Sequential analysis of lines of evidence—an advanced weight‐of‐evidence approach for ecological risk assessment

Abstract

Weight-of-evidence (WOE) approaches have been used in ecological risk assessment (ERA) for many years. The approaches integrate various types of data (e.g., from chemistry, bioassay, and field studies) to make an overall conclusion of risk. However, the current practice of WOE has several important difficulties, including a lack of transparency related to how each line of evidence is weighted or integrated into the overall weight-of-evidence conclusion. Therefore, a sequential analysis of lines of evidence (SALE) approach has been developed that advances the practice of WOE. It was developed for an ERA of chemical stressors but also can be used for nonchemical stressors and is equally applicable to the aquatic and terrestrial environments. The sequential aspect of the SALE process is a significant advancement and is based on 2 primary ideas. First, risks can be ruled out with the use of certain lines of evidence, including modeled hazard quotients (HQs) and comparisons of soil, water, or sediment quality with conservative soil, water or sediment quality guidelines. Thus, the SALE process recognizes that HQs are most useful in ruling out risk rather than predicting risk to ecological populations or communities. Second, the SALE process provides several opportunities to exit the risk assessment process, not only when risks are ruled out, but also when magnitude of effect is acceptable or when little or no evidence exists that associations between stressors and effects may be causal. Thus, the SALE approach explicitly includes interaction between assessors and managers. It illustrates to risk managers how risk management can go beyond the simple derivation of risk-based concentrations of chemicals of concern to risk management goals based on ecological metrics (e.g., species diversity). It also can be used to stimulate discussion of the limitations of the ERA science, and how scientists deal with uncertainty. It should assist risk managers by allowing their decisions to be based on a sequential, flexible, and transparent process that includes direct toxicity risks, indirect risks (via changes in habitat suitability), and the spatial and temporal factors that can influence the risk assessment.

Guidance on tiered risk assessment for plant protection products for aquatic organisms in edge‐of‐field surface waters

“…This process of combining available ‗lines of evidence' to form an integrated conclusion or risk characterisation is frequently referred to as ‗weight-of-evidence' assessment (e.g. EC, 2002c;Hull and Swanson, 2006). This term reflects the principle that the contribution of each line of evidence should be considered in proportion to its weight.…”

Ecological Risk Assessment of Sediments in Sydney Harbour, Nova Scotia, Canada

“…The LOE are integrated independently to allow for independent weighing of the evidence by risk managers and stakeholders. The general LOE approach of comparing sediment quality with conservative SQGs has been described by others and has been adopted by many organisations actively involved in development of ERA methodology, including the U.S. Environmental Protection Agency (US EPA), Society of Environmental Toxicology and Chemistry (SETAC), and European Centre for the Ecotoxicology and Toxicology of chemicals (ECETOC) (Suter, 1995;Hull and Swanson, 2006).…”

Assessment of eutrophication and phytoplankton community impairment in the Buffalo River Area of Concern

“…The use of a weight of evidence approach for ecological risk analysis was introduced in the early 1990s in support of Superfund assessments and increasingly is a popular method to evaluate and communicate various types of environmental and health risk (Krimsky, 2005;Hull and Swanson, 2006;Sanderson et al, 2006). Although a potentially valuable assessment tool, weight of evidence evaluations have been criticized, particularly in the public policy arena, as frequently using a process methodology that is low on transparency and high on subjectivity (Krimsky, 2005).…”