A wake-up call for ocean conservation

The world’s oceans are rapidly changing due to climate change, and our primary tool for marine conservation — Marine Protected Areas (MPAs) — may not be prepared to withstand the impact. For biodiversity data science, understanding this vulnerability is critical for developing effective, forward-looking conservation strategies.

Addressing the knowledge gap

Previous studies evaluating climate change effects on European MPAs often relied on simplistic metrics like climate velocity or sea surface temperature changes. These approaches neglected other biologically vital environmental factors, such as oxygen levels, pH, and primary productivity, which influence species distribution. Crucially, these earlier efforts also considered only climatic changes, disregarding species-specific niches. Without a comprehensive, multivariate assessment, the true exposure of threatened and commercially important species within MPAs remained an uncertainty.

Our main approach: multivariate climate novelty analysis

To close this gap, we conducted a multivariate climate novelty analysis on 398 threatened and commercially important species across 1,890 European MPAs. This approach assesses the future exposure of species to novel and extremely novel climatic conditions.

The analysis was based on four key distribution drivers for future climate novelty:

  • Maximum temperature
  • Average dissolved oxygen
  • pH
  • Primary productivity (PP)


We modeled conditions from the present day (2010–2020) to the end of the 21st century (2090–2100) under two contrasting Shared Socioeconomic Pathway (SSP) scenarios: a low emission (SSP1-1.9) and a high emission (SSP5-8.5) pathway.

Technological challenge: quantifying 'novel' climate

Our core technical challenge was to robustly quantify climate novelty—the deviation of future climate from present-day conditions within a species’ range—using multivariate data. We tackled this using the sigma dissimilarity (sigma) methodology, which is based on a modified multivariate standardized Euclidean distance (SED) metric. This innovative approach involved:

  • Converting SED to Mahalanobis distances to rescale variables relative to local intra-annual climate variation and reduce correlation-based variance.
  • Interpreting these distances using the chi distribution to account for the dimensionality effect of using multiple variables.


This methodology allowed us to categorize exposure based on sigma dissimilarity (sigma): similar climate (sigma<2), novel climate (sigma<4), and extremely novel climate (sigma>4). Species and MPAs with sigma>2 were considered at risk.

Main finding: the high-emission risk

Our results reveal a stark contrast between the two future scenarios:

  • Low-Emission Scenario (SSP1-1.9): Only approximately 6.5% of species and 0.5% of MPAs are projected to be at risk.
  • High-Emission Scenario (SSP5-8.5): The risk skyrockets, projecting that 80% of species and 87% of MPAs will be at risk due to novel or extremely novel conditions. Specifically, 67.6% of species are anticipated to be exposed to extremely novel conditions in at least one MPA.


Crucially, enclosed and semi-enclosed seas are identified as major hotspots of vulnerability:

  • Baltic Sea: Up to 100% of species are projected to be exposed to novel or even extremely novel conditions. More than half of its MPAs contain more than 50% of species at risk, with 380 MPAs showing 100% at-risk species.
  • Black Sea: Virtually all species in this region will be at risk, and all 14 IUCN-listed species are anticipated to be at risk.


In contrast, the Norwegian Sea, North-East Atlantic, and western parts of the Mediterranean and North Seas are expected to be less impacted, even under the high emission scenario.

Main implications for conservation and management

Our findings underscore the imperative value of adhering to the Paris Agreement to reduce global greenhouse gas emissions, which is crucial for the enduring effectiveness of European MPAs. For marine governance and conservation planning, our results have several critical implications:

  • Climate-Smart MPAs: We must move beyond business-as-usual and reactive management to proactive, anticipatory governance. This involves identifying and preserving climatically stable areas, such as those in the North-East Atlantic and Norwegian Sea, which can serve as refugia or essential migration corridors.
  • Prioritizing protection: Conservation strategies must prioritize vulnerable species for targeted protection, particularly in hotspot regions. Across all regions, over half of the species projected to be at risk belong to Chondrichthyes (sharks, rays, and skates).
  • Enhancing MPA effectiveness: Even in climatically stable regions, species are under pressure from human activities. The results highlight the need for comprehensive approaches that address both climate stability and other pressures, ensuring MPAs have strong protection and are not minimally regulated.
  • Evolving conservation frameworks: Our data provides a basis for expanding the roster of species safeguarded under the Habitats Directive and Regional Sea Conventions to encompass all threatened and commercially significant species susceptible to future climate alterations. Furthermore, MPA planners might need to contemplate safeguarding species that perform similar ecological functions but demonstrate greater resilience to impending climate shifts.