It is undisputed that single species toxicity tests are essential for obtaining information on
concentrations and durations of exposures to chemicals that result in changes in survival,
reproduction, physiology, biochemistry and the behaviour of individuals within a
particular species (Caims, 1983). However, the extent to which single species toxicity
tests can be used to predict effects in the natural environment and changes at higher levels
of biological organization are largely unknown (Kooijman, 1985). The ultimate aim of
laboratory tests is to predict the potential effects of toxicants in natural systems in order
to protect the structure and functioning of the ecosystem. Hence, the aim of this study was
to investigate the validity (i. e. the extent to which effects observed in the laboratory are
mirrored in a more natural environment) and ecological relevance (i. e. the extent to which
the tests are indicative of effects at higher levels of organisation which are important for
the structure and / or functioning of the ecosystem) of a particular single species laboratory test.
In order to increase its potential ecological relevance, both test species and response
criteria for the laboratory test were carefully selected. The effects of copper, a reference
toxicant, on energy budget parameters of Gammarus pulex, an important stream
detritivore, were investigated. The validity and ecological relevance of these energy budget tests, under more natural conditions, were investigated with the use of outdoor
artificial streams.
In the laboratory tests, consumption rates were identified as the most sensitive energy
budget parameter to copper-induced stress in G. pule-x. This decrease in consumption rates
was due to an increase in the body copper concentration of G. pulex via copper uptake
from the media, rather than to a rejection of copper contaminated food.
As well as being the most sensitive parameter to copper stress in laboratory tests,
consumption rates were identified as the most sensitive energy budget parameter
measured in animals deployed in the artificial streams. Further, there was no statistically
significant difference between the effects of copper on consumption rates of animals in
laboratory tests and in the artificial streams. Therefore tests based on consumption rates
as a response criterion were valid in more natural systems. Potential additional effects on
the consumption rates of G. pulex due to copper-induced reductions in food quality were
also indicated.
The ecological relevance of these tests was investigated by trying to predict higher level
effects in terms of growth and reproduction of whole organisms in the laboratory and in
terms of the structure and functioning of populations and communities in artificial
streams. Predictions regarding effects of copper on growth and reproduction of whole
organisms were fairly accurate in indicating trends but were unsuccessful in predicting
actual growth rates and reproductive output.
Population density and leaf processing were selected as response criteria indicative of
effects of copper on the structure and function of the freshwater ecosystem respectively. Effects on these two parameters were observed at the same concentration as adverse
effects on consumption rates, and no adverse effects were detected at concentrations
below those causing a decrease in consumption rates. Potential consequences, of the
observed effects of copper on G. pulax, for the structure and functioning of the freshwater
community are discussed.
iii
It is concluded that by careful selection of ecologically meaningful response criteria and
test species, laboratory tests based on individuals can be representative of potential impact of a toxicant at higher levels of organisation in the natural environment.
Consumption rates in G. pulex represent a sensitive, valid and ecologically relevant
response criterion for the determination of the effects of a pollutant in natural ecosystems