Is Australian wildlife fair game?
Box 1 | Kangaroo counting
The harvesting of kangaroos has generated considerable controversy. Some people say it could threaten the viability of wild populations, but others say a sustainable harvest will keep numbers in check while generating revenue for rural economies.
It’s an argument that good scientific data might help resolve. How many kangaroos (of various species) are there? Are numbers going up or down? How many might we cull without endangering species?
But there’s also a more basic question: how on earth do you count kangaroos? There are many different species; they occupy a wide range of habitats across much of the continent; and there are probably millions of them.
The answer is you don’t do it on earth – you do it from the air. Since the 1970s scientists have been counting kangaroos from aeroplanes, using a technique known as fixed-wing aerial survey. They don’t count every single kangaroo – they take samples, which are used to estimate the total population.
The exact details of the survey vary depending on the purpose of collection and the agency conducting it. The technique described here is just one variation and is based on that used for an Australia-wide survey.
Map marking
The first part takes place in the office. Using maps, the survey team divides Australia (or a given part of it, such as a State) into convenient blocks or strips: the Australia-wide survey used blocks of 1º latitude by 1º longitude in size. The area of this kind of block varies depending on latitude but is around 10,500 square kilometres. The team then marks out transects, or lines, in an east-west direction, wide enough apart so that the chance of kangaroos being ‘double-counted’ by moving from one transect to another is minimal.
Aerial survey
These transects form the basis of the sample survey, which is usually conducted in a high-wing single-engined Cessna. The plane maintains a height of 250 feet (about 76 metres), a speed of 100 knots (about 185 kilometres per hour) and a course that follows the transects marked on the map. Two people, one on each side of the aircraft, count the kangaroos they see in a 200-metre strip on the ground delineated by streamers attached to the wing-struts of the plane. Each counting period lasts for 97 seconds, during which the plane travels a distance of 5 kilometres. Thus, an area of 1 square kilometre (called a ‘unit’) is surveyed in each period by each observer.
Correction factors
It is very unlikely that even trained observers will count all the kangaroos in a unit – some might be hidden by vegetation and some might simply be missed because of the speed at which the plane is travelling. By comparing aerial counts with those obtained by ground-level surveys, scientists have demonstrated that a correction factor must be applied to the raw count to obtain an estimate of actual numbers. For red kangaroos, this is 2.29 in open vegetation, 2.36 in light cover, 2.43 in medium cover and 2.57 in dense cover. For western grey kangaroos, the correction factor is 4.8 in all habitats. Other correction factors apply to other species.
Experiments at ground level have also shown a relationship between temperature and the ‘countability’ of kangaroos, since the animals are less active (and therefore less visible) at high temperatures. The temperature correction factor is expressed in the following equation:
C = 1/(1474 – 0.0316x)
where C is the correction factor and x is the temperature (if it is above 15ºC).
An example
The following example is based on real data for the survey block 30-31º latitude and 144-145º longitude, an area of 10,663 square kilometres. Counts were made in two transects of 17 units (of 1 square kilometre) by two observers: these gave a total of 244 red kangaroos and 57 western greys in an area of 68 square kilometres (17 units × 2 observers × 2 transects). Note that there are statistical uncertainties associated with these counts. The 244 count has an uncertainty of about 6 per cent (because a count of another similar area might be expected to differ by up to this amount) and the 57 count similarly has an uncertainty of about 13 per cent. These uncertainties will affect the reliability of the final estimates of the number of kangaroos.
Correction factors
Applying the standard correction factors of 2.43 and 4.8 to the counts produces estimates of 593 and 274 for reds and western greys respectively. The average temperature during the count was 20ºC. Applying the temperature correction factor we get:
Number of red kangaroos = 593/(1.474-0.0316 × 20) = 704.
Number of western greys = 274/(1.474-0.0316 × 20) = 325.
These are the numbers of animals estimated to have actually been present during the survey in all the units combined.
Density of animals
By dividing these numbers by the area surveyed (68 square kilometres) we calculate the corrected average number of kangaroos per unit, or average density (R). Thus:
R (red kangaroos) is 704/68 = 10.4 animals per square kilometre.
R (western greys) is 325/68 = 4.8 animals per square kilometre.
Estimate of total numbers
The total number of animals in the block is represented by the equation:
E = RZ
Where E is the estimate of total numbers in the degree block, R is average density, and Z is the area of the block. Thus:
E (red kangaroos) = 10.4 × 10,663 = 111,000 animals.
E (western greys) = 4.8 × 10,663 = 51,000 animals.
where these numbers have been rounded off to allow for uncertainties in the original counts.
By adding the estimates for all the blocks together, we can produce an estimate for the total number of kangaroos in the region, State or continent. We can also use the data to generate an estimate of variance, which is a measure of reliability.
Review of survey technique
Scientists continue to review this survey technique. Helicopter surveys, while more expensive than fixed-wing surveys, may be more reliable. The correction factors, too, are widely debated, and research continues to refine them. This is important – the science of counting plays a fundamental role in our ability to manage kangaroos in a sustainable manner. We need to know we can count on the counters.
Box
Box 2. Principles relevant to the harvesting of native species
Related site
Estimating population size and density (Box 3 of Nova: Science in the news topic, A plague on the pest – rabbit calicivirus disease and biological control)
Posted February 2000.