Abstract
Habitat loss, disease, and climate change are catalysts causing mass amphibian declines globally. A recent assessment has suggested climate change is emerging as the new primary driver of these declines. Climate change is expected to exacerbate the threats already faced by amphibians. Island ecosystems are predicted to be the most vulnerable to the direct effects of climate change. How threatened amphibian species that live in these island ecosystems will respond to climate change is of concern. As an island ecosystem and home to vulnerable amphibians, it is essential to formulate a better understanding of how Aotearoa, New Zealand’s endemic Leiopelma frog genus, may respond. In this thesis, I aim to understand Leiopelma distributions at a broad spatial scale using species distribution modelling (SDM) and Hochstetter’s frog (L. hochstetteri) movements at a finer spatial scale through radio telemetry to inform conservation management.
Six species of Leiopelma were present in Aotearoa during the Holocene epoch (~12-0 Kya). The loss of the now-extinct Aurora, Markham’s, and Waitomo frogs occurred following significant human-induced habitat loss and the introduction of mammalian predators. In Chapter Two of this thesis, I explored the potential past and present distributions of Markham’s and Waitomo frogs using SDM. Moisture related predictor variables were informative of species presence, which aligned with the current life history that Markham’s frog was terrestrial and Waitomo frog was semi-aquatic, inferred from subfossil morphology and deposit type. Present-day models of these extinct species suggested large areas of potentially suitable habitat remain. In Chapter Three, I also utilised SDM and explored the potential past, present, and future distributions of the three living Leiopelma: Archey’s, Hamilton’s, and Hochstetter’s frogs. The models in this chapter built and improved upon previous Archey’s and Hochstetter’s frogs SDM efforts by integrating the subfossil record and including a detailed vegetation layer. Including the subfossil record increased the amount of occurrence data available for Hamilton’s frog, allowing the species to be modelled for the first time.
Model outputs across both chapters suggested forest refugia and land bridges may have supported and connected populations of all five Leiopelma modelled during the Last Glacial Maximum (LGM)(~32-19 Kya). Potentially suitable habitat reduced from the LGM to theii mid-Holocene (~6 Kya) for all species. Present-day outputs and the subfossil record suggest it is plausible that additional populations of Leiopelma could persist at low levels.
For the three living Leiopelma, integrating the subfossil record into models identified additional areas of potentially suitable habitat at all time periods. Future distributions (2050 and 2070) suggested potential range shifts for all three living species. A range expansion was suggested for the terrestrial Archey’s and Hamilton’s frogs. The semi-aquatic Hochstetter’s frog models showed a range shift southwards. Due to the fragmentation of populations and dispersal limitations of Leiopelma, conservation translocations will likely be required in the future for the survival of Leiopelma.
While understanding Leiopelma distributions at a broad scale is important, a pivotal step of SDM is interpreting model outputs in relation to known species ecology. Therefore, understanding the movement ecology of Leiopelma at a finer spatial scale is beneficial. In Chapter Four, I assessed the suitability of radio telemetry as a method to explore the movement ecology of Hochstetter’s frog. A simple harness design requires minor length adjustments to reduce abrasion before it can be considered suitable for frogs to carry transmitters. The field study also highlighted the need for consistent abrasion recording, and an abrasion scale was developed. Preliminary telemetry data revealed variation in individual Hochstetter’s frog activity areas (1.1 to 319.7 m2), and one seepage to seepage movement was recorded (16.9 m within a 12-hour period). Further research is required before home range can be confirmed, but this is the first study to demonstrate that radio telemetry is a promising tool to understand the movement and microhabitat preferences of Hochstetter’s frog in the future.
Altogether, this thesis exemplifies how understanding broad distributions of Leiopelma using SDM and fine scale movements through radio telemetry complement each other to improve future conservation suggestions. With a better understanding of how a species uses its habitat, potentially suitable locations identified by SDM can be surveyed more thoroughly and assessments of potential translocation sites can ensure the presence of suitable microhabitat features.