One reason for the legislative restrictions on feeding dolphins in many parts of the world is the putative increased risk of injury to dolphins conditioned to human interaction through food reinforcement. However, there are few empirical data to support this. Here, we present data for a population of Bottlenose Dolphins Tursiops sp. in Cockburn Sound, in the city of Perth, Western Australia, indicating higher incidence of boat strike injury and fishing line entanglement for dolphins conditioned to taking food from humans, compared to others in the population that were not conditioned. The data support prohibitions on feeding dolphins and rigorous enforcement of existing regulations.

A technique is described whereby the faeces of birds are analysed to determine the order (often the family) and the length of their anthropod prey. Equations are given enabling the dry weight and energy content of prey to be estimated from length data.

No abstract

The potential hazard of 1080 baiting for predators to 14 species of non-target mammals in the pastoral areas of Western Australia and a further six from Western Australia's Fitzgerald River National Park, was assessed by comparing projected doses of 1080 (based on consumption of non-toxic bait by captive animals in the absence of alternative food) with the approximate lethal dose of 1080 for each species. These figures suggested that individuals from 12 species were potentially at risk from crackle baits, while only individuals from Dasyurus hallucatus, Ningaui spp., Sminthopsis crassicaudata, Planigale maculata, one population of Leggadina forresti and one population of Sminthopsis ooldea were potentially endangered by meat baits. Tests using the native mammals Zyzomys argurus and Pseudomys hermannsbergensis and laboratory mice (Mus musculus) and laboratory rats (Rattus norvegicus) showed that individuals of all species reduced their consumption of toxic bait relative to non-toxic bait, although this did not prevent three of five rats and one of three P. hermannsbergensis from being killed.

The feeding habits of small birds in the understorey of a wet sclerophyll forest near Pemberton, W.A., were studied in May of 1979 and 1980. Observations were made of the height and substrate at which birds fed, and the sizes and types of invertebrates eaten were determined by faecal analysis. None of the seven most common species showed more than 55% overlap with each other in all four of these feeding dimensions. It is suggested that the inference of diets from purely observational data may give misleading results. An attempt to estimate foraging overlap between birds from their morphological similarity was unsuccessful.