Study area
The area of interest for this study comprised 15 local government areas (LGAs) located in South-East Queensland (Figure 1). South-East Queensland has a higher population of koalas compared with other areas of Queensland (Dique, Preece, Thompson and Villiers 2004), but has also a high urban development. The point locations of koala sighting events (details provided below) were plotted on a map of South-East Queensland, indicating relatively few sightings in the North and to the far West of Queensland. To develop an observation window that was tractable for modelling we constructed a convex hull around all koala sighting locations and then dilated the convex hull by 3 kilometers to accommodate the home range of koalas sighted on the border of the observation window (de Oliveira, Murray, de Villiers, & Baxter, 2014). Refinement of the study area in this way reduced the number of constituent LGAs from 15 to 11 and included parts of 3 remaining LGAs. The selected LGAs were in the Eastern and Central part of South-East Queensland. The northern LGA of the Fraser Coast was excluded due to an absence of koala sighting records. This reduced the study from 57,800 square kilometers to 30,500 square kilometers.
Figure 1. Local Government Areas (LGA) of South-East Queensland, Australia (N=15). The LGAs for which koala population densities were calculated are shown in the rectangle (Green areas represents the main koala conservation in South-East Queensland).

Koala sighting data

Incidental koala sightings reported in South-East Queensland between 1997 and 2013 (N = 14, 250 sightings) were retrieved from KoalaBASE, a database of clinical admission and sightings data of koalas, which developed by the University of Queensland’s School of Veterinary Science (www.koalabase.com.au). The dataset comprised of the date of the sighting as well as the longitude and latitude of the location, where the koala was observed. We assumed that if two koalas were reported in the same area in the same month, these could be repeated sightings of the same animal. The term ‘area’ in this context represents the estimated home range for a koala, which was assumed to be circular with an average size of 0.35 km2 (de Oliveira, et al. 2014). Thus, the minimum distance between two koalas was given by\(d=2\ \times\ \sqrt{A\ \div\ \pi}\), or 666 metres. Figure 2a shows a selection of koala sighting locations for month 1, which are minimum of 666 metres apart and excluded other sightings for the month, while Figure 2b shows a selection of koala sightings for month 2. Because home ranges of koalas can overlap (Ellis, et al. 2009), Figure 2c shows the two monthly data sets combined. In Figure 2c sightings are permitted to be closer than the minimum distance set for individual monthly observations, acknowledging that koala home ranges overlap here. Following this approach, monthly sightings observations for each year over the 17 years period were compiled, resulting in a total of 17 years × 12 months = 204 selections. By adopting this approach, we permitted a koala (i.e. the same koala or a different koala) to be seen in the same area at a maximum of 12 times per year.
Figure 2. Illustration of the method used to select incidental koala sightings to estimate koala sightings density. Circles represent an average home range area for a koala. Figure 2a shows a selection of koala sighting locations for month 1, while Figure 2b shows a selection of koala sightings for month 2. Figure 2c shows the two monthly data sets combined. Following this approach, monthly sightings observations for each year over the 17 years period were compiled and used a subset of sightings to estimate koala sightings density.