This paper challenges conclusions of Caughley et al. (1980) that the abundance of red kangaroos (Macropus rufus) in western New South Wales is solely due to lack of dingoes (Canis lupus dingo), and vice versa for neighbouring South Australia. A Dingo Barrier Fence divides the two different ecological systems, which have sheep in New South Wales and cattle in South Australia. This paper re-examines in particular whether there is an environmental gradient across the Fence that was dismissed by Caughley et al. This paper concludes to the contrary, that there is a strong environmental gradient. Our aerial surveys demonstrate significantly that habitats favouring red kangaroos are prevalent in New South Wales today, but are very scarce or absent in South Australian landscapes. Aerial surveys were used in both studies, but designs differed. Caughley et al. flew at right angles across the Fence on paths 28 km apart. Flights would have crossed the south-westerly streamlines rarely. Our flight lanes followed streamlines looking for floodouts, the favourite habitat of red kangaroos. Return lanes went between streamlines sampling other habitats. Counts of red kangaroos seen were made every 1.75 km, with the specific habitat also identified. Three extra factors are invoked in our study. One is that the low annual rainfalls translate into intrinsically low survival rates of pouch-young of red kangaroos, contrary to their abundance in New South Wales today. The other two are related to that current abundance also. There is now evidence for greatly increased run-off of rainfall from catchments onto the open plains in New South Wales. Also present is a very large shallow basin lying between catchments and the Dingo Barrier Fence. Streamlines enter it but none flow past its western rim. The above conclusions were confirmed during subsequent ground surveys over three years. Of eleven species of medium and large vertebrates seen in New South Wales, five were absent in South Australia. Three were kangaroos, and the others were feral pigs and goats. Emus are more abundant in New South Wales also. All of those species would be targets for dingoes, especially as alternate prey to rabbits that generate huge eruptions every decade or so. Red foxes (Vulpes vulpes) were in lower abundance in South Australia with dingoes present, as expected with meso-predator interactions. Feral cats (Felis catus) were in similar numbers on both sides of the Fence for unknown reasons. Competition between rabbits (Oryctolagus cuniculus) and sheep for food in New South Wales was shown to significantly reduce rabbit numbers in drought. That rabbits are perennially in lower densities there than in South Australia may be due to the higher densities of foxes than in South Australia. Historically, red kangaroos were rare in the region in the mid-1800s. Their abundance has arisen since European occupation. Thc species was rare on those open plains, and permanent water was scarce. Rabbit Haemorrhagic Disease reached the study-area in 1995. Its impact reduced rabbit populations to a rarity that prevails today on both sides of the Dingo Fence. Predation from dingoes, foxes and feral cats may assist continuance of low numbers of rabbits. Pastures, seedling trees and livestock will benefit, as will the kangaroos.

Shot samples of rabbits were collected over several years at 11 sites in Australia and one in New Zealand. At any one site, pregnancy rates and litter sizes vary both with age of mother and with time of year. Few rabbits become pregnant before the age of 19 weeks; pregnancy rate increases until the full adult rate is achieved about 27 weeks old. Litter size is also affected by age, adult litter size being reached by females at about 43 weeks old. Rates and timing of reproduction vary greatly from site to site in response to local conditions. The annual production of young per fully adult female is highest at Wairarapa, N.Z. (53 young) and lowest in the semiarid (17) and subalpine (15) environments. At six of the sites cohorts of marked rabbits provided estimates of seasonal survival rates; survival improves with age but becomes relatively constant in rabbits aged more than 24 weeks; survival patterns differ between sites. Life tables were constructed combining the information on reproduction and survival. Their possible usefulness is discussed.

Shortage of water in natural pastures led to a sharp decline in a large rabbit population in arid, northeastern South Australia. The pastures were dry and some rabbits drank at springs and water troughs. Further from water, rabbits climbed trees and shrubs to obtain succulent leaves and twigs. Rabbits provided with water maintained their weight and apparently survived better than those which did not drink. It seems unlikely that the rabbits lost weight because the water shortage reduced the amount of dry food they could eat. In caged rabbits, water shortage limits food intake but also results in low gut fill; whereas the wild rabbits had the normal amount of digesta in their guts. It is more likely that, as the pastures became dry, rabbits ate woody twigs and bark which were moist enough to meet their water requirements but contained too little digestible energy for maintenance. The water shortage apparently arose because rabbits were numerous and had eaten out the succulent pasture plants. Normally, it takes a long drought to reduce arid-zone plants to dry straw, and overgrazing is probably the usual cause of a lack of adequate water for rabbits.

The basic breeding biology of the rabbit flea Xenopsylla cunicularis Smit, 1957 is described, and comparisons are made with other closely related species of Xenopsylla and with Spilopsyllus cuniculi (Dale, 1878), another parasite of the wild rabbit. At 22�C and 80% relative humidity the eggs of X. cunicularis hatch 7-8 days after laying, and the larval stages last 8-9 days. Pupation lasts for a further 20 days for females and 27 days for males. However, at lower relative humidities (22�C, 50% RH) rates of development are slowed considerably. A method for rearing large numbers of X. cunicularis in the laboratory is described.

A three-year programme to eradicate Feral Cats Felis catus from the island of Baltra in the Galapagos archipelago achieved good results by initially poisoning with sodium monofluoroacetate (compound 1080) then trapping or shooting the remaining cats. The poisoning campaign removed 90% of the cats, its success being attributable to pre-baiting with unpolsoned baits to accustom cats to eating baits and placing enough baits to ensure that all cats encountered several baits within their home range. This, together with the use of metaclopromide (Pileran) as an anti-emetic, overcame a problem associated with poor retention of 1080 in thawed fish baits that limited the dose available to 1 mg 1080/bait, a quality insufficient to kill large cats. Removal of the remaining cats was delayed by a weather-induced irruption of Black Rats Rattus rattus and House Mice Mus musculus that enabled recruitment of kittens in 2002, but made cats more susceptible to trapping and shooting in 2003 when rodent populations collapsed. Since July 2003 no sign of a cat has been detected on Baltra despite extensive searching and monitoring throughout 2004. As cat abundance has decreased there have been more locally-bred Juvenile iguanas (Conolophus subcristatus) seen during annual censuses. However, such recruitment may reflect the increasing maturity and higher fecundity of iguanas repatriated from 1991 onwards rather than being a direct result of reduced cat predation alone. More time is necessary to determine the benefits of reduced cat predation on the Iguana population.