Accurate measurements of prey availability are vital to our understand
ing of foraging behaviour, particularly prey selectivity. In the prese
nt study, observations of shorebirds foraging for fiddler crabs on int
ertidal mudflats demonstrate that prey availability depends both on th
e temporal variation in crab activity and on the crabs' responses to t
he presence of foraging shorebirds. Our results suggest that measureme
nts of prey availability that do not specifically account for prey act
ivity patterns and their responses to predators are neccessarily inacc
urate. Furthermore, our results also show that tests for foraging sele
ctivity are extremely sensitive to the way in which prey availability
is measured and can even indicate active prey selectivity when more ac
curate measures of prey availability show predators to be non-selectiv
e. Because inaccurate measures of food resources greatly reduce our ab
ility to detect food preferences, greater care must be taken to accoun
t for prey activity patterns and their responses to predators in measu
rements of prey availability. (C) 1998 The Association for the Study o
f Animal Behaviour.