SOME DISTINGUISHING FEATURES OF MODELS OF SEARCH BEHAVIOR AND MATE CHOICE

We develop analytical predictions about how females should behave when either a (infinite time horizon) sequential search tactic or best-of- n rule is employed to search for mates. The likelihood that a male is rejected on first encounter with a female should decline with male qua lity if females use a sequential search tactic. If females employ a be st-of-n rule, however, the probability that a first encountered male i s rejected should not depend on male quality; the proportion of males of any quality rejected on first encounter with a female equals the pr oportion of females that perceive the optimal number of males to be sa mpled prior to mating to be more than one. The behavior of females usi ng either rule, however, depends on the distribution of quality among potential mates. We show that a variance-preserving increase of mean m ale quality results in an increase of the critical value of the thresh old of acceptance under a sequential search rule; that is, females bec ome more choosy. However, such a shift is not expected to change femal e search behavior if a best-of-n tactic is employed. A mean-preserving increase of the variance of male quality results in an increase of n among females that employ a best-of-n rule, whereas the number of male s sampled prior to mating may increase, decrease, or remain unchanged following such a shift if females use a sequential search rule. The th reshold acceptance criterion under sequential search, however, increas es (decreases) with a mean-preserving increase (reduction) of the vari ance of male quality. Patterns of resampling of males by females and t he position of accepted males in a search sequence have often been use d to distinguish between the best-of-n and sequential search tactics. Predictions of each pattern under a sequential search model depend on both whether the time horizon over which search takes place is assumed to be finite or infinite and whether or not recall (resampling) of pr eviously encountered males is permitted and sometimes resemble predict ions of a best-of-n model of search behavior. Because finite time hori zon models of sequential search are currently poorly developed, such i nformation is not likely to provide good evidence that females are usi ng either rule. The predictions that we derive, however, should be rel atively robust to assumptions about recall and the time horizon over w hich sampling takes place. An experimental protocol in which mate qual ity is altered may consequently provide a valuable approach to the stu dy of search behavior and mate choice.