Risk management of chemicals and contaminated land – from planning to verification - Environmental Deterioration and Contamination: Problems and their Management - page 274

- Various chemical forms of inorganic and organic contaminants can be formed

in the environment depending on pH, redox potential, and other chemicals
present;

- The impacted environmental compartments and phases may differ from each other

in sensitivity and risk buffering capacity;

- Sensitivity of the impacted species is rather different, and what is detrimental for

one, may be harmless for another;

- Adaptive potential of the ecosystem is high. Microorganisms can live with toxic

and otherwise harmful contaminants by tolerance and resistance or even use them
as energy substrate;

- The indicators selected for measuring adverse effects may or may not be represen-

tative;

- Measured toxicity end points give different results depending on the effect mech-

anism of the contaminant, for example, two contaminants having the same EC 50
value (the concentration which causes 50% inhibition in the measured end point,
e.g., respiration) may have a 100-fold difference in their no observed effect con-
centration (NOEC) value (the highest measured concentration without adverse
effect);

- Low-level and long-term effects and those which are not dose dependent cannot

be measured;

- Bioavailability of contaminants in the environment is a general problem. Reduced

or completely restricted bioavailability is rarely taken into consideration in the
risk value, mainly because irreversibility is completely uncertain, or better to
say, our knowledge is too little to reduce this kind of uncertainty. An innova-
tive approach can simulate the realistic worst case in microcosms or in other
test systems (see also Chapter 6 in Volume 2) what can be prognosed by pes-
simistic forecast. Another new method is the simulation of the biological effects on
contaminants in soil (biosolubilization, bioextraction, bioleaching, biomobiliza-
tion, membrane uptake, digestion, etc. of contaminants) using biomimetic testing
methods by the application of chemical solvents, sorbents and complexing agents,
which interact similarly with the contaminants as living organisms, or microbial
communities.

In addition to the above listed errors originating in the stochastic nature of the
environment and the uncertainties in the composition of the contaminant(s) and the
shortcomings of the testing methods, the uncertainties of the models used present
another problem. These uncertainties also originate from the heterogeneity of the
environment, the composition of the biota, species sensitivity distribution, etc.
The most significant error may be an error in the concept, which is typical still
today: limited knowledge and incorrect mapping of the problem reflected in a false
conceptual model.
These uncertainties appear in the models and in the ERA protocols as uncertainty
factors trying to avoid underestimation of the risk. When every compartment, every
single model includes an uncertainty factor, their addition or multiplication may result
in a multiple overestimate. The uncertainty factors should be minimized by improving
our knowledge, acquiring good quality data and information and applying statistical
methods which can exclude multiple overestimations.