Field experiments in southern Australia examined the spatial distribution of soil-borne disease inoculum within paddocks using DNA-based soil assays. Paddocks were divided into zones using cluster analysis for a range of combinations of digital data layers. Inoculum levels differed among zones in 33–64% of the 108 cases examined, depending on the zone model used. It was concluded that zone models used for precision agriculture (PA) most commonly in Australia (viz. zones based on cluster analysis of grain yield maps, ECa, and elevation, and zones based on satellite biomass imagery) were most suitable for partitioning inoculum distribution within paddocks. Generally there was a correlation between pre-sowing levels of inoculum and crop root damage and shoot biomass; however, there was not always a strong correlation between inoculum level and grain yield. There was some evidence that damage/unit soil inoculum varied among zones, but difficulties in predicting this a priori suggest that the damage rate should be assumed to be equal among zones.

It is suggested that crop managers divide paddocks into yield or management zones and test each zone before every crop, using an appropriate soil sampling protocol. The disease risk and yield potential for each zone should then be considered to decide whether differential management is feasible or warranted. A soil test based on one composite paddock sample gives a paddock average only, which in many cases gives insufficient information about varying inoculum levels for robust zone management. If testing of every zone is not possible, then zones with the highest risk to profit from disease damage should be tested, to minimise risk. As PA technologies and biological understanding of disease behaviour improve, crop managers will have greater opportunities to exploit the non-random spatial distribution of soil-borne disease inoculum in new and imaginative ways at the zone level.