RESUMO
Many researchers have speculated that infection dynamics of Sin Nombre virus are driven by density patterns of its major host, Peromyscus maniculatus. Few, if any, studies have examined this question systematically at a realistically large spatial scale, however. We collected data from 159 independent field sites within a 1 million-hectare study area in Nevada and California, from 1995-1998. In 1997, there was a widespread and substantial reduction in host density. This reduction in host density did not reduce seroprevalence of antibody to Sin Nombre virus within host populations. During this period, however, there was a significant reduction in the likelihood that antibody-positive mice had detectable virus in their blood, as determined by reverse-transcriptase polymerase chain reaction. Our findings suggest 2 possible causal mechanisms for this reduction: an apparent change in the age structure of host populations and landscape-scale patterns of host density. This study indicates that a relationship does exist between host density and infection dynamics and that this relationship concurrently operates at different spatial scales. It also highlights the limitations of antibody seroprevalence as a metric of infections, especially during transient host-density fluctuations.