Jason Munshi-South

Monogamy occurs in only 5% of mammalian species, but is significantly more common in the Euarchonta: primates, dermopterans, and treeshrews (15% spp.). However, many of these species do not breed monogamously, indicating the need to understand behavioral and genetic monogamy as separate evolutionary phenomena. I examined monogamy in the large treeshrew (Tupaia tana) in Sabah, Malaysia using radiotelemetry data from 46 individuals tracked during and after a fruit masting episode in 1990-1991, during a non-masting period from 2002-2004, and in a selectively logged forest from 2003-2004. I show that large treeshrews exhibit behavioral monogamy in all these ecological situations. However, behavioral monogamy is best characterized as dispersed pair-living, or "asocial monogamy", in this species because male-female pairs travel, forage, and sleep alone on their joint territories. Next, I use microsatellites and mitochondrial DNA d-loop haplotypes to analyze the genetic maternity and paternity of 24 T. tana offspring. I show one of the highest rates of extra-pair paternity (EPP) ever recorded for a behaviorally monogamous mammal. Over 40% of young were sired by males that were not the behavioral partner of their mother, and three litters exhibited evidence of multiple paternity. Comparative analysis of relative testis size in treeshrews and primates indicates that sperm competition is not associated with the high rates of EPP in T. tana, and that the evolution of monogamy is associated with the evolution of smaller testes. Finally, I find genetic evidence of female-biased dispersal and gene flow in large treeshrews. The vast majority of mammals exhibit the behavioral combination of polygyny and male-biased dispersal, but female-biased dispersal may evolve in monogamous species when females compete for ecological resources. In support of the local resource competition hypothesis, I find lower population assignment probabilities and pairwise relatedness for females than males. These results indicate that female T. tana are a mixture of philopatric residents and immigrants from other areas. Coalescent-based Bayesian analyses also show that historical female migration has been three times higher than the overall migration rate between primary and logged forest populations, providing evidence of female-biased gene flow.