Density Estimation

Wildlife abundance estimates are rarely a ‘total count’, as it is uncommon for all individuals within a population to be available (detectable) for counting at one time. Kangaroos in some small, open reserves in the ACT are an exception to this rule and it is possible in some circumstances to count each individual to get a total count. In other reserves, however, larger areas and dense vegetation mean a ‘sampling’ approach is required, where surveys collect data which allows a density to be estimated statistically with a ‘confidence interval’ which describes how much error is likely to be associated with the count.

Four main methods are used by the ACT Government to estimate the abundance of Eastern Grey Kangaroos in the ACT. These include direct counts, sweep counts, faecal pellet counts and walked line transect ‘distance’ sampling.

Direct counts: involve individual animals being counted from observers moving slowly through a reserve, coordinated by radio. Often, an additional observer is positioned on a hill overseeing progress and monitoring the movement and/or kangaroo groups to ensure no groups of animals are missed or counted twice.

Sweep counts: involve up to 30 surveyors spanning a reserve and moving forward as a line, tallying kangaroos as they pass through the line and into the counted area behind. Each observer counts kangaroos which pass between themselves and the person on their left, and the line is coordinated to ensure that all kangaroos pass through the line without being accidentally ‘herded’ out of the survey area.

Pellet counts: involve newly accumulated faecal pellets being counted in a large numbers of stratified survey quadrats (~240 per reserve). Kangaroo density is then mathematically estimated based on published or measured defecation rates.

Walked line transect ‘distance’ sampling: involves the distance and bearing to kangaroo groups being surveyed from approximately 40 km of transects per reserve, enabling a density to be estimated based on statistical modelling for the surveyed area. ‘Detection functions’ estimated by the Distance software consider the density of vegetation as part of the analysis, to allow for a higher likelihood of spotting kangaroos in open compared to forested habitat.

Details of each of these methods are described in more detail in Appendix 1 of the Eastern Grey Kangaroo: Controlled Native Species Management Plan. In general, an error rate less than 10-15% is considered desirable across all counting methods. All kangaroo counting procedures are approved under an animal ethics permit and have been reviewed by an external reviewer.

Photo: Kangaroo density is precisely measured from the density of faecal pellets compared to other sites with known kangaroo density and similar food supply.

For more information about this research call Access Canberra on 13 22 81.