Occupancy in continuous habitat

Abstract

Abstract. The probability that a site has at least one individual of a species ('occupancy') has come to be widely used as a state variable for animal population monitoring. The available statistical theory for estimation when detection is imperfect applies particularly to habitat patches or islands, although it is also used for arbitrary plots in continuous habitat. The probability that such a plot is occupied depends on plot size and home-range characteristics (size, shape and dispersion) as well as population density. Plot size is critical to the definition of occupancy as a state variable, but clear advice on plot size is missing from the literature on the design of occupancy studies. We describe models for the effects of varying plot size and home-range size on expected occupancy. Temporal, spatial, and species variation in average home-range size is to be expected, but information on home ranges is difficult to retrieve from species presence/absence data collected in occupancy studies. The effect of variable home-range size is negligible when plots are very large (.100 3 area of home range), but large plots pose practical problems. At the other extreme, sampling of 'point' plots with cameras or other passive detectors allows the true 'proportion of area occupied' to be estimated. However, this measure equally reflects home-range size and density, and is of doubtful value for population monitoring or cross-species comparisons. Plot size is ill-defined and variable in occupancy studies that detect animals at unknown distances, the commonest example being unlimited-radius point counts of song birds. We also find that plot size is ill-defined in recent treatments of ''multi-scale'' occupancy; the respective scales are better interpreted as temporal (instantaneous and asymptotic) rather than spatial. Occupancy is an inadequate metric for population monitoring when it is confounded with home-range size or detection distance.

Population mapping of gibbons in Kalimantan, Indonesia: correlates of gibbon density and vegetation across the species’ range

“…The primary aim of SECR is to estimate the population density of free-ranging animals (Efford et al 2004, Efford & Dawson 2012. Animals are assumed to be distributed independently in space and to oc cupy home ranges.…”

Can Camera Traps Monitor Komodo Dragons a Large Ectothermic Predator?

“…This will mean validating different model assumptions, reducing potential detection biases present within the existing study design, and most importantly, quantifying if site occupancy estimates are sensitive and accurate enough to provide good measures of changes in population status. Efford and Dawson [55] recently demonstrated several spatial related issues that can make estimating occupancy problematic and even inadequate. In particular, poor consideration of home range characteristics of animals, or situations where the effective area of each sampling location is unknown can cause considerable bias in occupancy estimates [55].…”

“…Efford and Dawson [55] recently demonstrated several spatial related issues that can make estimating occupancy problematic and even inadequate. In particular, poor consideration of home range characteristics of animals, or situations where the effective area of each sampling location is unknown can cause considerable bias in occupancy estimates [55]. For example, we know Komodo dragons have highly overlapping and variable home ranges.…”

Best Practice Framework and Principles for Monitoring the Effect of Coastal Development on Marine Mammals

“…Consequently, a relatively large area can be monitored for presence/absence or counts for the same total cost as a smaller area for capturerecapture surveys. While occupancy models (e.g., MacKenzie et al, 2006) may seem to offer a very efficient way to provide a snapshot of marine mammal distribution and the probability that they occupy certain types of habitats throughout their range, a critique of occupancy in continuous habitat by Efford and Dawson (2012) suggest that occupancy models may not always be informative. Occupancy is ill-defined when site sizes are arbitrarily defined in continuous habitat relative to the typically unknown home range sizes of marine mammals and estimates of the probability of occupancy by habitat types may amount to no more than estimates of relative density.…”