More than two-thirds of Australia is classified as arid or semi arid and receives less than 500 mm of rain annually. Throughout these arid areas, rivers and drainage lines (which often have their catchments in higher rainfall areas) act as arteries for vast amounts of water to flow into these dry landscapes. Many inland rivers, including the Paroo, Bulloo and Cooper Creeks, have highly variable flow rates and can go many years with very little or no flow. As a result of the hydrology of these areas, they contain different soil characteristics and vegetation communities from the surrounding arid landscapes. For example, soil nutrients are thought to be higher along drainage lines and on floodplains due to sediment deposition and decomposition of organic material. Also, vegetation in these areas needs to not only survive drought conditions (as a plant or propogule), but also periodic inundation and flood events. These characteristics are thought to make these areas more productive than the surrounding arid lanscapes and are thought to be important habitats for flora and fauna. There has been much debate over the importance of riparian corridors to vertebrate species. Several studies internationally in mesic areas have suggested that riparian corridors contain higher levels of biodiversity than the surrounding habitats, although this pattern is not replicated in all areas or for all taxonomic groups. Many of these studies suggest that the difference in biodiversity between riparian areas and adjacent habitats should be greater in xeric areas where the habitat differences between these areas are greatest. Other studies have suggested that diversity is not necessarily greater but riparian areas are important habitats for different species and may act as a refuge for some species. In Australia, very few studies have been conducted on the importance of riparian corridors to vertebrates in arid areas, despite these habitats being under threat from grazing, flow alteration and flow diversion. In this study I examined the role of the Field River in the Simpson Desert as a potential refuge for vertebrate species and compared diversity and population dynamics of vertebrates at dune and riverine sites. Specifically this study aimed to: 1) determine if species richness of vertebrates along the Field River was greater in riverine ecosystems than the surrounding dune habitats; 2) quantify how the riverine habitats differed in terms of floristics, vegetation complexity, soils and water availability from the surrounding landscapes; 3) determine the dynamics of terrestrial vertebrate populations along the Field River; and 4) determine if the abundance, diversity and demographics of small vertebrate populations along the Field River was different from populations in the sand dune fields on a broader landscape scale. These questions were examined on Ethabuka Station in the Simpson Desert, Queensland, from 2006 to 2008. The riparian corridor along the Field River was found to have a range of characteristics that make it unique from the surrounding dune habitats. Soils along the riverine corridor were found to be characterised by more than 20% clay while soils on the dune crest were characterised by no more than 5% clay. Further, soil carbon and nitrogen was significantly higher in the riverine corridor than in the dune habitats and decreased with distance from the catchment. Spinifex (Triodia basedowii) cover was low in the corridor but dominated the dune swale. Number of trees and cover by trees and non-spinifex grasses were also significantly higher in the riverine corridor when compared to the surrounding dunes. Following the rainfall, annual cover was also significantly greater and they persisted much longer than in the dunes. There was very little difference in the diversity or number of invertebrates between the different habitat types. Species richness was estimated to be highest (48 species) in the riverine centre and lowest in the floodplain (30 species). The riverine habitats had different species pools when compared to the dune habitats. Several species including Amphibolurus longirostris and Litoria rubella exclusively inhabited the riverine habitats while others such as the skinks Ctenotus ariadnae and Ctenotus dux were captured only in the dune habitats. Results from a Canonical Correspondence Analysis suggest that the distribution of some species in arid areas, such as the introduced House Mouse (Mus musculus), may be correlated with habitat characteristics associated with the riverine corridor, e.g. soil moisture or high annual cover. Abundance, body condition and reproduction of mammals along the Field River was generally driven by time. On a local scale, habitat had little effect on the abundance of the Sandy Inland Mouse (Pseudomys hermannsburgensis) and Lesser Hairy-footed Dunnart (Sminthopsis youngsoni), although M. musculus showed some preference for the riverine corridor. Abundance and reproduction of both rodent species increased following rainfall while S. youngsoni abundance was strongly seasonal with increases in abundance in autumn and winter. More than 45% of all captures were reptiles and this group was the most diverse and conspicuous fauna group along the riverine corridor. Abundance of the two most common lizard species, the Military Dragon (Ctenophorus isolepis) and Central-netted Dragons (Ctenophorus nuchalis), was significantly affected by time, with the abundance of both species decreasing dramatically 12 months after rainfall, possibly due to increased predation. Abundance of the Beaked Gecko, Rhynchoedura ornata, began to decrease prior to rainfall, suggesting a factor other than habitat structure or food availability causing the decline. The skink, Lerista labialis, showed strong seasonal trends in abundance and body condition, which suggested that within the desert riverine corridor the species was not reliant on rain-induced changes in food availability. To compare the effect of habitat on species richness, composition and population dynamics at a regional scale, sites along the riparian corridor were compared with dune sites located 500 m to 45 km from the Field River. On this scale, species species richness varied spatially with only one of the riverine sites having higher estimated species richness than the dune sites. Species turnover was greater in riverine sites despite the habitat data suggesting that primary production in these sites was more stable. Some species were able to persist at riverine sites at all times of the year but were only present in the community at dune sites following rainfall, suggesting that the riverine sites may be a more stable habitat for some species. Composition analysis revealed that dune and riverine habitats contain different communities, and some species, particularly Amphibolurus longirostris and Litoria rubella, probably rely on the river for their habitat requirements. Population dynamics including abundance, reproduction and body condition of seven species were compared between dune and riverine habitats. Only four, M. musculus, L. labialis, C. nuchalis and S. youngsoni, showed any significant affect of habitat on abundance. Abundance of C. nuchalis, S. youngsoni and L. labialis were generally more abundant in the dune habitats although the affect was only significant for some sampling sessions. The only species that was significantly more abundant in the riverine habitats than dune habitats was the introduced M. musculus. This species appears to be reliant on the riverine corridor, only appearing commonly in the dune habitats following rainfall. For most species, the number of juveniles captured increased following the rainfall in January 2007 but only in C. nuchalis and P. hermannsburgensis was there any significant difference between habitat types. For both species, generally more juveniles were captured in the dune habitats than in the riverine habitats. This study provided information on the role of the Field River to vertebrate fauna on Ethabuka Station. Although, the study focused only on one ephemeral desert river, it could be considered representative of similar desert rivers such as the Hay, Plenty and Hale rivers further west in central Australia. The riverine corridor did represent a different habitat to the surrounding sand dunes, and some characteristics, such as greater soil moisture, soil nutrients and annual cover, suggest that it may have greater and more stable primary production. Although the species studied did not generally have greater abundance and body condition, or reproduce more in the riverine corridor, many species including the Long-nosed Dragon (Amphibolurus longirostris), Desert Tree Frog (Litoria rubella) and Pygmy Mulga Monitor (Varanus gilleni) are likely to rely on the habitat that the riverine corridor provides. With expanding human populations placing greater pressure on rivers around the world, particularly in arid areas, it is imperative that we understand the ecology of these riparian systems so that effective management and conservation strategies can be developed.

More than two-thirds of Australia is classified as arid or semi arid and receives less than 500 mm of rain annually. Throughout these arid areas, rivers and drainage lines (which often have their catchments in higher rainfall areas) act as arteries for vast amounts of water to flow into these dry landscapes. Many inland rivers, including the Paroo, Bulloo and Cooper Creeks, have highly variable flow rates and can go many years with very little or no flow. As a result of the hydrology of these areas, they contain different soil characteristics and vegetation communities from the surrounding arid landscapes. For example, soil nutrients are thought to be higher along drainage lines and on floodplains due to sediment deposition and decomposition of organic material. Also, vegetation in these areas needs to not only survive drought conditions (as a plant or propogule), but also periodic inundation and flood events. These characteristics are thought to make these areas more productive than the surrounding arid lanscapes and are thought to be important habitats for flora and fauna. There has been much debate over the importance of riparian corridors to vertebrate species. Several studies internationally in mesic areas have suggested that riparian corridors contain higher levels of biodiversity than the surrounding habitats, although this pattern is not replicated in all areas or for all taxonomic groups. Many of these studies suggest that the difference in biodiversity between riparian areas and adjacent habitats should be greater in xeric areas where the habitat differences between these areas are greatest. Other studies have suggested that diversity is not necessarily greater but riparian areas are important habitats for different species and may act as a refuge for some species. In Australia, very few studies have been conducted on the importance of riparian corridors to vertebrates in arid areas, despite these habitats being under threat from grazing, flow alteration and flow diversion. In this study I examined the role of the Field River in the Simpson Desert as a potential refuge for vertebrate species and compared diversity and population dynamics of vertebrates at dune and riverine sites. Specifically this study aimed to: 1) determine if species richness of vertebrates along the Field River was greater in riverine ecosystems than the surrounding dune habitats; 2) quantify how the riverine habitats differed in terms of floristics, vegetation complexity, soils and water availability from the surrounding landscapes; 3) determine the dynamics of terrestrial vertebrate populations along the Field River; and 4) determine if the abundance, diversity and demographics of small vertebrate populations along the Field River was different from populations in the sand dune fields on a broader landscape scale. These questions were examined on Ethabuka Station in the Simpson Desert, Queensland, from 2006 to 2008. The riparian corridor along the Field River was found to have a range of characteristics that make it unique from the surrounding dune habitats. Soils along the riverine corridor were found to be characterised by more than 20% clay while soils on the dune crest were characterised by no more than 5% clay. Further, soil carbon and nitrogen was significantly higher in the riverine corridor than in the dune habitats and decreased with distance from the catchment. Spinifex (Triodia basedowii) cover was low in the corridor but dominated the dune swale. Number of trees and cover by trees and non-spinifex grasses were also significantly higher in the riverine corridor when compared to the surrounding dunes. Following the rainfall, annual cover was also significantly greater and they persisted much longer than in the dunes. There was very little difference in the diversity or number of invertebrates between the different habitat types. Species richness was estimated to be highest (48 species) in the riverine centre and lowest in the floodplain (30 species). The riverine habitats had different species pools when compared to the dune habitats. Several species including Amphibolurus longirostris and Litoria rubella exclusively inhabited the riverine habitats while others such as the skinks Ctenotus ariadnae and Ctenotus dux were captured only in the dune habitats. Results from a Canonical Correspondence Analysis suggest that the distribution of some species in arid areas, such as the introduced House Mouse (Mus musculus), may be correlated with habitat characteristics associated with the riverine corridor, e.g. soil moisture or high annual cover. Abundance, body condition and reproduction of mammals along the Field River was generally driven by time. On a local scale, habitat had little effect on the abundance of the Sandy Inland Mouse (Pseudomys hermannsburgensis) and Lesser Hairy-footed Dunnart (Sminthopsis youngsoni), although M. musculus showed some preference for the riverine corridor. Abundance and reproduction of both rodent species increased following rainfall while S. youngsoni abundance was strongly seasonal with increases in abundance in autumn and winter. More than 45% of all captures were reptiles and this group was the most diverse and conspicuous fauna group along the riverine corridor. Abundance of the two most common lizard species, the Military Dragon (Ctenophorus isolepis) and Central-netted Dragons (Ctenophorus nuchalis), was significantly affected by time, with the abundance of both species decreasing dramatically 12 months after rainfall, possibly due to increased predation. Abundance of the Beaked Gecko, Rhynchoedura ornata, began to decrease prior to rainfall, suggesting a factor other than habitat structure or food availability causing the decline. The skink, Lerista labialis, showed strong seasonal trends in abundance and body condition, which suggested that within the desert riverine corridor the species was not reliant on rain-induced changes in food availability. To compare the effect of habitat on species richness, composition and population dynamics at a regional scale, sites along the riparian corridor were compared with dune sites located 500 m to 45 km from the Field River. On this scale, species species richness varied spatially with only one of the riverine sites having higher estimated species richness than the dune sites. Species turnover was greater in riverine sites despite the habitat data suggesting that primary production in these sites was more stable. Some species were able to persist at riverine sites at all times of the year but were only present in the community at dune sites following rainfall, suggesting that the riverine sites may be a more stable habitat for some species. Composition analysis revealed that dune and riverine habitats contain different communities, and some species, particularly Amphibolurus longirostris and Litoria rubella, probably rely on the river for their habitat requirements. Population dynamics including abundance, reproduction and body condition of seven species were compared between dune and riverine habitats. Only four, M. musculus, L. labialis, C. nuchalis and S. youngsoni, showed any significant affect of habitat on abundance. Abundance of C. nuchalis, S. youngsoni and L. labialis were generally more abundant in the dune habitats although the affect was only significant for some sampling sessions. The only species that was significantly more abundant in the riverine habitats than dune habitats was the introduced M. musculus. This species appears to be reliant on the riverine corridor, only appearing commonly in the dune habitats following rainfall. For most species, the number of juveniles captured increased following the rainfall in January 2007 but only in C. nuchalis and P. hermannsburgensis was there any significant difference between habitat types. For both species, generally more juveniles were captured in the dune habitats than in the riverine habitats. This study provided information on the role of the Field River to vertebrate fauna on Ethabuka Station. Although, the study focused only on one ephemeral desert river, it could be considered representative of similar desert rivers such as the Hay, Plenty and Hale rivers further west in central Australia. The riverine corridor did represent a different habitat to the surrounding sand dunes, and some characteristics, such as greater soil moisture, soil nutrients and annual cover, suggest that it may have greater and more stable primary production. Although the species studied did not generally have greater abundance and body condition, or reproduce more in the riverine corridor, many species including the Long-nosed Dragon (Amphibolurus longirostris), Desert Tree Frog (Litoria rubella) and Pygmy Mulga Monitor (Varanus gilleni) are likely to rely on the habitat that the riverine corridor provides. With expanding human populations placing greater pressure on rivers around the world, particularly in arid areas, it is imperative that we understand the ecology of these riparian systems so that effective management and conservation strategies can be developed.