Rodent density indices from tracking tunnels, snap-traps and Fenn traps: do they tell the same story?

Comparisons were made of density indices of free-living populations of ship rats (Rattus rattus) in mixed forest in New Zealand by using footprint tracking tunnels and two kill-trapping methods. Tracking tunnels and snap-trap removal indices of rat densities showed similar trends when run on a 9 ha trapping grid, although immigration onto the grid occurred, thus violating one of the assumptions of the analysis. Tracking rates and snaptrap capture rates were not significantly correlated when run along a trapping line for a 12 month period, although tracking rates and the total number of rats caught in a trapping session were significantly correlated. Time series analysis showed that rat density indices from tracking tunnels and Fenn traps were significantly correlated when run for 27 consecutive months in a rat population with moderate density, but were not correlated in a low density rat population. The findings highlight the importance of habitat, sample size and target species behaviour in influencing relative density indices obtained from tracking tunnels, snap-traps and Fenn traps. Given the widespread use of rodent tracking tunnels in New Zealand, we suggest that tracking tunnels should only be used to compare relative abundance within similar habitat types, and should always be complemented with a second density index. The relationship between the commonly used density indices and true rodent population density requires urgent attention.