Analyses of gene flow among populations of deer mice (Peromyscus maniculatus) at sites near hantavirus pulmonary syndrome case-patient residences.

ABSTRACT

Gene flow and potential for Sin Nombre virus (SNV) trafficking of the deer mouse (Peromyscus maniculatus) was studied in Delta and Mesa counties of western Colorado (USA). The study areas included Grand Mesa and surrounding grazing and agricultural areas. This area has several natural potential barriers to rodent gene flow, including rivers, cliffs, and mountains. Ten study sites were utilized in a spatially nested design ranging from 0.65-81 km apart; four of these sites were at or near human hantavirus pulmonary syndrome (HPS) case-patient residences. One HPS case occurred on the north side of Grand Mesa in 1993; the other three (two confirmed, one presumptive) occurred on the south side of Grand Mesa between 1999-2000. Blood and tissue samples were collected from each of 221 deer mice captured from 1999-2000. Blood samples were tested for IgG antibody to SNV. At least one deer mouse had antibody to SNV at nine of 10 sites. Genomic DNA was isolated from tissue samples and alleles at six microsatellite loci were amplified by polymerase chain reaction (PCR). Polymorphisms were resolved on denaturing polyacrylamide gels and visualized by silver staining. Traditional population genetic analyses of this study indicated moderate population subdivision among the populations surveyed, slight evidence of isolation by distance, and that the Gunnison River system may hinder gene flow in this area. Application of assignment tests indicated that approximately 73-85% of mice were assigned to their population of capture. Many of the misclassifications occurred among sites less than 1 km apart; however, some long-distance misclassifications were noted. Additionally, some misclassifications were noted among study sites on different sides of the Gunnison River system, indicating that the riparian corridor of this system may facilitate some gene flow. Overall, these data indicate that SNV trafficking is more likely at the local level, but some long-distance trafficking may be possible, especially where select habitat variables favor long-distance movements.