Mesocosms consisting of physically and biologically intact segments of natu
ral communities are an ideal compromise between single species tests and ec
osystem experiments in the assessment of sediment contamination. Therefore,
large intact sediment cores, as mesocosms with naturally co-adapted commun
ities, would allow sediment contamination to be assessed using the replicab
ility and statistical power of laboratory techniques, while retaining much
of the ecological realism of field studies. This study investigates the col
lection and maintenance of such cores, collected from an unimpacted site in
Lake Erie. It demonstrates that box cores containing relatively undisturbe
d freshwater sediments can be brought back to the laboratory and maintained
for up to 8 weeks with little change in the resident benthic fauna. Feedin
g the systems is not required, nor is it deleterious to the indigenous faun
a. (C) 2001 Elsevier Science Ltd. All rights reserved.