In arid landscapes limited resources are concentrated into fertile patches that allow greater productivity than would otherwise be expected. Fertile patches are created and maintained by processes that modify soil topography and interrupt the flow of resources across the landscape. These processes may be abiotic or biotic in origin. Species that modify, maintain or create habitat have been termed ecosystem engineers. One group of soil disturbing ecosystem engineers creates fertile patches by modifying soil microtopography through foraging for food and creating habitat. This thesis examines the effects of soil foraging animals on resource concentration in arid environments and how effects are moderated by species and landscape. Chapter 1 provides an overview of the importance of fertile patches in arid systems and describes previous research on the effects of ecosystem engineers in arid environments. Chapter 2 examines how fertile patch creation by a well-studied arid zone engineer, the ant, varies between ant species and landforms, demonstrating that while we can generalise about the effects of ant nests on water flow and nutrient levels, differences in soil type, nest density and ant species across sites are likely to moderate these effects. Chapters 3, 4 and 5 compare the engineering effects of reintroduced native species, the Greater bilby (Macrotis lagotis) and Burrowing bettong (Bettongia lesueur) with the Sand goanna (Varanus gouldii) and the invasive European rabbit (Oryctolagus cuniculus) across three Australian landforms, revealing differences between reintroduced and invasive engineers at the both patch and landscape scales. At the landscape scale, differences in pit densities and resource concentration in foraging pits between landforms suggest that fertile patch development is not uniform. Chapter 6 examines whether the structure of the pit or the presence of litter per se explains observed increases in plant germination in foraging pits, and compares temperature and soil moisture conditions in bilby and bettong pits to the soil surface. Chapter 7 is a review that builds upon previous chapters, published and unpublished literature of the potential for reintroductions of ecosystem engineers to restore ecosystem function in degraded environments. Chapter 8 summarises the research presented in this thesis and its implications, and suggests directions for future work.

Arid Recovery, a fenced reserve free of feral predators in arid South Australia, has successfully reintroduced two critical weight range (CWR) mammals, greater bilbies (Macrotis lagotis) and burrowing bettongs (Bettongia lesueur) to facilitate the restoration of arid Australian ecosystems. This thesis evaluates the ecological roles of these reintroduced species and their relationship to ecosystem functioning and the restoration of these ecosystems. Surveys of foraging diggings and soil seed banks, and dietary analysis were used to measure impacts of these species on three main habitats within the Reserve. The results showed that bilbies and bettongs have three major roles in ecosystem functioning: consumers, ecosystem engineers, and dispersers of seeds and fungi. Both bilbies and bettongs were omnivorous, though their diets were distinctly different, with the bilbies focused more on invertebrates and seeds, while the bettongs consumed a greater proportion of coarser plant materials. The seed portion of the diets of both species during 2003-04 differed from a similar study three years previously, shortly after the animals had been reintroduced to Arid Recovery in 2000- 01. During 2003-04, the bilbies consumed a lower proportion of seeds of species of grasses and more of forbs, while seed consumption by bettongs narrowed to be almost exclusively the seeds of shrubs. The number of bilby and bettong diggings varied significantly both spatially and temporally, with averages of 7,530 ± 820 diggings ha-1 in Dunes, 10,560 ± 980 diggings ha-1 in Mulga, and 7,120 ± 610 diggings ha-1 in Swale. This resulted in an average of 2 to 3% disturbance of soil surface area, which is similar to or higher than reported for other Australian or overseas semi-fossorial species. The temporal variation in rates of digging was correlated with minimum daily temperatures and rainfall but not moonlight. This variation was also correlated with counts of bilby but not bettong tracks, suggesting that the temporal variability of diggings was related to levels of bilby activity. The persistence of bilby and bettong diggings differed between the three habitat types, with the majority of the diggings in the Mulga (94%) and Swale (87%) persisting for over 12 months, while Dune diggings filled much faster, with only 15% persisting over 12 months. These diggings were shown to accumulate and bury seeds and litter, and, under some conditions, enhance germination. The seeds of two plant species, the annual grass, Dactyloctenium radulans, and the shrub, Enchylaena tomentosa, germinated from bilby faecal pellets, and over a third of bilby faecal pellets contained fungal spores. Therefore bilbies have the potential to be dispersers of both seeds and fungi. All three major roles of the bilbies and bettongs in ecosystem functioning (consumers, ecosystem engineers, and dispersers of seeds and fungi) have the potential to affect the flows of organic, soil and water resources, and therefore vegetation structure and overall productivity. Soil seed bank densities differed between areas with and without bilbies and bettongs. However, the heterogeneity of the system made it difficult to confidently relate these differences to any particular effects of the bilbies and bettongs. Both bilbies and bettongs were able to locate and dig seeds buried 20 cm deep, and the caches of seed-harvester ants. Experiments showed that in areas of high digging density, 71 to 94% of seed rain accumulated and became buried in diggings. Since bilbies and bettongs have the potential to use buried seed resources, they have the potential to significantly affect soil seed banks through their consumption of seeds, redistribution of seeds through their digging activities and their interactions with other granivorous species. This study is a first step towards understanding the roles of reintroducing CWR mammals to arid ecosystems. Possible longer term effects of these reintroductions will depend on suitable regulation of animal numbers, and climatic patterns, as restorative effects of diggings would be greatest during periods of good rainfall, whereas droughts would slow restorative processes. Although the results of this research are unique for Arid Recovery, the principles of evaluating all major ecological roles of reintroduced species and their interactions with their environment could provide guidance for other reintroductions. These interactions are complex and would require longer-term studies over a range of conditions and locations to further understand the role of reintroducing CWR mammals to ecological restoration. / http://proxy.library.adelaide.edu.au/login?url= http://library.adelaide.edu.au/cgi-bin/Pwebrecon.cgi?BBID=1348600 / Thesis (Ph.D.) - University of Adelaide, School of Earth and Environmental Sciences, 2009

Arid Recovery, a fenced reserve free of feral predators in arid South Australia, has successfully reintroduced two critical weight range (CWR) mammals, greater bilbies (Macrotis lagotis) and burrowing bettongs (Bettongia lesueur) to facilitate the restoration of arid Australian ecosystems. This thesis evaluates the ecological roles of these reintroduced species and their relationship to ecosystem functioning and the restoration of these ecosystems. Surveys of foraging diggings and soil seed banks, and dietary analysis were used to measure impacts of these species on three main habitats within the Reserve. The results showed that bilbies and bettongs have three major roles in ecosystem functioning: consumers, ecosystem engineers, and dispersers of seeds and fungi. Both bilbies and bettongs were omnivorous, though their diets were distinctly different, with the bilbies focused more on invertebrates and seeds, while the bettongs consumed a greater proportion of coarser plant materials. The seed portion of the diets of both species during 2003-04 differed from a similar study three years previously, shortly after the animals had been reintroduced to Arid Recovery in 2000- 01. During 2003-04, the bilbies consumed a lower proportion of seeds of species of grasses and more of forbs, while seed consumption by bettongs narrowed to be almost exclusively the seeds of shrubs. The number of bilby and bettong diggings varied significantly both spatially and temporally, with averages of 7,530 ± 820 diggings ha-1 in Dunes, 10,560 ± 980 diggings ha-1 in Mulga, and 7,120 ± 610 diggings ha-1 in Swale. This resulted in an average of 2 to 3% disturbance of soil surface area, which is similar to or higher than reported for other Australian or overseas semi-fossorial species. The temporal variation in rates of digging was correlated with minimum daily temperatures and rainfall but not moonlight. This variation was also correlated with counts of bilby but not bettong tracks, suggesting that the temporal variability of diggings was related to levels of bilby activity. The persistence of bilby and bettong diggings differed between the three habitat types, with the majority of the diggings in the Mulga (94%) and Swale (87%) persisting for over 12 months, while Dune diggings filled much faster, with only 15% persisting over 12 months. These diggings were shown to accumulate and bury seeds and litter, and, under some conditions, enhance germination. The seeds of two plant species, the annual grass, Dactyloctenium radulans, and the shrub, Enchylaena tomentosa, germinated from bilby faecal pellets, and over a third of bilby faecal pellets contained fungal spores. Therefore bilbies have the potential to be dispersers of both seeds and fungi. All three major roles of the bilbies and bettongs in ecosystem functioning (consumers, ecosystem engineers, and dispersers of seeds and fungi) have the potential to affect the flows of organic, soil and water resources, and therefore vegetation structure and overall productivity. Soil seed bank densities differed between areas with and without bilbies and bettongs. However, the heterogeneity of the system made it difficult to confidently relate these differences to any particular effects of the bilbies and bettongs. Both bilbies and bettongs were able to locate and dig seeds buried 20 cm deep, and the caches of seed-harvester ants. Experiments showed that in areas of high digging density, 71 to 94% of seed rain accumulated and became buried in diggings. Since bilbies and bettongs have the potential to use buried seed resources, they have the potential to significantly affect soil seed banks through their consumption of seeds, redistribution of seeds through their digging activities and their interactions with other granivorous species. This study is a first step towards understanding the roles of reintroducing CWR mammals to arid ecosystems. Possible longer term effects of these reintroductions will depend on suitable regulation of animal numbers, and climatic patterns, as restorative effects of diggings would be greatest during periods of good rainfall, whereas droughts would slow restorative processes. Although the results of this research are unique for Arid Recovery, the principles of evaluating all major ecological roles of reintroduced species and their interactions with their environment could provide guidance for other reintroductions. These interactions are complex and would require longer-term studies over a range of conditions and locations to further understand the role of reintroducing CWR mammals to ecological restoration. / http://proxy.library.adelaide.edu.au/login?url= http://library.adelaide.edu.au/cgi-bin/Pwebrecon.cgi?BBID=1348600 / Thesis (Ph.D.) - University of Adelaide, School of Earth and Environmental Sciences, 2009