Risk assessment of pesticides for aquatic ecosystems is typically base
d on comparisons of exposure and effect concentrations at a variety of
levels (tiers). At the highest tier, effects assessment can involve g
enerating data under field conditions, typically in mesocosm experimen
ts. However, interpreting the ecological significance of effects measu
red in these studies can be difficult because ecological factors can i
nfluence the outcome of perturbations in the real world. The influence
of ecological factors is not readily addressed experimentally and so
a strategic modelling approach is proposed which may aid in defining a
cceptability of effects.