W. H. Verboom's research while affiliated with University of Western Australia and other places

This review draws attention to information from the literature and our own observations supporting the view that higher plants and micro-organisms display an intrinsic capacity to be proactively involved in pedogenetic processes. ‘Bioengineering’ of this kind is deemed to be spearheaded by principal deep-rooted tree and shrub species and to result in optimisation of command and conservation of water and nutrients within an ecosystem. Specific examples discussed in the paper include, the formation of silicon- or iron-based linings of vertical channels and pores, binding of sand on roots, generation of organically derived hydrophobicity, development of clay-based hardpans and texture-contrast seals, precipitation of silcrete, calcrete and ferricrete pavements, effective accessing and conservation of P resources, including mining by microbes and the biological cycling of Si and Al via plants and micro-organisms. In each case, definitive roles and mechanisms are suggested for the organisms involved, particularly in relation to formative effects relating to secretion of organic acids, dispersing agents and other classes of exudate. We introduce the term ‘phytotarium’ to connote the collective outcomes of the above biotic influences in construction and maintenance of niches peculiar to specific vegetation types and then review the evidence of imprinting of soil profiles due to operation of phytotaria. Examples given relate to the lateral and vertical facies encountered in certain contemporary soil profiles and paleosols with which we are familiar and are described in a companion paper.

Soil profiles and rooting morphologies were examined under an ecotone where open woodland of multi-stemmed, small, lignotuberous eucalypts (mallee) graded into proteaceous heath. Soils under the mallee showed a Solonetz-type seal which separated, hydrologically, the upper acidic horizon of bleached sand from lower alkaline horizons rich in calcrete, silcrete, finely divided carbonates and clay. Seal composition appeared to vary consistently with overlying species of mallee. The generally acidic lateritic profiles under heath were rich in pisolithic ferricretes and displayed Fe-coated root channels. Both sets of taxa exhibited dimorphic rooting patterns, with ectomycorrhizal roots and seal-penetrating, second-order tap roots developed on the extensive lateral roots of mallee versus a dominance of primary tap roots and cluster root development on laterals of Proteaceae. Overprinting of ferricrete by clays and silicified material was evident where mallee appeared to have invaded areas of heath. Examination of other contemporary lateral facies changes and vertically-stacked paleosol formations in the study region provided corroborating evidence of similar profile attributes, including presence of Fe- or Si-lined root channels, overprinting phenomena and consistency in occurrences of ferricrete and calcrete as expected of each class of vegetation. Observations were related to the concepts of bioengineering of soil profiles through activity of macroflora and associated micro-organisms as set out more generally in our companion review.

As part of a general soil and regolith mapping exercise across cleared and remnant bush land, radiometric data for distribution of potassium (K), uranium (U) and thorium (Th) were examined alongside relief models and floristic and soil surveys in test catchments at Elashgin and Toolibin in the Western Australian wheat-belt. The Elashgin survey showed that highly weathered low K soils co-concentrated U and Th and were vegetated mainly by cluster root-bearing Proteaceae and Casuarinaceae. In granitic soils ratios of U to Th were higher and cluster root bearing taxa much less prominent, except where ferricrete gravels were concentrated. Draping of radiometric imagery over a digital elevation model showed spiral waveforms of high and low U and Th signal which were largely independent of topography but demarcated different oligotrophic communities. General observations and a detailed 900-m transect along an aeolian sand plume at Toolibin showed very high U and Th in ferricrete gravels where Proteaceae were dominant, but failed to separate proteaceous versus myrtaceous shrublands on deep sands due to truncation of signal. Augering along the transect and examination of floristic, soil and signal composition of 32 sites in the Lake Toolibin catchment confirmed presence and degree of development of ferricretes in the rhizosphere of Proteaceae-dominated communities and showed high reactivity scores for Al in `Bs' horizons in profiles carrying such vegetation. Highly specific associations between Proteaceae and very high U and Th were generally evident on exposed ferricrete gravels. The data are discussed in relation to the effects of root exudates on ferricrete formation and destruction and how the broader spatial pattern of such processes might relate to competition for soil phosphorous.

This communication presents the hypothesis that certain Australian lateritic and related oligotrophic soils may have been partly derived biotically from soluble iron-rich complexes generated following secretion of low-molecular weight organic acids by phosphate-absorbing specialised proteoid (cluster) roots of proteaceous plants. Subsequent precipitation of the iron is then pictured as occurring onto the oxide rinds of developing laterite after consumption of the organic components of the complexes by soil bacteria. The hypothesis is f irst examined in relation to current theories of origins of laterites and the extent of the coincidences worldwide in past and present times between Proteaceae and oligotrophic soil types of lateritic character. The paper then provides more definitive lines of evidence supporting the hypothesis, based largely on recent studies by the authors in south-western Western Australia. This relates to (a) cases of definitive association in habitats rich in Proteaceae between zones of root proliferation and ferricrete layers in lateritic soils, (b) proximity in soil profiles between ferric deposits and current and ancestral root channels, (c) the recovery of citrate-consuming bacteria from soil profiles and specifically from ferricrete rinds and horizons accumulating sesquioxide organic matter and (d) distribution of iron and phosphorus within plant and soil profile components consistent with ferricrete rinds being generated by rhizosphere-mediated interactions of plants and microbes under conditions of severely limited availability of phosphorus. The mode of functioning of proteoid root clusters is then discussed, especially in relation to exudation of organic acid anions, uptake of phosphorus and the subsequent fate of organic anions and their metal ion complexes in the system. An empirically based scheme is presented indicating flow profiles for phosphorus and iron between soil, ferricrete rinds and bacterial and plant components. We then discuss possible carbon costs to proteaceous plant partners when accessing phosphorus under the nutrient-impoverished conditions typical of heathlands and open woodlands of Mediterranean-type ecosystems of Western Australia. The paper concludes with a critical overview of the hypothesis, particularly its implications regarding possible higher plant: microbial influences shaping soil and landscape evolution in the regions involved.