Assessing the Australian Soil Classification using cladistic analysis

Abstract

The Australian Soil Classification (ASC) has its roots in both the Handbook of Australian Soils and the Factual Key. The scheme’s use of mutually exclusive characteristics has led to Soil Orders containing a diverse range of soils, such as the Dermosols. The extent of these groupings has resulted in classes of soils sharing greater relationships with soils from other classes than they do with soils in the same class. Situations such as this arise from artificial classifications and highlight the need for natural classifications. Natural classifications accurately represent what is occurring in nature and are desirable because they represent evidence of a common history, process or mechanism. This study uses cladistics, a robust biological method that uncovers natural classifications, to assess the naturalness of the ASC. The analysis has the secondary aims of identifying natural soil orders and establishing which characters and tiers require revision. Two measures commonly used in cladistics, consistency index (CI) and retention index (RI), are used along with confidence levels generated by bootstrapping. The cladistic analysis undertaken consisted of coding 113 morphological and non-morphological characters used to identify 13 of the 14 Soil Orders in ASC into binary and multi-state matrices and analysis using a parsimony cladistic algorithm. The results suggest that, because of its low CI (0.196), the ASC is not a natural classification. However, certain Soil Orders of Organosols, Podosls and Vertosols, which all registered high CI, are natural. The analysis also indicated which soil morphological characters and Soil Orders require revision in order to make the ASC more natural, namely, soil colour and characters located in the Great Groups as well as Soil Orders such as Chromosols, Ferrosols and Dermosols. We conclude that cladistics offers a new avenue in discerning relationships between soils and in assessing the accuracy of, and identifying where improvements can be made in, the classifications used to identify them.