Lizard diversity on a rainforest–savanna altitude gradient in north-eastern Australia A. S. Kutt A,C , B. L. Bateman B and E. P. Vanderduys A A CSIRO Ecosystem Sciences, PMB PO, Aitkenvale, Qld 4814, Australia. B Centre for Tropical Biodiversity and Climate Change, School of Marine and Tropical Biology, James Cook University, Townsville, Qld 4811, Australia. C Corresponding author. Email alex.kutt@bigpond.com Abstract. Mountain ecosystems act as natural experiments for investigating the relationship between environmental heterogeneity and species diversity. A review of the global altitudinal distribution of reptiles identified a diverse range of patterns driven by climate and taxonomy. No Australian examples were included in this analysis. We addressed this gap by surveying the reptile assemblage along an altitude gradient from upland rainforest (~1000 m) through to open savanna woodlands (~350 m) in north-eastern Australia. Reptiles were sampled on four separate occasions between May 2006 and November 2007. Thirty-six species, representing seven families, were recorded along the gradient. As we used only diurnal active searching, snakes and nocturnal geckoes were probably under-sampled; thus we considered only lizards in the analysis of altitude pattern. Lizard species richness peaked at the mid-altitudes (600–900 m, 11–12 spp.) and abundance highest at the lower (<500 m) and higher (>800 m) zones. This pattern is likely a factor of both the increase in radiant heat sources (reduced canopy cover) and increased species packing due to the diversity of niches available (presence of rock cover and increase in saxicolous species). In the lower-altitude sites the high abundance of few species seems linked to the dominance of disturbance-tolerant species. We conclude that lizard richness and abundance patterns on this transect are not necessarily exhibiting a mid-domain effect, but instead are a function of species-specific ecological and habitat requirements. Additional keywords: tropical, species richness, abundance, niche, Carlia, ecotone. Introduction Mountain ecosystems can act as natural experiments for investigating processes that determine species assemblage along environmental gradients (Shmida and Wilson 1985). Rapid changes in altitude and associated habitat can provide insight into the relationship between environmental heterogeneity and local or regional species diversity (Bateman et al. 2010). There is often a strong relationship between altitude and changes in climate and vegetation (Körner 2007), so species assemblages can shift rapidly over relatively short distances (Bullock et al. 1995; van Ingen et al. 2008). These zones of rapid transition can result in areas of increased species diversity (Heaney 2001), though there is conflicting evidence and debate regarding the determinants of the diversity along such gradients (Rowe et al. 2010). Many studies of different biotic assemblages along altitudinal gradients have identified mid-altitude peaks in species richness (e.g. McCain 2009). These peaks can occur either at points of optimal environmental conditions or at locations where distinct vegetation communities occur in close proximity causing high species overlap and turnover (Lomolino 2001; Currie and Kerr 2008). Few Australian studies or examples have examined vertebrate patterns, although a study in north-eastern Queensland showed a peak in richness of small mammals towards the summit of the gradient and at the point of most contrasting environmental conditions and vegetation change (Bateman et al. 2010). Although there are varying hypotheses on the underlying cause of the mid-altitudinal peak, a simple universal explanation may be difficult to attain due to local environmental conditions, biogeographic history and anthropogenic impacts (Rowe et al. 2010). A review of global reptile distribution identified four altitudinal patterns of diversity: decreasing, low plateau, low plateau with a mid-altitude peak, and unimodal with a mid- altitude peak (McCain 2010). The distribution patterns were dependent on taxonomic and functional groups present, and temperature was the strongest environment correlate, which was not unexpected for ectothermic species (McCain 2010). Species- rich reptile faunas are often common at low altitudes (Barnosky et al. 2001), and in Australia, high reptile diversity occurs in relatively flat deserts, though these regions traverse a scope of intermediate altitude (~0–800 m) (Pianka 1969). In extensive tropical savannas the reptile fauna is largely undifferentiated in Eucalyptus woodland habitats, and patterns are more linked to distinct changes in substrate and terrain (Woinarski and Ó CSIRO 2011 10.1071/ZO11036 0004-959X/11/020086 CSIRO PUBLISHING www.publish.csiro.au/journals/ajz Australian Journal of Zoology, 2011, 59, 86–94