Estimating population growth rates from tracking tunnels

Tracking tunnels are widely used for monitoring small mammal populations, but inference on population size is hindered by the non-linear relationship between frequency of detection and abundance. The detection-abundance relationship can be used to relate the probability of a tunnel being tracked to the abundance of animals in the population and the population growth rate. We describe equations that show that a complementary-log-log link transformation for analysis of presence-absence tracking data allows estimation of small mammal population growth rates and proportional differences in abundance at different sites or under different treatments, as well as estimation of the rate at which animals pass through tunnels. We demonstrate the utility of this type of model to long-term temporal monitoring data of rat irruptions during beech mast seeding events, and show that the results are similar to theoretical growth rates of rat populations under such scenarios. This form of analysis yields biologically interpretable parameters and thus is an improvement on most current analytical techniques; however, it still requires the assumption that the probability of detection remains constant or is independently modelled, and it can still not be used to make inference on absolute population size.