Monitoring populations is a key component of wildlife conservation and management. This study aimed to evaluate the effectiveness of abundance indices used for estimating populations of the bush rat Rattus fuscipes. The study was conducted on Fraser Island, Queensland, Australia. A trapping grid consisting of 5 x 5 trap stations, 20 m apart, was used because this configuration was typical of those used for small mammal trapping in Australia. Capture-recapture data were collected from trapping 18 grids over four consecutive nights, immediately followed by two consecutive nights of conducting oil card and track indices. On 11 grids, model M_b was selected by the CAPTURE model selection procedure, indicating increased capture probability after first capture (trap-happy). The estimated trappable population size (N) on these grids was not precise with the standard error on average being 28 of N. On the other seven grids, the number of animals caught was too small (<3) to estimate model parameters. Themain problem of the grid trapping was that some grids caught too few animals to estimate N. Our data suggest that the minimum grid area for estimating useful N is one that would catch more than 8 individuals. The minimum grid area varies depending on thedensity of rats and should be determined by a pilot trapping study. The number of trapping occasions can be increased until a specific precision of N is reached. The correlation between track index and N was near the 0.05 significance level indicating tracks are potentially a reliable index for monitoring bush rat populations. The oil card index was not significantly correlated with N.
展开▼
