Waterlogging (WL) is one of the most damaging abiotic stresses, affecting 1,700 million hectares of land surface annually (Kaur et al., 2020). Under WL, saturation of soil pores with excessive water results in the development of anaerobic conditions with a subsequent reduction in root growth ( Figure 1A ; Pais et al., 2022). WL induces nutrient imbalances in soil by inducing chemical reduction of some nutrients, including nitrogen (N) (Steffens et al., 2005), thus leading to both nutrient deficiency and/or toxic buildups in soil. N is a very important mineral nutrient and plays a critical role in plant physiology; thus, nitrogen fertilization is adopted as one of the most essential principles for efficient crop production systems (Shah et al., 2021). Nitrogen application boosts crop yield (Shah et al., 2017; Shah et al., 2022); however, excessive application of N comes with several environmental issues. WL promotes soil N losses via runoff, leaching, and denitrification with a concomitant reduction in crop productivity, thus imposing economic and environmental implications. Thus, it is important to understand and improve nitrogen use efficiency (NUE) in plants under WL.

Roots uptake N from the soil in various forms, including amino acids, nitrate (NO₃ ⁻), and ammonium (NH₄ ⁺); however, NO₃ ⁻ is a major source of N in plants (Arduini et al., 2019). WL reduces root N uptake by altering root development (RD), root system architecture (RSA), and N availability in soil. The adoption of advanced agronomic N management techniques, including slow-release fertilizer, biochar application, or inoculation, plays a significant role in improving NUE under WL; however, the efficacy of any agronomic technique greatly depends on soil type and plant species. Moreover, recent development in genetics and breeding techniques have also shown tremendous potential in the development of crop cultivars with higher NUE under low N availability; however, the development of such cultivars is very complex due to genotype and environmental interactions. Moreover, a bottleneck has arisen in the collection of quality phenotypic data to advance crop breeding programs compared with genetic analysis. In this context, adoption of high-throughput root-phenotyping (HTRP) can provide blueprints for breeders to enhance N acquisition in roots under WL.

Several HTRP techniques enable us to phenotype and visualize the root performance under different growth conditions ( Figure 1B ); however, contemporary aboveground canopy-based crop phenotyping (GCCP) techniques account for N-deficiency-induced changes in vegetation index (VI) by measuring photosynthesis, chlorophyll contents, leaf temperature, and stay greenness. However, such GCCP data can be easily camouflaged by the multiple environmental factors that can directly or indirectly influence VI traits. Contrarily, roots being the first line of contact with N and WL, focusing on the establishment of HTRP at least at the early growth stage would be beneficial in determining the genetic basis of NUE in plants under WL.

Root system (RS) is very important in the context of N acquisition from soil, and several root traits such as root size, root length (RL), root density (Rd), and root distribution determine N acquisition from soil ( Figure 1C ; Garnett et al., 2009). Crop cultivars with larger RL and Rd uptake more N from soil (Ju et al., 2015), thus reducing N losses under WL. RSA is closely related to N uptake, and crop plants with steeper roots uptake more N from the soil (Zhan & Lynch, 2015). The duration of WL also influences the RS and N uptake (Malik et al., 2001); e.g., short-term WL reduced N uptake only in the bottom layer of the soil-filled pot, while long-term WL resulted in reduced N uptake in both the bottom and top layers of the pot (Dresbøll and Thorup-Kristensen, 2012). Root N uptake was more quickly recovered after short exposure to WL than after long exposure to WL, probably due to the production of new roots (Dresbøll and Thorup-Kristensen, 2012). However, even though N uptake was resumed after recovery from WL, oat roots exhibited reduced root biomass under WL, most likely due to the separation of dead root fragments (Brisson et al., 2002), advanced growth stage during recovery (Arduini et al., 2019), continuous N leakage from root tissues, or other detrimental effects of WL on RS (De San Celedonio et al., 2017). Nonetheless, a cultivar-specific relationship between RS and NUE was observed among two Chinese and one American variety of maize (Ju et al., 2015). The insufficient N uptake by roots under WL could also be due to low availability of N in soil (Nguyen et al., 2018), higher N losses, reduced RD (Brisson et al., 2002), and impaired NO₃ ⁻ uptake by roots (Pang et al., 2007). Thus, it is not practically easy to ascertain whether lower N availability to roots is the primary cause of reduced root growth under WL or vice versa. Therefore, it is important to consider factors such as cultivars, WL duration, WL method, and plant growth stage when performing HTRP.

Labeling the variations among genotypes and species that uphold improved root traits and integrating them into breeding programs for the development of N-efficient cultivars is a very demanding method. However, studying RSA is very challenging due to the complexity of accurately and precisely phenotyping RS under WL. Several HTRP techniques are being used to understand the relationship between RS and NUE under WL ( Figure 1B ); however, under field conditions, root phenotyping is still handled using a medium- to low-throughput platform (Araus et al., 2022). Different WL methods can also influence root phenotyping ( Figure 1D ). Under controlled conditions, growing plants in hydroponics or gel-based media provides an easy approach to monitoring root morphology; however, this is applicable only in early growth conditions (Langan et al., 2022). Sand culture is another HTRP technique to study RS for improved NUE and performance of root traits using scanners (Paez-Garcia et al., 2015). The pH level of soil and soilless cultures needs to be well monitored, as systems with pH instability and low buffer capacity affect N uptake and RD (Lager et al., 2010). Noninvasive measurements of RS for improved NUE under WL can also be examined using image technology, enabling 2D root growth accompanied by real-time gene expression relating to NUE in roots (Rellán-Álvarez et al., 2015). Other noninvasive techniques, including magnetic resonance imaging (MRI) and X-ray computed tomography (CT) (see glossary in Figure 1B ), assist in visualizing the physiological properties of roots (Mairhofer et al., 2013; Metzner et al., 2015). Nonetheless, technical complexities and high operation costs make these techniques less useful for large-scale phenotyping. The noninvasive microelectrode ion flux measurements (MIFE) technique was used to perform cell-based phenotyping for revealing QTL associated with hypoxia tolerance in barley (Gill et al., 2017) and understanding the N uptake by measuring the kinetics of NO₃ ⁻ and NH₄ ⁺ fluxes (Garnett et al., 2003).

At field conditions, several techniques have been applied for performing root phenotyping, such as shovelomics and soil coring (SC). Shovelomics also known as root crown phenotyping, consists of the manual digging and excavation of roots and up to 30 cm of rhizosphere (Wasson et al., 2020). SC also works as shovel omics does to some extent; however, SC consists of the extraction of cores from deeper soil using a corer, with a betting examination of RS (Wasson et al., 2014). For a better view of RS, SC is supplemented with a portable fluorescence imaging system known as BlueBox, which provides automatic root counting using image analysis software (Wasson et al., 2016). Geophysical platforms such as electrical resistance tomography and electromagnetic inductance are used to infer root growth under changes in soil water (Srayeddin & Doussan, 2009; Whalley et al., 2017). Moreover, ground-penetrating radar performs mapping of subsurface soil using radio wave pulses and detects RS under field conditions (Liu et al., 2017; Atkinson et al., 2019).

These HTRP techniques can be ineffective, laborious, and subject to soil conditions (soil types, WL duration, N in soil). Moreover, root extraction under WL is also very difficult due to the breakage of root fragments during extraction; thus, alternate approaches supplement HTRP, including phenotyping of aboveground traits. However, measuring aboveground traits can only infer root growth indirectly (Reynolds et al., 2012; Tracy et al., 2020). To understand root response, examination of the stable isotope composition of N in roots under WL can improve our understanding of the physiological basis of roots and NUE under WL. Having said that, isotopic signatures of oxygen in stem water were used as an indicator of water status in water-stressed roots (Kale Celik et al., 2018). Thus, this approach should be used along with other HTRP techniques.

Performing HTRP using imaging sensors (IS) and platforms goes on to grow exponentially, easing the bottleneck of root phenotypic data collection (Roitsch et al., 2019). IS such as red, green, and blue (RGB) sensors that take images within the wavelength range of 400–700 nm are termed visible IS, while IS that go beyond the visible wavelength are known as spectral IS (SIS) (Beisel et al., 2018; Bruning et al., 2020). In controlled conditions such as glasshouses or growth chambers, IS range from low-cost cameras to costly custom-made imaging setups (Tovar et al., 2018). Recently, Xia et al. (2019) used hyperspectral and RGB to phenotype WL in rape plants and found promising results. Nonetheless, the use of low-cost cameras may result in image noise; thus, to reduce image noise, image fragmentation must performed (Agata et al., 2007). Imaging plants under WL face other challenges due to the presence of extra water in a pot, which reflects the lights of IS and is due to unwanted algal growth. On the other hand, in field conditions, the use of unmanned ariel vehicles (UAV) and satellite-based imaging are the most popular imaging techniques (Li et al., 2014; Langan et al., 2022). Nonetheless, these imaging techniques also face challenges associated with soil heterogeneity and water drainage, so the use of machine learning (ML) has been suggested along with these imaging techniques to study WL in plants (Zhou et al., 2021). For 2D root images, tip locations have been identified using a deep network-based classifier scanned over an image to produce a location map (Pound et al., 2017). For 3D images, deep learning has been applied to the root–soil segmentation problem, where deep-learned features are used to drive a support vector machine classifying root/soil pixels (Douarre et al., 2016).

As mentioned before, RS plays a very important role in N uptake under WL, and using growth pouches to study root performance under WL or performing root phenotyping using the classical 2D imaging technique (Nagel et al., 2012) does not provide a clear understanding of the root development under WL. Thus, the use of tomographic techniques including CT scanning, MRI, or positron emission tomography has been successfully reported in the study of root phenotyping (Atkinson et al., 2019; Wasson et al., 2020). For instance, X-ray CT scanning was used to visualize the formation of aerenchyma under WL in the roots of barley (Kehoe et al., 2022). Therefore, the application of tomographic techniques can assist in root phenotyping under WL, thereby opening new opportunities for future studies. Though several other methods have been designed for root phenotyping by studying different root traits, including root surface area, crown roots, root length, and root density in soil core (Koyama et al., 2021), there is not any standard root phenotyping method to study different aspects of RSA under WL; therefore, the field of IS exhibits much to extend to the research community. Having said that, several image analysis software are available to quantify and visualize root systems ( Figure 1E ). A new initiative has been established to attempt to harness crop-management synergies using phenotyping, robotics, and computational technologies (http://www.phenorob.de/).

N fertilization has become the necessity of almost every intensive cropping system, and under WL conditions, crops face N deficiency. Thus, it is imperative to improve the ability of crops to improve NUE under limited N availability. Roots play a critical role in acquiring N from soil; thus, it is important to phenotype RS to highlight the root traits and their relationship with NUE under WL. Given that, the application of HTRP is intensifying due to the technical development and measurement of RS. The utilization of IS and noninvasive measurements of RS can facilitate improving NUE in roots under WL. Advances in ML further benefit analyzing root phenotyping data; however, under field conditions, high‐throughput analysis of root phenotyping remains subtle.

LH: Conceptualization, Writing – review & editing. YZ: Writing – original draft. JG: Writing – original draft. QP: Writing – original draft. ZZ: Writing – original draft. XD: Writing – review & editing. MT: Conceptualization, Writing – review & editing. YG: Conceptualization, Methodology, Writing – review & editing.