A gastric evacuation curve expresses how fast prey disappear from the
stomach, and empirical models are used generally for the relationship
between weight of prey remaining (W-t) and time (t) after a meal. Unfo
rtunately, empirical models are likely to have restricted applicabilit
y because their parameters often represent limited biological mechanis
ms. This paper develops a simple digestion model. The simplest form of
the model has four parameters; the digestion velocity (expressing enz
ymatic breakdown of prey), prey length, initial prey radius and the de
nsity of the prey. Two more parameters are included in an extended ver
sion; a time-delay before digestion starts and environmental temperatu
re. The approach is based on the assumption that prey digestion is a s
urface process in that digestive enzymes attack progressively deeper i
nto a prey of known size and shape so that the average digestion rate
is proportional to the prey radius r (m). This process is characterize
d by the digestion velocity d(s) (m s(-1)). Unknown parameters are est
imated with uncertainty using the maximum likelihood technique. Model
evaluation using published data sets demonstrated that the new model i
s flexible. Prey geometry is incorporated into the model and temperatu
re effects upon gastric evacuation are linked directly to the digestio
n velocity. (C) 1995 The Fisheries Society of the British Isles.