|dc.description.abstract||The identification of factors limiting the recovery of threatened bird species is an area of significant research in New Zealand, where high levels of endemism make protection of threatened species extremely important. Predation by introduced mammals is often assumed to be the most important limiting factor for populations of threatened bird species, and a number of methods have been developed and implemented to deal with predators. Pest-management operations have a long history of success in NZ, but can also have unexpected consequences for non-target species. The three most common mainland pest-management measures are trapping, poisoning, and predator-exclusion fencing. My study used the South Island robin (Petroica australis) as a model to investigate the costs and benefits of three predator control operations over a period of six years at three independent sites in Dunedin, NZ: Silverstream, where rodent trapping occurs; Silver Peaks, the site of an aerially dispersed cereal-bait 1080 operation with pre-feed; and Orokonui, a predator-free sanctuary. Chew track cards were used to track changes in relative abundances of ship rats (Rattus rattus), mice (Mus musculus), and brushtail possums (Trichosurus vulpecula) at Silverstream and Silver Peaks over the period of 2011 to 2014. I monitored known robin pairs and single birds at all sites over the 2013/14 and 2014/15 summer breeding periods and combined this with previous monitoring data to track changes in annual adult robin survival and juvenile and adult robin recruitment rates, as well as robin nesting success. I also filmed nest sites at Silverstream to determine nest predators (nests were not filmed at Silver Peaks due to nests being too high to access). Trapping was effective in reducing ship rat relative abundance, although possum relative abundance increased in parallel. Poisoning resulted in significant initial decreases of all monitored species. However, this was short-lived, with abundances of all monitored robin predators exceeding pre-operational numbers within a year, this being an outcome that has been observed in previous studies. Orokonui displayed high values for all robin metrics except adult recruitment. Silver Peaks displayed comparatively low rates of adult survival, high rates of adult recruitment, moderate rates of juvenile recruitment, and low rates of nesting success. Silverstream displayed high rates of adult survival and recruitment, but low rates of juvenile recruitment and nesting success. No significant differences in adult survival were detected between sexes, and no significant differences in nesting success were detected between incubation and nestling stages. Predation, especially by stoats, was found to be the primary limiting factor affecting nesting success, although no other metrics were thought to be significantly affected by predation. Some evidence for masking of competition effects by predation is presented.
The results of this thesis provide key insights into the efficacy of management for South Island robins as well as knowledge of the effects and interactions of predation and competition on a native bird species. This will be useful in future research and management strategies, helping better tailor predator-control regimes to target problem species, enabling rapid recovery of valuable species and preservation of New Zealand’s unique fauna and flora.||