% of nodes were isolated, whereas in Poor-condition reefs, this proportion nearly doubled to 25%, indicating a substantial breakdown of species interactions. Additional network properties corroborated this threshold-like transition: Moderate-condition networks had the highest edge connectivity (34.45) and maintained average path lengths (2.29) comparable to those in Excellent reefs, reflecting efficient information flow and robust taxonomic associations. By contrast, Poor-condition networks exhibited significantly longer average path lengths (2.46) and reduced clustering coefficients, key indicators of weakened community cohesion and impaired ecological connectivity. These patterns indicate that Moderate-condition reefs maintain higher network connectivity compared to both Excellent and Poor reefs, highlighting a nonlinear response to habitat degradation.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Role of module hubs and network robustness in fish communities\u003c/h2\u003e \u003cp\u003eIn fish community networks associated with coral reefs in Excellent and Moderate condition, module hubs, nodes with disproportionately high within-module connectivity, were consistently identified, whereas no such hubs were detected in Poor-condition reefs (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). These hubs indicate species that play central roles in maintaining intra-module cohesion and stabilizing interaction structures. The taxonomically annotated module hub MOTUs ranked within the top 30% in relative abundance for their respective reef conditions (Table S4), suggesting a strong correlation between numerical dominance and network centrality. These hub MOTUs were primarily assigned to 25 genera, including Epinephelus, Carcharhinus, and Cirrhitichthys. Among them, 14 MOTUs were identified as carnivores and 2 as herbivores, with trophic levels exceeding 3.8 for the carnivorous taxa, indicating their role as high-level predators. In contrast, their decline or disappearance in Poor-condition reefs likely disrupts trophic structuring and erodes the regulatory capacity of the community. Most module hubs were carnivorous taxa suggesting that higher trophic-level species play a key role in maintaining network integrity in healthier reefs.\u003c/p\u003e \u003cp\u003eNetwork robustness analysis further revealed significantly higher structural stability in fish communities associated with healthy reefs compared to degraded ones (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05; Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea). In particular, Poor-condition reefs exhibited the lowest robustness values across all categories (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Complementary natural connectivity analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb) showed that Excellent-condition networks maintained the highest overall resilience under simulated node removal, although Moderate-condition reefs exhibited superior resistance to intermediate disturbance levels (20\u0026ndash;40% node loss). In contrast, Poor-condition networks suffered rapid breakdown after just 20% of nodes were removed, despite initially appearing relatively connected.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e3.4 Stressors related with richness and network stability\u003c/h2\u003e \u003cp\u003eStructural equation modeling (SEM) revealed distinct pathways through which anthropogenic and environmental stressors influence coral reef fish communities (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea). Among these, fishing pressure exerted the most pronounced direct negative effect on fish species richness (β = -0.25, p\u0026thinsp;=\u0026thinsp;0.022), while ocean acidification, indicated by pH, showed a marginally significant direct negative impact (β = -0.24, p\u0026thinsp;=\u0026thinsp;0.053). Other environmental variables, including accumulated thermal stress and coastal construction, did not demonstrate significant direct effects on species richness. In contrast, coral habitat condition emerged as a critical positive driver of fish community network stability (β\u0026thinsp;=\u0026thinsp;0.96, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), underscoring the foundational role of reef health in maintaining ecological connectivity. Coastal construction also exhibited a modest but significant positive influence on network stability (β\u0026thinsp;=\u0026thinsp;0.09, p\u0026thinsp;=\u0026thinsp;0.003), potentially reflecting localized management or habitat modifications. Other factors, thermal stress, fishing pressure, and environmental variables aside from temperature, showed negligible direct effects on network stability.\u003c/p\u003e \u003cp\u003eNotably, fishing pressure significantly undermined coral habitat condition (β = -0.21, p\u0026thinsp;=\u0026thinsp;0.048), with accumulated thermal stress also approaching a marginally significant negative impact (β = -0.21, p\u0026thinsp;=\u0026thinsp;0.054). Coral habitat condition acted as a key mediator, channeling indirect effects from fishing pressure to fish biodiversity and community structure. For species richness, the indirect effect of fishing pressure via coral habitat condition was similar in magnitude to its direct effect. More strikingly, for network stability, the indirect influence of fishing pressure mediated through coral habitat condition surpassed both its direct and total effects, highlighting the central role of reef condition in buffering or amplifying anthropogenic impacts. Together, these findings indicate that fishing pressure primarily undermines reef fish networks indirectly by degrading coral habitat, reinforcing the central role of reef condition as a mediator of anthropogenic stress.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"4 Discussion","content":"\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Harnessing eDNA for Coral Reef Fish Monitoring: Opportunities and Limitations for Management\u003c/h2\u003e \u003cp\u003eCoral reefs on remote oceanic islands are ecologically isolated and highly sensitive to climate-induced degradation (Strona et al. \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Monitoring reef biodiversity in remote islands remains challenging due to logistical constraints, highlighting the need for alternative survey methods (Samoilys et al. \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Traditional underwater visual census (UVC) methods face limitations in these habitats due to strong currents, turbidity, and logistical challenges (Polanco Fern\u0026aacute;ndez et al. \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). As a diver-independent alternative, environmental DNA (eDNA) metabarcoding has gained traction. Despite its strengths, including cost-efficiency, taxonomic breadth, and the ability to detect rare or elusive species, the approach is constrained by incomplete reference databases, particularly for cryptobenthic fishes that dominate coral reef ecosystems (Brandl et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; G\u0026oacute;mez-Buckley et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn our study, only 45.6% of the 2,669 regionally documented species had matching barcodes in public repositories, underscoring this limitation. Nonetheless, eDNA metabarcoding detected 265 species from 21 water samples across seven reef sites, surpassing prior UVC records in the region (Shuting et al. \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Taxonomic coverage was extensive, spanning 33 orders, and included ecologically and functionally important coral reef taxa such as Acanthuridae and Serranidae, aligning with previous research ((Zhao et al. \u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e2023\u003c/span\u003e, \u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Although some cryptobenthic taxa remained unidentified, MOTUs provided reliable biodiversity proxies (Juhel et al. \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), allowing their ecological roles to be integrated into community-level and network-based analyses (Mathon et al. \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Malik et al. \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eFrom a conservation and management perspective, these findings emphasize that eDNA is not simply a substitute for traditional approaches but a complementary and often more practical tool for large-scale biodiversity assessment. Especially in remote, politically sensitive, or physically challenging regions, eDNA provides a scalable method to track biodiversity dynamics where conventional surveys are infeasible (Bohmann et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Stat et al. \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). The integration of eDNA with UVC or baited remote underwater video (BRUV) surveys could enhance both the breadth and depth of monitoring, creating hybrid approaches that are directly applicable to management programs (Evans et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Muenzel et al. \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Despite its limitations, eDNA provides a practical approach for monitoring reef fish assemblages, particularly where early detection of biodiversity change is needed (Rees et al. \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Nonlinear and Stage Responses of Fish Networks to Coral Habitat Loss: Early-Warning Signals for Management\u003c/h2\u003e \u003cp\u003eOur findings reveal a nonlinear response of coral reef fish diversity and ecological stability to habitat degradation, aligning with prior studies that suggest lagged responses of ichthyofaunal assemblages to Scleractinian coral loss (Wilson et al. \u003cspan citationid=\"CR91\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Adam et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). This pattern may be driven by two key mechanisms. First, the three-dimensional structural complexity of live coral provides a buffering effect in the early stages of degradation. Even during initial bleaching events, reefs may retain sufficient architectural integrity to support diverse microhabitats, modulate predator-prey interactions, and sustain interspecific competition (Darling et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Transitional states, such as the coexistence of bleached coral and algal cover, can temporarily support herbivorous and omnivorous fishes (Koester et al. \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). This is consistent with our ZIPI network analysis, which identified herbivorous species as module hub nodes in moderately degraded reef systems.\u003c/p\u003e \u003cp\u003eSecond, fish recruitment dynamics exhibit pronounced time-lag effects. Species that depend heavily on live coral during juvenile stages may persist into adulthood even after coral decline, delaying the onset of biodiversity loss (Wilson et al. \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). This may contribute to the temporary resilience of fish communities during the early stages of coral degradation observed in our study. Indeed, our results indicate that coral reef fish communities exhibit ecological network robustness during moderate degradation, supporting trophic models that predict sustained fish productivity under initial stress (Rogers et al. \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2018a\u003c/span\u003e, \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e2018b\u003c/span\u003e). The temporary increase in prey availability may even benefit large-bodied predators (Graham et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2007\u003c/span\u003e), which aligns with our finding that key MOTUs in moderately degraded reefs are primarily high-trophic-level species.\u003c/p\u003e \u003cp\u003eHowever, this resilience is fragile. Our analysis revealed that moderate degradation coincides with reduced energy flow velocity and lower species turnover (Morais et al. \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), suggesting that apparent stability masks underlying vulnerability. Once coral condition declines past a certain \u0026ldquo;refuge threshold,\u0026rdquo; resilience mechanisms break down (Morais et al. \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Severely degraded reefs exhibited markedly lower ecological network robustness, accompanied by the loss of keystone species and functional redundancy (Kerry and Bellwood \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). The disappearance of critical MOTUs likely reflects compounding stressors, including habitat collapse, delayed predator recruitment failure (Pratchett et al. \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), and intensified overfishing pressure (Sherman et al. \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Zhao et al. \u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eInterestingly, our robustness analysis indicates that some degree of natural connectivity persists even after critical MOTUs are lost, suggesting that reef fish communities may transition into new, alternative stable states (Fung et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Luza et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). However, these alternative states are more vulnerable to disturbance and support fewer ecological functions than those on healthier reefs. This insight has direct management implications: monitoring programs should move beyond coral cover alone and incorporate network-based indicators such as redundancy, hub persistence, or trophic flow velocity as early-warning signals of ecosystem collapse. These results suggest that early changes in network structure should be monitored to detect declines before major losses in community function occur (Bellwood et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2012\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e4.3 Linking Anthropogenic and Climatic Pressures to Reef Fish Decline: Implications for Multilateral Governance\u003c/h2\u003e \u003cp\u003eOur study highlights that coral reef condition is the most influential factor affecting the stability of reef fish ecological networks. However, this condition is itself shaped by a suite of anthropogenic and climatic stressors that exert both direct and indirect impacts. Among these, escalating fishing pressure in the South China Sea (SCS) has emerged as a dominant driver. Despite China\u0026rsquo;s seasonal fishing moratoriums covering over 40% of regional yields(Wu et al. \u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), the absence of coordinated governance among the ten littoral states (Yu and Chang \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Lai and Yu \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) has permitted unsustainable exploitation of remote reefs (Li et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Hong et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Our findings align with previous reports documenting trophic downgrading and biodiversity collapse in the Xisha (Zhao et al. \u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Qiu et al. \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e2024\u003c/span\u003e)and Nansha Islands (Dai et al. \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Kang et al. \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), where non-selective fishing methods such as trawling and gillnetting (Guan et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2021\u003c/span\u003e)disproportionately depleted carnivorous key MOTUs at degraded sites.\u003c/p\u003e \u003cp\u003eBeyond biomass removal, these extractive practices exacerbate habitat degradation. Anchoring, bottom trawling, and destructive methods (e.g., blast or cyanide fishing) inflict physical damage on coral structures, reducing reef complexity and fragmenting habitat mosaics (Pet-Soede et al. \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Arai \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Williams et al. \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Borland et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). This fragmentation is evident in coral cover gradients, from \u0026lt;\u0026thinsp;20% in heavily impacted nearshore Xisha reefs (Chen et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2007\u003c/span\u003e) to \u0026gt;\u0026thinsp;35% in relatively intact Nansha systems(Tkachenko and Hoang \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), reflecting shifts in fishing effort (Wu et al. \u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Such fragmentation interacts synergistically with thermal stress, as cumulative warming accelerates bleaching events and reef collapse (Gordon et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Asbury et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The resultant decline in structural complexity diminishes fish recruitment, dispersal, and connectivity, reinforcing negative feedback loops that erode resilience (Kubicek and Reuter \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2016\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThese findings underscore that reef condition is not an isolated ecological variable, but one embedded within a complex network of interacting anthropogenic and climatic drivers. Effective management will require coordinated actions, such as regulating fishing practices, protecting structurally complex reefs, and improving cooperation among neighboring countries. Island construction and restoration projects further complicate this picture. While often assumed to cause permanent coral loss (Tkachenko and Hoang \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), our model detected no significant negative effects on fish diversity (p\u0026thinsp;\u0026gt;\u0026thinsp;0.05) and, unexpectedly, a positive effect on ecological network stability (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05), likely reflecting post-construction recovery periods that allowed partial ecological reorganization. Nevertheless, these results reinforce a critical conservation principle: the long-term resilience of reef fish communities hinges on preserving coral habitat integrity. Without coordinated management that addresses both local pressures and warming events, reefs that appear stable today may still shift toward degraded states with reduced ecological function.\u003c/p\u003e \u003c/div\u003e"},{"header":"5 Conclusion and Management Implications","content":"\u003cp\u003eOur findings demonstrate that reef fish assemblages do not decline smoothly with coral degradation. Instead, modest early losses of live coral can trigger short-lived increases in biodiversity and connectivity, but once a critical threshold is surpassed, species richness, network stability, and key ecological functions collapse abruptly. By coupling high-resolution eDNA metabarcoding with ecological network and structural-equation analyses, we reveal that top-down control by apex predators and other stabilizing interactions erode progressively along degradation gradients. Fishing pressure and cumulative thermal stress emerged as major drivers, acting directly by removing fish biomass and bleaching corals, and indirectly by simplifying habitat architecture and food-web structure.\u003c/p\u003e \u003cp\u003eWhile fish communities may temporarily reorganize into stable but functionally compromised states, such resilience is fragile and potentially irreversible if stressors persist. Effective conservation therefore requires proactive, multilateral strategies that curb local anthropogenic impacts, mitigate global climate stress, and preserve structural complexity before reefs are pushed beyond ecological tipping points. Our results highlight three key applications:\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eIncorporating network metrics into monitoring frameworks as early-warning indicators of collapse.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eIntegrating eDNA into regional biodiversity monitoring programs, particularly for remote and politically sensitive reefs.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eDesigning multi-scalar governance frameworks, combining local enforcement with transboundary MPAs, to address both anthropogenic and climatic stressors.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003eSuch coordinated action is especially urgent in the South China Sea and other geopolitically complex regions where management jurisdictions overlap and climate extremes are intensifying. By linking eDNA-based biodiversity data with network-level ecological insights, our study provides a mechanistically grounded framework for anticipating coral reef collapse and guiding interventions that sustain resilience in the Anthropocene.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eConflict of Interest Statement\u003c/h2\u003e \u003cp\u003eThe authors declare that there are no financial or non-financial conflicts of interest that could have influenced the work reported in this paper.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eCong Zeng: Conceptualization, Methodology, Supervision, Writing - review \u0026amp; editing. Zheying Lin: Investigation, Formal analysis, Writing - original draft. Zhiyi Su, Yuanbin Zhao: Writing - review \u0026amp; editing. Yue Liu: Data curation, Formal analysis. Mingjie Li, Chengxuan Zou, Yue Zheng, Hongqiang Yang: Field investigation. Qiang Lin: Funding acquisition, Field investigation. Ling Cao: Conceptualization, Supervision, Writing - review \u0026amp; editing, Funding acquisition.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eWe gratefully acknowledge the financial support provided by the Ministry of Science and Technology of China (2022-24), the National Natural Science Foundation of China (42425603) and National Natural Science Foundation of China (42206082). The views, findings, and conclusions presented within this material are solely those of the authors and do not necessarily represent the perspectives of the funding organizations.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eAll environmental DNA (eDNA) amplicon sequencing data generated in this study have been deposited in the NCBI BioSample database under accession numbers SAMN50205399\u0026ndash;SAMN50205543 as part of submission SUB15490360.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAdam TC, Brooks AJ, Holbrook SJ, Schmitt RJ, Washburn L, Bernardi G (2014) How will coral reef fish communities respond to climate-driven disturbances? Insight from landscape-scale perturbations. Oecologia 176:285\u0026ndash;296\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlberdi A, Gilbert MTP (2019) hilldiv: an R package for the integral analysis of diversity based on Hill numbers. 545665\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlvarez-Filip L, Gill JA, Dulvy NK (2011) Complex reef architecture supports more small-bodied fishes and longer food chains on Caribbean reefs. Ecosphere 2:art118\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArai T (2015) Diversity and conservation of coral reef fishes in the Malaysian South China Sea. Rev Fish Biol Fisheries 25:85\u0026ndash;101\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAsbury M, Innes-Gold AA, Wulstein DM, Madin EMP, Madin JS, McManus LC (2024) Recovery potential of fish and coral populations following ecological disturbance. Ecosphere 15:e4915\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBellwood DR, Baird AH, Depczynski M, Gonz\u0026aacute;lez-Cabello A, Hoey AS, Lef\u0026egrave;vre CD, Tanner JK (2012) Coral recovery may not herald the return of fishes on damaged coral reefs. Oecologia 170:567\u0026ndash;573\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBellwood DR, Hughes TP, Folke C, Nystr\u0026ouml;m M (2004) Confronting the coral reef crisis. Nature 429:827\u0026ndash;833\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBohmann K, Evans A, Gilbert MTP, Carvalho GR, Creer S, Knapp M, Yu DW, de Bruyn M (2014) Environmental DNA for wildlife biology and biodiversity monitoring. Trends in Ecology \u0026amp; Evolution 29:358\u0026ndash;367\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBorland HP, Gilby BL, Henderson CJ, Leon JX, Schlacher TA, Connolly RM, Pittman SJ, Sheaves M, Olds AD (2021) The influence of seafloor terrain on fish and fisheries: A global synthesis. Fish and Fisheries 22:707\u0026ndash;734\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrandl SJ, Goatley CHR, Bellwood DR, Tornabene L (2018) The hidden half: ecology and evolution of cryptobenthic fishes on coral reefs. Biological Reviews 93:1846\u0026ndash;1873\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCardinale BJ, Duffy JE, Gonzalez A, Hooper DU, Perrings C, Venail P, Narwani A, Mace GM, Tilman D, Wardle DA, Kinzig AP, Daily GC, Loreau M, Grace JB, Larigauderie A, Srivastava DS, Naeem S (2012) Biodiversity loss and its impact on humanity. Nature 486:59\u0026ndash;67\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen G, Li Y, Chen X (2007) Species diversity of fishes in the coral reefs of South China Sea. Biodiversity Science 15:373\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eClements KD, Choat JH (2018) Nutritional Ecology of Parrotfishes (Scarinae, Labridae). Biology of Parrotfishes. CRC Press,\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCurd EE, Gal L, Gallego R, Silliman K, Nielsen S, Gold Z (2024) rCRUX: A rapid and versatile tool for generating metabarcoding reference libraries in R. Environmental DNA 6:e489\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDai X, Li Y, Cai Y, Gong Y, Zhang J, Chen Z (2022) Variations in Fish Community Structure at the Lagoon of Yongshu Reef, South China Sea between 1999 and 2016\u0026ndash;2019. Diversity 14:763\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDarling ES, Graham NAJ, Januchowski-Hartley FA, Nash KL, Pratchett MS, Wilson SK (2017) Relationships between structural complexity, coral traits, and reef fish assemblages. Coral Reefs 36:561\u0026ndash;575\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDeng Y, Jiang Y-H, Yang Y, He Z, Luo F, Zhou J (2012) Molecular ecological network analyses. BMC Bioinformatics 13:113\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDunne JA, Williams RJ, Martinez ND (2004) Network structure and robustness of marine food webs. Marine Ecology Progress Series 273:291\u0026ndash;302\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEmslie MJ, Bray P, Cheal AJ, Johns KA, Osborne K, Sinclair-Taylor T, Thompson CA (2020) Decades of monitoring have informed the stewardship and ecological understanding of Australia\u0026rsquo;s Great Barrier Reef. Biological Conservation 252:108854\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEvans NT, Shirey PD, Wieringa JG, Mahon AR, Lamberti GA (2017) Comparative Cost and Effort of Fish Distribution Detection via Environmental DNA Analysis and Electrofishing. Fisheries 42:90\u0026ndash;99\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFang H, Lyu X (2019) Reef Fish Identification of Nansha Islands. China Ocean University Press. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.hceis.com/home/book_view.aspx?id=11498\u003c/span\u003e\u003cspan address=\"https://www.hceis.com/home/book_view.aspx?id=11498\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFisher WS (2023) Relating fish populations to coral colony size and complexity. Ecological Indicators 148:110117\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFontoura L, Zawada KJA, D\u0026rsquo;agata S, \u0026Aacute;lvarez-Noriega M, Baird AH, Boutros N, Dornelas M, Luiz OJ, Madin JS, Maina JM, Pizarro O, Torres-Pulliza D, Woods RM, Madin EMP (2020) Climate-driven shift in coral morphological structure predicts decline of juvenile reef fishes. Global Change Biology 26:557\u0026ndash;567\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFung T, Seymour RM, Johnson CR (2011) Alternative stable states and phase shifts in coral reefs under anthropogenic stress. Ecology 92:967\u0026ndash;982\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eG\u0026oacute;mez-Buckley MC, Gallego R, Arranz V, Halafihi T, Stone K, Erdmann M, Tornabene LM (2023) Comparing anesthetic stations and environmental DNA sampling to determine community composition of cryptobenthic coral reef fishes of Vava\u0026rsquo;u, Kingdom of Tonga. Coral Reefs 42:785\u0026ndash;797\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGordon TAC, Harding HR, Wong KE, Merchant ND, Meekan MG, McCormick MI, Radford AN, Simpson SD (2018) Habitat degradation negatively affects auditory settlement behavior of coral reef fishes. Proceedings of the National Academy of Sciences 115:5193\u0026ndash;5198\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGove JM, Williams GJ, Lecky J, Brown E, Conklin E, Counsell C, Davis G, Donovan MK, Falinski K, Kramer L, Kozar K, Li N, Maynard JA, McCutcheon A, McKenna SA, Neilson BJ, Safaie A, Teague C, Whittier R, Asner GP (2023) Coral reefs benefit from reduced land\u0026ndash;sea impacts under ocean warming. Nature 621:536\u0026ndash;542\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGraham N a. J, Wilson SK, Jennings S, Polunin NVC, Robinson J, Bijoux JP, Daw TM (2007) Lag Effects in the Impacts of Mass Coral Bleaching on Coral Reef Fish, Fisheries, and Ecosystems. Conservation Biology 21:1291\u0026ndash;1300\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGraham NAJ, McClanahan TR, MacNeil MA, Wilson SK, Cinner JE, Huchery C, Holmes TH (2017) Human Disruption of Coral Reef Trophic Structure. Current Biology 27:231\u0026ndash;236\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGraham NAJ, Wilson SK, Jennings S, Polunin NVC, Bijoux JP, Robinson J (2006) Dynamic fragility of oceanic coral reef ecosystems. Proceedings of the National Academy of Sciences 103:8425\u0026ndash;8429\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuan Y, Zhang J, Zhang X, Li Z, Meng J, Liu G, Bao M, Cao C (2021) Identification of Fishing Vessel Types and Analysis of Seasonal Activities in the Northern South China Sea Based on AIS Data: A Case Study of 2018. Remote Sensing 13:1952\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHill TS, Hoogenboom MO (2022) The indirect effects of ocean acidification on corals and coral communities. Coral Reefs 41:1557\u0026ndash;1583\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHong X, Zhang K, Li J, Xu Y, Sun M, Xu S, Cai Y, Qiu Y, Chen Z (2024) Stock Assessment of the Commercial Small Pelagic Fishes in the Beibu Gulf, the South China Sea, 2006\u0026ndash;2020. Biology 13:226\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang D, Licuanan WY, Hoeksema BW, Chen CA, Ang PO, Huang H, Lane DJW, Vo ST, Waheed Z, Affendi YA, Yeemin T, Chou LM (2015) Extraordinary diversity of reef corals in the South China Sea. Mar Biodiv 45:157\u0026ndash;168\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHughes TP, Barnes ML, Bellwood DR, Cinner JE, Cumming GS, Jackson JBC, Kleypas J, van de Leemput IA, Lough JM, Morrison TH, Palumbi SR, van Nes EH, Scheffer M (2017) Coral reefs in the Anthropocene. Nature 546:82\u0026ndash;90\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIPCC (2022) Climate Change 2022: Impacts, Adaptation, and Vulnerability. Contribution of Working Group II to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change.Cambridge University Press.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJuhel J-B, Utama RS, Marques V, Vimono IB, Sugeha HY, Kadarusman, Pouyaud L, Dejean T, Mouillot D, Hocd\u0026eacute; R (2020) Accumulation curves of environmental DNA sequences predict coastal fish diversity in the coral triangle. Proc Biol Sci 287:20200248\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKang Z, Wang T, Li C, Zhao J, Shi J, Xie H, Liu Y (2024) Species composition and succession of coral reef fishes in Huaguang Reef, Xisha Islands. Water Biology and Security 3:100273\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKerry JT, Bellwood DR (2012) The effect of coral morphology on shelter selection by coral reef fishes. Coral Reefs 31:415\u0026ndash;424\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKerry JT, Bellwood DR (2015) Do tabular corals constitute keystone structures for fishes on coral reefs? Coral Reefs 34:41\u0026ndash;50\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKoester A, Gord\u0026oacute;\u0026ndash;Vilaseca C, Bunbury N, Ferse SCA, Ford A, Haupt P, A\u0026rsquo;Bear L, Bielsa M, Burt AJ, Letori J, Mederic E, Nancy E, Sanchez C, Waller M, Wild C (2023) Impacts of coral bleaching on reef fish abundance, biomass and assemblage structure at remote Aldabra Atoll, Seychelles: insights from two survey methods. Front Mar Sci 10:1230717\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKubicek A, Reuter H (2016) Mechanics of multiple feedbacks in benthic coral reef communities. Ecological Modelling 329:29\u0026ndash;40\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLai M, Yu W (2025) Analysis of fishery ecosystem structure in the South China Sea. JFSC 31:1336\u0026ndash;1350\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLamy T, Pitz KJ, Chavez FP, Yorke CE, Miller RJ (2021) Environmental DNA reveals the fine-grained and hierarchical spatial structure of kelp forest fish communities. Sci Rep 11:14439\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLefcheck JS (2016) piecewiseSEM: Piecewise structural equation modelling in r for ecology, evolution, and systematics. Methods in Ecology and Evolution 7:573\u0026ndash;579\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi M, Xu Y, Sun M, Li J, Zhou X, Chen Z, Zhang K (2023) Impacts of Strong ENSO Events on Fish Communities in an Overexploited Ecosystem in the South China Sea. Biology 12:946\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuza AL, Bender MG, Ferreira CEL, Floeter SR, Francini-Filho RB, Longo GO, Pinheiro HT, Quimbayo JP, Bastazini VAG (2024) Coping with collapse: Functional robustness of coral-reef fish network to simulated cascade extinction. Global Change Biology 30:e17513\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuza AL, Quimbayo JP, Ferreira CEL, Floeter SR, Francini-Filho RB, Bender MG, Longo GO (2022) Low functional vulnerability of fish assemblages to coral loss in Southwestern Atlantic marginal reefs. Sci Rep 12:17164\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMalik MDA, Ambariyanto A, Hartati R, Nursalim N, Kholilah N, Kurniasih EM, Anggoro AW, Prasetia R, Syamsyuni Y, Muh F, Cahyani NKD (2025) eDNA uncovers hidden fish diversity in the coral reef ecosystems of Karimunjawa National Park, Indonesia. Regional Studies in Marine Science 81:103945\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMathon L, Marques V, Mouillot D, Albouy C, Andrello M, Baletaud F, Borrero-P\u0026eacute;rez GH, Dejean T, Edgar GJ, Grondin J, Guerin P-E, Hocd\u0026eacute; R, Juhel J-B, Kadarusman, Maire E, Mariani G, McLean M, Polanco F. A, Pouyaud L, Stuart-Smith RD, Sugeha HY, Valentini A, Vigliola L, Vimono IB, Pellissier L, Manel S (2022) Cross-ocean patterns and processes in fish biodiversity on coral reefs through the lens of eDNA metabarcoding. Proc Biol Sci 289:20220162\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcCann KS (2000) The diversity\u0026ndash;stability debate. Nature 405:228\u0026ndash;233\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcWilliam M, Pratchett MS, Hoogenboom MO, Hughes TP (2020) Deficits in functional trait diversity following recovery on coral reefs. Proceedings: Biological Sciences 287:1\u0026ndash;9\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMeyer JM, Leempoel K, Losapio G, Hadly EA (2020) Molecular Ecological Network Analyses: An Effective Conservation Tool for the Assessment of Biodiversity, Trophic Interactions, and Community Structure. Front Ecol Evol 8:588430\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMiya M (2022) Environmental DNA Metabarcoding: A Novel Method for Biodiversity Monitoring of Marine Fish Communities. Annual Review of Marine Science 14:161\u0026ndash;185\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMiya M, Sato Y, Fukunaga T, Sado T, Poulsen JY, Sato K, Minamoto T, Yamamoto S, Yamanaka H, Araki H, Kondoh M, Iwasaki W (2015) MiFish, a set of universal PCR primers for metabarcoding environmental DNA from fishes: detection of more than 230 subtropical marine species. R Soc Open Sci 2:150088\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMontoya JM, Pimm SL, Sol\u0026eacute; RV (2006) Ecological networks and their fragility. Nature 442:259\u0026ndash;264\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMora C, Caldwell IR, Birkeland C, McManus JW (2016) Dredging in the Spratly Islands: Gaining Land but Losing Reefs. PLOS Biology 14:e1002422\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMorais RA, Depczynski M, Fulton C, Marnane M, Narvaez P, Huertas V, Brandl SJ, Bellwood DR (2020) Severe coral loss shifts energetic dynamics on a coral reef. Functional Ecology 34:1507\u0026ndash;1518\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMouillot D, Vill\u0026eacute;ger S, Parravicini V, Kulbicki M, Arias-Gonz\u0026aacute;lez JE, Bender M, Chabanet P, Floeter SR, Friedlander A, Vigliola L, Bellwood DR (2014) Functional over-redundancy and high functional vulnerability in global fish faunas on tropical reefs. Proceedings of the National Academy of Sciences 111:13757\u0026ndash;13762\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMuenzel D, Bani A, De Brauwer M, Stewart E, Djakiman C, Halwi, Purnama R, Yusuf S, Santoso P, Hukom FD, Struebig M, Jompa J, Limmon G, Dumbrell A, Beger M (2024) Combining environmental DNA and visual surveys can inform conservation planning for coral reefs. Proceedings of the National Academy of Sciences 121:e2307214121\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOksanen J, Simpson G L, Blanchet F G, Kindt R, Legendre P, O\u0026rsquo;Hara R B, Solymos P, Stevens M H H, Szoecs E, Wagner H, Barbour M, Bedward M, Bolker B, Borcard D, Borman T, Carvalho G, Chirico M, De Caceres M, Durand S, Evangelista H B A, FitzJohn R, Friendly M, Furneaux B, Hannigan G, Hill M O, Lahti L, Martino C, McGlinn D, Ouellette M H, Ribeiro Cunha E, Smith T, Stier A, Ter Braak C J F, Weedon J (2025) vegan: Community Ecology Package. R package version 2.8-0, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://vegandevs.github.io/vegan/\u003c/span\u003e\u003cspan address=\"https://vegandevs.github.io/vegan/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOl\u0026aacute;n-Gonz\u0026aacute;lez M, Reyes-Bonilla H, \u0026Aacute;lvarez-Filip L, P\u0026eacute;rez-Espa\u0026ntilde;a H, Olivier D (2020) Fish diversity divergence between tropical eastern pacific and tropical western Atlantic coral reefs. Environ Biol Fish 103:1323\u0026ndash;1341\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePang J, Ren G, Shi Q, Zhu H, Hu Y, Dong J, Ma Y (2021) Analysis of coral reef bleaching in Yongle Islands of Xisha from 2005 to 2018 based on sediment types change monitoring. MS 45:92\u0026ndash;106\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePet-Soede C, Cesar HSJ, Pet JS (1999) An economic analysis of blast fishing on Indonesian coral reefs. Environmental Conservation 26:83\u0026ndash;93\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePolanco Fern\u0026aacute;ndez A, Marques V, Fopp F, Juhel J-B, Borrero-P\u0026eacute;rez GH, Cheutin M-C, Dejean T, Gonz\u0026aacute;lez Corredor JD, Acosta-Chaparro A, Hocd\u0026eacute; R, Eme D, Maire E, Spescha M, Valentini A, Manel S, Mouillot D, Albouy C, Pellissier L (2021) Comparing environmental DNA metabarcoding and underwater visual census to monitor tropical reef fishes. Environmental DNA 3:142\u0026ndash;156\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePorebski S, Bailey LG, Baum BR (1997) Modification of a CTAB DNA extraction protocol for plants containing high polysaccharide and polyphenol components. Plant Mol Biol Rep 15:8\u0026ndash;15\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePozas-Schacre C, Casey JM, Brandl SJ, Kulbicki M, Harmelin-Vivien M, Strona G, Parravicini V (2021) Congruent trophic pathways underpin global coral reef food webs. Proceedings of the National Academy of Sciences 118:e2100966118\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePratchett MS, Berumen ML, Marnane MJ, Eagle JV, Pratchett DJ (2008) Habitat associations of juvenile versus adult butterflyfishes. Coral Reefs 27:541\u0026ndash;551\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePratchett MS, Blowes SA, Coker D, Kubacki E, Nowicki J, Hoey AS (2015) Indirect benefits of high coral cover for non-corallivorous butterflyfishes. Coral Reefs 34:665\u0026ndash;672\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePratchett MS, Coker DJ, Jones GP, Munday PL (2012) Specialization in habitat use by coral reef damselfishes and their susceptibility to habitat loss. Ecology and Evolution 2:2168\u0026ndash;2180\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePratchett MS, Hoey AS, Wilson SK (2014) Reef degradation and the loss of critical ecosystem goods and services provided by coral reef fishes. Current Opinion in Environmental Sustainability 7:37\u0026ndash;43\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQiu (2021) Checklist of the coral fish fauna of Xisha Islands, China. Biodiversity Data Journal 9:e63945\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQiu S, Liu X, Chen B, Wang Y, Liao J, Du J (2024) Reef fish diversity and community pattern of the Xisha Islands, South China Sea. MES 41:395\u0026ndash;401\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRees HC, Maddison BC, Middleditch DJ, Patmore JRM, Gough KC (2014) REVIEW: The detection of aquatic animal species using environmental DNA \u0026ndash; a review of eDNA as a survey tool in ecology. Journal of Applied Ecology 51:1450\u0026ndash;1459\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRogers A, Blanchard JL, Mumby PJ (2018a) Fisheries productivity under progressive coral reef degradation. Journal of Applied Ecology 55:1041\u0026ndash;1049\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRogers A, Blanchard JL, Newman SP, Dryden CS, Mumby PJ (2018b) High refuge availability on coral reefs increases the vulnerability of reef-associated predators to overexploitation. Ecology 99:450\u0026ndash;463\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSamoilys M, Alvarez-Filip L, Myers R, Chabanet P (2022) Diversity of Coral Reef Fishes in the Western Indian Ocean: Implications for Conservation. Diversity 14:102\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSantoso P, Setiawan F, Subhan B, Arafat D, Bengen DG, Iqbal Sani LM, Humphries AT, Madduppa H (2022) Influence of Coral Reef Rugosity on Fish Communities in Marine Reserves Around Lombok Island, Indonesia. Environ Biol Fish 105:105\u0026ndash;117\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSherman CS, Simpfendorfer CA, Pacoureau N, Matsushiba JH, Yan HF, Walls RHL, Rigby CL, VanderWright WJ, Jabado RW, Pollom RA, Carlson JK, Charvet P, Bin Ali A, Fahmi, Cheok J, Derrick DH, Herman KB, Finucci B, Eddy TD, Palomares MLD, Avalos-Castillo CG, Kinattumkara B, Blanco-Parra M-P, Dharmadi, Espinoza M, Fernando D, Haque AB, Mej\u0026iacute;a-Falla PA, Navia AF, P\u0026eacute;rez-Jim\u0026eacute;nez JC, Utzurrum J, Yuneni RR, Dulvy NK (2023) Half a century of rising extinction risk of coral reef sharks and rays. Nat Commun 14:15\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShi J, Li C, Wang T, Zhao J, Liu Y, Xiao Y (2022) Distribution Pattern of Coral Reef Fishes in China. Sustainability 14:15107\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShuting Q, Xinming L, Bin C, Yanguo W, Jianji L, Jianguo D (2022) Reef fish diversity and community pattern of the Xisha Islands,South China Sea. Haiyang Huanjing Kexue 41:395\u0026ndash;401\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStat M, Huggett MJ, Bernasconi R, DiBattista JD, Berry TE, Newman SJ, Harvey ES, Bunce M (2017) Ecosystem biomonitoring with eDNA: metabarcoding across the tree of life in a tropical marine environment. Sci Rep 7:12240\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStier AC, Chase TJ, Osenberg CW (2025) Fish services to corals: a review of how coral-associated fishes benefit corals. Coral Reefs 44:825\u0026ndash;834\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStrona G, Beck PSA, Cabeza M, Fattorini S, Guilhaumon F, Micheli F, Montano S, Ovaskainen O, Planes S, Veech JA, Parravicini V (2021) Ecological dependencies make remote reef fish communities most vulnerable to coral loss. Nat Commun 12:7282\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTkachenko KS, Hoang DT (2022) Concurrent effect of crown-of-thorns starfish outbreak and thermal anomaly of 2020 on coral reef communities of the Spratly Islands (South China Sea). Marine Ecology 43:e12717\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTylianakis JM, Lalibert\u0026eacute; E, Nielsen A, Bascompte J (2010) Conservation of species interaction networks. Biological Conservation 143:2270\u0026ndash;2279\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWen T, Xie P, Yang S, Niu G, Liu X, Ding Z, Xue C, Liu Y-X, Shen Q, Yuan J (2022) ggClusterNet: An R package for microbiome network analysis and modularity-based multiple network layouts. iMeta 1:e32\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWickham H (2016) ggplot2 Elegant Graphics for Data Analysis. Springer International Publishing, Cham\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilliams A, Althaus F, Maguire K, Green M, Untiedt C, Alderslade P, Clark MR, Bax N, Schlacher TA (2020) The Fate of Deep-Sea Coral Reefs on Seamounts in a Fishery-Seascape: What Are the Impacts, What Remains, and What Is Protected? Front Mar Sci 7:567002\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilson SK, Burgess SC, Cheal AJ, Emslie M, Fisher R, Miller I, Polunin NVC, Sweatman HPA (2008) Habitat Utilization by Coral Reef Fish: Implications for Specialists vs. Generalists in a Changing Environment. Journal of Animal Ecology 77:220\u0026ndash;228\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilson SK, Graham N a. J, Pratchett MS, Jones GP, Polunin NVC (2006) Multiple disturbances and the global degradation of coral reefs: are reef fishes at risk or resilient? Global Change Biology 12:2220\u0026ndash;2234\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWolfe K, Kenyon TM, Mumby PJ (2021) The biology and ecology of coral rubble and implications for the future of coral reefs. Coral Reefs 40:1769\u0026ndash;1806\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu W, Zhang P, Wang Q, Kang L, Su F (2024) Analysis of fishing intensity in the South China Sea based on automatic identification system data: A comparison between China and Vietnam. Marine and Coastal Fisheries 16:e10309\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYoung HS, McCauley FO, Micheli F, Dunbar RB, McCauley DJ (2024) Shortened food chain length in a fished versus unfished coral reef. Ecological Applications 34:e3002\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu C, Chang Y-C (2023) China\u0026rsquo;s Incentives and Efforts against IUU Fishing in the South China Sea. Sustainability 15:7255\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang C, Lei S, Wu H, Liao L, Wang X, Zhang L, Liu G, Wang G, Fang L, Song Z (2024) Simplified microbial network reduced microbial structure stability and soil functionality in alpine grassland along a natural aridity gradient. Soil Biology and Biochemistry 191:109366\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang SY, Speare KE, Long ZT, McKeever KA, Gyoerkoe M, Ramus AP, Mohorn Z, Akins KL, Hambridge SM, Graham NAJ, Nash KL, Selig ER, Bruno JF (2014) Is coral richness related to community resistance to and recovery from disturbance? PeerJ 2:e308\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao J, Li C, Wang T, Shi J, Song X, Liu Y (2023) Composition and Long-Term Variation Characteristics of Coral Reef Fish Species in Yongle Atoll, Xisha Islands, China. Biology 12:1062\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao J, Wang T, Li C, Shi J, Xie H, Luo L, Xiao Y, Liu Y (2024) Seven decades of transformation: evaluating the dynamics of coral reef fish communities in the Xisha Islands, South China Sea. Rev Fish Biol Fisheries 34:1261\u0026ndash;1281\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou J, Deng Y, Luo F, He Z, Tu Q, Zhi X (2010) Functional Molecular Ecological Networks. mBio 1:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1128/mbio.00169\u0026thinsp;\u0026ndash;\u0026thinsp;10\u003c/span\u003e\u003cspan address=\"10.1128/mbio.00169\u0026thinsp;\u0026ndash;\u0026thinsp;10\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"coral-reefs","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"core","sideBox":"Learn more about [Coral Reefs](http://link.springer.com/journal/338)","snPcode":"338","submissionUrl":"https://submission.nature.com/new-submission/338/3","title":"Coral Reefs","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Coral reef ecosystem, Fish community structure, Environmental DNA, Ecological network, South China Sea","lastPublishedDoi":"10.21203/rs.3.rs-9465484/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9465484/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eCoral reef fish communities are highly sensitive to changes in coral habitat condition, yet the ecological trajectories and non-linear responses of these communities remain poorly understood. In the biodiverse South China Sea, we used environmental DNA metabarcoding to profile reef-fish assemblages across sites categorized as Excellent, Moderate, or Poor in coral condition. Integrating alpha and beta diversity metrics with ecological network analysis and piecewise structural equation modeling (SEM), we quantified how fish communities shift along this degradation gradient. Moderate coral decline was associated with (i) a rise in species richness, (ii) higher network modularity, and (iii) an increase in core molecular operational taxonomic units (MOTUs). In contrast, severely degraded reefs exhibited abrupt collapses in species richness and network stability, compositional homogenization, and loss of core taxa, revealing a hysteretic, non-linear response. SEM identified coral condition as the dominant direct driver of fish community structure, mediated by habitat complexity, and as an indirect driver via fishing pressure. Intensified fishing simultaneously degraded coral (r = -0.21) and reduced species richness (r = -0.25), amplifying biodiversity loss. Our results reveal stage-dependent resilience limits beyond which reef-fish communities shift abruptly, underscoring the need for integrated management that addresses both global climate stressors and local anthropogenic impacts. Protecting coral integrity while curbing fishing pressure is essential to maintain ecosystem function and biodiversity, especially in geopolitically contested marine regions.\u003c/p\u003e","manuscriptTitle":"Environmental DNA reveals coral-driven shifts in reef fish networks: Implications for conservation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-14 08:10:08","doi":"10.21203/rs.3.rs-9465484/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"83496180539930791373502836855646026538","date":"2026-05-17T01:51:40+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"37417952617188003036113131691664741141","date":"2026-05-07T06:59:08+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-05-05T06:01:02+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-24T16:19:03+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-20T08:33:27+00:00","index":"","fulltext":""},{"type":"submitted","content":"Coral Reefs","date":"2026-04-20T01:32:51+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"coral-reefs","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"core","sideBox":"Learn more about [Coral Reefs](http://link.springer.com/journal/338)","snPcode":"338","submissionUrl":"https://submission.nature.com/new-submission/338/3","title":"Coral Reefs","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"a52f25a9-8787-4d3e-b684-a268c168a72f","owner":[],"postedDate":"May 14th, 2026","published":true,"recentEditorialEvents":[{"type":"reviewerAgreed","content":"83496180539930791373502836855646026538","date":"2026-05-17T01:51:40+00:00","index":15,"fulltext":""},{"type":"reviewerAgreed","content":"37417952617188003036113131691664741141","date":"2026-05-07T06:59:08+00:00","index":11,"fulltext":""},{"type":"reviewersInvited","content":"6","date":"2026-05-05T06:01:02+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-14T08:10:08+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-14 08:10:08","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9465484","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9465484","identity":"rs-9465484","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}