A systematic review of the effectiveness of protected areas at reducing threats to biodiversity

A systematic review of the effectiveness of protected areas at reducing threats to biodiversity | 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 A systematic review of the effectiveness of protected areas at reducing threats to biodiversity Katherine Pulido-Chadid, Helena N. Alves-Pinto, Harith Farooq, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7046705/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 10 You are reading this latest preprint version Abstract Background Protected areas (PAs) are central to global biodiversity conservation efforts, yet their effectiveness in reducing threats is context-dependent and poorly understood. This systematic review evaluates how well PAs mitigate anthropogenic threats to biodiversity by synthesizing findings from peer-reviewed and grey literature. Specifically, we assess how threat levels change in relation to protection status, over time, and in response to specific management interventions. The review aims to evaluate threat levels inside PAs relative to appropriate comparators or interventions, identify variation across threat types, and highlight key factors associated with both successful and unsuccessful conservation outcomes. Methods We conducted a comprehensive systematic review, screening 5,687 unique articles and including 105, that assessed changes in threat within at least one PA compared to other locations, time periods, or types of interventions. A descriptive, narrative synthesis was performed to evaluate the strength of study designs and the influence of methodological differences on reported outcomes. The synthesis focused on three key dimensions: (1) the proportion of studies reporting reduced threats inside PAs over time or relative to comparators; (2) effectiveness across different threat types; and (3) contextual and management-related factors influencing PA performance. Review findings Our review revealed three major findings: Firstly, PAs on average had lower threat levels compared to unprotected areas, but effectiveness was mixed. Most studies reported increasing threats both inside and outside PAs. Nonetheless, threats tended to be lower within PAs boundaries. Secondly, the threat reduction potential depends on the type of threat. PAs were more effective at reducing land-use and land-cover change (LULCC), fire, and hunting. Studies frequently reported lower levels of deforestation, agricultural expansion, and infrastructure development within PAs. However, PAs were less effective in addressing threats like invasive alien species, pollution, and recreation, likely due to external origins or indirect impacts that extend beyond formal boundaries. Thirdly, PA outcomes were shaped by multiple interrelated factors. Effectiveness was generally lower in areas with high baseline threats, high human population density, and easy access. In contrast, strong governance, adequate funding, robust enforcement, large PA size, and alignment with national development goals and governance were associated with more positive conservation outcomes. Conclusions Despite rising human-induced threats across landscapes, PAs demonstrated better performance than unprotected landscapes in reducing threats to biodiversity. However, their success is highly dependent on the context, including the type of threat, biophysical characteristics, and socio-economic context where PAs are located. This review highlights the urgent need to enhance the effectiveness of existing PAs and offers important insights for the planning and implementation of future ones. The potential of PAs to reduce threats is not solely determined by their design or location but also by how well they are integrated into the local socio-ecological systems in which they operate. Strengthening governance, improving management strategies, and ensuring the meaningful involvement of local stakeholders are essential to ensure long-term conservation outcomes. Additionally, this review underscores the need for standardized methodologies, consistent threat assessment metrics, and rigorous evaluation frameworks to better understand what drives PA effectiveness and to support evidence-based decision-making. Protected areas Conservation Threats to biodiversity Protected areas effectiveness Threat reduction in protected areas Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Background Currently, we are facing an unprecedented biodiversity crisis, with an increasing number of species facing extinction as we exceed the safe operating boundaries for biosphere integrity (Richardson et al., 2023). These planetary challenges are driven by a series of threats that affect biodiversity (IPBES, 2019). Direct threats are posed by human actions that cause degradation or loss of ecosystems and species populations, such as agriculture, infrastructure development, resource extraction, and ecosystem management. Additionally, threats encompass disturbances in natural systems whose dynamics have been altered by past or present human activity, including invasive species, pollution, natural disasters, and climate change (Salafsky et al., 2025). Ultimately, threats to biodiversity are shaped by indirect drivers rooted in societal values and behaviors, including patterns of production and consumption, trade practices, technological innovations, and governance structures, all of which vary across regions and contexts (IPBES, 2019). Protected areas (PAs) are among the most widely used tools for nature conservation, which aim to reduce both indirect and direct threats to biodiversity (Rodrigues & Cazalis, 2020; Watson et al., 2014). International agreements such as the Convention on Biological Diversity (CBD) have consistently called for increased PA coverage and improved management, resulting in expanded protected land during recent decades. As of 2024, approximately 17% of terrestrial ecosystems and less than 10% of marine ecosystems were covered by PAs or other effective area-based conservation measures (OECMs) (UNEP-WCMC, 2024). Despite the rising numbers and the global target of 30% coverage expected by 2030, as outlined in the Kunming-Montreal Global Biodiversity Framework (CBD, 2022), biodiversity loss continues, and human pressures inside and outside PAs are increasing (CBD, 2020; Geldmann et al., 2019; Jones et al., 2018; Leclère et al., 2020). As a result, the Kunming-Montreal Framework, aligned with the 2030 Agenda for Sustainable Development, emphasizes not only expanding PAs but also directly addressing the key threats driving biodiversity loss, as stated in Targets 1 to 8 (CBD, 2022). The ecological effectiveness of PAs depends on a combination of factors, including decisions made at the time of their establishment, such as location, extent, design, connectivity, and representativeness, as well as subsequent governance and management practices (Rodrigues & Cazalis, 2020). However, assessing their impact remains challenging, as PAs are embedded within complex socio-ecological systems where multiple interrelated factors influence the outcomes (Durán et al., 2022; Geldmann et al., 2025). Research shows that PAs generally reduce biodiversity decline compared to unprotected sites (Coetzee et al., 2014; Geldmann et al., 2013). Nonetheless, their effectiveness varies significantly depending on contextual factors, including socio-economic conditions and management regimes (Graham et al., 2021; Wauchope et al., 2022). For example, wildlife populations tend to be higher in PAs located in regions with stronger development indicators (Barnes et al., 2016). Another major challenge in evaluating PA impact is the presence of inherent location biases. Many PAs are established in remote areas with limited economic interest, making direct comparisons with unprotected lands problematic (Joppa & Pfaff, 2011). To address this limitation, studies have adopted counterfactual approaches, comparing outcomes before and after PA establishment and/or against comparable unprotected areas. These methods allow for more robust estimates of the ecological effects of PAs by modeling what would have occurred in their absence (Chen et al., 2023; Schleicher et al., 2020; Wauchope et al., 2022). The long-term conservation success of PAs relies not only on their initial configuration but also on their capacity to mitigate threats through effective management practices (Geldmann, 2023; Margules & Pressey, 2000; Rodrigues & Cazalis, 2020). PAs have been established in response to threats to biodiversity, but these threats are also the catalysts for creating management strategies for PAs. Analyzing how effectively PAs and their management strategies reduce threats can thus provide essential insights into the broader management effectiveness of PAs. Given the urgent need for effective PAs capable of reducing threats and countering ongoing biodiversity loss, we aim to identify whether and how PAs have succeeded in threat reduction, and what characteristics have contributed to or hindered their effectiveness by conducting a systematic literature review. 2.1 Objective of the Review A key objective of establishing PAs is to reduce the pressures that threaten biodiversity on unprotected lands and seas. This systematic review aims to test that premise by assessing the effectiveness of PAs in reducing threats to biodiversity through a synthesis of evidence from peer-reviewed and grey literature across all available anthropogenic threat types, regions, habitats, and study scales. To achieve this, we evaluate different measures of impact, incorporating spatial, temporal, and management intervention comparisons. The focus of this review is on how threats to biodiversity change in relation to protection status, evolve over time, and/or respond to implemented intervention strategies. Our main goal is to assess the effectiveness of PAs and associated interventions in reducing threats to biodiversity. In addition, we examine how effectiveness varies across different threat types and identify key factors contributing to conservation outcomes. Methods This systematic review followed the Guidelines and Standards for Systematic Reviews of the Collaboration for Environmental Evidence (Collaboration for Environmental Evidence, 2022). The methods of this review adhered to a published peer-reviewed systematic review protocol (Pulido-Chadid et al., 2023) with a few minor deviations (see section 3.1). 3.1 Deviations from the protocol Specifically, we explicitly excluded publications related to climate change, given the inherent challenges in accurately measuring its effects on PAs. Additionally, literature reviews were excluded to maintain a focus on primary sources of research. To optimize the review process, data extraction was conducted by a single reviewer (KP). Any uncertainties that arose were referred to a senior scientist for resolution, ensuring alignment with the inclusion criteria and the review’s objectives. Additionally, a random 10% subset of articles was independently reviewed by JG. Disagreements were minimal and were discussed and resolved through consensus. 3.2 Search for articles 3.2.1 Search sources, terms and strings: We conducted searches in Scopus and the Web of Science Core Collection bibliographic databases using the search string and search blocks (Table 1 ) adopted from Pulido-Chadid et al. (2023). The bibliographic database searches were conducted on the 14 of August of 2023. Additionally, we searched for grey literature using Google Scholar and the websites of the Wildlife Conservation Society (WCS), International Union for Conservation of Nature (IUCN), Global Environmental Facility (GEF), World Wide Fund for Nature (WWF), Conservation International, and BirdLife International. Table 1 Search string that will be used in the search of the literature in the Web of Science Core Collection and Scopus Topic Search string Threats to biodiversity threat $ OR "human impact*" OR "human pressure*" OR” anthropogenic impact*" OR "human activity" OR stressor OR “anthropogenic pressure” AND Measure the change of threats reduc* OR effective* OR impact AND Protected areas "protected area*" OR "conserv* area" OR "nature reserve" OR sanctuar* OR "national park*" OR "biosphere reserve*" OR "biodiversity reserve*" OR "wildlife habitat*" Adopted from: Pulido-Chadid et al., 2023. For grey literature, we adapted our approach based on available filtering options. On organizational websites, we used “protected areas” as the primary keyword and manually screened relevant literature. In Google Scholar, we searched for “protected area” AND “threats” and reviewed the first 50 results, as specified in the protocol (Pulido-Chadid et al. 2023). The keywords, search dates, websites accessed, and retrieved information are detailed in Additional File 2 from Pulido-Chadid et al. 2023. 3.2.2 Search limitations No restrictions were applied regarding the year of publication or document type of the articles, with reviews removed as part of the document screening process. Our searches were conducted in English. While we acknowledge that publications in other languages may be relevant (Amano et al., 2023; Amano & Sutherland, 2013), particularly given the diverse geographical scope of this review, non-English publications were not included, which may limit the representation of findings from certain regions. 3.2.3 Estimating the comprehensiveness of the search At the protocol stage, we tested the comprehensiveness of the search strings by assessing the search strategy comprehensiveness of 20 benchmark articles (see Pulido-Chadid et al., 2023). 3.2.4 Search results All records, including their bibliographic information, were exported to EndNote 20 (The EndNote Team, 2013), where deduplication removal was performed. The unique records were then exported to Covidence ( https://www.covidence.org/home ), where some additional duplicates were identified and removed. 3.3 Article screening and study eligibility criteria 3.3.1 Screening process Prior to the conclusive screening process, we conducted an initial consistency check to ensure that the eligibility criteria were interpreted and applied uniformly across reviewers. To achieve this, we randomly selected 600 abstracts to ensure sufficient representation of the identified literature. These abstracts were screened in a double-blinded process and independently by the reviewers (KP, AG, HAP, HF, EV, JG). To advance to full-text screening, we required a moderate agreement threshold (Kappa ≥ 0.6). Any discrepancies were discussed and resolved to improve consistency, and if necessary, the inclusion and exclusion criteria were refined. This process was conducted in three iterative rounds. In the first round, the 600 random subset was divided into two groups, one with two reviewers and another with four reviewers. We observed that the agreement was slightly higher with fewer reviewers, with a Kappa value of 0.42 in the two-reviewer group compared to 0.405 in the four-reviewer group. Discrepancies were analyzed, and refinements were made to the eligibility criteria. Following this, an intermediary step was introduced to further clarify the inclusion criteria. Two smaller tests were conducted, where a random selection of 15 abstracts was screened and carefully discussed twice among all reviewers. This resulted in an improvement in the Kappa value from 0.44 in the first test to 0.51 in the second round. In the final assessment, a new random subset of 675 abstracts was screened. These were divided into three groups of two reviewers, achieving the benchmark Kappa score of 0.6. Once a sufficient agreement was achieved, each article was categorized into one of three groups: (1) include, (2) exclude, or (3) maybe. As it was conducted in the consistency check, the “maybes” were then reevaluated by the lead author. Then full title and abstract screening process was conducted for all remaining articles among reviewers. The full-text analysis of all articles selected in the first step was screened first by the lead author. Any uncertainties that arose during the extraction were continuously referred to and discussed with the senior scientist for resolution, ensuring alignment with the inclusion criteria and the review’s objectives. Additionally, a randomly selected, blinded subset of 10% of articles was independently reviewed by the senior scientist. Disagreements were minimal and were resolved through consensus. A list of articles and grey literature excluded during full-text screening, along with reasons for exclusion, is provided in Additional File 1 . Articles excluded from the first stage (title and abstract screening) were not coded with reasons for exclusion, but are available. 3.3.2 Eligibility criteria To qualify for this review, publications were selected based on the Population, Intervention, Comparator, and Outcome (PICO) framework (Table 2 ). The inclusion criteria were designed to identify publications that investigate the effectiveness of PAs in controlling threats to biodiversity and contribute to a broader understanding of their role in biodiversity conservation. Publications were required to focus on anthropogenic threats, assess changes within at least one PA, and include a comparator(s) to evaluate changes over time, space, or management interventions (Table 2 ). The included studies employed a range of evaluation designs to assess threat changes in PAs. BACI (Before-After Control-Impact) designs were used in studies that incorporated both a temporal comparison of conditions before and after the implementation of an intervention (such as a management strategy, change in management authority, or PA designation) and a spatial comparison between the PA and surrounding areas or comparator sites. BA (Before-After) designs relied solely on temporal comparisons, evaluating changes before and after the intervention without including spatial control. CI (Control-Impact) designs focused on spatial comparisons between PAs and surrounding areas or comparator sites but did not specify the use of matching approaches. Matching is a statistical technique that pairs protected and unprotected areas with similar characteristics—such as habitat type or fishing pressure—to isolate the effect of PA protection (Ahmadia et al., 2015). These may include temporal comparisons, but only for the post-intervention period. Finally, CI (Matching) studies also employed spatial comparisons but explicitly used matching methods to select control sites. These studies may similarly include a temporal comparison after the intervention, but not necessarily a “before” period. Table 2 Eligibility criteria for article screening Category Inclusion criteria Exclusion criteria Population Areas experiencing threats to biodiversity A. Studies measuring the change of threat levels in PAs. - Threats must be anthropogenic and assessed directly or indirectly using proxies (e.g., Human Footprint Index). - Studies must mention threats explicitly or implicitly (in the abstract) or use synonyms. - Studies using surveys or interviews that quantify human perceptions of threat change (e.g., Threat Reduction Assessment). - Studies that do not address threats to biodiversity or measure changes in threat states. - Studies focusing on non-anthropogenic threats (e.g., geological events). - Climate change is excluded due to challenges in measuring PA effectiveness in mitigating or exacerbating its impacts. Intervention Establishing protected areas B. Studies assessing threats inside at least one PA. - Includes all PA types and designations for biodiversity conservation (e.g., national parks, wildlife sanctuaries, nature reserves, biosphere reserves). - Studies conducted at any scale (local to global) and across terrestrial, freshwater, and marine ecosystems. - Studies that do not focus on areas designated primarily for biodiversity conservation. - Studies on OECMs (Other Effective Area-Based Conservation Measures) that do not meet the definition of PAs. Comparator Difference in threat state C. Studies measuring changes in threats within PAs, using at least one comparator : C1. Temporal Comparison: Threat state assessed at two or more time points within PAs. C2. Management Comparison: Comparison of different PA management interventions over time or with control sites (C3). C3. Spatial Comparison: Threat changes in PAs assessed using spatial comparators: - PA vs. non-PA (e.g., non-protected locations, buffer zones, surrounding areas, land-use types, control sites). - PA vs. PA (e.g., within the same PA with a control site or threat/protection gradient, or across PAs with different categories/intervention types). Studies that do not assess changes in threats as an outcome of protection, management intervention, or spatial comparison (i.e., lacking a comparator: C1, C2, or C3). Outcome Assessment of threat change in PAs Selected studies must meet criteria A, B, and C and conduct a quantitative analysis aligned with the review’s objectives by addressing one of the following: • Evaluates changes in threats resulting from protection or interventions within PAs (Direct impact assessment of threats in PAs) • Compares threats or assesses the potential effectiveness of PAs with any comparator (C1, C2, C3) (Indirect or potential threat assessment). Metrics must be directly linked to the specific threat (e.g., invasive species: presence inside vs. outside PAs; deforestation: % forest cover). - Studies that do not assess threat states after an intervention. - Studies measuring proxies of threats that may be influenced by other factors (e.g., species abundance, population traits, vegetation cover, animal behavior, community composition). Thus, studies relying solely on biodiversity metrics to measure PA impact were excluded. Study design Studies employing comparison groups CI and/or BA or BACI designs. - Personal views, perspectives, theoretical studies, and models. - Observational studies with no controls or comparators. - Literature reviews. 3.3.3 Study validity assessment A critical appraisal process was conducted using the Collaboration for Environmental Evidence Critical Appraisal Tool (Version 0.3, Prototype) (Konno et al., 2021). This tool was adapted to align with the specific requirements of this study. Given the nature of the research, the assessment focused on key sources of bias, including risk of confounding biases, post-intervention/exposure selection biases, misclassified comparison biases, detection biases, outcome reporting biases, and outcome assessment biases. As part of the consistency-checking process, a random subset of full-text articles was critically appraised by the senior researcher to ensure alignment with the conducted appraisal and maintain assessment reliability. Articles identified as having a high risk of bias were excluded from the analysis. The adjusted appraisal template and individual study assessments are available in Additional File 2 . 3.3.4 Data coding and extraction strategy Data coding was conducted based on the PICO framework (Population, Intervention, Comparator, and Outcome), along with relevant metadata. To streamline the process, data extraction and critical appraisal were performed using a predefined coded template in Covidence, designed to ensure consistency and efficiency in the data extraction process. The decisions of the coded data, as well as the options for selection, can be found in Additional File 3 . Consistency checks were conducted in collaboration with the senior author to validate the accuracy and reliability of the extracted information. We systematically recorded study metadata, including author(s), year of publication, title, and whether the study explicitly aimed to evaluate effectiveness as a primary goal. We also collected data on study characteristics such as geographic region, ecosystem classification (terrestrial, freshwater, marine), study design type, comparator type, threat assessment methods, information on PAs, reported biodiversity threats, and study outcomes. A comparator refers to any unit, such as an area, period, or management intervention, used to evaluate the effectiveness of a PA. Comparators can include areas outside the PA, buffer zones surrounding the PA, PAs with varying levels of protection, or comparisons before and after an intervention. These comparators may be descriptive or loosely matched without strict equivalency, or they may involve the use of control with counterfactuals, which aim to represent what would have happened in the absence of the intervention (Table 3 ). Table 3 Comparator types selected Comparator type Description Inside - Outside Compares areas inside and outside the PA. PA - PA buffer Compares the interior of the PA with its surrounding buffer zone. PA - PA categories Compares outcomes between PAs with different levels of enforcement or protection categories. PA - Gradient Compares threat levels within PAs under different gradients of pressure, management strictness, or locations within the PA facing conditions influencing the outcomes. PA - Management intervention Compares threat levels inside PA(s) to measure the effectiveness of a management intervention PA - Other land use types Compares threat levels of PA(s) with other land use types in the surrounding landscape (eg. Indigenous lands) Outcome data were extracted by identifying measurable changes in threats, whether directly assessed within PAs or inferred from indicators such as human footprint indices (Sanderson et al., 2002; Venter et al., 2016a, 2016b) or modification indices (Kennedy et al., 2019), recognizing that threats can manifest at multiple levels, including drivers, sources, and mechanisms (Balmford et al., 2009). Specifically, we documented changes in threats both within PAs and in comparator (control) areas, as well as whether PAs were found to be more effective than their respective controls. We also extracted any reasons identified by the articles that explained the observed outcomes. Following the CEE Critical Appraisal Tool (Konno et al., 2021), we assessed the risk of bias for each study. The original appraisal categories were adapted to better align with the context and scope of this review. To ensure consistency in data extraction across the full set of papers, a subset of full-text papers was randomly selected and critically appraised by a senior researcher to verify alignment with the conducted appraisal. Discrepancies were identified, discussed, and resolved through consensus. In cases where one of the review authors was also an author of a study included in the review, they were not involved in its screening, appraisal, or selection of that article. Rules and conditions for data extraction Given the heterogeneity of results across articles, including differences in the number and type of threats analyzed (e.g., single threats, composite indices, or multiple threat types), as well as the number and characteristics of PAs assessed, we adapted our data extraction sheet accordingly, applying a set of predefined rules. When an article analyzed multiple threats, outcomes were recorded based on the availability of data for each threat individually. Additionally, when an article contained distinct intervention types (e.g., between PAs with different levels of strictness), data were extracted separately for each case. We refer to these individual comparisons as separate “studies.” As such, a single article could contribute with multiple studies to the review. In contrast, if an article assessed several PAs of the same category, it was treated as a single study, and only the overall result was extracted (Additional file 3). For outcome extraction, we focused on quantitative data and assigned a qualifier based on predefined criteria. Specifically, when a study examined changes in threats within a PA over time or between comparators, we categorized the outcomes based on the magnitude and direction of reported changes. Changes greater than 10% were classified as either "increased" or "decreased," depending on the direction of the reported outcome. When the change ranged between 5 and 10%, the outcome was classified as “slightly increased” or “slightly decreased”. Changes smaller than ± 5%, or cases where statistical tests indicated non-significant differences (e.g., p-values above the significance threshold), were classified as "neutral." These thresholds were established a priori to ensure consistency in interpreting and comparing results across studies. Reported threats were first coded according to the IUCN Threat Classification Scheme – Version 3.3 (IUCN, 2024). To facilitate analysis and interpretation, similar threat types were subsequently grouped into ten simplified categories, which we refer to as reported grouped threats (Table 4 ). The definitions and inclusion criteria for each coded response are available in Additional File 3, and full details of the data extraction process, including all variables collected from each study, are available in Additional File 2 . Table 4 Categorization of reported threats based on the IUCN threat classification scheme Reported threats Included IUCN threats LULCC-Infrastructure 1 Residential & commercial development, 4 Transportation & service corridors LULCC-Agriculture 2 Agriculture & aquaculture, grazing Fishing and Hunting 5.1 Hunting, gathering, fishing LULCC-Deforestation 5.3 Logging & wood harvesting Recreation 6.1 Recreational activities Fire 7.1 Fire & fire suppression Invasive Species 8 Invasive & other problematic species, genes & diseases Pollution 9 Pollution (Light pollution, chemical pollution) Multiple Threats Multiple threats with unspecified independent outcomes LULCC-Other Combined human pressure indices: HFI, Human pressure, Human modification, Temporal Human Pressure Index (THPI) 3.3.5 Potential effect modifiers/reasons for heterogeneity Effect modifiers contributing to heterogeneity in the results were identified during the full-text screening and recorded for all included studies. Where possible, we collected information on the methods used to assess the impact of potential effect modifiers. Given the nature of our study, various biogeographic, environmental, and socio-economic factors contributed to the heterogeneity of impacts reported across studies. Some potential effect modifiers identified in previous research included the category of the PA, governance type, geographical location, topographic features, PA size, date and period of establishment, the socioeconomic context of the country or state where the PA was located, and ecosystem type, among others. Where available, we recorded these potential effect modifiers in Additional File 2. 3.3.6 Data synthesis and presentation We conducted a descriptive, narrative synthesis of the included studies. This synthesis describes the robustness of the included studies and the potential impact of methodological differences on the reported outcomes. Thus, frequencies and proportions were calculated across all studies to analyze patterns in study metadata, methodological approaches, threat assessment techniques, and outcomes. Metadata analysis included variables such as geographical location, ecosystem type (terrestrial, freshwater, marine), and study scale (local, regional, global). To assess methodological variation, we examined study designs and comparator types, determining whether studies used temporal, spatial, management-based, or mixed comparators. We also evaluated the threat assessment methods used, distinguishing between remote sensing, field-based data, and indices or proxies, as well as the number of threats analyzed per study (e.g., single threats, composite indices, or multiple threats). The validity assessment of the articles was based on predefined criteria, and we computed frequencies and proportions to assess the risk of bias. Furthermore, we contrasted the findings from the validity assessment with the types of study designs and their respective comparators to gain deeper insights into the potential methodological factors contributing to increased risk of bias. To summarize the major findings of our systematic review, the data synthesis was structured into three key main results. First, we analyzed the number and proportion of studies reporting the magnitude of changes inside and outside PAs across time, and the number of times where PAs exhibited lower threat intensity compared to their counterfactuals. Second, we analyzed PA effectiveness across different threat types. As some studies reported outcomes for multiple threats, we extracted data for each threat individually. We refer to individual threats within studies as “cases” (Additional file 3). Third, we examined the factors influencing PA outcomes, as reported or discussed in the studies. These factors were categorized into five main drivers of PA effectiveness: location and baseline conditions, design, management, and governance. Within these categories, we explored subcriteria determining whether PAs had a positive, neutral, or negative impact on threat mitigation. Tables and figures summarizing the results are provided as supplementary information in the systematic review as well as the collected data. All data wrangling and statistical analyses were conducted using R (version 4.2.3) (R Core Team, 2021). Review findings 4.1 Review descriptive statistics Our search yielded 5,687 unique articles. After screening titles and abstracts based on the inclusion criteria and alignment with the review objectives, 407 (7%) articles were selected for full-text review. Of these, 296 (5%) that did not meet the inclusion criteria were excluded. A total of 111 (2%) articles were included after the full-text screening and critical appraisal. The most common reason for exclusion at the full-text screening was the absence of a comparator (n = 103), which also served as a pre-filter for critical appraisal. For example, studies assessing threat states at a single time point, lack of relevance to the review objectives, such as studies not involving PAs; and reliance solely on changes in biodiversity metrics to infer the impact of threats. Finally, articles with a high risk of bias were excluded according to our critical appraisal assessment, resulting in a total of 105 articles. Within these articles, 9 contained different intervention comparisons resulting in a total of 117 studies that were included in the analysis (Fig. 1 ). Geographical location and habitat types Across the reviewed articles we had studies from 133 countries, with Asia accounting for the largest share (30%), followed by Europe (19%), South America (16%), and Africa (15%), which showed relatively balanced representation. Research from China, Brazil, Spain, Italy, Tanzania, Ecuador, and Indonesia primarily shaped these continental patterns (Fig. 2 ). Terrestrial ecosystems were the focus of 71% (n = 82) of the articles, while 12.4% (n = 13) examined marine and coastal habitats and 5.7% (n = 6) focused on freshwater ecosystems. The spatial scale of the studies was relatively homogeneous, with 36.2% conducted at the local scale (n = 38), followed by regional (26.6%), national (21%), global (8.6%), and continental (7.7%) scales. Threat assessment methods, study designs and comparators Remote sensing was the most common method to assess threats, used in 68% (n = 71) of the articles. Composite threat indices, such as the human footprint index, were used in 19% (n = 20), while only 11.6% (n = 11) of the articles relied on field sampling or interviews. Eighty percent of the articles assessed a single threat (n = 84), while fewer articles examined two (n = 10), three (n = 5), four (n = 5), and five threats (n = 1). In terms of study design, CI was the most common approach being employed in 80% (n = 84) of the articles, including 14.3% (n = 15) that applied matching. BACI and BA study designs were used less frequently, found in 11.4% (n = 12) and 8.6% (n = 9) of articles, respectively. The most common comparator was an "Inside-Outside" used in 34% (n = 50) of the articles, followed by comparisons between PAs and their buffers (26.5%, n = 39), management interventions (17%, n = 25), and different PA categories (15.6%, n = 23). Studies using CI and CI matching methods used mainly inside-outside, buffer, and PA category comparisons. In contrast, BA and BACI designs were more commonly applied in studies focused on management interventions (Fig. 3 ). 4.2 Narrative synthesis including study validity assessment We identified confounding and selection biases as the primary sources of methodological limitations. Over half of the articles exhibited a medium risk of bias in confounding and selection criteria, while the risk of bias in misclassification, detection, reporting, and assessment was generally low (Fig. 4 , A). When examining the causes of confounding bias, we found that in 90.4% of the articles, the impact of the exposure or intervention was likely or seemingly likely to be confounded (Fig. 4 , B). However, 61.5% controlled for potential confounders. Notably, 17% of articles did not mention the omission of confounders (Fig. 4 , B). CI studies had the highest frequency of medium-risk bias for both confounding (n = 38) and selection biases (n = 45), suggesting that despite their widespread use, CI designs are particularly vulnerable to these biases. This was especially evident in Inside–Outside and PA–PA Buffer comparisons. In contrast, CI studies with matching exhibited lower overall frequencies of bias, indicating that the use of matching techniques may help mitigate confounding and selection issues. BA and BACI studies had lower overall bias risks, though they were also less frequently used (Fig. 4 , C). 4.3 Data synthesis Mixed effectiveness of PAs in mitigating threats On average, we found that PAs had lower threat levels compared to unprotected areas, with 65% of observations (87 out of 134) showing that PAs performed better than their outside comparator (Fig. 5 , C). However, many PAs still experienced increases in threats. Fifty-seven percent of observations (73 out of 129) reported rising threat levels inside PAs (Fig. 5 , A), while 75% (83 out of 110) reported increases in control sites or outside PAs (Fig. 5 , B). LULCC-related threats such as infrastructure development, agriculture and deforestation were the most frequently reported threats increasing within PAs, with increases observed in 87%, 64%, and 53% of observations, respectively. However, these threats also exhibited the highest rates of reduction, with deforestation decreasing in 39% of the observations, agriculture in 36%, and other LULCC types in 21%, indicating that PAs can play a role in slowing land-use change. Notably, LULCC has been the dominant focus of the literature, reflecting its prevalence as the most frequently analyzed threat type. In contrast, fire and invasive species have proven more challenging to address even inside PAs, which were reported to be increasing in 70% and 60% of the observations inside PAs respectively (Fig. 9, A). Outside PAs, the rate of threat increase was generally higher than inside, with infrastructure development and other LULCC threats showing increases in 100% and 74% of the observations, respectively, suggesting that PAs provide some level of containment for these pressures. While deforestation and agriculture also increased outside PAs, some studies reported evidence of reductions in these threats near PAs, possibly due to buffer zones or spillover effects from conservation measures. For example, Xu et al., (2022) found that overall forest cover increased and fragmentation decreased from baseline years, when reserves were designated, in both reserves and surrounding areas (Fig. 6 , B). PAs effectiveness varies by threat type PAs were found to be effective in mitigating certain threats when compared to unprotected areas. Most studies assessing LULCC-related pressures reported lower threat levels inside PAs compared to areas without protection, particularly for infrastructure (73%), deforestation (77%), and agriculture (57%). Additionally, fire (56%) and hunting (80%) were significantly less prevalent within PAs, reinforcing their role in reducing direct anthropogenic disturbances (Fig. 6 , C). However, the effectiveness of PAs in buffering against pollution and recreation-related pressures was less evident, likely due to the external origin of these threats or their manifestation as secondary impacts within protected boundaries or create secondary threats within the PAs themselves (Fig. 9). For instance,Mollmann et al. (2022) found that stream water quality inside PAs reflected broader regional pollution patterns, indicating that exogenous inputs continue to shape environmental conditions within PA borders. Likewise,Gonson et al. (2017) found that marine protected areas (MPAs) experienced higher tourism-related pressure indices on islets than nearby unprotected sites. Drivers of PAs effectiveness Our results show that PA outcomes are shaped by a combination of location, design, management, and governance factors (Fig. 7 ). PAs being located under high threat and accessibility were consistently reported as major obstacles to PA effectiveness. Similarly, high population density and permeable boundaries reduce PA effectiveness when external pressures, like pollution, continue to impact ecosystems despite protection. Landscape fragmentation and PAs with a small size were frequently associated with increased threats, highlighting the vulnerability of PAs to be highly impacted by location, baseline and design conditions. Conversely, well-resourced PAs with sufficient funding, strong policy, and effective enforcement mechanisms were more successful in achieving positive outcomes. Notably, in some cases, robust management interventions and strong policies were sufficient to counteract the negative effects of high accessibility, population density, and high-threat environments. For instance, Bleher et al., (2006), found that a shift in management authority led to reduced deforestation and improved performance in PAs with high level of enforcement. Similarly, (Eklund et al., (2022) reported that the suspension of management activities in Madagascar’s PAs during COVID-19 resulted in a 76–248% increase in fire incidents, with burning levels returning to normal once management resumed. In addition, large PA size was more frequently associated with positive conservation outcomes, whereas older PAs displayed mixed effects. Hirons et al. (2022) and Wu et al. (2022) found that larger reserves experienced less human modification and reduced anthropogenic pressure across tropical biomes and in Northeast China, respectively. In contrast, age had a more complex influence: older and smaller PAs in Africa were more vulnerable to deforestation (Bowker et al., 2017), and globally, older PAs saw increased human pressure (Geldmann et al., 2014). In Europe, age played a more positive role in naturalization near long-established reserves (Mingarro & Lobo, 2023) while its effect on fragmentation in Natura 2000 sites was mostly neutral (Lawrence et al., 2021). Local governance structures and national development indices demonstrated positive effects. For example, Mammides, (2020), found that lakes in North America and Europe experienced lower or even decreasing human pressure, while biodiverse lakes in tropical regions faced significantly higher and rapidly increasing human impact. 4.4 Review limitations Limitations of the Review Methods The heterogeneity of study designs, variation in measured outcomes, and the lack of standardized reporting of effect sizes and associated uncertainties made it impossible to quantitatively compare the magnitude of effects across studies. Therefore, we were unable to conduct a meta-analysis based on the studies. This reflects a broader challenge in conservation evidence synthesis, where inconsistent or incomplete reporting impedes cross-study comparisons (Pullin & Stewart, 2006); Bowler et al., 2020). A meta-analysis would have enabled a more quantitative synthesis by allowing standardized comparisons across different studies (Gurevitch et al., 2018), and potentially draw more targeted conclusions for policy and practice. Our search strategy was restricted to studies published in English, which may have led to the omission of relevant non-English literature. This language bias can result in regional underrepresentation, particularly of studies from non-English-speaking biodiversity-rich countries (Amano et al., 2023). Despite this limitation, our systematic review includes a diverse representation of countries, limiting the risk that our findings are not broadly applicable across different geographic and governance contexts. While the objective and questions of this study have arisen from the authors' scientific motivation, stakeholder engagement is crucial for effective conservation. Therefore, we acknowledge the need for stakeholder involvement in future research. Nonetheless, the study findings will significantly impact conservationists, PA managers, and decision-makers. By assessing the success of current management strategies and methods for addressing changes in threats to biodiversity, this research will improve our understanding of the effectiveness of PAs in addressing threats. Furthermore, the information gathered can serve as evidence to formulate regulations related to Protected Area Management and Evaluation (PAME) and identify PAs' main strengths and weaknesses in facing threats based on their geographical location and socio-economic-ecological characteristics. In addition, this review will provide significant findings for national decision-making. Limitations of the Evidence Base A significant portion of the evidence found relied on CI study designs, which were associated with the most frequent risk of bias. These biases were primarily due to confounding and selection bias, arising from the fact that they are often established in remote regions or areas of lower economic interest (Geldmann et al., 2025; Joppa & Pfaff, 2009; Margules & Pressey, 2000). This non-random placement introduces substantial limitations when comparing protected and unprotected sites, reducing the ability to isolate the true effect of protection. To address this challenge, matching approaches have been increasingly used to construct more comparable control groups and reduce bias in impact evaluations (Chen et al., 2023). In our review, CI studies that incorporated matching techniques exhibited lower bias risks compared to those that did not, demonstrating the value of such methodological adjustments in enhancing validity. Despite these improvements, most studies did not explicitly account for counterfactuals and thus were limited in their ability to fully address confounding variables. This omission can lead to overestimations of the effects of protection (Terraube et al., 2020). Given the inherent ecological, spatial, and socio-political heterogeneity of PAs, this remains a critical limitation in assessing conservation impact. These findings highlight the need for standardized spatial comparison protocols and counterfactual-based methodologies, to ensure that PA effectiveness is assessed with greater methodological rigor and cross-study comparability (Geldmann et al., 2025). Additionally, while our review could approximate the effectiveness of PAs against specific threats, we were unable to capture the complexity of simultaneous or interacting threats. Many studies examined individual pressures in isolation, whereas, in reality, PAs are often exposed to multiple, compounding human activities (Beebee & Griffiths, 2005; Hof et al., 2011; Menéndez-Guerrero & Graham, 2013; Stuart et al., 2004). For instance, the Qilian Mountain Nature Reserve in China was found to be effective in controlling population growth and land-use change, but ineffective in curbing infrastructure development, including road construction, mining, and hydropower projects (S. Li et al., 2022). Similarly, (Martínez-Fernández et al., 2015) found increasing pressure from urbanization both inside and outside PAs, while pressures from agriculture and deforestation were declining. This illustrates the importance of temporal dynamics in threat assessment, as current land uses may build upon earlier, unmeasured disturbances. Land-use transitions often occur sequentially, with past pressures shaping current vulnerability and management needs (Curtis et al., 2018). Review conclusions 5.1 Implications for Policy/Management This review provides evidence that PAs have played a role in mitigating threats to biodiversity, with most studies showing lower threat levels inside PAs compared to non-protected control sites. Although pressures such as deforestation, agriculture, infrastructure development, fire, and hunting continue to increase globally, PAs were generally associated with smaller increases or reductions in these threats relative to unprotected areas. However, the effectiveness of PAs is not consistent across all contexts. Outcomes are highly dependent on the type of threat and the broader context in which the PA operates. Better outcomes were associated with adequate funding, strong policy frameworks, effective enforcement, larger PA size and higher national development indices. In contrast, high accessibility, dense surrounding populations, and location in high-pressure environments were consistently linked to reduced effectiveness. Finally, meeting global biodiversity targets will require not only expanding PA networks but also ensuring that existing PAs are effective (G. Li et al., 2024) and capable of reducing threats. This includes investing in capacity building, enforcement, policy coherence, and local engagement to ensure that PAs function as effective tools for halting habitat loss and reducing anthropogenic pressures. Our review supports this by showing that well-managed PAs can achieve significantly better outcomes, even in high-pressure environments. Overall, our review confirms that while PAs can reduce biodiversity threats, their effectiveness is highly context-specific, and improving the tools and frameworks used to evaluate them is essential for guiding conservation strategies globally. 5.2 Implications for Research We identified several areas where the current evidence base on PA effectiveness could be strengthened, providing useful directions for future research. While most studies found that PAs were associated with lower threat levels compared to unprotected areas, their effectiveness was highly variable and strongly context-dependent. Factors such as location, baseline conditions, management capacity, and governance structures influenced outcomes but were often inconsistently reported or controlled for. These uncertainties highlight the importance of developing transparent and context-sensitive approaches to assessing PA effectiveness, particularly in studies comparing protected and unprotected areas. In this context, the use of rigorous counterfactual approaches such as matching methods that address potential confounders and control for the non-random location of PAs represents a promising avenue to improve impact attribution (Schleicher et al., 2019). Future research should prioritize the adoption of such causal inference techniques to better isolate the effectiveness of PAs on threat mitigation and their impact on conservation outcomes (Terraube et al., 2020). In addition, there is a need to improve consistency in how threats are measured. Several articles were excluded during the screening phase due to the absence of a defined or measurable threat assessment metric, highlighting the importance of clearly reported and comparable indicators across studies. In addition, there is a need to broaden the scope of data collection. While substantial progress has been made particularly through the use of remote sensing and LULCC analyses(Andam et al., 2008; Soares-Filho et al., 2023; Zhu et al., 2022) limited attention has been given to threats such as overexploitation, pollution and invasive species, which require in situ data collection and have been less included in large-scale PA evaluations (Hernández-Yáñez et al., 2016; Rodrigues & Cazalis, 2020; Schulze et al., 2018). Establishing robust, threat-specific metrics can enhance the consistency of assessments and enable more detailed insights into how PAs perform across different contexts. In addition, future research should move toward integrative and longitudinal study designs that can account for temporal dynamics and interactions between multiple threats. For example, areas that were previously deforested may now be experiencing urbanization, complicating the interpretation of PA effectiveness. Incorporating land-use history, threat transitions, and long-term monitoring would improve the accuracy of assessments. Moreover, greater standardization in the definition and selection of comparator areas, particularly regarding spatial scale and distance, would enhance the comparability and interpretability of PA evaluations. Finally, to ensure that findings are globally relevant, future research should aim to reduce geographic, linguistic, and thematic biases by increasing the inclusion of non-English language literature, underrepresented regions, and less frequently studied threat types (Amano et al., 2016). This will help build a more comprehensive, equitable, and policy-relevant understanding of PA effectiveness worldwide. While our review highlights the need to broaden linguistic and geographic coverage and to tackle understudied regions and threat types, it also reveals a stark evidence gap. We found only 105 studies worldwide that rigorously quantify the change of threat inside PAs. Given that there are more than 200,000 terrestrial PAs globally, this lack of data is alarming. Unless more impact assessment studies are conducted, PAs managers and policymakers will lack the data needed to allocate resources effectively or to adapt strategies. Only by filling this gap can we move beyond to a comprehensive understanding of how PAs deliver on their promise of biodiversity conservation. Declarations 6.1 Ethics approval and consent to participate Not applicable 6.2 Consent for publication Not applicable 6.3 Availability of data and materials All data generated or analyzed during this study are included in this published article [and its supplementary information files]. 6.4 Competing interests The authors declare that they have no competing interests 6.5 Funding K.P-C., H.F and J.G. were supported by The Danish Independent Research council (grant no. 0165-00018B) and E.V. by the Kone Foundation (grant no. 201803179). A.G. was supported by HORIZON MSCA Postdoctoral Fellowships (Project number: 101104696, THREATS). H.A.P was supported by HORIZON MSCA Postdoctoral Fellowship (Project number: 101063896, PROTECTEDTRADE). C.R. was supported by research grant 25925 from VILLUM FONDEN. 6.6 Authors' contributions KP-C and JG jointly conceived and designed the study. KPC, HAP, AG, HF, EV, and JG conducted the title and abstract screening and helped refine the inclusion criteria. KPC carried out the full-text screening, data extraction, and critical appraisal, with support from JG, who resolved uncertainties and verified the process. KPC conducted the analysis and drafted the systematic review. All authors provided feedback and revisions. All authors read and approved the final manuscript. References Ahmadia, G. N., Glew, L., Provost, M., Gill, D., Hidayat, N. I., Mangubhai, S., Purwanto, & Fox, H. E. (2015). Integrating impact evaluation in the design and implementation of monitoring marine protected areas. Philosophical Transactions of the Royal Society B: Biological Sciences , 370 (1681). https://doi.org/10.1098/RSTB.2014.0275 Amano, T., Berdejo-Espinola, V., Akasaka, M., de Andrade Junior, M. A. U., Blaise, N., Checco, J., Çilingir, F. G., Citegetse, G., Corella Tor, M., Drobniak, S. M., Giakoumi, S., Golivets, M., Ion, M. C., Jara-Díaz, J. P., Katayose, R., Lasmana, F. P. S., Lin, H. Y., Lopez, E., Mikula, P., … Zamora-Gutierrez, V. (2023). The role of non-English-language science in informing national biodiversity assessments. Nature Sustainability 2023 6:7 , 6 (7), 845–854. https://doi.org/10.1038/s41893-023-01087-8 Amano, T., & Sutherland, W. J. (2013). Four barriers to the global understanding of biodiversity conservation: wealth, language, geographical location and security. Proceedings of the Royal Society B: Biological Sciences , 280 (1756). https://doi.org/10.1098/RSPB.2012.2649 Andam, K. S., Ferraro, P. J., Pfaff, A., Sanchez-Azofeifa, G. A., & Robalino, J. A. (2008). Measuring the effectiveness of protected area networks in reducing deforestation. Proceedings of the National Academy of Sciences , 105 (42), 16089–16094. https://doi.org/10.1073/pnas.0800437105 Balmford, A., Carey, P., Kapos, V., Manica, A., Rodrigues, A. S. L., Scharlemann, J. P. W., & Green, R. E. (2009). Capturing the many dimensions of threat: comment on Salafsky et al. Conservation Biology : The Journal of the Society for Conservation Biology , 23 (2), 482–487. https://doi.org/10.1111/J.1523-1739.2009.01196.X Barnes, M. D., Craigie, I. D., Harrison, L. B., Geldmann, J., Collen, B., Whitmee, S., Balmford, A., Burgess, N. D., Brooks, T., Hockings, M., & Woodley, S. (2016). Wildlife population trends in protected areas predicted by national socio-economic metrics and body size. Nature Communications 2016 7:1 , 7 (1), 1–9. https://doi.org/10.1038/ncomms12747 Beebee, T. J. C., & Griffiths, R. A. (2005). The amphibian decline crisis: A watershed for conservation biology? Biological Conservation , 125 (3), 271–285. https://doi.org/10.1016/J.BIOCON.2005.04.009 Bleher, B., Uster, D., & Bergsdorf, T. (2006). Assessment of threat status and management effectiveness in Kakamega Forest, Kenya. Biodiversity and Conservation , 15 (4), 1159–1177. https://doi.org/10.1007/S10531-004-3509-3/METRICS Bowker, J. N., De Vos, A., Ament, J. M., & Cumming, G. S. (2017). Effectiveness of Africa’s tropical protected areas for maintaining forest cover. Conservation Biology , 31 (3), 559–569. https://doi.org/10.1111/COBI.12851 CBD. (2020). Global Biodiversity Outlook 5 . www.emdashdesign.ca CBD. (2022). Decision adopted by the conference of the parties to the Convention on Biological Diversity 15/4. Kunming-Montreal Global Biodiversity Framework . https://www.cbd.int/conferences/post20202CBD/WG8J/11/7,CBD/SBSTTA/23/9,CBD/SBSTTA/24/12andCBD/SBI/3/21,respectively. Chen, H., Zhang, T., Costanza, R., & Kubiszewski, I. (2023). Review of the approaches for assessing protected areas’ effectiveness. Environmental Impact Assessment Review , 98 . https://doi.org/10.1016/J.EIAR.2022.106929 Coetzee, B. W. T., Gaston, K. J., & Chown, S. L. (2014). Local Scale Comparisons of Biodiversity as a Test for Global Protected Area Ecological Performance: A Meta-Analysis. PLOS ONE , 9 (8), e105824. https://doi.org/10.1371/JOURNAL.PONE.0105824 Collaboration for Environmental Evidence. (2022). Guidelines and Standards for Evidence synthesis in Environmental Management: Vol. Version 5.1 (AS Pullin, GK Frampton, B Livoreil, & G Petrokofsky, Eds.). https://environmentalevidence.org/information-for-authors/ Curtis, P. G., Slay, C. M., Harris, N. L., Tyukavina, A., & Hansen, M. C. (2018). Classifying drivers of global forest loss. Science , 361 (6407), 1108–1111. https://doi.org/10.1126/SCIENCE.AAU3445/SUPPL_FILE/AAU3445_CURTIS_SM.PDF Durán, A. P., Barbosa, O., & Gaston, K. J. (2022). Understanding the interacting factors that determine ecological effectiveness of terrestrial protected areas. In Journal for Nature Conservation (Vol. 70, p. 126264). Urban & Fischer. https://doi.org/10.1016/j.jnc.2022.126264 Eklund, J., Jones, J. P. G., Räsänen, M., Geldmann, J., Jokinen, A. P., Pellegrini, A., Rakotobe, D., Rakotonarivo, O. S., Toivonen, T., & Balmford, A. (2022). Elevated fires during COVID-19 lockdown and the vulnerability of protected areas. Nature Sustainability 2022 5:7 , 5 (7), 603–609. https://doi.org/10.1038/s41893-022-00884-x Geldmann, J. (2023). Safeguarding biodiversity requires understanding how to manage protected areas cost effectively. One Earth , 6 (2), 73–76. https://doi.org/10.1016/j.oneear.2023.01.008 Geldmann, J., Barnes, M., Coad, L., Craigie, I. D., Hockings, M., & Burgess, N. D. (2013). Effectiveness of terrestrial protected areas in reducing habitat loss and population declines. In Biological Conservation (Vol. 161, pp. 230–238). https://doi.org/10.1016/j.biocon.2013.02.018 Geldmann, J., Jones, J. P. G., Wauchope, H., & Ferraro, P. J. (2025). Causal claims, causal assumptions and protected area impact. Nature 2025 638:8052 , 638 (8052), E40–E41. https://doi.org/10.1038/s41586-024-08512-8 Geldmann, J., Joppa, L. N., & Burgess, N. D. (2014). Mapping Change in Human Pressure Globally on Land and within Protected Areas. Conservation Biology , 28 (6), 1604–1616. https://doi.org/10.1111/COBI.12332 Geldmann, J., Manica, A., Burgess, N. D., Coad, L., & Balmford, A. (2019). A global-level assessment of the effectiveness of protected areas at resisting anthropogenic pressures. Proceedings of the National Academy of Sciences of the United States of America , 116 (46), 23209–23215. https://doi.org/10.1073/PNAS.1908221116/SUPPL_FILE/PNAS.1908221116.SAPP.PDF Gonson, C., Pelletier, D., Alban, F., Giraud-Carrier, C., & Ferraris, J. (2017). Influence of settings management and protection status on recreational uses and pressures in marine protected areas. Journal of Environmental Management , 200 , 170–185. https://doi.org/10.1016/J.JENVMAN.2017.05.051 Graham, V., Geldmann, J., Adams, V. M., Grech, A., Deinet, S., & Chang, H. C. (2021). Management resourcing and government transparency are key drivers of biodiversity outcomes in Southeast Asian protected areas. Biological Conservation , 253 , 108875. https://doi.org/10.1016/J.BIOCON.2020.108875 Gurevitch, J., Koricheva, J., Nakagawa, S., & Stewart, G. (2018). Meta-analysis and the science of research synthesis. Nature 2018 555:7695 , 555 (7695), 175–182. https://doi.org/10.1038/nature25753 Hernández-Yáñez, H., Kos, J. T., Bast, M. D., Griggs, J. L., Hage, P. A., Killian, A., Loza, M. I., Whitmore, M. B., & Smith, A. B. (2016). A systematic assessment of threats affecting the rare plants of the United States. Biological Conservation , 203 , 260–267. https://doi.org/10.1016/J.BIOCON.2016.10.009 Hirons, S., Matilda Collins, C., & Singh, M. (2022). Assessing variation in the effectiveness of IUCN protected area categorisation. What remotely sensed forest integrity and human modification reveals across the major tropical forest biomes. Ecological Indicators , 143 , 109337. https://doi.org/10.1016/J.ECOLIND.2022.109337 Hof, C., Araújo, M. B., Jetz, W., & Rahbek, C. (2011). Additive threats from pathogens, climate and land-use change for global amphibian diversity. Nature 2011 480:7378 , 480 (7378), 516–519. https://doi.org/10.1038/nature10650 IPBES. (2019). Global assessment report on biodiversity and ecosystem services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services . https://doi.org/10.5281/ZENODO.6417333 IUCN. (2024). IUCN - Threats Classification Scheme (Version 3.3) . https://www.iucnredlist.org/resources/threat-classification-scheme Jones, K. R., Venter, O., Fuller, R. A., Allan, J. R., Maxwell, S. L., Negret, P. J., & Watson, J. E. M. (2018). One-third of global protected land is under intense human pressure. Science , 360 (6390), 788–791. https://doi.org/10.1126/SCIENCE.AAP9565/SUPPL_FILE/AAP9565_JONES_SM.PDF Joppa, L. N., & Pfaff, A. (2009). High and Far: Biases in the Location of Protected Areas. PLOS ONE , 4 (12), e8273. https://doi.org/10.1371/JOURNAL.PONE.0008273 Joppa, L. N., & Pfaff, A. (2011). Global protected area impacts. Proceedings of the Royal Society B: Biological Sciences , 278 (1712), 1633–1638. https://doi.org/10.1098/RSPB.2010.1713 Kennedy, C. M., Oakleaf, J. R., Theobald, D. M., Baruch-Mordo, S., & Kiesecker, J. (2019). Managing the middle: A shift in conservation priorities based on the global human modification gradient. Global Change Biology , 25 (3), 811–826. https://doi.org/10.1111/GCB.14549 Konno, K., Livoreil, B., & Pullin, A. (2021). Collaboration for Environmental Evidence Critical Appraisal Tool Version 0.3 (Prototype) . Lawrence, A., Friedrich, F., & Beierkuhnlein, C. (2021). Landscape fragmentation of the Natura 2000 network and its surrounding areas. PLOS ONE , 16 (10), e0258615. https://doi.org/10.1371/JOURNAL.PONE.0258615 Leclère, D., Obersteiner, M., Barrett, M., Butchart, S. H. M., Chaudhary, A., de Palma, A., DeClerck, F. A. J., di Marco, M., Doelman, J. C., Dürauer, M., Freeman, R., Harfoot, M., Hasegawa, T., Hellweg, S., Hilbers, J. P., Hill, S. L. L., Humpenöder, F., Jennings, N., Krisztin, T., … Young, L. (2020). Bending the curve of terrestrial biodiversity needs an integrated strategy. Nature 2020 585:7826 , 585 (7826), 551–556. https://doi.org/10.1038/s41586-020-2705-y Li, G., Fang, C., Watson, J. E. M., Sun, S., Qi, W., Wang, Z., & Liu, J. (2024). Mixed effectiveness of global protected areas in resisting habitat loss. Nature Communications 2024 15:1 , 15 (1), 1–17. https://doi.org/10.1038/s41467-024-52693-9 Li, S., Su, S., Liu, Y., Zhou, X., Luo, Q., & Paudel, B. (2022). Effectiveness of the Qilian Mountain Nature Reserve of China in Reducing Human Impacts. Land 2022, Vol. 11, Page 1071 , 11 (7), 1071. https://doi.org/10.3390/LAND11071071 Mammides, C. (2020). A global assessment of the human pressure on the world’s lakes. Global Environmental Change , 63 , 102084. https://doi.org/10.1016/J.GLOENVCHA.2020.102084 Margules, C. R., & Pressey, R. L. (2000). Systematic conservation planning. Nature 2000 405:6783 , 405 (6783), 243–253. https://doi.org/10.1038/35012251 Martínez-Fernández, J., Ruiz-Benito, P., & Zavala, M. A. (2015). Recent land cover changes in Spain across biogeographical regions and protection levels: Implications for conservation policies. Land Use Policy , 44 , 62–75. https://doi.org/10.1016/J.LANDUSEPOL.2014.11.021 Menéndez-Guerrero, P. A., & Graham, C. H. (2013). Evaluating multiple causes of amphibian declines of Ecuador using geographical quantitative analyses. Ecography , 36 (7), 756–769. https://doi.org/10.1111/J.1600-0587.2012.07877.X Mingarro, M., & Lobo, J. M. (2023). European National Parks protect their surroundings but not everywhere: A study using land use/land cover dynamics derived from CORINE Land Cover data. Land Use Policy , 124 , 106434. https://doi.org/10.1016/J.LANDUSEPOL.2022.106434 Mollmann, V. H. D. S., Santos, S., Fernandes, G., Mossolin, E. C., Dalosto, M. M., Cardoso, S. M. V. S., Prestes, O. D., Zanella, R., & Bartholomei-Santos, M. L. (2022). Terrestrial protected areas do not fully shield their streams from exogenous stressors. Environmental Conservation , 49 (4), 215–224. https://doi.org/10.1017/S0376892922000261 Pulido-Chadid, K., Virtanen, E., & Geldmann, J. (2023). How effective are protected areas for reducing threats to biodiversity? A systematic review protocol. Environmental Evidence , 12 (1), 1–10. https://doi.org/10.1186/S13750-023-00311-4/TABLES/5 Pullin, A. S., & Stewart, G. B. (2006). Guidelines for systematic review in conservation and environmental management. Conservation Biology : The Journal of the Society for Conservation Biology , 20 (6), 1647–1656. https://doi.org/10.1111/J.1523-1739.2006.00485.X R Core Team. (2021). R: A Language and Environment for Statistical Computing . https://www.r-project.org/ Richardson, K., Steffen, W., Lucht, W., Bendtsen, J., Cornell, S. E., Donges, J. F., Drüke, M., Fetzer, I., Bala, G., von Bloh, W., Feulner, G., Fiedler, S., Gerten, D., Gleeson, T., Hofmann, M., Huiskamp, W., Kummu, M., Mohan, C., Nogués-Bravo, D., … Rockström, J. (2023). Earth beyond six of nine planetary boundaries. Science Advances , 9 (37). https://doi.org/10.1126/SCIADV.ADH2458 Rodrigues, A. S. L., & Cazalis, V. (2020). The multifaceted challenge of evaluating protected area effectiveness. Nature Communications 2020 11:1 , 11 (1), 1–4. https://doi.org/10.1038/s41467-020-18989-2 Salafsky, N., Relton, C., Young, B. E., Lamarre, P., Böhm, M., Chénier, M., Cochrane, E., Dionne, M., He, K. K., Hilton-Taylor, C., Latrémouille, C., Morrison, J., Raymond, C. V., Seddon, M., & Suresh, V. (2025). Classification of direct threats to the conservation of ecosystems and species 4.0. Conservation Biology , e14434. https://doi.org/10.1111/COBI.14434 Sanderson, E. W., Jaiteh, M., Levy, M. A., Redford, K. H., Wannebo, A. V, & Woolmer, G. (2002). The Human Footprint and the Last of the Wild: The human footprint is a global map of human influence on the land surface, which suggests that human beings are stewards of nature, whether we like it or not. BioScience , 52 (10), 891–904. https://doi.org/10.1641/0006-3568(2002)052[0891:THFATL]2.0.CO;2 Schleicher, J., Eklund, J., D. Barnes, M., Geldmann, J., Oldekop, J. A., & Jones, J. P. G. (2020). Statistical matching for conservation science. Conservation Biology , 34 (3), 538–549. https://doi.org/10.1111/COBI.13448 Schleicher, J., Peres, C. A., & Leader-Williams, N. (2019). Conservation performance of tropical protected areas: How important is management? Conservation Letters , 12 (5), e12650. https://doi.org/10.1111/CONL.12650 Schulze, K., Knights, K., Coad, L., Geldmann, J., Leverington, F., Eassom, A., Marr, M., Butchart, S. H. M., Hockings, M., & Burgess, N. D. (2018). An assessment of threats to terrestrial protected areas. Conservation Letters , 11 (3). https://doi.org/10.1111/CONL.12435 Soares-Filho, B. S., Oliveira, U., Ferreira, M. N., Marques, F. F. C., de Oliveira, A. R., Silva, F. R., & Börner, J. (2023). Contribution of the Amazon protected areas program to forest conservation. Biological Conservation , 279 , 109928. https://doi.org/https://doi.org/10.1016/j.biocon.2023.109928 Stuart, S. N., Chanson, J. S., Cox, N. A., Young, B. E., Rodrigues, A. S. L., Fischman, D. L., & Waller, R. W. (2004). Status and trends of amphibian declines and extinctions worldwide. Science , 306 (5702), 1783–1786. https://doi.org/10.1126/SCIENCE.1103538/SUPPL_FILE/STUART.SOM.PDF Terraube, J., Van doninck, J., Helle, P., & Cabeza, M. (2020). Assessing the effectiveness of a national protected area network for carnivore conservation. Nature Communications 2020 11:1 , 11 (1), 1–9. https://doi.org/10.1038/s41467-020-16792-7 The EndNote Team. (2013). EndNote (EndNote 20 Version). Clarivate. UNEP-WCMC. (2024, May). Explore the World’s Protected Areas . https://www.protectedplanet.net/en Venter, O., Sanderson, E. W., Magrach, A., Allan, J. R., Beher, J., Jones, K. R., Possingham, H. P., Laurance, W. F., Wood, P., Fekete, B. M., Levy, M. A., & Watson, J. E. M. (2016a). Global terrestrial Human Footprint maps for 1993 and 2009. Scientific Data 2016 3:1 , 3 (1), 1–10. https://doi.org/10.1038/sdata.2016.67 Venter, O., Sanderson, E. W., Magrach, A., Allan, J. R., Beher, J., Jones, K. R., Possingham, H. P., Laurance, W. F., Wood, P., Fekete, B. M., Levy, M. A., & Watson, J. E. M. (2016b). Sixteen years of change in the global terrestrial human footprint and implications for biodiversity conservation. Nature Communications 2016 7:1 , 7 (1), 1–11. https://doi.org/10.1038/ncomms12558 Watson, J. E. M., Dudley, N., Segan, D. B., & Hockings, M. (2014). The performance and potential of protected areas. Nature 2014 515:7525 , 515 (7525), 67–73. https://doi.org/10.1038/nature13947 Wauchope, H. S., Jones, J. P. G., Geldmann, J., Simmons, B. I., Amano, T., Blanco, D. E., Fuller, R. A., Johnston, A., Langendoen, T., Mundkur, T., Nagy, S., & Sutherland, W. J. (2022). Protected areas have a mixed impact on waterbirds, but management helps. Nature 2022 605:7908 , 605 (7908), 103–107. https://doi.org/10.1038/s41586-022-04617-0 Wu, H., Fang, S., Yang, Y., & Cheng, J. (2022). Changes in habitat quality of nature reserves in depopulating areas due to anthropogenic pressure: Evidence from Northeast China, 2000–2018. Ecological Indicators , 138 , 108844. https://doi.org/10.1016/J.ECOLIND.2022.108844 Xu, Y., Price, M., Yang, B., Zhang, K., Yang, N., Tang, X., Ran, J., Yi, Y., & Wang, B. (2022). Have China’s national forest reserves designated since 1990 conserved forests effectively? Journal of Environmental Management , 306 , 114485. https://doi.org/10.1016/J.JENVMAN.2022.114485 Zhu, Z., Qiu, S., & Ye, S. (2022). Remote sensing of land change: A multifaceted perspective. Remote Sensing of Environment , 282 , 113266. https://doi.org/10.1016/J.RSE.2022.113266 Additional Declarations No competing interests reported. Supplementary Files Additionalfile1.Excludedpublicationsfulltext.xlsx Additional File 1. Excluded publications at full text with reasons Additionalfile2.Dataextractionandcriticalappraisal.xlsx Additional File 2. Data extraction and critical appraisal Additionalfile3.Datacodingandextractionstrategy.docx Additional File 3. Data coding and extraction strategies Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 06 Sep, 2025 Reviews received at journal 06 Sep, 2025 Reviewers agreed at journal 06 Sep, 2025 Reviews received at journal 28 Jul, 2025 Reviewers agreed at journal 18 Jul, 2025 Reviewers agreed at journal 17 Jul, 2025 Reviewers invited by journal 12 Jul, 2025 Editor assigned by journal 07 Jul, 2025 Submission checks completed at journal 07 Jul, 2025 First submitted to journal 04 Jul, 2025 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-7046705","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":482208066,"identity":"2ea45312-85aa-41a1-998d-9e49433202b8","order_by":0,"name":"Katherine Pulido-Chadid","email":"data:image/png;base64,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","orcid":"","institution":"University of Copenhagen","correspondingAuthor":true,"prefix":"","firstName":"Katherine","middleName":"","lastName":"Pulido-Chadid","suffix":""},{"id":482208067,"identity":"f21bf9cb-0a8d-4669-aabe-819a89c4b73b","order_by":1,"name":"Helena N. Alves-Pinto","email":"","orcid":"","institution":"University of Copenhagen","correspondingAuthor":false,"prefix":"","firstName":"Helena","middleName":"N.","lastName":"Alves-Pinto","suffix":""},{"id":482208068,"identity":"cb429ede-b529-497a-a016-3fe1f7771c27","order_by":2,"name":"Harith Farooq","email":"","orcid":"","institution":"University of Copenhagen","correspondingAuthor":false,"prefix":"","firstName":"Harith","middleName":"","lastName":"Farooq","suffix":""},{"id":482208069,"identity":"ec9ebb18-c0e2-413a-9ab2-093536b1048e","order_by":3,"name":"Antonella Gorosábel","email":"","orcid":"","institution":"University of Copenhagen","correspondingAuthor":false,"prefix":"","firstName":"Antonella","middleName":"","lastName":"Gorosábel","suffix":""},{"id":482208070,"identity":"663c102c-ab30-4cee-99a1-8bcb614b7e26","order_by":4,"name":"Elina A. Virtanen","email":"","orcid":"","institution":"University of Helsinki","correspondingAuthor":false,"prefix":"","firstName":"Elina","middleName":"A.","lastName":"Virtanen","suffix":""},{"id":482208071,"identity":"f2517925-adc6-48e5-9528-16a8249f1d95","order_by":5,"name":"Neil D. Burgess","email":"","orcid":"","institution":"University of Copenhagen","correspondingAuthor":false,"prefix":"","firstName":"Neil","middleName":"D.","lastName":"Burgess","suffix":""},{"id":482208072,"identity":"60adc280-2ed2-4e60-bb5f-f45fddb3a1bd","order_by":6,"name":"Carsten Rahbek","email":"","orcid":"","institution":"University of Copenhagen","correspondingAuthor":false,"prefix":"","firstName":"Carsten","middleName":"","lastName":"Rahbek","suffix":""},{"id":482208073,"identity":"21a41403-fd75-4fa4-b659-6c462acb96a0","order_by":7,"name":"Jonas Geldmann","email":"","orcid":"","institution":"University of Copenhagen","correspondingAuthor":false,"prefix":"","firstName":"Jonas","middleName":"","lastName":"Geldmann","suffix":""}],"badges":[],"createdAt":"2025-07-04 12:23:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7046705/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7046705/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":86487148,"identity":"c3750525-0536-4835-ac18-cfa62afe1a12","added_by":"auto","created_at":"2025-07-11 08:26:20","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":109636,"visible":true,"origin":"","legend":"\u003cp\u003eROSES flow diagram illustrating the selection process of articles for the systematic review.\u003c/p\u003e","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7046705/v1/3f7c34c171ccd6278e8891d3.png"},{"id":86487821,"identity":"1269e30e-1853-49e6-a930-6dd39f0e3033","added_by":"auto","created_at":"2025-07-11 08:34:20","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":232342,"visible":true,"origin":"","legend":"\u003cp\u003eFrequency of countries involved in the extracted studies (8 records with global studies and records with multiple countries involved are not included in the map).\u003c/p\u003e","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7046705/v1/802a97e043bb400704153b53.png"},{"id":86488825,"identity":"9a4e9c51-d692-45e3-8013-d7da84b3cd6b","added_by":"auto","created_at":"2025-07-11 08:42:20","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":237920,"visible":true,"origin":"","legend":"\u003cp\u003eFrequency of comparator types used across different study designs in 105 reviewed articles. BA = Before–After; CI = Control–Intervention (Without matching); CI (Matching) = Control–Intervention (With matching); and BACI = Before–After Control–Intervention.\u003c/p\u003e","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7046705/v1/9b4b330dea573b4f819f3e35.png"},{"id":86487150,"identity":"9c125a8e-64d7-4272-8399-315c0d719849","added_by":"auto","created_at":"2025-07-11 08:26:20","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":83998,"visible":true,"origin":"","legend":"\u003cp\u003eCritical appraisal of the included articles. \u003cstrong\u003ePanel A \u003c/strong\u003esummarizes the main critical appraisal criteria. \u003cstrong\u003ePanel B\u003c/strong\u003e outlines reasons for confounding and selection bias. Confounding bias was assessed based on the likelihood of the intervention of the study to be confounded, if relevant confounders were controlled, and whether omissions were justified. Selection bias was evaluated based on whether study subjects or areas were randomly or systematically selected before the intervention. \u003cstrong\u003ePanel C s\u003c/strong\u003ehows the risk of confounding and selection bias across study designs.\u003c/p\u003e","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7046705/v1/0c3929b986de71203af91480.png"},{"id":86487152,"identity":"30cf4428-06ba-4b88-a310-f910913c086a","added_by":"auto","created_at":"2025-07-11 08:26:20","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":32149,"visible":true,"origin":"","legend":"\u003cp\u003eChanges in threat levels inside and outside PAs. (A) Proporiton of observations reporting changes in threat levels inside PAs, (B) outside PAs, and (C) comparisons between threat levels inside and outside. From the 117 studies where data was available for each question, we counted multiple threats as independent observations.\u003c/p\u003e","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7046705/v1/709cd632a8d42249592da5e6.png"},{"id":86487155,"identity":"90828e4e-9a80-4418-aa4b-a25603653690","added_by":"auto","created_at":"2025-07-11 08:26:20","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":75664,"visible":true,"origin":"","legend":"\u003cp\u003eChanges in threat levels inside and outside PAs by threat type. (A) Number of observations reporting changes in threat levels inside PAs, (B) outside PAs, and (C) comparisons between threat levels inside and outside. The length of each bar indicates the number of cases for which data were available, while the percentage values within the bars represent the proportion of cases attributed to each threat type.\u003c/p\u003e","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7046705/v1/518104384f95e4a873fd6cc0.png"},{"id":86487826,"identity":"e0c77285-8239-44ca-b5d8-df522e542314","added_by":"auto","created_at":"2025-07-11 08:34:20","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":53898,"visible":true,"origin":"","legend":"\u003cp\u003eEffects of drivers influencing PAs effectiveness, as synthesized from 103 studies. The bars represent the number of studies reporting each factor with the color and direction indicating whether the specific driver was associated with improved effectiveness (green) no effect (grey) or reduced effectiveness (orange).\u003c/p\u003e","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7046705/v1/085b478748c8a277cd5ba58b.png"},{"id":86490355,"identity":"4709ac1d-cd5a-4963-9e3e-88839f78f25e","added_by":"auto","created_at":"2025-07-11 08:58:22","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2001628,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7046705/v1/e8feb924-e59f-4a6f-922b-5d5adfcb2cdb.pdf"},{"id":86487822,"identity":"8b1d7a96-0305-453f-9127-cbe796c0793e","added_by":"auto","created_at":"2025-07-11 08:34:20","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":280444,"visible":true,"origin":"","legend":"\u003cp\u003eAdditional File 1. Excluded publications at full text with reasons\u003c/p\u003e","description":"","filename":"Additionalfile1.Excludedpublicationsfulltext.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7046705/v1/3671c63da8c47c9d787bc82a.xlsx"},{"id":86487162,"identity":"f2df8f20-ce5b-4068-9e25-2d9ee535d2d8","added_by":"auto","created_at":"2025-07-11 08:26:20","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":186428,"visible":true,"origin":"","legend":"\u003cp\u003eAdditional File 2. Data extraction and critical appraisal\u003c/p\u003e","description":"","filename":"Additionalfile2.Dataextractionandcriticalappraisal.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7046705/v1/8eddb88af806b9ad2a421f36.xlsx"},{"id":86487159,"identity":"06c30ef7-6e0c-4a97-b704-c20b16be2e27","added_by":"auto","created_at":"2025-07-11 08:26:20","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":49066,"visible":true,"origin":"","legend":"\u003cp\u003eAdditional File 3. Data coding and extraction strategies\u003c/p\u003e","description":"","filename":"Additionalfile3.Datacodingandextractionstrategy.docx","url":"https://assets-eu.researchsquare.com/files/rs-7046705/v1/f7f34e86b987e50004f9d6c1.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"A systematic review of the effectiveness of protected areas at reducing threats to biodiversity","fulltext":[{"header":"Background","content":"\u003cp\u003eCurrently, we are facing an unprecedented biodiversity crisis, with an increasing number of species facing extinction as we exceed the safe operating boundaries for biosphere integrity (Richardson et al., 2023). These planetary challenges are driven by a series of threats that affect biodiversity (IPBES, 2019). Direct threats are posed by human actions that cause degradation or loss of ecosystems and species populations, such as agriculture, infrastructure development, resource extraction, and ecosystem management. Additionally, threats encompass disturbances in natural systems whose dynamics have been altered by past or present human activity, including invasive species, pollution, natural disasters, and climate change (Salafsky et al., 2025). Ultimately, threats to biodiversity are shaped by indirect drivers rooted in societal values and behaviors, including patterns of production and consumption, trade practices, technological innovations, and governance structures, all of which vary across regions and contexts (IPBES, 2019).\u003c/p\u003e\u003cp\u003eProtected areas (PAs) are among the most widely used tools for nature conservation, which aim to reduce both indirect and direct threats to biodiversity (Rodrigues \u0026amp; Cazalis, 2020; Watson et al., 2014). International agreements such as the Convention on Biological Diversity (CBD) have consistently called for increased PA coverage and improved management, resulting in expanded protected land during recent decades. As of 2024, approximately 17% of terrestrial ecosystems and less than 10% of marine ecosystems were covered by PAs or other effective area-based conservation measures (OECMs) (UNEP-WCMC, 2024). Despite the rising numbers and the global target of 30% coverage expected by 2030, as outlined in the Kunming-Montreal Global Biodiversity Framework (CBD, 2022), biodiversity loss continues, and human pressures inside and outside PAs are increasing (CBD, 2020; Geldmann et al., 2019; Jones et al., 2018; Lecl\u0026egrave;re et al., 2020). As a result, the Kunming-Montreal Framework, aligned with the 2030 Agenda for Sustainable Development, emphasizes not only expanding PAs but also directly addressing the key threats driving biodiversity loss, as stated in Targets 1 to 8 (CBD, 2022).\u003c/p\u003e\u003cp\u003eThe ecological effectiveness of PAs depends on a combination of factors, including decisions made at the time of their establishment, such as location, extent, design, connectivity, and representativeness, as well as subsequent governance and management practices (Rodrigues \u0026amp; Cazalis, 2020). However, assessing their impact remains challenging, as PAs are embedded within complex socio-ecological systems where multiple interrelated factors influence the outcomes (Dur\u0026aacute;n et al., 2022; Geldmann et al., 2025). Research shows that PAs generally reduce biodiversity decline compared to unprotected sites (Coetzee et al., 2014; Geldmann et al., 2013). Nonetheless, their effectiveness varies significantly depending on contextual factors, including socio-economic conditions and management regimes (Graham et al., 2021; Wauchope et al., 2022). For example, wildlife populations tend to be higher in PAs located in regions with stronger development indicators (Barnes et al., 2016).\u003c/p\u003e\u003cp\u003eAnother major challenge in evaluating PA impact is the presence of inherent location biases. Many PAs are established in remote areas with limited economic interest, making direct comparisons with unprotected lands problematic (Joppa \u0026amp; Pfaff, 2011). To address this limitation, studies have adopted counterfactual approaches, comparing outcomes before and after PA establishment and/or against comparable unprotected areas. These methods allow for more robust estimates of the ecological effects of PAs by modeling what would have occurred in their absence (Chen et al., 2023; Schleicher et al., 2020; Wauchope et al., 2022).\u003c/p\u003e\u003cp\u003eThe long-term conservation success of PAs relies not only on their initial configuration but also on their capacity to mitigate threats through effective management practices (Geldmann, 2023; Margules \u0026amp; Pressey, 2000; Rodrigues \u0026amp; Cazalis, 2020). PAs have been established in response to threats to biodiversity, but these threats are also the catalysts for creating management strategies for PAs. Analyzing how effectively PAs and their management strategies reduce threats can thus provide essential insights into the broader management effectiveness of PAs. Given the urgent need for effective PAs capable of reducing threats and countering ongoing biodiversity loss, we aim to identify whether and how PAs have succeeded in threat reduction, and what characteristics have contributed to or hindered their effectiveness by conducting a systematic literature review.\u003c/p\u003e\u003cdiv id=\"Sec2\" class=\"Section2\"\u003e\u003ch2\u003e2.1 Objective of the Review\u003c/h2\u003e\u003cp\u003eA key objective of establishing PAs is to reduce the pressures that threaten biodiversity on unprotected lands and seas. This systematic review aims to test that premise by assessing the effectiveness of PAs in reducing threats to biodiversity through a synthesis of evidence from peer-reviewed and grey literature across all available anthropogenic threat types, regions, habitats, and study scales. To achieve this, we evaluate different measures of impact, incorporating spatial, temporal, and management intervention comparisons. The focus of this review is on how threats to biodiversity change in relation to protection status, evolve over time, and/or respond to implemented intervention strategies. Our main goal is to assess the effectiveness of PAs and associated interventions in reducing threats to biodiversity. In addition, we examine how effectiveness varies across different threat types and identify key factors contributing to conservation outcomes.\u003c/p\u003e\u003c/div\u003e"},{"header":"Methods","content":"\u003cp\u003e This systematic review followed the Guidelines and Standards for Systematic Reviews of the Collaboration for Environmental Evidence (Collaboration for Environmental Evidence, 2022). The methods of this review adhered to a published peer-reviewed systematic review protocol (Pulido-Chadid et al., 2023) with a few minor deviations (see section 3.1).\u003c/p\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e3.1 Deviations from the protocol\u003c/h2\u003e\u003cp\u003eSpecifically, we explicitly excluded publications related to climate change, given the inherent challenges in accurately measuring its effects on PAs. Additionally, literature reviews were excluded to maintain a focus on primary sources of research.\u003c/p\u003e\u003cp\u003eTo optimize the review process, data extraction was conducted by a single reviewer (KP). Any uncertainties that arose were referred to a senior scientist for resolution, ensuring alignment with the inclusion criteria and the review\u0026rsquo;s objectives. Additionally, a random 10% subset of articles was independently reviewed by JG. Disagreements were minimal and were discussed and resolved through consensus.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003e3.2 Search for articles\u003c/h2\u003e\u003cdiv id=\"Sec6\" class=\"Section3\"\u003e\u003ch2\u003e3.2.1 Search sources, terms and strings:\u003c/h2\u003e\u003cp\u003eWe conducted searches in Scopus and the Web of Science Core Collection bibliographic databases using the search string and search blocks (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) adopted from Pulido-Chadid et al. (2023). The bibliographic database searches were conducted on the 14 of August of 2023. Additionally, we searched for grey literature using Google Scholar and the websites of the Wildlife Conservation Society (WCS), International Union for Conservation of Nature (IUCN), Global Environmental Facility (GEF), World Wide Fund for Nature (WWF), Conservation International, and BirdLife International.\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\u003eSearch string that will be used in the search of the literature in the Web of Science Core Collection and Scopus\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"2\"\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\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u003cp\u003eTopic Search string\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eThreats to biodiversity\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003ethreat\u003cspan\u003e$\u003c/span\u003e OR \"human impact*\" OR \"human pressure*\" OR\u0026rdquo; anthropogenic impact*\" OR \"human activity\" OR stressor OR \u0026ldquo;anthropogenic pressure\u0026rdquo; AND\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMeasure the change of threats\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003ereduc* OR effective* OR impact AND\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eProtected areas\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\"protected area*\" OR \"conserv* area\" OR \"nature reserve\" OR sanctuar* OR \"national park*\" OR \"biosphere reserve*\" OR \"biodiversity reserve*\" OR \"wildlife habitat*\"\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eAdopted from: Pulido-Chadid et al., 2023.\u003c/p\u003e\u003cp\u003eFor grey literature, we adapted our approach based on available filtering options. On organizational websites, we used \u0026ldquo;protected areas\u0026rdquo; as the primary keyword and manually screened relevant literature. In Google Scholar, we searched for \u0026ldquo;protected area\u0026rdquo; AND \u0026ldquo;threats\u0026rdquo; and reviewed the first 50 results, as specified in the protocol (Pulido-Chadid et al. 2023). The keywords, search dates, websites accessed, and retrieved information are detailed in Additional File 2 from Pulido-Chadid et al. 2023.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec7\" class=\"Section3\"\u003e\u003ch2\u003e3.2.2 Search limitations\u003c/h2\u003e\u003cp\u003eNo restrictions were applied regarding the year of publication or document type of the articles, with reviews removed as part of the document screening process. Our searches were conducted in English. While we acknowledge that publications in other languages may be relevant (Amano et al., 2023; Amano \u0026amp; Sutherland, 2013), particularly given the diverse geographical scope of this review, non-English publications were not included, which may limit the representation of findings from certain regions.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec8\" class=\"Section3\"\u003e\u003ch2\u003e3.2.3 Estimating the comprehensiveness of the search\u003c/h2\u003e\u003cp\u003eAt the protocol stage, we tested the comprehensiveness of the search strings by assessing the search strategy comprehensiveness of 20 benchmark articles (see Pulido-Chadid et al., 2023).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec9\" class=\"Section3\"\u003e\u003ch2\u003e3.2.4 Search results\u003c/h2\u003e\u003cp\u003eAll records, including their bibliographic information, were exported to EndNote 20 (The EndNote Team, 2013), where deduplication removal was performed. The unique records were then exported to Covidence (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.covidence.org/home\u003c/span\u003e\u003cspan address=\"https://www.covidence.org/home\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), where some additional duplicates were identified and removed.\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e3.3 Article screening and study eligibility criteria\u003c/h2\u003e\u003cdiv id=\"Sec11\" class=\"Section3\"\u003e\u003ch2\u003e3.3.1 Screening process\u003c/h2\u003e\u003cp\u003e Prior to the conclusive screening process, we conducted an initial consistency check to ensure that the eligibility criteria were interpreted and applied uniformly across reviewers. To achieve this, we randomly selected 600 abstracts to ensure sufficient representation of the identified literature. These abstracts were screened in a double-blinded process and independently by the reviewers (KP, AG, HAP, HF, EV, JG). To advance to full-text screening, we required a moderate agreement threshold (Kappa\u0026thinsp;\u0026ge;\u0026thinsp;0.6). Any discrepancies were discussed and resolved to improve consistency, and if necessary, the inclusion and exclusion criteria were refined.\u003c/p\u003e\u003cp\u003eThis process was conducted in three iterative rounds. In the first round, the 600 random subset was divided into two groups, one with two reviewers and another with four reviewers. We observed that the agreement was slightly higher with fewer reviewers, with a Kappa value of 0.42 in the two-reviewer group compared to 0.405 in the four-reviewer group. Discrepancies were analyzed, and refinements were made to the eligibility criteria. Following this, an intermediary step was introduced to further clarify the inclusion criteria. Two smaller tests were conducted, where a random selection of 15 abstracts was screened and carefully discussed twice among all reviewers. This resulted in an improvement in the Kappa value from 0.44 in the first test to 0.51 in the second round. In the final assessment, a new random subset of 675 abstracts was screened. These were divided into three groups of two reviewers, achieving the benchmark Kappa score of 0.6. Once a sufficient agreement was achieved, each article was categorized into one of three groups: (1) include, (2) exclude, or (3) maybe. As it was conducted in the consistency check, the \u0026ldquo;maybes\u0026rdquo; were then reevaluated by the lead author. Then full title and abstract screening process was conducted for all remaining articles among reviewers.\u003c/p\u003e\u003cp\u003eThe full-text analysis of all articles selected in the first step was screened first by the lead author. Any uncertainties that arose during the extraction were continuously referred to and discussed with the senior scientist for resolution, ensuring alignment with the inclusion criteria and the review\u0026rsquo;s objectives. Additionally, a randomly selected, blinded subset of 10% of articles was independently reviewed by the senior scientist. Disagreements were minimal and were resolved through consensus.\u003c/p\u003e\u003cp\u003eA list of articles and grey literature excluded during full-text screening, along with reasons for exclusion, is provided in \u003cem\u003eAdditional File 1\u003c/em\u003e. Articles excluded from the first stage (title and abstract screening) were not coded with reasons for exclusion, but are available.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section3\"\u003e\u003ch2\u003e3.3.2 Eligibility criteria\u003c/h2\u003e\u003cp\u003eTo qualify for this review, publications were selected based on the Population, Intervention, Comparator, and Outcome (PICO) framework (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The inclusion criteria were designed to identify publications that investigate the effectiveness of PAs in controlling threats to biodiversity and contribute to a broader understanding of their role in biodiversity conservation. Publications were required to focus on anthropogenic threats, assess changes within at least one PA, and include a comparator(s) to evaluate changes over time, space, or management interventions (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe included studies employed a range of evaluation designs to assess threat changes in PAs. BACI (Before-After Control-Impact) designs were used in studies that incorporated both a temporal comparison of conditions before and after the implementation of an intervention (such as a management strategy, change in management authority, or PA designation) and a spatial comparison between the PA and surrounding areas or comparator sites. BA (Before-After) designs relied solely on temporal comparisons, evaluating changes before and after the intervention without including spatial control. CI (Control-Impact) designs focused on spatial comparisons between PAs and surrounding areas or comparator sites but did not specify the use of matching approaches. Matching is a statistical technique that pairs protected and unprotected areas with similar characteristics\u0026mdash;such as habitat type or fishing pressure\u0026mdash;to isolate the effect of PA protection (Ahmadia et al., 2015). These may include temporal comparisons, but only for the post-intervention period. Finally, CI (Matching) studies also employed spatial comparisons but explicitly used matching methods to select control sites. These studies may similarly include a temporal comparison after the intervention, but not necessarily a \u0026ldquo;before\u0026rdquo; period.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eEligibility criteria for article screening\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"3\"\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\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCategory\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eInclusion criteria\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eExclusion criteria\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePopulation\u003c/p\u003e\u003cp\u003eAreas experiencing threats to biodiversity\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003eA. Studies measuring the change of threat levels in PAs.\u003c/b\u003e\u003c/p\u003e\u003cp\u003e- Threats must be anthropogenic and assessed directly or indirectly using proxies (e.g., Human Footprint Index).\u003c/p\u003e\u003cp\u003e- Studies must mention threats explicitly or implicitly (in the abstract) or use synonyms.\u003c/p\u003e\u003cp\u003e- Studies using surveys or interviews that quantify human perceptions of threat change (e.g., Threat Reduction Assessment).\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e- Studies that do not address threats to biodiversity or measure changes in threat states.\u003c/p\u003e\u003cp\u003e- Studies focusing on non-anthropogenic threats (e.g., geological events).\u003c/p\u003e\u003cp\u003e- Climate change is excluded due to challenges in measuring PA effectiveness in mitigating or exacerbating its impacts.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eIntervention\u003c/p\u003e\u003cp\u003eEstablishing protected areas\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003eB. Studies assessing threats inside at least one PA.\u003c/b\u003e\u003c/p\u003e\u003cp\u003e- Includes all PA types and designations for biodiversity conservation (e.g., national parks, wildlife sanctuaries, nature reserves, biosphere reserves).\u003c/p\u003e\u003cp\u003e- Studies conducted at any scale (local to global) and across terrestrial, freshwater, and marine ecosystems.\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e- Studies that do not focus on areas designated primarily for biodiversity conservation.\u003c/p\u003e\u003cp\u003e- Studies on OECMs (Other Effective Area-Based Conservation Measures) that do not meet the definition of PAs.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eComparator\u003c/p\u003e\u003cp\u003eDifference in threat state\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003eC. Studies measuring changes in threats within PAs, using at least one comparator\u003c/b\u003e:\u003c/p\u003e\u003cp\u003eC1. Temporal Comparison: Threat state assessed at two or more time points within PAs.\u003c/p\u003e\u003cp\u003eC2. Management Comparison: Comparison of different PA management interventions over time or with control sites (C3).\u003c/p\u003e\u003cp\u003eC3. Spatial Comparison: Threat changes in PAs assessed using spatial comparators:\u003c/p\u003e\u003cp\u003e- PA vs. non-PA (e.g., non-protected locations, buffer zones, surrounding areas, land-use types, control sites).\u003c/p\u003e\u003cp\u003e- PA vs. PA (e.g., within the same PA with a control site or threat/protection gradient, or across PAs with different categories/intervention types).\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eStudies that do not assess changes in threats as an outcome of protection, management intervention, or spatial comparison (i.e., lacking a comparator: C1, C2, or C3).\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOutcome\u003c/p\u003e\u003cp\u003eAssessment of threat change in PAs\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003eSelected studies must meet criteria A, B, and C\u003c/b\u003e and conduct a quantitative analysis aligned with the review\u0026rsquo;s objectives by addressing one of the following:\u003c/p\u003e\u003cp\u003e\u0026bull; Evaluates changes in threats resulting from protection or interventions within PAs (Direct impact assessment of threats in PAs)\u003c/p\u003e\u003cp\u003e\u0026bull; Compares threats or assesses the potential effectiveness of PAs with any comparator (C1, C2, C3) (Indirect or potential threat assessment).\u003c/p\u003e\u003cp\u003eMetrics must be directly linked to the specific threat (e.g., invasive species: presence inside vs. outside PAs; deforestation: % forest cover).\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e- Studies that do not assess threat states after an intervention.\u003c/p\u003e\u003cp\u003e- Studies measuring proxies of threats that may be influenced by other factors (e.g., species abundance, population traits, vegetation cover, animal behavior, community composition). Thus, studies relying solely on biodiversity metrics to measure PA impact were excluded.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStudy design\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eStudies employing comparison groups CI and/or BA or BACI designs.\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e- Personal views, perspectives, theoretical studies, and models.\u003c/p\u003e \u003cp\u003e- Observational studies with no controls or comparators.\u003c/p\u003e\u003cp\u003e- Literature reviews.\u003c/p\u003e\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=\"Sec13\" class=\"Section3\"\u003e\u003ch2\u003e3.3.3 Study validity assessment\u003c/h2\u003e\u003cp\u003eA critical appraisal process was conducted using the Collaboration for Environmental Evidence Critical Appraisal Tool (Version 0.3, Prototype) (Konno et al., 2021). This tool was adapted to align with the specific requirements of this study. Given the nature of the research, the assessment focused on key sources of bias, including risk of confounding biases, post-intervention/exposure selection biases, misclassified comparison biases, detection biases, outcome reporting biases, and outcome assessment biases. As part of the consistency-checking process, a random subset of full-text articles was critically appraised by the senior researcher to ensure alignment with the conducted appraisal and maintain assessment reliability. Articles identified as having a high risk of bias were excluded from the analysis. The adjusted appraisal template and individual study assessments are available in \u003cem\u003eAdditional File 2\u003c/em\u003e.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section3\"\u003e\u003ch2\u003e3.3.4 Data coding and extraction strategy\u003c/h2\u003e\u003cp\u003eData coding was conducted based on the PICO framework (Population, Intervention, Comparator, and Outcome), along with relevant metadata. To streamline the process, data extraction and critical appraisal were performed using a predefined coded template in Covidence, designed to ensure consistency and efficiency in the data extraction process. The decisions of the coded data, as well as the options for selection, can be found in \u003cem\u003eAdditional File 3\u003c/em\u003e. Consistency checks were conducted in collaboration with the senior author to validate the accuracy and reliability of the extracted information.\u003c/p\u003e\u003cp\u003eWe systematically recorded study metadata, including author(s), year of publication, title, and whether the study explicitly aimed to evaluate effectiveness as a primary goal. We also collected data on study characteristics such as geographic region, ecosystem classification (terrestrial, freshwater, marine), study design type, comparator type, threat assessment methods, information on PAs, reported biodiversity threats, and study outcomes.\u003c/p\u003e\u003cp\u003eA comparator refers to any unit, such as an area, period, or management intervention, used to evaluate the effectiveness of a PA. Comparators can include areas outside the PA, buffer zones surrounding the PA, PAs with varying levels of protection, or comparisons before and after an intervention. These comparators may be descriptive or loosely matched without strict equivalency, or they may involve the use of control with counterfactuals, which aim to represent what would have happened in the absence of the intervention (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eComparator types selected\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"2\"\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\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eComparator type\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDescription\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eInside - Outside\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCompares areas inside and outside the PA.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePA - PA buffer\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCompares the interior of the PA with its surrounding buffer zone.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePA - PA categories\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCompares outcomes between PAs with different levels of enforcement or protection categories.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePA - Gradient\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCompares threat levels within PAs under different gradients of pressure, management strictness, or locations within the PA facing conditions influencing the outcomes.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePA - Management intervention\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCompares threat levels inside PA(s) to measure the effectiveness of a management intervention\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePA - Other land use types\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCompares threat levels of PA(s) with other land use types in the surrounding landscape (eg. Indigenous lands)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eOutcome data were extracted by identifying measurable changes in threats, whether directly assessed within PAs or inferred from indicators such as human footprint indices (Sanderson et al., 2002; Venter et al., 2016a, 2016b) or modification indices (Kennedy et al., 2019), recognizing that threats can manifest at multiple levels, including drivers, sources, and mechanisms (Balmford et al., 2009). Specifically, we documented changes in threats both within PAs and in comparator (control) areas, as well as whether PAs were found to be more effective than their respective controls. We also extracted any reasons identified by the articles that explained the observed outcomes.\u003c/p\u003e\u003cp\u003eFollowing the CEE Critical Appraisal Tool (Konno et al., 2021), we assessed the risk of bias for each study. The original appraisal categories were adapted to better align with the context and scope of this review. To ensure consistency in data extraction across the full set of papers, a subset of full-text papers was randomly selected and critically appraised by a senior researcher to verify alignment with the conducted appraisal. Discrepancies were identified, discussed, and resolved through consensus. In cases where one of the review authors was also an author of a study included in the review, they were not involved in its screening, appraisal, or selection of that article.\u003c/p\u003e\u003cp\u003e\u003cb\u003eRules and conditions for data extraction\u003c/b\u003e\u003c/p\u003e\u003cp\u003eGiven the heterogeneity of results across articles, including differences in the number and type of threats analyzed (e.g., single threats, composite indices, or multiple threat types), as well as the number and characteristics of PAs assessed, we adapted our data extraction sheet accordingly, applying a set of predefined rules.\u003c/p\u003e\u003cp\u003eWhen an article analyzed multiple threats, outcomes were recorded based on the availability of data for each threat individually. Additionally, when an article contained distinct intervention types (e.g., between PAs with different levels of strictness), data were extracted separately for each case. We refer to these individual comparisons as separate \u0026ldquo;studies.\u0026rdquo; As such, a single article could contribute with multiple studies to the review. In contrast, if an article assessed several PAs of the same category, it was treated as a single study, and only the overall result was extracted (Additional file 3).\u003c/p\u003e\u003cp\u003eFor outcome extraction, we focused on quantitative data and assigned a qualifier based on predefined criteria. Specifically, when a study examined changes in threats within a PA over time or between comparators, we categorized the outcomes based on the magnitude and direction of reported changes. Changes greater than 10% were classified as either \"increased\" or \"decreased,\" depending on the direction of the reported outcome. When the change ranged between 5 and 10%, the outcome was classified as \u0026ldquo;slightly increased\u0026rdquo; or \u0026ldquo;slightly decreased\u0026rdquo;. Changes smaller than \u0026plusmn;\u0026thinsp;5%, or cases where statistical tests indicated non-significant differences (e.g., p-values above the significance threshold), were classified as \"neutral.\" These thresholds were established a priori to ensure consistency in interpreting and comparing results across studies.\u003c/p\u003e\u003cp\u003eReported threats were first coded according to the IUCN Threat Classification Scheme \u0026ndash; Version 3.3 (IUCN, 2024). To facilitate analysis and interpretation, similar threat types were subsequently grouped into ten simplified categories, which we refer to as reported grouped threats (Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The definitions and inclusion criteria for each coded response are available in Additional File 3, and full details of the data extraction process, including all variables collected from each study, are available in \u003cem\u003eAdditional File 2\u003c/em\u003e.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eCategorization of reported threats based on the IUCN threat classification scheme\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"2\"\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\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eReported threats\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eIncluded IUCN threats\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLULCC-Infrastructure\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1 Residential \u0026amp; commercial development,\u003c/p\u003e\u003cp\u003e4 Transportation \u0026amp; service corridors\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLULCC-Agriculture\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2 Agriculture \u0026amp; aquaculture, grazing\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFishing and Hunting\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5.1 Hunting, gathering, fishing\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLULCC-Deforestation\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5.3 Logging \u0026amp; wood harvesting\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRecreation\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e6.1 Recreational activities\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFire\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e7.1 Fire \u0026amp; fire suppression\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eInvasive Species\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e8 Invasive \u0026amp; other problematic species, genes \u0026amp; diseases\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePollution\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e9 Pollution (Light pollution, chemical pollution)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMultiple Threats\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMultiple threats with unspecified independent outcomes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLULCC-Other\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCombined human pressure indices: HFI, Human pressure, Human modification, Temporal Human Pressure Index (THPI)\u003c/p\u003e\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=\"Sec15\" class=\"Section3\"\u003e\u003ch2\u003e3.3.5 Potential effect modifiers/reasons for heterogeneity\u003c/h2\u003e\u003cp\u003eEffect modifiers contributing to heterogeneity in the results were identified during the full-text screening and recorded for all included studies. Where possible, we collected information on the methods used to assess the impact of potential effect modifiers. Given the nature of our study, various biogeographic, environmental, and socio-economic factors contributed to the heterogeneity of impacts reported across studies.\u003c/p\u003e\u003cp\u003eSome potential effect modifiers identified in previous research included the category of the PA, governance type, geographical location, topographic features, PA size, date and period of establishment, the socioeconomic context of the country or state where the PA was located, and ecosystem type, among others. Where available, we recorded these potential effect modifiers in Additional File 2.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec16\" class=\"Section3\"\u003e\u003ch2\u003e3.3.6 Data synthesis and presentation\u003c/h2\u003e\u003cp\u003eWe conducted a descriptive, narrative synthesis of the included studies. This synthesis describes the robustness of the included studies and the potential impact of methodological differences on the reported outcomes. Thus, frequencies and proportions were calculated across all studies to analyze patterns in study metadata, methodological approaches, threat assessment techniques, and outcomes.\u003c/p\u003e\u003cp\u003eMetadata analysis included variables such as geographical location, ecosystem type (terrestrial, freshwater, marine), and study scale (local, regional, global). To assess methodological variation, we examined study designs and comparator types, determining whether studies used temporal, spatial, management-based, or mixed comparators. We also evaluated the threat assessment methods used, distinguishing between remote sensing, field-based data, and indices or proxies, as well as the number of threats analyzed per study (e.g., single threats, composite indices, or multiple threats).\u003c/p\u003e\u003cp\u003eThe validity assessment of the articles was based on predefined criteria, and we computed frequencies and proportions to assess the risk of bias. Furthermore, we contrasted the findings from the validity assessment with the types of study designs and their respective comparators to gain deeper insights into the potential methodological factors contributing to increased risk of bias.\u003c/p\u003e\u003cp\u003eTo summarize the major findings of our systematic review, the data synthesis was structured into three key main results. First, we analyzed the number and proportion of studies reporting the magnitude of changes inside and outside PAs across time, and the number of times where PAs exhibited lower threat intensity compared to their counterfactuals. Second, we analyzed PA effectiveness across different threat types. As some studies reported outcomes for multiple threats, we extracted data for each threat individually. We refer to individual threats within studies as \u0026ldquo;cases\u0026rdquo; (Additional file 3). Third, we examined the factors influencing PA outcomes, as reported or discussed in the studies. These factors were categorized into five main drivers of PA effectiveness: location and baseline conditions, design, management, and governance. Within these categories, we explored subcriteria determining whether PAs had a positive, neutral, or negative impact on threat mitigation. Tables and figures summarizing the results are provided as supplementary information in the systematic review as well as the collected data. All data wrangling and statistical analyses were conducted using R (version 4.2.3) (R Core Team, 2021).\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e"},{"header":"Review findings","content":"\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\u003ch2\u003e4.1 Review descriptive statistics\u003c/h2\u003e\u003cp\u003eOur search yielded 5,687 unique articles. After screening titles and abstracts based on the inclusion criteria and alignment with the review objectives, 407 (7%) articles were selected for full-text review. Of these, 296 (5%) that did not meet the inclusion criteria were excluded. A total of 111 (2%) articles were included after the full-text screening and critical appraisal. The most common reason for exclusion at the full-text screening was the absence of a comparator (n\u0026thinsp;=\u0026thinsp;103), which also served as a pre-filter for critical appraisal. For example, studies assessing threat states at a single time point, lack of relevance to the review objectives, such as studies not involving PAs; and reliance solely on changes in biodiversity metrics to infer the impact of threats. Finally, articles with a high risk of bias were excluded according to our critical appraisal assessment, resulting in a total of 105 articles. Within these articles, 9 contained different intervention comparisons resulting in a total of 117 studies that were included in the analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eGeographical location and habitat types\u003c/b\u003e\u003c/p\u003e\u003cp\u003eAcross the reviewed articles we had studies from 133 countries, with Asia accounting for the largest share (30%), followed by Europe (19%), South America (16%), and Africa (15%), which showed relatively balanced representation. Research from China, Brazil, Spain, Italy, Tanzania, Ecuador, and Indonesia primarily shaped these continental patterns (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eTerrestrial ecosystems were the focus of 71% (n\u0026thinsp;=\u0026thinsp;82) of the articles, while 12.4% (n\u0026thinsp;=\u0026thinsp;13) examined marine and coastal habitats and 5.7% (n\u0026thinsp;=\u0026thinsp;6) focused on freshwater ecosystems. The spatial scale of the studies was relatively homogeneous, with 36.2% conducted at the local scale (n\u0026thinsp;=\u0026thinsp;38), followed by regional (26.6%), national (21%), global (8.6%), and continental (7.7%) scales.\u003c/p\u003e\u003cp\u003e\u003cb\u003eThreat assessment methods, study designs and comparators\u003c/b\u003e\u003c/p\u003e\u003cp\u003eRemote sensing was the most common method to assess threats, used in 68% (n\u0026thinsp;=\u0026thinsp;71) of the articles. Composite threat indices, such as the human footprint index, were used in 19% (n\u0026thinsp;=\u0026thinsp;20), while only 11.6% (n\u0026thinsp;=\u0026thinsp;11) of the articles relied on field sampling or interviews. Eighty percent of the articles assessed a single threat (n\u0026thinsp;=\u0026thinsp;84), while fewer articles examined two (n\u0026thinsp;=\u0026thinsp;10), three (n\u0026thinsp;=\u0026thinsp;5), four (n\u0026thinsp;=\u0026thinsp;5), and five threats (n\u0026thinsp;=\u0026thinsp;1).\u003c/p\u003e\u003cp\u003eIn terms of study design, CI was the most common approach being employed in 80% (n\u0026thinsp;=\u0026thinsp;84) of the articles, including 14.3% (n\u0026thinsp;=\u0026thinsp;15) that applied matching. BACI and BA study designs were used less frequently, found in 11.4% (n\u0026thinsp;=\u0026thinsp;12) and 8.6% (n\u0026thinsp;=\u0026thinsp;9) of articles, respectively.\u003c/p\u003e\u003cp\u003eThe most common comparator was an \"Inside-Outside\" used in 34% (n\u0026thinsp;=\u0026thinsp;50) of the articles, followed by comparisons between PAs and their buffers (26.5%, n\u0026thinsp;=\u0026thinsp;39), management interventions (17%, n\u0026thinsp;=\u0026thinsp;25), and different PA categories (15.6%, n\u0026thinsp;=\u0026thinsp;23). Studies using CI and CI matching methods used mainly inside-outside, buffer, and PA category comparisons. In contrast, BA and BACI designs were more commonly applied in studies focused on management interventions (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec19\" class=\"Section2\"\u003e\u003ch2\u003e4.2 Narrative synthesis including study validity assessment\u003c/h2\u003e\u003cp\u003eWe identified confounding and selection biases as the primary sources of methodological limitations. Over half of the articles exhibited a medium risk of bias in confounding and selection criteria, while the risk of bias in misclassification, detection, reporting, and assessment was generally low (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, A). When examining the causes of confounding bias, we found that in 90.4% of the articles, the impact of the exposure or intervention was likely or seemingly likely to be confounded (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, B). However, 61.5% controlled for potential confounders. Notably, 17% of articles did not mention the omission of confounders (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, B).\u003c/p\u003e\u003cp\u003eCI studies had the highest frequency of medium-risk bias for both confounding (n\u0026thinsp;=\u0026thinsp;38) and selection biases (n\u0026thinsp;=\u0026thinsp;45), suggesting that despite their widespread use, CI designs are particularly vulnerable to these biases. This was especially evident in Inside\u0026ndash;Outside and PA\u0026ndash;PA Buffer comparisons. In contrast, CI studies with matching exhibited lower overall frequencies of bias, indicating that the use of matching techniques may help mitigate confounding and selection issues. BA and BACI studies had lower overall bias risks, though they were also less frequently used (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, C).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec20\" class=\"Section2\"\u003e\u003ch2\u003e4.3 Data synthesis\u003c/h2\u003e\u003cp\u003e\u003cb\u003eMixed effectiveness of PAs in mitigating threats\u003c/b\u003e\u003c/p\u003e\u003cp\u003eOn average, we found that PAs had lower threat levels compared to unprotected areas, with 65% of observations (87 out of 134) showing that PAs performed better than their outside comparator (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, C). However, many PAs still experienced increases in threats. Fifty-seven percent of observations (73 out of 129) reported rising threat levels inside PAs (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, A), while 75% (83 out of 110) reported increases in control sites or outside PAs (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, B).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eLULCC-related threats such as infrastructure development, agriculture and deforestation were the most frequently reported threats increasing within PAs, with increases observed in 87%, 64%, and 53% of observations, respectively. However, these threats also exhibited the highest rates of reduction, with deforestation decreasing in 39% of the observations, agriculture in 36%, and other LULCC types in 21%, indicating that PAs can play a role in slowing land-use change. Notably, LULCC has been the dominant focus of the literature, reflecting its prevalence as the most frequently analyzed threat type. In contrast, fire and invasive species have proven more challenging to address even inside PAs, which were reported to be increasing in 70% and 60% of the observations inside PAs respectively (Fig.\u0026nbsp;9, A).\u003c/p\u003e\u003cp\u003eOutside PAs, the rate of threat increase was generally higher than inside, with infrastructure development and other LULCC threats showing increases in 100% and 74% of the observations, respectively, suggesting that PAs provide some level of containment for these pressures. While deforestation and agriculture also increased outside PAs, some studies reported evidence of reductions in these threats near PAs, possibly due to buffer zones or spillover effects from conservation measures. For example, Xu et al., (2022) found that overall forest cover increased and fragmentation decreased from baseline years, when reserves were designated, in both reserves and surrounding areas (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e, B).\u003c/p\u003e\u003cp\u003e\u003cb\u003ePAs effectiveness varies by threat type\u003c/b\u003e\u003c/p\u003e\u003cp\u003ePAs were found to be effective in mitigating certain threats when compared to unprotected areas. Most studies assessing LULCC-related pressures reported lower threat levels inside PAs compared to areas without protection, particularly for infrastructure (73%), deforestation (77%), and agriculture (57%). Additionally, fire (56%) and hunting (80%) were significantly less prevalent within PAs, reinforcing their role in reducing direct anthropogenic disturbances (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e, C). However, the effectiveness of PAs in buffering against pollution and recreation-related pressures was less evident, likely due to the external origin of these threats or their manifestation as secondary impacts within protected boundaries or create secondary threats within the PAs themselves (Fig.\u0026nbsp;9). For instance,Mollmann et al. (2022) found that stream water quality inside PAs reflected broader regional pollution patterns, indicating that exogenous inputs continue to shape environmental conditions within PA borders. Likewise,Gonson et al. (2017) found that marine protected areas (MPAs) experienced higher tourism-related pressure indices on islets than nearby unprotected sites.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eDrivers of PAs effectiveness\u003c/b\u003e\u003c/p\u003e\u003cp\u003eOur results show that PA outcomes are shaped by a combination of location, design, management, and governance factors (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). PAs being located under high threat and accessibility were consistently reported as major obstacles to PA effectiveness. Similarly, high population density and permeable boundaries reduce PA effectiveness when external pressures, like pollution, continue to impact ecosystems despite protection. Landscape fragmentation and PAs with a small size were frequently associated with increased threats, highlighting the vulnerability of PAs to be highly impacted by location, baseline and design conditions.\u003c/p\u003e\u003cp\u003eConversely, well-resourced PAs with sufficient funding, strong policy, and effective enforcement mechanisms were more successful in achieving positive outcomes. Notably, in some cases, robust management interventions and strong policies were sufficient to counteract the negative effects of high accessibility, population density, and high-threat environments. For instance, Bleher et al., (2006), found that a shift in management authority led to reduced deforestation and improved performance in PAs with high level of enforcement. Similarly, (Eklund et al., (2022) reported that the suspension of management activities in Madagascar\u0026rsquo;s PAs during COVID-19 resulted in a 76\u0026ndash;248% increase in fire incidents, with burning levels returning to normal once management resumed.\u003c/p\u003e\u003cp\u003eIn addition, large PA size was more frequently associated with positive conservation outcomes, whereas older PAs displayed mixed effects. Hirons et al. (2022) and Wu et al. (2022) found that larger reserves experienced less human modification and reduced anthropogenic pressure across tropical biomes and in Northeast China, respectively. In contrast, age had a more complex influence: older and smaller PAs in Africa were more vulnerable to deforestation (Bowker et al., 2017), and globally, older PAs saw increased human pressure (Geldmann et al., 2014). In Europe, age played a more positive role in naturalization near long-established reserves (Mingarro \u0026amp; Lobo, 2023) while its effect on fragmentation in Natura 2000 sites was mostly neutral (Lawrence et al., 2021).\u003c/p\u003e\u003cp\u003eLocal governance structures and national development indices demonstrated positive effects. For example, Mammides, (2020), found that lakes in North America and Europe experienced lower or even decreasing human pressure, while biodiverse lakes in tropical regions faced significantly higher and rapidly increasing human impact.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec21\" class=\"Section2\"\u003e\u003ch2\u003e4.4 Review limitations\u003c/h2\u003e\u003cp\u003e\u003cb\u003eLimitations of the Review Methods\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe heterogeneity of study designs, variation in measured outcomes, and the lack of standardized reporting of effect sizes and associated uncertainties made it impossible to quantitatively compare the magnitude of effects across studies. Therefore, we were unable to conduct a meta-analysis based on the studies. This reflects a broader challenge in conservation evidence synthesis, where inconsistent or incomplete reporting impedes cross-study comparisons (Pullin \u0026amp; Stewart, 2006); Bowler et al., 2020). A meta-analysis would have enabled a more quantitative synthesis by allowing standardized comparisons across different studies (Gurevitch et al., 2018), and potentially draw more targeted conclusions for policy and practice.\u003c/p\u003e\u003cp\u003eOur search strategy was restricted to studies published in English, which may have led to the omission of relevant non-English literature. This language bias can result in regional underrepresentation, particularly of studies from non-English-speaking biodiversity-rich countries (Amano et al., 2023). Despite this limitation, our systematic review includes a diverse representation of countries, limiting the risk that our findings are not broadly applicable across different geographic and governance contexts.\u003c/p\u003e\u003cp\u003eWhile the objective and questions of this study have arisen from the authors' scientific motivation, stakeholder engagement is crucial for effective conservation. Therefore, we acknowledge the need for stakeholder involvement in future research. Nonetheless, the study findings will significantly impact conservationists, PA managers, and decision-makers. By assessing the success of current management strategies and methods for addressing changes in threats to biodiversity, this research will improve our understanding of the effectiveness of PAs in addressing threats. Furthermore, the information gathered can serve as evidence to formulate regulations related to Protected Area Management and Evaluation (PAME) and identify PAs' main strengths and weaknesses in facing threats based on their geographical location and socio-economic-ecological characteristics. In addition, this review will provide significant findings for national decision-making.\u003c/p\u003e\u003cp\u003e\u003cb\u003eLimitations of the Evidence Base\u003c/b\u003e\u003c/p\u003e\u003cp\u003eA significant portion of the evidence found relied on CI study designs, which were associated with the most frequent risk of bias. These biases were primarily due to confounding and selection bias, arising from the fact that they are often established in remote regions or areas of lower economic interest (Geldmann et al., 2025; Joppa \u0026amp; Pfaff, 2009; Margules \u0026amp; Pressey, 2000). This non-random placement introduces substantial limitations when comparing protected and unprotected sites, reducing the ability to isolate the true effect of protection. To address this challenge, matching approaches have been increasingly used to construct more comparable control groups and reduce bias in impact evaluations (Chen et al., 2023). In our review, CI studies that incorporated matching techniques exhibited lower bias risks compared to those that did not, demonstrating the value of such methodological adjustments in enhancing validity.\u003c/p\u003e\u003cp\u003eDespite these improvements, most studies did not explicitly account for counterfactuals and thus were limited in their ability to fully address confounding variables. This omission can lead to overestimations of the effects of protection (Terraube et al., 2020). Given the inherent ecological, spatial, and socio-political heterogeneity of PAs, this remains a critical limitation in assessing conservation impact. These findings highlight the need for standardized spatial comparison protocols and counterfactual-based methodologies, to ensure that PA effectiveness is assessed with greater methodological rigor and cross-study comparability (Geldmann et al., 2025).\u003c/p\u003e\u003cp\u003e Additionally, while our review could approximate the effectiveness of PAs against specific threats, we were unable to capture the complexity of simultaneous or interacting threats. Many studies examined individual pressures in isolation, whereas, in reality, PAs are often exposed to multiple, compounding human activities (Beebee \u0026amp; Griffiths, 2005; Hof et al., 2011; Men\u0026eacute;ndez-Guerrero \u0026amp; Graham, 2013; Stuart et al., 2004). For instance, the Qilian Mountain Nature Reserve in China was found to be effective in controlling population growth and land-use change, but ineffective in curbing infrastructure development, including road construction, mining, and hydropower projects (S. Li et al., 2022). Similarly, (Mart\u0026iacute;nez-Fern\u0026aacute;ndez et al., 2015) found increasing pressure from urbanization both inside and outside PAs, while pressures from agriculture and deforestation were declining. This illustrates the importance of temporal dynamics in threat assessment, as current land uses may build upon earlier, unmeasured disturbances. Land-use transitions often occur sequentially, with past pressures shaping current vulnerability and management needs (Curtis et al., 2018).\u003c/p\u003e\u003c/div\u003e"},{"header":"Review conclusions","content":"\u003cdiv id=\"Sec23\" class=\"Section2\"\u003e\u003ch2\u003e5.1 Implications for Policy/Management\u003c/h2\u003e\u003cp\u003eThis review provides evidence that PAs have played a role in mitigating threats to biodiversity, with most studies showing lower threat levels inside PAs compared to non-protected control sites. Although pressures such as deforestation, agriculture, infrastructure development, fire, and hunting continue to increase globally, PAs were generally associated with smaller increases or reductions in these threats relative to unprotected areas.\u003c/p\u003e\u003cp\u003eHowever, the effectiveness of PAs is not consistent across all contexts. Outcomes are highly dependent on the type of threat and the broader context in which the PA operates. Better outcomes were associated with adequate funding, strong policy frameworks, effective enforcement, larger PA size and higher national development indices. In contrast, high accessibility, dense surrounding populations, and location in high-pressure environments were consistently linked to reduced effectiveness.\u003c/p\u003e\u003cp\u003eFinally, meeting global biodiversity targets will require not only expanding PA networks but also ensuring that existing PAs are effective (G. Li et al., 2024) and capable of reducing threats. This includes investing in capacity building, enforcement, policy coherence, and local engagement to ensure that PAs function as effective tools for halting habitat loss and reducing anthropogenic pressures. Our review supports this by showing that well-managed PAs can achieve significantly better outcomes, even in high-pressure environments.\u003c/p\u003e\u003cp\u003eOverall, our review confirms that while PAs can reduce biodiversity threats, their effectiveness is highly context-specific, and improving the tools and frameworks used to evaluate them is essential for guiding conservation strategies globally.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec24\" class=\"Section2\"\u003e\u003ch2\u003e5.2 Implications for Research\u003c/h2\u003e\u003cp\u003eWe identified several areas where the current evidence base on PA effectiveness could be strengthened, providing useful directions for future research. While most studies found that PAs were associated with lower threat levels compared to unprotected areas, their effectiveness was highly variable and strongly context-dependent. Factors such as location, baseline conditions, management capacity, and governance structures influenced outcomes but were often inconsistently reported or controlled for. These uncertainties highlight the importance of developing transparent and context-sensitive approaches to assessing PA effectiveness, particularly in studies comparing protected and unprotected areas. In this context, the use of rigorous counterfactual approaches such as matching methods that address potential confounders and control for the non-random location of PAs represents a promising avenue to improve impact attribution (Schleicher et al., 2019). Future research should prioritize the adoption of such causal inference techniques to better isolate the effectiveness of PAs on threat mitigation and their impact on conservation outcomes (Terraube et al., 2020).\u003c/p\u003e\u003cp\u003eIn addition, there is a need to improve consistency in how threats are measured. Several articles were excluded during the screening phase due to the absence of a defined or measurable threat assessment metric, highlighting the importance of clearly reported and comparable indicators across studies. In addition, there is a need to broaden the scope of data collection. While substantial progress has been made particularly through the use of remote sensing and LULCC analyses(Andam et al., 2008; Soares-Filho et al., 2023; Zhu et al., 2022) limited attention has been given to threats such as overexploitation, pollution and invasive species, which require in situ data collection and have been less included in large-scale PA evaluations (Hern\u0026aacute;ndez-Y\u0026aacute;\u0026ntilde;ez et al., 2016; Rodrigues \u0026amp; Cazalis, 2020; Schulze et al., 2018). Establishing robust, threat-specific metrics can enhance the consistency of assessments and enable more detailed insights into how PAs perform across different contexts.\u003c/p\u003e\u003cp\u003eIn addition, future research should move toward integrative and longitudinal study designs that can account for temporal dynamics and interactions between multiple threats. For example, areas that were previously deforested may now be experiencing urbanization, complicating the interpretation of PA effectiveness. Incorporating land-use history, threat transitions, and long-term monitoring would improve the accuracy of assessments. Moreover, greater standardization in the definition and selection of comparator areas, particularly regarding spatial scale and distance, would enhance the comparability and interpretability of PA evaluations.\u003c/p\u003e\u003cp\u003eFinally, to ensure that findings are globally relevant, future research should aim to reduce geographic, linguistic, and thematic biases by increasing the inclusion of non-English language literature, underrepresented regions, and less frequently studied threat types (Amano et al., 2016). This will help build a more comprehensive, equitable, and policy-relevant understanding of PA effectiveness worldwide.\u003c/p\u003e\u003cp\u003eWhile our review highlights the need to broaden linguistic and geographic coverage and to tackle understudied regions and threat types, it also reveals a stark evidence gap. We found only 105 studies worldwide that rigorously quantify the change of threat inside PAs. Given that there are more than 200,000 terrestrial PAs globally, this lack of data is alarming. Unless more impact assessment studies are conducted, PAs managers and policymakers will lack the data needed to allocate resources effectively or to adapt strategies. Only by filling this gap can we move beyond to a comprehensive understanding of how PAs deliver on their promise of biodiversity conservation.\u003c/p\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003ch2\u003e6.1 Ethics approval and consent to participate\u003c/h2\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003ch2\u003e6.2 Consent for publication\u003c/h2\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003ch2\u003e6.3 Availability of data and materials\u003c/h2\u003e\n\u003cp\u003eAll data generated or analyzed during this study are included in this published article [and its supplementary information files].\u003c/p\u003e\n\u003ch2\u003e6.4 Competing interests\u003c/h2\u003e\n\u003cp\u003eThe authors declare that they have no competing interests\u003c/p\u003e\n\u003ch2\u003e6.5 Funding\u003c/h2\u003e\n\u003cp\u003eK.P-C., H.F and J.G. were supported by The Danish Independent Research council (grant no. 0165-00018B) and E.V. by the Kone Foundation (grant no. 201803179). A.G. was supported by HORIZON MSCA Postdoctoral Fellowships (Project number: 101104696, THREATS). H.A.P was supported by HORIZON MSCA Postdoctoral Fellowship (Project number: 101063896, PROTECTEDTRADE). C.R. was supported by research grant 25925 from VILLUM FONDEN.\u003c/p\u003e\n\u003ch2\u003e6.6 Authors' contributions\u003c/h2\u003e\n\u003cp\u003eKP-C and JG jointly conceived and designed the study. KPC, HAP, AG, HF, EV, and JG conducted the title and abstract screening and helped refine the inclusion criteria. KPC carried out the full-text screening, data extraction, and critical appraisal, with support from JG, who resolved uncertainties and verified the process. KPC conducted the analysis and drafted the systematic review. All authors provided feedback and revisions. All authors read and approved the final manuscript.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAhmadia, G. N., Glew, L., Provost, M., Gill, D., Hidayat, N. I., Mangubhai, S., Purwanto, \u0026amp; Fox, H. E. (2015). Integrating impact evaluation in the design and implementation of monitoring marine protected areas. \u003cem\u003ePhilosophical Transactions of the Royal Society B: Biological Sciences\u003c/em\u003e, \u003cem\u003e370\u003c/em\u003e(1681). https://doi.org/10.1098/RSTB.2014.0275\u003c/li\u003e\n\u003cli\u003eAmano, T., Berdejo-Espinola, V., Akasaka, M., de Andrade Junior, M. A. U., Blaise, N., Checco, J., \u0026Ccedil;ilingir, F. G., Citegetse, G., Corella Tor, M., Drobniak, S. M., Giakoumi, S., Golivets, M., Ion, M. C., Jara-D\u0026iacute;az, J. P., Katayose, R., Lasmana, F. P. S., Lin, H. Y., Lopez, E., Mikula, P., \u0026hellip; Zamora-Gutierrez, V. (2023). The role of non-English-language science in informing national biodiversity assessments. \u003cem\u003eNature Sustainability 2023 6:7\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e(7), 845\u0026ndash;854. https://doi.org/10.1038/s41893-023-01087-8\u003c/li\u003e\n\u003cli\u003eAmano, T., \u0026amp; Sutherland, W. J. (2013). Four barriers to the global understanding of biodiversity conservation: wealth, language, geographical location and security. \u003cem\u003eProceedings of the Royal Society B: Biological Sciences\u003c/em\u003e, \u003cem\u003e280\u003c/em\u003e(1756). https://doi.org/10.1098/RSPB.2012.2649\u003c/li\u003e\n\u003cli\u003eAndam, K. S., Ferraro, P. J., Pfaff, A., Sanchez-Azofeifa, G. A., \u0026amp; Robalino, J. A. (2008). Measuring the effectiveness of protected area networks in reducing deforestation. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e, \u003cem\u003e105\u003c/em\u003e(42), 16089\u0026ndash;16094. https://doi.org/10.1073/pnas.0800437105\u003c/li\u003e\n\u003cli\u003eBalmford, A., Carey, P., Kapos, V., Manica, A., Rodrigues, A. S. L., Scharlemann, J. P. W., \u0026amp; Green, R. E. (2009). Capturing the many dimensions of threat: comment on Salafsky et al. \u003cem\u003eConservation Biology : The Journal of the Society for Conservation Biology\u003c/em\u003e, \u003cem\u003e23\u003c/em\u003e(2), 482\u0026ndash;487. https://doi.org/10.1111/J.1523-1739.2009.01196.X\u003c/li\u003e\n\u003cli\u003eBarnes, M. D., Craigie, I. D., Harrison, L. B., Geldmann, J., Collen, B., Whitmee, S., Balmford, A., Burgess, N. D., Brooks, T., Hockings, M., \u0026amp; Woodley, S. (2016). Wildlife population trends in protected areas predicted by national socio-economic metrics and body size. \u003cem\u003eNature Communications 2016 7:1\u003c/em\u003e, \u003cem\u003e7\u003c/em\u003e(1), 1\u0026ndash;9. https://doi.org/10.1038/ncomms12747\u003c/li\u003e\n\u003cli\u003eBeebee, T. J. C., \u0026amp; Griffiths, R. A. (2005). The amphibian decline crisis: A watershed for conservation biology? \u003cem\u003eBiological Conservation\u003c/em\u003e, \u003cem\u003e125\u003c/em\u003e(3), 271\u0026ndash;285. https://doi.org/10.1016/J.BIOCON.2005.04.009\u003c/li\u003e\n\u003cli\u003eBleher, B., Uster, D., \u0026amp; Bergsdorf, T. (2006). Assessment of threat status and management effectiveness in Kakamega Forest, Kenya. \u003cem\u003eBiodiversity and Conservation\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(4), 1159\u0026ndash;1177. https://doi.org/10.1007/S10531-004-3509-3/METRICS\u003c/li\u003e\n\u003cli\u003eBowker, J. N., De Vos, A., Ament, J. M., \u0026amp; Cumming, G. S. (2017). Effectiveness of Africa\u0026rsquo;s tropical protected areas for maintaining forest cover. \u003cem\u003eConservation Biology\u003c/em\u003e, \u003cem\u003e31\u003c/em\u003e(3), 559\u0026ndash;569. https://doi.org/10.1111/COBI.12851\u003c/li\u003e\n\u003cli\u003eCBD. (2020). \u003cem\u003eGlobal Biodiversity Outlook 5\u003c/em\u003e. www.emdashdesign.ca\u003c/li\u003e\n\u003cli\u003eCBD. (2022). \u003cem\u003eDecision adopted by the conference of the parties to the Convention on Biological Diversity 15/4. Kunming-Montreal Global Biodiversity Framework\u003c/em\u003e. https://www.cbd.int/conferences/post20202CBD/WG8J/11/7,CBD/SBSTTA/23/9,CBD/SBSTTA/24/12andCBD/SBI/3/21,respectively.\u003c/li\u003e\n\u003cli\u003eChen, H., Zhang, T., Costanza, R., \u0026amp; Kubiszewski, I. (2023). Review of the approaches for assessing protected areas\u0026rsquo; effectiveness. \u003cem\u003eEnvironmental Impact Assessment Review\u003c/em\u003e, \u003cem\u003e98\u003c/em\u003e. https://doi.org/10.1016/J.EIAR.2022.106929\u003c/li\u003e\n\u003cli\u003eCoetzee, B. W. T., Gaston, K. J., \u0026amp; Chown, S. L. (2014). Local Scale Comparisons of Biodiversity as a Test for Global Protected Area Ecological Performance: A Meta-Analysis. \u003cem\u003ePLOS ONE\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e(8), e105824. https://doi.org/10.1371/JOURNAL.PONE.0105824\u003c/li\u003e\n\u003cli\u003eCollaboration for Environmental Evidence. (2022). \u003cem\u003eGuidelines and Standards for Evidence synthesis in Environmental Management: Vol. Version 5.1\u003c/em\u003e (AS Pullin, GK Frampton, B Livoreil, \u0026amp; G Petrokofsky, Eds.). https://environmentalevidence.org/information-for-authors/\u003c/li\u003e\n\u003cli\u003eCurtis, P. G., Slay, C. M., Harris, N. L., Tyukavina, A., \u0026amp; Hansen, M. C. (2018). Classifying drivers of global forest loss. \u003cem\u003eScience\u003c/em\u003e, \u003cem\u003e361\u003c/em\u003e(6407), 1108\u0026ndash;1111. https://doi.org/10.1126/SCIENCE.AAU3445/SUPPL_FILE/AAU3445_CURTIS_SM.PDF\u003c/li\u003e\n\u003cli\u003eDur\u0026aacute;n, A. P., Barbosa, O., \u0026amp; Gaston, K. J. (2022). Understanding the interacting factors that determine ecological effectiveness of terrestrial protected areas. In \u003cem\u003eJournal for Nature Conservation\u003c/em\u003e (Vol. 70, p. 126264). Urban \u0026amp; Fischer. https://doi.org/10.1016/j.jnc.2022.126264\u003c/li\u003e\n\u003cli\u003eEklund, J., Jones, J. P. G., R\u0026auml;s\u0026auml;nen, M., Geldmann, J., Jokinen, A. P., Pellegrini, A., Rakotobe, D., Rakotonarivo, O. S., Toivonen, T., \u0026amp; Balmford, A. (2022). Elevated fires during COVID-19 lockdown and the vulnerability of protected areas. \u003cem\u003eNature Sustainability 2022 5:7\u003c/em\u003e, \u003cem\u003e5\u003c/em\u003e(7), 603\u0026ndash;609. https://doi.org/10.1038/s41893-022-00884-x\u003c/li\u003e\n\u003cli\u003eGeldmann, J. (2023). Safeguarding biodiversity requires understanding how to manage protected areas cost effectively. \u003cem\u003eOne Earth\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e(2), 73\u0026ndash;76. https://doi.org/10.1016/j.oneear.2023.01.008\u003c/li\u003e\n\u003cli\u003eGeldmann, J., Barnes, M., Coad, L., Craigie, I. D., Hockings, M., \u0026amp; Burgess, N. D. (2013). Effectiveness of terrestrial protected areas in reducing habitat loss and population declines. In \u003cem\u003eBiological Conservation\u003c/em\u003e (Vol. 161, pp. 230\u0026ndash;238). https://doi.org/10.1016/j.biocon.2013.02.018\u003c/li\u003e\n\u003cli\u003eGeldmann, J., Jones, J. P. G., Wauchope, H., \u0026amp; Ferraro, P. J. (2025). Causal claims, causal assumptions and protected area impact. \u003cem\u003eNature 2025 638:8052\u003c/em\u003e, \u003cem\u003e638\u003c/em\u003e(8052), E40\u0026ndash;E41. https://doi.org/10.1038/s41586-024-08512-8\u003c/li\u003e\n\u003cli\u003eGeldmann, J., Joppa, L. N., \u0026amp; Burgess, N. D. (2014). Mapping Change in Human Pressure Globally on Land and within Protected Areas. \u003cem\u003eConservation Biology\u003c/em\u003e, \u003cem\u003e28\u003c/em\u003e(6), 1604\u0026ndash;1616. https://doi.org/10.1111/COBI.12332\u003c/li\u003e\n\u003cli\u003eGeldmann, J., Manica, A., Burgess, N. D., Coad, L., \u0026amp; Balmford, A. (2019). A global-level assessment of the effectiveness of protected areas at resisting anthropogenic pressures. \u003cem\u003eProceedings of the National Academy of Sciences of the United States of America\u003c/em\u003e, \u003cem\u003e116\u003c/em\u003e(46), 23209\u0026ndash;23215. https://doi.org/10.1073/PNAS.1908221116/SUPPL_FILE/PNAS.1908221116.SAPP.PDF\u003c/li\u003e\n\u003cli\u003eGonson, C., Pelletier, D., Alban, F., Giraud-Carrier, C., \u0026amp; Ferraris, J. (2017). Influence of settings management and protection status on recreational uses and pressures in marine protected areas. \u003cem\u003eJournal of Environmental Management\u003c/em\u003e, \u003cem\u003e200\u003c/em\u003e, 170\u0026ndash;185. https://doi.org/10.1016/J.JENVMAN.2017.05.051\u003c/li\u003e\n\u003cli\u003eGraham, V., Geldmann, J., Adams, V. M., Grech, A., Deinet, S., \u0026amp; Chang, H. C. (2021). Management resourcing and government transparency are key drivers of biodiversity outcomes in Southeast Asian protected areas. \u003cem\u003eBiological Conservation\u003c/em\u003e, \u003cem\u003e253\u003c/em\u003e, 108875. https://doi.org/10.1016/J.BIOCON.2020.108875\u003c/li\u003e\n\u003cli\u003eGurevitch, J., Koricheva, J., Nakagawa, S., \u0026amp; Stewart, G. (2018). Meta-analysis and the science of research synthesis. \u003cem\u003eNature 2018 555:7695\u003c/em\u003e, \u003cem\u003e555\u003c/em\u003e(7695), 175\u0026ndash;182. https://doi.org/10.1038/nature25753\u003c/li\u003e\n\u003cli\u003eHern\u0026aacute;ndez-Y\u0026aacute;\u0026ntilde;ez, H., Kos, J. T., Bast, M. D., Griggs, J. L., Hage, P. A., Killian, A., Loza, M. I., Whitmore, M. B., \u0026amp; Smith, A. B. (2016). A systematic assessment of threats affecting the rare plants of the United States. \u003cem\u003eBiological Conservation\u003c/em\u003e, \u003cem\u003e203\u003c/em\u003e, 260\u0026ndash;267. https://doi.org/10.1016/J.BIOCON.2016.10.009\u003c/li\u003e\n\u003cli\u003eHirons, S., Matilda Collins, C., \u0026amp; Singh, M. (2022). Assessing variation in the effectiveness of IUCN protected area categorisation. What remotely sensed forest integrity and human modification reveals across the major tropical forest biomes. \u003cem\u003eEcological Indicators\u003c/em\u003e, \u003cem\u003e143\u003c/em\u003e, 109337. https://doi.org/10.1016/J.ECOLIND.2022.109337\u003c/li\u003e\n\u003cli\u003eHof, C., Ara\u0026uacute;jo, M. B., Jetz, W., \u0026amp; Rahbek, C. (2011). Additive threats from pathogens, climate and land-use change for global amphibian diversity. \u003cem\u003eNature 2011 480:7378\u003c/em\u003e, \u003cem\u003e480\u003c/em\u003e(7378), 516\u0026ndash;519. https://doi.org/10.1038/nature10650\u003c/li\u003e\n\u003cli\u003eIPBES. (2019). \u003cem\u003eGlobal assessment report on biodiversity and ecosystem services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services\u003c/em\u003e. https://doi.org/10.5281/ZENODO.6417333\u003c/li\u003e\n\u003cli\u003eIUCN. (2024). \u003cem\u003eIUCN - Threats Classification Scheme (Version 3.3)\u003c/em\u003e. https://www.iucnredlist.org/resources/threat-classification-scheme\u003c/li\u003e\n\u003cli\u003eJones, K. R., Venter, O., Fuller, R. A., Allan, J. R., Maxwell, S. L., Negret, P. J., \u0026amp; Watson, J. E. M. (2018). One-third of global protected land is under intense human pressure. \u003cem\u003eScience\u003c/em\u003e, \u003cem\u003e360\u003c/em\u003e(6390), 788\u0026ndash;791. https://doi.org/10.1126/SCIENCE.AAP9565/SUPPL_FILE/AAP9565_JONES_SM.PDF\u003c/li\u003e\n\u003cli\u003eJoppa, L. N., \u0026amp; Pfaff, A. (2009). High and Far: Biases in the Location of Protected Areas. \u003cem\u003ePLOS ONE\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e(12), e8273. https://doi.org/10.1371/JOURNAL.PONE.0008273\u003c/li\u003e\n\u003cli\u003eJoppa, L. N., \u0026amp; Pfaff, A. (2011). Global protected area impacts. \u003cem\u003eProceedings of the Royal Society B: Biological Sciences\u003c/em\u003e, \u003cem\u003e278\u003c/em\u003e(1712), 1633\u0026ndash;1638. https://doi.org/10.1098/RSPB.2010.1713\u003c/li\u003e\n\u003cli\u003eKennedy, C. M., Oakleaf, J. R., Theobald, D. M., Baruch-Mordo, S., \u0026amp; Kiesecker, J. (2019). Managing the middle: A shift in conservation priorities based on the global human modification gradient. \u003cem\u003eGlobal Change Biology\u003c/em\u003e, \u003cem\u003e25\u003c/em\u003e(3), 811\u0026ndash;826. https://doi.org/10.1111/GCB.14549\u003c/li\u003e\n\u003cli\u003eKonno, K., Livoreil, B., \u0026amp; Pullin, A. (2021). \u003cem\u003eCollaboration for Environmental Evidence Critical Appraisal Tool Version 0.3 (Prototype)\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eLawrence, A., Friedrich, F., \u0026amp; Beierkuhnlein, C. (2021). Landscape fragmentation of the Natura 2000 network and its surrounding areas. \u003cem\u003ePLOS ONE\u003c/em\u003e, \u003cem\u003e16\u003c/em\u003e(10), e0258615. https://doi.org/10.1371/JOURNAL.PONE.0258615\u003c/li\u003e\n\u003cli\u003eLecl\u0026egrave;re, D., Obersteiner, M., Barrett, M., Butchart, S. H. M., Chaudhary, A., de Palma, A., DeClerck, F. A. J., di Marco, M., Doelman, J. C., D\u0026uuml;rauer, M., Freeman, R., Harfoot, M., Hasegawa, T., Hellweg, S., Hilbers, J. P., Hill, S. L. L., Humpen\u0026ouml;der, F., Jennings, N., Krisztin, T., \u0026hellip; Young, L. (2020). Bending the curve of terrestrial biodiversity needs an integrated strategy. \u003cem\u003eNature 2020 585:7826\u003c/em\u003e, \u003cem\u003e585\u003c/em\u003e(7826), 551\u0026ndash;556. https://doi.org/10.1038/s41586-020-2705-y\u003c/li\u003e\n\u003cli\u003eLi, G., Fang, C., Watson, J. E. M., Sun, S., Qi, W., Wang, Z., \u0026amp; Liu, J. (2024). Mixed effectiveness of global protected areas in resisting habitat loss. \u003cem\u003eNature Communications 2024 15:1\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(1), 1\u0026ndash;17. https://doi.org/10.1038/s41467-024-52693-9\u003c/li\u003e\n\u003cli\u003eLi, S., Su, S., Liu, Y., Zhou, X., Luo, Q., \u0026amp; Paudel, B. (2022). Effectiveness of the Qilian Mountain Nature Reserve of China in Reducing Human Impacts. \u003cem\u003eLand 2022, Vol. 11, Page 1071\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(7), 1071. https://doi.org/10.3390/LAND11071071\u003c/li\u003e\n\u003cli\u003eMammides, C. (2020). A global assessment of the human pressure on the world\u0026rsquo;s lakes. \u003cem\u003eGlobal Environmental Change\u003c/em\u003e, \u003cem\u003e63\u003c/em\u003e, 102084. https://doi.org/10.1016/J.GLOENVCHA.2020.102084\u003c/li\u003e\n\u003cli\u003eMargules, C. R., \u0026amp; Pressey, R. L. (2000). Systematic conservation planning. \u003cem\u003eNature 2000 405:6783\u003c/em\u003e, \u003cem\u003e405\u003c/em\u003e(6783), 243\u0026ndash;253. https://doi.org/10.1038/35012251\u003c/li\u003e\n\u003cli\u003eMart\u0026iacute;nez-Fern\u0026aacute;ndez, J., Ruiz-Benito, P., \u0026amp; Zavala, M. A. (2015). Recent land cover changes in Spain across biogeographical regions and protection levels: Implications for conservation policies. \u003cem\u003eLand Use Policy\u003c/em\u003e, \u003cem\u003e44\u003c/em\u003e, 62\u0026ndash;75. https://doi.org/10.1016/J.LANDUSEPOL.2014.11.021\u003c/li\u003e\n\u003cli\u003eMen\u0026eacute;ndez-Guerrero, P. A., \u0026amp; Graham, C. H. (2013). Evaluating multiple causes of amphibian declines of Ecuador using geographical quantitative analyses. \u003cem\u003eEcography\u003c/em\u003e, \u003cem\u003e36\u003c/em\u003e(7), 756\u0026ndash;769. https://doi.org/10.1111/J.1600-0587.2012.07877.X\u003c/li\u003e\n\u003cli\u003eMingarro, M., \u0026amp; Lobo, J. M. (2023). European National Parks protect their surroundings but not everywhere: A study using land use/land cover dynamics derived from CORINE Land Cover data. \u003cem\u003eLand Use Policy\u003c/em\u003e, \u003cem\u003e124\u003c/em\u003e, 106434. https://doi.org/10.1016/J.LANDUSEPOL.2022.106434\u003c/li\u003e\n\u003cli\u003eMollmann, V. H. D. S., Santos, S., Fernandes, G., Mossolin, E. C., Dalosto, M. M., Cardoso, S. M. V. S., Prestes, O. D., Zanella, R., \u0026amp; Bartholomei-Santos, M. L. (2022). Terrestrial protected areas do not fully shield their streams from exogenous stressors. \u003cem\u003eEnvironmental Conservation\u003c/em\u003e, \u003cem\u003e49\u003c/em\u003e(4), 215\u0026ndash;224. https://doi.org/10.1017/S0376892922000261\u003c/li\u003e\n\u003cli\u003ePulido-Chadid, K., Virtanen, E., \u0026amp; Geldmann, J. (2023). How effective are protected areas for reducing threats to biodiversity? A systematic review protocol. \u003cem\u003eEnvironmental Evidence\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(1), 1\u0026ndash;10. https://doi.org/10.1186/S13750-023-00311-4/TABLES/5\u003c/li\u003e\n\u003cli\u003ePullin, A. S., \u0026amp; Stewart, G. B. (2006). Guidelines for systematic review in conservation and environmental management. \u003cem\u003eConservation Biology : The Journal of the Society for Conservation Biology\u003c/em\u003e, \u003cem\u003e20\u003c/em\u003e(6), 1647\u0026ndash;1656. https://doi.org/10.1111/J.1523-1739.2006.00485.X\u003c/li\u003e\n\u003cli\u003eR Core Team. (2021). \u003cem\u003eR: A Language and Environment for Statistical Computing\u003c/em\u003e. https://www.r-project.org/\u003c/li\u003e\n\u003cli\u003eRichardson, K., Steffen, W., Lucht, W., Bendtsen, J., Cornell, S. E., Donges, J. F., Dr\u0026uuml;ke, M., Fetzer, I., Bala, G., von Bloh, W., Feulner, G., Fiedler, S., Gerten, D., Gleeson, T., Hofmann, M., Huiskamp, W., Kummu, M., Mohan, C., Nogu\u0026eacute;s-Bravo, D., \u0026hellip; Rockstr\u0026ouml;m, J. (2023). Earth beyond six of nine planetary boundaries. \u003cem\u003eScience Advances\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e(37). https://doi.org/10.1126/SCIADV.ADH2458\u003c/li\u003e\n\u003cli\u003eRodrigues, A. S. L., \u0026amp; Cazalis, V. (2020). The multifaceted challenge of evaluating protected area effectiveness. \u003cem\u003eNature Communications 2020 11:1\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(1), 1\u0026ndash;4. https://doi.org/10.1038/s41467-020-18989-2\u003c/li\u003e\n\u003cli\u003eSalafsky, N., Relton, C., Young, B. E., Lamarre, P., B\u0026ouml;hm, M., Ch\u0026eacute;nier, M., Cochrane, E., Dionne, M., He, K. K., Hilton-Taylor, C., Latr\u0026eacute;mouille, C., Morrison, J., Raymond, C. V., Seddon, M., \u0026amp; Suresh, V. (2025). Classification of direct threats to the conservation of ecosystems and species 4.0. \u003cem\u003eConservation Biology\u003c/em\u003e, e14434. https://doi.org/10.1111/COBI.14434\u003c/li\u003e\n\u003cli\u003eSanderson, E. W., Jaiteh, M., Levy, M. A., Redford, K. H., Wannebo, A. V, \u0026amp; Woolmer, G. (2002). The Human Footprint and the Last of the Wild: The human footprint is a global map of human influence on the land surface, which suggests that human beings are stewards of nature, whether we like it or not. \u003cem\u003eBioScience\u003c/em\u003e, \u003cem\u003e52\u003c/em\u003e(10), 891\u0026ndash;904. https://doi.org/10.1641/0006-3568(2002)052[0891:THFATL]2.0.CO;2\u003c/li\u003e\n\u003cli\u003eSchleicher, J., Eklund, J., D. Barnes, M., Geldmann, J., Oldekop, J. A., \u0026amp; Jones, J. P. G. (2020). Statistical matching for conservation science. \u003cem\u003eConservation Biology\u003c/em\u003e, \u003cem\u003e34\u003c/em\u003e(3), 538\u0026ndash;549. https://doi.org/10.1111/COBI.13448\u003c/li\u003e\n\u003cli\u003eSchleicher, J., Peres, C. A., \u0026amp; Leader-Williams, N. (2019). Conservation performance of tropical protected areas: How important is management? \u003cem\u003eConservation Letters\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(5), e12650. https://doi.org/10.1111/CONL.12650\u003c/li\u003e\n\u003cli\u003eSchulze, K., Knights, K., Coad, L., Geldmann, J., Leverington, F., Eassom, A., Marr, M., Butchart, S. H. M., Hockings, M., \u0026amp; Burgess, N. D. (2018). An assessment of threats to terrestrial protected areas. \u003cem\u003eConservation Letters\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(3). https://doi.org/10.1111/CONL.12435\u003c/li\u003e\n\u003cli\u003eSoares-Filho, B. S., Oliveira, U., Ferreira, M. N., Marques, F. F. C., de Oliveira, A. R., Silva, F. R., \u0026amp; B\u0026ouml;rner, J. (2023). Contribution of the Amazon protected areas program to forest conservation. \u003cem\u003eBiological Conservation\u003c/em\u003e, \u003cem\u003e279\u003c/em\u003e, 109928. https://doi.org/https://doi.org/10.1016/j.biocon.2023.109928\u003c/li\u003e\n\u003cli\u003eStuart, S. N., Chanson, J. S., Cox, N. A., Young, B. E., Rodrigues, A. S. L., Fischman, D. L., \u0026amp; Waller, R. W. (2004). Status and trends of amphibian declines and extinctions worldwide. \u003cem\u003eScience\u003c/em\u003e, \u003cem\u003e306\u003c/em\u003e(5702), 1783\u0026ndash;1786. https://doi.org/10.1126/SCIENCE.1103538/SUPPL_FILE/STUART.SOM.PDF\u003c/li\u003e\n\u003cli\u003eTerraube, J., Van doninck, J., Helle, P., \u0026amp; Cabeza, M. (2020). Assessing the effectiveness of a national protected area network for carnivore conservation. \u003cem\u003eNature Communications 2020 11:1\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(1), 1\u0026ndash;9. https://doi.org/10.1038/s41467-020-16792-7\u003c/li\u003e\n\u003cli\u003eThe EndNote Team. (2013). \u003cem\u003eEndNote\u003c/em\u003e (EndNote 20 Version). Clarivate.\u003c/li\u003e\n\u003cli\u003eUNEP-WCMC. (2024, May). \u003cem\u003eExplore the World\u0026rsquo;s Protected Areas\u003c/em\u003e. https://www.protectedplanet.net/en\u003c/li\u003e\n\u003cli\u003eVenter, O., Sanderson, E. W., Magrach, A., Allan, J. R., Beher, J., Jones, K. R., Possingham, H. P., Laurance, W. F., Wood, P., Fekete, B. M., Levy, M. A., \u0026amp; Watson, J. E. M. (2016a). Global terrestrial Human Footprint maps for 1993 and 2009. \u003cem\u003eScientific Data 2016 3:1\u003c/em\u003e, \u003cem\u003e3\u003c/em\u003e(1), 1\u0026ndash;10. https://doi.org/10.1038/sdata.2016.67\u003c/li\u003e\n\u003cli\u003eVenter, O., Sanderson, E. W., Magrach, A., Allan, J. R., Beher, J., Jones, K. R., Possingham, H. P., Laurance, W. F., Wood, P., Fekete, B. M., Levy, M. A., \u0026amp; Watson, J. E. M. (2016b). Sixteen years of change in the global terrestrial human footprint and implications for biodiversity conservation. \u003cem\u003eNature Communications 2016 7:1\u003c/em\u003e, \u003cem\u003e7\u003c/em\u003e(1), 1\u0026ndash;11. https://doi.org/10.1038/ncomms12558\u003c/li\u003e\n\u003cli\u003eWatson, J. E. M., Dudley, N., Segan, D. B., \u0026amp; Hockings, M. (2014). The performance and potential of protected areas. \u003cem\u003eNature 2014 515:7525\u003c/em\u003e, \u003cem\u003e515\u003c/em\u003e(7525), 67\u0026ndash;73. https://doi.org/10.1038/nature13947\u003c/li\u003e\n\u003cli\u003eWauchope, H. S., Jones, J. P. G., Geldmann, J., Simmons, B. I., Amano, T., Blanco, D. E., Fuller, R. A., Johnston, A., Langendoen, T., Mundkur, T., Nagy, S., \u0026amp; Sutherland, W. J. (2022). Protected areas have a mixed impact on waterbirds, but management helps. \u003cem\u003eNature 2022 605:7908\u003c/em\u003e, \u003cem\u003e605\u003c/em\u003e(7908), 103\u0026ndash;107. https://doi.org/10.1038/s41586-022-04617-0\u003c/li\u003e\n\u003cli\u003eWu, H., Fang, S., Yang, Y., \u0026amp; Cheng, J. (2022). Changes in habitat quality of nature reserves in depopulating areas due to anthropogenic pressure: Evidence from Northeast China, 2000\u0026ndash;2018. \u003cem\u003eEcological Indicators\u003c/em\u003e, \u003cem\u003e138\u003c/em\u003e, 108844. https://doi.org/10.1016/J.ECOLIND.2022.108844\u003c/li\u003e\n\u003cli\u003eXu, Y., Price, M., Yang, B., Zhang, K., Yang, N., Tang, X., Ran, J., Yi, Y., \u0026amp; Wang, B. (2022). Have China\u0026rsquo;s national forest reserves designated since 1990 conserved forests effectively? \u003cem\u003eJournal of Environmental Management\u003c/em\u003e, \u003cem\u003e306\u003c/em\u003e, 114485. https://doi.org/10.1016/J.JENVMAN.2022.114485\u003c/li\u003e\n\u003cli\u003eZhu, Z., Qiu, S., \u0026amp; Ye, S. (2022). Remote sensing of land change: A multifaceted perspective. \u003cem\u003eRemote Sensing of Environment\u003c/em\u003e, \u003cem\u003e282\u003c/em\u003e, 113266. https://doi.org/10.1016/J.RSE.2022.113266\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"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":"environmental-evidence","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"enev","sideBox":"Learn more about [Environmental Evidence](http://environmentalevidencejournal.biomedcentral.com)","snPcode":"13750","submissionUrl":"https://submission.springernature.com/new-submission/13750/3","title":"Environmental Evidence","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Protected areas, Conservation, Threats to biodiversity, Protected areas effectiveness, Threat reduction in protected areas","lastPublishedDoi":"10.21203/rs.3.rs-7046705/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7046705/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eProtected areas (PAs) are central to global biodiversity conservation efforts, yet their effectiveness in reducing threats is context-dependent and poorly understood. This systematic review evaluates how well PAs mitigate anthropogenic threats to biodiversity by synthesizing findings from peer-reviewed and grey literature. Specifically, we assess how threat levels change in relation to protection status, over time, and in response to specific management interventions. The review aims to evaluate threat levels inside PAs relative to appropriate comparators or interventions, identify variation across threat types, and highlight key factors associated with both successful and unsuccessful conservation outcomes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe conducted a comprehensive systematic review, screening 5,687 unique articles and including 105, that assessed changes in threat within at least one PA compared to other locations, time periods, or types of interventions. A descriptive, narrative synthesis was performed to evaluate the strength of study designs and the influence of methodological differences on reported outcomes. The synthesis focused on three key dimensions: (1) the proportion of studies reporting reduced threats inside PAs over time or relative to comparators; (2) effectiveness across different threat types; and (3) contextual and management-related factors influencing PA performance.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eReview findings\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOur review revealed three major findings:\u003c/p\u003e\n\u003cp\u003eFirstly, PAs on average had lower threat levels compared to unprotected areas, but effectiveness was mixed. Most studies reported increasing threats both inside and outside PAs. Nonetheless, threats tended to be lower within PAs boundaries. Secondly, the threat reduction potential depends on the type of threat. PAs were more effective at reducing land-use and land-cover change (LULCC), fire, and hunting. Studies frequently reported lower levels of deforestation, agricultural expansion, and infrastructure development within PAs. However, PAs were less effective in addressing threats like invasive alien species, pollution, and recreation, likely due to external origins or indirect impacts that extend beyond formal boundaries. Thirdly, PA outcomes were shaped by multiple interrelated factors. Effectiveness was generally lower in areas with high baseline threats, high human population density, and easy access. In contrast, strong governance, adequate funding, robust enforcement, large PA size, and alignment with national development goals and governance were associated with more positive conservation outcomes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDespite rising human-induced threats across landscapes, PAs demonstrated better performance than unprotected landscapes in reducing threats to biodiversity. However, their success is highly dependent on the context, including the type of threat, biophysical characteristics, and socio-economic context where PAs are located. This review highlights the urgent need to enhance the effectiveness of existing PAs and offers important insights for the planning and implementation of future ones.\u003c/p\u003e\n\u003cp\u003eThe potential of PAs to reduce threats is not solely determined by their design or location but also by how well they are integrated into the local socio-ecological systems in which they operate. Strengthening governance, improving management strategies, and ensuring the meaningful involvement of local stakeholders are essential to ensure long-term conservation outcomes.\u003c/p\u003e\n\u003cp\u003eAdditionally, this review underscores the need for standardized methodologies, consistent threat assessment metrics, and rigorous evaluation frameworks to better understand what drives PA effectiveness and to support evidence-based decision-making.\u003c/p\u003e","manuscriptTitle":"A systematic review of the effectiveness of protected areas at reducing threats to biodiversity","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-11 08:26:15","doi":"10.21203/rs.3.rs-7046705/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-09-06T09:22:38+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-06T09:19:29+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"95028451766901184529789261178085110724","date":"2025-09-06T08:59:03+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-28T06:25:22+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"223692747731919345419242886310224547051","date":"2025-07-18T20:41:27+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"75034028105888375631000123756800863326","date":"2025-07-17T11:33:51+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-07-12T09:00:03+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-07-07T08:54:43+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-07-07T08:52:44+00:00","index":"","fulltext":""},{"type":"submitted","content":"Environmental Evidence","date":"2025-07-04T12:11:10+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"environmental-evidence","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"enev","sideBox":"Learn more about [Environmental Evidence](http://environmentalevidencejournal.biomedcentral.com)","snPcode":"13750","submissionUrl":"https://submission.springernature.com/new-submission/13750/3","title":"Environmental Evidence","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"eeda1458-17ef-4ef6-b2ea-b046f6526f32","owner":[],"postedDate":"July 11th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-10T22:53:19+00:00","versionOfRecord":[],"versionCreatedAt":"2025-07-11 08:26:15","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7046705","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7046705","identity":"rs-7046705","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}