Salmon and Complexity: Challenges to Assessment

Salmon in the U.S. Pacific Northwest are in widespread decline despite countless environmental assessment studies and billions of dollars spent. Having been involved in environmental assessment for more than three decades, I am forced to conclude that this decline tells us that our established practices of assessment and management are fundamentally deficient. Rather than studying the salmon, we should examine our own practices. These practices presume that, if individual actions are found to be beneficial through analytical assessments, the cumulative outcomes of many actions will also be beneficial. This “linear” presumption is embedded in institutions, analytical methods, and assessment practices. For a whole class of emerging problems, including declining salmon, this presumption is fundamentally wrong. Declining salmon provide a warning that our own analytical habits of thought and notions of progress are leading to outcomes that are both destructive and contrary to our best intentions. This paper is a response to this warning.     

Ecological Risk Assessment and Ecological Epidemiology for Contaminated Sites

After 20 years of development, ecological risk assessment is widely accepted. However, it is evolving in response to a variety of technical and societal pressures. First, pressure for greater simplicity and standardization arise from the expectation that risk assessments should require little time and resources but be defensible. Second, the advance of the environmental sciences and increasing awareness of the complexity of ecological responses generate pressure for greater realism. Third, the dominance of human health risk assessment generates a pressure to integrate ecological risk assessment with that dominant field. Fourth, the demand for cost-benefit analysis creates pressure for integration with environmental economics. Finally, the need to connect the practice of ecological epidemiology with risk-based decision-making creates a pressure of the formation of a single integrated ecological assessment practice.     

Restoring Wild Salmon to the Pacific Northwest: Framing the Risk Question

In the Pacific Northwest of the United States, it is urgent to assess accurately the various options proposed to restore wild salmon. For the past 150 years, a variety of analytic approaches have been employed to assess the ecological consequences of salmon management options. Each approach provided useful information to decision makers, but each also suffered from limitations, some relatively minor, others sufficient to undermine any potential utility. Risk assessment has become the most widely used analytic approach to evaluate environmental policy options. To date its use in ecological policy has been largely constrained to evaluating relatively simple technical questions (e.g., regulatory actions associated with specific chemicals or hazardous waste sites). Recently, however, there has been interest in applying risk assessment to more complex ecological policy problems (e.g., the decline of wild salmon in the Pacific Northwest). Although its use has become commonplace and widely accepted, especially among regulatory and land management agencies, risk assessment remains contentious. The most heated debates revolve around delineating the specific meaning of risk; that is, framing the risk “question” to be answered.     

Issues Related to Chemical Analysis,Data Reporting,and Use: Implications for Human Health Risk Assessment of PCBs and PBDEs in Fish Tissue

Recent reports in the scientific literature and the media, related to elevated levels of polychlorinated biphenyls (PCBs) and polybrominated diethyl ethers (PBDEs) in farmed and wild salmon have had significant impacts on public opinion and consumer behavior, influencing the sales of farmed salmon in North America and Europe. The assessment of contaminants in fatty fish, an important source of omega-3 fatty acids, is therefore an exercise in balancing risks and benefits. Human health risk assessors and risk managers will benefit from an understanding of the level of uncertainty that is integrated into all aspects of evaluating risk in this context. Significant variability exists in the way in which analyses are conducted, how data are reported, and how they are used in risk assessments. We conducted an analytical review of PCB and PBDE data in farmed and wild salmon, and identified critical issues having implications on human health risk assessment from fish consumption. These issues include: analytical methodologies used, quantification issues, reporting of QA/QC information, tissue sampling, nature of tissue analyzed, and laboratory competence. This article reviews and outlines these issues, discusses their implications for human health risk assessment, and recommends the consistent application of analytical fish tissue data in human health risk assessment.     

The Ecological Component of Ecological Risk Assessment: Lessons from a Field Experiment

To improve ecological relevance, regulatory agencies are promoting assessments of effects at higher levels of organization, an objective that requires an understanding of current ecological theories. One such theory, hierarchy theory, contends that the effects of a disturbance acting at one level of organization (e.g., population) are not, as a rule, transmitted to higher levels of organization (e.g., community). Conversely, effects at higher levels of organization only occur if lower level variables have been affected. Further, responses to disturbance depend on disturbance history. In this study, I determined the effects of a disturbance treatment at the population, guild, and community levels of organization for vegetation in five wetlands with a disturbance history ranging from highly to rarely disturbed. The 2-year field experiment revealed that the effects of the disturbance treatment were most strongly felt at the population level of organization in wetlands without a history of disturbance. These observed impacts took place against a backdrop of constant change. Thus, the eventual disappearance of treatment effects was not due to a return to the pre-treatment state, but rather a return to a trajectory similar to that exhibited by the control plots. The implications of these results for ecological risk assessment are: (1) the observed effects of a stressor in a system cannot be extrapolated to other systems unless they have similar disturbance histories, (2) detecting effects before they become serious requires monitoring at lower levels of organization, (3) recovery to a naturally innate state is not a viable concept, and (4) the traditional approach of using one post-treatment measurement to determine if reference and impact sites differ is of very questionable value.     

Ecological Risk Assessment for Terrestrial Ecosystems: The Summary of Discussions and Recommendations from the Adelaide Workshop (April 2004)

The Ecological Risk Assessment (ERA) workshop focused on the anthropogenic contaminants in the terrestrial environment, addressing various components of the ERA process. These included sources, exposure pathways, bioavailability, and toxicity to receptor organisms as well as risk communication. It was concluded that although the overseas experience on ERA for terrestrial ecosystems (e.g., International Standards or guidelines from the European Union and the United States) is very useful, it is not directly applicable to Australia due to the differences in receptor organisms, contaminants, soil, and environmental conditions. Workshop discussions stressed on the need for making ERA locally relevant (in terms of choice of receptor organisms, sampling strategy, and bioavailability considerations). The workshop discussions highlighted the need of better appreciation of both the similarities and the clear differences between aquatic and terrestrial ERAs. The availability of reliable data, problems with databases, estimation of bioavailability, and extrapolations from laboratory to field were noted among the key limitations. ERA—being inherently complex and involving a range of environmental compartments, target receptor, and exposure pathways—at a minimum requires a multidisciplinary approach to address the complexities. Bringing a multidisciplinary team together often proves a major challenge in ERA. The delegates called for continued efforts in this area and formation of a network or working group in Australia.     

Proposed Local Ecological Impact Categories and Indicators for Life Cycle Assessment of Aquaculture

In this study we discuss impact categories and indicators to incorporate local ecological impacts into life cycle assessment (LCA) for aquaculture. We focus on the production stages of salmon farming—freshwater hatcheries used to produce smolts and marine grow‐out sites using open netpens. Specifically, we propose two impact categories: impacts of nutrient release and impacts on biodiversity. Proposed indicators for impacts of nutrient release are (1) the area altered by farm waste, (2) changes in nutrient concentration in the water column, (3) the percent of carrying capacity reached, (4) the percent of total anthropogenic nutrient release, and (5) release of wastes into freshwater. Proposed indicators for impacts on biodiversity are (1) the number of escaped salmon, (2) the number of reported disease outbreaks, (3) parasite abundance on farms, and (4) the percent reduction in wild salmon survival. For each proposed indicator, an example of how the indicator could be estimated is given and the strengths and weaknesses of that indicator are discussed. We propose that including local environmental impacts as well as global‐scale ones in LCA allows us to better identify potential trade‐offs, where actions that are beneficial at one scale are harmful at another, and synchronicities, where actions have desirable or undesirable effects at both spatial scales. We also discuss the potential applicability of meta‐analytic statistical techniques to LCA.     

Ecological Stability: Reality,Misconceptions, and Implications for Risk Assessment

Over a long time frame, an ecological system may not exhibit constancy due to successional and evolutionary changes in the species composing the system. However, over shorter time frames an ecological system exhibits a certain degree of constancy (i.e., varies within defined bounds). Traditionally, ecologists considered this short-term constancy to reflect a “balance of nature,” which was viewed akin to the simple homeostatic dynamics of physiological systems. This is an appealing perspective because the disruption of the system's “balance” (i.e., its ”health“) can be ascertained by comparing the system's current state after the imposition of a perturbation with the societally desired state (i.e., baseline). Recently, ecologists have started to develop a much more complex, and perhaps more realistic, perspective regarding ecosystem dynamics, which does not depend upon homeostasis with a single baseline state. This new view includes stochastic variation, nonlinear dynamics and alternative states, and poses a challenge for assessing environmental “health” and the risk of creating “unhealthy” ecological systems     

Public Health Risk Assessment Linked to Climatic and Ecological Change

Disturbances of climatic and ecological systems can present risks to human health, which are becoming more evident from health studies linked to climate variability, landuse change and global climate change. Waterborne disease agents, such as Giardia cysts and Cryposporidium oocysts have been positively correlated with rainfall. El Niño-related extreme weather conditions can have a significant impact on vector- and water-borne diseases. The linkages between weather, terrestrial ecology and human health have been discovered for some diseases, such as rodent-borne hantavirus. Marine ecology also plays a role in determining human health risks, such as from cholera, and other enteric pathogens. Deforestation and ensuing changes in landuse, human settlement, commercial development, road construction, and water control systems singly, and in combination have been accompanied by increases in or emergence of diseases like malaria and schistosomiasis in some regions of the world. Long-term climate change may increase the frequency of heat waves and potentially air pollution episodes, increase the number of extreme weather events, cause coastal flooding and salination of fresh water aquifers, and displace coastal settlements. Ultimately, a two-pronged approach (empirical and modeling studies) is required to better understand these linkages between climato-logical and ecological change as determinants of disease.     

Toxicological Considerations of Contaminants in the Terrestrial Environment for Ecological Risk Assessment

The terrestrial environment acts as a “sink” for contaminants that have been purposely or accidentally released into the environment. Science and policy that support protective measures for terrestrial ecosystems have run behind those of aquatic toxicology and water quality concerns. As a result ecological risk assessment (ERA) involving terrestrial environments tends to be conducted at a simplistic level, relying on numeric targets (soil quality criteria) as a basis for decision-making. However, soil criteria for ecological receptors are somewhat deficient in terms of the numbers available and the data that supports these numbers. Direct toxicity assessments (DTA) for terrestrial environments, such as those used for water quality evaluations, can provide additional useful information about the toxicity and bioavailability of mixtures of contaminants present in soils. This article outlines the approaches used for assessing the toxicity of soil contaminants in terrestrial environments and critiques their advantages and pitfalls.     

Terrestrial Metals Bioavailability: A Comprehensive Review and Literature-Derived Decision Rule for Ecological Risk Assessment

Interstudy variation among bioavailability studies is a primary deterrent to a universal methodology to assess metals bioavailability to soil-dwelling organisms and is largely the result of specific experimental conditions unique to independent studies. Accordingly, two datasets were established from relevant literature; one includes data from studies related to bioaccumulation (total obs = 520), while the other contains data from studies related to toxicity (total obs = 1264). Experimental factors that affected toxicity and bioaccumulation independent of the effect of soil chemical/physical properties were statistically apportioned from the variation attributed to soil chemical/physical properties for both datasets using a linear mixed model. Residual bioaccumulation data were then used to develop a non-parametric regression tree whereby bootstrap and cross-validation techniques were used to internally validate the resulting decision rule. A similar approach was employed with the toxicity dataset as an independent external validation. A validated decision rule is presented as a quantitative assessment tool that characterizes typical aerobic soils in terms of their potential to sequester common divalent cationic metal contaminants and mitigate their bioavailability to soil-dwelling biota.     

Risk Assessment of Tropospheric Ozone: Human Health,Natural Resources,and Ecology

Ozone is an unusual trace gas in the atmosphere, presenting a challenge for risk assessors and risk managers. The challenge can be traced to the gas’ complex chemistry in the atmosphere (exposure), toxicology in biological systems (response), and the fledgling enterprise of risk assessment for widely distributed, highly reactive pollutants. This paper addresses the (i) co-evolution of the scientific data underlying ozone risk assessment on human health, natural resources (crops and managed forests), and unmanaged ecosystems, (ii) similarities and differences in risk assessment among these receptors, and (iii) utility of indicators in risk assessment. The scientific community has developed a sound database to underpin the ozone risk assessment, although the breadth and depth differ markedly among the three receptors. There are similarities in ozone risk assessment among human health, natural resources, and ecology, including features of exposure (e.g., temporal variation), response of plants and humans (e.g., sensitive cohorts), and integration of exposure and response (e.g., importance of peak and cumulative exposures). Equally important are the notable differences, and the more prominent are scaling of exposure-response relationships, air quality monitoring, economic valuation, and models to complement more traditional experimental approaches. Of the three receptors, the status of indicators for conducting ecological ozone risk assessment is the weakest.     

Microbial Endpoints: The Rationale for their Exclusion as Ecological Assessment Endpoints

The functional importance of bacteria and fungi in terrestrial systems is recognized widely. However, microbial population, community, and functional measurement endpoints change rapidly and across very short spatial scales. Measurement endpoints of microbes tend to be highly responsive to typical fluxes of temperature, moisture, oxygen, and many other noncontaminant factors. Functional redundancy across broad taxonomic groups enables wild swings in community composition without remarkable change in rates of decomposition or community respiration. Consequently, it is exceedingly difficult to relate specific microbial activities with indications of adverse and unacceptable environmental conditions. Moreover, changes in microbial processes do not necessarily result in consequences to plant and animal populations or communities, which in the end are the resources most commonly identified as those to be protected. Therefore, unless more definitive linkages are made between specific microbial effects and an adverse condition for typical assessment endpoint species, microbial endpoints will continue to have limited use in risk assessments; they will not drive the process as primary assessment endpoints.     

An Integrated Case Study for Evaluating the Impacts of an Oil Refinery Effluent on Aquatic Biota in the Delaware River: Bivalve Bioavailability Studies

The objectives of this study were to (1) determine the bioavailability of polycyclic aromatic hydrocarbons (PAHs) and other non-polar organics in resident brackish water clams (Rangia cuneata) at selected sites near an oil refinery; (2) determine if the tissue burdens were causing adverse effects to the clams, and (3) evaluate potential seasonal variations from reproduction in clams taken from the same beds in the spring and fall. Clams were evaluated from three beds located in the refinery discharge plume (near-field stations), three beds located up river outside of the Refinery effluent plume (north far-field), and three beds down river (south far-field) of the Refinery plume. Total PAH concentrations in the tissues of the near-field clams were significantly higher than in the clams located at the far-field stations in both the spring and fall. Total PAH concentrations of the near-field clams were significantly higher in the spring than the fall. No difference was found in total PAHs in the spring or fall in the far-field clams. Total pesticide and total PCB concentrations were significantly higher in the spring than the fall at all stations. The highest concentrations of both pesticides and PCBs were found at the north far-field stations. A tissue residue concentration analysis and three theoretical approaches for estimating detrimental effects to clams in both the near- and far-field suggested that no adverse effects should occur from total PAHs, total pesticides, or total PCBs. Some uncertainty, however, was associated with the theoretical approaches. An estimate of clam density in each clam bed showed that Rangia were growing and reproducing at all stations.     

The Intersection of Independent Lies: Increasing Realism in Ecological Risk Assessment

In 1966, Levins presented a philosophical discussion on making inference about populations using clusters of models. In this article we provide an overview of model inference in ecological risk assessment, discuss the benefits and trade-offs of increasing model realism, show the similarities and differences between Levins' model clusters and those used in ecological risk assessment, and present how risk assessment models can incorporate Levins' ideas of truth through independent lies. Two aspects of Levins' philosophy are directly relevant to risk assessment. First, confidence in our interpretation of risk is increased when multiple risk assessments yield similar qualitative results. Second, model clusters should be evaluated to determine if they maximize precision, generality, or realism or a mix of the three. In the later case, the evaluation of each model will differ depending on whether it is more general, precise, or realistic relative to the other models used. We conclude that risk assessments can be strengthened using Levins' idea, but that Levins' caution that model outcome should not be mistaken for truth is still applicable.     

An Ecological Risk Assessment Methodology for Screening Discharge Alternatives of Produced Water

Previous studies on Ecological Risk Assessment (ERA) of produced water relied on the use of deterministic hydrodynamic models. The assessment was usually carried out in the North Sea context using a model such as the Chemical Hazard Assessment and Risk Management (CHARM), or in the North American context based on the output of a hydrodynamic model such as the Cornell Mixing Zone Expert System (CORMIX). In both these cases, however, probabilistic analysis has not been employed, particularly, to account for uncertainty associated with hydrodynamic models in the ERA study. In fact, it is the hydrodynamic model that has a direct linkage to the selection of the discharge alternatives. Apart from the monitoring purposes, in this article, it is suggested that criteria for evaluating discharge alternatives of produced water in a marine environment might incorporate an awareness of ecological risks by incorporating engineering and toxicological aspects. An ERA methodology consisting of problem formulation, analysis, and risk characterization is discussed in light of evaluating the discharge alternatives. A probabilistic analysis using Latin Hypercube Sampling (LHS)–based Monte Carlo (MC) simulations was employed. A depiction of associated risks for an area comparable to a regulatory mixing zone of typical effluent discharges is presented.