A review of the ICLARM fish feeding model MAXIMS is presented, including a
literature review of previous publications using the model. This model calc
ulates feeding period, ingestion and evacuation rates and daily ration from
sets of 24-h stomach content field data without the need for laboratory st
udies to determine any of these parameters. The best fit to the data is det
ermined by means of a lowest sum of squared residuals (SSR) approach. The m
odel has four main routines, giving the option of either one (Model 1) or t
wo (Model 2) feeding periods and ingestion rates that are either constant (
Models 1.1 and 2.1) or inversely dependent on the level of stomach fullness
(Models 1.2 and 2.2). When both the constant and inversely dependent model
are fitted to the same data set, nearly the same fit and a slightly lower
SSR value is obtained using the latter model; the mathematical reasons for
and statistical significance of this phenomenon are discussed. The authors
attempt to give biological guidelines on the choice of model for different
fish species depending on feeding type and environmental circumstances, as
well as a mathematical way to test the goodness of fit of the Models 1.1/2.
1 curves against that of the Models 1.2/2.2 curves so that these biological
consideration may gain statistical support. An improved version using the
software package SAS for Windows is also presented, calculating confidence
limits for the various parameters so that curves for different data sets ma
y be compared against each other. (C) 1999 Elsevier Science B.V. All rights
reserved.