In pesticide bioassays, especially those with neurotoxic agents, effects on animals are typically grouped into classes according to behaviour, such as normal and affected behaviour, which may range from unstable walking behaviour, to unable to move, to mortality. Generally, recovery is observed in all these effect classes, except the last. Mortality, however, disturbs the analysis of the recovery processes because it decreases the number of animals that otherwise could have shown a reversible effect. We consider that this interaction between mortality and other, reversible, effects is a conceptual problem, and give arguments in favour of analysing changes in behaviour and mortality as two independent, simultaneously occurring neurotoxic syndromes. As an illustration, two data sets are analysed in both ways and these show that marked differences may exist between conclusions reached by the two viewpoints. The consequences thereof are discussed in relation to toxico‐kinetic explanations for neurotoxicant effects on behaviour and mortality.
© 1999 Society of Chemical Industry