Climate change has multiple effects on Baltic Sea species, communities and ecosystem functioning through changes in physical and biogeochemical environmental characteristics of the sea. Associated indirect and secondary effects on species interactions, trophic dynamics and ecosystem function are expected to be significant. We review studies investigating species-, population- and ecosystem-level effects of abiotic factors that may change due to global climate change, such as temperature, salinity, oxygen, pH, nutrient levels, and the more indirect biogeochemical and food web processes, primarily based on peer-reviewed literature published since 2010.
For phytoplankton, clear symptoms of climate change, such as prolongation of the growing season, are evident and can be explained by the warming, but
otherwise climate effects vary from species to species and area to area.
Several modelling studies project a decrease of phytoplankton bloom in
spring and an increase in cyanobacteria blooms in summer. The associated
increase in
In deep benthic communities, continued eutrophication promotes high sedimentation and maintains good food conditions for zoobenthos. If nutrient abatement proceeds, improving oxygen conditions will first increase zoobenthos biomass, but the subsequent decrease of sedimenting matter will disrupt the pelagic–benthic coupling and lead to a decreased zoobenthos biomass. In the shallower photic systems, heatwaves may produce eutrophication-like effects, e.g. overgrowth of bladderwrack by epiphytes, due to a trophic cascade. If salinity also declines, marine species such as bladderwrack, eelgrass and blue mussel may decline. Freshwater vascular plants will be favoured but they cannot replace macroalgae on rocky substrates. Consequently invertebrates and fish benefiting from macroalgal belts may also suffer. Climate-induced changes in the environment also favour establishment of non-indigenous species, potentially affecting food web dynamics in the Baltic Sea.
As for fish, salinity decline and continuing of hypoxia is projected to keep cod stocks low, whereas the increasing temperature has been projected to favour sprat and certain coastal fish. Regime shifts and cascading effects have been observed in both pelagic and benthic systems as a result of several climatic and environmental effects acting synergistically.
Knowledge gaps include uncertainties in projecting the future salinity level, as well as stratification and potential rate of internal loading, under different climate forcings. This weakens our ability to project how pelagic productivity, fish populations and macroalgal communities may change in the future. The 3D ecosystem models, food web models and 2D species distribution models would benefit from integration, but progress is slowed down by scale problems and inability of models to consider the complex interactions between species. Experimental work should be better integrated into empirical and modelling studies of food web dynamics to get a more comprehensive view of the responses of the pelagic and benthic systems to climate change, from bacteria to fish. In addition, to better understand the effects of climate change on the biodiversity of the Baltic Sea, more emphasis should be placed on studies of shallow photic environments.
The fate of the Baltic Sea ecosystem will depend on various intertwined environmental factors and on development of the society. Climate change will probably delay the effects of nutrient abatement and tend to keep the ecosystem in its “novel” state. However, several modelling studies conclude that nutrient reductions will be a stronger driver for ecosystem functioning of the Baltic Sea than climate change. Such studies highlight the importance of studying the Baltic Sea as an interlinked socio-ecological system.
Global climate change affects the marine ecosystem through ocean warming, acidification and deoxygenation and through changes in nutrient loading and water circulation, which may all impact marine biological processes from genes to populations, communities and ecosystems (Brierley and Kingsford, 2009; Henson et al., 2017). The biological consequences range from shifts in species abundance and distributions, changes in dispersal patterns and modification of species interactions to altered food webs and decreasing ocean productivity (Hoegh-Guldberg and Bruno, 2010; Philippart et al., 2011; Doney et al., 2012; Burrows et al., 2019). The changes in biological processes also affect marine ecosystem services and threaten human food security, especially in the most vulnerable areas (Barange et al., 2014).
Climate change also has multiple effects on the Baltic Sea, impacting species, communities and ecosystem functioning. As in the ocean, the effects are usually mediated via climate-affected oceanographic or biogeochemical processes and via associated indirect effects on species interactions, trophic dynamics and ecosystem function mechanisms. These potentially affect the biota inhabiting the Baltic Sea, as well as the human society (Paasche et al., 2015; Hyytiäinen et al., 2019; Pihlainen et al., 2020; Stenseth et al., 2020).
The effects of climate change on the Baltic Sea ecosystem may differ from those projected for the oceanic areas, as the Baltic Sea differs in many respects from the oceans and even from the coastal ecosystems surrounding the other regional seas and oceans. The communities of the Baltic Sea are formed of a peculiar combination of marine, limnetic and brackish-water taxa. The long winter and the strong seasonal cycle give the area subarctic properties, especially in the northern areas. The Baltic Sea has also been shown to warm up faster than most other sea areas of the world (Belkin, 2009; Sherman et al., 2009), albeit with large differences between sub-basins (Kniebusch et al., 2019; Dutheil et al., 2021).The Baltic Sea is also strongly affected by its watershed, which is more than 4 times larger than its surface area and is inhabited by ca. 85 million people (Omran and Negm, 2020). The marine ecosystem therefore receives excess nutrients and other elements and contaminants from the land via rivers, through the air, and by leaking from the sediments of the Baltic Sea. Furthermore, the irregular inflows of more saline and oxic North Sea water, which at specific basin-wide weather conditions enter the Baltic sea through the Danish Straits (Matthäus and Schinke, 1994; Lehmann et al., 2022) and influence the state and functioning of the Baltic Sea.
All of these pathways of chemical elements and oceanographic and biogeochemical processes may be affected by global climate change and the quasi-cyclic climate phenomena such as the North Atlantic Oscillation (NAO). It has also been suggested that impacts and symptoms of global climate change are accumulating faster in the Baltic Sea than in other coastal areas of the oceans and that Baltic Sea thus can be considered as “a time machine for the future coastal ocean” (Reusch et al., 2018).
However, attribution of the observed ecosystem changes to global (anthropogenic) climate change is challenging because of the multiple synergistic effects between climate and other environmental drivers, such as eutrophication, harmful substances, habitat modification, fishing and introduction of non-indigenous species, which all may have strong impacts on ecosystems and their functioning in time and space (Reusch et al., 2018; Stenseth et al., 2020; Bonsdorff, 2021). Therefore, profound knowledge of the mechanisms and processes governing the Baltic Sea ecosystem under climate change are vital for the understanding and management of the Baltic Sea (Reusch et al., 2018; Bonsdorff, 2021; Blenckner et al., 2021).
The overall effects of climate change on the Baltic Sea have been reviewed in earlier synthesis studies (The BACC Author Team, 2008; The BACC II Author Team, 2015), in which climate impacts on the marine ecosystem were also assessed (Dippner et al., 2008; Viitasalo et al., 2015). Since then, a wealth of field, experimental and modelling studies have shed more light onto the complex interactions between the climate change and the Baltic Sea system (Meier et al., 2022b).
In this paper, we review research on climate change effects on the Baltic Sea species, habitats, and ecosystem functioning, primarily based on research published in 2010–2021. We include both studies investigating direct effects of climate-related parameters on organisms, as well as studies that investigate the more indirect processes affecting the structure and functioning of the Baltic Sea ecosystem through biogeochemistry and food web interactions. Evidence is compiled from empirical field studies that show past changes and responses of species, populations, and communities to climate-affected parameters such as temperature, salinity, oxygen and pH. A large number of experimental studies, investigating species responses to the same parameters in microcosms or mesocosms, are reviewed. Studies investigating the complex effects of climate change on the interactions between species and trophic groups, i.e. phytoplankton, bacteria, cyanobacteria, zooplankton and fish, as well as algae or vascular plants and invertebrates grazing on them, are also analysed. Modelling studies, based on coupled oceanographic–biogeochemical models or other types of species-level or food web models, are reviewed. Based on the published research we draw conclusions about the role of climate-driven environmental variables on shaping the structure and functioning of the Baltic Sea ecosystem and identify knowledge gaps and current issues of dissensus. Areas in need of more research are recommended.
We review studies that shed light to the possible climate effects on the Baltic Sea ecosystem by studying oceanographic and biogeochemical parameters which have been projected to change due to climate change. As such changes may be affected by both anthropogenic global climate change and natural climate variations, it is first necessary to define certain key terms used in this review.
By “global climate change” we refer to the past and contemporary increase in global temperature
caused by anthropogenic emissions of
For “ecosystem functioning” we use Tilman's (2001) definition, “the rate, level, or temporal dynamics of one or more ecosystem processes such as primary production, total plant biomass, or nutrient gain, loss, or concentration”. By “functional diversity” we mean “the range and value of those species and organismal traits that influence ecosystem functioning” (Tilman, 2001). A “functional group”, is “a set of species that have similar traits and that thus are likely to be similar in their effects on ecosystem functioning” (Tilman, 2001).
With “biogeochemical processes”, we refer to various biogeochemical cycles and processes, which often involve cycling and transfer of allochthonous or autochthonous essential nutrients and/or minerals and organic carbon and which are either driven or influence biological activity in species. With “trophic dynamics” we refer to interactions between trophic levels or functional groups, such as phytoplankton, bacteria, cyanobacteria, nanoflagellates and microflagellates, microzooplankton and mesozooplankton, zoobenthos, and fish, as well as the algae, vascular plants and invertebrates living amongst them.
“Trophic efficiency” is defined as “the efficiency of energy flow between trophic levels and is the percentage of energy from a trophic level that is used by the organisms of the next trophic level for growth and reproduction” (Hine, 2019).
The search for relevant papers was implemented mainly using Web of Science
(WoS) website tool (
Some papers from 2021 and 2022 were found and downloaded with an unstructured search performed with Google Scholar, as this website tool includes more recent publications than WoS. In some cases, references before 2010 were also included if it was necessary to back up the statements with older studies.
The search resulted in over 500 papers, of which many were not relevant to the current review, i.e. were not concerning effects of climate change on species, habitats or ecosystem functioning in the Baltic Sea. The most relevant studies were saved into library groups of EndNote X9.2 reference management software (Clarivate Analytics), and the contents were scrutinised in more detail.
Because of the focus period, 2010–2021, the review is not a full systematic review of all research done on climate change effects on the Baltic Sea ecosystem this far. Also, certain taxonomic groups and study types were less thoroughly reviewed than others. Fish studies in particular were not comprehensively scrutinised because the complex responses of fish populations to climate, eutrophication and fisheries have recently been addressed by a large number of studies and would merit their own review. Also, we have not reviewed all experimental studies that have dealt with environmental variables that may change with climate change. Our goal is to highlight the variety of field, experimental and modelling studies and to summarise what can be concluded from the recent evidence on the possible effects of climate change on the Baltic Sea.
Climate change may have direct effects on the physiology and phenology on phytoplankton through physical and chemical parameters and indirectly through hydrodynamics, e.g. stratification and availability of light and nutrients. Top-down forces, i.e. grazing on phytoplankton, may also be modified in various ways if grazer populations change.
The growing season of phytoplankton has been significantly prolonged with
warming temperatures during the recent decades. A satellite-based study
suggested that the length of the period with chlorophyll concentration of at
least 3
The spring species communities have also shifted from dominance of early
blooming diatoms to later blooming dinoflagellates and the mixotrophic
ciliate
Some studies have attributed the springtime shifts in phytoplankton phenology and community structure to changes in environmental conditions driven by global climate change. A 15-year study (2000–2014) using FerryBox observations, covering the area between Helsinki (Gulf of Finland) and Travemünde (Mecklenburg Bight), confirmed that spring bloom intensity was mainly determined by winter nutrient concentration, while bloom timing and duration co-varied with meteorological conditions. The authors conclude that the bloom magnitude has been affected by the reduction of nutrient loading from land, while bloom phenology can also be modified by global climate change affecting seasonal oceanographic and biogeochemical processes (Groetsch et al., 2016).
It has also been suggested that in the future climate higher temperatures and less ice will cause an earlier bloom of both diatoms and dinoflagellates, with increased dinoflagellate dominance (Hjerne et al., 2019). Experimental (mesocosm) evidence supports findings that warming up of water and changes in light conditions will accelerate the spring bloom, induce a decline in peak biomass and favour small size cells, either directly or via increased grazing by copepods (Sommer et al., 2012). On the other hand, this development may be counteracted by increases of windiness and cloudiness, which have also been projected by certain modelling studies (Hjerne et al., 2019). Recent studies have however indicated that the projections for spring and summer wind and radiation are uncertain (Christensen et al., 2022), and future weather changes and associated spring bloom dynamics therefore remain obscure.
Climate change effects, i.e. temperature increase, salinity decline and
acidification have been shown to have variable results on the toxic
dinoflagellate
Climate change also increases concentration of water carbon dioxide, a
compound necessary for primary production, and ocean acidification (OA)
could therefore enhance productivity of phytoplankton. However, the results
of experimental studies investigating effects of
There are also studies that have indicated a connection between phytoplankton and the North Atlantic Oscillation (NAO). A decline in the intensity of NAO in the 1990s was suggested to have been caused by less cloudy conditions, giving more irradiance and less windy conditions, inducing stronger stratification of surface water (Hjerne et al., 2019). If the shifts are driven by variations in NAO or the Baltic Sea Index (BSI, a regional index similar to the NAO), they may be temporary and reversible, whereas shifts caused by global climate change may be more enduring.
In the northern Baltic Proper, Åland Sea and the Gulf of Finland, the
biomasses of Chrysophyceae, Prymnesiophyceae and Cyanophyceae have increased
and the phytoplankton biomass maximum, which in the 1980s was in spring and
mainly consisted of diatoms, is now in July–August and is dominated by
filamentous cyanobacteria (Suikkanen et al., 2013). This shift was
explained by a complex interaction between eutrophication, climate-induced
warming, and increased top-down pressure, as well as changes in
It is obvious that climatic influences are intertwined with other processes and parameters affecting phytoplankton, especially anthropogenic nutrient loading from land and internal loading of nutrients from the sediments. There is however a discrepancy in the relative effects of eutrophication, climate change, and other environmental and anthropogenic factors in explaining past variations in phytoplankton communities and biomass. Also, several studies have identified complex variations in phytoplankton communities that cannot be easily explained by any of the studied factors or environmental parameters.
A study comparing historic phytoplankton communities from 1903–1911 with the present ones (1993–2005) in the northern Baltic Proper and the Gulf of Finland observed an undefined “period effect”, characterised by a decline in diatoms and increase in dinoflagellates, that was not explained well by the available environmental variables (temperature, salinity, and general regional climatological data). Although data on biogeochemical parameters was not available for the period 1903–1911, the authors interpreted the observed community change as evidence of the direct and/or indirect influence of eutrophication (Hällfors et al., 2013).
A study investigating summer phytoplankton time series (HELCOM monitoring 1979–2012) across the Baltic Sea found that there were no common interannual patterns. Instead, the class trends, e.g. that of cryptophytes, may be affected by anomalies in the BSI, although a mechanistic explanation for the relationship could not be found (Griffiths et al., 2020). Other studies did not find any explanation for the observed changes in the biovolumes of different taxa, e.g. decrease in diatoms and increase in certain dinoflagellate taxa, and concluded that phytoplankton community in the Baltic Sea is not in a steady state (Olli et al., 2011), or noted that stochastic dynamics at local scales confound any commonalities between phytoplankton groups (Griffiths et al., 2020).
To sum up, the past changes in phytoplankton community composition have been very variable and usually cannot be explained by a single factor. Some clear signs of climate change, such as prolongation of the growing season are evident, and can be explained by the warming and associated biogeochemical processes, but the changes in species and communities vary from area to area and have multiple reasons, including climate change, changes in nutrient dynamics and changes in trophic interactions.
Filamentous diazotrophic cyanobacteria benefit from warm temperatures and stratified water, and they tend to bloom during the hottest and calmest periods of summer (Munkes et al., 2021). Several modelling studies suggest that the climate-induced increase in stratification (Liblik and Lips, 2019), together with potentially increasing hypoxia and consequent release of phosphorus from the anoxic sediments, will increase cyanobacteria blooms in the Baltic Sea (Meier et al., 2011a; Neumann et al., 2012; Chust et al., 2014; Lessin et al., 2014; Andersson et al., 2015; Ryabchenko et al., 2016).
Many field studies have also stated that cyanobacteria have already
increased along with the warming of the Baltic Sea. In the northern Baltic
Proper, Åland Sea and the Gulf of Finland, the biomasses of Cyanophyceae
have increased, which has been explained by an interaction between warming,
eutrophication and increased top-down pressure on species of the spring
bloom, as well as changes in
Also, in the Gulf of Bothnia, eutrophication and cyanobacteria have
increased in summer (Fleming-Lehtinen et al., 2015; Kuosa et al., 2017),
and extensive cyanobacteria blooms have in the past few years been detected
with satellite methods in the Bothnian Sea, an area usually devoid of such
phenomena (unpublished monitoring and satellite records collected by the
Finnish Environment Institute). The increase of cyanobacteria in the
Bothnian Sea has been attributed to an increased freshwater flow and, since
2000, to an increased intrusion of more saline and phosphorus rich Baltic
Proper water into the Bothnian Sea. These changes have increased
stratification, lowered oxygen conditions, and led to a decline in
It has also been suggested that the various drivers of climate change may
contribute to increase blooms and toxicity of cyanobacteria in the Baltic
Sea. For instance, the intracellular toxin concentration of the
cyanobacterium
A few long-term studies have not found an increase in cyanobacteria during
the past. Two recent studies compiling monitoring data from the Baltic Sea
for 1979–2012 (Griffiths et al., 2020) and 1979–2017 (Olofsson et
al., 2020) did not find any evidence for an overall increase of
diazotrophic filamentous cyanobacteria during this period. Biovolume of the
hepatotoxic
Hypothetically, ocean acidification could benefit cyanobacteria through
increased availability of carbon dioxide in water. The available studies do
not give a definitive answer however. When
If the biomasses of
To sum up, there are species-specific responses to climate change and associated oceanographic parameters within cyanobacteria. Several field and modelling studies suggest that the climate-induced increase in temperature and stratification, together with increasing hypoxia and release of phosphorus from the sediments, has increased cyanobacteria biomass and will also continue to favour cyanobacteria blooms in the future. However, the results of certain empirical and experimental studies give a more multifaceted picture of cyanobacteria response to climate change. The past increase in cyanobacteria is not as obvious as might be expected, responses vary from species to species, and processes affecting the amount of cyanobacteria in the Baltic Sea can be modified, counteracted or amplified by various environmental processes and food web interactions.
The Baltic Sea mesozooplankton species originate either from marine or
freshwater environments, and some are typically found in brackish water. It is
therefore plausible that they respond to long-term variations in
oceanographic parameters. Several field studies have confirmed that marine
copepod species (e.g.,
Environmental impacts on the physiology of the more sensitive species may
also affect the reproductive success of zooplankton (Möller et al.,
2015). The increase of euryhaline taxa has been directly or indirectly
attributed to the temperature increase (Mäkinen et al., 2017).
It has also been suggested that species that reside in the upper water
layers, such as the copepod
The effects of climate-driven variations in temperature and ocean
acidification (OA) on zooplankton have been studied experimentally. In
Changes in zooplankton functional groups, such as a shift from raptorial
and suspension-feeding copepods and cladocerans to a dominance of small
filter-feeding rotifers and cladocerans, have also been shown as results of
warming (Suikkanen et al., 2013; Jansson et al., 2020). OA also promoted
the growth of suspension-feeding cladocerans because of a
Furthermore, a switch from predominantly herbivorous feeding by copepods to predation on ciliates has been observed in a field study in the southern and central Baltic Sea during cyanobacterial blooms (Loick-Wilde et al., 2019). This was caused by decomposing of the otherwise unpalatable filamentous cyanobacteria and an associated increase of the bacteria, nanoflagellates and ciliates (Hogfors et al., 2014). Warming may also increase zooplankton grazing on medium to large algae, which could contribute to a change towards smaller-sized phytoplankton species (Klauschies et al., 2012; Paul et al., 2015). It is therefore possible that the dominant traits of zooplankton communities will change if climate-induced warming and reduced salinity trends continue. It has also been suggested, from experimental (mesocosm) evidence, that warming speeds up the growth of copepods but leaves phytoplankton unaffected, which shortens the time lag between phytoplankton and zooplankton. This may lead to a larger and earlier zooplankton peak and increase the possibility of zooplankton controlling phytoplankton, which may lead to a reduced phytoplankton biomass under warm temperature (Paul et al., 2016).
Sufficient supply of essential compounds such as amino acids (AA) produced
by phytoplankton and cyanobacteria is essential for the growth and
productivity of zooplankton grazers. A field study performed in the Baltic
Proper shows that, during a warm summer, thermophilic rotifers and
cladocerans (e.g.
Little is known about the adaptation capabilities of zooplankton against
physicochemical stress, but some degree of temperature adaptation has been
demonstrated experimentally for the copepod
To sum up, a shift towards smaller-sized zooplankton and a stronger linkage between mesozooplankton and the microbial food web is probable in a warmer Baltic Sea. A decline in certain marine species has also been projected, but this will depend on the future velocity of salinity decline, patterns of stratification, realised time lag between phytoplankton and zooplankton peaks, predation pressure by fish, and the possible adaptation of zooplankton species to the subtle changes in salinity.
Bacteria are key components of the ecosystem, as they decompose organic material and serve as food for heterotrophic nanoflagellates and the associated microbial food web. They affect the nutrient and carbon dynamics of the marine ecosystem, and it is therefore possible that climate impacts on bacteria may radiate to the structure and functioning of the entire Baltic Sea ecosystem.
The effects of climate-induced changes in environmental factors to pelagic
bacteria and the other components of the microbial food web have been
studied experimentally. The effects of projected ocean acidification (OA) on
bacteria have been studied alone and also in combination with other abiotic
variables, such as temperature (OAW) and salinity (OAS). OA alone had a
limited impact on spring bloom microbial communities (sampled from the sea
area around the island Öland in the Baltic Proper and kept in 100 L
mesocosms for 21 d), but when combined with increased temperature,
certain bacterial phylotypes, such as betaproteobacteria, increased. It was
suggested that synergistic effects of increased temperature and
acidification selectively promote growth of specific bacterial populations
(Lindh et al., 2013). In the southern Baltic Sea (Kiel Bight) the impact
of OA was studied in 1400 L indoor mesocosms for 21–24 d.
Acidification only affected a few operational taxonomic units (OTUs), such as
In an OAS experiment (4 L aquaria, 12 d) using a natural summer
microplanktonic community, the biovolume of heterotrophic bacteria declined
when
Experimental studies have demonstrated that complex food web responses to
climate change may also arise. In the Quark, located in the Gulf of Bothnia, an increase in dissolved organic matter (DOM) enhanced respiration and abundance of bacteria, whereas an increase in temperature (from 12 to 15
As for microzooplankton (MZP), the effects of OA and warming seem to be
mostly beneficial. OA does not have a negative effect on MZP, probably
because estuarine MZP are adapted to a large natural variability in
To sum up, different components of the microbial food web show very variable responses to climate-induced changes in temperature, salinity and pH. Bacteria growth is generally favoured by increasing temperature, but mixed effects are common, and indirect processes affecting decay and availability of organic matter and abundances of species predating on bacteria are also important. This highlights the importance of considering the effects of abiotic factors and the delicate indirect food web effects on the dynamics of the microbial food web and the pelagic ecosystem in general.
Long-term changes in Baltic Sea macroalgae and charophytes have mostly been explained by combined or synergistic effects of changes in salinity, wind exposure, nutrient availability and water transparency (Gubelit, 2015; Blindow et al., 2016; Eveleens Maarse et al., 2020; Rinne and Salovius-Laurén, 2020) and by biotic interactions (Korpinen et al., 2007).
For the brown alga bladderwrack
The direct effects of climate-induced changes in temperature, salinity and
ocean acidification (OA) on bladderwrack
The timing of temperature stress is however important for the damage
experienced by algae. Experiments done with bladderwrack
Ocean acidification combined with warming (OAW) may also act in concert with
hypoxia in areas where upwellings bring hypoxic water close to the surface.
In a 3 d experiment simulating an upwelling event, hypoxic water
caused severe mortality of
Climate-induced decline in salinity may affect communities via its direct effect on the physiology of individual populations and species. A retreat towards the south–south-west has been predicted for marine species such as bladderwrack and eelgrass and for species affiliated to them (Vuorinen et al., 2015). Species distribution modelling studies have suggested that this mainly salinity-induced decrease of bladderwrack will cause habitat fragmentation with large effects on the biodiversity and ecosystem functioning of the shallow water communities of the northern Baltic Sea (Takolander et al., 2017a; Jonsson et al., 2018; Kotta et al., 2019).
It is not certain to what degree
It has also been shown that
Similar experiments on climate change effects to those done with bladderwrack have
also been made with other macroalgae and certain vascular plants. In field
mesocosm experiments, OA increased the growth of the opportunistic green
alga
Salinity decline is projected to decrease the distributional ranges of the
marine eelgrass
Certain species may be favoured by the projected climate change. Lowering of
salinity generally favours vascular plants originating from freshwater, and
temperature increase favours thermophilic species, such as charophytes
(Torn et al., 2020). In mesocosm studies made in
Kõiguste Bay, photosynthesis of charophytes (
To sum up, recent studies suggest that changes in species composition of macroalgae and vascular plants are likely due to temperature, pH and salinity changes. Climate change, in conjunction with other environmental changes (especially eutrophication) may also influence carbon storage in both macroalgae and vascular plants in the Baltic Sea (Röhr et al., 2016; Takolander et al., 2017a; Jonsson et al., 2018; Salo et al., 2020; Bobsien et al., 2021). It has been projected that macroalgae will decline in hard bottoms and vascular plants increase in the more sheltered soft-bottom areas (Torn et al., 2020). Because algae and plants mostly occupy different habitats, the possible increase of vascular plants or charophytes cannot counteract the negative effects of the disappearance of macroalgae from hard-bottom areas. Consequently the invertebrates, fish and birds benefiting from habitats formed by macroalgae also will suffer from the climate change.
As with other species groups, projecting the fate of macroalgae and vascular plants is challenging. This is caused by the uncertainties in projections concerning salinity and stratification (Lehmann et al., 2022), discrepancy on which physicochemical factors determine the distribution of invertebrates, unknown adaptation capabilities of algae and plants, and uncertainties concerning future trophic interactions within macroalgae and vascular plant communities.
Soft-bottom benthic communities are dependent on several hydrographic and
biogeochemical variables, and parameters that change with climatic
variations have been shown to drive the long-term progression of zoobenthic
communities (Weigel et al., 2015; Rousi et al., 2019; Ehrnsten, 2020). In
the SW coast of Finland, a drastic community change took place, with
amphipods being replaced by Baltic clam
Many marine invertebrates will directly and indirectly suffer from
decreasing salinity. In experiments simulating projected changes in
temperature and salinity, the survival of the isopod
Ocean acidification has various effects on benthic invertebrates. The size
and time to settlement of pelagic larvae of the Baltic clam
Several modelling studies have suggested that climate-induced changes in temperature, salinity and eutrophication, affecting oxygen levels and food availability for benthos, drive the development of benthic communities and their biomass in the future (Ehrnsten et al., 2019a, b). A physiological fauna model linked to a 3D coupled hydrodynamic–ecological model projected that, in areas previously burdened by hypoxia, benthic biomass will increase (until year 2100) by up to 200 % after re-oxygenating bottom waters, whereas in permanently oxygenated areas the macrofauna biomass will decrease by 35 % due to lowered food supply to the benthic ecosystem (Timmermann et al., 2012). In another modelling study, zoobenthic production decreased in the coastal zones and gradually also decreased in the more offshore areas, with increasing temperature and declining salinity and bottom oxygen regardless of the nutrient load scenarios (Weigel et al., 2015). The fate of zoobenthos also depends on human intervention, i.e. success of nutrient-reduction schemes. For instance, it has been projected that if the HELCOM BSAP is implemented, the biomass of benthic animals, and hence food for benthic-eating fish, will first increase and then decrease (Ehrnsten et al., 2020).
There are very few modelling studies focusing on invertebrates inhabiting
shallower hard-bottom habitats. One study, where experimental work and
species distribution modelling were combined, projected a decline of the
isopod
One modelling study also investigated how
It is notable that hypoxia, which is a key factor affecting zoobenthos, is by no means limited to the deep basins of the Baltic Sea (Conley et al., 2011). The archipelagos of the northern Baltic Sea are especially prone to hypoxia due to their complex topography and limited water exchange (Virtanen et al., 2019). Increasing sea surface temperature will strengthen stratification and enhance mineralisation of organic matter by microbes, which may increase the release of phosphorus from sediments (Puttonen et al., 2016) and lead to a “vicious circle of eutrophication” (Vahtera et al., 2007). The sheltered archipelago areas and enclosed bays may therefore become “climate change hotspots” (Queiros et al., 2021), where zoobenthic communities are most drastically changed as well.
To sum up, zoobenthic communities are affected by all environmental parameters that are projected to change with climate change, i.e., temperature, salinity, pH and oxygen, as well as benthic–pelagic coupling. However, the effects are not unidirectional, and several processes may amplify or counteract the possible changes. The magnitude of the future salinity decline is unclear, and other factors, such as decreasing ice cover and changes in future wind conditions (of which no consensus exists), may also affect nutrient and oxygen dynamics of the Baltic Sea. Also, there may be feedback effects on sediment oxygen levels, as different benthic species have different bioirrigation activities (Norkko et al., 2012; Guy-Haim et al., 2018). Such processes, which are dependent on traits of a few species, may be of particular importance in low-diversity systems such as the northern Baltic Sea (Gladstone-Gallagher et al., 2021).
It is often suggested that global climate change favours invasions of non-indigenous species (NIS) worldwide (Jones and Cheung, 2015). This is plausible because an increase in temperature will open new niches and induce a poleward shift of the ranges of species inhabiting tropical and temperate sea areas. In the Baltic Sea, it has been shown that non-native species typically occur in areas characterised by high temperatures, reduced salinity, high proportion of soft seabed, and decreased wave exposure, whereas most native species display an opposite pattern (Jänes et al., 2017). This suggests that the former areas are more prone to climate-induced range expansion of non-native species than the latter. This is consistent with the hypothesis of climate change hotspots, which suggests that some coastal areas may be more susceptible to effects of climate change than others (Queiros et al., 2021).
Modelled scenarios of temperature and salinity have been used to project how changes in the abiotic environment could affect NIS already present in the Baltic Sea. One modelling study suggests an increase in Ponto-Caspian cladocerans in the pelagic community and an increase in dreissenid bivalves, amphipods and mysids in the coastal benthic areas of the northern Baltic Sea until 2100 (Holopainen et al., 2016).
To sum up, the global climate change induces many environmental changes that may favour establishment of NIS in the Baltic Sea. However, attribution of the observed establishments to the climate change is difficult. It has even been claimed that there is no conclusive evidence that NIS will gain significant advantage from environmental alterations caused by climate change (Henseler et al., 2021). Stochastic processes related to maritime transport and other types of human activities are obviously important for the chances of NIS to be introduced and established into a given sea area. Long-term surveys and comparisons with areas where NIS have not been established are needed to distinguish climate-related effects from other ecosystem-level drivers (Bailey et al., 2020).
Fish populations in the Baltic Sea are influenced by various environmental and anthropogenic factors, including nutrition, predation, habitat destruction, and fisheries but also by climatic variations.
Sprat probably benefits from global climate change because increasing spring and summer temperatures have in empirical studies been observed to increase survival of sprat eggs and larvae (Voss et al., 2012) and in modelling studies to increase productivity and biomass of sprat (R. Voss et al., 2011; Mackenzie et al., 2012; Niiranen et al., 2013).
For herring the results are more variable. The growth rate of herring larvae is positively affected by temperature (Hakala et al., 2003), but weight at age and stock biomass of herring adults has in several studies been linked to the availability of food, mainly determined by the abundance of marine copepods and competition with sprat (Flinkman et al., 1998; Möllmann et al., 2003; Casini et al., 2011; Heikinheimo, 2011; Otto et al., 2014b). In modelling studies both an increase (Bartolino et al., 2014) and a short-term decrease (until 1950) (Niiranen et al., 2013) of herring populations have been projected.
Both herring and sprat populations probably benefited from the eutrophication during the 1950s to 1980s (Eero et al., 2016), which is the same period that the Baltic Sea eutrophication status changed from good to poor (Andersen et al., 2017; Murray et al., 2019). Since then, sprat biomass has varied independently of nutrient dynamics and has been more strongly affected by climatic variation and top-down control, i.e. cod predation and fisheries (Eero et al., 2016).
Based on experimental and modelling studies, future climatic variations may affect Baltic cod through their effects on water temperature, salinity, oxygen, and pH, as well as nutrients, which indirectly affect both the availability and quality of food (Limburg and Casini, 2019; Möllmann et al., 2021). The responses of cod larvae to ocean acidification (OA) have been studied experimentally, also in combination with warming (OAW). In some studies, no effects of OA or OAW on hatching, survival or development rates of cod larvae were found (Frommel et al., 2013), while in others mortality of cod larvae doubled when they were treated with high-end projections of OA (based on RCP8.5). When the projected increase of mortality was included into a stock-recruitment model, recruitment of western Baltic cod declined to only 8 % of the baseline recruitment (Stiasny et al., 2016), suggesting a dramatic effect of OA on cod populations.
A thorough review including long-term data and modelling demonstrated how predation, fishing, eutrophication and climate have sequentially affected eastern Baltic cod during the past century (Eero et al., 2011). In the early decades of the 20th century, cod reproduction was successful but seal predation and food availability kept the size of cod stock at a moderate level. From the 1940s, fishing replaced seal predation in controlling cod population, whereas the slowly increasing eutrophication had a minor positive influence on cod spawning stock biomass in 1950s to 1970s. In the late 1970s, a series of large saline inflows increased the salinity of the Baltic Sea and kept oxygen conditions in the deep basins favourable for cod. Consequently, reproduction peaked in 1978–1982 and, as fishing pressure was also temporarily low, the spawning stock biomass increased to a record-breaking level of ca. 700 000 t in 1980–1984 (Eero et al., 2011). After this peak period, cod stock started to decline due to a drastic reduction of the water volume where conditions are sufficiently saline and oxic for survival of cod eggs and larvae, the “cod reproductive volume” (RV). The decline of RV was associated with a stagnation period with low oxygen caused by a combination of anthropogenic eutrophication and climate-induced paucity of major saline inflows. Since then, the productivity of cod stocks has remained low (Eero et al., 2020), and the average maximum length of cod individuals has also been constantly declining (Orio et al., 2021). The reason for low growth may have been the low availability of both benthic and pelagic food (Neuenfeldt et al., 2020). Alternatively, a long-term exposure to low-oxygen conditions may affect body chemistry (Limburg and Casini, 2019) and decrease digestion rate and food consumption of cod (Brander, 2020). The physiological hypothesis is strengthened by the observed increase in depth distribution of cod and consequent dwelling of cod in low oxygen water (Casini et al., 2021).
Several studies project low abundances of cod towards the end of the century due to the climate- and eutrophication-induced decrease of RV (Eero et al., 2011, 2020; Gårdmark et al., 2013; Niiranen et al., 2013; Wåhlström et al., 2020). It has also been speculated that seal predation could contribute to keeping cod stocks low. However, although seal predation can cause damage to cod fisheries in coastal areas (Blomquist and Waldo, 2021), it has been concluded that the increased seal predation is a less important factor for the future size of fish stocks in the Baltic Sea than climate, eutrophication and fisheries (Mackenzie et al., 2011; Tomczak et al., 2021).
There is some disagreement on the effect of fisheries on cod stocks in the future. Earlier studies suggested that fisheries limitations may well enable stock recovery even in a “cod-hostile” environment (Cardinale and Svedäng, 2011; Heikinheimo, 2011). Certain recent modelling studies have however been less optimistic and projected that cod productivity will remain low due to the large impact of environmental drivers, especially oxygen and availability of food (Eero et al., 2020). For the western Baltic cod (inhabiting the Danish straits and the Arkona Sea) it has even been suggested that cod is now beyond a tipping point, with severe ecological, economic, and social consequences. At a critical moment, fisheries management failed to fully consider the changed environmental conditions, and climatic factors now prevent the recovery of cod stocks (Möllmann et al., 2021).
Increasing seawater temperature has also made it possible for certain warm-water Atlantic species, such as anchovy (Alheit et al., 2012), sole and turbot (Sparrevohn et al., 2013) to occur more abundantly in Kattegat and the southernmost Baltic Sea. Such northward and eastward migrations of these warm-water species may be caused by both global climate change and by variations in the Northern Hemisphere temperature anomalies (NHA), North Atlantic Oscillation (NAO), the Atlantic Multidecadal Oscillation (AMO), as well as contraction of the subpolar gyre (Alheit et al., 2012; Sparrevohn et al., 2013).
As for coastal freshwater fish, the distribution of pikeperch (
To sum up, temperature, salinity, oxygen and pH have a big impact on Baltic fish recruitment and growth, and as all these variables respond to climatic variations it seems evident that fish communities in the Baltic Sea will undergo changes, with the open-sea ecosystem remaining dominated by clupeids and certain freshwater fish increasing in coastal areas (Reusch et al., 2018; Stenseth et al., 2020; Möllmann et al., 2021). Together with other environmental changes, especially eutrophication, changes in fish populations may lead to altered food web dynamics (Eero et al., 2021), necessitating ecosystem-based management of fisheries and socio-ecological adaptation (Woods et al., 2021).
The Baltic Sea ecosystem is impacted by climate-induced changes in the physical and biogeochemical environment in various ways. Climatic changes affect species and populations directly and indirectly, also impacting micro-evolution of species and having synergistic effects on other environmental drivers such as eutrophication and hypoxia (Wikner and Andersson, 2012; Niiranen et al., 2013; Ehrnsten et al., 2020; Pecuchet et al., 2020; Schmidt et al., 2020). In synergy, these impacts have already boosted the emergence of “novelty” in the system and profoundly altered pathways of energy (Ammar et al., 2021). This development will probably continue, at least if the environmental conditions of the Baltic Sea continue to change as projected by modelling studies. Below, recent findings regarding climate impacts on structure and functioning on the Baltic Sea ecosystem are summarised.
For the global ocean it has been projected that climate change will decrease both primary and secondary production because of intensified stratification and decreased availability of nutrients in the surface layer (Blanchard et al., 2012; IPCC, 2019). The effects of climate change on the Baltic Sea ecosystem may however be different because of the special hydrographical characteristics, peculiar communities, strong seasonal cycle, and the strong dependency of the Baltic Sea of both its watershed and the adjacent North Sea.
In the Baltic Sea, changes in ice conditions, water temperature, density stratification, and especially supply of nutrients through rivers and from the sediment, affect the nutrient dynamics and primary productivity in both coastal areas and the open sea. Different species however respond in different ways to changes in the environmental parameters, and both increases and decreases in primary production have been reported and projected along with climate-induced changes in the environment.
Climate change will most probably mean milder winters, and if soils remain
thawed, more nutrients will leak from the terrestrial areas into the
freshwater system. The nutrient load into the sea will probably increase,
especially in the northern Baltic Sea where precipitation is probably
increasing the most (Lessin et al., 2014; Huttunen et al., 2015;
Christensen et al., 2022) but also in the southern Baltic Sea (M. Voss et
al., 2011). It has also been projected that the total phosphorus loading
(from terrestrial areas of Finland) will increase relatively more than that
of nitrogen (Huttunen et al., 2015) and, together with the
internal loading of phosphorus from sediments (Lessin et al., 2014;
Stigebrandt et al., 2014; Stigebrandt and Anderson, 2020), phosphorus
availability to primary producers may increase. If the
In the central Baltic Sea, increased spring water temperature causes,
together with increased irradiation and enhanced wind-induced mixing of the
surface layer, an earlier but less intense spring bloom. In summer, in
contrast, an increase in temperature is coupled with increased thermal
stratification, which is projected to favour production of cyanobacteria
(Meier et al., 2011a; Neumann et al., 2012; Chust et al., 2014; Andersson
et al., 2015). Intensified blooms of cyanobacteria are expected especially
if hypoxia will prevail and internal loading will decrease the
Several modelling studies project an increase in total phytoplankton
concentration (chlorophyll, in
Nutrient abatement may however counteract climate effects. For instance, in
Kattegat in mid-1990s, reduction of nutrient loading led to a shift from a
highly eutrophic state, characterised by small phytoplankton species and low
water transparency, to an improved state, with a larger share of diatoms,
decreased phytoplankton biomass and increase in water transparency
(Lindegren et al., 2012). An opposing trend has taken place in the
Bothnian Sea. Because of the lack of halocline and lower anthropogenic
nutrient loading, the Bothnian Sea has thus far remained in a relatively
good condition. However, since the year 2000 the Bothnian Sea has also shown
symptoms of eutrophication (Kuosa et al., 2017), and open-sea
cyanobacteria blooms have also become more common in recent years, due to a
“leaking” of phosphorus-rich water from the central Baltic Sea through the
Åland Sea (Rolff and Elfwing, 2015; Ahlgren et al., 2017). The
connection of this process to climate change is not certain. Rather, the
severe hypoxia of the central Baltic Sea has brought the anoxic layer so
close to the sill separating the Baltic Proper from the Åland Sea that
flow of nutrient-rich water across the Åland Sea is at times possible.
Whether or not the proceeding climate change will amplify the ongoing
eutrophication of the Bothnian Sea remains to be seen, but if temperature
stratification increases and
Several recent modelling studies conclude that nutrient abatement according to HELCOM BSAP will in the long run counteract the climate-induced increase in nutrient loading and lead to decreased eutrophication (Meier et al., 2018; Ehrnsten et al., 2019a; Murray et al., 2019; Pihlainen et al., 2020). Based on oceanographic–biogeochemical modelling, it has also been suggested that hypoxia will eventually diminish (Meier et al., 2021) and that extreme cyanobacteria blooms will no longer occur in the future, if nutrient loadings will be lowered according to BSAP despite the proceeding climate change (Meier et al., 2019a).
To sum up, the fate of the level of primary production and level of eutrophication will depend on various intertwined factors and processes and on the development of both the climate and society. Changes in primary production will impact interactions between the main trophic levels, i.e. phytoplankton, detritus, zoobenthos, detritivores, benthivores, grazers, zooplanktivores and piscivores (Kiljunen et al., 2020; Kortsch et al., 2021).
Recycling and build-up of carbon within the ecosystem determines the overall productivity and biomass of different trophic levels. Several studies suggest fundamental changes in trophic dynamics and eventually in the pathways of carbon in the Baltic Sea.
A climate- and nutrient-load-driven model reconstruction of the Baltic Sea state from 1850 to 2006 suggests that the shift from spring to summer primary production is accompanied by an intensification of pelagic recycling of organic matter (Gustafsson et al., 2012). In mesocosm studies, warming accelerated (southern Baltic Sea) phytoplankton spring bloom and increased carbon-specific primary productivity (Sommer and Lewandowska, 2011; Sommer et al., 2012; Paul et al., 2016). The total phytoplankton biomass decreased because increased stratification decreased nutrient flux to the surface layer however (Lewandowska et al., 2012, 2014). Furthermore, in stratified conditions the relative importance of pathways of carbon through the microbial food web increased because copepods switched to feed more on ciliates than phytoplankton. Decrease in ciliates in turn increased the amount of heterotrophic nanoflagellates grazing on bacteria. Decreases in bacteria may reduce remineralisation and thus decrease availability of nutrients for phytoplankton (Lewandowska et al., 2014). On the other hand, decreasing of bacteria would also decrease competition for nutrients between bacteria and phytoplankton, which could counteract the negative effects of diminishing remineralisation on phytoplankton.
It has also been projected that, in addition to nutrients, the flow of dissolved organic matter (DOM) into the Baltic Sea will increase in the future climate (M. Voss et al., 2011; Strååt et al., 2018). Precipitation will increase, especially in the northern areas, and by using long-term time series from 1994 to 2006, it was shown that climate change has increased discharge of terrestrial DOM into the middle part of the Gulf of Bothnia. This provided additional substrate for bacteria, which maintained bacterial biomass production despite reduced phytoplankton production (Wikner and Andersson, 2012). This suggests that increased humic-rich river inflow may counteract climate change induced eutrophication in the coastal waters (Andersson et al., 2013).
Experimental studies have also demonstrated increased microbial activity and biomass with increasing DOM and temperature (Ducklow et al., 2010), although different bacteria taxa respond differently to the simultaneous increase of DOM and temperature (Lindh et al., 2015). Increase of DOM and bacteria may be detrimental to primary production as bacteria compete for nutrients with phytoplankton and as the brownification of water reduces light availability. Consequently, the carbon flow shifts towards microbial heterotrophy, which may induce a decrease in both phytoplankton productivity and biomass and lead to a promotion of the microbial food web and other heterotrophic organisms (Wikner and Andersson, 2012; Andersson et al., 2013). Especially if stratification increases, cycling of carbon through the microbial food web increases pelagic recycling and may also decrease vertical flux of organic matter to zoobenthos (Ehrnsten et al., 2020).
It has been suggested that climate change may also decrease fish productivity. In areas where climate change increases the supply of allochthonous DOM into the system, and where increasing stratification reduces the transport of nutrients from deeper waters, phytoplankton production may decline and the trophic pathways from bacteria and flagellates through ciliates to copepods may strengthen (Aberle et al., 2015). When the system shifts towards heterotrophy, the food web efficiency declines (Båmstedt and Wikner, 2016), and if zooplankton also becomes dominated by smaller-sized plankton (Dahlgren et al., 2011; Suikkanen et al., 2013; Jansson et al., 2020), there will be less suitable food available for planktivorous fish. If sedimentation of organic matter will also be reduced, zoobenthos production will decrease and there will be less food for benthic-eating fish. Eventually the total fish production may decrease.
Results of experimental studies have not equivocally confirmed this hypothesis. A study performed in a large biotest area artificially heated by the cooling waters of the Forsmark nuclear power plant, in the southern Bothnian Sea, found that warming of water may lead to increased species turnover and decreased compositional stability of diatom, macrophyte and invertebrate communities (Hillebrand et al., 2010). Certain mesocosm studies, simulating effects of climate change in the pelagic ecosystem, have also found that the production and biomass of both copepods and fish (three-spined sticklebacks) can remain high because the positive effects of increasing temperature and increasing availability of DOC override the negative effects of decreasing food web efficiency on copepod production (Lefébure et al., 2013).
Furthermore, many Baltic Sea copepods are omnivorous and can opportunistically switch between suspension feeding on flagellates and raptorial feeding on ciliates (Kiørboe et al., 1996). Such a flexible feeding strategy stabilises the system and can sustain copepod production even under lower phytoplankton production. This flexibility, and the fact that heterotrophic production increases with high DOC availability, suggests that fish production may be supported even when a relatively large amount of carbon flows through the microbial food web (Lefébure et al., 2013).
To sum up, a reorganisation of pathways of carbon is possible in the Baltic Sea due to the climate change. However, the system is complex due to several counteracting and interacting processes and large uncertainties in key processes, such as stratification and nutrient loads from land and the sediments (Meier et al., 2019c; Saraiva et al., 2019), and both increases and decreases of secondary producers have been demonstrated in field, experimental and modelling studies. The complexity of the system has been highlighted by a thorough review that illustrated how changes in benthic–pelagic coupling may induce ecosystem-wide consequences via increasing sedimentation of organic matter, inducing hypoxic conditions and internal loading of nutrients (Griffiths et al., 2017).
If climate change induces an increase in allochthonous nutrient loads, consequences can be expected in the communities of algae and vascular plants in the shallow photic zone. The shallow-water food webs based on macroalgae and seagrasses may also be affected by the indirect effects of climate change, mediated through interactions between algae and their grazers.
The effects of late summer heatwaves on algae and invertebrates living
amongst bladderwrack
Similar results were obtained in an artificially heated biotest basin
(Forsmark nuclear power plant) in the Gulf of Bothnia, where the biomass of
the non-native gastropod grazer
Decline of bladderwrack will affect other species due to declining
availability to habitat and food (Takolander et al., 2017a; Jonsson et
al., 2018; Kotta et al., 2019). Connectivity between bladderwrack
populations and organisms inhabiting patches of bladderwrack may also
decline (Jonsson et al., 2020; Virtanen et al., 2020). However, perhaps
due to the complex biotic interactions in the sublittoral ecosystem, there
are very few modelling studies that have attempted to project the fate of
the algal and invertebrate communities inhabiting the shallow photic zone of
the Baltic Sea. Only one study has used a combination of experimental work
and modelling to study the effects of climate change on invertebrates. A
decline of the isopod
To sum up, temperature and salinity changes have been projected to affect species interactions in hard and soft bottoms in the sublittoral zone. Both summer heatwaves and cold season warming can induce novel trophic interactions that produce eutrophication-like effects, e.g. overgrowth of bladderwrack by epiphytes, in the photic zone dominated by macroalgae, even without an increase in nutrient loading. However, as macroalgae are very much dependent on water clarity, the future level of eutrophication will also affect the fate of the shallow-water communities in the Baltic Sea. The complexity of the system, uncertainty of the oceanographic projections, and unknown adaptation capabilities of species, make it challenging to project the future food web interactions in the sublittoral ecosystem.
In the 1980s a partly climate-induced regime shift was recorded with drastic changes in the central Baltic food web, including phytoplankton, zooplankton and pelagic planktivores and their main predator, Baltic cod (Möllmann et al., 2009; Lindegren et al., 2010a). In 1980–2000, a decline in “reproductive volume” (RV), contributed to the decline of cod population (Hinrichsen et al., 2011; Casini et al., 2016; Bartolino et al., 2017) and induced cascading effects on planktivorous fish and zooplankton (Casini et al., 2008). The different effects of temperature and salinity on sprat and cod (see above) also resulted in a spatial mismatch between these species, which further released sprat from cod predation and contributed to the increase of sprat stocks in the central Baltic Sea (Eero et al., 2012; Reusch et al., 2018). As herring is an inferior competitor for food, and food availability per individual declined, the condition of herring declined (Möllmann et al., 2003; Casini et al., 2010). Transition to a lower saline Baltic Sea, and associated decline of marine copepods (Hänninen et al., 2015), also contributed to the observed halving of (3-year-old) herring weight at age, from 50–70 g in the late 1970s to 25–30 g in the 2000s (Dippner et al., 2019). The described regime shift has also been partly questioned, as the descriptions of the shift did not cover the entire food web (Yletyinen et al., 2016).
A factor that has been less often considered when studying reasons of cod decline is the interaction with another benthic predator, flounder. Flounder may be both prey for larger cod and a competitor for the small and juvenile ones. Now that cod size has declined, cod predation on flounder has decreased, releasing competition for benthic food again. This has caused more spatial overlap between flounder populations and the remaining small sized cod, and created more intense competition between flounder and the small-sized cod, further contributing to the decline in body condition of cod (Orio et al., 2020).
Multi-species modelling studies have concluded that both fishing and climate
strongly affects the size of cod stocks. If fishing is intense but climate
remains unchanged, cod declines but not very dramatically, while if climate
change proceeds as projected (according to the intermediate-high A2
scenario), cod goes extinct in two models out of seven, even with the present
low fishing effort (Gårdmark et al., 2013). Different combinations of
climate change and eutrophication scenarios may however yield very different
outcomes. Medium
The above studies have mostly considered the ecosystem of the central Baltic. In other basins, the associated processes and species interactions may be different. For example, in the Bothnian Bay, salinity was also a major driver for changes in populations of planktivorous fish, but the species involved were different. Here the decline of spawning-stock biomass of herring, observed in 1980–2013, was explained by a simultaneously increased competition with vendace, a limnic species that had increased with lowering salinity (Pekcan-Hekim et al., 2016).
In Kattegat, the western Baltic Sea, where the ecosystem is more oceanic than in the other parts of the Baltic Sea, a regime shift was detected in mid-1990s. Here the shift was explained by global climate change, cyclic climate phenomena, and human intervention. First, a reduction in anthropogenic nutrient loading led into a shift from a eutrophic ecosystem state to an ecosystem characterised by decreasing phytoplankton and zooplankton biomass, dominance by small-sized fish in the pelagial, and an increase in macroalgae and filter-feeding molluscs on hard bottoms and other benthic animals in the soft sediments (Lindegren et al., 2012). Second, the positive phases of NAO and BSI enabled an inflow of oxygenised water from the North Sea, which improved conditions for zoobenthos, including the commercially important Norway lobster. A climate-induced increase of sea surface temperatures contributed to the improved flatfish growth and survival in the shallow nursery areas (Lindegren et al., 2012). Decreasing fishing may also have contributed to the increase of gadoid and flatfish populations, but its relative importance is difficult to distinguish from other co-occurring effects.
To sum up, regime shifts are usually a result of several environmental, climatic and anthropogenic effects acting synergistically on the entire ecosystem. The climate-driven changes in temperature and salinity have been identified as key drivers for the significant rise of “novelty” in both abiotic conditions and biotic assemblages in several basins of the Baltic Sea (Ammar et al., 2021), but human contributions, i.e., anthropogenic eutrophication or its alleviation, have been a factor (Reusch et al., 2018). The recent research confirms that climate change induces multiple direct and indirect effects on species and communities and affects nutrient and carbon dynamics of the Baltic Sea ecosystem. However, despite the major structural changes, the overall food web complexity in the central Baltic Sea has remained surprisingly stable (Yletyinen et al., 2016). The relatively small changes may be explained by the fact that responses to climate change are not uniform or unidirectional but vary from species group to another, within groups and even between sibling species. Species-specific responses, many feedbacks, altered trophic pathways and the possibility of species-level adaptation combine to make projections concerning the state of the ecosystem and trophic effects challenging.
The main challenge when analysing effects of climate change on the Baltic Sea is the possible synergistic effects of climate with other environmental drivers, such as eutrophication, harmful substances and the introduction of non-indigenous species, which also may have profound impacts on ecosystems and their functioning (Reusch et al., 2018; Stenseth et al., 2020; Bonsdorff, 2021). Consequently there are numerous knowledge gaps, bottlenecks and issues of dissensus that weaken our ability to project the future biological processes, such as primary and secondary productivity, benthic–pelagic coupling and hypoxia, interactions between phytoplankton, zooplankton, and fish populations; and geographic shifts in macroalgal and invertebrate communities.
Attribution of the observed phenomena to climate change is challenging because of the collinear, intertwined and interacting processes. It is especially difficult to distinguish the effects of anthropogenic global climate change from those of quasi-cyclic phenomena, such as the NAO or BSI, or from other more stochastic variations in climate. This is partly due to the slow pace of climatic variations and time lags between physical and chemical variations and ecosystem responses. A fairly small number of studies have investigated a period long enough to cover any larger number of NAO periods. Research into the long-term dynamics of the food webs in particular is still scarce (Törnroos et al., 2019; Pecuchet et al., 2020; Kortsch et al., 2021).
Field studies have ended up with different conclusions concerning past and present changes of the environment and the biota and their causes depending on time periods and data scrutinised. For instance, certain studies note that cyanobacteria have increased (Suikkanen et al., 2013; Kuosa et al., 2017), while others do not find proof for such a phenomenon (Griffiths et al., 2020; Olofsson et al., 2020). Different periods studied, sparse sampling, varying species responses, and changes in phenology rather than total biomass, may explain some of the discrepancies between studies. The tendency of filamentous cyanobacteria to float during calm weather may also bias our view on the total biomass cyanobacteria in the sea, especially if low wind periods become more frequent.
Experimental studies are useful in pinpointing causative relationships, but their small spatial scales, short duration and simple food webs make upscaling of results to natural systems difficult. Experiments usually only last for a few days or weeks and study one or few species at a time. Reproducing natural patterns of environmental variability is also challenging. When mesocosms of hundreds of litres and natural communities are used, it may be difficult to simulate seasonal processes extending over several life cycles of the studied organisms. Even the most sophisticated multi-stressor experiments, which use levels of environmental stressors projected by modelling studies, tend to use constant stress levels.
A few mesocosm studies have exposed the communities to near-natural environmental conditions and have been able to shed light on the complex dynamics of the Baltic Sea ecosystem, e.g. the responses of the microbial food web to changes of environmental variables affected by the climate change. In studies made in the Gulf of Bothnia, bacterial, phytoplankton and zooplankton production increased with additions of inorganic carbon, and the systems remained net autotrophic. In contrast, when both nutrients and DOC was increased, only bacterial and zooplankton production increased, driving the system to net heterotrophy (Andersson et al., 2013; Båmstedt and Wikner, 2016). Increased heterotrophy led to a decreased fatty acid content and lower individual weight in the zooplankton (Dahlgren et al., 2011). With the combined treatment of elevated temperature and terrestrial nutrient loads, fish production (of three-spined sticklebacks) also increased, with terrestrial rather than autotrophic carbon being the main energy source (Lefébure et al., 2013). The complex responses indicate that to provide useful inferences about physiological and population-level responses of organisms to climate change, experimental work should use full communities, apply naturalistic exposure regimes, and investigate effects of stress at spatial and temporal scales appropriate to the species studied (Gunderson et al., 2016).
Ecosystem modelling using coupled oceanographic–biogeochemical models has advanced greatly in the past 15 years, but significant challenges remain. Projections of sea surface temperature and ice conditions can be held as relatively reliable, but there are still large uncertainties in projecting salinity, stratification, hypoxia and hence the rate of internal loading (Meier et al., 2022a). Also, natural variability is a larger source of uncertainty in future projections of hypoxia than previously understood (Meier et al., 2021). Because salinity, stratification and oxygen strongly affect many Baltic Sea organisms, it is difficult to project the fate of plankton and benthos communities with certainty. This uncertainty especially concerns marine species such as cod, bladderwrack, eelgrass, and blue mussel, which in many studies have been projected to decrease in the northern basins of the Baltic Sea. Further, uncertainties are imposed by complex biogeochemical processes in the terrestrial and freshwater ecosystems, as well as by unknown development of national economies and farming practices (Huttunen et al., 2015), especially in coastal areas strongly affected by nutrient loading.
Ecosystem models rarely consider complex biological interactions and feedback effects caused, for example, by multi-species predatory or intraguild relationships. Inclusion of such effects would require parameterizing the 3D ecosystem models with experiments and results from multi-species food web models that operate on the level populations rather than carbon flows. Also, models cannot at present consider potential adaptation capabilities of species, as little is known about them. Several recent studies have however pointed out, for example, that macroalgae (Rothäusler et al., 2018; Rugiu et al., 2018a) and zooplankton (Karlsson and Winder, 2020) have phenotypic plasticity and potential for adaptation against gradual changes in the abiotic environment.
Food web models offer useful tools for assessing the relative effects of climate, eutrophication and other human impacts, including fisheries, on the structure of the Baltic Sea ecosystem. They could potentially take into account characteristics of species and their responses to changes in the environment. The current models however mostly concern the pelagic ecosystems (e.g. cod–sprat–herring–zooplankton food chain) and there are major gaps for key trophic groups, such as macrophytes and macrozoobenthos (Korpinen et al., 2022) and the microbial food web.
The 3D ecosystem models, food web models and 2D spatial modelling would benefit from integration. Species distribution models (SDMs) can be produced at a fine spatial scale, even a few tens of metres (Virtanen et al., 2018), and in climate change studies they can be parameterised with 3D model results (Jonsson et al., 2018; Kotta et al., 2019). In the future, food web models involving relevant coastal taxa could also be used to fill in the missing links between the large-scale (3D) processes and detailed spatial patterns identified by the 2D models.
Assessing climate effects in a smaller spatial scale would be useful because shallow and sheltered bays, lagoons and estuaries may be more susceptible to climate change effects than deeper offshore areas, and may appear as “climate change hotspots”, where climate change drives the ecosystem towards a new state (Queiros et al., 2021). The existing coupled oceanographic–biogeochemical modelling studies however typically have a horizontal resolution of 1 or 2 nautical miles (ca. 2 or 4 km) and thus cannot easily be used for projecting local variations in temperature, salinity and stratification within the archipelago or inside estuaries. A bottleneck for high-resolution 3D models is the poor availability of high-resolution pan-Baltic bathymetries and forcing data (e.g. wind fields). For the SDMs, in turn, a major constraint is in many areas the poor availability of detailed species and habitat mapping data, as well as availability of high-resolution data on benthic substrates. Considering population-level effects on spatial patterns of species would also require estimation of connectivity between sea areas, a research field that is also underdeveloped in the Baltic Sea (Berglund et al., 2012; Jonsson et al., 2020; Virtanen et al., 2020). Consequently, no study has thus far considered how climate change affects microclimatic patterns in the Baltic Sea or how different species and habitats may respond to such local variations.
Due to the above challenges, there are certain discrepancies concerning our view on the effects of climate change on the structure and function of the Baltic Sea ecosystem. Some of these issues are highlighted below.
Increased primary production and phytoplankton biomass (measured in
chlorophyll
Several ecosystem models also predict an increase of cyanobacteria. As
cyanobacteria blooms are favoured by warm, stabile and conditions and low
Recent awareness of marine heatwaves and their potential impacts on the marine ecosystem has increased our knowledge on how climate change may impact pelagic, benthic and littoral communities in the ocean (Pansch et al., 2018; Saha et al., 2020). More studies on the responses of pelagic and benthic organisms of the Baltic Sea to heatwaves would increase our understanding of the population-level consequences of short-term variability in environmental parameters. Research on effects of climate change would also benefit from methodological diversity. For example, more extensive use of biochemical and genetic methods, such as biomarkers (Turja et al., 2014, 2015; Villnäs et al., 2019), stable isotopes (Voss et al., 2000; Gorokhova et al., 2005; Morkune et al., 2016; Lienart et al., 2021), compound-specific isotope analyses (Ek et al., 2018; Weber et al., 2021) or metabarcoding (Leray and Knowlton, 2015; Bucklin et al., 2016; Klunder et al., 2022), as well as development of remote sensing methods (Huber et al., 2021), could yield novel information on stress levels experienced by organisms and environmental niches preferred by species. Such information would allow validation of the biogeochemical models under different environmental and climate scenarios.
There is some bias in the focus organisms and habitats studied. While experiments on planktonic organisms and soft-bottom animals are relatively abundant, experiments on macroalgae, vascular plants and invertebrates inhabiting hard bottoms are less abundant, and studies focusing on the entire food web are scarce. In general, empirical and modelling studies focusing on climate effects on shallow photic habitats are less abundant than those on the pelagic and deep benthic habitats (Tedesco et al., 2016). Very few studies have investigated the shallow water ecosystems holistically, including macroalgae and microalgae, invertebrates, and fish at the same time. Those that have done so have revealed complex interactions and multiple feedbacks between species and ecosystem components (Svensson et al., 2017; Salo et al., 2020). Also, while there are ample monitoring data on pelagic and deep benthic communities, similar long-term records are very sporadic for communities associated with key habitat-forming species such as bladderwrack, eelgrass, blue mussel on hard bottoms, and vascular plants growing on soft sediments. This lack of empirical data and subsequent modelling studies hampers our understanding of the long-term responses of sublittoral communities to climate change.
Furthermore, there is a large body of literature published on sea ice algae and sea ice ecology in the Baltic Sea (Granskog et al., 2006; Tedesco et al., 2017; Thomas et al., 2017), and all of them are relevant for studying winter ecology. However, few of them have directly assessed the effects of climate change on ice ecology in the Baltic Sea. More empirical and modelling studies including quantitative projections on the effect of diminishing sea ice to biodiversity and functioning of the Baltic Sea ecosystem in winter and spring would therefore be desirable.
To sum up, there are still several significant knowledge gaps and issues of dissensus in our understanding of the effects of climate change on the Baltic Sea ecosystem. To fill these gaps, the results and conclusions from the experimental work should be better integrated into the wider empirical and modelling studies of food web dynamics, and more emphasis should be placed on studying effects of climate change on less studied environments, such as the microbial food web, sea ice communities and the sublittoral ecosystem. Such studies would provide a more comprehensive view of the responses of the pelagic and benthic systems to climate change in both the open sea and the benthic system, from bacteria to fish (Kortsch et al., 2021). Also, continuation of both spatial mapping programmes and long-term ecological studies will be crucial for validating experimental results and for developing ecosystem models, advancing our understanding of environmental and meteorological drivers of the Baltic Sea ecosystem on large spatial and temporal scales.
Climate change has an obvious potential to affect entire marine food webs, from coastal to offshore areas, from shallow to deep waters and from pelagic to benthic systems. Climate change can also induce changes in species distributions and proportions, and key nodes and linkages in the food webs may be altered or lost (Lindegren et al., 2010b; Niiranen et al., 2013; Leidenberger et al., 2015; Griffiths et al., 2017; Kotta et al., 2019; Gårdmark and Huss, 2020). As many ecosystem services are dependent on the state of the entire ecosystem (Hyytiäinen et al., 2019), a long-term decline in provision of ecosystem services to humans is possible. It is therefore indispensable to increase our understanding of the consequences of climate change on the socio-ecological system of the Baltic Sea and its surrounding marine regions (Stenseth et al., 2020).
The direct and indirect effects of climate-change-related parameters on species, communities and the ecosystem are summarised in Table 1 based on research done since 2010. While results are variable, some conclusions can be drawn from the evidence this far.
Summary of research findings and conclusions on the anticipated
effects of climate change (CC) effects in the Baltic Sea for selected
variables. The table only shows studies published in 2011–2021 and some of the
studies referred to in the text are not included. For earlier studies, see
Dippner et al. (2008) and Viitasalo et al. (2015). Observations,
experimental simulations or modelled projections are categorised using the following abbreviations: T stands for temperature
increase, S stands for salinity decline, TSO
Continued.
Continued.
Continued.
As for the eutrophication status of the Baltic Sea, it can be concluded that the ecological status of the Baltic Sea has not significantly improved despite a decrease in anthropogenic nutrient loading since the 1980s (Fleming-Lehtinen et al., 2015; Andersen et al., 2017), largely due to the pervasive internal loading (Murray et al., 2019; Stigebrandt and Anderson, 2020). Success of nutrient abatement largely determines the future state of the Baltic Sea (Hyytiäinen et al., 2019; Ehrnsten et al., 2020), but climate change may delay or even counter the improvement of the ecosystem state (Bonsdorff, 2021).
Climate-induced increases of nutrient loading and enhancing of internal loading of phosphorus have been hypothesised to promote phytoplankton and cyanobacteria production and to maintain the “vicious circle of eutrophication” (Vahtera et al., 2007), and several modelling studies indeed project an increase in both total phytoplankton biomass and cyanobacteria blooms in the future (Meier et al., 2011a; Funkey et al., 2014).
The eutrophication process may however be counteracted by various factors. Increase of DOM flowing via the rivers may decrease both primary and secondary production, at least in the Gulf of Bothnia (Wikner and Andersson, 2012; Andersson et al., 2013), and certain cyanobacteria may be negatively affected by increased temperature and ocean acidification (Paul et al., 2018). Thus, changes in structure and functioning of phytoplankton and cyanobacteria communities are probable, but the narrative that global climate change will inevitably increase phytoplankton biomass and cyanobacteria blooms and amplify the eutrophication of the Baltic Sea, may be too simplistic and needs to be refined by reconsidering the climate effects on food web processes and nutrient and carbon dynamics.
For the deep benthic communities, climate change effects are also not straightforward. If salinity declines, the majority of marine species will suffer, but according to the latest analyses undisputable evidence is lacking for a future decline in the salinity of the Baltic Sea (Lehmann et al., 2022; Meier et al., 2022b). Improvement of oxygen conditions may first promote higher zoobenthos biomasses, but eventually increasing stratification will weaken benthic–pelagic coupling and reduce food availability for benthic organisms. If nutrient abatement also proceeds favourably, biomass of zoobenthos will start to decline (Ehrnsten et al., 2020).
In the shallower photic benthic systems, nutrient increase probably enhances
eutrophication, and if salinity also declines, habitat-forming marine
species, such as bladderwrack, eelgrass and blue mussel, probably decline in
the northern Baltic Sea (Vuorinen et al., 2015; Jonsson et al., 2018;
Kotta et al., 2019). As both eutrophication and increasing temperature
favour filamentous algae, continued major changes in the sublittoral
communities can be expected, including negative effects of such algal
aggregations (Arroyo and Bonsdorff, 2016). Of particular concern is the
potential loss from rocky substrates of the habitat-forming bladderwrack and
red macroalgae. Freshwater vascular plants will be favoured by freshening of
the Baltic Sea, but they cannot replace the marine macroalgae on rocky
sublittoral because they only grow on soft substrates. On the other hand,
salinity projections are still uncertain (Lehmann et al., 2022), and
even if salinity declined,
As for fish, responses also depend on species. Salinity decline and hypoxia increase will most probably have negative consequences on cod stocks (Gårdmark et al., 2013), whereas the increasing temperature has been projected to favour sprat (Mackenzie et al., 2012) and certain coastal fish (Bergström et al., 2016). Again, as projections for salinity, stratification and oxygen levels are uncertain, the future fate of fish populations cannot be projected with certainty.
The global climate change induces many environmental changes that may favour establishment of NIS in the Baltic Sea. Opportunistic and thermophilic species occupying soft sediments are the most probable winners. It is notable that it is extremely difficult to eradicate a marine NIS after it has found a suitable niche in the Baltic Sea. As the effects of NIS on both the ecosystem and the society are usually negative, their spreading should be prevented already before they enter the Baltic Sea, by effectively eradicating NIS from ballast waters of ships and other possible vectors.
Climate change is obviously not the only factor determining the fate of the Baltic Sea in the future. Several modelling studies have concluded that nutrient reductions will be a stronger driver for ecosystem functions in the Baltic Sea than climate change (Friedland et al., 2012; Niiranen et al., 2013; Ehrnsten et al., 2019a; Pihlainen et al., 2020; Meier et al., 2021). In moderate nutrient loading scenarios climate change will also play a role, but under full implementation of BSAP, the environmental state of the Baltic Sea is projected to become significantly improved and hypoxia reduced by the end of the century (Meier et al., 2018, 2021; Saraiva et al., 2018, 2019). Despite the many uncertainties concerning the effects of climate and eutrophication on the state of the Baltic Sea (Munkes et al., 2021), it can be concluded that continued abatement of anthropogenic nutrient loading, combined with sustainable fisheries, seem to be the most reliable, albeit slow, measures to solve the grand challenges of the Baltic Sea (Meier et al., 2018; Murray et al., 2019).
Several studies have focused on studying the effects of climate change on the future state of the Baltic Sea, and the ecosystem modelling studies already provide valuable results that are directly usable in decision-making concerning mitigation of eutrophication under climate change. In contrast, studies concerning effects of climate change on biodiversity of the Baltic Sea are lagging behind and are hampered by model uncertainties (e.g. for salinity) and by the current inability of models to consider the complex interactions between species and multiple feedbacks between trophic levels. Long-term and modelling studies focusing on shallow photic environments, which harbour the highest biodiversity in the Baltic Sea, are especially sparse. This is a major drawback in a situation where all major environmental policies, including the UN Convention on Biological Diversity and EU Biodiversity Strategy for 2030, urge for halting ongoing biodiversity loss. To designate effective measures to safeguard biodiversity, including a climate-smart expansion of the protected area network, a better understanding of the effects of climate change on the sublittoral ecosystem is urgently needed.
Knowledge of the mechanisms and processes governing the Baltic Sea ecosystem under climate change have recently accumulated and already provide information that can be used to design adaptation tools and mitigation measures for the Baltic Sea (Reusch et al., 2018). It is necessary to continue studying the Baltic Sea as a socio-ecological system, responding to both environmental and societal changes (Bauer et al., 2018, 2019; Hyytiäinen et al., 2019), and to continue the dialogue with human society, in order to attune to the future changes ultimately driven by the ocean itself (Stenseth et al., 2020).
This review paper refers exclusively to published research articles and their data. We refer the reader to the cited literature for access to data.
MV designed the review article, wrote main parts of the paper, produced the table and edited the manuscript versions for review. EB contributed with texts and comments.
The contact author has declared that neither they nor their co-author has any competing interests.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This article is part of the special issue “The Baltic Earth Assessment Reports (BEAR)”. It is not associated with a conference.
This work has been done under the auspices of the Finnish Environment Institute, Marine Research Centre (Markku Viitasalo) and is a part of the Åbo Akademi University Strategic Profile “The Sea” (Erik Bonsdorff). We thank the Baltic Earth secretariat for inviting us to write this review and three anonymous reviewers for constructive comments significantly improving earlier versions of the manuscript.
This research has been supported by the projects SmartSea (Academy of Finland, Strategic Research Council, grant nos. 292985 and 314225) and FutureMARES (EU Horizon 2020, grant no. 869300) for Markku Viitasalo and by the Åbo Akademi Foundation for Erik Bonsdorff.
This paper was edited by Markus Meier and reviewed by three anonymous referees.