Comparison of occupancy modeling and radiotelemetry to estimate ungulate population dynamics

Radiotelemetry and unmarked occupancy modeling have been used to estimate animal population growth, but have not been compared for ungulates. We compared white-tailed deer (Odocoileus virginianus) population growth estimates from radiomarked individuals and occupancy modeling of unmarked individuals and evaluated advantages and disadvantages of each method. Estimates of population growth were obtained using remote camera (N = 54/year) detection/non-detection occupancy surveys of unmarked deer and from survival and recruitment data of radiomarked adult females (N = 87) and neonate fawns (N = 127) in a predominantly forested region of the Upper Peninsula of Michigan, USA, 2009-2011. We hypothesized that occupancy models and radiotelemetry data would have similar population growth trends because both methods sampled the same temporally closed population. Percent changes in camera trap data generally reflected finite population growth (lambda) of radiomarked deer which increased (lambda = 1.10 +/- A 0.01) from 2009 to 2010, but decreased (lambda = 0.87 +/- A 0.02) from 2010 to 2011. Also, unmarked adult female abundance and fawn:adult female ratios generally reflected trends in radiomarked deer survival and recruitment. Royle-Nichols occupancy model abundance estimates had wide confidence intervals, which may preclude using this method from accurately estimating deer population growth. Radiotelemetry provided more precise population growth estimates, while allowing collection of vital rates and location data. However, the Royle-Nichols occupancy model may be preferred to radiotelemetry because it reflected yearly variation in population growth with reduced labor and no invasive marking. Researchers should consider the objectives and logistics of their study when choosing a specific method.