Abundance and microhabitat use of Leiopelma archeyi in relation to land-use : a thesis presented in partial fulfilment of the requirements for the degree of Masters of Science at Massey University, Palmerston North, New Zealand
Habitat disturbance has been recognised as a significant factor contributing to biodiversity decline worldwide. In New Zealand, land-use activities on conservation land and privately-owned property have been disputed due to their potential effects on vulnerable species. Currently, there is limited information regarding the impact of land-use activities on populations of Archey’s frogs (Leiopelma archeyi), a small ground-dwelling amphibian. Historical exploration mining in Wharekirauponga (WKP), and roading and housing-related activities in Mahakirau Forest Estate (MFE) in the Coromandel Peninsula, were the land-use activities focused on in this thesis. Disturbance from these activities included site clearance, which removed >50% of vegetation in ca.100m² areas, during the 1980s, 1990s and 2010-16. My primary aims were to investigate whether these disturbances have affected the current abundance and microhabitat use of L. archeyi. I surveyed 16 pairs of disturbed and undisturbed sites for emerged frogs over three consecutive nights. These sites were 10 m x 10 m. Captured frogs were photographed, measured and weighed for individual identification and recaptures noted. Vegetation of sites was characterised using reconnaissance (RECCE) plot vegetation descriptions. To predict frog presence within the 100 m² areas, microhabitat features were assessed in 1.5 m³ plots where frogs were found and randomly-selected plots where frogs were not found. A purpose-built closed-mark recapture model was developed to calculate detection probabilities that were used to estimate abundance within disturbed and undisturbed sites. This modelling incorporated a disturbance effect and habitat characteristics assumed to be important to L. archeyi. Vegetation composition from the RECCE data and finer habitat features were characterised using ordination techniques and coefficients of determination, both to assess how vegetation was affected by disturbance and assess whether this was a useful predictor of frog abundance and presence. The probability of finding a frog within a 1.5 m³ plot was analysed using a generalised linear model (GLM) with logit link function. Sites disturbed by historical exploration mining, roading and housing-related activities did not have significantly lower abundance of L. archeyi than undisturbed sites i.e. those with no vegetation clearance after 1980. Abundance estimates were instead correlated with higher elevation and with plant species typically associated with mature forest, which were found in both disturbed and undisturbed sites. Ordination techniques used to assess vegetation composition revealed differences among sites that could be due to forest succession and replantation in disturbed sites. Mature forest species, such as tree ferns and rewarewa (Knightia excelsa), which were associated with higher frog abundance, also provided the substrates that frogs were most frequently found on. These plant species contribute to greater leaf litter depth, which was the only microhabitat variable that differed significantly between 1.5 m³ plots where frogs were present or absent. In comparison, sites that had been replanted with Kānuka (Kunzea robusta) and kauri (Agathis australis) did not provide substantial leaf litter. These results do not imply that land-use activities have no immediate effect on L. archeyi populations. However, over the time elapsed since disturbance, frogs of various age classes likely re-colonised sites after leaf litter build-up and reached densities determined by the local habitat quality. Because the disturbances considered were on a small scale (ca. 100 m² for mining exploration, roading and housing), frogs would have been available for re-colonisation from the surrounding landscape. Abundance estimation accounting for detection probability was a valuable method to increase our understanding of the historical effects of disturbance on L. archeyi and to determine the habitat features that instead influence abundance. Further, this information should be used in translocation site selection and survey methodology to predict abundance and presence of frogs in other areas. Based on my results I recommend that rehabilitation of sites following disturbance should be undertaken by natural forest succession and should emphasise retention and restoration of leaf litter.