2. Assessment of climate change effects on Nordic marine ecosystems

Marine ecosystems in the Baltic Sea have adapted to the brackish environment for thousands of years and are sensitive to hydrographic and biogeochemical changes. These long-term climate changes come on top of other pressures from pollution, shipping, fishing etc. In a recent study (Wåhlström et al. 2022) it was shown that the climate change impact (from temperature increase, salinity decrease, and ice cover decrease) on marine ecosystems in Swedish waters was of similar magnitude as the combined impacts of all present human pressures on the marine environment. Climate change will result in novel environmental conditions that will change ecosystem composition (Blenckner et al. 2021), modify biodiversity, and change the risks of invasive species entering the region. Cod is an example of a species which depends on specific salinity and oxygen conditions for reproduction and is also an example of a species where nutrient remediation measures can help to increase climate resilience (Wåhlström et al. 2020).

The scenarios we have modelled and provided data on will have several near-future appliances. They will be used to perform habitat modelling and species distribution modelling (Fredriksson et al. 2024) for present and “door knocking” marine invasive species. In addition to habitat modelling, currents will be used to study existing and changing connectivity and advective spread of invasive species. The habitat modelling and dispersal understanding will in turn facilitate the possibility to initiate control activities at the introduction of an invasive species or very early in the invasion curve and thereby decrease the rate of growth and expansion of the invasive species. The dispersal patterns and connectivity of invasive species are, however, influenced by factors beyond national borders (e.g., water currents, rising temperatures, salinity). It is therefore crucial to adopt a regional approach to develop effective mitigation strategies. Here this common understanding through scenarios and a broader regional perspective as well as bi-lateral initiatives are essential to ensure timely, precise, and cost-effective actions. Besides initiating bi-lateral and regional control, the understanding will also locally guide municipalities responsible e.g. for shielding Marine Protected Areas.

Climate change is expected to increase the water temperature in the Limfjord by up to 5 °C in the worst-case scenario (SSP5-8.5). In addition, salinity is reduced due to increased runoff and thus the stratification of the water column becomes stronger in all scenarios. In the two most optimistic future scenarios (SSP1-2.6, SSP2-4.5), the expected improvements in bottom oxygen and Chl a concentration during implementation of Water Management Plans were counteracted by changes in physics (warming, stronger stratification). Therefore, it will be necessary to increase nutrient reductions in the Water Management Plans to achieve good ecological status under the influence of climate change. In the worst-case future scenario (SSP5-8.5), the oxygen depletion in the Limfjord will increase with negative consequences especially for benthic mussels. Mussel farming in the water column was intensified in all scenarios and showed a high potential for harvesting, while benthic mussels suffered from reduced food supply and oxygen depletion (Maar et al. 2024). 

Due to climate change and increasing demands for a more sustainable production of mussels, it is likely that mussel farming will become the primary source of mussels in the industry in the future rather than dredging of mussels in the wild. The applied modelling scenarios can be used to inform managers, mussel farmers, fishermen and the local population about future mussel harvesting options, potential changes in ecosystem health and to find solutions to adapt to climate change impacts.

For both the Barents and Norwegian seas pronounced changes are expected to arise from climate change. The degree, and in some cases also the direction of change, is however, highly different between scenarios. The most significant ecological impacts from climate change are, also in these regions, through higher sea temperatures. Temperature changes will decrease both ice cover and thickness, and also mixed layer depth, in most cases decreasing it. This and other effects of temperature change will likely alter nutrient mixing and availability. In the Barents Sea, our results show that change in mixed layer depth is a main driver for changes in net primary production (NPP). For the Norwegian Sea, an initial strong decrease in mixed layer depth until 2030 is particularly notable in our results. Caution should be shown to the latter though, as we cannot rule out that this is an erroneous effect of model initialization artifacts.

The projected alterations in primary and gross secondary production (GSP) will affect fish stocks, some negatively and some positively, and thus alter ecosystem dynamics in many ways. Projected increase in GSP in the Barents Sea might positively impact fish recruitment and cause fish species like Atlantic cod and haddock to expand their distribution northwards and eastwards. In the Norwegian Sea our results show significant decrease in both NPP and GSP, especially in coastal areas. This may potentially negatively impact recruitment to important fish stocks there, like Norwegian spring-spawning herring.

We have not projected future development of marine heat waves (MHWs), but with general temperature increase, MHWs are expected to become more frequent, longer lasting and, more intense (although the actual values depend on the base period on which MHWs are calculated). Relatively little work has been done on MHWs in the Nordic sea area, but recent and ongoing research point to increasing number and severity of both surface and near-bottom events over large parts of the region. 

The projections conducted with the NoBA Atlantis model are the first with an end-to-end ecosystem model using downscaled regional physical forcing. They project significantly different results from global models, showing that some local species respond positively to increased temperatures, although with high uncertainties connected to functional responses and temperature tolerances. It should be noted that different approaches have yielded quite different results. The negative response in cod and capelin, the lack of response in the pelagic species in the Norwegian sea, and the unclear response of the polar cod found in the NoBa Atlantis projections contrast with the findings in other expert based studies in the region (Kjesbu et al. 2021;  Sandø et al. 2024). This highlights the complexity and uncertainty in long-term projections for higher trophic levels, like fish stocks. In Nilsen et al. (2024) important processes such as density-dependent movement-recruitment relationships (e.g. for the pelagic fish stocks in the Norwegian/Barents sea, and the polar cod), the lack of knowledge for some of the species, the lack of adaptation, and the importance of indirect and direct trophic interactions are discussed. The results of the study point towards a need for both modelling and empirical studies to go more in depth in some of these challenges, and to improve our possibility to provide, e.g., fisheries managers with projections that can be used for adaptations of fisheries to climate change. Models like Atlantis can also be used to detect and explain potential ‘surprises’ driven by indirect and direct predator-prey interactions, which may not necessarily be clear from expert-based studies founded on single-species information.