Rum rats: ecology and behaviour

The Isle of Rum hosts approximately 25% of the global population of Manx shearwaters, Puffinus puffinus. Potential for negative impacts on the shearwater colonies from introduced Norway rats, Rattus norvegicus, has highlighted the need to better understand the ecology of this island invader. Abundance and population estimates are often central to effective conservation and invasive species management. Despite their potentially catastrophic impacts on seabird populations, the ecology of Norway rats on islands is poorly understood. Developing an appropriate tool to measure rat population size is therefore important. Furthermore, an understanding of the movement patterns and diet of introduced Norway rats is vital to establishing the nature of their relationship with native species. Optimisation (transect duration/length) of two common rodent activity survey techniques (tracking plates and chocolate-flavoured wax blocks) was carried out to derive parameters that resulted in a standard deviation of the bootstrapped distribution of the mean (precision) of less than 0.2. Activity survey scores were then calibrated against capture-mark-recapture population estimates for linear (coast, river, wall) and non-linear (grassland, heath, woodland) sites. An island-wide population estimate of Norway rats on Rum was calculated using extrapolated density estimates for linear features (boundary strip method) and non-linear vegetation types (SECR models). Home range size of 17 Norway rats, across three study sites (coast, heath, river) was estimated. Jacob’s Index was used to compare vegetation availability (home range) with use (location points), and Spearman’s correlation for a relationship between capture rates (C100TN) and distance to linear features. Stomach content analysis was used to describe the diet of 29 rats across three ecotopes (coast, grassland, urban) on Rum. Activity indices reached an acceptable level of precision after three days and a grid size of 35 x 10m2 from tracking plate (linear and non-linear sites) and wax block (non-linear sites only) surveys. Total population size of Norway rats on Rum was estimated at 11,844 ± 5,685 for non-linear/linear sites combined. Mean home range size (MCP100) was estimated at 2.32 ± 0.86 hectares, with differences between sex and site found. Distance to coast had a strong negative correlation with mean capture rates. Vegetation and invertebrates were the most frequent/abundant food types across all ecotopes. Calibrated tracking plates are preferable for the use of estimating population size of Norway rats on islands. Coastal environments are likely to be an important resource for introduced rats; woodland may offer intra-island refugia to rats on Rum. Overall, Norway rats are unlikely to be a current threat to Manx shearwaters on Rum.