Ship rat density in urban Dunedin and the development of a non-invasive estimation method

In New Zealand ship rats (Rattus rattus) are one of the major threats to endemic fauna and flora. Rural ship rat populations have been implicated in the ongoing decline and extinction of many species of endemic wildlife. The role ship rats have in structuring urban ecosystems, directly through predation, and indirectly through food and habitat competition is poorly understood in New Zealand. Understanding the role of ship rats in the urban environment is impeded by a lack of information on their distribution and robust estimates of their density. Rat presence and distribution across different urban habitats was determined by the identification of genus-specific bite marks on wax blocks. The results from the wax block survey suggest that rats are either absent from, or at very low densities within the housed residential sites sampled in this study. In urban bush fragments rats were detected infrequently using wax blocks. High rates of non-target species interference may obscure the rate of rat detection in urban areas. Density is a fundamental biological parameter, however unbiased density estimation can be extremely difficult for certain species. Ship rats are nocturnal and highly dispersed, which makes them particularly difficult to sample using conventional techniques. Currently the most accurate and reliable estimates of absolute ship rat density are obtained through cage-trapping and spatially explicit capture-recapture analysis. This sampling method is both laborious and intrusive. Invasive sampling methods are also not always suitable for use in urban areas. This study describes the application of a non-invasive genetic technique for the estimation of urban ship rat density. Individual genotyping of ship rats was facilitated by analysing nine microsatellite loci amplified from the tissue of ship rat hair follicles. Hair samples were collected using hair-snag tubes (220 mm lengths of 65 mm PVC down pipe). Hair samples were retained on adhesive coated rubber bands that partly occluded the opening at both ends of the hair tubes. Hair tubes were baited with peanut butter and set in a known array. Ship rat density was estimated using spatially explicit capture-recapture analysis (DENSITY 4.1). Maximum likelihood was used to fit a range of candidate models to the spatial dimensions of hair tube re-visitation data. The efficacy of the hair tube methodology was initially verified in the Orongorongo Valley on a well studied population of ship rats. In the Orongorongo Valley the density estimate of 1.17 ± 0.42 (SE) rats/ha was in accordance with recent cage-trapping estimates from the same sampling grid. Very low densities (0.26 ± 0.10 (SE) ha) of ship rats were found in Dunedin urban bush fragments. The overall effect of ship rats as predators on urban birdlife is inferred to be much less than in rural areas, where higher ship rat densities exist. If rats exist in high densities within urban Dunedin it seems likely they do so within small pockets of favourable habitat i.e. areas that are not frequently controlled, where food is abundant or where domestic cat densities are low. Systematic sampling and genetic profiling of ship rat hair for spatially explicit density estimation requires fewer human resources than cage-trapping and provides robust estimates of absolute density, but involves increased costs in laboratory analysis