Silcrete: an Australian perspective

We know how most rocks are formed. Silcrete is something of an enigma, for although there are many interpretations of the origins of individual silcrete bodies, such as those in the Paris Basin, England, Botswana and central Australia, an overarching hypothesis capable of explaining all occurrences is still to be found. This paper reviews the literature of research on predominantly Australian silcretes as well as reviewing their occurrence, mineralogy, geochemistry and petrology. Silcrete ages and paleoclimatic significance are also reviewed. Most silcretes are formed low in landscapes along fluvial tracts or lakes but, some may form at breakaway margins as a result of lateral groundwater movement. Following silicification and landscape inversion, many silcretes are left high in the landscape. Most silcretes must form in climates where there is an abundance of water, perhaps seasonally, and of organic acids. The age of a silcrete can be constrained by the fossils it may contain; ages of Australian silcretes so established range through most of the Cenozoic. Lacking fossil evidence, sediments of known stratigraphic age that have been silcreted can only provide a maximum age for the silcrete. Many silcretes in eastern Australia are overlain by basalt, but the age of the basalt can only give the minimum age of the silcreted host below it, not of the silcrete. Silcretes commonly exhibit a number of fabrics; externally glerp structures (also called cockade, ropy or botryoidal), and internally pedogenic and geopetal titaniferous grain-cap fabrics. We conclude that silcretes are formed by the precipitation of silica in various forms, almost always along with titania as anatase, at the time of cementation. Anatase occurs either where it is precipitated or by illuviation, commonly becoming concentrated as geopetal caps or coatings on larger detrital framework grains. This implies that the fluids moving the cementing components largely move downward through the silcreted host. Alternating Ti-rich and Ti-poor laminae in the caps show this process can be repetitious.