The effects of season, population density and habitat on group density and group size in the eastern grey kangaroo were examined. In an open forest habitat, both group density and group size increased significantly with population density, but no seasonal variation was detected for either parameter. The relationship with population density was logarithmic for group density and exponential for group size. In a tall shrubland habitat both group density and group size increased with population density, but the range of population densities sampled was too narrow for the nature of the relationships to be determined. Group density was significantly higher, and group size significantly lower, in the tall shrubland than in the open forest. Possible reasons for this difference in spatial dispersion between habitats are discussed.

Recognition of individual large mammals by natural features is an ancient practice, the usefulness of which has recently become appreciated in field studies. We show that such recognition is possible and repeatable, under particular circumstances, with macropodid marsupials in the field in Australia. In trials, there was a 98% agreement between observers in identification of individuals of both eastern grey kangaroos Macropus giganteus and red-necked wallabies M. rufogriseus. Characters used to identify individuals were first definitive ('oddity') characters, and then specific sets of character-states. However, in time observers came to recognise individuals 'on sight', by unconscious summation of character-states. Recognition skills could be readily learned by observers, allowing almost complete populations of these species to be known and the component members studied individually. The technique should be applicable in studies of other macropodid populations.

A broad-scale ground survey of macropod abundance in north-eastern New South Wales was undertaken in 1989-90. The survey area was stratified into an eastern and western region, and within regions by habitat on the basis of vegetation and topographic criteria. Macropod density in each stratum was estimated from walked line transect counts for five species: Macropus giganteus, Macropus robustus, Macropus parryi, Macropus rufogriseus and Wallabia bicolor. Within sampled strata, macropod density was substantially higher in the western region than in the eastern region. Macropus parryi occurred in only trace numbers in the eastern region, with the number of sightings too small for reliable density estimation. No M. parryi were sighted in the western region. Macropus giganteus and M. rufogriseus were the dominant species in sampled strata in the eastern region, while M. giganteus and M. robustus were dominant in the west. Overall, pre-survey activities of habitat digitising and liaison with landholders required approximately the same effort as actual survey work. Liaison with landholders required relatively more time in the eastern region, where human population was higher, than in the western region. Incomplete coverage of planned transect routes, due mainly to impenetrable vegetation, lead to a positive bias of no more than 10-20% in population estimation for M. giganteus and M. robustus, and an unknown bias for the other species. The use of broad vegetation information on 1 : 100 000 map sheets for habitat stratification limited the scale at which analysis of sightability could be undertaken. Macropus giganteus and M. robustus are commercially harvested in the western region. Annual commercial harvest rates in the three years prior to the survey were conservatively estimated to be less than 0.9% for M. robustus and less than 5.4% for M. giganteus.

The effects of season, population density and habitat on aspects of group formation in the eastern grey kangaroo were examined. The rate at which groups join and split increased exponentially with population density, but did not vary with season or habitat. In general, large males and females with young-at-foot were seen alone more frequently than small males, females with pouch young or females with no young. The high frequency oflone large males may be due to a tendency to move alone from one group to another to check the oestrous condition of females, but females with young-at-foot appear to be relatively solitary. Association among population classes in groups of two or more was highly variable.

This paper introduces a series of papers on the ecology, social organisation and behaviour of populations of sympatric macropods (Macropodoidea : Marsupialia) in north-eastern New South Wales. The study site, in the valley of Wallaby Creek, covers partly tree-cleared cattle-grazed pastures and also wet and dry forest communities; 10 species of macropods live there. The valley has a moderately high rainfall (1023 mm per annum), falling predominately in summer, and an equable climate of cool winters and warm summers. Soils derived from sedimentary and basaltic rocks and alluvium support naturally diverse plant associations further diversified by clearing and establishment of pasture and weed species. Macropods favouring open country occupy the pastures, which can also be used by cover-dependent species where pasture abuts forest or remnant patches of cover. Composition of the macropod community has changed since development of the pasture zone. Dingoes, major predators of some of the macropods, are abundant, and all exotic mammals other than cattle are rare. Populations of two of the macropod species are habituated to approach by observers, and close observation, on foot, of undisturbed animals has become our common study technique. A 1-ha grid has been established over 3.7 km2 of the study site to facilitate exact location of animals and observations. The suitability of the macropod populations for this kind of study results from the attitudes of the landholders.