Decision making in crickets: Neuronal correlates of behavioral choice

Animals and humans have to make decisions in choice situations of variable context to maximize their utility function. These decisions are often complicated by the fact that there are several available options which differ in a number of attributes, and that no single option is superior over others. The present proposal suggests a model system which allows testing predictions following absolute and relative decision making both in behavior and at the neuronal level, using female crickets and their decision in acoustic mate choice behavior. Since previous work has shown that mate choice in females is based on the evaluation of activity in only two identified pairs of interneurons, we can use a combination of behavioural and neurophysiological tools to test a number of current hypotheses on decision making. We will determine the response- and trade-off functions for different attributes of an acoustic signal in a choice situation and can then test, whether the decision circuitry in the female is based on absolute standards or on relative ones. These experimental series will enable us to determine whether the processing of acoustic signals takes places in a serial or parallel manner with respect to recognition and localization. Arena trials and outdoor tests will link decision making in a standard environment to a more complex and realistic context. Linking the results of behavioural tests to neuronal activity we will be able to test different hypotheses about the neural computation based on the discharge patterns in two opposing neurons underlying decision making.

Animals and humans have to make decisions in choice situations of variable context to maximize their utility function. These decisions are often complicated by the fact that there are several available options which differ in a number of attributes, and that no single option is superior over others. Decision making in humans suggests that the magnitude of human preferences and our rank ordering of options can vary with the number and type of alternatives. The same problem also exists for many animals when making decisions about the best option for mating partners. Rational choice theory predicts that animals assign each signal a fixed value on a single dimension. The probability of choosing one signal over the other should be a monotonic function of the respective values and result in transitive choices. A violation of transitivity in choice behavior would suggest comparative rather than absolute decision making. The present project used the model system of female crickets and their choice for mates to test the concepts of transitivity and independence from irrelevant alternatives, because they make clear predictions about choice behavior, if decision making is based on absolute or relative evaluation of available cues. We tested female choice both in the laboratory and under field conditions, and found a violation of transitivity, suggesting a comparative evaluation of available signals by female crickets. We also found that irrelevant alternative signals, which are not preferred by females, may nevertheless play an important role in decision making.