Targeting ocean conservation outcomes through threat reduction

The study addresses how to target ocean conservation actions to achieve measurable reductions in global species extinction risk. Despite rapid expansion of Marine Protected Areas (MPAs), marine biodiversity loss continues, partly because designations often minimize socioeconomic conflict rather than maximize biodiversity outcomes and are not always effectively managed. Without focusing on threat reduction, the 30% protection target risks being met without biodiversity recovery. Oceans face accelerating, compounding impacts from human activities and climate change, with many mid-range regions under threat and iconic megafauna among the most imperiled. Reliable, globally relevant marine biodiversity metrics are scarce, hindering accountability and mainstreaming of responsibilities across sectors. The Species Threat Abatement and Restoration (STAR) metric quantifies the relative importance of minimizing different threats in different locations, based on IUCN Red List extinction risk and Area of Habitat (AOH). Previously available only for terrestrial species, this work extends the STAR threat abatement component to the marine realm to enable governments, businesses, and civil society to prioritize, target, and measure actions toward SDG 14/15, the post-2020 Global Biodiversity Framework, and the BBNJ treaty.

Overview: The marine STAR threat abatement (STARt) layer was produced to align with the terrestrial STAR approach, enabling cross-realm use. Steps included species selection, spatial data preparation, threat assessment and attribution, STARt computation, and spatial analyses.

Species selection:

• Source: IUCN Red List of Threatened Species™ database (version 2021.2 for range polygons; species data to October 2022).
• Included categories: Near Threatened (NT), Vulnerable (VU), Endangered (EN), and Critically Endangered (CR). Least Concern (LC) species excluded (weight=0). Data Deficient (DD) excluded due to uncertainty.
• Filter: Species coded by IUCN as occurring in marine habitats (biome_marine = TRUE), possibly also in terrestrial/freshwater realms.
• Taxonomic scope: All threatened and NT species within comprehensively assessed groups (≥80% of taxa assessed) using IUCN summary statistics and API, yielding 1698 species initially.
• Range data filters per IUCN mapping standards: presence = "Extant" and "Possibly Extinct"; origin = "Native", "Reintroduced", "Assisted Colonization". Four species lacking suitable polygons were excluded; polygons reviewed for 164 species in comprehensively assessed groups.

Area of Habitat (AOH):

• Built a crosswalk between IUCN Red List habitat classifications and IUCN Global Ecosystem Typology biomes (global raster layers), including major and minor occurrences.
• Combined matching habitat rasters to generate species-specific AOH.
• Cropped AOH to marine biomes (GET Level 3) to avoid overlap with terrestrial STAR and to isolate marine-relevant portions for species spanning multiple realms.

Threat attribution and expected decline:

• Used IUCN Threat Classification Scheme to list ongoing threats for each species.
• For each species–threat, used Red List scope (population proportion affected) and severity (decline rate) to estimate expected percentage population decline over 10 years or three generations, based on a standardized matrix (Table 1 in the paper).
• Where scope/severity were unknown (e.g., Anthozoa, Hydrozoa, Liliopsida, Magnoliopsida, Myriil), assigned median values: scope = Majority (50–90%); severity = Slow, Significant Declines. This procedure applied to 430 species; 18 with negligible severity across all threats (total decline=0) were excluded. After further checks, 30 species were removed, yielding a final set of 1646 species.

STARt calculation:

• For grid cell i (5 km × 5 km) and threat t: T_{i,t} = sum over species of (P_i × W × C_t), normalized as per terrestrial methodology. Here, P_i is the proportion of a species’ global current AOH in cell i; W is IUCN Red List category weight (NT=100, VU=200, EN=300, CR=400); C_t is the fraction of expected population decline attributable to threat t for that species (the threat’s expected decline divided by the sum across all threats for that species). The STARt score per cell is the sum across species; per-threat scores are disaggregated similarly.
• Assumption: Threats are uniformly distributed across each species’ AOH (spatial variation in threat magnitude not modeled).

Spatial analyses and mapping:

• Resolution: 5 km × 5 km global grid.
• Extracted statistics by country and EEZ using Natural Earth boundaries (1:50m) and Marine Boundaries Geodatabase; also summarized for Protected Areas (WDPA), Key Biodiversity Areas (KBAs), Important Marine Mammal Areas (IMMAs), and Large Marine Ecosystems (LMEs).
• Software: RStudio; packages terra, exactextractr, tidyverse, sf; mapping with terra, tidyverse, naturalearthR, maptools.
• Visualization: Classified STAR scores into categories (Very Low >0–0.1; Low 0.1–1; Medium 1–10; High 10–100; Very High >100 per 5 km cell). For combined land–sea products, overlapping species from terrestrial STAR layers were removed from marine layers to avoid double counting.

Data availability: IUCN Red List (ranges, habitats, threats); KBAs/Protected Areas databases; Marine Boundaries Geodatabase (EEZs); NOAA marine ecosystem spatial data.

• Coverage and taxa: 1646 species included (NT and threatened: CR=171, EN=293, VU=684, NT=498) spanning 11 classes, 62 orders, 192 families, 552 genera. Composition: 78% strictly marine (n=1277); 11% marine–terrestrial (n=184); 4% marine–freshwater (n=49); 7% all three realms (n=111). Largest groups: sharks and rays (n=490), reef-building corals (n=401), bony fishes (n=282), birds (n=252), mammals (n=62).
• Spatial patterns: Marine STAR scores computed globally at 5 km × 5 km resolution. Most (95%) marine cells fall into the Very Low category (>0–0.1 per cell). Marine STAR score range noted as 9.67–80.204 per cell (5 km).
• Concentration of opportunity: Only 0.001% of cells (24 cells ≈600 km²) addressing climate change (cells with STARt=10) account for almost 60% of the global marine STARt, reflecting restricted-range and highly threatened species.
• National contributions: Indonesia holds the largest share (11.5%) of global marine STARt, followed by Australia (6.9%), Mexico (4.1%), the Philippines (3.6%), Brazil (3.5%), and China (3.1%). Areas Beyond National Jurisdiction (high seas) comprise 5.7% of global STARt, spread across 42% of ocean area.
• Density hot spots: Highest STAR density (per km² of EEZ) occurs in smaller EEZs; notably Singapore (very high per-km² density), Belize, and Gibraltar. The top 10 countries by STAR density account for 3.7% of the global STAR score.
• Large Marine Ecosystems (LMEs): The Indonesian Sea LME contains 26% of the global total; the Canadian High Arctic–North Greenland LME has the lowest share (0.004%). Highest per-km² STAR scores in the Gulf of California and East China Sea (≈0.011 per km²). Polar systems have low totals and densities due to lower richness and larger species ranges.
• Protection status overlap: 24.9% of global marine STAR score lies within WDPA-listed protected areas (covering 10.2% of the area with marine STAR). Only 2.8% of global marine STAR is within no-take/fully protected areas. KBAs account for 10.8% of total marine STAR (covering 4.3% of the area), EBSAs overlap 21.2% of the area covered by marine STAR, and IMMAs overlap 4.0% of the area and capture 26.1% of the marine STAR for mammals (n=42).
• Threat importance (share of global marine STAR): • Biological resource use—Fishing & harvesting aquatic resources (code 5.4): 43.0% • Invasive non-native/alien species/diseases (8.1): 5.4% • Climate change—Habitat shifting & alteration (11.1): 4.7% • Climate change—Temperature extremes (11.3): 4.6% • Commercial & industrial areas (1.2): 4.5% • Housing & urban areas (1.1): 4.4% • Pollution—Industrial & military effluents (9.2): 4.3% • Pollution—Agricultural & urban water (9.1): 3.2% • Residential & commercial development—Urbanization (1.3): 2.9% • Pollution—Agricultural & forestry effluents (9.3): 2.8% (Subtotal listed = 79.8%)
• Policy relevance: The vast majority (≈75%) of marine STAR lies outside protected areas and very little within no-take zones (≈2.7–2.8%), indicating substantial opportunity and need for non-MPA interventions and better placement and management of MPAs. Fishing pressure reduction offers the single largest potential reduction in extinction risk.

Extending the STAR threat abatement metric to marine biodiversity identifies where and which threat-mitigation actions can most effectively reduce global extinction risk. The findings reveal a strong spatial and threat-specific concentration of opportunity: a small fraction of locations and a small number of threat classes (principally unsustainable fishing) account for a large share of global risk-reduction potential. This supports a shift from area targets alone toward targeted, outcome-focused management actions (e.g., establishing and enforcing fishing limits, bycatch mitigation) and informs prioritization for protected area placement and designation of no-take zones where they can deliver the greatest benefits. The results also demonstrate that many high-opportunity areas are outside current protected areas and that cross-jurisdictional coordination is critical, given species’ ranges span multiple EEZs and ABNJ.

Disaggregating STAR by threat and geography helps governments, businesses, and conservation organizations set science-based, location- and threat-specific targets aligned with the GBF, SDGs 14/15, and the BBNJ Agreement. The metric complements other site-based designations (KBAs, EBSAs, IMMAs) and can be calibrated with local data to improve decision-making. Notably, climate-related threats already account for a substantial portion of STAR, but temporal lags in threat manifestation and assessment imply that proactive climate mitigation and adaptation will be increasingly necessary to achieve extinction risk reductions. Overall, the marine STAR layer provides a practical, globally consistent tool to direct investments and policies toward the most consequential actions and places for marine biodiversity recovery.

This paper introduces the first global marine implementation of the STAR threat abatement metric, enabling identification and quantification of where actions can most reduce extinction risk for marine species. Key contributions include: (i) a harmonized methodology aligned with terrestrial STAR, (ii) a global, 5 km grid product disaggregable by threat and geography, (iii) evidence that reducing unsustainable fishing offers the greatest opportunity (≈43% of global STAR), and (iv) a clear demonstration that most action opportunities lie outside current protected areas, with very limited coverage by no-take zones. The metric can guide governments, the private sector, and civil society in setting and tracking science-based targets under the GBF, SDGs, and BBNJ.

Future priorities: (1) develop a marine STAR restoration component (analogous to terrestrial STAR restoration) as time-series habitat data improve; (2) expand taxonomic coverage, especially for deep-sea and polar taxa; (3) refine spatial representation of threat footprints and incorporate variation in threat magnitude, duration, frequency, and intensity; (4) improve treatment of climate threats given temporal lags; (5) continue harmonization across terrestrial, freshwater, and marine STAR layers; and (6) routinely calibrate STAR with local species occurrence and threat data to enhance accuracy for site-level decisions.

• Data gaps and biases: Many marine taxa, especially deep-sea and polar species, remain underassessed; Data Deficient species were excluded. This may bias STAR toward better-studied regions and taxa.
• Threat attribution uncertainty: Scope and severity are not mandatory in Red List assessments; where missing, median values were imputed, introducing uncertainty. Eighteen species with negligible severity were excluded.
• Uniform threat distribution assumption: Threats were assumed uniformly distributed across a species’ AOH; spatial heterogeneity in threat intensity was not modeled.
• Climate threat lag: IUCN Red List captures threats over a 10-year or three-generation horizon; emerging climate impacts may be underestimated, making STAR a lagging indicator for some species.
• Protection overlap interpretation: Overlaps with MPAs, KBAs, EBSAs, IMMAs are constrained by the current, incomplete coverage and designation processes in the marine realm.
• Restoration not included: The marine restoration (STAR restoration) component was not developed due to lack of historical habitat extent data from remote sensing, limiting guidance on restoration opportunities.
• Methodological inconsistencies and typos: Minor inconsistencies in notation and some reported values reflect constraints of source data and text but do not alter the core methodology and results.