My nascent ideas on the application of economics/voting theory to decision making in animals, e.g. during choice between different palatable foods.
During the execution of motivated behaviors, animals often have to identify and choose between two or more objects objects: between food sources (e.g. hi-calorie vs. lo-calorie, or non-toxic vs. toxic), between possible mates (e.g. pair bonding in priarie voles), between shelters, and offspring (own offspring vs. unrelated offspring).
Of course, it should be noted that not all species need to make the same choices. For example, a grazing herbivore or a highly specialized predator may not be able or need to make food choices. Promiscuous species (e.g. montane vole) do not form mate preferences (discriminate between mates?) Solitary mothers of altricial offspring (e.g. rats) do not distinguish between their offspring and others, while herd mothers of precocial offspring (e.g. sheep) need to recognize their own.
Mechanism of Decision Making
The standard view seems to be that when an animal is confronted by a choice of foods, the decision to choose one food over the other is determined by a rational weighing of different factors (i.e. algebraic summation of neural or synaptic weights), leading to a threshold choice for one food over another. For example, when given a choice between two palatable solutions, perhaps dopamine is released in the n. accumbens in response to both flavors; due to prior experience, however, the flavored solution with, e.g., the higher caloric content might induce more dopamine release than the other solution, triggering a choice of the caloric flavor.
Rather than an algebraic summation of preference/aversion value, I hypothesize a group of neurons/synapses are "voting" on the choice of foods. Furthermore, the system of voting may not necessarily be a "first past the post" system in which the plurality of neurons in favor of one solution "wins"; alternative systems of voting are possible. The Economist had an article describing the work of Saari that employed as an example voting for beverage choice.
Releasing Stimuli vs. Choice
As shown by Tinbergen et al., animals can be "hard-wired" to respond with particular behaviors to fairly specific "releasing" stimuli. The classic example is the response of gulls to artificial eggs: when an egg is separated from the nest, the gull will retrieve it. Parametric studies showed that round eggs were better than cubic eggs at eliciting retrieval, large (even abnormally large) eggs were better than small eggs, and green eggs were better than other colors (even the natural mottled gray coloration).
Operationally, the animal is making a decision to choose one stimulus (small cubic egg) over a second stimulus (large green egg). But is there there a distinction when the choice is determined by the pre-programmed salience of a releasing stimuli? Is there a different process at work when the stimuli are more diverse, less pre-programmed, or have learned associations?
Useful references to track down:
Donald G. Saari on economic models of voting.