Why do we need this project? 

​​Even though genetic variability in root architecture (i.e., root structure) is known to exist in sorghum (Borrell et al., 2014; Mace et al., 2012; Singh et al., 2012), and wheat (Alahmad et al., 2019; Hassouni et al., 2018; Richard et al., 2015), rarely these traits have been functionally related to differences in crop water use, yield, or yield stability in the field.

 

Lack of success can be attributed to multiple factors including:

• the presence of interactions and seasonal stress dynamics between genotypes (G) and environment (E) on root structural traits limit the value of glasshouse, tubes, root chamber, pots, and lysimeter studies.

• that root structural traits determined early in the season on plants in vegetative stages tend to show poor correlation with the same traits determined later in the season.

• that there has been a focus on characterising root structural traits rather than to understand how the rooting system function and respond to environmental and management cues i.e., functional traits.

• that the phenotyping of structural root traits is expensive, only feasible on small number of treatments, and subject to large errors when sampled in the field; and

• that the predominant focus of approaches has been limited to characterise the average value of a trait, which overlooks the fact that the rooting system is highly plastic, and that different genotypes show different degree of such as plasticity under stress.

​

Advances in functional high-throughput phenotyping (HTP) of root traits in the field

In contrast to structural phenotyping approaches, progress was recently made in the development of functional HTP of root traits in the field, i.e., assessing function rather than form. These new tools have shown to be cheap, quick, highly accurate (Zhao et al., 2022), and functionally related to changes. In our approach we combine time-lapsed electromagnetic induction (EMI) surveys, drone imagery, crop-ecophysiology principles, and machine learning techniques (see Zhao et al., 2022), to build a 3D representations of root growth and root activity in the soil profile. Here we propose that  useful characterisations of the rooting system need to be made understanding root function in the field, instead of  producing complex and expensive visualisations of the rooting system growing in pots or using MRI or XRay facilities. Our methods were developed by the project Optimising Sorghum Agronomy in the Northern Grains Region (UOQ 1808-001RTX), applied to map plant available water capacity in farmers fields (UOQ2304-006RTX), and published by Zhao et al., (2022; 2023a; 2023b). The approach is now also being applied in the ARC EC Industry project Drought tolerance in sorghum: the root of the solution in partnership with GenTech (Pioneer Seeds) and AirbornInsight.

 

​Briefly, we conduct time-lapse electromagnetic induction (EMI) surveys to quantify the spatiotemporal variation of the soil volumetric water content (v, cm3 cm-3) from each 10 cm of soil and down up to 3m (depending on site conditions), at key times in crop cycle. A DUALEM-21S (Dualem Inc., Milton, ON, Canada) is used to collect soil apparent electrical conductivity (ECa), which is a function of soil moisture content. The instrument is dragged 3m to the right of a four-wheel all-terrain vehicle that traversed the field along the transect in the middle of each plot. The vehicle travels at a speed of ~4 km h-1 and collects two ECa readings per second producing a highly detailed characterisation of layered soil water. Assuming 4m long plots and no stops, the approach has a work capacity of 1,000 plots h-1, at the cost of running an ATV at low speed, and calibrating the sensor for soil type from gravimetric determinations of soil moisture.

​​

​After calibration for soil type, the surveyed values of layered water use and canopy size (derived from drone imagery) are used to calculate a root activity factor for each soil layer, down to the maximum rooting depth. The maximum rooting depth is estimated as the deepest depth that shows changes in soil moisture between the two consecutive surveys i.e., root activity. The indices derived from the surveys have been validated for the presence of roots using (soil coring). Our results show that our indices are closely related to measured root length density in the field (Zhao et al., 2023a) across a wide range of environmental conditions, and to differences between hybrids in yield, yield components and their stability (Zhao et al., 2023a and b).

​