Camera traps are an effective method for identifying individuals and determining the sex of spotted-tailed quolls (Dasyurus maculatus gracilis)

Abstract

We compared two bait station techniques for determining the sex and identifying individual spotted-tailed quolls (Dasyurus maculatus gracilis) using images taken by camera traps. One method used bait in a plastic mesh bag and the other was a new method using a raised bait canister to entice the quolls to stand on their hind legs and present their ventral surface to the camera. Individuals were identified from multiple images of their unique spot pattern, and sex was determined from ventral images. The bait bag method was better for detecting quolls and both methods performed similarly in allowing observers to identify individuals from images. However, the bait canister method was superior for determining sex of individuals. Using this new bait canister method, individual identification was possible in 202 out of 206 detection events and the sex of 81% (47 of 58) of identified individuals was confidently assigned from multiple detections. This bait station design can therefore provide additional data on individual quolls and reduces the need for more invasive live-trapping techniques. This methodology could be adapted for other mammals in Australia and worldwide.

Camera‐trapping density estimates suggest critically low population sizes for the Wet Tropics subspecies of the spotted‐tailed quoll (Dasyurus maculatus gracilis)

“…Continued annual monitoring using standardized camera‐trapping methodologies (e.g. Rowland et al, 2020a) should be undertaken at Windsor Tableland, Carbine Tableland, Lamb Range (particularly regarding extremely low numbers), and South Johnstone. Surveys in the latter area should also extend to adjacent high‐elevation habitat of the Walter Hill Range and uplands on the northern side of Tully Gorge, to better resolve the distribution and size of this subpopulation.…”

“…All cameras were programmed to take three still images per trigger, with either a 0 or 1 s gap between trigger events. The differences between camera models were considered minimal for detecting quolls because they are more readily detectable being a medium‐sized mammal and because activity time at bait stations is generally 2–10 min (see Rowland et al, 2020a). At each camera station, a single camera was placed on a tree at a height of 10–50 cm above the ground, facing horizontally but with a slight downwards angle towards a bait.…”

“…Individual quolls were identified from their distinct spot patterns, sexed, and tagged from camera‐trap images using the camera‐trap software program Camelot (Hendry & Mann, 2018), following the protocols of Rowland et al (2020a). Similarly, as per Rowland et al (2020a), we defined a single detection event as any detection of a quoll within 10 min of a previous detection of that same individual. A separate detection event occurred when a different quoll appeared in the 10‐minute period or when the same individual was photographed more than 10 min after the previous detection event of that individual.…”

“…'bait bag'), or bait in an open-ended, enclosed PVC cylinder (i.e. 'bait canister') (Rowland et al, 2020a). The bait bag method was used in earlier surveys but was generally superseded by the bait canister method in later surveys (Table 1) due availability as well as the better resolution on individual identification and sex offered by the bait canister method (Rowland et al, 2020a).…”

Optimising camera trap height and model increases detection and individual identification rates for a small mammal, the numbat (

“…Camera traps may also be useful in identifying individual animals for species with unique markings, such as tigers, leopards, hyenas and Australian dasyurids (Karanth 1995;Jackson et al 2006;Hohnen et al 2013;Tichon et al 2017;Rowland et al 2020). Identifying individuals allows the use of capture-recapture modelling to obtain robust population size estimates (Royle et al 2014).…”