Development of Tools for Integrated Monitoring and Assessment of Hazardous Substances and Their Biological Effects in the Baltic Sea

The need to develop biological effects monitoring to facilitate a reliable assessment of hazardous substances has been emphasized in the Baltic Sea Action Plan of the Helsinki Commission. An integrated chemical–biological approach is vitally important for the understanding and proper assessment of anthropogenic pressures and their effects on the Baltic Sea. Such an approach is also necessary for prudent management aiming at safeguarding the sustainable use of ecosystem goods and Services. The BEAST project (Biological Effects of Anthropogenic Chemical Stress: Tools for the Assessment of Ecosystem Health) set out to address this topic within the BONUS Programme. BEAST generated a large amount of quality-assured data on several biological effects parameters (biomarkers) in various marine species in different sub-regions of the Baltic Sea. New indicators (biological response measurement methods) and management tools (integrated indices) with regard to the integrated monitoring approach were suggested.


INTRODUCTION
Marine and coastal ecosystems around the globe are facing the threat of multiple anthropogenic activities causing severe impacts on their health. Among these, the effects of hazardous substances can be observed at all biological levels; consequently, their impacts must be considered when assessing the health status of marine ecosystems (Downs and Ambrose 2001). To increase the significance of impact assessments, there is now a need to bridge the gap between the measured concentrations of chemicals and their biological effects at different biological organization levels. Suitable tools and monitoring strategies have been developed and applied in various sea areas (Allen and Moore 2004;Broeg et al. 2005;Broeg and Lehtonen 2006;Moore et al. 2006;Dagnino et al. 2007;Strand 2007;Davies and Vethaak 2012).
Biological responses measured at individual, cellular, and subcellular level are commonly referred to as biomarkers. As demonstrated in numerous studies, biomarkers have diagnostic power with regard to both exposure to and effects of contaminants (Broeg et al. 1999;Sturve et al. 2005;Lang et al. 2006). Integrated indices and similar approaches (expert systems, environmental prognostics, etc.) based on the measurement of a set of biomarkers have recently been developed and tested in the field. Application of these methods allows for comparisons of the ''health status'' of populations (i.e., an integrated estimate of the performance capability of a population based on measurements of potential dysfunctions or damage at different biological levels, a definition applied for the purpose of this study) inhabiting different locations. Furthermore, a mathematically derived integrated index consisting of chemical, biomarker, and population/community data can facilitate an overall assessment of the ecosystem health of an area, which is obviously also affected by eutrophication, habitat disturbance, overfishing, noise, etc. Such indices have also been shown to be useful to communicate the results of complex interactions to environmental managers and the public (ICES 2007).
The Baltic Sea suffers from a multitude of stressors generated by human activities. Besides the steadily progressing eutrophication and related problems, the threat of ecological effects caused by hazardous substances has not receded, despite the reduced environmental concentrations of some ''classical'' contaminants. In fact, the profile of chemical exposure has shifted to include a much wider range of different types of compounds used by humankind for various purposes (consumer chemicals, pharmaceuticals, plastic additives, surfactants, etc.). This has led to a situation where there is uncertainty of the true environmental risk posed by these new chemical mixtures with steadily elevating concentrations in the marine environment, and in terms of how to detect and assess the significance of the problem. The necessity to include biological effects methods in the monitoring and assessment toolbox has therefore gained weight markedly.
The EU 5FP BEEP project (Biological Effects of Environmental Pollution in Marine Coastal Ecosystems, 2001Ecosystems, -2004 was the first extensive joint effort for the development of tools and approaches for monitoring the biological effects in the Baltic Sea area . Requirements for an integrated monitoring program were identified and considered in relation to the assessment of ecosystem health . The next major international collaboration project BEAST (Biological Effects of Anthropogenic Chemical Stress: Tools for the Assessment of Ecosystem Health) executed under the Baltic Sea BONUS program continued on the main tracks of BEEP by testing and validating the biomarker methodology. However, the focus was now increasingly directed to the development of integrated assessment of chemical pollution and tools contributing to the holistic ecosystem health assessments (Table 1).

BASIC RESEARCH: DEVELOPMENT OF NEW METHODS AND TESTING OF POTENTIAL TARGET SPECIES IN DIFFERENT SUBREGIONS OF THE BALTIC SEA
The basic research component of BEAST focused mainly on field studies regarding biological effects of hazardous substances. Measurements were made at various levels of biological organization in local target organisms considered promising for environmental monitoring purposes. Fourteen field campaigns were performed during 2009-2011 in the Belt Sea, Gulf of Gdansk, G. of Riga, G. of Finland, and G. of Bothnia subregions. Key species representing algae, zooplanktonic and benthic crustaceans, bivalves, gastropods, and fish were collected and analyzed for a wide range of biological effects by applying a variety of novel (Table 2) and established methods. Bioassays using crustacean amphipod species and sediments collected from the subregions as well as analyses of selected chemical contaminants such as polycyclic aromatic hydrocarbons (PAH), organochlorines, organotins, and trace metals from sediments and/or tissues were performed to assess the degree of contamination of the studied areas.
Field transplantation studies using caged blue mussels were carried out in selected coastal areas. In addition to the field studies, a range of specific experimental studies were carried out to investigate combined effects of contaminants and other environmental stressors including hypoxia, salinity, pH, and eutrophication.
Many of the biological effects methods applied in this project have been recommended by, for example, expert groups of the International Council for the Exploration of the Sea (ICES) and The Convention for the Protection of the Marine Environment of the North-East Atlantic (OSPAR) to be applied in the integrated monitoring of contaminants (ICES 2007(ICES , 2011bDavies and Vethaak 2012). However, they were still considered to be in need for more research, e.g., for the establishment of assessment criteria (background and threshold values) in different species. Besides the effects of contaminants, there was a quest for the examination of possible impacts of salinity and other environmental variables on the responses. Some of these methods were eventually selected to compose the set of methods recommended for the CO-RESET project of the Helsinki Commission for the protection of the Baltic Sea (HELCOM) for the strengthening of biological effects monitoring that has been running clearly behind the progress going on in other sea areas of Europe HELCOM 2010).

DEVELOPMENT OF MONITORING: DATABASE, SELECTION OF ENDPOINTS, AND INTEGRATED ASSESSMENT
BEAST worked in close collaboration with HELCOM CORESET (HELCOM 2012a, b) and the ICES Study Group for the Development of Integrated Monitoring and Assessment of Ecosystem Health in the Baltic Sea (SGEH) to develop the application of bioeffect tools in the monitoring and assessment of impact of anthropogenic contaminants on the Baltic Sea ecosystem. This resulted in the preparation of scientific background documents, including Baltic Sea specific assessment criteria for different biomonitoring species and biological effects, and recommendations for biological effects methods to be used as core and candidate indicators for the revised monitoring program of the Baltic Sea and the implementation of the EU Marine Strategy Framework Directive (MSFD) (HELCOM 2012a, b). The main achievements included also the construction of the database ''BonusHAZ,'' jointly with the BONUS BALCOFISH project, and exploring of tools for integrated assessment between contaminant levels, different biomarkers, and/or species.  Dabrowska et al. (2012Dabrowska et al. ( , 2013 Testing and validation of integrated monitoring approaches, indices, and expert systems with regard to their applicability for the Baltic Sea, taking into account the specific biotic and abiotic characteristics of the different subregions and different contaminant burdens

Database
The ''BonusHAZ'' database forms, to date, the most comprehensive data collection on biological effects of hazardous substances in the Baltic Sea region, consist of 60 different parameters for more than 600 single specimens of different fish and invertebrate species in different subregions of the Baltic Sea. As of 2010, BEAST has a formal arrangement with ICES to use their database code lists (RECO), and the ''BonusHAZ'' is planned to become available in public domain at a later stage.

Core Methods and Assessment Criteria
With the new database and along with information from various national initiatives and monitoring activities in the Baltic Sea countries as well as improved scientific understanding of cause-and-effect relationships, sufficient data are now available and used to develop Baltic Sea-specific assessment criteria for biological effects of contaminants. In fact, this is a prerequisite for including any method in a monitoring program. A set of specific biological effects indicators tailored for the Baltic Sea has now been selected, based on knowledge obtained during scientific research in the area, e.g., within the BEEP, BEAST, and BALCOFISH projects, as well as from the work performed in relevant ICES and OSPAR expert groups. The chosen core indicators listed below describe either effects caused by mixtures of contaminants or specific contaminants. They also indicate adverse effects at different biological levels, i.e., molecular/biochemical/cellular (''early warning'') or individual/population (health and reproductive impairments) levels. For these indicators, cause-effect relationships have been studied, and assessment criteria have been established. Furthermore, being important for method of harmonization across the sea areas of Europe, these are the established methods recommended by ICES expert groups, and many have already been included in other regional monitoring programs, e.g., under OSPAR and MEDPOL, or in national monitoring programs (Denmark, Sweden, and/or Germany). The recommended core methods are: • General stress caused by a range of contaminants (''early warning''): lysosomal membrane stability in fish, bivalves, or amphipods; • Effects caused by genotoxic contaminants (''early warning''): induction of micronuclei in fish, bivalves, or amphipods; • Reproductive success impairments caused by a range of contaminants: embryo aberrations in fish (eelpout) or amphipods; • General health status: Fish Disease Index based on externally visible fish diseases, macroscopic liver neoplasms, and liver histopathology.
In addition, two contaminant-specific biological effects indicators, imposex in marine prosobranch gastropods and PAH metabolites in fish, have been included as part of the core indicators for TBT and PAH, respectively. Finally, a few methods have been placed in the candidate list, including intersex or vitellogenin induction in male fish (endocrine disruption), acetylcholinesterase activity (neurotoxicity), and ethoxyresorufin-O-deethylase activity (biotransformation). The future application of biological effects indicators in a combined monitoring strategy strengthens the assessment of contaminants that are approaching critical concentration levels giving rise to pollution effects in the Baltic Sea. Inclusion of biological effects indicators into the HEL-COM Cooperative Monitoring in the Baltic Marine Environment (COMBINE) program in the future also allows for the detection of combined effects of complex mixtures of hazardous substances, which is the norm for most sea areas. In using only contaminant-specific effect methods, such as imposex in prosobranch gastropods or PAH metabolites in fish, a much larger number of indicators Cardiac activity in crabs and mussels Recovery time of heart rate after standardized teststimulus, i.e., experimentally lowered salinity or temperature, was used as an integrated physiological biomarker of general health in organisms for studying pollution effects Cellular immune responses in mussels Biomarkers for immune responses (i.e., total and differential haemocyte count, phagocytic activity and apoptosis) were studied in mussels from areas characterized by variable salinities and different contaminant profiles

Mytilus edulis
Belt Sea Höher et al. (2012) AMBIO 2014, 43:69-81 would have to be applied to cover the effects of all relevant groups of hazardous substances present in the marine environment. In addition, most of them would still be neglected since some of the process-related chemicals released to the Baltic Sea are presently unidentified. Effect indicators with a wider detection spectrum pinpoint the need for further investigations to identify the chemicals causing the effects observed. Besides fulfilling the requirement of the Baltic Sea Action Plan (BSAP), inclusion of core indicators of biological effects into the HEL-COM COMBINE program will be an important step toward harmonization of environmental assessments of different sea areas in Europe; eventually, this will allow for comparisons on larger scales and, thus, implementing important main goals of the MSFD.

Testing of Integrated Approaches
At present, various methodologies are either available or under development to monitor and assess pollution effects and ecosystem health in marine and coastal waters. Integrated indices and similar approaches (e.g., multivariate approaches, expert systems, and decision support systems) based on the measurement of a set of biomarkers have recently been developed and applied in the North Sea, the Northeast Atlantic, and the Mediterranean Sea (Cajaraville et al. 2000;Beliaeff and Burgeot 2002;Moore et al. 2004;Broeg and Lehtonen 2006;Dagnino et al. 2007;Viarengo et al. 2007;Hagger et al. 2009;Marigómez et al. 2013). During BEAST, feasibility of application of these approaches for integrated assessment between contaminant levels, different biomarkers, and/or species was explored (Höher et al. 2012;Dabrowska et al. 2012Dabrowska et al. , 2013Turja et al. in press). These studies showed different patterns in biological responses between the more-and less-contaminated sites, often also with discrimination between effects caused by contaminants and salinity. One example of the data integration methods was the application of the Integrated Biomarker Assessment Tool (IBAT), developed jointly by BEAST and BALCOFISH, and tested on eelpout (Strand et al. unpublished data). The IBAT tool allows for comparisons of input data for the measured biological effect parameters for which assessment criteria have been developed, and an overall Integrated Biomarker Assessment Score (IBAS) can thus be calculated. IBAS includes weighted score values depending on the biological response level of the respective effects, and therefore, considers the biological significance of the effects observed, i.e., subcellular (Score 1), cellular/ tissue (Score 2), or whole organism responses (Score 3). IBAS summarizes all weighted scores by the formula IBAS = (R X)/(sqrt n), which follows the principles of the indicator-based integrative assessment tool CHASE, developed during the HELCOM Integrated Thematic Assessment of Hazardous Substances in the Baltic Sea (HELCOM 2010). CHASE has been used to integrate the status of contamination by individual chemicals and biological effects at specific sites or areas into a single status value termed ''contamination ratio.'' The assessment of the IBAS value follows the ''traffic light'' assessment scale as applied in the CHASE tool, i.e., \1 (low impact), 1 to \5 (moderate impact), and [5 (high impact). Subsequently, IBAT has the potential to be incorporated as an improvement into the CHASE tool for future assessments. IBAS also has the potential to be used for a single-species health index if enough data for relevant biological effect indicators are available, or for several species as an integrative measure for the all the observed biological effects in a particular area.
An example of the use of IBAT is given in Fig. 1, showing the outcome of a study on eelpout in selected coastal areas in Denmark. The IBAS indicates a relatively high impact of pollution in eelpout at two of the stations investigated, whereas at two others, the impact is moderate. Only at one station, a low impact of pollution is indicated. Considering the available information on contaminant levels at the respective stations, the impact levels of pollution in eelpout defined by IBAT are considered feasible.
The integration of several effect biomarkers measured at different levels of biological organization up to the level of reproduction has been a further challenge with respect to incorporating all information into an integrated assessment and to provide a science-based ''traffic light'' score. Thus, IBAT has been developed based on data from all biological effects indicators, and it can be used to perform an overall assessment of risk level for pollution effects at a specific location. The application of the integrated pollution-response model and the use of a decision support system for the ranking of responses from the cellular to the reproduction level in amphipods have proven to be promising approaches (Löf et al. in prep.).
Another example of integrated approaches performed under BEAST is given in Fig. 2. Biological effects data obtained from females of herring collected from all the studied subregions were fed into a simple algorithm set-up, the Integrated Biomarker Response (IBR) (Beliaeff and Burgeot 2002;Broeg and Lehtonen 2006). The results of the IBR approach demonstrate the value of integrated methods in describing the combined environmental stress experienced by organisms at different field sites in relation to each other.

OPERATIONAL AND PRACTICAL ISSUES IN ECOSYSTEM HEALTH ASSESSMENTS: SAMPLING, INTERCALIBRATION, AND QUALITY ASSURANCE
In addition to basic scientific research and method testing, the BEAST project set out to prepare methodological guidelines and standard operating procedures for relevant biological effects techniques, and, in association with that, to organize and execute training and intercalibration workshops and activities.

Guidelines and Standard Operating Procedures
The implementation of guidelines and standard operating procedures is a fundamental quality assurance requirement for monitoring and assessment. Although guidelines for many biological effects techniques already exist (largely developed  (5) used for the index: catalase activity (oxidative stress); acetylcholinesterase activity (neurotoxicity), glutathione S-transferase activity (biotransformation phase II); lysosomal membrane stability (general stress); and histopathology (general health). The dashed line indicates the mean index value for this data by ICES), they often lack coherence and do not sufficiently address the specific environmental conditions of the Baltic Sea. Therefore, it was necessary to develop specific guidelines and standard operating procedures for those biological effects techniques identified by BEAST as being useful for the Baltic Sea. Guidelines and instructions generated were subsequently transferred to draft standard operating procedures, focusing on the core and candidate indicators identified by HELCOM CORESET (Table 3). These standard operating procedures will be published once the results of the present revision of HELCOM's monitoring program with regard to indicators for future monitoring and the structure of the monitoring guidelines become available. It is envisaged that standard operating procedures for other techniques applied and assessed in the BEAST project will be developed at a later stage. Further research is needed to integrate the BEAST draft standard operating procedures with those on chemical measurements that are partly existing (e.g., under HELCOM COMBINE) or are currently under development. The preparation of the guidelines and standard operating procedures can be considered as a major milestone for the implementation of a qualityassured integrated monitoring and assessment program for the Baltic Sea that will meet the goals and ecological objectives of HELCOM BSAP and EU MSFD in relation to hazardous substances.

Training and Intercalibration
Training of personnel and intercalibration of methodologies applied in research and monitoring are essential for their implementation in joint monitoring and assessment programs. Together with the establishment and use of guidelines and standard operating procedures they form the essential components of quality assurance. The importance of quality assurance, especially in international monitoring programs with contributions from several national institutes/laboratories into a common data pool, has been clearly recognized.
Aspects covered during BEAST intercalibration and training exercises included sampling of biotic and abiotic matrices, processing and conservation of samples, on board examination of the health status of organisms, laboratorybased analysis of samples, and data treatment (Table 4). While intercalibration was mainly meant as an internal process (involving BEAST, and, partly, BALCOFISH partners) to improve the comparability of results, many of the workshops arranged were open to both project partners and external interested colleagues, e.g., the intercalibration of PAH metabolite measurement in fish bile attracted several non-BEAST specialist laboratories from the OSPAR area (Kammann et al. 2013). Determination of acetylcholinesterase activity in fish and bivalves Lehtonen K. and Gercken J.
Determination of EROD activity in fish Vuorinen P., Tuvikene A., Dabrowska H., and Lang T. Overall, the BEAST training/intercalibration activities contributed to the enhancing of quality assurance, capacity building, and networking in the Baltic Sea region. It is emphasized that internal and external training and intercalibration have to be understood as dynamic processes that need to be carried out on a continuous basis as an important part of marine monitoring and assessment. It is, thus, essential that the revised HELCOM COMBINE program (as well as the future monitoring under the MSFD) encompasses an operative quality assurance component.

CONCLUSIONS AND FUTURE CHALLENGES
The research activities carried out in the BEAST project were focused on the further development of existing biological effects techniques to be applied in Baltic Sea subregions. The results show that hazardous substances are currently causing a wide variety of biological responses in the key species selected and must therefore be regarded as a continuous threat to the health of the Baltic Sea ecosystem. Based on these results, BEAST identified a set of core and candidate biological effects techniques (indicators) for future monitoring of hazardous substances and their effects, and developed and tested integrated assessment strategies, including numeric criteria for the assessment of the environmental status of the Baltic Sea. Furthermore, an extensive database, comprising conceptual and methodological background documents as well as quality assurance components, was developed, all of which are required in the implementation processes concerning integrated monitoring at the national and international levels. Recommendations made by BEAST were taken up by the HELCOM CORESET project and international expert fora, and their implementation in respect of national monitoring is under discussion at present. However, despite the progress achieved, additional work still remains to be done: • Some biological effects techniques regarded as promising (e.g., the candidate indicators above) still need to be further developed and validated. • A stronger linkage between measurements of contaminants and biological effects at various levels of biological organization needs to be achieved to facilitate integrated monitoring and assessment. This will require new approaches, addressing, e.g., source identification, means of discharge to the sea, chemical behavior in the environment, biotic pathways and responses (biological degradation, biomagnification, bioaccumulation, and toxicity), and fate (chemical degradation, deposition). • An improvement of current assessment concepts and criteria is required to be able to tackle ecologically relevant problems such as toxicity of chemical mixtures, multiple stress, and multilevel effects. This can be reached by applying and evaluating new methods and model approaches (e.g., passive sampling, integrated biological effects methods, distribution and fate models, decision trees, substance flow analyses, etc.) and establishing cause-effect links between different levels of biological organization by dedicated laboratory and field studies with ecologically relevant target species. Integrated indicators and indices as well as AC should still be developed.
• The assessment and prediction of ecological and socioeconomic consequence of hazardous substances should be improved. Databases and models from BEAST and earlier projects can be exploited to create scenarios for biological effects. These will supply important information for socioeconomic analyses of changes, e.g., in fish populations, remediation measures of contaminated sediment and waste water-treatment plant effluents, and relevance for human food quality. Evaluations of the cost-efficiency of different monitoring and assessment methods, and the benefits of policy implementation and abatement strategies, as well as exploration of the efficiency of reduction measures with regard to impacts observed in the marine environment are urgent tasks.
Finally, although the BEAST project contributed to the international networking of institutions in charge of monitoring and assessment of hazardous substances in the Baltic Sea, there is still a requirement for enhanced international coordination and harmonization; especially in the light of the requirements of the EU MSFD concerning regional cooperation, this will be a challenging task to accomplish in the coming years.