Abstract
Conservation translocations are increasingly used to mitigate biodiversity loss, yet their outcomes remain variable and difficult to predict. This uncertainty is compounded by limited post-release monitoring and a lack of understanding of the ecological processes that influence outcomes during and after translocation. Predation is widely recognised as a key driver of translocation failure, particularly during the establishment phase, but risk is often assessed using simplified presence-absence frameworks that overlook the spatial, temporal, and behavioural processes that govern predator-prey interactions. In Aotearoa New Zealand, where translocations are a central conservation tool and native predators are frequently excluded from release sites as a precaution, understanding how these processes influence outcomes is critical. This is particularly important for endemic lizards, which remain underrepresented in translocation research. Understanding the threat posed by one of their native predators, the weka (Gallirallus australis), which is itself of conservation concern, is a high research priority for informing conservation planning in Aotearoa.
In this thesis, I investigated how the spatial and behavioural ecology of weka determines patterns of overlap and encounter rates with translocated skinks. I also examined how these interactions relate to translocation outcomes. Using a multi-scale approach, I first synthesised weka home range sizes across Aotearoa from a range of studies spanning 14 sites. I then collected and analysed data on weka home ranges, habitat selection and activity patterns on Secretary Island, Fiordland. Secretary Island represents a relatively intact ecological context with few invasive mammals, a high-density weka population, and an elevational gradient from forest to alpine habitats. In addition, I examined two lizard translocation case studies in Fiordland that addressed different sources of uncertainty in translocation outcomes. The first case study focused on short-term translocation outcomes of mahogany skinks (Oligosoma aff. inconspicuum "mahogany") in the alpine zone of Secretary Island, assessing whether resident weka preyed on translocated skinks. The second case study examined longer-term translocation outcomes of Te Kākahu skinks (O. tekakahu) in a subalpine environment on Anchor Island, which lacks both introduced and native predators. This case study provides a contrasting perspective on translocation outcomes in the absence of predation, including other risks such as habitat suitability.
Across Aotearoa, weka home range size varied by more than two orders of magnitude (<1 ha to >100 ha), including substantial variation within sites. Ecological predictors such as island versus mainland site, invasive mammal presence, and habitat type did not show consistent relationships with home range size, reflecting strong context dependence and limiting their utility for predicting encounter rates with prey. On Secretary Island, habitat selection and activity patterns were non-random, with weka showing strong selection for tussock habitats and significantly higher movement rates during the day. Habitat selection varied across the diel cycle, with the strongest differences occurring at dusk and night, demonstrating that weka space use is dynamic across both space and time.
Within the same system on Secretary Island, the mahogany skink population was monitored at their release site over two summers, and weka were not observed preying on skinks. Skinks persisted at the release site for a minimum of one year, and some dispersed at least 25 metres away. Peak activity levels of skinks and weka did not overlap in relation to time of day, daylight type, and temperature. These results indicate that weka presence did not translate into realised predation risk, and that interactions were constrained by encounter processes, providing insight into early post-translocation dynamics under predation risk.
In a longer-term context, a five-year post-translocation study of the Te Kākahu skink on predator-free Anchor Island showed strong evidence of population establishment. A high proportion of founders were recaptured over five years, body condition improved following release, and island-born juveniles were detected, with multiple generations observed. Using a stage-based framework adapted to the species’ specific life history traits, the translocation was successful to at least the point of reproduction and showed strong evidence of progressing towards population growth, emphasising the importance of longer-term monitoring for evaluating translocation success.
Together, my findings demonstrate that predation risk to translocated lizards is not determined solely by weka presence, but is mediated by spatial, temporal, and behavioural processes that shape encounter probability. Translocation outcomes are therefore highly context-dependent, reflecting interactions among both predator and prey ecology, habitat structure, environmental conditions, and demographic processes. My thesis highlights the importance of evaluating translocation success across multiple stages and timeframes aligned with species’ life histories, and demonstrates that native predators such as weka do not necessarily pose uniform risk to translocated populations. It also provides the first assessment of lizard translocations in alpine and subalpine habitats, represents one of the first studies focused on lizards in Fiordland, and offers a detailed analysis of weka ecology across an elevational gradient that includes the alpine zone. By moving beyond simplified approaches towards process-based understanding, this research supports more flexible and adaptive approaches to translocation planning, where risks are assessed in ecological context and management decisions are informed by both empirical evidence and ongoing monitoring.