Ecological Risk Assessment of Diffuse and Local Soil Contamination Using Species Sensitivity Distributions - Dealing with Contaminated Sites: From Theory Towards Practical Application - page 643

100

US, fish
logY = 1.1163*log(msPAF) - 0.4661
R 2 = 65.80%, F = 177, p = 10 -23

UK, Invertebrates
logY = 1.6239*log(msPAF) - 0.0825
R 2 = 87.01%, F = 883; p= 10 -60

10

1

0.1

0.01

0.01 0.1 1 10 100 0.01 0.1 1 10 100

Potentially Affected Fraction of species (msPAF(EC 50 ), as %)

Fig. 14.5 Association between (SSD-) predicted (X-axis, msPAF) and observed fraction of species
loss attributed to those mixtures (Y-axis), derived from diagnostic modeling on (bio)monitoring
data concerning fish species in Ohio (U.S.) rivers, and freshwater invertebrate species in England
and Wales. The toxic pressure data are based on SSD-output for various contaminants, yielding an
msPAF based on EC50 data for each site. The Y-data are based on a diagnostic method, separating
out the probable contribution of actual contaminant mixtures from other stressors as causes of
species loss (De Zwart et al. 2006 ). Data points refer to XY-consolidated averages. All the sites with
(at the first decimal level of their logs) comparable values are represented by one single cross. The
dotted line represents the ideal hypothetical condition, where predicted fraction of species affected
equals observed species' loss attributed to toxicant mixtures. The dotted grey areas indicate that
there is a range where both X and Y values are lower than realistic values (any species' loss can
per se not be lower than one, regardless of the chosen taxonomic resolution). The solid lines were
fitted to the observations and were highly significant. Data from De Zwart and Posthuma ( 2006 )
and De Zwart et al. ( 2008a )

1. There is an association between predicted (X) and observed species loss (Y),
with a highly positive slope: apparently increased acute toxic pressures imply
increased probability of species loss in the field.
2. The predicted species loss (X) is generally higher than the observed loss
attributed to mixtures (Y), indicating species loss due to other stressors and/or an
association between EC50 and species loss that is generally not 1:1, apart from
the highest levels of mixture exposure.
3. At the highest toxic pressure levels, the predicted and observed values tend to be
similar, at least for the invertebrates.
4. There is huge variability (more or fewer species may be lost than predicted)
probably due to natural variability and other stressors.
5. There are no observations where there is low toxic pressure (X) but an associated
high species loss (no data upper left corner), nor is there a high predicted loss
and no observed loss (no data lower right corner).