Ecological implications of habitat fragmentation and restoration for the gecko Hoplodactylus maculatus

Abstract

Habitat modification has created heterogeneous landscapes that either restrict animals to remnants of suitable habitat or challenges them to use live created matrices. To successfully predict effects of landscape change, such as habitat fragmentation and reforestation, it is important to understand the habitat choice and movement patterns of representative taxa. New Zealand geckos are functionally important parts of endemic ecosystems; however, much is unknown about their ecology and population biology. Furthermore, sampling difficulties have restricted their study. Here I compare the efficiency and reliability of common sampling methods for the New Zealand common gecko, Hoplodactylus maculatus, and determine its habitat choice in three vegetation types (grassland, young replanted forest, older remnant forest) on Mana Island and Stephens Island in Cook Strait, New Zealand. I then investigate gecko distributions in relation to habitat edges. Artificial cover objects (ACOs) prove to be a useful method to survey and monitor gecko populations, especially in younger forests. Hoplodactylus maculatus is abundant in young and older forests, however, mostly absent from grassland. Demographic proportions differ between sites, in that juveniles are more abundant in earlier successional vegetation than in remnant patches. Body condition is overall lower in replanted native vegetation and in grassland than in remnant forest, while more individuals are affected by external parasitism by mites (Neotrombicula spp.) in remnant forests. Geckos are predominantly found above ground in forest patches, and particularly at the edges. Demographic patterns differ between habitats. Adult geckos are found at similar abundances throughout forest patches while juveniles are most common at habitat patch edges. Significantly fewer geckos of all age classes were found in the grassland than in either of the forest habitats. Of the few geckos that were found in grassland, the majority were juveniles. This may imply density dependent natal dispersal into suboptimal habitat. This research shows that ACOs are a useful method for sampling Hoplodactylus maculatus. Gecko demographics and body condition are not randomly distributed and specific habitat preferences were identified. Geckos are mostly absent from grasslands and, as natal dispersal is directed towards habitat edges into this matrix, habitat fragmentation is predicted to affect this taxon negatively. Restoration of native forest will positively influence abundance of Hoplodactylus maculatus, while connectivity of native vegetation will aid dispersal and distribution.