Modelling Plant Variety Dependent Least Limiting Water Range (LLWR)

Drought stress is a major limiting factor for yield on a global scale (Solh and van Ginkel, 2014), with drought effects being predicted to become more severe with increasing global temperatures (IPCC, 2014). Climate change is also expected to increase the frequency and severity of floods leading to root oxygen stress (Trenberth, 2011). At the same time, current agricultural practises are increasingly relying on heavy machinery leading to soil compaction and changes in soil structure (Chamen et al., 2003), reducing the rate of cell division in the root meristem, and decreasing cell expansion (Bengough and Mullins, 1990). As such, in order to reduce yield losses it is essential to understand the complex interaction between oxygen stress, water stress and mechanical stress (Mohammadi et al., 2010). The least limiting water range (LLWR) is one such model which relates the above-mentioned soil stressors in order to estimate the soil moisture range in a particular soil for which plants should be less limited in terms of growth. However, the extent to which the LLWR considers the influence of root traits in changing its boundaries is currently limited. In order to be able to assess the effects of root trait variability on the LLWR boundaries while manipulating the LLWR soil stressors a minirhizotron based system (RS) was developed. This cheap (~£10 per unit), acrylic based, A3 sized system enabled in situ imaging of roots and root hairs. Destructive sampling methods were also used to determine root border cell numbers and root tip geometry. To further optimise the process of data collection, Rcpp based image processing algorithms were developed to obtain automated estimates of the root traits of root length, root hair, root border cells and root tip eccentricity to further increase the efficiency of the RS phenotyping platform. To test how contrasting root traits influence the LLWR a plant phenotyping experiment was performed comparing four spring barley (Hordeum vulgare L.) varieties, Optic, KWS Sassy, Derkado and Golden Promise. Root growth rates both in the vertical and horizontal directions all increased with increasing water availability and decreasing substrate density. Root hair area did not vary significantly among treatments and between variaties. Root border cell count and root tip eccentricity increased with increasing substrate density but did not vary significantly across varieties. A root micro-trait based linear interaction model was developed to describe average root growth rates and it was demonstrated that root growth rates on average follow a linear patern for values >= 8 mm day-1. Root micro-traits mostly failed to correlate well with root growth rates except for a negative assosiation with root tip geometry (cor = -0.4192, p = 2e-05**)