Root traits offer a largely untapped resource for researchers and breeders aiming to improve yield in wheat due to their role as a primary interface for water and nutrient acquisition. Recent studies have shown that breeding for root traits is increasingly feasible with the development of high-throughput phenotyping platforms and molecular markers that can be used by researchers and breeders (Munns et al. 2012; Alahmad et al. 2020; Wasson et al. 2020; Makhoul et al. 2020; Ober et al. 2021; Rambla et al. 2022). Phenotyping methods are essential for exploring and discovering beneficial root traits that could be incorporated into future cultivars, helping to identify marker-trait associations, and candidate genes useful for marker-assisted selection (Gregory et al. 2009; Kuijken et al. 2015).
Recent reviews highlight a variety of root phenotyping methods that are available (Kuijken et al. 2015; Paez-Garcia et al. 2015; Topp et al. 2016; Atkinson et al. 2019; Tracy et al. 2020). However, the study of crop root biomass and allocation at a population scale in the field is still obstructed by the challenges related to root phenotyping (Wasson et al. 2012; Joshi et al. 2017; Martins et al. 2019). Controlled environment phenotyping, both destructive and non-destructive, is an alternative to field-based methods. The ability to study root distribution non-destructively is crucial to investigate factors that influence root system development. This is because non-invasive phenotyping allows traits to be measured over time, allowing association between early to late root development and association between root and shoot development. Also, plants with desired root traits can be grown to maturity for generation advance or crossing in breeding programs.
Rhizoboxes, thin boxes with a transparent side to visualise part of the root system, provide a non-invasive root system phenotyping method. The original ‘root box’ was created by Julius von Sachs in the 19th century (Sachs 1865). This work inspired the development of several rhizobox systems available today, perhaps re-pioneered by Dinkelaker (1989). The basic rhizobox system is made from PVC boxes equipped with transparent root observation windows with a glass side covered with a black PVC sheet to avoid any exposure to light (Liao et al. 2006; Palta et al. 2007; Neumann et al. 2009; Benlloch-Gonzalez et al. 2014). Depending on the plant species and the developmental stage studied, different rhizobox sizes are available, containing soil amounts ranging from one hundred grams up to 80–100 kg (Neumann et al. 2009). Currently, rhizoboxes are widely used for root research, including in high-throughput automated platforms (Nagel et al. 2012; Chen et al. 2015; Krzyzaniak et al. 2021).
These systems allow researchers to test plant growth behaviour in soil with a variety of conditions, including with soil moisture probe entry ports, waterlogging or low moisture stress, or with contrasting nutrient availability conditions and soil temperature (Dresbøll et al. 2013; Jin et al. 2015; Avramova et al. 2016; Durand et al. 2016; Schmidt et al. 2018; Zhang et al. 2019). Furthermore, the rhizoboxes enable the quantification of root system dynamics and kinetics, providing a way to study root traits through time and in a non-destructive way (Nagel et al. 2012; Chochois et al. 2015; Adu et al. 2017; Wu et al. 2018; Arsova et al. 2020). While these modern rhizobox methods represent the current state-of-the-art platforms to phenotype roots in controlled environments, they can be expensive due to the cost of customised components, assembly, and maintenance.
Coupling the rhizobox with image-based approaches using scanners, cameras and microscopes, provides a set of easily extractable quantitative traits (e.g., specific root length, root tissue density, average root diameter, branching intensity, etc; Freschet et al. 2021). Two-dimensional root scanning/imaging is currently the most widely used technique. WinRhizo (Regent Instrument Inc., Ville de Québec, QC Canada) is still the most popular software package to analyse root images (Delory et al. 2021). However, software such as RootPainter (Smith et al. 2020; Han et al. 2021) and RhizoVision Explorer (Seethepalli et al. 2021) are becoming increasingly popular. Their free and open-source, high-performance and complementary software tools support efficient and effective image analysis to support root research (Delory et al. 2021). In addition, RhizoVision Explorer offers the opportunity to estimate root traits from images acquired from a flatbed scanner or camera (Seethepalli and York 2020). RhizoVsion Explorer also allows for semi-automated image processing whereas the majority of image analysis requires labour-intensive manual inputs from the user, such as selecting points or tracing lines on the root.
Besides all the tools illustrated above, novel technologies are still required to characterize the complexity of root system architecture with more affordable, high-quality, non-destructive, and time-efficient methods. Low-cost, customizable methods to characterize root growth in soil over time have been developed (Chen et al. 2011; Joshi et al. 2017; Schmidt et al. 2018; Nguyen and Stangoulis 2019). However, these methods have limitations, mostly related to the number of rhizoboxes that can be screened, typically up to 20 soil-filled rhizoboxes (Liao et al. 2006; Palta et al. 2007; Neumann et al. 2009; Benlloch-Gonzalez et al. 2014). Nevertheless, for more comprehensive analysis and screening, there are specialised platforms that require growing plants in specific laboratories internationally. Furthermore, this could be expensive for preliminary screening studies of the root system. The GROWSCREEN-Rhizo (Nagel et al. 2012) is capable of automatically imaging roots and shoots of plants grown in soil-filled rhizoboxes (up to a volume of ~ 18 L) at a throughput of 60 rhizoboxes per hour. This platform includes 72 rhizoboxes, but many researchers cannot afford to establish this platform due to the cost associated with materials and assembly. One option could be outsourcing phenotyping using the GROWSCREEN-Rhizo, however, this is usually expensive and time consuming for preliminary screening studies of the root system, thus, would be more advantageous to screen using an affordable phenotyping method that could be conducted locally.
The present study had the goals to develop a rhizobox phenotyping method that (1) can be easily and readily built by researchers locally for use in laboratories and greenhouses; (2) supports extraction of root traits through image analysis using a free and open-source software; (3) integrates phenotyping of shoot traits such as, above-ground biomass, tiller number and root shoot ratio; and (4) enables the differentiation of genotypic effects for root and shoot traits. These efforts are intended to support root research and crop breeding programs that want to implement straightforward in-house root phenotyping capabilities.