Integrating diverse model results into decision support for good environmental status and blue growth.

Sustainable environmental management needs to consider multiple ecological and societal objectives simultaneously while accounting for the many uncertainties arising from natural variability, insufficient knowledge about the system's behaviour leading to diverging model projections, and changing ecosystem. In this paper we demonstrate how a Bayesian network- based decision support model can be used to summarize a large body of research and model projections about potential management alternatives and climate scenarios for the Baltic Sea. We demonstrate how this type of a model can act as an emulator and ensemble, integrating disciplines such as climatology, biogeochemistry, marine and fisheries ecology as well as economics. Further, Bayesian network models include and present the uncertainty related to the predictions, allowing evaluation of the uncertainties, precautionary management, and the explicit consideration of acceptable risk levels. The Baltic Sea example also shows that the two biogeochemical models frequently used in future projections give considerably different predictions. Further, inclusion of parameter uncertainty of the food web model increased uncertainty in the outcomes and reduced the predicted manageability of the system. The model allows simultaneous evaluation of environmental and economic goals, while illustrating the uncertainty of predictions, providing a more holistic view of the management problem.

Composition and Seasonality of Membrane Transporters in Marine Picoplankton.

In this study, we examined transporter genes in metagenomic and metatranscriptomic data from a time-series survey in the temperate marine environment of the Baltic Sea. We analyzed the abundance and taxonomic distribution of transporters in the 3µm-0.2µm size fraction comprising prokaryotes and some picoeukaryotes. The presence of specific transporter traits was shown to be guiding the succession of these microorganisms. A limited number of taxa were associated with the dominant transporter proteins that were identified for the nine key substrate categories for microbial growth. Throughout the year, the microbial taxa at the level of order showed highly similar patterns in terms of transporter traits. The distribution of transporters stayed the same, irrespective of the abundance of each taxon. This would suggest that the distribution pattern of transporters depends on the bacterial groups being dominant at a given time of the year. Also, we find notable numbers of secretion proteins that may allow marine bacteria to infect and kill prey organisms thus releasing nutrients. Finally, we demonstrate that transporter proteins may provide clues to the relative importance of biogeochemical processes, and we suggest that virtual transporter functionalities may become important components in future population dynamics models.

The meagre future of benthic fauna in a coastal sea-Benthic responses to recovery from eutrophication in a changing climate.

Nutrient loading and climate change affect coastal ecosystems worldwide. Unravelling the combined effects of these pressures on benthic macrofauna is essential for understanding the future functioning of coastal ecosystems, as it is an important component linking the benthic and pelagic realms. In this study, we extended an existing model of benthic macrofauna coupled with a physical-biogeochemical model of the Baltic Sea to study the combined effects of changing nutrient loads and climate on biomass and metabolism of benthic macrofauna historically and in scenarios for the future. Based on a statistical comparison with a large validation dataset of measured biomasses, the model showed good or reasonable performance across the different basins and depth strata in the model area. In scenarios with decreasing nutrient loads according to the Baltic Sea Action Plan but also with continued recent loads (mean loads 2012-2014), overall macrofaunal biomass and carbon processing were projected to decrease significantly by the end of the century despite improved oxygen conditions at the seafloor. Climate change led to intensified pelagic recycling of primary production and reduced export of particulate organic carbon to the seafloor with negative effects on macrofaunal biomass. In the high nutrient load scenario, representing the highest recorded historical loads, climate change counteracted the effects of increased productivity leading to a hyperbolic response: biomass and carbon processing increased up to mid-21st century but then decreased, giving almost no net change by the end of the 21st century compared to present. The study shows that benthic responses to environmental change are nonlinear and partly decoupled from pelagic responses and indicates that benthic-pelagic coupling might be weaker in a warmer and less eutrophic sea.

Food web and fisheries in the future Baltic Sea.

We developed numerical simulations of potential future ecological states of the Baltic Sea ecosystem at the end of century under five scenarios. We used a spatial food web (Ecospace) model, forced by a physical-biogeochemical model. The scenarios are built on consistent storylines that describe plausible developments of climatic and socioeconomic factors in the Baltic Sea region. Modelled species diversity and fish catches are driven by climate- and nutrient load-related changes in habitat quality and by fisheries management strategies. Our results suggest that a scenario including low greenhouse gas concentrations and nutrient pollution and ecologically focused fisheries management results in high biodiversity and catch value. On the other hand, scenarios envisioning increasing societal inequality or economic growth based on fossil fuels, high greenhouse gas emissions and high nutrient loads result in decreased habitat quality and diminished biodiversity. Under the latter scenarios catches are high but they predominantly consist of lower-valued fish.

Model uncertainty and simulated multispecies fisheries management advice in the Baltic Sea.

Different ecosystem models often provide contrasting predictions (model uncertainty), which is perceived to be a major challenge impeding their use to support ecosystem-based fisheries management (EBFM). The focus of this manuscript is to examine the extent of model disagreements which could impact management advice for EBFM in the central Baltic Sea. We compare how much three models (EwE, Gadget and a multispecies stock production model) differ in 1) their estimates of fishing mortality rates (Fs) satisfying alternative hypothetical management scenario objectives and 2) the outcomes of those scenarios in terms of performance indicators (spawning stock biomasses, catches, profits). Uncertainty in future environmental conditions affecting fish was taken into account by considering two seal population growth scenarios and two nutrient load scenarios. Differences in the development of the stocks, yields and profits exist among the models but the general patterns are also sufficiently similar to appear promising in the context of strategic fishery advice. Thus, we suggest that disagreements among the ecosystem models will not impede their use for providing strategic advice on how to reach management objectives that go beyond the traditional maximum yield targets and for informing on the potential consequences of pursuing such objectives. This is especially true for scenarios aiming at exploiting forage fish sprat and herring, for which the agreement was the largest among our models. However, the quantitative response to altering fishing pressure differed among models. This was due to the diverse environmental covariates and the different number of trophic relationships and their functional forms considered in the models. This suggests that ecosystem models can be used to provide quantitative advice only after more targeted research is conducted to gain a deeper understanding into the relationship between trophic links and fish population dynamics in the Baltic Sea.

Combined effects of global climate change and regional ecosystem drivers on an exploited marine food web.

Changes in climate, in combination with intensive exploitation of marine resources, have caused large-scale reorganizations in many of the world's marine ecosystems during the past decades. The Baltic Sea in Northern Europe is one of the systems most affected. In addition to being exposed to persistent eutrophication, intensive fishing, and one of the world's fastest rates of warming in the last two decades of the 20th century, accelerated climate change including atmospheric warming and changes in precipitation is projected for this region during the 21st century. Here, we used a new multimodel approach to project how the interaction of climate, nutrient loads, and cod fishing may affect the future of the open Central Baltic Sea food web. Regionally downscaled global climate scenarios were, in combination with three nutrient load scenarios, used to drive an ensemble of three regional biogeochemical models (BGMs). An Ecopath with Ecosim food web model was then forced with the BGM results from different nutrient-climate scenarios in combination with two different cod fishing scenarios. The results showed that regional management is likely to play a major role in determining the future of the Baltic Sea ecosystem. By the end of the 21st century, for example, the combination of intensive cod fishing and high nutrient loads projected a strongly eutrophicated and sprat-dominated ecosystem, whereas low cod fishing in combination with low nutrient loads resulted in a cod-dominated ecosystem with eutrophication levels close to present. Also, nonlinearities were observed in the sensitivity of different trophic groups to nutrient loads or fishing depending on the combination of the two. Finally, many climate variables and species biomasses were projected to levels unseen in the past. Hence, the risk for ecological surprises needs to be addressed, particularly when the results are discussed in the ecosystem-based management context.

Biological ensemble modeling to evaluate potential futures of living marine resources.

Natural resource management requires approaches to understand and handle sources of uncertainty in future responses of complex systems to human activities. Here we present one such approach, the "biological ensemble modeling approach," using the Eastern Baltic cod (Gadus morhua callarias) as an example. The core of the approach is to expose an ensemble of models with different ecological assumptions to climate forcing, using multiple realizations of each climate scenario. We simulated the long-term response of cod to future fishing and climate change in seven ecological models ranging from single-species to food web models. These models were analyzed using the "biological ensemble modeling approach" by which we (1) identified a key ecological mechanism explaining the differences in simulated cod responses between models, (2) disentangled the uncertainty caused by differences in ecological model assumptions from the statistical uncertainty of future climate, and (3) identified results common for the whole model ensemble. Species interactions greatly influenced the simulated response of cod to fishing and climate, as well as the degree to which the statistical uncertainty of climate trajectories carried through to uncertainty of cod responses. Models ignoring the feedback from prey on cod showed large interannual fluctuations in cod dynamics and were more sensitive to the underlying uncertainty of climate forcing than models accounting for such stabilizing predator-prey feedbacks. Yet in all models, intense fishing prevented recovery, and climate change further decreased the cod population. Our study demonstrates how the biological ensemble modeling approach makes it possible to evaluate the relative importance of different sources of uncertainty in future species responses, as well as to seek scientific conclusions and sustainable management solutions robust to uncertainty of food web processes in the face of climate change.

Reconstructing the development of Baltic sea eutrophication 1850-2006.

A comprehensive reconstruction of the Baltic Sea state from 1850 to 2006 is presented: driving forces are reconstructed and the evolution of the hydrography and biogeochemical cycles is simulated using the model BALTSEM. Driven by high resolution atmospheric forcing fields (HiResAFF), BALTSEM reproduces dynamics of salinity, temperature, and maximum ice extent. Nutrient loads have been increasing with a noteworthy acceleration from the 1950s until peak values around 1980 followed by a decrease continuing up to present. BALTSEM shows a delayed response to the massive load increase with most eutrophic conditions occurring only at the end of the simulation. This is accompanied by an intensification of the pelagic cycling driven by a shift from spring to summer primary production. The simulation indicates that no improvement in water quality of the Baltic Sea compared to its present state can be expected from the decrease in nutrient loads in recent decades.

Impact of climate change on ecological quality indicators and biogeochemical fluxes in the Baltic sea: a multi-model ensemble study.

Multi-model ensemble simulations using three coupled physical-biogeochemical models were performed to calculate the combined impact of projected future climate change and plausible nutrient load changes on biogeochemical cycles in the Baltic Sea. Climate projections for 1961-2099 were combined with four nutrient load scenarios ranging from a pessimistic business-as-usual to a more optimistic case following the Helsinki Commission's (HELCOM) Baltic Sea Action Plan (BSAP). The model results suggest that in a future climate, water quality, characterized by ecological quality indicators like winter nutrient, summer bottom oxygen, and annual mean phytoplankton concentrations as well as annual mean Secchi depth (water transparency), will be deteriorated compared to present conditions. In case of nutrient load reductions required by the BSAP, water quality is only slightly improved. Based on the analysis of biogeochemical fluxes, we find that in warmer and more anoxic waters, internal feedbacks could be reinforced. Increased phosphorus fluxes out of the sediments, reduced denitrification efficiency and increased nitrogen fixation may partly counteract nutrient load abatement strategies.

Extremes of temperature, oxygen and blooms in the Baltic sea in a changing climate.

In the future, the Baltic Sea ecosystem will be impacted both by climate change and by riverine and atmospheric nutrient inputs. Multi-model ensemble simulations comprising one IPCC scenario (A1B), two global climate models, two regional climate models, and three Baltic Sea ecosystem models were performed to elucidate the combined effect of climate change and changes in nutrient inputs. This study focuses on the occurrence of extreme events in the projected future climate. Results suggest that the number of days favoring cyanobacteria blooms could increase, anoxic events may become more frequent and last longer, and salinity may tend to decrease. Nutrient load reductions following the Baltic Sea Action Plan can reduce the deterioration of oxygen conditions.

Ecological quality boundary-setting procedures: the Gulf of Riga case study.

Two approaches for setting ecological class boundaries, response curves and a simplified mathematical boundary-setting protocol, were tested for coastal, transitional and open waters in the Gulf of Riga, Baltic Sea. The simplified mathematical boundary-setting protocol defines acceptable ecological status based on expert judgment by a uniform relative deviation from reference conditions. In contrast, response curves derive class boundary definitions from observed changes in biological quality elements along environmental pressure gradients for class boundary definitions. Identification of relevant environmental pressures for the construction of response curves was based on a conceptual model of eutrophication in the Gulf of Riga. Response curves were successfully established for summer chlorophyll a and transparency, as well as for macrozoobenthos abundance in the Central Gulf, macrozoobenthos biotic coefficient in the Southern Gulf, and maximum depth of phytobenthos in the Northern Gulf. In the Gulf of Riga response curves almost always permitted a larger deviation from reference conditions than the 50% deviation applied for the simplified mathematical boundary-setting protocol. The case study clearly demonstrated that class boundary definitions should take into account the sensitivity of the target water body. Also, the class boundaries for different ecological quality elements were internally more consistent than those derived by the simplified mathematical boundary-setting protocol.