POPULATION STRUCTURE OF DIPODOMYS INGENS (HETEROMYIDAE): THE ROLE OF SPATIAL HETEROGENEITY IN MAINTAINING GENETIC DIVERSITY.

The giant kangaroo rat, Dipodomys ingens (Heteromyidae), is an endangered rodent that inhabits approximately 3% of its estimated historic range. Its current distribution is centered in two geographic areas, situated about 150 km apart, in south-central California. We sequenced a 293 base-pair fragment at the 5' end of the control region in 95 giant kangaroo rats from nine localities to examine the genetic structure of extant populations. We determine that mutations in this section of the control region follow a negative binominal distribution, rather than a Poisson. However, the distance between haplotypes is small enough that the difference between a tree that corrects for the non-Poisson distribution of mutations and one that does not, is minimal. This implies that the use of methods that assume a Poisson distribution of mutations, such as those based on coalescent theory, are justified. We find that the correlation between levels of genetic diversity and estimated census size is poor. This suggests that population sizes have fluctuated over time or that populations have not been isolated from one another, or both. We also examine the hierarchical structure of populations and find that the southern populations are not genetically subdivided but that there is significant subdivision between northern and southern populations and between some northern subpopulations. The phylogeographic relationship between northern and southern populations can primarily be attributed to isolation by distance, although the time since divergence between them appears to be less than the age of either. To examine the phylogeographic relationships in more detail we construct a minimum spanning tree based on Tamura-Nei gamma-corrected distances and superimpose on it the geographic position of haplotypes. This reveals that there is more genetic distance between some northern haplotypes than between any northern and southern haplotypes, despite the geographic distance separating north from south and the larger size of the southern population. It also reveals that one northern population, in the Panoche Valley, contains old allelic lineages and shares ancestral polymorphism with several other populations. It also shows that two, small, geographically remote populations contain a surprising amount of genetic diversity, but that different population/geographic processes have affected the structure of that diversity. We estimate the average migration rate among all populations to be 7.5 per generation, and conclude that a disproportionate number of migration events involve gene flow with one northern population, the Panoche Valley. We find evidence for the hypothesis that there has been an increase in population size in the remaining populations in the north and suggest that the Panoche Valley could play a role in these expansions. Finally we discuss the probabilitiy that the genetic structure of the southern populations has been affected by fluctuations in size. These results are briefly compared to other studies on the genetic structure of rodent populations.