Integrated Environmental Assessment and Management — Volume 13, Number 2—pp. 376–386 376
Received: 6 March 2016
|
Revised: 28 March 2016
|
Accepted: 14 September 2016
Health & Ecological Risk Assessment
Toward a Conceptual Approach for Assessing Risks from Chemical Mixtures and Other Stressors to Coastal Ecosystem Services Kristian Syberg,*y Thomas Backhaus,z Gary Banta,y Peter Bruce,§ Mikael Gustavsson,z € mo € ,§ Henriette Selck,y and Jonas S Gunnarsson§ Wayne R Munns Jr,k Robert Ra yDepartment of Science and Environment, Roskilde University, Roskilde, Denmark zDepartment of Biological and Environmental Sciences, University of Gothenburg, Gothenburg, Sweden §Department of Ecology, Environment and Plant Sciences, Stockholm University, Stockholm, Sweden kAtlantic Ecology Division, US Environmental Protection Agency, Narragansett, Rhode Island
EDITOR’S NOTE: Roskilde University in Denmark hosted the international workshop “Environmental Risk – Assessing and Managing Multiple Risks in a Changing World,” November 16-17, 2015, as part of its annual ‘SUNRISE’ series of conferences and workshops that feature groundbreaking science. The goal of this workshop was to develop a holistic perspective for assessing and managing risks from the multiple stressors and natural hazards that impact ecosystems and the humans who rely on them. Such a perspective is critical as—in a finite world with limited resources—it is paramount that major, multiple risks be appropriately addressed. All 3 articles from the workshop, including the overarching Focus Article in Environmental Toxicology and Chemistry, are available on the IEAM web site in the virtual issue “SUNRISE Environmental Risk Workshop, Roskilde 2015.“
ABSTRACT Growth of human populations and increased human activity, particularly in coastal areas, increase pressure on coastal ecosystems and the ecosystem services (ES) they provide. As a means toward being able to assess the impact of multiple stressors on ES, in the present study we propose an 8-step conceptual approach for assessing effects of chemical mixtures and other stressors on ES in coastal areas: step A, identify the relevant problems and policy aims; step B, identify temporal and spatial boundaries; step C, identify relevant ES; step D, identify relevant stressors (e.g., chemicals); step E, translate impacts into ES units; step F, assess cumulative risk in ES units; step G, rank stressors based on their contribution to adverse effects on ES; and step H, implement regulation and management as appropriate and necessary. Two illustrative case studies (Swedish coastal waters and a coastal lagoon in Costa Rica) are provided; one focuses on chemicals that affect human food supply and the other addresses pesticide runoff and trade-offs among ES. The 2 cases are used to highlight challenges of such risk assessments, including use of standardized versus ES-relevant test species, data completeness, and trade-offs among ES. Lessons learned from the 2 case studies are discussed in relation to environmental risk assessment and management of C 2016 SETAC chemical mixtures. Integr Environ Assess Manag 2017;13:376–386. Keywords: Environmental risk assessment Multiple stressors Coastal areas Ecosystem services
INTRODUCTION Ecosystem services (ES), such as food production and nutrient cycling, are essential for human well-being and economic wealth (Costanza et al. 1997). However, ecosystems are facing increasing anthropogenic pressures from human population growth, overexploitation of resources, This article includes online-only Supplemental Data. * Address correspondence to
[email protected] Published 23 September 2016 on wileyonlinelibrary.com/journal/ieam.
Integr Environ Assess Manag 2017:376–386
Chemical contamination
environmental pollution, habitat change, introduced species, and climate change (Lotze et al. 2006). Current environmental assessment frameworks address either individual stressors (e.g., REACH for chemicals) (European Commission [EC] 2006) or struggle to provide the information needed for the ecosystem-based management they are intended to support (e.g., Marine Strategy Framework Directive) (Soma et al. 2015). ES has been suggested as a common currency for assessing the impact of chemical pollution and other stressors in environmental risk assessments (ERAs) (Munns et al.
DOI: 10.1002/ieam.1849
C 2016 SETAC
Risk of Chemical Mixtures to Ecosystem Services—Integr Environ Assess Manag 13, 2017
2009; Apitz 2013; Selck et al. 2016; Munns, Poulsen et al. 2016). Such a common currency enables integration and comparison of the impact of different types of stressors on ecosystems and may thus serve as a means for managers to assess risk more holistically and to help prioritize regulatory measures and management actions. As a means toward developing models to assess impact of multiple stressors on ES, in the present study we propose a conceptual approach to assess the impacts of stressors including chemical mixtures on ES. We apply the model to 2 coastal ecosystem case studies, one in Sweden and the other in Costa Rica. Although these 2 case studies involve only chemical contaminants, the approach is not restricted to chemical mixtures; it is equally applicable to these and other stressors such as eutrophication, invasive species, habitat change or loss, and climate change. Based on the conceptual model and the 2 cases used, we furthermore discuss aspects and challenges to be considered both regarding development of assessment methods and in conducting future assessments of impacts on coastal ecosystems. Ecosystem services and the proposed conceptual approach The ES concept offers a framework for linking vital ecosystem functions to associated benefits they provide to the environment and humans (Schr€ oter et al. 2014). Given that ecosystems are generally exposed to a range of stressors, a key challenge is to identify ecological endpoints that are suitable for assessment of chemical mixtures as well as other stressors, and how these can be implemented in regulatory and management frameworks. A more holistic approach to management that focuses on ecosystems (i.e., ES) rather than individual receptors or endpoints will provide a better foundation for ERA (Munns et al. 2009; Apitz 2013; Munns, Rea et al. 2016). A
377
challenge for such holistic ERAs is, however, that using one ES may lead to a negative impact on another ES, and such antagonistic processes can be difficult to compare because they are often assessed on different scales and with different quantification metrics; for example, a pesticide used to increase crop yield might negatively impact key species in adjacent ecosystems (e.g., streams receiving runoff). We propose a stepwise conceptual approach (Figure 1) as an approach toward: 1) assessing the impacts of multiple stressors on ES, explicitly illustrated with 2 cases concerning chemical mixtures; and 2) providing information necessary for managing those stressors. This process involves 8 steps denoted A through H: Step A: The 1st step involves identifying the problems to be addressed in the study—that is, framing the problem within an ecological context as well as identifying the potential ES that may be affected. Policy aims and possible regulatory measures also need to be identified. Step B: Identify temporal and spatial boundaries; this step aims to identify the geographical area of interest and the relevant timeframe for an assessment (e.g., from a single release in a small oil spill, to seasonal contaminants such as runoff of plant-protection products, to continuous effluent from industries or households, to global persistent organic pollutants). Apart from knowledge about exposure pathways and stressors of concern, pragmatic aspects such as data from existing monitoring programs can frame spatial and temporal boundaries. Step C: Identify relevant ES; determine which ES may be impacted in the geographical area of interest during
Figure 1. Steps needed to frame the environmental impact of chemical mixtures on an ecosystem services (ES) scale in a regulatory and management context. Step A: Identify relevant problems and policy aims. Step B: Determine the boundaries for the assessment both spatially and temporally. Step C: Identify relevant ES. Step D: Identify relevant stressors (in this case, chemicals of potential concern, but not limited to these). Step E: Translate stressor impacts into ES units by identifying relevant endpoints. Step F: Assess the risk, based on ES units, thus enabling step G. Step G: Rank the impact of the different stressors for step H. Step H: Implement regulation and management.
Integr Environ Assess Manag 2017:376–386
wileyonlinelibrary.com/journal/ieam
C 2016 SETAC
378
Integr Environ Assess Manag 13, 2017—K Syberg et al.
the relevant timeframe. The overall aim is to preserve the structure and function of an exposed ecosystem rather than setting thresholds for specific chemicals of concern, and as such, the vital ES should be identified early in the process with the involvement of communities of interest (e.g., industry, nongovernmental organizations, environmental managers; see Selck et al. 2016). Step D: Identify relevant stressors including chemicals that impact the ES of concern. The relevant stressors will vary depending on the ES in focus, which means that conceptual models linking stressors to effects on ES are critically important in this step. Step E: Translate stressor-related impacts into ES units (stressor to ES translation in general is discussed in Selck et al. 2016). Step F: Assess cumulative risk in ES units. Step G: Rank the stressors based on their contributions to adverse effects on ES. This step supports the part of the management process where priority is assigned to different stressors. Step H: Implement regulation and management. In close collaboration with communities of interest, determine and implement regulation and management appropriate for the case as presented. In the present study, we illustrate some of the aspects and dilemmas that researchers may encounter when conducting assessments of impacts on ES, through the application of the conceptual approach to 2 case studies. The 1st case study shows how fish chemical body-burden data can be used to assess impacts from chemical mixtures on food provisioning (i.e., on food safety), a key ES in coastal ecosystems. The 2nd case study concerns a watershed with banana plantations in Costa Rica. Rather than quantifying impacts on a specific ES, this 2nd case illustrates how an ES-driven assessment can identify trade-offs among ES. In addition to representing a range of coastal systems, the 2 case studies differ in decision context: retrospective assessment of complex chemical burden for the ES food provision in the Swedish case, and management of active pesticide use in the Costa Rican case. The 2 case studies discussed in the present study are previous studies not originally designed with the proposed approach in mind. They are used in the present study as examples to illustrate aspects of the suggested approach as well as to pinpoint future directions for method development in assessment of multiple stressors to ES. The case studies are presented with sufficient detail to illustrate how the approach would work, whereas full risk assessments (chemical and other stressor risks) in the 2 coastal areas are beyond the scope of the present study. In the long run, an assessment designed initially to incorporate such a conceptual approach would identify key ES explicitly and include a more holistic appraisal of relevant stressors on those ES. This would include nonchemical stressors, such as eutrophication and changes
Integr Environ Assess Manag 2017:376–386
to the physical environment (e.g., impact of dams or removal of vegetation in streams). In the present study, the case studies involving chemical mixtures serve as examples of how the effects of multiple stressors may be quantified and integrated into a common currency (i.e., ES). Such assessment should then serve as a means for managers to simplify the process of assessing risks and prioritizing actions in both regulatory and management contexts. Ecosystem services in coastal areas Coastal areas, defined geographically as the region from the coast to the continental shelf at a depth of 200 m (Martnez et al. 2007), comprise important ecosystems, such as seagrass meadows, estuaries, sandflats and mudflats, salt marshes, and mangroves. Throughout history, coastal areas have provided human societies with important ES such as food (e.g., fish and shellfish harvests, production of algae). Coastal communities have also benefited from a range of less obvious coastal ecosystem functions, such as protection from flooding by mangrove forests and salt marshes, and the natural water treatment in estuaries through organic matter degradation and nutrient cycling (Barbier et al. 2011). Each ecosystem component may provide multiple services: Transitional waters such as estuaries and coastal lagoons are highly productive mixing zones between freshwater and seawater. Many important processes occur where terrestrial organic matter and nutrients are processed and retained, aided by the hydrological conditions of estuarine circulation. These processes support high levels of productivity (including commercially relevant species) and nutrient transformation and removal, through denitrification and other processes (Piehler and Smyth 2011).
CHEMICAL IMPACTS ON COASTAL ECOSYSTEMS The chemicals that impact coastal ecosystems can generally be divided into 2 groups: chemicals that resident species are exposed to with temporally variable concentrations (e.g., pesticides in runoff water during rainfall) and chemicals that persist in the environment for a prolonged period and cause chronic exposure, for example, persistent organic pollutants such as PCBs. To identify which chemicals pose a risk to a specific coastal ES, it is necessary to assess the physicochemical properties of chemicals. Properties such as their hydrophobicity and persistence determine their fate. The sources, such as atmospheric deposition, runoff from land, effluents from wastewater treatment plants (household and industry) and direct discharges, are equally important. Monitoring programs that include the sediment compartment provide a good basis for assessing chemical legacies (historic emissions), as well as both current and emerging chemicals, and temporal changes in loading to a coastal area. Chemicals characterized as hydrophobic compounds (i.e., log Kow > 5) are expected to persist longer and bioaccumulate more in aquatic organisms than water-soluble chemicals. In contrast, water-soluble compounds have a higher potential to be transported away from the coastal zone. Hydrophobic
DOI: 10.1002/ieam.1849
C 2016 SETAC
Risk of Chemical Mixtures to Ecosystem Services—Integr Environ Assess Manag 13, 2017
379
Table 1. List of ecosystem services and ecological parameters supporting ecosystem services, in the Madre de Dios watershed, which are impacted by pesticide mixtures from surrounding banana plantations Ecosystem services
Supporting parameters
Examples
Food production
Viable fish stocks Species diversity Macrophyte cover for fish spawning Benthic invertebrates Primary production
The production of fish and clean drinking water
Cultural services
Primary production Species diversity
The aesthetic and scientific values of the ecosystem
Ecotourism
Viable fish stocks Species diversity
Recreational fishing and swimming
Habitat maintenance
Viability of gene pools Macrophyte cover for fish spawning
The ability of the ecosystem to serve as a nursery and transport route for organisms and nutrients
compounds will concentrate at sediment depositional sites (sites with lower water velocity and/or higher sediment organic matter concentration). Because of these characteristics, chemicals that are persistent, bioaccumulative, and toxic are of particular concern (EC 2006). A key challenge for ERA and management is the presence of mixtures rather than single compounds in the environment (Munaron et al. 2012; Sobek et al. 2016). In a recent review, 132 of 4445 chemicals were identified as potential persistent, bioaccumulative, and toxic (PBT)/very persistent very bioaccumulative (vPvB) substances (B€ ohnhardt 2013). We identified 70 of these as specific individual compounds (i.e., having a Chemical Abstracts Service number). A total of 1175 papers have been published concerning these 70 chemicals, only 13 of which address mixture toxicity (Supplemental Data Table 1). This illustrates a general lack of knowledge of mixtures. The data illustrate a general lack of knowledge of these mixtures of concern. It is therefore highly relevant to establish an approach with which ecosystem impacts of these mixtures can be assessed, preferably in a holistic context. In the absence of experimental data, the effects of mixtures can be initially screened using component-based mixture models that apply concentration addition and/or response addition (sometimes also termed independent action) models. The concentration addition model uses data from single substances to derive risk values for individual chemicals (Altenburger et al. 2000; Kortenkamp et al. 2009; Syberg et al. 2009; Backhaus and Faust 2012; Syberg and Hansen 2016) and can be combined with the response addition model in a 2-step model for risk assessment of more complex mixtures (de Zwart and Posthuma 2005). However, these models can only be applied when the mixture components are defined, that is, if concentrations of the relevant chemicals are known and can be related to individual toxicity estimates. Thus, these models rely on a single-substance assessment for each mixture component. The application of either mixture model further assumes that the toxic mode of action of all relevant chemicals are known, and incorrect determination of chemicals’ modes of action may lead to underestimated toxicity risks if the response Integr Environ Assess Manag 2017:376–386
addition model is applied. This is the reason why the concentration addition model has been recommended as the best default model for ERA (Greco et al. 1995; Kortenkamp et al. 2009). Case study: Swedish coastal waters Data used in this case study are part of an ongoing biomonitoring effort along the Swedish west coast analyzing the body burden of chemicals in eelpout (Zoarces viviparous) in 3 different fjord ecosystems. More information can be found in Bohuskustens Vattenvårdsf€ orbunds (BVVF 2015). We link these body-burden data to the potential risk to a specific ES, food provision, to provide an example of how data from chemical monitoring programs can be used to assess chemical risk to an essential ES. The following subheadings correspond to steps A through H as shown in Figure 1. Step A: Identify problems and policy aims. Fish exposed to low levels of bioaccumulating compound may contain tissue concentrations of contaminants that influence the suitability of caught fish as a food source for humans. Thus, chemical contamination along the Swedish west coast is a potential problem for food safety. Because the European Marine Strategy Framework Directive descriptor 9 states, “Contaminants in fish and other seafood for human consumption do not exceed levels established by Community legislation or other relevant standards” (EC 2008), the policy aim clearly is that food safety is not impaired in the area. Step B: Identify temporal and spatial boundaries. Three different fjord ecosystems provide the spatial boundaries: Kosterfjord (58˚520 77.500 N 11˚070 31.900 E), Brofjord (58˚ 210 83.900 N 11˚260 66.600 E), and Danafjord (57˚400 27.300 N 11˚ 400 35.900 E). Data published by Bohuskustens Vattenvårdsf€ orbunds Kontrollprogram (BVVF 2015) were provided to us by Gothenburg’s Region Association of Local Authorities in Excel format (Supplemental Data Table 2). The temporal boundary for the assessment is the year of the monitoring, but the boundary of the ecosystem preservation goes beyond a single year. Because the chemicals included in
wileyonlinelibrary.com/journal/ieam
C 2016 SETAC
380
Integr Environ Assess Manag 13, 2017—K Syberg et al. Table 2. Contribution of each pesticide to total cumulative risk in msPAF
Type Fungicides
Herbicides
Insecticides
Pesticide Name
Primary Producers
Algae
Aquatic Plants
Fish and Arthropods
Arthropods
Fish
Chlorothalonil
0.1%
0.0%
0.1%
0.0%
0.0%
0.0%
Difenoconazole
5.9%
9.1%
—
6.8%
—
67.8%
Propiconazole
0.0%
0.1%
—
0.0%
0.0%
0.0%
Azoxystrobin
3.2%
6.8%
—
0.0%
0.0%
0.0%
Butachlor
0.2%
0.2%
—
0.0%
0.0%
0.0%
Ametryn
52.8%
60.2%
—
0.0%
—
0.0%
Terbutryn
0.0%
0.1%
—
0.0%
—
0.0%
Oxyfluorfen
1.4%
1.3%
—
0.0%
—
—
Diuron
36.2%
20.7%
99.9%
0.0%
0.0%
0.0%
Hexazinone
0.0%
—
—
0.0%
—
0.0%
Bromacil
0.1%
1.4%
—
0.0%
—
—
Ethoprophos
—
—
—
9.4%
36.0%
16.1%
Terbufos
—
—
—
0.7%
2.5%
0.3%
Carbaryl
0.0%
0.0%
—
4.3%
7.8%
0.0%
Chlorpyrifos
0.0%
0.0%
—
48.3%
35.3%
15.7%
Diazinon
0.0%
0.0%
—
30.5%
18.4%
0.0%
Pesticides required to explain >90% of cumulative Potentially affected fraction (PAF) to a species group are in boldface. Pesticides without noteworthy contribution (
