To conserve an endangered species, we need to provide suitable habitat, shelter, prey items, and other resources (see e.g.
[1–4]). Prey availability may be one of the most critical issues, especially for predators with specialized diets
[5, 6]. If management plans for endangered species include the restoration of habitat, we need to know if the endangered taxon itself is vagile enough to locate and colonise the newly-available sites. Evaluating the likelihood that significant prey species also will colonise restored areas is also important; if they do not do so (perhaps because of poor dispersal capacity), otherwise-suitable habitat may be unable to support populations of the endangered taxon.
The Broad-headed snake (Hoplocephalus bungaroides, Elapidae) is a small elapid snake restricted to rocky areas (sandstone plateaux) within a 200 km radius of Sydney, in south-eastern Australia
. These snakes were abundant at the time of European colonisation 200 years ago, but have now disappeared from most of its former range
[7, 8]. The threatening processes include habitat degradation and fragmentation resulting from the removal and destruction of critical shelter sites (especially, exfoliated rock that forms thermally-suitable retreat sites during the coldest parts of the year:
), forest overgrowth
[3, 4, 9] and illegal collection of animals for the pet trade
. Efforts at habitat restoration have produced encouraging results, with the snakes and their lizard prey rapidly colonising sites by themselves where artificial rocks have replaced stolen natural rocks
 and where trimming of vegetation has allowed increased sunlight penetration
[3, 4]. However, these studies have focused on sites very close to extant populations of snakes and their prey; the prospectus for successful colonisation of more distant sites remains unclear.
For relatively isolated habitat patches to be colonised, both the snakes and their prey must be able to reach them. Landscape-genetic analyses have confirmed that broad-headed snakes often move between adjacent outcrops (distance between outcrops: 0.9 to 10.7 km), and thus are likely to rapidly find any restored habitat patches
. The probability of colonisation by the snakes’ prey species has not been studied, and is the subject of the present paper. Broad-headed snakes consume a diversity of vertebrate prey taxa, but the most important taxon (especially during cooler months of the year, when the snakes are restricted to rock outcrops) is the velvet gecko (Oedura lesueurii, Diplodactylidae:
). Indeed, velvet geckos comprised 70% of prey items consumed by juvenile H. bungaroides. Like H. bungaroides,
O. lesueurii is restricted to rock outcrops
[13, 14]. The predator–prey interaction between these two taxa presumably has been a long-running one, because geckos from populations sympatric with this snake species are reported to display a suite of antipredator tactics not seen in conspecific geckos from populations allopatric to broad-headed snakes (
; but see
 for data that challenge this conclusion). Local coadaptation is likely only when gene flow is restricted between populations (e.g.
[15, 17, 18]), allowing the evolution of spatial heterogeneity in relevant traits.
To evaluate the history of this predator–prey interaction, we need to know the timeline not only for the predator’s evolution
 but also for the prey’s evolution (current study). Because O. lesueurii is an important prey species for H. bungaroides, we also need to evaluate the potential for O. lesueurii to colonise newly restored areas of rocky habitat. We can clarify this issue with a study of landscape genetics (e.g., what are the spatial scales of current and historical rates of gene flow?) and direct measures of dispersal, based on mark-recapture fieldwork.