Response of lizard assemblages in the Mercury Islands, New Zealand, to removal of an introduced rodent: the kiore (Rattus exulans)

Abstract

Predation by kiore (Rattus exulans) has long been assumed to reduce the abundance and species diversity of lizards on islands around New Zealand. However, proof of this is hard to find, partly because it is difficult to compare lizard populations on different islands. In this paper I review methods for obtaining contemporaneous samples of lizard populations, compare the relative abundances of lizards on islands with and without kiore, and describe the responses of lizard populations to the eradication of kiore. On islands in the Mercury Group, northern New Zealand, kiore are absent from Middle I., present on Stanley I., and formerly present (now eradicated) from Korapuki I. Comparison of islands with kiore versus those without kiore shows that there are differences in addition to those of species richness of lizards. On Middle I. in 1986 and 1987, 65% of the lizard biomass comprised nocturnal/crepuscular species, whereas these species represented only 2% of the biomass on Korapuki I. The circumstantial differences between islands were tested before and after eradication of the kiore from Korapuki I. In the forested parts of this island some lizard species cannot coexist with kiore, and lizard populations in forest changed little in the three years following removal of rats. In coastal areas, however, habitat structure influences vulnerability of lizards to kiore, so removal of rats produced rapid changes in habitat use, population structure and productivity of some lizards. The data indicate that the effects of kiore depend on the size of the island, the extent of modification of habitat, and the availability of habitat refuges. A model presenting a series of hypotheses on the effects of kiore on the terrestrial reptiles of New Zealand islands is presented.