Jump to main content.

National Information
Region 8 Risk Assessment
Ecological Risk Assessment pages

Risk Characterization

On This Page

Introduction

In this step, the risk assessor combines information on exposure and toxicity to predict the probability that adverse effects will occur in each of the groups of receptors selected for evaluation, with an estimate (when possible) as to whether the effects are likely to result in a population-level impact.

In general, there are three basic strategies that may be used in the risk characterization process:

Ideally, risk characterization is based on a weight of evidence (WOE) approach in which results from several alternative strategies are used.  For example, results from the HQ approach may be compared to the results from site-specific toxicity tests and to results from site-specific demographic studies of receptor density and diversity.  Each of these alternative approaches has some advantages and limitations, as discussed below, and the strongest conclusions usually come from a synthesis of observations across several different approaches (weight of evidence).  Resources and guidance documents are provided at the end of the following discussion under Resources.

Hazard Quotient Approach

One approach for evaluating ecological risks from environmental contaminants is to predict the potential for adverse effects by comparing estimated levels of exposure of various environmental receptors to appropriate Toxicity Reference Values (TRVs).  TRVs are usually derived from review of published toxicity studies, and identify concentration values or dose values that correspond to a No-Observed-Adverse-Effect Level (NOAEL) and/or a Lowest-Observed-Adverse-Effect-Level (LOAEL).  Each TRV is both chemical-specific and species-specific.  The comparison takes the form of a ratio, referred to as the Hazard Quotient (HQ), as follows:

HQ = Site Exposure Level / TRV

If the value of the HQ is less than or equal to one (1E+00), it is believed that no unacceptable impacts will occur in the exposed population of receptors.  If the value of the HQ exceeds 1E+00, then an unacceptable impact may occur, with the predicted likelihood and/or severity of the impacts increasing as the value of the HQ increases.

In some cases, two alternative TRV values are available for a particular metal for a particular receptor, one based on the NOAEL and one based on the LOAEL.  These two alternative TRV values reflect the range of uncertainty that exists in the actual threshold between the presence and absence of an adverse effect.  If the HQ based on the NOAEL does not exceed a value of one, it is concluded that the chemical does not pose a hazard.  If the HQ based on the LOAEL exceeds a value of one, it is expected that the chemical could pose a significant hazard.  If the HQ based on the LOAEL is less than one but the HQ based on the NOAEL is greater than one, the chemical is probably close to a level that could cause adverse effects, but whether or not significant effects would actually occur cannot be judged with certainty.

When the HQ approach is used to evaluate risks to a population of receptors, it is important to understand that there may be a wide range of exposure levels experienced by different members of a population.  Because of this natural variability in exposure, two different types of exposure estimates are often used to describe the range of exposures within a population:

The chief advantages of this HQ-based approach are:

There are also limitations to the HQ approach that stem from uncertainties in both the numerator (the estimate of dose or exposure) and the denominator (the TRV) used to calculate the HQ.  For example, TRV values are based on toxicity tests performed under laboratory conditions, which may or may not account for factors that can influence (either increasing or decreasing) toxicity in the field (e.g., reduced bioavailability, interaction with other chemicals, combined stress from other sources).  Also, some TRVs are based on limited and sometimes internally inconsistent toxicity data, and TRVs may not be available for all receptors of concern at a site.  Therefore, some TRVs may be relatively uncertain, especially when extrapolation of findings across different species is required.  In addition, estimation of actual exposure levels is often difficult (especially for terrestrial receptors) due to lack of site-specific data on intake rates, home ranges, etc. 

Because of these potential limitations, the HQ approach is best considered to be a screening-level means of evaluation.  That is, if an HQ is above a level of concern, this is an indication that effects may be occurring, but further studies (e.g., direct observation of exposed receptors) are sometimes needed in order to confirm if this is really the case.

[Top of page]

Site-Specific Toxicity Studies

A second way to evaluate the toxicity of environmental contaminants on environmental receptors is to perform toxicity studies using contaminated media from the site.  This may be done either in the field (e.g., placing caged fish in the river and observing whether any adverse effect occurs) or in the laboratory using media collected on the site (e.g., placing water or sediment from the site in an aquarium and exposing fish or benthic macroinvertebrates to the medium, growing plants in site soils, administering site soils to laboratory mice). 

The chief advantage of this approach is that site-specific conditions that can influence toxicity are usually taken into account.  A potential disadvantage is that, if toxic effects are observed to occur when test organisms are exposed to site media, it is usually not possible to specify which chemical(s) is (are) responsible for the effect.  Rather, the results of the toxicity testing reflect the combined effect of the mixture of chemicals present in the site medium.  In addition, it is often difficult to test the full range of environmental conditions that may occur at the site across time and space, either in the field or in the laboratory.

When available, site-specific toxicity data are usually used to modify or replace default TRV values.  These site-specific TRVs are used, in turn, to calculate site-specific HQ values that are likely to be more accurate than those based on default or literature-based TRVs.  (Resources for site-specific toxicity testing are provided below.)

Demographic Studies of Ecosystem Status

A third approach for evaluating impacts of environmental contamination on ecological receptors is to make direct observations on the receptors in the field, seeking to determine whether any receptor population has unusually low numbers of individuals, or whether the diversity (number of different species) of a particular category of receptors (e.g., plants, benthic organisms, birds) is lower than expected.

The chief advantage of this approach is that direct observation of community status does not require making the numerous assumptions and estimates needed in the HQ approach.  However, there are also a number of important limitations to this approach. 

The most important of these is that both the abundance and diversity of an ecological population depend on many site-specific factors (e.g., habitat suitability, availability of food, predator pressure) and it is often difficult to know what the expected (un-impacted) abundance and diversity of an ecological population should be in a particular area.  This problem is generally approached by seeking an appropriate "reference area" (either the site itself before the impact occurred or some similar site that has not been impacted) and comparing the observed abundance and diversity in the reference area to that for the site.  However, it is sometimes difficult to locate reference areas that are truly a good match for all of the important habitat variables at the site; comparisons based on this approach do not always establish firm cause-and-effect conclusions regarding the impact of environmental contamination on a receptor population.

This problem is further complicated by the natural variability in population parameters over time -- measurements of diversity and abundance at any one point in time may not be representative of long-term average values.  Thus, comparisons between a site area and a reference area are of greatest value when based on cumulative observations over long periods of time.

Data on population and community structure are especially valuable for hypothesis testing.  In general, the basic hypothesis to be tested is that there is a direct correlation between the concentration of contaminants in the environment and the level of effect observed in exposed populations.  One way to test this hypothesis is to plot a measure of population or community status (e.g., number of fish per acre) as a function of one or more measures of environmental contamination.  If the contaminants are causing an effect on the population-based measurement endpoint, then it is expected that there will be an observable trend in the measurement endpoint as a function of the environmental endpoint. 

If a trend is observed, and if the trend is statistically significant, this may be taken as good evidence for a cause-and-effect relationship.  However, the converse is not necessarily true.  That is, absence of a statistically significant trend is not necessarily proof that metals are having no effect.  This is because the measurement endpoint may depend not only on metals but on numerous other variables (e.g., water temperature, flow rate, prey abundance, habitat quality) and the effect of metals may be partly or entirely obscured by variability in the effects of these other stressors. 

This problem is further complicated by the difficulty usually encountered in obtaining accurate measurements of population status and/or environmental contamination levels.  If one or both values are not known precisely, this "measurement error" can prevent the detection of cause and effect relationships, which do exist.  Thus, if no significant relationship is detected between a community endpoint and the level of environmental contamination, the correct interpretation is that the effect of the contamination (if any) is not sufficiently large to be detected in the face of other independent variables and/or measurement error. 

Resources for demographic studies are provided below.

Weight of Evidence Evaluation

As discussed above, each of the methods available for evaluating potential impacts of environmental pollution on ecological receptors has advantages and limitations.  For this reason, conclusions based on only one method of evaluation may be misleading.  Therefore, the best approach for deriving reliable conclusions is to combine the findings of all methods for which data are available, taking the relative strengths and weaknesses of each method into account.  If the methods all yield similar conclusions, confidence in the conclusion is greatly increased.  If different methods yield different conclusions, then a careful review must be performed to identify the likely basis of the discrepancy and to decide which method is more likely to yield the correct conclusion.

[Top of page]

Resources

Site-Specific Toxicity Testing
General Guidance
Testing the Toxicity of Sediments to Benthic Macroinvertebrates
Testing the Toxicity of Soil on Soil Invertebrates
Testing the Toxicity of Soil on Plants
Testing the Toxicity of Surface Water on Fish and Other Aquatic Species
Demographic Guidance

[Top of page]

Local Navigation

Jump to main content.