Ecological Risk Assessment of Diffuse and Local Soil Contamination Using Species Sensitivity Distributions - Dealing with Contaminated Sites: From Theory Towards Practical Application - page 663

or 100) to set a “safe” soil quality standard in the context of soil protection. In
comparison to the concept of uncertainty factors, SSDs are more refined.
Within the SSD-context itself, however, two tiers can be considered:
1. a basic use of SSDs (one well-fitted model for each contaminant); and
2. a refined use of SSDs.
In the refined use , the risk assessor considers the risks of contaminants for var-
ious subgroups of organisms. That is: the input data set is split in two or more
subgroups before fitting two or more SSDs. For example, in the case of soil contam-
ination with an insecticide, one can make an SSD for arthropods and one for “other
species”. There are three reasons for following a refined approach: (1) the problem
definition requires that the assessor follow the refined approach, (2) additional data
or insights trigger the refined approach (e.g., a specific Mode of Action is identified)
and (3) statistical findings like misfit of models to the data show the basic use to be
inadequate.
The refined approach is illustrated in Fig. 14.14 .A basic assessment for the pho-
tosynthesis inhibitor atrazine would suggest a moderate fraction of all species to be
impacted (approximately 75%), with a considerable misfit of the model to the data.
A refined assessment shows an appropriate fit of the three submodels to the data sets,
and suggests that the major fraction of primary producers (estimate: 100%) would
suffer from the same exposure. This leads to the subsequent ecological implication,
that the whole system would be at risk, due to the selective risk for the primary
producers followed by indirect effects on species depending on those primary pro-
ducers. For contaminants with non-specific Modes of Action (like narcotic action),
the refined assessment matters much less, as shown in the examples on benzene.
SSDs are quite simple as compared to higher tier methods such as food chain
modeling (in which one considers transfer of a contaminant through a food chain
from plant eater to top predator) or population or food web modeling (see e.g.,
Forbes et al. 2001 , 2008 ). But, in contrast to the latter methods, SSD-approaches are
versatile, there are many data available to use SSDs in practice, they are easy to use,
they quickly yield insights that may be helpful directly in Risk Management or for
focusing a next assessment step (see below), and they have an intuitive interpretation
(fraction of species affected).
14.11 Weight-of-Evidence and Tiered Use of SSD Output
SSD output can be used as a single line of evidence or in a weight-of-evidence setting
(multiple lines of evidence). This is treated in detail in Rutgers and Jensen ( Chapter
15 of this topic), where the toxic pressure output of SSD-modeling is used with
toxicity tests of contaminated soils and field observations, to assess actual risks of
high contaminations.