Research

INTRODUCTION

Anthropogenic climate change is already causing long-term changes in oceanographic conditions that affect marine ecosystems and fisheries production (IPCC 2023). Current scientific knowledge of climate change impacts and fisheries vulnerabilities is based on many disparate sources of information, including the monitoring and modeling analyses such as species-specific exposure studies (Nye et al. 2009, Payne et al. 2016), long-term monitoring of fisheries biomass and harvest (Gallo et al. 2022, Marshak and Link 2024), consideration of socioeconomic factors (Allison et al. 2009, Colburn et al. 2016), and predictive impacts (Cheung et al. 2010, Kleisner et al. 2017). Quantitative scientific information generated through formalized processes, such as research and/or the application of scientific methodology (e.g., monitoring programs, retrospective assessments, and predictive models), has been shown to be capable of providing robust hindcasts and forecasts and informing local to international fisheries management. However, such data may not be the only suitable approach for understanding climate impacts and informing decision-making processes (Cvitanovic et al. 2015, Lima et al. 2017). Obtaining quantitative scientific information demands long-term monitoring programs and, often, complex research models. Global assessments of fishing nations indicate that areas with higher levels of climate vulnerability frequently lack the resources or capacity to collect traditional scientific data (Allison et al. 2009, Blasiak et al. 2017). This limitation can pose a greater risk to fisheries management because of the slower response to climate change threats. Additionally, in areas with long histories of resource use and exploitation but lower capacity for data collection and monitoring, the relatively shorter historical baseline of scientific surveys may not capture directional changes and long-term variability (Ban et al. 2018). From a spatial perspective, a scale mismatch has been commonly recognized as a barrier to fisheries management, where the spatial scale of scientific surveys and modeling approaches fails to align with that of management action (Cope and Punt 2011). Furthermore, conventional research approaches that are detached from users’ needs and values often fail to adequately address the growing complexity of climate challenges, posing questions of how to reconcile the supply of scientific information with other forms of knowledge to improve effective decision-making (McNie et al. 2016).

A major recent advance is the adoption of more collaborative approaches that acknowledge and integrate knowledge of multiple disciplines and users, including scientific information, institutional expert knowledge, and local fishermen’s knowledge, as part of the assessment and management of fisheries under climate change (Carroll et al. 2023, Mills et al. 2023, Cannon et al. 2024). In the context of this study, institutional expert knowledge (hereafter referred to as expert knowledge) refers to ecological or socioeconomic knowledge held by fisheries experts (i.e., managers, policy-makers, researchers, and NGO representatives), accumulated through their own management or research experience, communications with colleagues and stakeholders, and personal knowledge and experience of climate change impacts and vulnerabilities (Teck et al. 2010, Marvin et al. 2020). Local fishermen’s knowledge (hereafter referred to as local knowledge) refers to ecological knowledge held by place-based fishing communities, derived from their on-the-water observations, intergenerational experience, and personal perceptions about the socioeconomic status of their businesses and livelihoods (Lima et al. 2017, Ban et al. 2018).

Expert and local knowledges share commonalities with scientific information by recognizing the interconnection of living and physical entities and striving to understand environmental, ecological, and human drivers that influence the abundance and distribution of species (Ban et al. 2018). They also provide valuable complementary insights to scientific data, addressing gaps in biological, socioeconomic, and management information. For instance, experts and local resource users can offer long historical baselines and rich social-ecological knowledge within specific cultural contexts (Gauvreau et al. 2017, Ban et al. 2018). Such engagement with stakeholders in knowledge production can further foster partnerships between knowledge producers and users, enhancing the uptake of ocean science across the science–policy interface (Pendleton et al. 2023). Over recent years, these benefits have catalyzed extensive discussions around producing useful information and integrating diverse knowledge systems to advance climate-ready fisheries science and management in the literature and in practice (Lomonico et al. 2021, Mills et al. 2023).

What is reflected in action is the development of participatory approaches for climate vulnerability assessments (CVAs). Vulnerability is defined as “the degree to which a system is susceptible to, and unable to cope with, adverse effects of climate change, including climate variability and extremes” (IPCC 2007). The flexibility of CVA frameworks, which can accommodate both qualitative and quantitative data, has enabled their widespread application using various knowledge types in the assessment process. For example, Boyce et al. (2022) used global marine species distribution data and biological trait information (e.g., maximum body length and thermal safety margin) to assess species’ climate risk. Hare et al. (2016) employed expert knowledge to compare and rank species’ relative vulnerability in the Northeast U.S. Continental Shelf Large Marine Ecosystem. Macusi et al. (2021) took a bottom-up participatory approach to understand fishers’ perceived impacts of climate change and the vulnerability of small-scale fisheries in the Philippines. In particular, collaborative and participatory approaches have gained increasing popularity since 2012 (Li et al. 2023). Among 65 studies on fishery CVAs, over half have integrated expert and local knowledge, with nearly 30% combining multiple forms of knowledge (Li et al. 2023). This trend underscores the growing recognition of the value of complementary information in improving research and management outcomes.

Nonetheless, most of these participatory studies tend to broadly describe participants’ perspectives on species or socioeconomic vulnerability (e.g., Gómez Murciano et al. 2021, Macusi et al. 2021), and others utilize scientific information as background material or additional context to support expert or fishermen’s assessments (e.g., Hare et al. 2016, Carrol et al. 2023). Much work remains to fully understand the utility and limitations of various knowledge forms. In the context of climate-ready fisheries management, three key questions arise: (1) What are the strengths and weaknesses of different forms of knowledge? (2) What are the potential pitfalls of relying solely on one form? (3) What are the possible solutions to synergize multiple knowledge systems to facilitate equitable and effective adaptation to climate change?

To address these questions, we used China as a case study that is situated within an interesting fisheries decision-making context. As the world’s largest seafood producer and consumer, China faces challenges in effectively managing and conserving its fisheries resources amid growing anthropogenic impacts, demands, and changing ocean conditions (FAO 2024, Li et al. 2024). The country has developed and implemented a suite of fisheries management tools to combat these challenges, including input and output controls. Decision-making has traditionally followed a top-down mechanism reliant on a command-and-control approach (Shen and Heino 2014, Su et al. 2020). Marine policies are formulated at the national level and subsequently implemented through provincial, municipal, and county administrations. Historically, expert knowledge has played a central role in framing environmental problems and proposing management solutions, such as determining fishery closure seasons and locations, through a so-called “brainstorming-type of decision-making mode” (Hu 2013:633). This process involves a broad range of experts from academia, think tanks, NGOs, and government bureaucracies working to build consensus on policy directions. Although experts contribute domain-specific knowledge and act as science arbiters and credible sources of policy legitimization, concerns persist regarding the adequacy and accountability of their judgment, particularly when addressing complex challenges (Shen et al. 2022). Recognizing these limitations, China’s recent fisheries management reforms have emphasized the integration of cutting-edge science and stakeholder engagement to bridge the information gap. Examples include local pilot fisheries testing new monitoring techniques, such as electronic monitoring and port-based data collection, to improve fishing data quality and quantity (Zhu et al. 2021) as well as the establishment of statutory mechanisms that incorporate expert involvement, risk assessment, and public participation in decision-making processes (Su et al. 2022).

Thus, unlike most previous studies that focused on presenting the results of participatory CVAs, our study aimed to investigate how scientific data, expert knowledge, and local knowledge converge or diverge in assessing vulnerability in fisheries social-ecological systems under climate change. We tested two hypotheses: (1) data-driven and knowledge-driven approaches (i.e., scientific data versus expert and local knowledge) may yield different results in fishery CVAs, and (2) these inconsistencies can be attributed to factors such as measurement indicators, information sources, data quality, or perception bias.

To explore these hypotheses, we compiled and collated scientific data, expert knowledge, and local perceptions of fishermen regarding species and socioeconomic vulnerability to climate change through desktop research, expert questionnaires, and fishermen interviews from fisheries across China. By comparing these data, we identified key differences, their contributing factors, and their effects on the results of CVAs. The results of our study provide key lessons for future knowledge production efforts in CVAs and add to the knowledge base regarding strengths and weaknesses of different forms of knowledge for enhancing resilience within fisheries management.

MATERIALS AND METHODS

Definition of ecological and socioeconomic vulnerability

We adopted the definition of vulnerability from the IPCC Fifth Assessment Report (AR5), which describes vulnerability as an internal property of a system, determined by its sensitivity and adaptive capacity (Lavell et al. 2012). In the context of fisheries social-ecological systems, vulnerability can be used to assess and characterize the state of species (ecological dimension) and industries/societies/fishing communities (socioeconomic dimension) under climate change impacts (Metcalf et al. 2015). Accordingly, we further defined vulnerability in these two dimensions. Specifically, ecological vulnerability refers to the vulnerability of species or taxa, influenced by sensitivity traits that determine the extent to which they are affected by climate variability or change, and adaptive capacity traits that reflect their ability to adjust to such changes. Socioeconomic vulnerability pertains to the attributes of a social system that influence its capacity to anticipate or respond to climate-induced changes as well as its ability to minimize, cope with, and recover from their consequences.

Data-driven approach

The data-driven approach broadly employed a systematic literature review and database searching methods to collate readily available data and information to characterize ecological and socioeconomic vulnerability. These data were compiled and analyzed during 2022–2023, independent of expert or stakeholder perceptions.

The methodological details and results were previously described and discussed in Li et al. (2024). Briefly, in this data-based assessment, we focused on China’s domestic marine fisheries on a national level. The assessment consisted of two components. The ecological vulnerability assessment focused on 28 taxonomic groups, which were selected based on their commercial importance (contributed to more than 80% of total landings), expert opinions, and data availability. We used a trait-based approach from Hare et al. (2016) and applied 11 categories of biological traits identified from a global review of vulnerability indicators (Li et al. 2023). The sensitivity traits encompassed (1) prey specificity, (2) population growth, (3) early life stage, (4) habitat and environment, (5) stock status, and (6) sensitivity to acidification. The adaptive capacity traits encompassed (1) reproduction, (2) larval dispersal, (3) adult mobility, (4) conservation, and (5) stock enhancement. For each taxon, we searched for numerical values and biological information relevant to the identified traits. Data sources included peer-reviewed literature in both English and Chinese, and online databases such as FishBase (https://www.fishbase.org), SeaLifeBase (https://www.sealifebase.se/search.php), and Sea Around Us (https://www.seaaroundus.org). The scoring process followed the protocol and criteria developed by previous studies (Morrison et al. 2015, Li et al. 2024), using a semi-quantitative approach to score sensitivity and adaptive capacity traits for each taxon on a scale of 1 (low sensitivity or adaptive capacity) to 4 (very high sensitivity or adaptive capacity; Appendix 1). The scores were then combined across biological traits based on the following equation to derive a relative taxon-specific climate vulnerability score (Vul_t).

(1)

where (ES_t,i) and (EA_t,i) represent the scores of ecological sensitivity and adaptive capacity for each trait i for taxon t, and n is the total number of traits.

During the literature review process, we also documented data quality to demonstrate data uncertainty and gaps. A score of 0 was assigned to the traits with no available data, 0.5 indicated that data were collected from a limited number of sources and/or from sources of limited reliability (e.g., only one data source or the study area is beyond the China Seas), and 1 represented data of both high quality and strong reliability (i.e., validated by multiple sources).

The other component of data-driven vulnerability assessment focused on the socioeconomic dimension of fisheries. This assessment was conducted on a provincial level to align with China’s governance and administration structure. A total of 11 provinces were evaluated, covering all the coastal regions in Mainland China (Fig. 1A). Like ecological vulnerability, we adopted a trait-based approach to measure socioeconomic sensitivity and adaptive capacity of coastal regions to potential changes in fisheries resources induced by climate change impacts (Li et al. 2024). The socioeconomic sensitivity traits (or indicators) encompassed (1) fishery dependence, (2) economic dependence, (3) food dependence, and (4) infrastructure. The socioeconomic adaptive capacity indicators encompassed (1) governance, (2) learning, (3) assets, (4) social organization, and (5) flexibility (Appendix 1). For each province, we searched for numerical values and socioeconomic information relevant to the identified indicators. Data sources included publicly available reports and published social and economic statistics and census data. Based on this information, we scored socioeconomic indicators for each province based on predefined scoring criteria (Li et al. 2024, Appendix 1). This process generated a relative score ranging from 1 (low sensitivity or adaptive capacity) to 4 (very high sensitivity or adaptive capacity). The scores were then combined across socioeconomic indicators based on the following equation to derive a relative province-specific climate vulnerability score (Vul_p).

(2)

where (SS_p,i) and (SA_p,i) represent the scores of socioeconomic sensitivity and adaptive capacity for each indicator i for province p, and n is the total number of indicators.

Expert knowledge-driven approach

We conducted an online questionnaire to collate expert knowledge and perceptions about climate vulnerability in China’s fisheries. The identification of expert participants was built from the China Fisheries Learning Network, a program established to facilitate knowledge sharing and cross-institution collaboration to support China’s fisheries management. As of August 2023, 326 individuals had joined the Network, hailing from diverse organizations including management agencies, universities, research institutions, and NGOs. We leveraged the Network member list and adopted a two-stage quota sampling approach to identify expert participants. Initially, we divided the population into three subpopulations based on their working regions (i.e., Bohai and Yellow Sea, East China Sea, and South China Sea). Within each stratum, we further categorized the subpopulations into groups of managers, scientists, and NGO representatives and used non-probability sampling techniques to select samples within each group. This process generated a distribution list of 89 individuals. Email invitations and questionnaires were sent out in August 2023. A total of 33 participants (37% response rate) were recruited for this survey.

The expert questionnaire was made up of four parts: (1) demographic information, (2) general perceptions about climate change impacts on fisheries, (3) scoring of ecological vulnerability, and (4) scoring of socioeconomic vulnerability (Appendix 1). To ensure comparability between data-driven and knowledge-driven approaches, we applied the same assessment unit and scale for the expert survey. Specifically, experts were asked to score ecological vulnerability for 28 pre-identified taxonomic groups and socioeconomic vulnerability for the 11 coastal provinces included in the data-driven approach. Importantly, during the scoring process, we provided the definition of vulnerability but did not prescribe specific indicators or criteria to guide the experts’ assessment. Instead, we asked them to explain the attributes and criteria they considered when assigning scores (n = 17 provided scoring criteria for taxon vulnerability, and n = 11 provided scoring criteria for socioeconomic vulnerability). This approach allowed for a comparison of the factors influencing the results across different methodologies and helped enhance understanding of the utility and nuances of expert knowledge in CVAs.

Furthermore, we asked participants to rate their confidence in scoring, as a potential factor to be compared with data quality, offering insights into gaps between scientific data and expert knowledge. We also inquired about the information sources that supported their scoring and any additional resources or guidance they felt would enhance their assessment. This information was used to help identify areas for improving data accessibility and refining assessment processes to support more effective expert engagement in future CVA practices.

Local knowledge-driven approach

To further explore local fishermen’s perceptions and traditional knowledge about climate change impacts, we designed and conducted semi-structured interviews that specifically targeted fishermen. The rationale for this approach lies in the unique insights that local fishermen can provide for identifying on-the-ground vulnerabilities, which are often overlooked in data-driven or expert-driven assessments. Given the time and effort required for in-person interviews, we focused on fishing communities in Shandong Province—a coastal region surrounded by the Bohai and Yellow Sea, identified as high climate risk in our data-based approach (Li et al. 2024). By prioritizing fishermen in this region, we sought to understand how those most affected perceive climate change impacts and bridge the gap between national and local perspectives.

The recruitment process used a snowball sampling approach, starting with an initial set of participants who had participated in previous social surveys. These initial participants were then asked to refer others they knew who met the survey’s inclusion criteria (i.e., they must be adults 18 years of age or older and must have experience in the fishing industry). The semi-structured interviews were conducted in seven fishing ports during August 2023, resulting in the recruitment of 24 participants (Fig. 1B).

The interview questions were structured similarly to the expert surveys to enable a comparative analysis (Appendix 1). However, because of the difference in scale between the two surveys (national versus provincial) and the fact that fishermen tend to hold place-based knowledge and may have low literacy on academic jargon (e.g., the concept and terminology of vulnerability), we adjusted the questions to emphasize local conditions and observations. For instance, to assess ecological vulnerability, we asked fishermen about observed changes in their target species. For the socioeconomic dimension, we focused on individual- and household-level sensitivity and adaptive capacity such as their reliance on fishing for livelihoods, flexibility in shifting fishing locations or occupations, and access to resources and information for adaptation. Both the expert and fishermen surveys were approved by the Institutional Review Board of the Office of Research Compliance at Stony Brook University (Approval Number: IRB2023-00268).

Comparison between approaches

Given the epistemological nature of the study, we explored how these diverse approaches—scientific data, institutional expert knowledge, and local fishermen’s knowledge—generate, interpret, and apply information differently in the context of fisheries CVAs. We compared ecological and socioeconomic vulnerability assessment results derived from data- and expert-driven approaches, given their comparable scales (both national), and then used fishermen's local knowledge at a finer scale to complement these comparisons.

First, we quantitatively compared taxon vulnerability scores derived from scientific data with those from the expert questionnaire. Prior to analysis, scores were standardized to a scale of zero to three. To identify taxon assemblages that may show distinct patterns between different approaches, we applied a k-means clustering algorithm to the dataset combining both the average data-derived and expert scores. The optimal number of clusters was determined based on the elbow method that identified clusters based on the total within-cluster sum of squares (Kaufman and Rousseeuw 2009). Once the cluster count was established, the k-means clustering algorithm was applied. An ANOVA analysis was conducted to assess the significance of differentiation among the identified assemblages.

To explore potential reasons for these differences, we examined the indicators and criteria used to assess ecological and socioeconomic vulnerability across approaches. We extracted relevant information from expert responses and compared it with the indicators and scoring criteria applied in the data-driven approach to identify their commonalities and differences. Additionally, a linear regression analysis was performed to test the hypothesis that landing volume may influence experts’ scoring (i.e., taxa and provinces with higher landings are more likely to be rated as highly vulnerable by experts because of their greater economic importance, which may attract more expert attention and familiarity). Furthermore, in assessing ecological vulnerability, we compared the quality of scientific data and the confidence levels of expert knowledge for each taxon, aiming to examine how uncertainties in input information might affect CVA outcomes.

Finally, local knowledge from fishermen was compared against the other two data sources. Biological information and observed changes in target species were qualitatively analyzed and linked to the relevant taxa assessed through national-level approaches. We first ranked species based on their frequency as reported target species by fishermen and then created information sheets for each species, summarizing trends in their status and observed changes. By linking these target species to the previously identified cluster groups, we compared fishermen’s observations with vulnerability scores derived from the data-driven and expert assessments and evaluated whether local knowledge could address existing knowledge gaps. For socioeconomic vulnerability, fishermen’s responses were categorized into nine variables in line with the structure used in the data-driven approach. This information was then qualitatively compared to the socioeconomic vulnerability results derived from the other approaches.

RESULTS

Demographics of survey participants

The expert participants represented three ocean regions around China with a relatively even distribution (Fig. 2). The majority were affiliated with universities and research institutions (73%), followed by those working at management agencies (21%) and NGOs (6%). Career stages spanned from early career (27%) to senior career (30%), with the largest proportion at the mid-career stage (43%). The survey captured a wide range of expertise, including natural sciences (e.g., biology, ecology, stock assessment conservation and restoration) and social sciences (e.g., fisheries management and policy, resource economics). Notably, participants with natural science expertise and academic or research backgrounds dominated the sample because the field typically emphasizes the ecological aspect of fisheries and there exists a larger pool of researchers compared to those affiliated with management agencies and NGOs.

The fishermen participants recruited were all male and had extensive experience in the fishing industry: 12% had 10–20 years of experience, 46% had 21–30 years, and 42% had been involved for more than 30 years (Fig. 2). Although the interviewees were all male, it is important to note that women play critical but often overlooked roles in the fisheries sector in the region, particularly in fish processing, trading, and other post-harvest activities along the supply chain. In addition, fishermen’s demographic profiles showed a relatively low education level, with only 10% holding education beyond junior high school. The household size ranged from medium (50% had households of three members) to large (37% had households of four–six members). While this measure reflected formal education, we acknowledge that fishermen may possess extensive informal or experiential knowledge relevant to adaptation.

Ecological vulnerability

By comparing taxon vulnerability scores derived from data-driven and expert knowledge-driven approaches, we identified four clusters (p < 0.05, Fig. 3). Cluster 1 represented taxa with low-to-moderate vulnerability scores in both approaches, indicating consistent results. These taxa contributed to 0.87–3.1% of total landings and shared biological traits such as high production and mobility. Cluster 2 represented taxa rated higher in the data-driven approach compared to expert assessments. These taxa had low contributions to total landings (0.64–2.46%) and included diverse groups such as pelagic species (e.g., sand lance), cephalopods (e.g. cuttlefish and octopus), crustaceans (e.g., Scylla and Southern rough shrimp), and benthic or demersal fish (e.g., grouper and spiny head croaker). Cluster 3 showed the largest difference, where expert scores were significantly higher compared to data-driven scores. Two taxa were included, anchovy and sardine, contributing to 6.62% and 1.08% of total landings, respectively. Cluster 4 represented the taxa that received moderately higher expert scores than data-driven scores. These included high-yielding taxa, contributing to 1.47–8.87% of total landings, such as hairtail, scad, and swimming crab, which experts recognized as moderately vulnerable, while scientific data indicated low vulnerability. A regression analysis of the relationship between taxa landing volumes and score differences showed a positive and significant correlation (p < 0.05; Fig. A1.1).

To further investigate factors contributing to these differences, we compared traits considered in the two assessment approaches (Fig. 4). The data-driven approach identified early life stage survival and settlement requirements, stock status, and larval dispersal as key traits influencing vulnerability. In contrast, expert assessments focused more on population growth and habitat specificity, with stock status being the criterion showing the greatest overlap between the two approaches. We also found that experts often diverged in their scoring criteria and interpretations. For instance, some experts argued that high population growth rates indicate greater vulnerability, particularly for pelagic fish such as anchovy and sardine, whereas the data-driven approach considered high population growth rates as a sign of resilience to climate change because of rapid recovery and reproduction. Mobility also elicited differing views. Experts suggested that highly mobile species might be more susceptible to warming temperatures for their broader geographic range, while the data-driven approach, based on the IPCC framework, considered this factor under the exposure domain, which describes the spatial presence of species that could be adversely affected. Instead, mobility was classified as an adaptive trait indicating greater resilience given the species’ ability to move to favorable areas in response to changing environmental conditions.

Expert knowledge also differed in its integration of diverse information sources. Experts drew on personal observations, scientific research, and fisheries-related datasets (Fig. 5). Meanwhile, they highlighted several gaps that hinder their assessments (Fig. 5). For instance, access to fisheries monitoring, socioeconomics, and climate impacts datasets was often limited because of institutional silos. Data gaps also persisted across spatiotemporal scales. Most experts did not directly engage with fishermen and emphasized the importance of incorporating local fishermen’s knowledge into assessments. Additionally, there was a lack of familiarity among participants with the concept of vulnerability and associated methodologies that limited the consistency and accuracy of their assessments.

Uncertainty in data quality and expert knowledge was another significant factor influencing differences in results (Fig. 6). When comparing data quality and expert confidence levels across taxa, four distinct patterns emerged: Cluster A showed higher expert confidence than data quality. Cluster C showed the opposite trend with higher data quality than expert confidence. Clusters B and D showed closer alignment between data quality and expert confidence, but differing overall certainty levels—Cluster D showed low-to-moderate certainty, while Cluster B showed moderate-to-high certainty. Linking the differences in taxon vulnerability scores to information uncertainty revealed further insights (Fig. 6). Taxa in Cluster 4 (Fig. 3), characterized by higher expert scores than data scores for climate vulnerability, consistently showed moderate levels of expert confidence, despite variability in data quality (Fig. 6). A positive correlation was identified between the differences in vulnerability scores and the differences in expert knowledge and data quality (p < 0.05), indicating that greater confidence in expert knowledge of a taxon contributed to a higher weighting of that taxon’s climate vulnerability.

Lastly, we incorporated local fishermen’s knowledge to complement the assessment (Table 1). Survey results identified crustaceans (e.g., swimming crab, mantis shrimp) and cephalopods (e.g., octopus) as primary target taxa in the region. Interestingly, these taxa had low-to-moderate data quality and expert confidence in previous assessments. By linking fishermen’s observations to expert and data scores, we found that most target taxa were rated as having low or moderate vulnerabilities, with one exception: sand lance. All fishermen noted sand lance as a declining species, aligning with its high vulnerability in the data-driven approach, though experts rated it as moderately vulnerable.

Fishermen provided nuanced insights into the trends in target species, including declines in traditional fisheries resources (e.g., croakers, Penaeus, mackerel) and the rise of emerging crustacean fisheries (Table 1). They also reported notable changes in catch, fish size, ex-vessel prices, and fishing locations. While there were varying opinions about those changes, the majority observed a declining or fluctuating trend in population and expressed concerns about growing uncertainty in the industry. However, they remain unsure whether changes were driven by climate change impacts or overexploitation.

In addition, fishermen identified local species that emerged as increasingly abundant catches in the area but were not included in the data-driven approach, such as Pholis fangi (a gunnel fish noted as having high resilience and low fishing vulnerability in FishBase) and yellow goosefish (a demersal fish with low resilience and very high fishing vulnerability, per FishBase; Table 1).

Socioeconomic vulnerability

The comparison of expert and data scores for socioeconomic vulnerability showed four distinct patterns (Fig. 7). Cluster I had only one province, Hainan, rated as highly vulnerable in the data-driven approach but considered moderately vulnerable by experts. In Cluster II, provinces were rated as highly vulnerable by experts but low to moderate in socioeconomic datasets. This cluster included four provinces, Shandong, Guangdong, Zhejiang, and Liaoning—primarily high-landing regions. Cluster III showed consistent results between the two methods, with four provinces (Guangxi, Jiangsu, Tianjin, and Shanghai) receiving consistently low vulnerability scores. Hebei and Fujian were classified into Cluster IV, with moderate data-driven scores and low expert scores. A regression analysis was conducted to examine the relationship between landing volume in each province and the score difference (Fig. A1.2). The results indicated no significant relationship.

When examining the methods and indicators considered in scoring, we found that only a few socioeconomic variables, such as fishery dependence, infrastructure, and flexibility, were used by experts (Fig. 8). Interestingly, experts also considered biophysical indicators when assessing socioeconomic vulnerability. For instance, geographic features were often cited, with participants suggesting that provinces in semi-closed seas might be more vulnerable because fish populations could struggle to migrate to suitable areas under changing ocean conditions. This comparison revealed significant differences in assessment indicators applied by data-driven versus expert-driven approaches.

Local fishermen’s perspectives further corroborated some of the experts’ assessments, particularly regarding fishery dependence in Shandong Province (Table 2). Nearly 90% of fishermen stated that more than half of their annual household revenue depended on fishing, with 24% indicating no alternative income sources. Approximately 50% of the interviewed fishermen expressed unwillingness to change fishing locations or gear, even if their target species declined or their business became less profitable. Based on these variables, their adaptive flexibility appeared limited.

On a positive note, fishermen reported access to multiple resources and social networks that could support their adaptation efforts (Table 2). All interviewees indicated they had various information sources for fishing areas and weather conditions and were active members of local fishing associations, suggesting strong social cohesion and community interactions. However, fishermen expressed concerns about reduced fishing hours and increased costs due to extreme weather events. They also identified several areas needing support to improve their resilience, including technical assistance and training on vessel maintenance and fishing techniques, financial support for social security, and policy measures to diversify and boost household income.

DISCUSSION

There is growing recognition of the complementary nature of multiple knowledge systems and the importance of incorporating them into CVAs and other tools for fisheries management. However, the literature on knowledge integration has often depicted it as straightforward endeavor rather than a complex and dynamic process with multiple interpretations and large uncertainties (Fazey et al. 2014, Cvitanovic et al. 2015). Consequently, while many theoretical frameworks, such as participatory CVA approaches, provide a promising mechanism for improving knowledge flow and addressing information gaps, they seldom address the fundamental differences between knowledge systems that hinder the process, nor do they provide insights into how synergies in knowledge production can be effectively and iteratively improved (Cvitanovic et al. 2015). Our findings highlight the critical utilities and limitations of various knowledge systems as well as the factors that influence the effectiveness and efficiency of knowledge production for fishery CVAs.

Differences in results and potential explanations

Our results support the hypothesis that data-driven and knowledge-driven approaches can yield different outcomes in assessing climate vulnerability of fisheries. These differences were particularly evident in taxon-level vulnerability assessments. Experts tended to assign higher vulnerability scores to pelagic fish and high-landing taxa, such as scad, mackerel, squid, chub mackerel, swimming crab, Acetes, Penaeus, and hairtail. In contrast, fishermen had mixed opinions about the vulnerability of high-landing taxa in their local waters. For example, swimming crab, mantis shrimp, and hairtail were perceived as more resilient than mackerel and croaker. Moreover, scientific data suggested that low-landing taxa, such as cuttlefish, croaker, and leatherjacket, were more vulnerable.

Although inconsistencies in the scale and locations of assessments between the local knowledge-driven approach and the other two methods limit direct comparability, we included fishermen from high-risk regions to evaluate whether the results derived from scientific information and expert assessments align with the observations from those most impacted. The findings show a divergence in results, indicating the potential issue of interpreting broad-scale CVA outcomes to inform local adaptation and decision-making. This discrepancy also underscores the importance of engaging local knowledge to validate findings and provide contextualized interpretations for more effective management strategies.

We identified four factors that contributed to the differences observed in vulnerability scores. First, the significant positive relationship between taxa landing volume and score difference suggests that experts tend to assign higher vulnerability scores to taxa with higher landings, likely because of their greater familiarity and research attention devoted to economically important fisheries resources. This underscores the need for careful interpretation of expert assessments, especially for the less-studied species that may hold significant social or cultural importance. Second, there was a notable divergence between the subjective criteria used by experts and the structured, objective criteria applied in the data-driven approach. Experts’ assessments often emphasized population growth and habitat or environmental conditions, likely reflecting their expertise in fish biology, ecology, and stock assessment, or local species characteristics. However, this may limit the transferability of CVAs to other areas. Conversely, the data-driven approach integrated diverse indicators and methods from a comprehensive literature review, offering a more holistic perspective and enabling reproducibility and cross-regional comparisons. Third, we observed various knowledge gaps by comparing different datasets. For example, the adaptive nature of human knowledge enabled experts and fishermen to synthesize information from multiple sources and across multiple generations more easily than the data-driven approach. This capacity allowed their assessments to incorporate broader insights from stakeholders beyond the confines of singular datasets. On the other hand, limitations to accessing certain datasets have limited people’s understanding of fisheries vulnerabilities to climate change impacts. We find these gaps are particularly pronounced at the ecosystem, fishery, and community levels. Finally, inconsistent data quality and underlying uncertainties in knowledge influenced results. For instance, the spiny head croaker was assessed differently by experts compared to the data-driven approach, with the discrepancy stemming from variations in uncertainty levels of data quality and expert confidence.

In addition to taxon-specific differences, notable discrepancies were observed in the socioeconomic vulnerability assessments of coastal provinces, particularly in high-landing provinces, such as Zhejiang, Shandong, Guangdong, Liaoning, and Hainan. Although landing volume was not a significant factor differentiating expert and data-derived scores, the findings revealed fundamentally different scoring criteria across approaches. Experts’ assessments relied heavily on biophysical traits and appeared to conflate ecological and socioeconomic components, which diverge from traditional CVA frameworks that separate these steps (Metcalf et al. 2015, Payne et al. 2021). This blending may be attributed to the imbalance of expertise among survey participants; the majority specialized in biology or stock assessment, whereas fewer than 5% had backgrounds in social science or economics. The unfamiliarity of most participants (~ 88%) with fishery CVAs and their assessment processes seems to have further exacerbated the discrepancies.

Strengths and weaknesses of knowledge systems

This study shows the distinct contributions of various knowledge systems in informing climate vulnerability of fisheries social-ecological systems. Building on prior theoretical advancements on the differences and complementarities among scientific data, expert, and local knowledge, as well as our empirical analysis, we further summarize their respective strengths and weaknesses in the context of climate-ready fisheries (Fig. 9).

Scientific data

The data-driven approach offers significant advantages, particularly in providing a robust, structured, and repeatable assessment supported by a wide array of datasets. These can include biological traits, documented climate change impacts, and socioeconomic data that describe human dimensions of vulnerabilities at varying scales. By synthesizing existing resources and leveraging established methodologies, this approach ensures good objectivity, cross-region transferability, and precise categorization in the assessment process, with indicators and criteria tested in previous studies.

However, caveats are obvious. Many databases operate at large spatial scales that may not fully reflect local specificity, which is critical for effective local assessments and adaptation (Cordier et al. 2024). This issue is compounded by outdated data in some sources. For instance, the yellow goosefish, a local species identified as of growing commercial and ecological importance in the Yellow Sea, demonstrated contrasting vulnerability interpretations. While FishBase, referencing Cheung et al. (2005), classified it as highly vulnerable based on global-scale assessments, recent studies focused on the Yellow Sea highlighted its high recovery potential and resilience to fishing and environmental changes (Sun et al. 2021). Additionally, datasets collected through local fisheries monitoring programs often require substantial resources and effort, which can be expensive and labor-intensive. Such constraints can compromise data quality and quantity, further limiting the evaluation of critical indicators such as early life history traits, which emerges as a significant data gap in this study.

Institutional expert knowledge

Experts often act as science arbiters and offer knowledge that is cost-effective and reliable and that could address some of the gaps in scientific data (Shen et al. 2022). They possess deep knowledge of certain species and areas, particularly those of high economic importance. Collecting expert knowledge can be low cost, which can be useful for decision-making that requires rapid assessment (Teck et al. 2010). Yet, this approach may require some time and coordination for meaningful engagement. Furthermore, expert knowledge can synthesize diverse information sources and can quickly identify data gaps and new information in the empirical research domain.

Nevertheless, there are important limitations to relying solely on expert knowledge. First, experts’ judgment can be highly influenced by individual expertise, subjective interpretations, and information silos, leading to variability and uncertainties in assessments. This perception bias is also a key barrier to effective knowledge exchange and production among experts and decision-makers (Cvitanovic et al. 2015). In this study, both ecological and socioeconomic vulnerability assessments leaned heavily on biological factors, reflecting the predominance of fisheries biologists among participants. Furthermore, people may take very different approaches to scoring and distinct interpretations of climate vulnerability. The lack of clear goals, standardized guidelines, and ground rules can limit transparency, reproducibility, and transferability of assessments and, ultimately, may render CVAs less useful for management.

Local fishermen’s knowledge

Fishermen’s knowledge provides invaluable frontline experience and observations of local biophysical changes, characterized by a high degree of responsiveness because of their daily, on-the-water experiences (Ban et al. 2018, Carroll et al. 2023). This feature also helps support rapid adaptation to complex and urgent crises like climate change (Ford et al. 2015). Incorporating fishermen’s decades of observations on declining and emerging fisheries allows us to gain a more nuanced, contextualized perspective than the other two approaches. We see particular value in fishermen’s insights for understanding individual and household-level socioeconomic vulnerabilities. These insights can help fill data and knowledge gaps identified in other approaches and enable the development of targeted and tailored strategies to support local adaptation.

However, fishermen's knowledge also presents several limitations and considerations. Engagement and trust-building could require significant time and resources (Fleming et al. 2020). Although local knowledge excels in capturing past and present experiences especially when fishermen have been involved in the industry for a long time, it is less effective at extrapolating these observations to predict outcomes under future conditions. Its inherent place-based, value-laden nature poses challenges for generalization and broader-scale applications. An example from our research is the fact that fishermen from different ports voiced different opinions of the status and vulnerabilities of the same species. These differences may reflect actual local variations or simply differing perceptions. This divergence has also made it challenging for direct comparison with the two other knowledge systems. As noted in previous research, the qualitative and descriptive features of local and Indigenous knowledge complicate systematic comparison and validation with other quantitative forms of knowledge (Jones et al. 2024). Our semi-quantitative methods that incorporated local knowledge as supplementary information present a possible solution to this problem. However, such approaches may still imply that local knowledge is secondary and less valid. Future research should focus on developing techniques that recognize and integrate the equal value of these fundamentally different ways of knowing (Jones et al. 2024).

Implications for incorporating multiple knowledge forms in CVAs

Decades of research on the role of science in transformative fisheries policy have demonstrated that the most successful scientific programs are collaborative, incorporating various stakeholders and knowledge sources to identify challenges and co-develop solutions (Singh et al. 2021). Incorporating the strengths and weakness of different knowledge systems with the lessons we learned from our approaches enables us to discuss strategies for generating more meaningful and actionable knowledge for fishery CVAs. It is important to identify solutions at various phases of the CVA process that can enhance synergies among multiple knowledge forms toward the goals of improving CVA science and supporting climate-ready fisheries management.

Co-design of CVAs

This phase involves scientists and stakeholders collaboratively defining the CVA approach and the roles of various knowledge forms. The process should address fundamental questions such as which types of knowledge will be incorporated, what methods are to be taken to collect this information, and at which stages knowledge holders will be engaged.

The integration of diverse knowledge forms should occur early in the CVA process and be goal-oriented (Mistry and Berardi 2016, Norström et al. 2020, Mason et al. 2023). Given the distinct utilities of scientific data, expert knowledge, and local knowledge, CVA practitioners should identify which information aligns with the specific objectives and scope of research or management context, whether at the national or local level or for forward- or backward-looking analyses. Early engagement should also clarify the role of the expert or local knowledge at different stages of CVA processes. For example, interviews with fishermen can provide local insights into emerging species of concern that may not be captured by experts or national datasets, as was the case in our study. Similarly, experts’ institutional knowledge and their regional expertise can identify available literature, filter credible sources of information, or weigh factors for assessments, complementing data-driven approaches. In addition, methods for incorporating multiple knowledge forms can vary greatly, leading to different uses and interpretations of information (Marvin et al. 2020). In previous studies, experts were often engaged during the scoring stage, where they evaluated datasets and reports to directly inform decisions (Hare et al. 2016). In contrast, our study adopted a comparative approach to evaluate consistency across data- and knowledge-driven methods, which could help policymakers identify the gaps, biases, and uncertainties in different knowledge sources before translating them into management decisions.

Knowledge production

During this phase, participants contribute information and knowledge to support the assessment. To ensure meaningful engagement and robust evaluation, we make the following recommendations based on our findings.

The inclusion of appropriate and representative participants is key to the success of participatory CVAs. While most studies have emphasized broad stakeholder integration (Cooke et al. 2021), our findings reveal the importance of a balanced representation of expertise alongside diversity. In fisheries, a field traditionally dominated by natural sciences and male participants, including marginalized groups and underrepresented expertise in participatory approaches would greatly enhance the reliability, representativeness, and equity of CVAs.

Clear technical guidelines and structured, sequential processes are crucial components of knowledge production (Mason et al. 2023). Our approach intentionally avoided standardized methods for scoring indicators and criteria, instead encouraging participants to share their perspectives. While this method allowed for a comparative analysis of data- and knowledge-driven criteria, it also introduced variability that rendered some outcomes less applicable to CVAs and management. Recognizing this limitation, we suggest future CVAs should incorporate clearly defined, context-specific guidelines tailored to the language and needs of different stakeholder groups, such as clear demonstrations of scoring criteria for scientists, policymakers, and fishermen. Interactive processes that facilitate learning and discussion of key CVA concepts and terminologies among participants could further enhance the robustness and applicability of results.

Interpretation and delivery of research outcomes

Although the integration of expert and local knowledge into CVA research has become more common, their roles in the post-assessment stage are rarely documented. Our findings highlight the potential for continued engagement with these knowledge holders during later stages to amplify the utility of research outputs and enhance policy uptake.

One promising avenue is the collaborative interpretation of research outcomes. In particular, interpreting differences in results can help better understand the root causes of divergences. Our comparison of data quality and expert confidence levels suggests three potential ways for engagement: (1) cross-validation in areas where multiple knowledge sources have high certainty but produce conflicting results, which can allow for verification and reconciliation of discrepancies through stakeholder discussions; (2) gap identification in areas with low certainty across all information sources, signaling the need for further research or the inclusion of additional perspectives; and (3) knowledge sharing and exchange to foster mutual learning in areas where one knowledge system demonstrates greater certainty than another, which could also help break down the information solos associated with the inaccessibility of information to potential users and create opportunities to build a shared understanding of complex challenges.

Finally, knowledge producers—such as scientists, managers, policymakers, NGOs, and fishermen—are often the end-users of CVA products. Insights generated through fishery CVAs can guide their priorities for research, management, and adaptation strategies. Engaging these stakeholders during the delivery phase of CVA results also ensures the users develop a good understanding of the research content and have a strong sense of ownership in the research product (Cvitanovic et al. 2015), which can foster broader dissemination within their networks, raise the awareness of others, and catalyze collective action to address climate challenges in fisheries.

LITERATURE CITED

Allison, E. H., A. L. Perry, M. C. Badjeck, W. N. Adger, K. Brown, D. Conway, A. S. Halls, G. M. Pilling, J. D. Reynolds, N. L. Andrew, and N. K. Dulvy. 2009. Vulnerability of national economies to the impacts of climate change on fisheries. Fish and Fisheries 10(2):173-196. https://doi.org/10.1111/j.1467-2979.2008.00310.x

Ban, N. C., A. Frid, M. Reid, B. Edgar, D. Shaw, and P. Siwallace. 2018. Incorporate Indigenous perspectives for impactful research and effective management. Nature Ecology and Evolution 2(11):1680-1683. https://doi.org/10.1038/s41559-018-0706-0

Blasiak, R., J. Spijkers, K. Tokunaga, J. Pittman, N. Yagi, and H. Österblom. 2017. Climate change and marine fisheries: least developed countries top global index of vulnerability. PLoS ONE 12(6):e0179632. https://doi.org/10.1371/journal.pone.0179632

Boyce, D. G., D. P. Tittensor, C. Garilao, S. Henson, K. Kaschner, K. Kesner-Reyes, V. Lam, E. Cheung, L. Morissette, M. Coll, and B. Worm. 2022. A climate risk index for marine life. Nature Climate Change 12(9):854-862. https://doi.org/10.1038/s41558-022-01437-y

Cannon, S. E., J. W. Moore, M. S. Adams, T. Degai, E. Griggs, J. Griggs, K. Klein, E. Anderson, R. Brown, A. Hill, and Indigenous Data Sovereignty Workshop Collective. 2024. Taking care of knowledge, taking care of salmon: towards Indigenous data sovereignty in an era of climate change and cumulative effects. FACETS 9:1-21. https://doi.org/10.1139/facets-2023-0135

Carroll, G., J. G. Eurich, K. D. Sherman, R. Glazer, M. T. Braynen, K. A. Callwood, R. Stump, J. Brodziak, A. Bryan, T. Curtis, and S. Haukebo. 2023. A participatory climate vulnerability assessment for recreational tidal flats fisheries in Belize and The Bahamas. Frontiers in Marine Science 10:1177715. https://doi.org/10.3389/fmars.2023.1177715

Cheung, W. W., V. W. Lam, J. L. Sarmiento, K. Kearney, R. Watson, D. Zeller, and D. Pauly. 2010. Large-scale redistribution of maximum fisheries catch potential in the global ocean under climate change. Global Change Biology 16(1):24-35. https://doi.org/10.1111/j.1365-2486.2009.01995.x

Cheung, W. W., T. J. Pitcher, and D. Pauly. 2005. A fuzzy logic expert system to estimate intrinsic extinction vulnerabilities of marine fishes to fishing. Biological Conservation 124(1):97-111. https://doi.org/10.1016/j.biocon.2005.01.017

Colburn, L. L., M. Jepson, C. Weng, T. Seara, J. Weiss, and J. A. Hare. 2016. Indicators of climate change and social vulnerability in fishing dependent communities along the Eastern and Gulf Coasts of the United States. Marine Policy 74:323-333. https://doi.org/10.1016/j.marpol.2016.04.030

Cooke, S. J., V. M. Nguyen, J. M. Chapman, A. J. Reid, S. J. Landsman, N. Young, S. Carriere, C. A. Semeniuk, L. Bishop, T. D. Beard Jr., and R. Arlinghaus. 2021. Knowledge co-production: a pathway to effective fisheries management, conservation, and governance. Fisheries 46(2):89-97. https://doi.org/10.1002/fsh.10512

Cope, J. M., and A. E. Punt. 2011. Reconciling stock assessment and management scales under conditions of spatially varying catch histories. Fisheries Research 107(1-3):22-38. https://doi.org/10.1016/j.fishres.2010.10.002

Cordier, J. M., L. Osorio-Olvera, P. Y. Huais, A. N. Tomba, F. Villalobos, and J. Nori. 2024. Capability of big data to capture threatened vertebrate diversity in protected areas. Conservation Biology 39(1):e14371. https://doi.org/10.1111/cobi.14371

Cvitanovic, C., A. J. Hobday, L. van Kerkhoff, S. K. Wilson, K. Dobbs, and N. A. Marshall. 2015. Improving knowledge exchange among scientists and decision-makers to facilitate the adaptive governance of marine resources: a review of knowledge and research needs. Ocean and Coastal Management 112:25-35. https://doi.org/10.1016/j.ocecoaman.2015.05.002

Fazey, I., L. Bunse, J. Msika, M. Pinke, K. Preedy, A. C. Evely, E. Lambert, K. Hastings, S. Morris, J. Reed, and M. S. Reed. 2014. Evaluating knowledge exchange in interdisciplinary and multi-stakeholder research. Global Environmental Change 25:204-220. https://doi.org/10.1016/j.gloenvcha.2013.12.012

Fleming, A., E. Ogier, A. J. Hobday, L. Thomas, J. R. Hartog, and B. Haas. 2020. Stakeholder trust and holistic fishery sustainability assessments. Marine Policy 111:103719. https://doi.org/10.1016/j.marpol.2019.103719

Food and Agriculture Organization of the United Nations (FAO). 2024. The state of world fisheries and aquaculture: blue transformation in action. FAO, Rome, Italy. https://doi.org/10.4060/cd0683en

Ford, J. D., G. McDowell, and T. Pearce. 2015. The adaptation challenge in the Arctic. Nature Climate Change 5(12):1046-1053. https://doi.org/10.1038/nclimate2723

Gallo, N. D., N. M. Bowlin, A. R. Thompson, E. V. Satterthwaite, B. Brady, and B. X. Semmens. 2022. Fisheries surveys are essential ocean observing programs in a time of global change: a synthesis of oceanographic and ecological data from US West Coast Fisheries surveys. Frontiers in Marine Science 9:757124. https://doi.org/10.3389/fmars.2022.757124

Gauvreau, A. M., D. Lepofsky, M. Rutherford, and M. Reid. 2017. “Everything revolves around the herring”: the Heiltsuk-herring relationship through time. Ecology and Society 22(2):10. https://doi.org/10.5751/ES-09201-220210

Gómez Murciano, M., Y. Liu, V. Ünal, and J. L. Sánchez LIzaso. 2021. Comparative analysis of the social vulnerability assessment to climate change applied to fisheries from Spain and Turkey. Scientific Reports 11(1):13949. https://doi.org/10.1038/s41598-021-93165-0

Hare, J. A., W. E. Morrison, M. . Nelson, M. M. Stachura, E. J. Teeters, R. B. Griffis, C. A. Griswold, E. Methratta, D. Alexander, and J. D. Scott. 2016. A vulnerability assessment of fish and invertebrates to climate change on the Northeast US Continental Shelf. PLoS ONE 11(2):e0146756.

Hu, A. 2013. The distinctive transition of China’s five-year plans. Modern China 39(6):629-639. https://doi.org/10.1177/0097700413499129

Intergovernmental Panel on Climate Change (IPCC). 2007. Climate change 2007: impacts, adaptation and vulnerability. Working group II contribution to the Intergovernmental Panel on Climate Change, fourth assessment report. Cambridge University Press, Cambridge, UK. https://www.ipcc.ch/site/assets/uploads/2018/03/ar4_wg2_full_report.pdf

Intergovernmental Panel on Climate Change (IPCC). 2023. Climate change 2023: synthesis report. Contribution of working groups I, II and III to the sixth assessment report of the Intergovernmental Panel on Climate Change. H. Lee and J. Romero, editors. IPCC, Geneva, Switzerland. https://doi.org/10.59327/IPCC/AR6-9789291691647

Jones, B. L., R. O. Santos, W. R. James, S. V. Costa, A. J. Adams, R. E. Boucek, L. Coals, L. C. Cullen-Unsworth, S. Shephard, and J. S. Rehage. 2024. New directions for Indigenous and local knowledge research and application in fisheries science: lessons from a systematic review. Fish and Fisheries 25(4):647-671. https://doi.org/10.1111/faf.12831

Kaufman, L. and P. J. Rousseeuw. 2009. Finding groups in data: an introduction to cluster analysis. John Wiley and Sons, Hoboken, New Jersey, USA. https://doi.org/10.1002/9780470316801

Kleisner, K. M., M. J. Fogarty, S. McGee, J. A. Hare, S. Moret, C. T. Perretti, and V. S. Saba. 2017. Marine species distribution shifts on the US Northeast Continental Shelf under continued ocean warming. Progress in Oceanography 153:24-36. https://doi.org/10.1016/j.pocean.2017.04.001

Lavell, A., M. Oppenheimer, C. Diop, J. Hess, R. Lempert, J. Li, R. Muir-Wood, and S. Myeong. 2012. Climate change: new dimensions in disaster risk, exposure, vulnerability, and resilience. Pages 25-64 in C. B. Field, V. Barros, T. F. Stocker, D. Qin, D. J. Dokken, K. L. Ebi, M. D. Mastrandrea, K. J. Mach, G.-K. Plattner, S. K. Allen, M. Tignor, and P. M. Midgley, editors. Managing the risks of extreme events and disasters to advance climate change adaptation. Cambridge University Press, Cambridge, UK. https://www.ipcc.ch/site/assets/uploads/2018/03/SREX-Chap1_FINAL-1.pdf https://doi.org/10.1017/CBO9781139177245.004

Li, Y., M. Sun, K. M. Kleisner, K. E. Mills, and Y. Chen. 2023. A global synthesis of climate vulnerability assessments on marine fisheries: methods, scales, and knowledge co-production. Global Change Biology 29(13):3545-3561. https://doi.org/10.1111/gcb.16733

Li, Y., M. Sun, X. Yang, M. Yang, K. M. Kleisner, K. E. Mills, Y. Tang, F. Du, Y. Qiu, Y. Ren, and Y. Chen. 2024. Social-ecological vulnerability and risk of China’s marine capture fisheries to climate change. Proceedings of the National Academy of Sciences 121(1):p.e2313773120. https://doi.org/10.1073/pnas.2313773120

Lima, M. S. P., J. E. L. Oliveira, M. F. de Nóbrega, and P. F. M. Lopes. 2017. The use of local ecological knowledge as a complementary approach to understand the temporal and spatial patterns of fishery resources distribution. Journal of Ethnobiology and Ethnomedicine 13:1-12. https://doi.org/10.1186/s13002-017-0156-9

Lomonico, S., M. G. Gleason, J. R. Wilson, D. Bradley, K. Kauer, R. J. Bell, and T. Dempsey. 2021. Opportunities for fishery partnerships to advance climate-ready fisheries science and management. Marine Policy 123:104252. https://doi.org/10.1016/j.marpol.2020.104252

Macusi, E. D., K. L. Camaso, A. Barboza, and E. S. Macusi. 2021. Perceived vulnerability and climate change impacts on small-scale fisheries in Davao Gulf, Philippines. Frontiers in Marine Science 8:597385. https://doi.org/10.3389/fmars.2021.597385

Marshak, A. R., and J. S. Link. 2024. Responses of fisheries ecosystems to marine heatwaves and other extreme events. PLoS ONE 19(12):e0315224. https://doi.org/10.1371/journal.pone.0315224

Marvin, H. J., E. van Asselt, G. Kleter, N. Meijer, G. Lorentzen, L. H. Johansen, R. Hannisdal, V. Sele, and Y. Bouzembrak. 2020. Expert-driven methodology to assess and predict the effects of drivers of change on vulnerabilities in a food supply chain: aquaculture of Atlantic salmon in Norway as a showcase. Trends in Food Science and Technology 103:49-56. https://doi.org/10.1016/j.tifs.2020.06.022

Mason, J. G., S. J. Weisberg, J. L. Morano, R. J. Bell, M. Fitchett, R. B. Griffis, E. L. Hazen, W. D. Heyman, K. Holsman, K. M. Kleisner, et al. 2023. Linking knowledge and action for climate-ready fisheries: emerging best practices across the US. Marine Policy 155:105758. https://doi.org/10.1016/j.marpol.2023.105758

McNie, E. C., A. Parris, and D. Sarewitz. 2016. Improving the public value of science: A typology to inform discussion, design and implementation of research. Research Policy 45(4):884-895. https://doi.org/10.1016/j.respol.2016.01.004

Metcalf, S. J., E. I. van Putten, S. Frusher, N. A. Marshall, M. Tull, N. Caputi, M. Haward, A. J. Hobday, N. J. Holbrook, S. M. Jennings, et al. 2015. Measuring the vulnerability of marine social-ecological systems: a prerequisite for the identification of climate change adaptations. Ecology and Society 20(2):35. https://doi.org/10.5751/ES-07509-200235

Mills, K. E., D. Armitage, J. G. Eurich, K. M. Kleisner, G. T. Pecl, and K. Tokunaga. 2023. Co-production of knowledge and strategies to support climate resilient fisheries. ICES Journal of Marine Science 80(2):358-361. https://doi.org/10.1093/icesjms/fsac110

Mistry, J., and A. Berardi. 2016. Bridging Indigenous and scientific knowledge. Science 352(6291):1274-1275. https://doi.org/10.1126/science.aaf1160

Morrison, W. E., M. W. Nelson, J. F. Howard, E. J. Teeters, J. A. Hare, R. B. Griffis, J. D. Scott, and M. A. Alexander. 2015. Methodology for assessing the vulnerability of marine fish and shellfish species to a changing climate. National Oceanic and Atmospheric Administration Technical Memorandum NMFS-OSF-3, Silver Spring, Maryland, USA. https://www.st.nmfs.noaa.gov/Assets/ecosystems/climate/documents/TM%20OSF3.pdf

Norström, A.V., C. Cvitanovic, M. F. Löf, S. West, C. Wyborn, P. Balvanera, A. T. Bednarek, E. M. Bennett, R. Biggs, A. de Bremond, and B. M. Campbell. 2020. Principles for knowledge co-production in sustainability research. Nature Sustainability 3(3):182-190. https://doi.org/10.1038/s41893-019-0448-2

Nye, J. A., J. S. Link, J. A. Hare, and W. J. Overholtz. 2009. Changing spatial distribution of fish stocks in relation to climate and population size on the Northeast United States continental shelf. Marine Ecology Progress Series 393:111-129. https://doi.org/10.3354/meps08220

Payne, M. R., M. Kudahl, G. H. Engelhard, M. A. Peck, and J. K. Pinnegar. 2021. Climate risk to European fisheries and coastal communities. Proceedings of the National Academy of Sciences 118(40):e2018086118. https://doi.org/10.1073/pnas.2018086118

Payne, N. L., J. A. Smith, D. E. van der Meulen, M. D. Taylor, Y. Y. Watanabe, A. Takahashi, T. A. Marzullo, C. A. Gray, G. Cadiou, and I. M. Suthers. 2016. Temperature dependence of fish performance in the wild: links with species biogeography and physiological thermal tolerance. Functional Ecology 30(6):903-912. https://doi.org/10.1111/1365-2435.12618

Pendleton, L. H., S. J. Alexandroff, A. Clausen, J. O. Schmidt, and H. I. Browman. 2023. Co-designing marine science for the ocean we want. ICES Journal of Marine Science 80(2):342-346. https://doi.org/10.1093/icesjms/fsad018

Shen, G. M., and M. Heino. 2014. An overview of marine fisheries management in China. Marine Policy 44:265-272. https://doi.org/10.1016/j.marpol.2013.09.012

Shen, Y., M. U. Ieong, and Z. Zhu. 2022. The function of expert involvement in China’s local policy making. Politics and Policy 50(1):59-76. https://doi.org/10.1111/polp.12450

Singh, G. G., H. Harden-Davies, E. H. Allison, A. M Cisneros-Montemayor, W. Swartz, K. M. Crosman, and Y. Ota. 2021. Will understanding the ocean lead to “the ocean we want”?. Proceedings of the National Academy of Sciences 118(5):e2100205118. https://doi.org/10.1073/pnas.2100205118

Su, S., Y. Tang, B. Chang, W. Zhu, and Y. Chen. 2020. Evolution of marine fisheries management in China from 1949 to 2019: how did China get here and where does China go next? Fish and Fisheries 21(2):435-452. https://doi.org/10.1111/faf.12439

Su, S., Y. Tang, J. P. Kritzer, and Y. Chen. 2022. Using systems thinking to diagnose science-based fisheries management in China. Marine Policy 138:104974. https://doi.org/10.1016/j.marpol.2022.104974

Sun, Y., C. Zhang, Y. Tian, and Y. Watanabe. 2021. Age, growth, and mortality rate of the yellow goosefish Lophius litulon (Jordan, 1902) in the Yellow Sea. Journal of Oceanology and Limnology 39(2):32-740. https://doi.org/10.1007/s00343-019-9216-4

Teck, S. J., B. S. Halpern, C. V. Kappel, F. Micheli, K. A. Selkoe, C. M. Crain, R. Martone, C. Shearer, J. Arvai, B. Fischhoff, G. Murray, R. Neslo, and R. Cooke. 2010. Using expert judgment to estimate marine ecosystem vulnerability in the California Current. Ecological Applications 20(5): 1402-1416. https://doi.org/10.1890/09-1173.1

Zhu, W., Z. Lu, Q. Dai, K. Lu, Z. Li, Y. Zhou, Y. Zhang, M. Sun, Y. Li, and W. Li. 2021. Transition to timely and accurate reporting: an evaluation of monitoring programs for China’s first Total Allowable Catch pilot fishery. Marine Policy 129:104503. https://doi.org/10.1016/j.marpol.2021.104503