Water and nutrient availability, along with the efficiency of their use by crops, are major determinants of agricultural productivity (Fageria et al., 2008; Sinclair & Rufty, 2012). However, unchecked demand for natural resources, such as land, fertilizer, and water, to support global agricultural production has led to widespread ecosystem degradation and biodiversity loss (Foley et al., 2005; Williams et al., 2020). Thus, the demand for stable agricultural yields amid globally changing climate and decreasing natural resource availability necessitates more resource‐efficient and climate‐resilient crops and agronomic practices.

Plant growth is neither linear over time (Paine et al., 2012) nor uniform across genotypes within a species (Mortlock & Hammer, 2000; Sugiyama et al., 2013). Measurements taken at single time points thus fail to capture the dynamic variation of plant growth across developmental stages and environmental conditions (Granier & Vile, 2014). In addition to being generally higher throughput than destructive measurements, noninvasive phenotyping approaches provide a window into temporal dynamics along a developmental trajectory and in response to changing environmental conditions. Immense progress has been made in high‐throughput phenotyping to capture plant traits noninvasively, particularly through the use of advanced imaging platforms (Li et al., 2020; Shakoor et al., 2017). For example, by noninvasively tracking growth over time in various Setaria spp., in which visual leaf area strongly correlates with above‐ground biomass, Fahlgren and colleagues uncovered species‐specific growth dynamics across time and environmental conditions that would have otherwise been missed by endpoint measurements alone (Fahlgren et al., 2015).

In contrast to phenotyping approaches that capture data from only certain plant parts, whole‐plant phenotyping can identify phenotypes and growth conditions that are optimal for both above‐ and below‐ground tissues and is thus likely more valuable for breeding more resilient and resource‐efficient crops (Chochois et al., 2015). Because of the relative difficulty and time required to measure traits at the whole‐plant level (Li et al., 2020), measurements are often simplified to include only above‐ground plant tissue. However, approaches that exclude root biomass are potentially limited in value. For example, identifying genetic variation in traits such as biomass accumulation and water use efficiency (WUE) is particularly important in the context of water‐limited conditions in which the ratio of root:shoot biomass is often substantially different than under water‐replete conditions (Benjamin et al., 2014; Hubick et al., 1986; Xu et al., 2015). Even in well‐watered conditions, the ratio of root:shoot biomass partitioning can be different among genotypes (Chenu et al., 2018; Chochois et al., 2015), further highlighting the importance of including below‐ground biomass to accurately quantify resource use efficiency (Leakey et al., 2019).

Various approaches can phenotype below‐ground traits noninvasively. For example, root growth can be measured in vivo using magnetic resonance imaging (MRI) (Rascher et al., 2011), X‐ray computed tomography (CT) scanning (Zeng et al., 2021), or with luminescence‐based reporters (Rellán‐Álvarez et al., 2015). When combined with non‐destructive techniques to measure above‐ground tissues, such approaches could in theory measure whole‐plant growth over time. However, these approaches often require large upfront costs, state‐of‐the‐art technologies, or specific technical expertise (Großkinsky et al., 2015). There are reports of cheaper and simpler approaches to measure whole‐plant traits non‐destructively (Fletcher et al., 2018). However, the authors found that data obtained non‐destructively did not correlate well with values from plants harvested for end‐point measurements.

To address the various limitations of measuring whole‐plant traits non‐destructively, we developed a low‐cost and versatile approach that can noninvasively measure whole‐plant physiology over time. As a proof of concept, we measured biomass accumulation and water use over time to directly quantify whole‐plant WUE. Plant WUE, broadly defined as the ratio of biomass accumulation to water use via transpiration (Vadez et al., 2014), represents an important metric to optimize for improving plant resilience and agricultural sustainability. Various methods exist to quantify WUE, ranging from instantaneous measurements at the leaf scale (Vadez et al., 2014) to long‐term quantification at the canopy scale (Lambers & Oliveira, 2019). Because of the difficulty in measuring whole‐plant WUE, it is common to extrapolate these data from leaf‐level measurements or via proxies. For example, quantifying leaf gas exchange rates can provide a measure of both photosynthetic and transpiration rates. However, instantaneous leaf‐level WUE data are often in disagreement with whole‐plant measurements at longer timescales (Medrano et al., 2015). Leaf carbon isotope discrimination (CID) is often used to estimate whole‐plant WUE as CID is strongly correlated with WUE in C₃ species (Farquhar et al., 1982). However, the strength of this correlation is greatly reduced in C₄ species, including important crops such as maize and sorghum (Henderson et al., 1998; Vadez et al., 2014). Moreover, quantifying CID is relatively slow and expensive (Vadez et al., 2014), thus necessitating simpler and more widely applicable methods.

In contrast to these methods, our approach, which uses hydroponic cultivation, allows for the direct and non‐destructive measurement of whole‐plant WUE through direct quantification of biomass accumulation and water use. We show that hydroponic cultivation results in WUE equal to that of soil‐grown plants and then demonstrate the temporal and contextual versatility of our approach by measuring whole‐plant WUE every two days on unstressed and osmotically stressed seedlings. In doing so, we identified relationships between root zone acidification and aspects of photosynthesis on whole‐plant growth and WUE over time. Our approach can be implemented using cheap, basic components and requires no specific technical expertise. It can be implemented for any non‐aquatic, vascular plant species and can therefore contribute broadly to identifying and engineering superior crop germplasm.

We developed a simple, cheap, and modular system to grow and phenotype whole‐plant traits non‐destructively over time. A detailed, step‐by‐step protocol can be found online (protocols.io; dx.doi.org/10.17504/protocols.io.q26g7p6m1gwz/v1). Briefly, a hole is drilled into screw‐on caps and a cut‐off pipette tip of approximately equal diameter is inserted into the hole of the screw‐on cap (Figure 1a). Soil is then inserted into the snugly‐fit pipette tips, and seeds are sown into the soil‐filled tips. Plants are initially grown in “open” hydroponic conditions such that all samples are grown in a single reservoir whose water is exposed to the air (Figure 1b). Upon reaching a desired size, age, or developmental stage, tops are screwed onto tubes and the system becomes “closed” such that water loss from the tube is restricted to uptake by the plant (Figure 1c). By unscrewing the tops, the entire plant can be removed, weighed, and then screwed back onto the tube for continued growth (Figure 1c–d). When grown this way, whole‐plant biomass accumulation and water use can be assayed repeatedly over time (Figure 2a–d). Moreover, the modularity of our approach allows for precise control of the root environment to investigate a variety of phenotypes, including the relationship of whole‐plant biomass accumulation and water use in response to changing abiotic or biotic conditions.

As a proof‐of‐concept of this approach, we sought to compare genotype‐specific variation in biomass accumulation, water use, and whole‐plant WUE of various sorghum accessions when grown either in soil or using our closed‐system hydroponic approach. Sorghum (Sorghum bicolor) is a highly productive and relatively stress‐resistant C₄ crop (Ananda et al., 2020; Khalifa & Eltahir, 2023). It is among the world's most important cereal crops and is utilized for both food and bioenergy (Ananda et al., 2020; Khalifa & Eltahir, 2023). Genotype‐specific variation exists in whole‐plant WUE in soil‐grown sorghum (Balota et al., 2008; Xin et al., 2008). We selected five sorghum genotypes (Table 1) to represent a diverse range of geographic origins, taxonomic races, variation in transpiration rates measured at the leaf level (Balota et al., 2008), and variation in WUE measured at the whole‐plant level (Balota et al., 2008; Xin et al., 2008).

We first asked whether our method of quantifying WUE is consistent with established methods (Cernusak et al., 2007; Xin et al., 2009) by comparing the WUE of plants grown in our hydroponic system to plants grown in soil. Because most traditional soil‐grown approaches to directly measure WUE are destructive, we compared end‐point biomass accumulation and water use over the course of 8 days after reaching the third leaf stage in either hydroponic or soil‐growth conditions. All genotypes accumulated more biomass and consumed more water when grown in soil compared with when grown hydroponically (Figure S2A–B). However, WUE after 8 days was indistinguishable between soil and hydroponic conditions for all genotypes (Figure S2C).

Next, we asked whether evaporative loss of water over time was negligible compared to plant water use, which is required to accurately infer plant water use from changes in tube weight over time. Water loss over time due to evaporation was measured in tubes from which whole plants were removed by cutting off roots and shoots to eliminate transpiration. Over the course of 7 days, plant‐less samples lost an average of 0.6 mL of water because of evaporation (Figure S3), equivalent to approximately 0.7% of average water use in plants grown in control conditions over the same period. Evaporative water loss in our hydroponic system was therefore determined to be negligible compared with transpirational water uptake.

Having validated our system, we next sought to compare the effect of osmotic stress on cumulative biomass accumulation, water use, and WUE across five sorghum accessions (Table 1). Specifically, we asked how these traits vary over time for each accession and whether genotype‐specific differences in plant physiology manifest in unstressed or osmotically stressed conditions. Plants were grown in tubes containing either nutrient solution alone or nutrient solution with 10 mM mannitol to induce osmotic stress. Osmotic stress resulted in decreased biomass accumulation, water use, and WUE in all genotypes, compared with control conditions (Figure 2). There were no genotype‐specific differences in the rates of biomass accumulation, water use, or WUE under control conditions (Figure 2a,c,e). However, genotype‐specific differences were observed in both the time and extent to which mannitol stress impacted growth and WUE. For example, Grassl and Tx7078 accumulated biomass at a greater rate than the other genotypes when grown in mannitol (Figure 2a). Additionally, change in biomass accumulation in mannitol conditions relative to control was observed slightly earlier in BTx623 compared to all other genotypes (Figure 2b). Despite this slight temporal difference, the overall rate of biomass accumulation in BTx623 under osmotic stress was similar to that of osmotically‐stressed RTx430 and LTA (Figure 2a). Mannitol stress resulted in the greatest reduction in water use rates in RTx430 and the smallest decrease in Tx7078 (Figure 2c). However, there were no genotype‐specific temporal differences in average water use rates under osmotic stress compared to controls (Figure 2d). With regard to WUE rates, stress‐induced reductions were most pronounced in LTA (Figure 2e). This greater reduction in WUE was independent from the time at which WUE of osmotically stressed plants dropped below control levels. For example, WUE of RTx430 and BTx623 plants grown in mannitol was decreased compared to control by Day 4, whereas osmotic stress did not reduce Tx7078 WUE until Day 8, later than all other genotypes (Figure 2f). Collectively, these results suggest that the onset of osmotic stress does not always mirror its magnitude on plant physiology over time. Additionally, these results indicate that Grassl and Tx7078 are more tolerant to osmotic stress than RTx430 and that LTA is less efficient in using water compared with other accessions when osmotically stressed at the third leaf stage.

To better understand how osmotic stress affects the relationship between biomass accumulation and water use, and to identify which underlying component is a stronger predictor of changes in WUE, we regressed biomass accumulation by water use (Figure S4A), WUE by biomass accumulation (Figure S4B), and WUE by water use (Figure S4C). There was a strong positive correlation between biomass accumulation and water use for all genotypes in both control and mannitol‐treated samples, though this relationship was stronger in control conditions (Figure S4A). A positive correlation was also found between WUE and biomass accumulation for all genotypes under both control and osmotically stressed conditions. However, the correlation was stronger for all genotypes when exposed to osmotic stress (Figure S4B). Conversely, WUE could only be explained in part by water use in certain genotypes and under certain conditions (Figure S4C). In LTA, for example, WUE increased linearly with water use only under osmotic stress, whereas there was no relationship between WUE and water use in either condition for Tx7078 (Figure S4C). For all genotypes in which a positive correlation was found between both biomass accumulation and water use with WUE, biomass accumulation was the stronger predictor for increases in WUE (Figure S4B–C). These results suggest that mannitol‐induced osmotic stress reduces WUE and that in both stressed and unstressed conditions, biomass accumulation is a stronger determinant of WUE than water use.

Having observed genotype‐specific differences in biomass accumulation, water use, and WUE rates specifically in osmotically‐stressed conditions, we next sought to identify possible mechanisms for this context‐specific difference in plant growth and physiology. Osmotic stress can impair photosynthetic performance (Chen et al., 2011; Grzesiak et al., 2006). We therefore asked whether genotype‐specific differences in biomass accumulation, water use, and WUE under mannitol stress could be explained by differences in photosynthetic efficiency. Surprisingly, Photosystem II (PSII) maximum quantum efficiency (F _V /F _M ) of the third true leaf decreased to the greatest extent in both Tx7078 and RTx430 plants in response to osmotic stress (Figure 3a). Similar to the effects of osmotic stress on biomass accumulation and water use, the time at which F _V /F _M levels dropped compared to control varied by genotype (Figure 3b). Specifically, F _V /F _M in mannitol‐grown LTA plants was reduced relative to control levels by Day 4, which was earlier than all other genotypes (Figure 3b) and further demonstrates that the onset of stress does not always correlate with its magnitude. These results suggest that photosynthetic capacity Fv/Fm at the third leaf stage was not strongly associated with plant growth or WUE under osmotic stress, as Grassl and Tx7078 plants had equal rates of biomass accumulation and WUE under osmotic stress despite differences in the effect of osmotic stress on their PSII quantum efficiency.

We next asked if there were any differences between the accessions in their root physiology in response to osmotic stress. By exuding a diverse set of compounds from their roots, plants actively modify their root zone to facilitate nutrient uptake and in response to abiotic stress, including mannitol‐induced osmotic stress (Darko et al., 2019; Falhof et al., 2016; Marschner et al., 1986). One of the ways in which plants optimize their rhizosphere properties is via soil acidification (Ehrenfeld et al., 2005; Marschner et al., 1986). Growing plants in closed tubes makes it possible to directly and easily measure rhizosphere acidification. We therefore asked whether and when there were any genotype‐specific differences in pH change of the media every 2 days (delta‐pH). The pH of media with or without mannitol was measured before adding it to tubes, and after 2 days of plant growth (Figure 3c and Data S1). Any solution remaining in the tubes after every 2 days was then discarded and fresh nutrient solution was added to ensure sufficient water for plant growth for the subsequent 2 days. After 2 days of growth in closed tubes, there was no effect of osmotic stress on delta‐pH in any genotype ; both control and mannitol‐supplemented media were acidified by 1.0‐1.5 in all genotypes (Figure 3c). However, control plants stopped acidifying their media within 4 to 6 days, whereas the mannitol‐supplemented media was consistently acidified throughout the experiment by plants of all genotypes (one‐sample t‐test; (alpha = 0.05), Figure 3c, and Data S2). These results suggest that prolonged root zone acidification may be an adaptive response conserved among sorghum accessions in response to osmotic stress at the seedling stage.

We have demonstrated how non‐destructive, whole‐plant phenotyping can be employed using simple and readily available components. We showcase the versatility of our approach by directly measuring multiple traits across a diverse set of sorghum accessions in unstressed and osmotically stressed conditions. Collectively, we illustrate how this method can identify genotype‐specific temporal dynamics in plant growth across multiple organs and physiological processes.

Plant phenotypes are dynamic manifestations of genetics, environmental conditions, and developmental stage. Phenotypic plasticity is evidenced by changes in growth of both above‐ and below‐ground tissues. Non‐destructive, whole‐plant phenotyping facilitates quantification of dynamic growth processes that are more challenging to identify via destructive approaches and is therefore crucial for improving our understanding of plant growth and engineering more resilient and resource‐efficient crops. However, noninvasive, whole‐plant phenotyping approaches are often expensive and technically complex. To address these limitations, we developed a cost‐effective, simple, and modular approach that can measure whole‐plant physiology over time. Unlike soil‐grown conditions, hydroponic cultivation allows us to physically isolate roots and return them to their growth environment non‐destructively. In doing so, entire plant biomass can be measured directly and repeatedly. Additionally, if each plant is grown physically separated from other plants and evaporation is prevented, plant water use can also be measured directly. These capabilities alone are of practical value for the study of plant growth and development given that these traits are not readily quantifiable in a direct, non‐destructive manner for soil‐grown plants. Moreover, quantification of these traits over time allows for direct measurements of whole‐plant WUE over time. Previous studies have demonstrated creative approaches to quantifying WUE non‐destructively (Fahlgren et al., 2015; Fletcher et al., 2018). However, these methods either ignore root biomass contributions (Fahlgren et al., 2015) or do not correlate well with WUE values from harvested soil‐grown plants (Fletcher et al., 2018).

Comparing our hydroponic system to soil‐grown plants, we observed that plant growth and resource usage were strongly influenced by abiotic conditions. This was made clear by the greater biomass accumulation and water use of soil‐grown plants compared with those grown hydroponically. Interestingly, WUE was uniform across growth conditions. This suggests that biomass accumulation at this stage of development necessitates a proportionally equal amount of water use, regardless of absolute growth rates. What might explain slower growth rates in the hydroponic system compared with potted soil? The hydroponic growth conditions used in this study were chosen to be suitable for a diverse range of genotypes at a specific developmental stage, rather than being designed to optimize growth. However, irrigation scheduling, nutrient composition, and abiotic conditions for hydroponic growth such as media oxygenation could all be refined to optimize a given metric, as desired.

We chose to examine the five varieties of sorghum (Table 1) because, previously, genotype‐specific differences in whole‐plant WUE have been observed in soil‐grown sorghum (Mortlock & Hammer, 2000; Xin et al., 2008, 2009). We specifically focused on identifying genotype‐specific variation at the seedling stage as sorghum biomass accumulation at the time of harvest is most affected by drought‐induced osmotic stress when stress occurs at the seedling stage (Mastrorilli et al., 1999), similar to the heightened sensitivity of many crop species to stresses at early stages in their development, such as maize (Ge et al., 2012; Kang et al., 2000), rice (Zhao et al., 2014), and tomato (Sivakumar et al., 2020). Under non‐stressed conditions, LTA previously exhibited greater whole‐plant WUE than Tx7078 when grown in soil, but equal WUE to BTx623 and RTx430 (Xin et al., 2008; Xin et al., 2009). That we did not observe genotype‐specific differences in WUE in either soil‐ or hydroponically grown sorghum in control conditions could reflect the different ages and developmental stages of the plants being assayed. We measured plant growth starting at the third leaf stage, corresponding to plants that were approximately 2 weeks old, whereas plants were approximately 4–6 weeks old (corresponding to the eighth leaf stage) in Xin et al. (2008) and Xin et al. (2009). The differences in WUE we observed compared with previous reports further support the idea that WUE varies by developmental stage (Fahlgren et al., 2015). We did, however, observe genotype‐specific differences in WUE when plants were grown under osmotic stress. In light of the previously identified superiority of LTA in terms of its WUE when grown in soil under non‐stressed conditions, it was somewhat surprising that LTA was most sensitive to osmotic stress in terms of its WUE. Although WUE of many species, including sorghum (Xie & Su, 2012), increases in response to drought stress in soil‐grown conditions (Peters et al., 2018), mannitol‐induced osmotic stress decreases WUE in multiple species (Cha‐Um & Kirdmanee, 2008; Darko et al., 2019), similar to what we observed with sorghum in this study. Collectively, these findings highlight the importance of considering multiple developmental stages and abiotic conditions when studying plant resilience.

Because WUE is underpinned by both biomass accumulation and water use, much work has been done to identify which of these components should be optimized to improve crop WUE. Whether WUE is even a desirable agronomic trait is actually a contentious matter (Blum, 2009; Condon et al., 2004). This is because, in many crop species, increased WUE is usually achieved through reduced stomatal conductance, which in turn results in decreased carbon fixation and lower yields (Blum, 2009; Vadez et al., 2014). Other work, however, has demonstrated that higher WUE via reduced stomatal conductance does not necessitate a yield penalty if there is a concomitant increase in photosynthetic rates (Vadez et al., 2014). These contradictory findings suggest that the contributions of growth and water use to WUE are both complex and are underpinned by both genotypic and environmental factors. In all such cases, however, it is generally agreed that increasing, rather than decreasing soil water use, is a key breeding strategy for improving both crop yield and potentially WUE (Blum, 2009; Condon, 2020). This is because increased plant water use would support more carbon capture and reduce the fraction of soil water lost to evaporation. Moreover, increased root growth could facilitate water capture from deeper in the soil profile where water is less prone to evaporative losses (Condon, 2020). These findings suggest that breeders should aim to identify genotypes that maximize water use and biomass accumulation, as these traits should support both high yields and plant resilience during water stress (Blum, 2009; Condon, 2020). This idea is supported by the findings in this study, given that WUE positively correlated with both increased biomass accumulation and water use and that these relationships were stronger under osmotic stress. Moreover, biomass accumulation, rather than water use, was a stronger predictor of WUE in all genotypes, which has previously been observed in soil‐grown sorghum (Xin et al., 2009). These results further support the idea that breeders should aim to identify the fastest growing lines to both optimize yields and WUE while bearing in mind the complex interactions of WUE with plant genetics, developmental stage, and abiotic conditions.

The ability to track plant growth over time non‐destructively can reveal biological insights that could be missed through endpoint analysis alone (Fahlgren et al., 2015). For example, average WUE of mannitol‐treated RTx430 and BTx623 plants was already lower than control plants by Day 4, yet F _V /F _M in those same plants was not yet distinguishable from control plants, indicating that WUE in sorghum is negatively affected by osmotic stress before maximum quantum efficiency of Photosystem II is affected. Also, because biomass accumulation was already reduced by Day 4 in mannitol‐treated BTx623 plants compared with control plants, something other than a reduction in Fv/Fm was impairing growth rates in that accession. End‐point analysis would not have revealed how osmotic stress impacted growth and Fv/Fm differently at different times among the genotypes. Non‐destructive phenotyping also brought to light that stress response onset does not necessarily correlate with stress response magnitude, as evidenced by the late, but severe drop in F _V /F _M in Tx7078 plants compared with the other genotypes. Previous work utilizing non‐destructive phenotyping techniques has also identified a disconnect between stress response onset and magnitude in Arabidopsis (Awlia et al., 2016). Collectively, these observations further support the value of non‐destructive approaches to capturing the dynamic nature of plant phenotypes across time and environments (Granier & Vile, 2014).

Plants actively modify the soil properties of their growing environment to optimize growth by improving nutrient uptake and mobilization and to regulate soil microbial dynamics (Falhof et al., 2016; Marschner et al., 1986; Shen et al., 2013). Depending on the biotic and abiotic conditions in the soil, roots can exude sugars, amino acids, enzymes, organic acids, and phytosiderophores to optimize conditions for symbiotic rhizospheric microorganisms (Ryan et al., 2009) or for direct nutrient uptake (Marschner et al., 1986; Shen et al., 2013). The non‐destructive nature of our closed‐system hydroponic approach allowed us to more readily identify a connection between root zone acidification and biomass accumulation than had we grown plants in soil for destructive end‐point analyses. For example, after 2 days of growth in closed tubes, both control and mannitol‐treated plants of all genotypes acidified their root zones equally. At this time point, there was also no difference in total biomass accumulation between control and mannitol‐treated samples in any genotype. Although root zone acidification gradually decreased over time in control samples of all genotypes, it remained fairly constant for mannitol‐treated plants. That mannitol‐treated plants continued to acidify their root zone throughout the experiment suggests a greater and prolonged investment of root exudates to actively modify their root environment. One explanation could be that mannitol stress can impair the uptake and translocation of phosphorus (Resnik, 1970) and potassium (Slama et al., 2007). Interestingly, the average pH after every 2 days of growth in mannitol was 5.5–5.8 (Data S1), which is a range that favors plant uptake of all essential nutrients (Pennisi & Thomas, 2005; Reed, 1996). Thus, root zone acidification in response to mannitol stress could be an adaptive response to improve mineral nutrient uptake and mobility.

Various aspects of the dynamic effects of root zone conditions on plant growth remain poorly understood (Kardol et al., 2013). Additionally, many plant responses to environmental conditions are age‐dependent (Bond, 2000; Panter & Jones, 2002). Thus, although hydroponic cultivation is an established approach in plant biology research (Asao, 2012), the isolation and modularity of the root zone environment afforded by our approach open the door to additional avenues of experimentation for directly studying dynamic and age‐dependent processes. For example, researchers could finely control the characteristics of the solution in each tube to investigate the temporal effects of mineral nutrition or biotic stress on whole‐plant development and physiology and to analyze how these characteristics differentially affect shoot versus root growth. Our approach also makes it easier to investigate how changes in root architecture and morphology over time influence plant growth. Although currently our approach does not distinguish shoot from root biomass, it could easily be paired with imaging‐based phenotyping technologies such as PhenoBox (Czedik‐Eysenberg et al., 2018) or PlantCV (Gehan et al., 2017) to quantify both components separately. Future studies that incorporate such imaging technologies could directly quantify the relationship between root zone acidification and root growth characteristics.

Another future application would be to examine the effect of root zone oxygenation on growth. Some plant species require more oxygen in their root zone than others (Asao, 2012; Narsai et al., 2011). Supplementing hydroponic nutrient solution with hydrogen peroxide can increase the availability of dissolved oxygen in the media and can increase nutrient uptake and chlorophyll content in various species (Butcher et al., 2017; Liu et al., 2022). Although suitable concentrations of hydrogen peroxide would need to be established to avoid phytotoxicity, its use in the nutrient solution would not require any physical modifications to the closed‐hydroponic design. Additionally, tubes could be opened more frequently than every 2 days to further increase root zone oxygenation. Although 50‐mL tubes were sufficiently large for sorghum plants at the third leaf stage, larger tubes could be used for either older or larger species, as required.

Efforts to identify and engineer more resource‐efficient and resilient crops will benefit from phenotyping approaches that can noninvasively quantify whole‐plant growth and physiology over time. Our approach allows for inherently greater throughput than traditional approaches and can, in its entirety, be implemented using cheap and readily available components. Its versatility for studying various aspects of plant growth and physiology makes it a valuable method for plant scientists in studying plant growth as a whole and identifying superior and more resilient crops.

A detailed, step‐by‐step protocol can be found at protocols.io; dx.doi.org/10.17504/protocols.io.q26g7p6m1gwz/v1.

An R notebook containing scripts for visualizing and analyzing data from these experiments can be found at https://rpubs.com/danny_ginzburg/1143647.

Daniel N. Ginzburg conceived of and designed the closed‐hydroponic apparatus. Daniel N. Ginzburg and Seung Y. Rhee conceived of the experimental design. Daniel N. Ginzburg and Jack A. Cox performed the growth experiments. Jack A. Cox and Seung Y. Rhee provided intellectual contribution and input into the manuscript organization. Daniel N. Ginzburg wrote the manuscript, and Seung Y. Rhee edited the manuscript. Seung Y. Rhee advised Daniel N. Ginzburg and Jack A. Cox.

The authors declare no conflict of interest.