Tasmanian organic soils

<p dir="ltr">The character, extent and location of Tasmanian organic soils have been largely overlooked in Australian soil classification and taxonomy, with only a loose interpretation of northern hemisphere organic soil classifications applied. The aim of present work is to produce a classification of the organic soils, based on measurable soil properties and to then relate the characteristic organic soil properties to environmental factors. The relationship between organic soil characteristics and environmental factors will enable predictive mapping of the occurrence and organic content of organic soils in Tasmania. Tasmanian organic soils were sampled across 127 sites yielding a total of 1159 soil pits. Soil and environmental characteristics were recorded for each soil pit. Unsupervised clustering of the soil characteristics from each soil pit distinguished 23 organic soil groups. A classification key for identifying the 23 clusters was produced using the soil characteristics, soil organic carbon, humification, soil total nitrogen and organic soil depth.Dominant environmental factors influencing the 23 clusters were found, through vector analysis, smooth plate spline contouring and multinomial log-linear modelling to be: vegetation, burn frequency, topography, geology, altitude and climate. In order to predict the location and occurrence of the soils and soil characteristics produced through unsupervised clustering, the dominant environmental factors were subsequently used to provide cluster centroids for a supervised clustering. The resulting 41 soil groups were found to be distinguishable in terms of vegetation type, geology, topography and microtopography. The supervised clusters were found to perform better than the available vegetation classifications in predicting the unsupervised clusters. Organic soil carbon, bulk density and depth were used to model organic soil carbon stocks in Tasmania and provide a geographic context for the supervised and unsupervised soil clusters. Stepwise regression of soil organic carbon, showed slope as the dominant predictor across organic soil producing vegetation types. The regression models allowed for mapping of organic soil areal extent and soil organic carbon stocks in Tasmania, producing a value of 3,072 Tg of soil organic carbon over 8, 974 km².<br>Suggested changes to the Australian Soil Classification for the order organosol include the addition of folic, lignic, arenic and argyllic to the differentiae. Suggested family criteria include: humification of surface tiers, organic horizon thickness, botanical composition of surface layers, botanical composition of dominant layers and acidity classes below pH _CA 4.6. Changes to landform labels are also suggested.</p>