Population viability of Rorippa columbiae: Multiple models and spatial trend data

In managing rare species for conservation, we often need to make decisions based on existing, scanty data sets. Our ability to use such data effectively may be critical to our success in managing individual populations. We used viability analysis to examine the population dynamics of a threatened plant species, Rorippa columbiae, on Pierce Island in Washington (U.S.A.). Although the island was protected in part because of the presence of R. columbiae, monitoring suggested that plant density declined on the island after the preserve was established. To understand the implications of this decline, we generated a number of hypotheses about factors that might be important in controlling population dynamics in R. columbiae: density dependence, size dependence, spatial environmental heterogeneity among sites, and dispersal and colonization of new sites. Using statistical and biological criteria, we were able to rule out the hypotheses of size-dependent population growth rates and dispersal among sites, but the importance of spatial heterogeneity and density dependence was ambiguous. Models with and without spatial heterogeneity make radically different predictions about the future of R. columbiae on Pierce Island: with spatial heterogeneity we expect persistence, albeit at a limited number of sites, whereas in a spatially uniform environment we expect extinction in the near future. This emphasizes the uncertain future of this population but tells us that ecological differences among sites and correlates of population vigor are important data to collect in future studies Our analysis also points to the importance of maintaining multiple working hypotheses about controls of population growth when interpreting population trends from limited data.