Laurence T.P. Jassogne1, Rob Davidson2, Ann McNeill2, and David Chittleborough2. (1) University of Western Australia, c/o University of Adelaide, Waite Campus, Glen Osmond, Australia, (2) University of Adelaide, Waite Campus, Glen OSmond, Australia
Sodic soils (principally Natrixeralfs) are common in the rainfall cereal cropping zone of southern Australia. The strong texture contrast between the E and Btn horizons, and the high bulk density and low plant available water capacity of the upper B horizon are inimical to penetration by many crop plants the root systems of which are primarily confined to the A and upper E horizons. Many native perennial plants appear to thrive on these ‘hostile' soils. Field inspection of some native species revealed that their roots are commonly located in the B horizon as well. We hypothesize that the roots of crop and pasture plants that do penetrate the Btn manage this by following macro and meso channels formed originally by native plants and are now composed of decaying organic matter. These macropores form pathways of least resistance through the dense E-Btn interface and, furthermore, provide nutrients for living roots. The work presented here is part of a study into the physico-chemical nature of B horizons and their amelioration to allow greater root exploration. The aim of our study was to characterise the porosity of sodic duplex soils in three dimensions (3D), identify differences in root behaviour between a native species (saltbush), a perennial pasture plant (lucerne) and an annual oil seed crop (canola) and investigate the relationship between pre-existing pores and root exploration of the Btn horizons. Intact cores of length 50 cm and diameter 15 cm were taken in the field. Computer tomography (CT) was used to characterise the macroporosity in those intact cores non-invasively and non-destructively. The size of the pixels was 0.3 mm in the x and y dimension. The distance between two subsequent slices was 0.4 mm. The CT images were segmented applying an adaptive threshold technique based on indicator Kriging. The porosity, pore size distribution, pore length, pore inclination and tortuosity were calculated (3D). After scanning of the intact cores, one plant was allowed to grow in each of for 10 weeks. A core of each plant was scanned every 3 weeks to visualise root exploration. Root length, root volume and root distribution were calculated and the root structures were visualised in 3D. Finally, the scans of the intact cores were compared with the scans of the same cores with a plant grown in them. Results show the importance of the association of roots with pre-existing soil structure, especially old root channels.