Abstract:
Karori Sanctuary (252 ha) is a fenced restoration site in Wellington, New Zealand from which all species of introduced mammals have been eradicated except house mice (Mus musculus). In 2006, the endemic New Zealand frog Leiopelma pakeka was transferred to Karori Sanctuary as part of a long term plan to restore the site's original biota. This was a significant event in that it was the first re-introduction of a New Zealand frog to a mainland site, the first New Zealand amphibian translocation for the purpose of restoration and the first time L. pakeka were released into habitat also occupied by an introduced mammal. An adaptive management regime facilitated research within the constraints of a community restoration project for which only a small population (n=60) was made available for release. Two groups (n = 30) were released into mouse-proof enclosures in February and October, 2006. Survival was high (97%) and frogs maintained a healthy body condition. Breeding was not detected during the first year and this was attributed to an inappropriate sex ratios that were restructured in April 2007 when half of the frogs (n= 29) were removed from the enclosures and released into forest habitat. The survival, condition and recruitment of frogs living inside and outside of the mouse-proof enclosures were compared. Both groups initially had a similar recapture rate, but after one year, just one frog (3%) was recaptured outside the enclosure compared with 27 adults (93%) and fourteen juveniles captured within the enclosure. In March 2009, 26 of the 29 individuals originally released into the enclosure were recaptured and a further ten juveniles were captured for the first time. No individuals have been sighted outside the enclosure since March 2008. Post-release movements did not explain the apparent decline of the population living outside of the enclosure. The mean distance dispersed during the first month after release (3.4 +/- 0.05 m) did not significantly increase after eight months (4.2 +/- 0.05 m) and the maximum-recorded dispersal distance was 7.0 m. The centre of activity of the nine frogs captured > 5 occasions were all within 3 m of the release site and kernel estimates of high habitat usage clustered around artificially constructed rock piles. Predation by house mice and/or native species such as little spotted kiwi (Apteryx owenii) were considered the most likely explanation for the failure to recapture frogs outside of the enclosure, especially those frogs that appeared to have settled at the release site.
The extremely low number of individuals released outside of the enclosure exacerbated the impact of processes acting on the founding population. Recommendations are provided for the next adaptive management stage and include transferring an additional 100 frogs from Maud Island for release into forest habitat outside of the mouse-proof enclosure. Post-release movements should be restricted and all potential predators except house mice excluded. The population within the enclosures should be retained as is. Finally, the viability including L. pakeka in attempts to reconstruct mainland communities is examined.
