Australia’s RootBot Maps How 36 Wheat Varieties Use Water
A two-season Australian field experiment has shown that wheat varieties differ not only in how much water they use, but also in when they use it and how deeply they draw it from the soil. That distinction matters because two crops can consume similar totals yet carry very different risks when flowering and grain filling arrive.
The University of Queensland announced the findings on 3 November 2026 after experiments at Gatton in 2024 and 2025. Researchers assessed 36 wheat varieties at different plant densities using electromagnetic-induction sensors on the autonomous high-clearance RootBot and drone-based sensing. The results were presented at the Australian Agronomy Conference in Darwin and ACB Breeders Week on the Gold Coast.
A map through depth and time
The platform combines above-ground and below-ground observations. Drones track canopy development and radiation capture. RootBot repeatedly carries electromagnetic-induction sensors through plots, allowing the team to follow changes in soil moisture at multiple depths without excavating every root system.
A single soil reading describes one moment. Repeated measurements create a trajectory: when the crop draws water, which layers supply it and whether the profile is depleted early or conserved for later stages. UQ reported differences among cultivars in total water use, timing and depth.
The early-canopy trade-off
Higher plant density typically increased early-season water use, radiation capture and canopy development. Faster cover can intercept more sunlight, build biomass and potentially suppress weeds. But the same early advantage can draw down stored water before reproductive stages that contribute most directly to yield.
The practical question is therefore not simply whether a variety uses more or less water. It is whether the timing and rooting pattern fit the season, soil profile and sowing density. Aggressive early growth may be rewarded on a deep profile with reliable late rain, yet become a liability in a stored-water system facing a dry finish.
The university release did not provide variety rankings, numeric yield gains or evidence that RootBot had already produced a drought-tolerant cultivar. The reported result is a phenotyping capability and evidence of cultivar differences. Calling it a proven yield technology would go beyond the available data.
Why genetics and management must be tested together
Breeders often screen yield and canopy traits while agronomists alter sowing rate, planting date and nutrition. Water-use timing connects those decisions. A genotype that disappoints at one density may fit another; a line with deeper late extraction helps only where the soil holds accessible water at depth.
That creates a genotype-by-management tool. Instead of labelling one variety universally drought tolerant, researchers can ask which variety, at which density, on which soil and rainfall pattern leaves enough water for yield-critical stages.
Similar trials have been conducted with Agriculture Victoria, Victoria's Department of Energy, Environment and Climate Action, and Western Australia's Department of Primary Industries and Regional Development. Those sites are essential because UQ expects water-use traits to vary with soil profile and climate. Multi-site work will show whether signatures are stable or location-specific.
Beyond wheat
The team has deployed the technology in sorghum, chickpea and barley and plans to expand into canola and maize. It also aims to evaluate nitrogen-use efficiency by matching active root distribution with soil water and nitrogen availability.
That extension is agronomically logical: nutrient movement and uptake depend on moisture. Measuring both could help separate a crop that merely looks vigorous from one that uses scarce water and fertilizer at the time and depth most likely to support harvestable yield.
What would prove the platform's value
The engineering challenge is simplification. Research robots can collect dense measurements, but breeders and agronomists need a reliable workflow across thousands of plots. Sensor calibration, soil-specific interpretation, repeatability and cost per plot will decide whether the method scales.
The scientific test is sharper: do these water-use traits predict yield stability across seasons and locations better than simpler canopy or soil measurements? The strongest next disclosure would pair RootBot profiles with variety-level yield, phenology, root depth and soil-water balance, then test those relationships outside Gatton.
Sources and method
Primary source: University of Queensland, “Unpacking wheat’s water use for drought research” (3 November 2026).
Agrilinkage reviewed the dated university announcement and separated reported observations from proposed uses. No peer-reviewed paper or variety-level yield table was linked, so this article does not infer yield improvements or name a winning cultivar.
Comments