Commercial harvesting of Kangaroos in Australia
by Tony Pople and Gordon Grigg
Department of Zoology, The University of Queensland
for Environment Australia, August 1999
Chapters 10,11,12 and 13 and Appendix 1 provided by staff at Environment Australia
BIOLOGY OF THE HARVESTED SPECIES
It should be obvious that a good working knowledge of the biology of any harvested species is required in order to have any sustainable commercial exploitation put onto a sound basis and, indeed, a requirement in the regulations (Paragraph 5.1.a) is
'.... that there is available to the Designated Authority sufficient information concerning the biology of each species subject to the management program and the role of that species in the ecosystems in which it occurs, to enable the Designated Authority to evaluate a management program for that species.'
In the case of the harvested kangaroo species, especially the larger ones, there is a vast amount of information about their biology in general and their ecology in particular, and some of it is reviewed in this document. A comprehensive review would be a very daunting task. All we can do here is to try to provide an overview and to refer the interested reader to other general and specific sources. A very useful starting point is Hume et al. (1989), in Volume 1B of The Fauna of Australia (Walton and Richardson 1989). Other general reference material is listed at the end of Chapter 1.
Changes to macropod populations since the arrival of Europeans.
This topic was reviewed by Calaby and Grigg (1989). The overall results are summarised in Table 1. Of the 50 or so macropod species in Australia, 30 have suffered significant declines in geographic range or numbers and 14 are thought to have increased or remained essentially stable. Six species have declined to extinction and an additional four are extinct on the mainland. Only about half, 28 species, remain common where they now occur and some of them have become very common indeed, to the extent of being regarded as pests by many landholders.
Habitat changes are likely to have been the major causative agents of change, whether positive or negative, the most important being the clearing of forests and habitat changes associated with pastoralism and grazing, with predation by introduced feral animals thought to be of lesser significance.
All of the species which are harvested on the mainland and for which there is are management programs approved by the federal government are species which are thought to be more abundant now than they were at the time of European arrival.
The harvested species, their habitats and distributions
The five species which are the subject of the present overview are abundant over a generally broad area of the continent. In particular, the three most abundant species (M. rufus, M. giganteus and M. fuliginosus) which make up 95% of the commercial harvest are particularly common over the sheep and cattle grazing pastures of Australia (Figures 3, 4 and 5). Red kangaroos are the most arid adapted and their distribution is more inland than the two species of greys. The distributions of wallaroos and whiptail wallabies are shown in Figure 6 and Figure 7.
It is noteworthy that the highest densities occur within that part of Australia known as the sheep rangelands (Figure 8), which have about 15% of Australia's sheep, and that the densities of red kangaroos drop off quite sharply outside the dingo-proof fences (Figure 9). Part of the reason for that may be that the fences were constructed near the limits of the habitat judged suitable for sheep, and that red kangaroos too may do better within those limits. Mainly, however, it is thought that the presence of dingoes beyond the fences exerts control over kangaroo numbers (Caughley et al. 1980; Newsome 1990; Pople et al. in prep), especially as watering points and pasture are available outside the fences as well.
Within the sheep rangelands, the provision of permanent watering points meant that kangaroos were now more likely to be limited by food than water (Oliver 1986). This would have had a profound effect upon their distribution as well as their abundance (Newsome 1965a). It has been suggested that sheep and cattle also improved the habitat of kangaroos through facilitative grazing; creating a sub-climax pasture (Newsome 1975). Such a relationship needs to be contrasted with the strong evidence of competition between kangaroos and domestic stock below certain pasture levels (Edwards 1989; see above). These changes to the environment would have been most pronounced in the late 1800s when average sheep numbers in the rangelands of New South Wales were nearly twice what they are today (Caughley 1976). Other changes were also wrought upon Australia's rangelands following European settlement. Numerous species of eutherian herbivores and predators were introduced and became established in the wild. At the same time numerous native small mammal species disappeared and many are now extinct. As Caughley (1987b) explained, not only was the habitat modified, but the ecological system was 'changed beyond recognition'. The current distribution and abundance of kangaroos may therefore bear only a vague resemblance to what it was prior to European settlement.
Figure 3. Density and distribution of red kangaroos determined from aerial surveys in 1980-82 (after Caughley 1987b).
Density and distribution of eastern grey kangaroos determined from aerial surveys in 1980- 82 (after Caughley 1987b).
Density and distribution of western grey kangaroos determined from aerial surveys in 1980- 82 (after Caughley 1987b).
Figure 6. Distribution of common wallaroos (after Strahan 1996).
Figure 7. Distribution of whiptail wallabies (after Southwell and Fletcher 1989).
Figure 8. Sheep distribution in Australia in 1977 (after Caughley 1987b).
Figure 9a. Australian states.
Figure 9b. Major vermin-proof fences in Australia.
Macropus rufus (red kangaroo)
In geographic terms, red kangaroos occur across the continent west of the Great Dividing Range, but excluding Cape York, Arnhem Land, the Kimberley region, the south-west corner and Tasmania (Figure 3). This distribution coincides generally with mulga and mallee scrub, shrubland, woodland, grassland and desert, vegetation typical of the dry inland. M. rufus appears to prefer open plains with scattered trees. It is a primary consumer preferring to graze on green grasses and dicotyledonous plants. It is very mobile and is unrestricted by pasture fences. The range of red kangaroos reflects an interaction between mean annual rainfall and mean annual temperature, lying between the 400 mm isocline of mean annual rainfall in the south and 700 mm in the warmer north of the country (Caughley et al. 1987b). Within this range, the distribution of red kangaroos has been described for the sheep rangelands of New South Wales (Sinclair 1980), Western Australia (Short et al. 1983) and South Australia (Cairns et al. 1991). The common pattern in each state was for the highest densities to occur in the more arid regions, usually within mulga woodlands. Low densities occurred in the less productive and poorly watered mallee country, and in cultivated land where there is little available shelter (Short and Grigg 1982). Areas comprising a mosaic of land systems also appear to be important to red kangaroos because they provide a habitat in which animals can seek the best pasture and most appropriate shelter (Pople 1989).
Most scientists are of the opinion that the clearance of trees, provision of artificial watering points and reduction of dingo numbers to facilitate the grazing of domestic stock in the pastoral zone have 'improved' the habitat for M. rufus and hence resulted in a general population increase from pre-European times (Russell 1974; Newsome 1975; Caughley et al. 1980; Squires 1982; Grigg 1982). Intensive agriculture, on the other hand, is not beneficial to the species (Grigg 1982; Short and Grigg 1982), but little of M. rufus habitat has been altered by intensive agriculture. Apart from these density changes, there have been no obvious shifts in the boundaries of the distribution of red kangaroos during this century.
Macropus giganteus (eastern grey kangaroo)
Eastern grey kangaroos have an almost continuous distribution down the eastern seaboard where annual rainfall exceeds 250 mm but has little seasonal trend or where summer rain exceeds that in winter (Caughley et al. 1987b). This distribution includes all of Queensland (except western Cape York), New South Wales, Victoria and north-eastern Tasmania (Figure 4 ). Its habitats, from the inland plains to the coast, include semi-arid mallee scrub to woodland and forest (Caughley 1964; Calaby 1966; Bell 1973; Russell 1974; McCann 1975; Taylor 1980; Hill 1981; Southwell 1987; Strahan 1995). Pastoral development has led to a marked increase in populations of this species.
The occupied habitat of grey kangaroos is approximately the whole of their potential habitat. In the more arid areas they may be confined to narrow belts of woodland bordering watercourses (Poole 1975) and densities may be particularly low away from rivers and creeks in areas developed extensively for cropping and intensive pastoral activities (Short and Grigg 1982). Eastern grey kangaroos also inhabit areas of higher rainfall between the inland plains and the coast, particularly semi-arid mallee scrub, shrub woodland and forests.
Historically, the western boundary of the distribution of the eastern grey kangaroo was probably determined by increasing aridity. However, since European settlement, it has moved substantially farther west as a result of the establishment of numerous semi-permanent watering points for stock. According to Caughley et al. (1984) this westwards expansion is occurring at a rate of 3-4 km year-1. Cairns and Coombs (1992) described the most recent extension of the western boundary.
The ranges of the eastern grey and the western grey overlap over a large area of central and western New South Wales and a small area of south-western Queensland (Caughley et al. 1984).
Macropus fuliginosus (western grey kangaroo)
Macropus fuliginosus was confirmed as a separate species from M. giganteus only after detailed investigation of electrophoretic, serological, morphological and reproductive evidence (Kirsch and Poole 1967, 1972). The need for such approaches to confirm the difference between them is a reflection of the very similar biology of the two species. Poole (in Strahan 1995) in reviewing information on M. fuliginosus commented that many aspects of the species' biology are so similar to M. giganteus that they hardly needed to be described separately.
Based on morphology, two subspecies have been recognised, M. fuliginosus fuliginosus on Kangaroo Island and M. f. melanops on the mainland (Poole et al. 1990). The mainland subspecies has been further divided, but appears best represented as a cline rather than distinct subspecies (Poole in Strahan 1995).
The western grey kangaroo is, perhaps, named inappropriately because the species actually occurs across the south of the continent, with a distribution extending northwards through western New South Wales and into a small area of southern central Queensland(Figure 5). It is absent from Tasmania. This distribution corresponds to areas of aseasonal or winter rainfall (Caughley et al. 1987b). Where M. fuliginosus overlaps in its range with M. giganteus, the latter is more abundant. Both species have similar habitat preferences and M. fuliginosus, too, has been advantaged by pastoralism but disadvantaged by intensive agriculture (Short and Grigg 1982).
Macropus robustus (common wallaroo or Euro)
Macropus robustus has the widest distribution of the larger macropodid species (Figure 6). It occurs across the entire mainland continent, and is only absent from the extreme northern and southern portions of the continent (Russell 1974; Strahan 1995). Within this distribution it is regarded as abundant (Strahan 1995).
Common wallaroos show considerable variation in external characteristics such as coat colour, coat texture and ear length. In the most recent review of the species, Richardson and Sharman (1976) suggested that the number of sub-species recognised should be reduced to a maximum of four which they considered reflected the extremes of the variability present. Only one of these sub- species occurs in South Australia, two are found in Queensland and three occur in Western Australia.
The validity of at least two of these sub-species (M. robustus robustus and M. robustus erubescens) is questionable as both forms intergrade into one another over a broad area of Queensland and consequently do not fulfil the criteria established by Mayr (1966) as requisite for recognition of variants as sub-species. Rather, the situation as found in Queensland is more consistent with the concept of clinal variation.
Macropus robustus occupies a wide range of habitats but prefers areas where steep escarpments, rocky hills or stony rises (Calaby 1966; Kirkpatrick 1968; Russell 1974; McCann 1975; Taylor 1985; Strahan 1995). Such terrain has many caves or overhangs which may provide some shelter from extreme temperatures. Newsome (1975) considered that the alteration of vegetation communities to sub-climax spinifex by the grazing of sheep in northwestern Western Australia has enabled M. robustus to invade previously unoccupied valley areas. In a reassessment of this, Newsome (1994) suggested that the increase in wallaroo numbers was, at least in part, due to dingo control (see below).
Macropus parryi (whiptail wallaby)
The whiptail wallaby has a north-eastern coastal distribution extending from Cooktown to northern New South Wales (Figure 7). M. parryi is common in undulating terrain with an open forest and grass understorey (Strahan 1995). Partial clearing of forest for cattle grazing appears to have benefited M. parryi populations by increasing the extent of the forest edge (Calaby 1966; McCann 1975; Southwell 1987; Strahan 1995). Total clearance of forest has been detrimental to the species in some areas (Johnson in Strahan 1995), particularly in the southern part of its range (Figure 7). but these represent only a small proportion of its range. Several authors have described qualitatively M. parryi having preferences for undulating or hilly country with open forest and a grass understorey (Calaby 1966; Bell 1973; McCann 1975; Strahan 1995). Recent detailed quantitative studies by Southwell (1987), and Southwell and Fletcher (1988) have confirmed the earlier qualitative descriptions.
A comprehensive review of what is known about the diets of the harvested species has been provided by Dawson (1989) while, in the same volume, Sanson (1989) reviewed dentition in relation to macropod diets. All are herbivores, whose role in the ecosystem is therefore that of a primary consumer. The large kangaroos are all specialist grass eaters, with red kangaroos, perhaps by necessity, being less specialised than the others.
Many detailed dietary studies have been undertaken on M. rufus (Griffiths and Barker 1966; Chippendale 1968; Storr 1968; Bailey et al. 1971; Griffiths et al. 1974; Ellis 1976; Barker 1987; Edwards et al. 1995), indicating a preference for green herbage including grasses and dicotyledonous plants.
Detailed studies of the diet of M. giganteus have been conducted in different areas and under different conditions by Kirkpatrick (1965, 1985), Griffith and Barker (1966), Griffith et al. (1974), Southwell (1981) and Taylor (1983b). The species may be said to be a selective grass eater, with the extent of its selectivity yet to be established (Dawson 1989). Nevertheless, specific food preferences appear to be retained even during drought when available plant species are fewer (Barker 1987; Strahan 1995).
Concerning M. fuliginosus, Coulson and Norbury (1988) found that, like M. giganteus, it feeds mainly on grasses. Norbury (1987), working in northwestern Victoria, found that they ate more than 75% grass in a mixed pasture but, as pasture biomass declined, shifted to forbs and shrubs. Barker (1987) described a similar shift from forbs and grasses to shrubs for western greys feeding on pastures in western New South Wales and southern Queensland. The diet of sheep showed a similar shift. This contrasted with red kangaroos and eastern grey kangaroos which continued to feed on grasses and forbs as pasture biomass declined.
Detailed dietary studies on M. robustus were undertaken by Ealey and Main (1967), Storr (1968), Ellis (1976), Squires (1982), and Taylor (1983b). Taylor (1983b) found that, in the tablelands of New South Wales, M. robustus had a broadly similar diet to M. giganteus, consisting primarily of grasses. In the arid Pilbara region of Western Australia, M. robustus was found to concentrate on spinifex (Ealey and Main 1967). The species is thus a grazer.
There have been few studies on the diet of M. parryi, however, Strahan (1995) and Bell (1973) described the species as a grazer.
Reproductive biology of the Macropodoidea has been reviewed in a book by Tyndale-Biscoe and Renfree (1987) and a series of papers in Grigg et al. (1989). Dawson (1995) has provided a semi-popular review.
Kangaroos are typical marsupials in that gestation is short, typically 30-35 days, the foetus does not implant until the last few days of gestation and most of the pre-natal development depends upon the consumption of yolk. The embryo is born at an early stage of development in comparison to placental mammals and, unless it finds its way to the pouch and one of the four teats, further development cannot occur. This means that kangaroos expend far less energy in reproduction than do placentals, because in placentals so much of the reproductive cost is pre-natal. Unfavourable environmental conditions may lead to the young being abandoned at this early stage, before the maternal energetic investment has become high.
Young are produced singly but, in most macropods, a second young, conceived following a post-partum oestrus, is held in embryonic diapause during lactation. This capacity means that, should environmental conditions precipitate a loss of the pouch young, it can be replaced quite rapidly. This is one of the features which enable such rapid recovery of kangaroo populations after drought conditions, and can be considered one of their many spectacular adoptations to life in harsh environments.
Another remarkable feature of macropods is their capacity to have three young simultaneously at different stages of development, with one in diapause, one pouch young and one young-at-foot. Also remarkably, the latter two will be suckling milk of different composition, from different teats, simultaneously.
There is a spectacular series of colour photographs showing reproductive events in the Tammar Wallaby in Renfree et al. (1989).
Comparative data which are relevant particularly to considerations of population dynamics of the harvested species of macropods are presented in Table 2.
There has been much research on the reproductive biology of M. rufus (Frith and Sharman 1964; Newsome 1964a, 1964b, 1965b; Sharman 1964; Sharman and Pilton 1964; Clark 1966; and see review by Tyndale-Biscoe and Renfree 1987). Breeding may occur year round, except in very poor seasons. Females are typically 26 kg when adult, live for >20 years (Bailey and Best 1992), and become sexually mature at about 28 months. They come into oestrus at approximately 35 day intervals and so are potentially fertile throughout the year. Periods of extreme drought however, may lead to suppression of the oestrous cycle. Females can come into breeding condition almost immediately after drought breaking rains. Pregnancy does not interrupt recurrence of oestrus. The female may give birth 33 days after mating and may mate again a day or two after this. The embryo resulting from this post- partum mating remains a quiescent blastocyst until the previous young is about to leave the pouch or is lost prematurely (embryonic diapause).
In M. giganteus, breeding occurs throughout the year but there is a peak of births in summer. Females are typically 32 kg when adult, live for >20 years, and become sexually mature at 14-28 months. The oestrous cycle is 46 days and the gestation period 36 days. Quiescent blastocysts are found occasionally in this species.
Reproductive biology of M. fuliginosus shows some minor differences from M. giganteus: the mean lengths of oestrous cycle (35 days) and gestation (30.5 days) are shorter, and M. fuliginosus does not exhibit embryonic diapause (Poole, in Strahan 1995). Breeding may occur year round, except in very poor seasons. Females are typically 27 kg when adult, live for >20 years, and become sexually mature at 18-24 months.
Reproductive biology of M. robustus has been studied by Ealey (1963), Sadlier (1965), Kirkpatrick (1968) and Poole and Merchant (1987). Under normal conditions, females breed continuously giving birth to a single young every 8 to 9 months. The gestation period is approximately 33 days. Breeding may occur year round, with a peak October-March. Females are typically 23 kg when adult, may live for >20 years, and become sexually mature at about 18-24 months.
Reproductive physiology of M. parryi has been studied by Merchant (in Calaby and Poole 1971) and Maynes (1973) and reproductive behaviour has been studied by Kaufmann (1974). The oestrous cycle is 41-44 days, and gestation takes 34-38 days. The young become detached from the teat when about 23 weeks old, and vacate the pouch at about 37 weeks. Breeding may occur year round, with a peak in summer, October-March. Females are typically 11 kg when adult, live for >12 years, and become sexually mature at about 18-36 months.
Behaviour and social organisation
Social organisation in the Macropodoidea, including the large kangaroos, has been reviewed by Croft (1989). He made the generalisation that 'the striking feature of the social organisations of many species of macropodoidea is their disorder.... age/sex classes intermingle haphazardly, feeding ranges are not defended, mating is promiscuous.' He referred to the large grazers as 'aggregated and gregarious'. Forage availability seems to play a bigger role than sociality in the distribution of females, while males have home ranges which are usually larger than and overlap those of females. Larger, higher ranked males have larger home ranges. Intersexual conflict is more common between males than females (Croft 1989). The extent of a continuing association between a large male and a number of females appears to have been exaggerated.
Among the harvested species there are far more similarities than differences. M. rufus is gregarious (Kirkpatrick 1967), but, although relatively large groups may sometimes form, these groups are unstable in their composition (Croft 1980). The only enduring social relationship is between the mother and her young. The mating system of M. rufus appears to be based on polygamy (Croft 1980).
Macropus giganteus, too, is gregarious (Southwell 1984a), forming groups which are unstable in their composition (Southwell 1984b). It has a polygamous mating system (Jarman and Southwell 1986).
Macropus robustus is less gregarious than other large kangaroo species (Kirkpatrick 1968; Croft 1981; Taylor 1982). Social behaviour has been studied by Croft (1981), and is nevertheless broadly similar to that for other large kangaroo species. Groups are highly labile in composition and the only enduring social relationship is between a female and its progeny. The mating system, too, is similar to the other large kangaroos.
The behaviour of M. parryi has been studied in detail by Kaufmann (1974). It is a social species (Kaufmann 1974 and Southwell unpublished data), living in groups of up to 50 individuals comprised of unstable sub-groups. M. parryi is polygamous (Kaufmann 1974), with only a few of the males mating with the many females in a group. The only long-term social bond is between the female and her young.
Movement in kangaroos has been studied by tagging individuals with ear tags or a collar around the neck, relying on shooters to return tags for a small reward, and by radiotelemetry. Movement of kangaroos in two study areas has been studied in close detail, at Fowlers Gap, 100 km north of Broken Hill (Dawson, Croft and colleagues) and Kinchega National Park, 200 km south east of Broken Hill (Priddel). Mature adults of all of the species are generally sedentary, with immatures more mobile. Tagging studies have shown that individuals sometimes very large distances, for reasons not yet known. Radio telemetry studies have shown detail of daily and short term movements. Daily movements involve, commonly, the movement from day-time shelter sites to foraging areas, and back. Dawson (1995) has provided an informative and very readable review of movement in kangaroos.
Red kangaroos have received the closest attention (Frith 1964; Bailey 1971; Denny 1980; Croft 1991a; Priddel 1987; Priddel et al. 1988a, 1988b; Norbury et al. 1994) and, although they are more mobile than the other species, the majority of the population is relatively sedentary, moving distances of no more than 10km. Individual home ranges are strongly overlapping, with those of males being larger than those of females. A small proportion of animals may move distances of tens or hundreds of kilometres. Some of these movements are permanent and are presumably related to dispersal. Younger individuals, particularly are mobile and are less likely to remain in an area where they are radio- collared for very long. Some movements, however, are related to movement away from a drought-stricken area in search of better feed and, in such a case, individuals apparently may show strong site fidelity and return to the same home range in due course.
Both eastern and western greys are less mobile than reds. Studies of eastern greys by Jarman and Taylor (1983) and Jarman and Southwell (1986) indicate that the species occupies well-defined, highly overlapping home ranges. Few individuals have been shown to disperse, those which do being young males.
Western greys were studied by Priddel (1987) and Priddel et al. (1988a, 1988b) and show the same general patterns, with individuals occupying relatively small home ranges that overlap extensively.
Studies of movement in wallaroos, by Ealey (1967), Croft (1991b), Jarman and Taylor (1983) and Clancy and Croft (1989) have shown that they, too, are relatively sedentary, occupying small home-ranges that overlap broadly with those of other individuals. Clancy and Croft (1989) found that males of M. r. erubescens in the Fowlers gap area progressively shifted their centres of activity within their home ranges on a short term basis, a trait shown by some of the females as well. Movements were, however, quite small- scale, within a couple of kilometres and home ranges remained stable from year to year.
Movements of M. parryi are restricted to small home ranges which overlap extensively with the ranges of other individuals in a group.
(A review of the effects of harvesting on genetic diversity is given in Chapter 6)
An assessment of the genetic structure of red kangaroo (Clegg et al. submitted; Hale et al. in prep.) and common wallaroo (Hale et al. in prep.) populations throughout Australia has enabled potential management units to be identified. At a continental scale, red kangaroos show little structuring. However, at a regional level, distinct, populations could be identified and these were of smaller geographic range in areas with greater habitat complexity (Figure10). Genetic structuring of common wallaroo populations generally conforms to earlier divisions into sub-species based largely on morphology (Figure 6). However, clear population structuring occurs within these divisions. In order to offset any effects of harvesting on genetic diversity , these genetically distinct populations should be managed as separate units (Moritz 1994).
Figure 10. Geographic range of genetic populations of the red kangaroo. Light shading indicates the species' range. Cross-hatching identifies sampling regions. The dotted lines join regions that could not be distinguished on the basis of allozyme or microsatellite genotype frequency (after Clegg et al. submitted).
In a review of diseases of wild macropods, Speare et al. (1989) pointed out that, while dynamics of kangaroo populations are driven by environmental fluctuations, most individuals die of specific diseases. However, a distinction needs to be made between the impact of disease at the level of the individual, and the local and regional population. A number of diseases which are significant agents of mortality in captive populations, either have not been reported in wild populations or are of lesser importance. Some diseases have zoonotic significance, including salmonellosis, leptospirosis and Q fever, and can potentially affect products of the kangaroo industry.
However, Andrews (1988), reported in Dawson (1995), found that, of over 200,000 kangaroo carcasses inspected for export as game meat, <0.7% were considered to have some form of pathological condition. Rates for domestic animals are considerably (see Chapter 7). The nematode parasite causing the bulk of the rejections of kangaroo meat, Pelecitus roemeri, is harmless to humans. Lesions on hides of red kangaroos in the northern half of Queensland may cause substantial losses to shooters and dealers. Gross and histological examination of a large number of carcasses by Kelly (1997), identified ticks, Amblyomma trigutttatum, as the likely cause. Less vigorous preparation of skins, which converts tick damage into actual perforations, may reduce the problem.
A number of epizootics have been reported in wild kangaroos. Apparent epidemics of lumpy jaw have occurred in the Murchison area of Western Australia several times this century (Tomlinson and Gooding 1954). Localised epidemics of coccidiosis resulted in the deaths of many juvenile eastern grey kangaroos trapped by rising flood waters (Barker et al. 1972). Malnutrition and high densities were thought to make younger animals particularly susceptible when exposed to large numbers of oocysts. In 1990, following heavy rain and flooding in the Thompson-Barcoo-Cooper river system in western Queensland, Clancy et al. (1990) reported significant mortalities of red kangaroos, eastern grey kangaroos and wallaroos, with mortality rates declining away from the river. Aerial surveys suggested a reduction in the red kangaroo population of >60% in an area of 10,000 km2. The mortalities coincided with outbreaks of sandflies, Austrosimulium pestilens, and necropsies on carcasses suggested arbovirus infection (Speare et al. 1990). At the same time, similar surveys by Choquenot (1991) in northwestern New South Wales found no reductions in populations of either red or grey kangaroos. Flood-related epizootics do not appear to be an inevitable consequence of flooding. While flood-related mortalities may be significant for local populations, regional populations may actually show a net increase in size resulting from increased food supply.
Outbreaks of choroid blindness have also been reported in all three kangaroo species and common wallaroos. Western grey kangaroos appear to be the worst affected. An outbreak occurred in western New South Wales in 1994-5, spreading rapidly into northern Victoria and southern South Australia, but abating in the cooler months. It was detected in Western Australia in late 1995. The disease was attributed to a Wallal virus, an orbivirus, with biting insects Culicicoides austropalpalus and C. dycei as the main vectors. Aerial surveys suggested that the impact of the disease at a population level was minor. Anecdotal reports of blindness in kangaroos date back to the 1930s, suggesting the disease is not new.
Before European settlement, large fluctuations in kangaroo numbers may not have been as common a feature of semi-arid Australia. Outside the sheep rangelands where dingoes are common, kangaroo densities are generally substantially lower (Caughley et al. 1980), suggesting that dingoes are an important limiting factor. Trends in red kangaroo (and emu) numbers on either side of the dingo fence in South Australia further suggest that dingoes are capable of regulating kangaroo populations (Pople et al. in prep.). Following drought, red kangaroos remained at low densities outside the fence where dingo numbers appeared to increase. Inside the fence, where dingoes are rare, red kangaroo populations showed a typical post-drought recovery. Prey switching appears to be an important mechanism for these dynamics. Throughout much of southern and central Australia, dingoes prey primarily on rabbits, but, as rabbit numbers decline, dingoes increasingly take larger prey, including red kangaroos. They also feed on carcasses of cattle that have died during the drought (Corbett and Newsome 1987; Newsome 1990; Corbett 1995). Even in the absence of rabbits, dingoes appear capable of strongly limiting red kangaroo populations and possibly regulating them (Thomson 1992). Other factors will also be important, including the increased vulnerability of kangaroo populations during drought when they become concentrated around areas of persistent food and water (Newsome 1965a).
How the composition of red kangaroo populations is altered by dingo predation is yet to be quantified. In northern New South Wales, unstable age distributions of swamp wallabies (Wallabia bicolor), red-necked wallabies and eastern grey kangaroos were all considered by Robertshaw and Harden (1989) to be the result of dingo predation that targeted juveniles. Selective predation by dingoes has been reported also on old male and juvenile wallaroos in northern Western Australia (Oliver 1986) and on juvenile red kangaroos in northwestern New South Wales (Shepherd 1981). Although these data are sketchy, there appear to be marked differences between the composition of macropod populations preyed upon by dingoes, commercially harvested populations and populations where predation is insignificant.
While dingoes are the only significant non-human predator of older animals, foxes and perhaps wedge-tailed eagles may prey upon juvenile kangaroos and wallaroos (Robertshaw and Harden 1989). Banks (unpublished data) reported an increase in eastern grey kangaroo numbers following fox control compared with a decline in kangaroo numbers in non-removal sites. Banks' results suggested that the kangaroo population was limited by fox predation on juveniles, with a higher proportion of females with young-at foot in sites where foxes were removed. Banks also reported a change in kangaroo foraging behaviour, where predation risk was reduced, with animals feeding in smaller groups and in more exposed habitats.