A Simple, Quantitative Determination of the Appropriate Duration of Site Closure to Improve Occupancy Modelling for Nomadic Species

Aim

Occupancy models are ubiquitous in ecological and biogeographical applications, but they rely upon an assumption of site closure that is sometimes applied across many weeks or even months. Nomadic species challenge this approach because they are likely to exhibit within-season movements that violate the assumption of closure. Assuming closure for nomadic species across a full season likely results in depressed estimates of detection probability and inflated estimates of occupancy that can obscure important habitat associations. However, selecting an appropriate duration over which to assume site closure can also be a challenge, especially with continuous survey methods, like acoustic surveys or camera traps, where ‘surveys’ and ‘seasons’ can be ambiguous.

Innovation

We present a simple framework for quantitatively assessing the most appropriate duration over which to assume site closure by directly comparing models with differing season lengths. We demonstrate our framework using simulated data, as well as a passive acoustic monitoring dataset of a nomadic bird, the Clark's nutcracker, collected during one summer across over 25,000 km² of the Sierra Nevada, California. We applied dynamic occupancy models to data collected during a single calendar season to optimise the occupancy approach for the nutcracker.

Main Conclusions

We used our method to select an appropriate duration over which to assume site closure for the Clark's nutcracker, and our simulations indicated that this method is robust under a variety of scenarios. Our empirical and simulated results suggest that quantitatively determining the duration of the closure assumption can enable a more accurate understanding of the population dynamics and habitat use of nomadic species, especially when surveys are conducted with emerging continuous, or near-continuous monitoring technologies. Our relatively simple framework can be used to improve occupancy modelling for other nomadic species, which will ultimately improve the efficacy of conservation measures taken to protect these species.