The choice of a place to live and reproduce is crucial for species� survival in providing them with adequate resources and shelter from predators or climatic conditions. Determining habitat suitability in endangered species is important for the success of translocation as a conservation tool. In addition, understanding mechanisms (source/sink system versus ecological traps) that drive habitat selection in translocated animals may be critical to population viability. In New Zealand, where ecosystems are highly vulnerable to extinction, habitat restoration on predator-free off-shore islands is an important recovery tool. Therefore, there is a need to understand the relationship between the establishment of the translocated populations and the characteristics of their new environment. Previous research indicated that re-introduced populations of Stewart Island robin (Petroica australis rakiura - Toutouwai) and South Island saddleback (Philesternus carunculatus carunculatus - Tieke) on Ulva Island (Stewart Island), New Zealand, showed preferences for coastal habitats that were characterized by low-lying dense vegetation and open ground cover. In this study, we further investigated territorial establishment in these two populations since re-introduction and looked at how birds utilised the landscape. I hypothesised that sites colonised soon after re-introduction were of high quality and later on, birds moved into unsuitable habitats. I defined habitat quality at a micro-scale in terms of vegetation structure, nest characteristics and food availability. I modeled bird presence and nesting success in relation to habitat components to determine factors in the environment that influenced breeding site selection and contributed to successful nesting in these two species. I discussed results in comparison to similar bird-habitat models developed for the South Island saddleback population on Motuara Island (Marlborough Sounds) and examined explanatory variables in each model. Translocated birds in the three studied populations first established territories in coastal scrub, and in the following years moved into larger coastal forest stands. Although vegetation structure was the primary variable explaining site selection in these populations, vegetation composition should still be considered important as it dictated the suitability of nesting substrate and the availability of food items. There was no evidence that first-colonised areas were more suitable habitats, and I concluded that these cases could not be used as examples of ecological traps. Instead, results suggested that with increased density robins and saddlebacks on Ulva have more recently settled in sites less suitable to nesting and foraging, thus underlying a source/sink structure. However, the sparse distribution of food items on Motuara contributed to a lack of territorial behavior and environmental effect on breeding success; therefore a source/sink system could not be confirmed in this population. I recommended that future translocation sites give preference to mixed-size stands with broadleaved species that are characterised by dense canopy below 4 m height and with suitable cavities in live trees. Lastly, due to robins� and saddlebacks� attraction to conspecifics and their territorial behavior, resources evenly distributed across the landscape could also increase their survival and reproductive success.

The choice of a place to live and reproduce is crucial for species� survival in providing them with adequate resources and shelter from predators or climatic conditions. Determining habitat suitability in endangered species is important for the success of translocation as a conservation tool. In addition, understanding mechanisms (source/sink system versus ecological traps) that drive habitat selection in translocated animals may be critical to population viability. In New Zealand, where ecosystems are highly vulnerable to extinction, habitat restoration on predator-free off-shore islands is an important recovery tool. Therefore, there is a need to understand the relationship between the establishment of the translocated populations and the characteristics of their new environment. Previous research indicated that re-introduced populations of Stewart Island robin (Petroica australis rakiura - Toutouwai) and South Island saddleback (Philesternus carunculatus carunculatus - Tieke) on Ulva Island (Stewart Island), New Zealand, showed preferences for coastal habitats that were characterized by low-lying dense vegetation and open ground cover. In this study, we further investigated territorial establishment in these two populations since re-introduction and looked at how birds utilised the landscape. I hypothesised that sites colonised soon after re-introduction were of high quality and later on, birds moved into unsuitable habitats. I defined habitat quality at a micro-scale in terms of vegetation structure, nest characteristics and food availability. I modeled bird presence and nesting success in relation to habitat components to determine factors in the environment that influenced breeding site selection and contributed to successful nesting in these two species. I discussed results in comparison to similar bird-habitat models developed for the South Island saddleback population on Motuara Island (Marlborough Sounds) and examined explanatory variables in each model. Translocated birds in the three studied populations first established territories in coastal scrub, and in the following years moved into larger coastal forest stands. Although vegetation structure was the primary variable explaining site selection in these populations, vegetation composition should still be considered important as it dictated the suitability of nesting substrate and the availability of food items. There was no evidence that first-colonised areas were more suitable habitats, and I concluded that these cases could not be used as examples of ecological traps. Instead, results suggested that with increased density robins and saddlebacks on Ulva have more recently settled in sites less suitable to nesting and foraging, thus underlying a source/sink structure. However, the sparse distribution of food items on Motuara contributed to a lack of territorial behavior and environmental effect on breeding success; therefore a source/sink system could not be confirmed in this population. I recommended that future translocation sites give preference to mixed-size stands with broadleaved species that are characterised by dense canopy below 4 m height and with suitable cavities in live trees. Lastly, due to robins� and saddlebacks� attraction to conspecifics and their territorial behavior, resources evenly distributed across the landscape could also increase their survival and reproductive success.