ABSTRACT

Title of Dissertation: Reticulitermes flavipes (Isoptera: Rhinotermitidae) colonies: reproductive lifespans, caste ratios, nesting and foraging dynamics, and genetic architecture

Catherine Everitt Long, Doctor of Philosophy, 2005

Dissertation Directed By: Professor Barbara L. Thorne Department of Entomology

The eastern subterranean termite, Reticulitermes flavipes (Kollar) (Isoptera: Rhinotermitidae) is a major decomposer of wood and a significant pest of lumber and paper. Despite its economic and ecological importance, key aspects of its colony dynamics are poorly understood. In 1993, laboratory colonies were initiated with alate pairs. In 2003, the colonies contained an average of 12,600 individuals. Ninety-seven percent of kings and 72% of primary queens survived; 29% of colonies contained replacement female reproductives. In addition to providing unprecedented demographic information regarding colony growth rate and longevity, lifespan of founding reproductives, and the response of colonies to the loss of primary kings and queens, these complete colonies demonstrated foraging and nesting activities of socially intact families. The laboratory colonies foraged in multi-resource networks. Travel between the resource nodes was observed, and after 30 weeks all spatial networks were censused. None of the castes was distributed equally among the three resources. Reproductives, which were found in a satellite node in 71% of colonies, and brood did not share the same node a significant portion of the time, suggesting that the nesting strategy was polydomous rather than monodomous. Mark-recapture data indicate that workers were significantly more likely to be found in the resource where they had been located previously, indicating i) they feed non-randomly among the multiple resources and ii) they feed extensively at one location rather than shuttling regularly between satellite- and central nodes (as in a central-place foraging model). Such site fidelity violates an assumption of the Lincoln Index, leading to significant misestimation of actual colony population totals established by census. Worker genotypes, as determined by microsatellite markers, indicate that despite the absence of obvious physical isolation, genetic differentiation had occurred among the workers in one of the queenless colonies. Inbreeding coefficients generated by the queenless colony genotypes did not differ significantly from the predicted F-statistics for colonies of similar breeding structure, confirming that sampled workers can accurately estimate the breeding regime of field populations.