The New Zealand archipelago is under constant invasion pressure by a great variety of introduced species, with one of the most notorious island pests being the black rat (Rattus rattus). Conservation managers particularly need to know if individuals detected after an eradication are survivors or re-invaders, particularly when islands exist in a meta-population with ongoing reinvasion.
Goat Island (<10ha) is an island off the east coast of New Zealand, which is separated from the mainland by a small 50m water channel at low tide. Black rats first invaded the island before the 1970s and despite repeated annual eradication attempts, rats are constantly redetected on the island.
Using molecular tools this study aims to determine which proportion of these individuals are survivors or swimming re-invaders. Tissue samples from the island and the adjacent mainland have been collected during the years 2011-2013, with eradications carried out on the island in 2012&2013. For comparison we included the microsatellite dataset collected in 2005 by Russell et al. We assessed the genetic diversity and population structure using 14 polymorphic microsatellite loci.
We combined Bayesian clustering approaches (STRUCTURE) with conventional F-statistics to assess the gene flow and structure between putative source and sink populations. Contrary to our expectations, the pairwise FST estimates show significant genetic differentiation between the sampling sites. The clustering results suggest that all sampled individuals belong to three different gene pools (K=3), with two distinct clusters being found for the island population. Interestingly, the island individuals share no ancestry with the distinct mainland population, despite the close proximity to the mainland.
This might be explained by the water barrier clearly restricting gene flow and acting as dispersal barrier. However it is more likely an incumbent advantage of the resident island population resisting new mainland invaders, until its own density is suppressed by annual control programmes.