Estimating population density and habitat selection of rats (Rattus spp.) and abundance of birds on the Otago Peninsula

Abstract

The introduction of rats (Rattus spp.) and conversion of native forest and scrub have had a devastating effect on the biodiversity of New Zealand’s fauna and flora, threatening many native bird species with extinction. The Otago Peninsula is an area of land characterised by its fragmented landscape and presence of introduced mammalian predators, including rats. This study aimed to estimate the abundance of rats at three sites of fragmented vegetation on the Otago Peninsula using spatially explicit capture-recapture (SECR) and noninvasive genetic sampling. Hair-snag tube grids and microsatellite genotyping were used in an attempt to identify individual rats and estimate population density. The results remain inconclusive in terms of estimating rat population density due to the high degree of genotyping error and inability to identify recaptures. As such, a simple estimate of rat population density at each site was obtained through the inclusion of a boundary strip around the hair-snag tube grid and using a minimum number alive of rats at each site. The boundary strip was calculated to establish an effective trap area to reduce the edge effect in the density calculation. The highest density of rats was found to be at Okia, with approximately 0.9 rats/ha. Leith Walk had a density of 0.7 rats ha-1, while Hooper’s Inlet had an estimated density of 0.3 rats ha-1. In addition to this, a habitat selection analysis was also undertaken. Number of tiers was used as an index of vegetation complexity and was found to be positively associated with probability of rat presence (p=0.020), as was percentage groundcover (p=0.006). While rat species could not be discerned from genotyping, these results suggest the species detected was likely ship rat (Rattus rattus). Distance to coast, presence or absence of fruiting trees, distance to freshwater, maximum canopy height, and site were all found to be nonsignificant variables in predicting rat presence. A tracking tunnel line at each site also gave a rough index of rat abundance. Very low tracking rates of 10% at Hooper’s Inlet and Okia confirmed rat presence in the area. No rat tracks were recorded at Leith Walk. There was no correlation between rat tracking rates and the number of rats detected from the hair-snag tube grids. Mice (Mus musculus) were frequently detected using both sampling methods, with tracking rates of 100% at all three sites. Using 5-minute bird counts 24 bird species were counted across the same three sites in which the rat sampling was conducted. The most abundant bird species were bellbird (Anthornis melanura) at Leith Walk, and goldfinch (Carduelis carduelis) at Hooper’s Inlet and Okia. Distance sampling allowed for analysis of ten of the most abundant species in DISTANCE software to estimate population density of these species. Fantails (Rhipidura fuliginosa) had the highest population density among bird species at Leith Walk (1.043 ha-1), whereas silvereyes (Zosterops lateralis lateralis) had the highest estimated densities of species at both Leith Walk (0.175 ha-1) and Okia (0.163 ha-1). These results provide a foundation for further study of rat density estimations and bird abundance on the Otago Peninsula, and can be used as a baseline when evaluating the impact of future rat eradication, which has been proposed to meet the goal of a Predator Free Peninsula by 2050.