In vitro evaluation
To assess what the minimum concentration of pesticide is able to generate an effect on the water organisms, toxicity tests can be carried out in laboratories on series of individuals belonging to different species. Those tests, often normalized, enable to estimate the intrinsic toxic potential of a substance (= danger). Besides, they enable to compare the toxicity of several substances. Most of the time, they are carried out on species considered as sensible and representative of different links of the food chain (microscopic algae, herbivorous invertebrates such as daphnies, fish such as trout for instance). Moreover, those species have to be easily bred in a laboratory. The aim of these tests is to determine the link between the exposure (to the concentration) and the effects (relation of effect-concentration) in order to estimate various magnitudes, such as CL or CE50, NOEC or LOEC (figure 1).
Figure 1. Example of a theoretical relation between the concentration of a substance and its effects assessed on lots of organisms at the laboratory (CL50 = lethal concentration for 50% of the exposed individuals; CE50 = effective concentration which reduces for 50% the intensity of a biological process such as the growth of a population for instance; NOEC = No Observed Effect Concentration; LOEC = Lowest Observed Effect Concentration).
However, the laboratory are completed in simplified conditions (one substance only, one species only and no sediments for instance) and do not allow to estimate only part of the direct effects of the tested substances. The simplicity of the experimental conditions do not allow to consider the phenomenons which, within an ecosystem, can put up (presence of stages or species more sensible, (bio)masses) or on the contrary reduce (absorption on the sediments or hanging material, accelerated degradation) the bioavailability and/or the toxicity of the contaminants. The use of experimental food chains can enable to bring out contamination phenomenons from food. The use of microcosms, laboratory miniature ecosystems with water, sediment and one or several animal or vegetable species, enable to get closer to natural environmental conditions and consider various phenomenons such as the absorption of toxic substances on sediments for example.
In natural conditions, investigation means go from the setting up of experimental ecosystems (canals, ponds) to the contamination of natural environments, passing by the use of enclosed portions of a natural ecosystem. Nowadays, the word "mesocosm" (artificial ecosystems placed in natural environmental conditions) is used to designate equally canals or artificial rivers, experimental ponds or enclosures. Those systems represent an intermediate degree of organisation between the simplicity of laboratory tests and the complexity of natural environments. The mesocosms enable in particular to assess the effects of substances tested on various levels of biological organization thanks to the measure of different qualitative and/or quantitative answers. It is also possible to bring out the indirect and direct effects of toxic substances on the various compartments of water ecosystems. Moreover, the phenomenons which reduce or increase the bioavailability of the contaminants, and may therefore influence their toxicity, can be taken into consideration. The aim of these mesocosms is not to mimic a precise natural ecosystem. We have to constitute systems which have their own characteristics, in which diverse and realistic scenarios - on an ecological point of view- will be put up (present organisms, nature of the contamination, etc). As such, they allow to gain relevant data, in satisfactory feasibility and reliability conditions.
The biotests constitute an essential first step in order to gauge the toxicity of a contaminant on representative species of different trophic scales. Conversely, the studies in micro- or mesocosms can enable to identify ecological process or functions which will be relevant to mention in the risk analysis (long term or indirect effects, for instance). Along with these experimental approaches, mathematic model-making tools allowing to simulate an answer of the biological systems to the toxic substances are developed for several levels of biological organization (individuals, population, community). The combined use of experimental tools and mathematic tools enables to approach the problematic of the necessary change of scale to go from effects measured on an individual scale to those which happen on a population and community scale. The model-making is also a good means to analyze and predict the effects of the mixing of the substances.
(translation: Alice Gautreau)
