Ecological Risk Assessment of Diffuse and Local Soil Contamination Using Species Sensitivity Distributions - Dealing with Contaminated Sites: From Theory Towards Practical Application - page 665

This weight-of evidence approach is commonly used in tiered assessments.
Higher tiered assessments are applied when the uncertainty in the SSD-output is
so large, or the costs of possible interventions are so high, that one wants more
refined assessments. For example, when a set of 100 sites is ranked using 100 cal-
culated values of msPAF, one may have sufficient budget to remediate only 10 sites.
Hence, one may select a top-20 of sites using SSDs only, and then apply the more
costly weight-of-evidence scheme of Chapter 15 by Rutgers and Jensen, this topic
to give a detailed and experimentally based ranking of only those sites. This is more
cost-effective than weight-of-evidence testing of all 100 sites.
14.12 Key Strengths and Limitations of SSDs
SSD-modeling has strengths and weaknesses, especially in comparison to other
methods.
First, SSDs are easy to use, requiring only simple statistical software and appro-
priate ecotoxicity data. Such software and data are available. Second, SSDs combine
ecotoxicity data for a contaminant in a way that is easily interpreted, namely in
terms of a potentially affected fraction of species. Third, these estimated fractions
can be aggregated to a single estimated toxic pressure for cases of contamination
with mixtures; this is of key importance to solve various problems in practical Risk
Assessment. Fourth, such toxic pressures appear to be related to field impacts, in
the sense that increased predicted toxic pressures relate to increased species loss
and impacts (though much more evidence is needed), despite the influence of other
stressors that affect biotic communities. Fifth, SSD-based outcomes appear to be
robust, in the sense that they can be used for relative ranking an array of com-
pounds, and arrays of contaminated sites. Sixth, SSD-modeling can be used for
very different environmental problems; a key strength is versatility : the concept can
be used for two major purposes, derivation of formal soil quality standards and
site-specific Ecological Risk Assessment. The latter can be done on the basis of
measured or expected concentrations. This permits the comparison of alternative
Risk Management scenarios to the existing conditions. Seventh, the major strength
may be that SSDs can be of help to make Risk Assessment and Risk Management
processes cost effective. Rather than 100 experimental tests, or 100 higher-tier pop-
ulation or food web models for 100 cases, SSDs can help to quickly select the top
20 of risky sites. This will cause a substantial cost reduction when the higher-tier
assessments are done for only the selected sites, and the Risk Management activities
are applied to sites with clearly significant risks.
The limitations of the SSD-approach relate mainly to the issue that the model is
statistical and contains no ecological interactions. Critics are right when they say
that the method has eco-deficiencies, but they would be wrong if they concluded
from that fact that SSD- output has no practical use in Ecological Risk Assessment
and Risk Management (see use examples below). SSD-output appears to result in
consistent patterns of relative rankings, both amongst contaminants and amongst
contaminated sites. This means that the weaknesses apparently do not invalidate the
relative interpretations of SSD-output (the rankings).