Identifying and managing soil salinity at multiple spatial scales on King Island, Tasmania

Soil salinity is a major determinant of agricultural productivity in many regions of Australia. Soil salinity is also spatially variable. This thesis examines the application of electromagnetic induction geophysical techniques and coincident soil sampling to (1) represent the areal extent and magnitude of soil salinity on the agricultural areas of King Island, Tasmania, (2) constitute its major causes; and (3) address mitigation strategies. An automated electromagnetic induction meter used in the vertical dipolar mode (EM31v) was used to capture apparent total soil conductivity data over 15,420 Ha of the island. A total of 447 soil samples were obtained from sixty-one soil sample holes typically to 350 cm below surface. Ordinary least squares-based regression methods were used to predict average ECe at the soil sample sites using the conductivity data (ECa) assessed by the EM31v (R2 = 0.76, p-value = 0.0001). A local, exponential semivariogram kriging model was developed to interpolate average ECe to 350 cm depth across the surveyed area. An analysis of geographic information layers, further terrain modelling, and climatic estimates of salt accessions were used to isolate the geological, geomorphological and climatic determinants of soil ECe on the island. Across the island the major source of salt is from west coast generated sea spray. Down to 350 cm, the highest average soil conductivity (ECe of 8 dS/m and above) were found to occur in soils formed on Proterozoic granite, Proterozoic shale and undifferentiated Quaternary sediments. A long-term climate prediction, compelled by climate change forecasts of less rainfall to flush salt from these lithologies by 2030, is that the risk of salinisation in these areas will increase (by 10 %). Terrain morphology was found to be a good predictor of high ECe on Proterozoic shale, but was found to be unrelated to ECe on other lithologies. More generally, high ECe in soils formed on granite were observed to occur at the valley floors and toward the crests of hills. Elevation and geological data were used to estimate the spatial location of high ECe to 350 cm depth across the entire island (103,000 hectares). Soil salinity is a major determinant of agricultural productivity in many regions of Australia. Soil salinity is also spatially variable. This thesis examines the application of electromagnetic induction geophysical techniques and coincident soil sampling to (1) represent the areal extent and magnitude of soil salinity on the agricultural areas of King Island, Tasmania, (2) constitute its major causes; and (3) address mitigation strategies. An automated electromagnetic induction meter used in the vertical dipolar mode (EM31v) was used to capture apparent total soil conductivity data over 15,420 Ha of the island. A total of 447 soil samples were obtained from sixty-one soil sample holes typically to 350 cm below surface. Ordinary least squares-based regression methods were used to predict average ECe at the soil sample sites using the conductivity data (ECa) assessed by the EM31v (R2 = 0.76, p-value = 0.0001). A local, exponential semivariogram kriging model was developed to interpolate average ECe to 350 cm depth across the surveyed area. An analysis of geographic information layers, further terrain modelling, and climatic estimates of salt accessions were used to isolate the geological, geomorphological and climatic determinants of soil ECe on the island. Across the island the major source of salt is from west coast generated sea spray. Down to 350 cm, the highest average soil conductivity (ECe of 8 dS/m and above) were found to occur in soils formed on Proterozoic granite, Proterozoic shale and undifferentiated Quaternary sediments. A long-term climate prediction, compelled by climate change forecasts of less rainfall to flush salt from these lithologies by 2030, is that the risk of salinisation in these areas will increase (by 10 %). Terrain morphology was found to be a good predictor of high ECe on Proterozoic shale, but was found to be unrelated to ECe on other lithologies. More generally, high ECe in soils formed on granite were observed to occur at the valley floors and toward the crests of hills. Elevation and geological data were used to estimate the spatial location of high ECe to 350 cm depth across the entire island (103,000 hectares).