Host Range Evolution of Cleptoparasite Meloe franciscanus (Coleoptera: Meloidae)

Abstract

Chemosensory signals play a major role in reproductive isolation and mate location in both invertebrate and vertebrate species. Closely related species often produce similar but distinct signals by varying the ratios or minor components in pheromone blends to avoid interference in their communication channels, and cross-attraction among congeners. However, exploitation of reproductive signals by predators and parasites also may provide strong selective pressure on semiochemical communication signals. For example, bolas spiders mimic the pheromones of several moth species to attract their prey within reach and parasitic phoretic blister beetle larvae mimic the pheromone of their female host bees to attract male bees as the first step in their being transported to the nests of their hosts. In both cases, there is strong selection pressure on the host to discriminate real mates from aggressive mimics, and conversely, on the predator, parasite or parasitoid to track and locally adapt to the evolving signals of their hosts. Chapter 1 investigates parasite host adaptation to multiple hosts across the geographic range of cleptoparasite Meloe franciscanus (Coleoptera: Meloidae). Meloe franciscanus mimics the pheromone chemistry and mate location behavior of its hosts, different species of solitary bees in the genus Habropoda. The larvae of M. franciscanus hatch synchronously and aggregate to collectively produce an olfactory signal that mimics the host female’s sex pheromone to attract male bees to pseudocopulate. We report that M. franciscanus ’ deceptive signal is locally host-adapted in its chemical composition and ratio of components, with host bees from each allopatric population preferring the deceptive signals of their sympatric parasite population. Further, in different locales, the parasite has adapted its aggregation height to the height at which male bees typically patrol while searching for females. In Chapter 2 we tested the hypothesis that all four allopatric populations of Meloe franciscanus are the same species parasitizing each host bee at different localities. These four populations had been defined as a single species based on morphological characters of both the adult and larvae. We examined all four populations using two mitochondrial genes, cox-1 and 16S mtDNA and three nuclear genes 28S, ITS-2 rDNA and EF1-α. We used Meloe strigulosus, Meloe proscarabaeus and Tenebrio molitor as outgroups. Analysis using Bayesian and MP methods of single genes and concatenated mtDNA sequences (cox-1 and 16S) and concatenated nuclear sequences (28S, ITS-2, EF1-α) and total evidence resulted in alternative tree topologies separate species for each population. Chapter 3 explores the utility of multiple functional traits as filters for predicting host range of the parasite Meloe franciscanus including host-parasite spatial overlap, host-parasite temporal overlap, local host population abundance and reliability across time and host physiological suitability using size as a proxy for bee nest provisions. Several functional traits of host bees tested were useful in predicting host range including: local host population abundance and reliability, and host mass as a proxy for nest provisions. Semiochemical signals and behavioral traits, which where examined in Chapter 1 were also important filters for host range for the solitary bee nest parasite M. franciscanus. Taken together, my dissertation shows that M. franciscanus is locally adapted to its solitary bee host Habropoda pallida in the Mojave Desert and H. miserabilis in Oregon. It customizes its chemical blend to match the female sex pheromone of its local host. In addition its cooperative aggregation behavior has been fixed to match the patrolling behavior of the local host bee. Using Bayesian and parsimony analysis of molecular sequences from two mitochondrial and three nuclear genes, four parasite populations studied are inferred to be a single species, M. franciscanus.