Temporal dynamics of climate change exposure and opportunities for global marine biodiversity

Marine organisms are particularly sensitive to warming due to narrow thermal safety margins and close tracking of thermal limits across their latitudinal ranges. Climate change is driving range shifts and community turnover, with declines and local extirpations near equatorward range edges and species gains at higher latitudes. While many studies project redistribution of suitable habitats by mid- to late-century, a key gap is the lack of temporal detail on when and how fast thermally driven habitat losses (exposure) and gains (opportunities) will occur. Understanding the balance and relative timing of exposure versus opportunity is crucial because early and persistent opportunities can facilitate range expansions and resilience, whereas exposure can drive contractions and local losses. The study aims to model the global temporal dynamics of thermal exposure and opportunity for marine species assemblages through 2100, focusing on the epipelagic layer and providing an early-warning framework for monitoring and conservation.

Prior work shows marine ectotherms track thermal isotherms more closely than terrestrial ectotherms and have narrower safety margins, leading to higher rates of local extirpation, colonisation, richness change and turnover in the oceans. Projections using climate velocity and species distribution models indicate future losses of suitable habitats in the tropics and gains at higher latitudes, with increased species invasions and turnover. However, many studies focus on one or two future time slices, limiting insight on timing and abruptness. Recent analyses using annual or daily climate projections have explored exposure timing but have paid less attention to the emergence and persistence of thermal opportunities. Studies also highlight that other abiotic factors (oxygen, pH) and biotic interactions can modulate realised shifts, and that dispersal and ocean currents shape coupling between climate change and biogeographical responses. This work integrates these insights by quantifying year-by-year exposure and opportunity dynamics within a common framework.

Scope and species: The analysis considers 21,696 marine species (33 phyla; chordates 44%—mostly fish—molluscs 22%, arthropods 16%, cnidarians 6%, echinoderms 4%). Species occurring exclusively below 200 m were excluded to focus on the epipelagic layer; 88% of species occur within 0–50 m depth. Climate data: Yearly sea surface temperature (SST) projections were derived from nine ISIMIP3b-selected CMIP6 models (CanESM5, CNRM-CM6-1, CNRM-ESM2-1, EC-Earth3, IPSL-CM6A-LR, MIROC6, MPI-ESM1-2-HR, MRI-ESM2-0, UKESM1-0-LL). Historical (1850–2014) and future (2015–2100) monthly SSTs for SSP1-2.6, SSP2-4.5, SSP5-8.5 were averaged to annual means on an equal-area 100 km grid. Species distributions: AquaMaps predicted ranges were used with a probability threshold ≥0.5 to define presence, aligning with standard practice. Sensitivity analyses compared ≥0.5 vs >0 thresholds (higher exposure under SSP5-8.5 with >0 due to larger ranges). Ranges were rasterised to the 100 km grid. Realised thermal niche limits: For each species, annual mean SST across the historical period within the species’ range was used to estimate Tmax and Tmin. Outliers were removed at cell and species levels using ±3 SD filters. The mean of remaining annual maxima and minima across the range yielded realised upper (Tmax) and lower (Tmin) thermal limits. Definition of exposure and opportunities: Exposure occurs when, within a species’ native range cell, annual mean SST exceeds Tmax or falls below Tmin for at least five consecutive years; the exposure year is the first of the run. Opportunity arises when, in a previously unsuitable cell near the species’ range that meets constraints, SST falls within [Tmin, Tmax] for at least five consecutive years; the opportunity year is the first of the run. Constraints on opportunities: Opportunities were restricted to cells satisfying (a) depth suitability: overlap between cell depth (ETOPO1) and species minimum–maximum depth, with pelagic species allowed in shallower cells; (b) dispersal buffer: cells within a buffer distance from the species’ current range corresponding to average (10 km yr−1 → 860 km) and high (50 km yr−1 → 4300 km) dispersal rates over 2015–2100; (c) niche unfilling filter: cells whose temperature fell within [Tmin, Tmax] but AquaMaps classified as unsuitable were excluded. Opportunities could arise within the buffer and outside current range, and also within current range only after a prior exposure event in that cell. Assemblage metrics: For 41,220 assemblages (100 km grid cells), six metrics were computed using annual series: (i) magnitude of exposure (maximum number of resident species exposed; expressed as proportion of local richness); (ii) magnitude of opportunity (maximum number of opportunities; proportion of local richness); (iii) timing of exposure (median year of exposure events); (iv) timing of opportunity (median year of opportunity events); (v) abruptness of exposure (percent of total exposure events occurring within the decade of maximum exposure); (vi) abruptness of opportunity (percent of total opportunities occurring within the decade of maximum opportunity). Results are reported as medians across the nine climate models. Persistence and transience of opportunities: Each opportunity’s duration was tracked to 2100. Persistent opportunities remain open through 2100. Transient opportunities close if five consecutive years of temperatures beyond limits occur post-emergence. Transients were classified as cold-exposed (closure due to below Tmin) or warm-exposed (closure due to above Tmax). Year of emergence and duration distributions were summarised for each SSP and by transient type. Sensitivity analyses repeated key results with 50 km yr−1 buffers. Computation: All analyses were conducted in R 4.3.3. Data sources: CMIP6 (via ISIMIP3b), AquaMaps distributions and depth ranges, ETOPO1 bathymetry.

- Spatial patterns: Exposure is concentrated in the tropics, while most opportunities arise in temperate and polar regions. Assemblages generally show either high exposure or high opportunity, rarely both. Thermal exposure and opportunity together affect >10% of current species richness for 26% (SSP1-2.6), 34% (SSP2-4.5), and 61% (SSP5-8.5) of assemblages. - Emissions dependence: Reducing emissions from SSP5-8.5 to SSP1-2.6 reduces exposure by ~100-fold, while opportunities are reduced by about half. Assemblages with ≥10% of species exposed decline from 28% (SSP5-8.5) to <1% (SSP1-2.6). Under 10 km yr−1 dispersal, 17% of assemblages have opportunities ≥10% of current richness across all SSPs; under 50 km yr−1, this rises to 42%, 47%, and 53% for SSP1-2.6, SSP2-4.5, SSP5-8.5, respectively. - Temporal dynamics: Opportunities begin immediately (circa 2015) and increase gradually with similar trajectories across scenarios until ~2040, then continue rising, with more opportunities under higher emissions. Exposure starts later and occurs more abruptly; until ~2060, exposure is similar across scenarios, after which it accelerates under SSP2-4.5 and dramatically under SSP5-8.5, exceeding opportunities by >7× by 2100 in SSP5-8.5. - Abruptness and timing: Exposure becomes more abrupt with higher emissions; opportunities are less abrupt under higher emissions (more spread through the century). The tropics generally exhibit more abrupt exposure and opportunity. Timing of both shifts later with higher emissions, but many assemblages still see early opportunities under SSP5-8.5, especially in tropical–temperate transition zones, the South Pacific, and the Arctic. - Persistence of opportunities: Most opportunities persist to 2100: 79% (SSP1-2.6), 91% (SSP2-4.5), 97% (SSP5-8.5). Across emissions and dispersal scenarios, 76–97% of opportunities persist. Higher dispersal rates increase the number of opportunities 3–4× but maintain similar persistence proportions. - Transient opportunity mechanisms: Under SSP1-2.6 and SSP2-4.5, most transients close due to cold exposure; under SSP5-8.5, warm exposure dominates transient closures. Warm-exposed transients concentrate in the tropics and North Atlantic and generally emerge earlier and last longer (especially under SSP5-8.5) than cold-exposed transients. Cold-exposed transients are more common in temperate and polar regions, arise later and last longer under low emissions, and may increase migrant mortality risk. - Taxonomic patterns: Chordates contribute the largest share of projected exposures and opportunities in most cells, followed by molluscs and arthropods; contributions track local taxonomic richness composition. - Regional proportional change: In absolute numbers, most opportunities emerge in temperate zones; proportionally, polar regions undergo greater changes in opportunities, particularly under higher dispersal.

The results show that climate-driven thermal opportunities are already emerging and, through mid-century, accumulate faster than exposure across most scenarios. This temporal asymmetry helps explain widespread observations of faster poleward range expansions relative to equatorward contractions. Regions with high opportunity but low exposure may still face significant ecological disruption from novel species interactions, demonstrating that low exposure does not equate to low risk. Conversely, tropical regions face later but more abrupt exposure, with strong risks of near-synchronous exposure of a large fraction of resident species under high emissions. Reducing greenhouse gas emissions substantially curtails exposure, especially in the tropics, altering the balance from exposure-dominated late-century outcomes to opportunity-dominated near- and mid-term changes in temperate and polar seas. The high persistence of opportunities increases the likelihood that they translate into colonisations, potentially providing resilience (via range expansions) to some species and benefits to human communities relying on shifting resources. However, the realisation of opportunities depends on dispersal pathways, ocean currents, biotic interactions, and other abiotic constraints (oxygen, pH), which may modulate colonisation success and impacts. The framework provides an early warning system for monitoring where and when thermal changes are likely to occur, enabling adaptive conservation and management interventions.

This study introduces a unified, annual-resolution framework to quantify when and where marine species assemblages will face thermal exposure and gain thermal opportunities at the ocean surface. Opportunities generally arise earlier and more gradually, are concentrated at higher latitudes, and largely persist to 2100, suggesting they will drive near- and mid-term changes, even under rapid emissions reductions. Under continued high emissions, late-century dynamics shift toward abrupt, widespread exposure, especially in tropical regions. The approach and outputs can inform early warning, monitoring, and adaptive management, highlighting hotspots for impending colonisations and extirpations. Future research should integrate additional abiotic variables, biotic interactions, and vertical thermal structure to assess three-dimensional habitat dynamics and to connect thermal opportunities and exposure to realised biodiversity outcomes.

- Environmental scope: Analyses use sea surface temperature only and focus on the epipelagic layer; vertical dynamics and deeper habitats are not modelled, limiting extrapolation beyond the surface layer. - Niche estimation: Realised thermal limits derived from historical distributions may be narrower than fundamental tolerances and can be affected by past range shifts; population-level adaptation and plasticity are not explicitly modelled. - Opportunity realisation: The framework identifies thermal suitability, not actual colonisations; does not incorporate species-specific dispersal pathways, currents as barriers/facilitators, biotic interactions, or additional abiotic constraints (e.g., oxygen, pH, salinity), potentially overestimating effective opportunities. - Exposure criteria: Requiring five consecutive years beyond limits may underestimate risk for sensitive taxa (e.g., corals) and miss shorter damaging events; conversely, broader niches or acclimation could delay exposure. - Dispersal representation: Opportunities are constrained within static dispersal buffers (10 and 50 km yr−1) rather than annual incremental spread; rare long-distance events are implicitly allowed but routes are not modelled. - Species data: Use of AquaMaps probability threshold (≥0.5) affects range size and thus exposure/opportunity counts; although sensitivity analyses show broad robustness, SSP5-8.5 exposure is higher with lower thresholds. Cells classified as unsuitable by AquaMaps are excluded to account for niche unfilling, which may exclude thermally suitable but otherwise constrained sites.