We have introduced a framework that enables the identification of the impor
tant elements in complex nutritional systems, and the quantification of the
interactions among them. These interactions include those among the multip
le constituents of the ingesta, as well as between behavioural (ingestive)
and physiological (post-ingestive) components of nutritional homeostasis. T
he resulting descriptions provide a powerful means to generate and test hyp
otheses concerning the mechanisms, ecology and evolution of nutritional sys
tems. We provide an overview of the key concepts involved in our scheme, an
d then introduce four examples in which the framework is used to develop an
d test hypotheses. In the first example we use comparative methods based on
a data set of 117 insect species to test a prediction about the relationsh
ip between evolving an association with bacterial endosymbionts and the com
position of the optimal diet. Second, using two species of locusts (a grass
specialist and a generalist), we consider the relationship between an anim
al's diet breadth and the decision rules employed when feeding on foods con
taining suboptimal protein :carbohydrate values. Third, we introduce a math
ematical model that predicts the dose-response properties of gustatory syst
ems in the context of nutritional homeostasis. Finally, we consider the int
eraction between tannic acid and macronutrient balance in the diet of locus
ts.