Fitzpatrick, RobertSmernik, RonMosley, LukeStiglingh, Andrea Donné2023-04-122023-04-122022https://hdl.handle.net/2440/137881Competition and predation pressures from introduced mammals (e.g. cats, foxes and rabbits) have been significant contributors to the high rates of small mammal extinctions observed in Australia over the past 200-250 years. Exclusion fences and the creation of feral-free nature reserves (‘safe havens’ for threatened native mammal species) are increasingly incorporated into wildlife conservation strategies and invasive species management programs in both Australia and New Zealand. Fence damage (metal dissolution and rusting) sustained on galvanised-steel fences, caused by contact with a corrosive soil environment, is a common management issue and can create opportunities for feral animal incursions onto conservation reserves if the netting has been sufficiently degraded. To reduce the frequency of fence netting repair and replacement due to corrosion damage, soil corrosion risk mapping can be used to pre-emptively identify ‘high maintenance’ areas to avoid when installing new fencing/reserves and identify the level of corrosion-resistance necessary for fence materials used in different environments. Zinc loss from galvanised steel fencing measured over time provides a useful proxy of soil corrosivity potential. In lieu of long-term fence experiments, the expected rates of electrochemical reactions in soils (and consequently the degree of fence corrosion risk) within specific environments can be inferred from a range of soil physicochemical indicators. These include soil electrical resistivity, soil drainage characteristics and soil pH. Methods used to quantify soil corrosion risk are often highly specific to the soil environment in which they have been developed, as the relative contributions of the aforementioned soil attributes to the overall corrosivity potential of the environment is strongly dependent on environmental conditions (e.g., climate). Consequently, no standard method for quantifying the corrosivity potential of a soil towards buried infrastructure exists in the literature. Furthermore, no method currently exists that classifies corrosion risks towards fencing in surface soils, nor has a risk classification method specifically been developed for use in Australian soils. This thesis outlines soil classification and risk mapping approaches that can be used by land managers to assess the degree of fence corrosion risk on their properties to inform the design and location of exclusion fences used in conservation programs. These approaches incorporate inexpensive soil testing and open-source soil data with corrosion risk classification methods developed specifically for South Australian soils. These approaches include a qualitative risk classification index developed for use in arid soil environments (Chapter 2) and an updated, and more generalised quantitative risk classification tree (Chapter 3). The quantitative decision tree was used to provide an evaluation of fencing material suitability in arid soil environments (Chapter 4) and as part of a Geographic Information System (GIS) assessment case-study (Chapter 5), which outlines the processes for mapping soil corrosivity risk using open-source soil survey data. Preliminary corrosion product compositional data was used to identify some of the underlying soil processes driving soil corrosivity potential in different environments (Chapter 3). Initial experiments were conducted in three coastal and two anthropogenic acid sulfate soils, to identify key electrochemical processes occurring in these highly acidic and potential acidic soils (Chapter 6).enFence corrosion; soil corrosivity; exclusion fencing; corrosion index; risk mappingThesis