| Formatted abstract
|
This thesis documents the findings of a study which had the purpose of testing the accuracy of the predictions provided by the Landscape Function Analysis (LFA) Soil Surface Condition (SSC) indicators. This study, which undertook investigations at four mine sites (Gregory, Brocks Creek, Nabarlek and Gove) located in tropical and sub-tropical Australia, was a component of a broader project, titled "Indicators of Ecosystem Rehabilitation Success (Stage 2), ACMER Project No. 31". A total of nine mine sites in Australia and Indonesia were involved overall, encompassing a climatic range from 200 mm to 4000 mm annual precipitation. LFA, the method investigated by this study, has been proffered as a useful rehabilitation monitoring tool for the mining industry. LFA uses quickly observed soil surface features to estimate the status of three important soil functional processes which have been proposed as indicators of ecosystem rehabilitation success. In order for this method to gain acceptance, the indicators must be shown to be accurate. This study was designed to test the veracity of predictions provided by the indices of soil stability, infiltration and nutrient cycling, using conventional scientific methods and measurements. Regression analysis was used to assess the nature of the relationships between the indices and the measured variables. Testing of the soil stability index was achieved using measures of aggregate stability. Saturated flow infiltration was measured in the field to obtain data for testing the accuracy of the infiltration index. Measures of soil respiration, microbial biomass and soil nutrient levels were used to assist in verifying the nutrient cycling index. This study revealed that the method, in its entirety, was only reliable at the Gove mine site, where each of the indices was successfully validated. It was reasoned that the high level of control over topsoil management and consistent rehabilitation techniques over time at this site, facilitated the expression of satisfactory relationships between the indices and their measures. At the three remaining mine sites, the accuracy of the estimates provided by the indices varied. The stability index was successfully verified at Nabarlek, partially verified at Gregory and not verified at Brocks Creek. Technical difficulties were experienced at Gregory and Brocks Creek while measuring infiltration, and thus the index was not able to be fully tested. The infiltration index was not verified at Nabarlek. With respect to the nutrient cycling index, it was only successfully verified by measures of soil respiration and mineralisable N at Gregory. In conclusion, it was found that the method was not suitable for application as a generic monitoring tool. The accuracy of the estimates provided by the indices varied considerably from mine site to mine site. The method was shown to be most successful where there was a high degree of substrate homogeneity across sites and where rehabilitation techniques had remained relatively constant over the range of rehabilitation sites investigated.
|
