2 INDEXES OF SOIL-STRUCTURE BASED ON PREDICTION OF SOIL-WATER PROCESSES

Evaluation of soil structure should reflect the nature and degree of s oil physical limitations to land use for local climatic conditions. Th is can be achieved through mechanistic simulation of soil-plant-atmosp here processes. We propose two alternative indices of soil structure, based on predicted infiltration. The potential runoff index (PRI) is t he runoff predicted from a 1-in-20-yr average recurrence interval stor m event of 30-min duration using an event-based soil water infiltratio n model that utilizes Richards' equation. The runoff recurrence index (RRI) is the average recurrence interval of a 30-min duration storm of intensity just sufficient to result in runoff. These indices were app lied to data from 37 sites (mainly Palexeralfs, but including Natrixer alfs, Rhodoxeralfs, and Haploxeralfs) in southeastern Australia, where rainfall partitioning between infiltration and runoff can significant ly affect agricultural production. Sites with different land use histo ries were ranked on the basis of the RRI. Woodland sites have a signif icantly greater mean ranking (mean ranking [RRI] = 33.5), indicating m ore favorable soil structure, compared with less conservative agricult ural land uses such as heavily grazed pasture (mean ranking = 11.7) or intensively cultivated cropping where stubble was not retained (mean ranking = 12.4). Both indices integrate the effects of soil structure, as represented through soil hydraulic properties, with local rainfall characteristics. They also account for soil horizon interactions infl uencing infiltration. The saturation-excess runoff generation predicte d for 12 of the 37 sites indicates that structural amelioration of B h orizon soil should be a high priority at these sites.