Evolution and function of complex signaling in spider courtship behavior

Abstract

Complex signaling, while common throughout the animal kingdom, is poorly understood and generally understudied. This dissertation provides a framework from which one can approach studies of complex signal function. The semantics involved in studies of signal evolution and complex signaling are discussed and complex signaling is broken down into two categories: multiple signals and multimodal signals. While single signals ultimately reflect the pressures of both tactical and strategic design, so must complex signals. I explore hypotheses relating to the function of complex signaling in terms of both tactical and strategic design pressures and then explore hypotheses allowing for the possibility of inter-signal interactions. Predictions of some tactical design hypotheses of complex signaling suggest that multimodal signals can act independently, as back-ups to each other in the presence of environmental noise, or they can interact in a variety of ways. I tested the tactical back-up hypothesis, among others, using the bimodal signaling wolf spider Schizocosa uetzi. Results suggest that the signals are not independent and that there is an inter-signal interaction in which the vibratory signal redirects and focuses a female's attention to the visual signal. A comparative approach explored the tactical and strategic design components of multimodal signaling across three species of Schizocosa, one unimodal signaling species (S. avida, vibration only) and two bimodally signaling species (S. uetzi and S. stridulans). Signal non-independence was found for both bimodally signaling species but the pattern of inter-signal interaction differed between the species. Inter-population interactions between divergent populations that possess independent, derived traits can also be affected by inter-signal interactions. Results from an empirical study using two divergent populations of the jumping spider Habronattus pugillis in inter-populational reciprocal crosses were compared to predictions of different models of sexual selection. Our results show a xenophilic mating preference with one population of females mating more frequently with foreign males than local males, while the second population of females showed no such difference. While this pattern is inconsistent with Fisherian selection and does not completely fit the predictions of pure sensory exploitation, it is completely consistent with one pattern predicted from a process of antagonistic coevolution.