ABSOLUTE VERSUS RELATIVE MEASUREMENTS OF SEXUAL SELECTION: ASSESSING THE CONTRIBUTIONS OF ULTRASONIC SIGNAL CHARACTERS TO MATE ATTRACTION IN LESSER WAX MOTHS, ACHROIA GRISELLA (LEPIDOPTERA: PYRALIDAE).

When females choose a mate among a group of signaling males concentrated in a small area, a male's mating success is often determined not only by his absolute attractiveness but by the attractiveness of his neighbors as well. Multivariate analyses of sexual selection measurements based on absolute values of predictor variables are then misleading, because such analyses assume that the fitness of a given individual is not influenced by others. We addressed this problem of relative fitness in sexual selection by developing two adjustments of the predictor variables in the multivariate analyses by including group means in addition to absolute values and by using relative values, deviations from group means. In the lesser wax moth, Achroia grisella (Lepidoptera: Pyralidae), aggregated males produce ultrasonic signals attractive to females in the vicinity of honey bee colonies. Playback experiments showed that females prefer signals whose pulses are louder, longer, delivered at a faster rate, and include lengthy silent gaps within pulse pairs (long asynchrony intervals). To measure sexual selection on ultrasonic signals, attractiveness was assessed by observing the number of females orienting toward given males in four-choice trials. The various signal characters of the males were computed from ultrasound recordings. Multivariate regression and nonparametric analyses revealed that peak amplitude, asynchrony interval, and pulse rate were the direct targets of selection. Nonlinear stabilizing and correlational selection were also detected. Comparisons of results from multivariate analyses based on absolute values and on the two relative adjustments indicated that the quality of a four-male group did not affect the strength or direction of linear selection but changed the strength of nonlinear selection. Both relative adjustments improved the prediction of male attractiveness. Three-dimensional surface plots generated by the nonparametric regression analyses showed that attractiveness increased monotonically with any combination of the signal characters. This graphical technique also showed that female choice in A. grisella was based on a relative, as opposed to a threshold, decision rule. Overall, male attractiveness for the population remained consistent between testing days. However, attractiveness values for some individual males changed between successive testing days. In these cases, the change in attractiveness was related only to the change in peak amplitude.