COS 83-10
Dispersal kernel heritability in Callosobruchus maculatus and implications for the velocity of range expansion

Wednesday, August 13, 2014: 4:40 PM
Carmel AB, Hyatt Regency Hotel
Brad Ochocki, BioSciences, Rice University, Houston, TX
Adam Geiger, Ecology and Evolutionary Biology, Rice University, Houston, TX
Tom E. X. Miller, BioSciences, Rice University, Houston, TX
Background/Question/Methods

Current theory suggests that individuals in an expanding population should become spatially sorted by dispersal ability, since the best dispersers are the ones most likely to travel to the leading edge of an expansion. This spatial sorting leads to assortative mating by dispersal ability and -- if dispersal is a heritable trait -- accelerating rates of range expansion. Although this process of ‘spatial selection’ is potentially widespread, empirical evidence is sparse. Here, we tested whether the dispersal kernel (the probability distribution of dispersal distance) is a heritable trait in laboratory-reared lines of the beetle Callosobruchus maculatus, a widespread agricultural pest. We also estimated the covariance between the dispersal kernel and life-history traits relevant to low-density population growth, including fecundity, fertility, interspecific competitive ability, and female host preference.

To estimate the narrow sense heritability of the dispersal kernel, we measured the net displacement of individuals with known pedigree. We established the pedigree with controlled breeding trials; when offspring from the breeding trials reached adulthood, we measured their net displacement after two hours of dispersal across a one-dimensional array. We approximated the dispersal kernel with a Poisson distribution, and used generalized linear models to estimate the heritability of the kernel.

Results/Conclusions

Preliminary analyses estimate the narrow sense heritability to be 0.15, with a 95% credible interval of 0.05 to 0.30. We used the deviance information criterion (DIC) to compare the fit of the generalized linear model (an ‘animal model’) to a null model with no additive genetic effect, and the animal model performed moderately better (DIC_animal = 5554.2, DIC_null = 5564.0). Contrary to findings previously reported in this system, dispersal was found to be slightly female biased.

These results suggest that dispersal is a heritable trait in C. maculatus, and that naturally expanding populations of C.maculatus are likely to be subject to spatial selection. A spatial population dynamics model indicated that the level of dispersal heritability that we quantified is expected to have an accelerating effect on spatial expansion. Interestingly, the laboratory-reared lines of C. maculatus used in this study maintain dispersal heritability even though they are inbred and have endured repeated population bottlenecks -- a genetic history that is likely to be paralleled by many invasive organisms. Our results highlight the potentially important but under-studied influence of evolution on the ecological dynamics of range expansion.