A comprehensive rhizobox pipeline for analyzing pepper root system architecture under well-watered and water deficit conditions

A comprehensive rhizobox pipeline for analyzing pepper root system architecture under well-watered and water deficit conditions | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Method Article A comprehensive rhizobox pipeline for analyzing pepper root system architecture under well-watered and water deficit conditions Chee Gang Ngui, Leah McHale, Christine Sprunger, Kristin L. Mercer This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7697278/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Drought stress can significantly impede plant productivity, adversely impacting crop yields. The root system is an important plant organ contributing to drought resistance mechanisms. Therefore, assessing root systems under drought stress conditions can provide insights to identify root traits associated with enhanced drought resistance. When seeking dense and high-quality root data, root phenotyping can be complex, costly, and time-consuming. The objectives of this study were to establish a method to grow chili pepper plants in a soil-based rhizobox container under water deficit conditions and compare two methods for collecting two-dimensional root trait data from their roots. Method We grew two chile peppers ( Capsicum annuum ) accessions in soil-based rhizobox containers to analyze the responses of root architecture traits under well-watered and water-deficit conditions during the vegetative stage. The root traits were phenotyped using two different methods. The first method involved non-destructive in-box imaging of roots in situ through acrylic glass while the plant grew. The second method involved scanning destructively harvested and washed roots—the gold standard for root measurements. For the first method, we developed a pipeline for rhizobox studies to demonstrate the response of root system architecture to water deficit over time and assessed the quality of non-destructive in-box imaging methods as compared to scans of destructively harvested and washed roots. We used a relatively large rhizobox (53.34 cm in width x 78.73 cm in height) into which we established and maintained well-watered and water deficit conditions based on the field capacity and permanent wilting point of the soil (Bodner et al. 2017; Cassel & Nielsen 1986).Our in-box root imaging pipeline captures high-resolution root images with an affordable camera that can achieve a maximum resolution of 9152 x 6944 pixels, as well as high-quality root segmentation using a robust graphical user interface-based software called RootPainter (Smith et al. 2022). Results Root growth decreased under water deficit compared to well-watered conditions. There were strong positive relationships between total root length using the washed scanned method and the in-box imaging method. The same was observed for root perimeter and most of the total root length distinct root diameter classes, but not for average root diameter. Some of these relationships weakened under water deficit conditions. In addition, we also found a strong relationship between root biomass and total root length using both phenotyping methods. Conclusion Overall, we developed a rhizobox pipeline for phenotyping the root system architecture of chile pepper plants under both well-watered and water-deficit conditions. We showed that measurements taken via non-destructive in-box imaging strongly predict those taken directly on washed scanned roots, with the added benefit of allowing repeated measurements over time. Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction When experiencing drought, plant function can be severely disrupted, resulting in stunted growth, wilting, increased vulnerability to disease infection, and yield reduction (Seleiman et al., 2022). Monitoring belowground responses to drought could be critical for overall plan resilience. For instance, plant roots are the loci for plant-soil interactions, take up water and nutrients for plant development, and anchor the plant. Root system architecture (RSA) is the physical arrangement and traits of roots within the soil. RSA can be influenced by the allocation of water and nutrients, altering the development of lateral roots, elongation of root tips, and biomass allocation (Correa et al. 2019 ). Phenotyping plant response under water deficit and identifying plant traits that contribute to sustaining productivity in water-limited environments are essential for genetic improvement of crop plants (Li., et al., 2014). Measuring RSA traits in plants grown under a range of conditions can help us evaluate genetic resources for important traits, such as drought tolerance. Yet research on improving plant root systems is relatively scarce compared to that of aboveground traits due to the difficulty of capturing RSA measurements (Takahashi & Pradal, 2021 ). Root studies can be classified by their use of destructive and non-destructive phenotyping (Li et al., 2022 ). Destructive phenotyping involves collecting the root system from the soil. The drawbacks include the time-consuming nature of the methods, the fact that you can only phenotype the root at a single timepoint. Non-destructive phenotyping allows for assessment of the same traits on the same plant over time (i.e., without destructive harvesting). In non-field-based root growing methods, soil media (like pots, rhizoboxes, or rhizotrons) or soil-less media (like agar, pouches, and hydroponic systems) can be used for both destructive and non-destructive phenotyping (Li et al., 2022 ). Non-field-based root-growing methods can simplify root phenotyping; however, they may fail to replicate natural conditions, especially when small or irregularly shaped containers are used. Such setups often create uneven abiotic conditions, particularly in the deeper soil layers. (Mašková & Klimeš, 2020 ; Li et al., 2022 ; Paez-Garcia et al., 2015 ; Kuijken et al., 2015). Moreover, most of these non-field-based methods, when used non-destructively, allow for the application of high throughput root phenotyping techniques and measurements of RSA over time (Li et al., 2022 ). Phenotyping techniques for RSA can be separated into systems that allow for two-dimensional (2D) and three-dimensional (3D) views of the root system. The 2D techniques mainly use a camera or flatbed scanner to visualize the root system, either non-destructively in transparent containers or from destructively harvested, washed roots (Li et al., 2022 ). The main advantage of 2D root phenotyping techniques is that the root imaging is affordable and simple to establish compared to 3D root phenotyping techniques, such as X-ray computed tomography, magnetic resonance imaging, electrical impedance tomography, neutron tomography, and ground penetrating radar, which are expensive. Moreover, it can be challenging to develop the technology in individual laboratories (Li et al., 2022 ). Studies employing rhizoboxes use soil as a medium and employ non-destructive 2D root phenotyping. The advantage of a rhizobox is that it allows us to observe the root growth in a soil environment over time by tracing or imaging roots through the transparent side of the box as the roots grow. One disadvantage is that rhizoboxes are constructed to have very little space between front and back acrylic panels, which can constrain natural root growth. Nevertheless, Mašková and Klimeš ( 2020 ) reported no significant difference between the total root biomass of 4-week-old plants grown in flat pots (rhizobox, 19.5cm height x 15 cm width x 1 cm inner diameter) and a more typical square pot (upper size 7 × 7 cm, bottom size 5 × 5 cm, height 8 cm); they did suggested that researchers be cautious of the limitations of rhizoboxes on natural root growth (Mašková and Klimeš, 2020 ). In the past, limitations of 2D root phenotyping with soil media in rhizoboxes included the lack of high-resolution root images and low accuracy of root segmentation (i.e., the process used to separate roots from the surrounding soil in the image). This was primarily due to the expense of high-quality cameras and the underdeveloped state of segmentation software (Li et al., 2022 ). Such technical difficulties can now be overcome, with affordable cameras achieving a resolution of 9152 x 6944 pixels and high-quality root segmentation achieved using robust graphical user interface-based software such as RootPainter (Smith et al. 2022 ). Rhizobox experiments under well-watered (WW) and water deficit (WD) conditions are few, but have been used to develop phenotyping methods with sugar beet, (Bodner et al., 2017 ;Rambla et al, 2023) and to measure the responses of root traits, such as total root length, root dry weight, and root-to-shoot ratio, in soybeans grown in minirhizotrons (24.5 × 24.5 × 2.5 cm) (Dayoub et al. 2021 ). Surprisingly, in this latter study, they found that there was no effect of WD (Dayoub et al. 2021 ). Thus, we still lack a clear understanding of how WD manipulations in rhizoboxes may influence RSA traits via non-destructive methodology, especially in vegetable crops. By growing two distinct chile pepper accessions in relatively large rhizoboxes (53.5 cm (W) x 78.5 cm (H) x 1 cm (D)), under WW and WD conditions, we were able to non-destructively assess 2D RSA and compare those results to destructive assessments of 2D RSA at the end of the experiment. We had three objectives. 1) To measure biological responses in RSA traits to our WW and WD treatments over the course of the experiment. 2) To determine the degree to which non-destructive in-box 2D imaging of RSA traits can predict and possibly replace 2D scans of destructively harvested and washed roots. 3) To assess the feasibility of utilizing traits, such as total root length (TRL) measured through non-destructive in-box 2D imaging methods to predict destructively harvested root biomass. With these analyses, we aimed to provide guidance for researchers trying to improve and streamline their phenotyping pipeline without sacrificing data quality. Method and materials Plant materials We selected two distinct chile pepper accessions, Ca0045 and PI586666, as our model plants for the rhizobox study due to their differences in aboveground characteristics and origin location (McCoy et al., 2022). PI586666 is a commercial accession with a New Mexican pod type provided by the Plant Genetic Resources Conservation Unit of the USDA, Griffin, Georgia. Ca0045 is a Costeño Rojo landrace accession collected from a plantation along the coast of the Mexican state of Oaxaca (personal communication, L. Jardón Barbolla). All the seeds for accessions were grown in the greenhouse, Columbus, Ohio. Seed sterilization, germination, transplantation, and growing conditions We sterilized seeds with 0.825 % of sodium hydrochloride for ten minutes and rinsed with distilled water. We germinated the seeds in sanitized 11 cm x 11 cm square germination boxes with 75 g of autoclaved sand (Quikrete Co., Atlanta, GA, USA) and 15 ml of distilled water. We placed the seeds in the Conviron G30 germination chamber (Conviron, Winnipeg, Manitoba, Canada) for seven days. Germinated seeds were ready to transplant into rhizoboxes once the hypocotyl of the seedling turned green and the radicle reached 2 to 3 cm with the majority of the cotyledon part was enclosed by the seed coat (Figure 1e and f), then we transplanted the seedling into the soil-filled rhizobox. We found that earlier or later transplanting reduced the rate of seedling emergence from the soil. Normally, seedlings of both accessions emerged from the soil within 3 to 4 days. We applied 0.8 grams of slow-release 14 (nitrogen) -14 (phosphorus)-14 (potassium) fertilizer (Osmocote, Scotts Miracle-Gro, Marysville, OH, USA) to the soil surface after transplanting was completed. We grew the seedlings in rhizoboxes in the Howlett Greenhouse at the Ohio State University in Columbus, Ohio. The average and maximum temperature per run differed (see Supplementary Material 2), depending on the time of year and the use of shade cloth. For instance, in the summer (i.e., runs 4 and 5), the average and maximum temperature were 24°C and 36°C, respectively, so we placed shade cloth over the rhizoboxes to reduce direct heat. The rhizobox size was 53.5 cm (W) x 78.5 cm (H) x 1 cm (D), while the soil filling volume was approximately 47 cm (W) x 74 cm (H) x 1 cm (D). The description of building these rhizoboxes can be found in Supplementary Material 4. The rhizobox was supported by a customized white and black poly cover (Figure 1a). Each rhizobox was placed on a greenhouse bench with a custom PVC rhizobox holder based on the design suggested in Schmidt et al. (2018), but larger (Figure 1d and g). We kept the rhizoboxes at a 45° angle to maximize root visibility through the back panel (Bontpart et al. 2020). After transplanting, the exposed soil was covered by a white poly film to decrease evaporation (i.e., maintain plant available water in the soil) and allow us to achieve a long-term gradual reduction in water availability (Figure 1b) (Osmolovskaya et al., 2018). Once the seedling emerged, we opened a small area in the middle of the white poly film to allow it to grow toward the light. The experiment lasted for a duration ranging from 54 to 61 days. Soil preparation We used a soil mixture of 70% topsoil and 30% dark grey mason sand (by volume) in the rhizoboxes. First, we sifted the topsoil (Mr. Mulch, Columbus, OH) through a 0.02mm test sieve to remove large particles—a slow process (40 minutes for 5-gallons). Fine mason sand (smaller than 0.02mm in diameter) was most suitable since its dark grey color facilitated root segmentation. The sand and sifted soil were autoclaved separately in plastic bags to steam for an hour to kill weed seeds. The autoclave took 55 minutes to reach 200°F and the soil was exposed to a constant 200°F for 5 minutes (Bitarafan et al., 2022) and then air-dried in an open pan. Experimental design We ran a factorial experiment assessing the effects of chile pepper accession and irrigation level in factorial combination. We performed multiple runs over time; within each run, we had one or multiple replicates separated into spatial blocks. Thus, in the entire experiment, the total number of runs was five, which were divided into seven blocks (Supplementary Material 2). Within each block, each rhizobox was an experimental unit since a rhizobox was the scale at which we applied one replicate of the factorial combinations of accession and irrigation treatment. The high levels of mortality in some runs and for some treatment combinations resulted in uneven representation across blocks. The irrigation application in runs four and five resulted in no plant mortality so, by the final runs, our methodology stabilized. However, due to lack of WD treatment representation across blocks, block 7 was assigned with WD. As described below, we included blocks in our analyses when using the ANOVA approach, but not when we used the regression approach. Data Collection The data was collected from scanned washed roots collected at the end of the experiment, as well as in-box imaging root traits. The experiment was terminated 61 days after transplanting in run 1 but 55 days after transplanting in the other runs. For objective 1, repeated measures analyses of TRL captured by in-box imagining was analyzed from 25 days after transplanting to day 55. The image of the root was taken every Monday, Wednesday, and Friday. In the fifth run, only the WD treatment was applied since there was less representation of WD treatments in runs 3 and 4. Manipulation of water and soil needed for the water deficit and well water treatments. In this section, we incorporated Bodner et al. (2017), Passioura (2006) and Cassel & Nielsen (1986) methods to comprehensively measure soil and water requirements to implement WW and WD effects on the plants. First, we determined the weight of the topsoil and mason sand combination by adding the air-dried topsoil into a beaker until it reached 350 ml and measured the weight. We did the same method for mason sand until it reached150 ml. Based on the weight, we knew the weight of 70% topsoil and 30% mason sand by volume. Next, we measured soil bulk density (Bodner et al., 2017). Subsequently, we determined field capacity (FC) and permanent wilting point (PWP) of our soil. To do this, we used a pressure plate device as described in Supplementary Material 1 . Typically, FC is measured at a water potential of 33 kilopascals (kPa). However, in our study, we used a lower water content at a matrix potential of 8 kpa due to the height of our rhizobox (Passioura, 2006). Finally, we determined the plant available water (PAW) for the soil mixture for the WW and WD treatments (Bodner et al., 2017). The full procedure is described in Supplementary Material 1. Soil filling and irrigation treatments adjustment for well-watered and water deficit treatments To fill the rhizobox with soil, we began by measuring the initial weight of the empty rhizobox on a floor scale. The first weight of soil was based on soil volume (47 cm × 74 cm × 1 cm) with a balance scale. We then mixed mason sand, topsoil and a specified volume of water in a bucket. The volume of water was based on the PAW of irrigation treatments. Adding water at this stage prevented the separation of mason sand and topsoil, ensuring a homogeneous mixture. Using a funnel, we carefully added the homogenized soil mixture into the upright rhizobox, gently swaying the rhizobox to ensure the soil settled uniformly and firmly. Due to bowing of the acrylic side panels, we added extra of soil mixture and water (the water added was based on the weight of the soil and treatment) to ensure a uniform height of 74 cm in each rhizobox. Therefore, the final weight of soil and water can vary slightly for each rhizobox. The process of adding the soil into each rhizobox took approximately 45 minutes. We used the concept of PAW to determine the exact amount of water required for each treatment, with slight adjustments to the percentage of PAW across different blocks or runs (Supplementary Material 3). This adjustment was necessary because young chile pepper seedlings in rhizoboxes showed high sensitivity to WD stress. After transplanting, we added water to the opening section of the rhizobox. In WW, the percentage of PAW for all the runs were increased after transplant, except run 1, which remained the same. The percentage of PAW in WW treatment was maintained until the experiment ended. In WD treatment, the PAW of all the runs were elevated after transplant except for run 1. Then, starting at day 18, we reduced the PAW percentage. Every 7 days, we reduced the PAW 10% until it reached 40%, which we then maintained until the experiment ended (Supplementary Material 3). Runs 1, 2, 3, 4, and 5 had 33%, 0%, 50%, 0% and 0% plant mortality, respectively (Supplementary Material 2). In-box root imagining In the root image-capturing process, we used the 64MP Hawkeye camera (Arducam, China) mounted with a Raspberry Pi 4 Model B (Raspberry PI, United Kingdom) used as a low-cost computer to produce high-quality root images with 9152 x 6944 pixels. The camera hardware and software setup were described in the user guide (https://docs.arducam.com/Raspberry-Pi-Camera/Native-camera/PiCamera2-User-Guide/). We took the root images at an image station we constructed to ensure all rhizoboxes at all time points were imaged in the same position, using the same light source and quantity (Refer to Supplementary Material 5 for image station structure.). We captured root images from both the front and back panels of the rhizoboxes; more roots were visible from the back panel, which faced down. To process the root images, we used two free software tools. The first was RootPainter (Smith et al., 2022), which was designed for root and soil segmentation and allowed us to obtain the image of the isolated root system from the rhizobox for further root analysis. We used 530 root images captured at various developmental stages of chile pepper in both irrigation treatments as a training data set to generate the final best model to extract the root image for further analysis. We used two methods to visualize the accuracy of root and soil segmentation. We checked the model’s accuracy using the Sørensen–Dice coefficient (Zou et al., 2004) in the metrics plot of RootPainter. We also compared the raw and segmented images by eye to determine the accuracy of the segmented image relative to the actual roots. We then used RhizoVision Explorer (Seethepalli and York, 2022) to analyze root images and quantifying RSA traits. Before conducting the analysis, we described the procedure for measuring the root image’s pixels per millimeter in Supplementary Material 7. The root traits extracted by RhizoVision Explorer from our in-box images and that we used for analyses related to objectives 1-3 were total root length (B_TRL, cm); root perimeter (B_PE, cm), cm; average root diameter (B_AVR); total root length for roots with 0 to 0.5 mm diameter (B_RD 0-0.5, cm); total root length for roots with 0.5 to 1 mm diameter (B_RD 0.5-1, cm); total root length for roots with 1 to 2 mm diameter (B_RD 1-2, cm) and total root length for roots with above 2 mm diameter (B_RD >2, cm). All the features in RhizoVision Explorer are described in the manual (Seethepalli & York 2020). Root traits such as B_TRL and B_PE from the front and back panels of each root sample were subsequently combined (e.g., Total root length in sample 1 = TRL, front panel + TRL, back pan Root washing and scanning washed roots At the end of the experiment, we washed the roots using methods from Sprunger et al. (2018), losing few roots. First, we placed a 1 mm mesh mat on the drain to capture all the broken roots. We then laid the rhizobox on top of two bars and used a garden hose nozzle to wash away the soil (Figure 2b). Once the roots were clean, we removed the shoot biomass. We cut multiple of the secondary roots to separate the root system into sections to facilitate the scanning process. Then we placed the root sections in ziplock bags in the refrigerator and sprayed with deionized water to prevent them from drying out prior to scanning. To scan the roots, we submerged them in a water-filled tray to untangle the roots and scanned them using an EPSON Expression 10000 XL scanner (Epson, Japan) connected to the computer (Figure 2b). Winrhizo software (Regent Instruments Inc, Canada) was utilized for capturing the scanned root image. Prior to scanning, we configured the settings in Winrhizo, setting the resolution at 400 dot per inch (dpi) and defining the image acquisition parameters with a width 30 cm and a length 40 cm. Once the scanned root image acquired, it loaded to RhizoVision Explorer (Seethepalli & York 2020). Before conducting the root analysis, we performed various settings for image preprocessing and feature extraction. In the image preprocessing phase, we selected the “broken root” analysis mode since the scanned whole root system appeared in disconnected pieces. We used a resolution of 400 dpi and set the image thresholding level to 200. The “filter non-root object” was set to 0.5, we left all other boxes unticked. For feature extraction, we applied a root pruning threshold of 1 and set the roots into four different diameter ranges. It is important to note that the specific settings in RhizoVision Explorer may vary depending on factors such as the resolution of the root image, the density of the roots, and the presence of debris (soil, organic matter and unwanted materials). The scanned washed root traits were the following: total root length (S_TRL, cm); root perimeter (S_PE, cm), cm; average root diameter (S_AVR); total root length with 0 to 0.5 mm diameter (S_RD 0-0.5, cm); total root length with 0.5 to 1 mm diameter (S_RD 0.5-1, cm); total root length with 1 to 2 mm diameter (S_RD 1-2, cm) and total root length with above 2 mm diameter (S_RD >2, cm) . Root biomass Once roots were scanned, we removed the excess water on the roots with a paper towel and measured the root fresh weight (RFW, g). Then, we obtained root dry weight (RDW, g) by placing them in an oven at 65°C for three days and reweighing (López-Serrano et al. 2019). Statistical analyses All statistical analyses were performed in SAS software (version 9.4) (SAS Institute Inc., 2013). All general linear mixed model analyses were checked for the assumptions of normal distribution of residuals, constant variance of residuals, and independence of errors, as well as the presence of outliers. Similarly, for all linear regression analyses, we checked the assumptions of linearity between predictor and response variables, normal distribution of residuals, constant variance of residuals, independence of errors, the presence of outliers, and the influence of leverage (extreme values of predictor variables). We performed general linear model analyses for all three objectives. For objective one, we took a mixed model ANOVA approach using Proc MIXED. In this analysis, the fixed factors in our model were accession (Ca0045 and PI586666) and irrigation treatment (WW and WD), while the random factor was block. The response variables were B_TRL, S_TRL, RFW and RDW. The mean separation was done on least squares means using Tukey’s honestly significant difference (HSD) test with α=0.05. Natural log transformation was performed on S_TRL and RFW to normalize residuals. No transformation was needed for B_TRL and RDW due to fit in the assumption of linear regression. Analyses for objective one also included repeated measurement analysis with an in-box imaging method using a general linear mixed model in Proc MIXED. In our general linear mixed model, we included the fixed factors of accession (Ca0045 and PI586666), irrigation treatment (WW and WD), and day (25, 27, 29, 32, 34, 36, 39, 41, 43, 46, 48, 50, 53 and 55 day), as well as their interactions. We also included block as a random factor. Since repeated measures analyses require appropriate variance covariance structures, we tested structures for our model and found the first order interdependence covariance structure resulted in the lowest AIC (Akaike's information criterion) and BIC (Bayesian information criterion) and was significantly better than the simpler structure (Unstructured, compound symmetry, and heterogeneous compound symmetry) using a log likelihood test. Tukey’s honestly significant difference (HSD) test with α=0.05 was utilized for mean separation. Natural log transformation was executed on B_TRL trait to normalize residuals. For the second objective, we took a regression approach running a general linear model using Proc REG to determine the degree to which RSA traits from the scanning washed root method could be well-predicted, and potentially replaced, by RSA traits from in-box imaging. Therefore, the predictor variable (X) was always an RSA trait from our in-box imaging and the response variable (Y) was an RSA trait from our scans of washed roots. The seven analyses were as follows: 1) predicting S_TRL from B_TRL; 2) predicting S_PE from B_PE; 3) S_AVRL from B_ AVRL; 4) predicting S_RD 0-0.5 from B_RD 0-0.5; 5) predicting S_RD 0.5-1 from B _RD 0.5-1; 6) predicting S_RD 1-2, cm from B_RD 1-2; and 7) predicting S_RD above 2 cm from B_RD above 2 cm. These analyses were performed separately on data from WW and WD treatments, as well as with both treatments combined (TC). For objective three, the goal was exploring the effectiveness of fresh and dry weight in predicting total root length in both scanning and in-box imaging. The response variable was either S_TRL or B_TRL. The predictor variable was either RFW or RDW. We performed four sets of analyses here: 1) predicting B_TRL from RFW ; 2) predicting B_TRL from RDW; 3) predicting S_TRL from RFW; and 4) predicting S_TRL. For objectives 2 and 3, analyses of WW and TC datasets for some traits better fit the assumptions of linear regression with natural log-log transformation. However, we present the analyses and interpretations using untransformed data because the R 2 values and p-values of the models showed little difference between the untransformed and natural log transformation datasets and interpretation of untransformed data is more straightforward. Results For objective 1, we found a significant effect of the irrigation treatment on all the response variables analyzed (Table 1). Table 1. F values and their significance for general linear mixed model results of end of experiment root system architecture traits. The numerator and denominator degree of freedom for each trait was 1 and 15, respectively. n No transformation. z Natural log transformation. Tukey’s HSD test with **** (p value ≤ 0.0001), *** (p value ≤ 0.001), ** (p value ≤ 0.01), * (p value ≤ 0.05), and ns (p value > 0.05) used to determine the significant effects. S_TRL cm (total root length for scanned washed roots), B_TRL cm (total root length, in box imaging), RFW g (root fresh weight) and RDW g (root dry weight). Scanned washed roots (cm) In-box image (cm) Root biomass (g) S_TRL z B_TRL n RFW z RDW n Accession (A) 0.91 (ns) 1.06 (ns) 0.16 (ns) 0.22 (ns) Irrigation (I) 60.62 (****) 6.44 (*) 31.32 (***) 11.28 (**) A x I 1.55 (ns) 0.83 (ns) 0.09 (ns) 0.23 (ns) The back-transformed means for S_TRL (from scanned washed roots) were 6063 cm (WW) and 1619 cm (WD). In other words, TRL (from scanned washed roots) was 3.7 × longer under the WW treatment than under the WD treatment (Table 2). Furthermore, the WW and WD treatments for B_TRL (from in-box imaging) were 2441 cm and 895 cm, respectively; B_TRL (In-box imaging) was 2.7 X longer in WW treatment than in WD treatment (Table 2). Additionally, the back-transformed means of RFW were 8.41 g (WW) and 3.06 g (WD), which RFW was 2.7 X greater under WW conditions than under WD conditions (Table 2.3). Likewise, the RDW were 0.5 g (WW) and 0.1 g (WD), or 5 X greater under WW conditions than under WD conditions (Table 2). Table 2. Least square means (and s.e.) and mean separation for irrigation treatment effect at the end of experiment. n No transformation. z Natural log transformation. WW (irrigation treatment effect for well-watered) and WD (irrigation treatment effect for water deficit). Superscript letters on least squares means used to indicate significant mean separation using Tukey’s HSD Test (P=0.05). The natural transformation means were back-transformed. Means sharing a letter are not significantly different. S_TRL cm (total root length for scanned washed roots), B_TRL cm (total root length, in box imaging), RFW g (root fresh weight) and RDW g (root dry weight). Scanned washed roots (cm) In-box image (cm) Root biomass (g) S_TRL z B_TRL n RFW z RDW n Irrigation WW 6063.2 (0.23) a 2441.3 (551.07) a 8.41 (0.2) a 0.51 (0.12) a WD 1619.7 (0.22) b 894.7 (499.06) b 3.06 (0.19) b 0.10 (0.10) b Our repeated measurement analysis identified a significant effect of the day and irrigation treatment interaction (Table 3). In-box imaging showed that the B_TRL increased over time under WW and WD conditions; however, by day 50 and for the rest of the experiment, B_TRL in WD was lower than in the WW (Figure 3). Thus, in-box imaging was able to discern changes in an important RSA trait over time and identify when our experimental treatments caused its value to diverge. Table 3. General linear mixed model results of repeated measures analysis for TRL measured by in-box imaging of chile pepper roots grown in rhizoboxes in the greenhouse at Ohio State University, Columbus, OH. The numerator degree of freedom (N DF), denominator degree of freedom (D DF), F value, and P value. z (natural log transformation). B_TRL, cm (total root length, in-box imaging). B_TRL cm Z Effects N df D df F value P value Accession (A) 1 13.9 0.21 ns Irrigation (I) 13 47.1 93.71 **** Day (D) 1 14.2 0.15 ns A x D 13 47.1 0.99 ns A x I 1 13.5 0 ns D x I 13 47.1 13.34 **** A x I x D 13 47.1 0.4 ns We found that RSA traits measured with in-box images mostly predicted values from washed and scanned roots accurately. Both treatments combined (TC) for all traits, all relationships were positive and significant (Supplementary Material 6). For instance, for total root length, a slope of 2.2 (± se = 0.09), means that for every 1 cm increase in B_TRL (in-box imaging), we found a 2.2 cm increase in S_TRL (washed root scan) (Supplementary Material 6a). Moreover, most overall analyses resulted in very high (>0.90) or high (>0.70) R 2 values, with total root length (0.96), perimeter length (0.94), and total root length (0-0.5mm diameter) (0.95) in the very high R 2 category (Supplementary Material 6a, b, and d) and total root length for 0.5-1 mm diameter (0.85), 1-2 mm diameter (0.72) and >2 mm diameter (0.82) in the high R 2 category (Supplementary Material 6e-g). Only the analysis of average root diameter had a low R 2 value (0.27) (Supplementary Material 6c). When we performed these regression analyses separately on plants grown in the WW and WD conditions (Figure 4), we found that the separate models did as good a job of predicting the data for TRL, perimeter, and total root length (0-0.5mm diameter) (Figure 4a, b, and d) as they had for the overall dataset (Supplementary Material 6a, b, and d). However, for the other traits, the model fit the data much better under WW than under WD (Figure 4c and e-g). In particular, R 2 values in the WD analyses plummeted for average root diameter (0.09, Figure 4c), total root length for 1-2 mm diameter (0.35, Figure 4f), and >2 mm diameter (0.05, Figure 4g) and slopes for average root diameter and total root length for >2 mm diameter became non-significant under WD. Thus, predicting results of washed scanned roots from in-box imaging was less reliable under WD than under WW. All linear regression analyses predicting either fresh or dry root biomass from total root length (either from in-box imaging and scanning methods) showed significant positive relationships and very high (>0.9) or high (>0.7) R 2 values (Figure 5). However, as mentioned above, all R 2 values from WD analyses (0.61-0.72) were lower than those from the WW (0.78-0.9) and TC (0.84-0.92) analyses (Figure 5). Discussion The effectiveness of irrigation on the morphological traits of chile peppers When discussing plant resource allocations, Bloom et al. (1985) implies that a plant can expand as long as the benefit of growth outweighs the cost, which means the cost of growing more biomass is relatively low. However, as resources (water, light or nutrients) become limited, the cost of producing more growth increases. As this relates to our results, the TRL (from in-box imaging and scanning methods) and root biomass were stunted under WD conditions. This finding confirmed that our imposed moisture treatments allowed us to assess the difference in biological responses under WW and WD treatments by the end of the experiment, but also over time. Sprunger et al. ( 2018 ), observed that perennial intermediate wheatgrass increased root biomass as N inputs increased, supporting Bloom’s theory. It is important to note that the theoretical framework that Bloom et al. (1985), as well as the empirical support for that theory provided by Sprunger et al. ( 2018 ) were based on field studies Our work with rhizoboxes, however different, also supports the same fundamental resource allocation patterns. These effects of WD on root growth may be primarily attributed to water limitation, which disrupts metabolism and physiological processes in plants, leading to the slowing of growth of new meristematic cells required for cell elongation. (Smirnoff, 1993 ; Foyer et al., 1994 ; Hirt & Shinozaki 2003; Wajhat et al., 2023). The exploration of WW and WD treatments in rhizobox experiments is relatively limited in existing research. However, Bodner et al. ( 2017 ) conducted a rhizobox experiment examining WW and WD treatments that focused on method development. The study did not discuss the biological response of sugar beet root under WW and WD conditions. However, it did mention some physiological aspects, such as sugar beet cultivars with larger root surface areas having higher average stomatal conductance and keeping their stomata open longer. Scanning of washed root traits can be substituted with in-box root imaging Analyzing scans of washed roots is considered the gold standard for accurately quantifying RSA in plant science. This experiment was the first instance of comparing results from scanned washed roots to those found through in-box root imaging. We found a strong relationship between TRL measured using in-box images and scans of washed roots overall and when analyzed by irrigation treatment. This means we can accurately predict TRL that would have been obtained from scanning washed roots based on the values calculated from in-box imaging. The same predictive ability is true for other traits. The values from in-box imaging of root perimeter, TRL for root diameters 0-0.5 mm, and TRL for root diameters 0.5-1 mm also predicted values from scanned washed roots well with R 2 of ≥ 91%, ≥ 85% and ≥ 70%, respectively. Despite this strong predictive relationship, the in-box imaging has some drawbacks in comparison to scanning washed roots. All root traits from in-box imaging underestimated the true measurements from scanned washed roots. Scanning washed roots allows for more accurate root trait measurement of the entire root system. By contrast, in-box imaging had limitations due to its two-dimensional nature. This method cannot fully capture all roots; as some remain hidden (Li et al., 2022 ). Yet, our analysis indicates that the degree to which roots remain hidden is predictable. Moreover, to minimize the underestimation, we captured root traits from both the front and back panels when using in-box imaging. Thus, in-box imaging can be used instead of scanning of washed roots to reduce labor costs of RSA experiments. Four distinct TRL root diameter classes showed the same pattern, with R 2 in WD analyses being lower than in WW and overall analyses. The high R 2 in TC was driven by the high R 2 in WW. In the WD analysis, the R 2 value was lower than that for WW for TRL for root diameters of 1–2 mm and greater than 2 mm. Moreover, some predictions (e.g., for TRL of roots greater than 2 mm) were not significant at all. In contrast, the WW and TC analyses always yielded significant predictions. We also found when the TRL root diameter thickness increased, the R 2 values decreased. These might relate to the TRL root diameter distribution for in-box imaged and scanned washed roots. The TRL measurements from in-box imaging predominantly fell within the 0-0.5 mm and 0.5-1 mm diameter ranges. In contrast, the TRL from scanned washed roots was mainly within the 0.5-1 mm and 1–2 mm diameter ranges. In addition to the average root diameter, our models performed well in predicting variation for the WW analyses. However, they were less effective for the WD and TC analyses due to low R² values and a non-significant slope for WD. It is important to note the limitations of in-box imaging and scanned washed roots, which might contribute to inaccuracies in measuring average root diameter and TRL root diameter classification. For instance, by visualizing the root growth from the in-box imaging, there were multiple misrepresentations. First, if two separate roots grew close together or up against one another, RootPainter and RhizoVision Explorer sometimes recognized them as a single root with a thicker diameter, leading to inaccuracies in assigning TRL to different root diameter classes. Figure D.1 shows an example of where two separate roots joined together were misinterpreted as a single root with a thicker diameter. The average root diameter from in-box imaging was twice that of scanned washed roots (see Table). Second, when the plant is actively growing, a given length of root may have visible root hairs that can widen the root diameter estimated by RootPainter and RhizoVision Explorer. Those same roots, once washed, may scan as having a narrower diameter Challenges of irrigation treatments The approach we used for imposing WW and WD treatments in our study was mostly based on the methods outlined in Bodner et al. ( 2017 ), and Cassel & Nielsen (1986). Bodner et al. ( 2017 ) measured the PAW under WW and WD conditions specifically for rhizoboxes. Cassel & Nielsen (1986) discussed using a pressure plate to determine the soil's field capacity and permanent wilting point. However, we made modifications to the PAW levels for our treatments over the course of the study. The rhizobox study in Bodner et al. ( 2017 ) had 80% and 40% of PAW for their WW and WD treatments, respectively, and the study was performed in a controlled environment room which offered better control of environmental conditions than we had in the greenhouse. For instance, temperature and vapor pressure deficit values in the controlled environmental room were precisely managed to meet the needs of the WW and WD treatments. (Bodner et al. 2017 ). By contrast, in our greenhouse, we were unable to maintain the temperature and vapor pressure deficit values recommended by Bodner et al. ( 2017 ). In fact, the greenhouse temperature fluctuates. For instance, in run 3, the maximum temperature reached 34°C, with an average of 23.4°C. During our preliminary experiment, 18 days after transplanting, we decreased the PAW from 80% to 40% in the WD treatment, while maintaining the WW at 80% PAW. This resulted in a high mortality rate for the plants in WD treatments due to the high daytime temperatures in the greenhouse and the rapid drying of the upper soil layer led to a sudden onset of severe WD stress. To address this issue, we adjusted the PAW percentage by gradually decreasing it for the WD treatments in runs 1 to 5. Nevertheless, there were plant losses in runs 1 and 3 due to high daytime temperatures. However, no chili pepper plants died in run 2, likely because the rhizoboxes were positioned near the cooling pad. In runs 4 and 5 were conducted in summer with a maximum temperature of 36°C, we used shade cloth to reduce heat stress and adjusted the PAW levels. These changes eliminated plant mortality. By using shade cloth in high temperatures and starting with higher PAW levels before gradually decreasing them, we successfully established healthy seedlings before exposing them to WD stress. The feasibility of using both root phenotyping methods to predict the root biomass Setting up an image station and in-box image analysis pipeline, as well as scanning washed roots, can be labor intensive, so being able to accurately estimate root biomass from TRL would be beneficial. Additionally, no study has assessed whether both phenotyping methods can predict root biomass in a rhizobox. Based on our findings, using dry root weight gives a more accurate estimate of root biomass compared to using root fresh weight when predicting from TRL in root phenotyping studies. Similarly, for cell mass determination, measuring wet weight can be less precise due to liquid in the cell. However, dry weight, provides a more accurate measure by removing all the water content in the cell (Godbey. 2022). In a study on root growth for marigold and zinnia plants in clear cylindrical plexiglass tubes (7.6 cm diameter and 1.8 m tall) found that that there was a strong correlation between the total root length (visible roots) measured from a digital camera and the dry root weight (Judd et al. 2015 ). Although the plant containers of rhizobox and cylindrical plexiglass tube had different designs. Our rhizobox study, along with the study by Judd et al. ( 2015 ), demonstrated a strong relationship between total root length and root dry weight. Besides that, predicting root biomass from TRL measured using in-box imaging had lower R2 than scanning washed roots, possibly because some roots were hidden and unable to be captured by in-box imaging. It was unclear why the prediction of root biomass under WD had lower R2 than the WW condition. It may be due to the smaller root systems in WD. These smaller root systems were more prone to root loss during root washing and the hidden root from in-box imaging, leading to less accurate biomass predictions. In contrast, larger root systems might be less susceptible to root loss, resulting in higher R2 values in WW condition. Best practices for quantifying root traits via rhizoboxes Our use of acrylic panels in the rhizobox design introduced challenges during soil filling, impacting soil distribution. First, we explored the soiling-filling method introduced by Bodner et al. ( 2017 ). The rhizobox used by Bodner et al. ( 2017 ) was 100 cm high x 30 cm wide x 1 cm deep (internally). The back panel was made from 15 mm thick rigid grey PVC, and the front panel was composed of 6 mm thick hard mineral glass. This setup allowed them to effectively fill the rhizobox with soil by opening it and adding soil horizontally to the back panel. While our rhizobox had acrylic panels for specific reasons – lower cost, lighter weight, and higher shatter resistance compared to glass -- it was less rigid than glass. We tested the soil filling method used by Bodner et al. ( 2017 ) with our rhizobox (two 0.9-inch-thick acrylic panels); it resulted in the creation of large air gaps within the soil (Figure D.2). This was because it was difficult to achieve a completely flat soil surface during horizontal filling, especially considering that acrylic panels are less rigid compared to glass. This process appears to induce deformations in the acrylic panels when assembling the rhizobox, ultimately leading to the unintended formation of these air gaps within the soil. To address this problem, we implemented a soil-filling method as described in the methods section, which effectively eliminated air gaps within the soil." (Figure D.3). While the acrylic may have slightly deformed using our method (which can be seen in the variation in actual soil volume of 3100 to 3600 cm³ per box, rather than 3478 cm³, it is impossible to completely avoid deformation due to use of the less rigid acrylic glass. Conclusion In our rhizobox experiment, we successfully developed a comprehensive rhizobox pipeline capable of phenotyping the root architecture system of chile pepper plants, thus eliciting biological responses under both WW and WD treatments. We found that in-box imaging metrics can predict final scan metrics across WW and WD plants for many RSA traits. However, our study also uncovered some limitations. For instance, inbox imaging was not successful at determining average root diameter and TRL of root diameter above 2 mm in WD analysis. Moreover, dry root weight were effective predictors of total root length from the in-box imaging and scanning washed root methods in WW, WD and TC. By employing this affordable and user-friendly pipeline, researchers will be able to readily conduct root experiments using a rhizobox pipeline. Abbreviations 2D : 2-dimensional 3D : 3-dimensional ANOVA : Analysis of variance B_TRL : Total root length from root imaging B_PE : Root perimeter from root imaging B_AVR : Average root diameter from root imaging B_RD 0-0.5 : Total root length for roots with 0 to 0.5 mm diameter from root imaging B_RD 0.5-1 : Total root length for roots with 0.5 to 1 mm diameter from root imaging B_RD 1-2 : Total root length for roots with 1 to 2 mm diameter from root imaging B_RD >2 : Total root length for roots with above 2 mm diameter from root imaging Ca : Ca0045, Costeño Rojo landrace chile pepper accession CC : Equal number of accessions in both treatments per run cm : Centimeter D : Inner diameter between two panels e.u : Experimental unit dpi : dot per inch FC : Field capacity g: Gram H : Height kPa ; Kilopascals PAW : plant available water PWP : Permanent wilting point PI: PI 586666, commercial chile pepper accession RDW : Root dry weight RFW : Root fresh weight RSA : Root system architecture TC : Both treatments combined TRL : Total root length W : Width WD : Water deficit treatment WW : Well-watered treatment X : Predictor variable Y : Response variable Declarations Acknowledgements We would like to express our gratitude to J. McCoy, L. Connolly, H. Scheppler, E. Cigany, K. Fulcher, M. Lincicome and Vivian Bernau for their valuable assistance in carrying out the research and D. Francis for helpful guidance. We thank Mike Anderson and Jackie Rothgery for their assistance in preparing the greenhouse space and managing pest. We are especially grateful to our collaborator from Mexico, L. Jardón Barbolla, for providing the germplasm Ca0045 used in this study. We also acknowledge partial support for this research from the National Institute of Food and Agriculture (AFRI grant 20186701327555), the Center for Applied Plant Sciences at Ohio State University, and the Ohio Agricultural Research and Development Center (OARDC) to KLM and LKM (OSU HCS 25-XX). References Bitarafan, Z., Kaczmarek-Derda, W., Berge, T. W., Tørresen, K. S., & Fløistad, I. S. (2022). Soil steaming to disinfect barnyardgrass-infested soil masses. Weed Technology , 36 (1), 177–185. https://doi.org/10.1017/wet.2021.107 Bodner, G., Alsalem, M., Nakhforoosh, A., Arnold, T., & Leitner, D. (2017). RGB and Spectral Root Imaging for Plant Phenotyping and Physiological Research: Experimental Setup and Imaging Protocols. Journal of Visualized Experiments , 126 , 56251. https://doi.org/10.3791/56251 Bodner, G., Nakhforoosh, A., & Kaul, H.-P. (2015). Management of crop water under drought: A review. Agronomy for Sustainable Development , 35 (2), 401–442. https://doi.org/10.1007/s13593-015-0283-4 Bucksch, A., Burridge, J., York, L. M., Das, A., Nord, E., Weitz, J. S., & Lynch, J. P. (2014). Image-Based High-Throughput Field Phenotyping of Crop Roots. PLANT PHYSIOLOGY , 166 (2), 470–486. https://doi.org/10.1104/pp.114.243519 Cassel, D. K., & Nielsen, D. R. (2018). Field Capacity and Available Water Capacity. In A. Klute (Ed.), SSSA Book Series (pp. 901–926). Soil Science Society of America, American Society of Agronomy. https://doi.org/10.2136/sssabookser5.1.2ed.c36 Chen, X., Zhu, Y., Ding, Y., Pan, R., Shen, W., Yu, X., & Xiong, F. (2021). The relationship between characteristics of root morphology and grain filling in wheat under drought stress. PeerJ , 9 , e12015. https://doi.org/10.7717/peerj.12015 Correa, J., Postma, J. A., Watt, M., & Wojciechowski, T. (2019). Soil compaction and the architectural plasticity of root systems. Journal of Experimental Botany , 70 (21), 6019–6034. https://doi.org/10.1093/jxb/erz383 Crossett, R. N., Campbell, D. J., & Stewart, H. E. (1975). Compensatory growth in cereal root systems. Plant and Soil , 42 (3), 673–683. https://doi.org/10.1007/BF00009951 Dayoub, E., Lamichhane, J. R., Schoving, C., Debaeke, P., & Maury, P. (2021). Early-Stage Phenotyping of Root Traits Provides Insights into the Drought Tolerance Level of Soybean Cultivars. Agronomy , 11 (1), 188. https://doi.org/10.3390/agronomy11010188 El-Soda, M., Malosetti, M., Zwaan, B. J., Koornneef, M., & Aarts, M. G. M. (2014). Genotype × environment interaction QTL mapping in plants: Lessons from Arabidopsis. Trends in Plant Science , 19 (6), 390–398. https://doi.org/10.1016/j.tplants.2014.01.001 Foyer, C. H., Lelandais, M., & Kunert, K. J. (1994). Photooxidative stress in plants. Physiologia Plantarum , 92 (4), 696–717. https://doi.org/10.1111/j.1399-3054.1994.tb03042.x Judd, L. A., Jackson, B. E., & Fonteno, W. C. (2015). Rhizometer: An Apparatus to Observe and Measure Root Growth and Its Effect on Container Substrate Physical Properties Over Time. HortScience , 50 (2), 288–294. https://doi.org/10.21273/HORTSCI.50.2.288 Hirt, H., & Shinozaki, K. (2003). Plant responses to abiotic stress (Vol. 4). Springer Science & Business Media.Kuijken, R. C. P., Van Eeuwijk, Fred. A., Marcelis, L. F. M., & Bouwmeester, H. J. (2015). Root phenotyping: From component trait in the lab to breeding: Table 1. Journal of Experimental Botany , 66 (18), 5389–5401. https://doi.org/10.1093/jxb/erv239 Kücke, M., Schmid, H., & Spiess, A. (1995). A comparison of four methods for measuring roots of field crops in three contrasting soils. Plant and Soil , 172 (1), 63–71. https://doi.org/10.1007/BF00020860 Lesmes-Vesga, R. A., Cano, L. M., Ritenour, M. A., Sarkhosh, A., Chaparro, J. X., & Rossi, L. (2022). Rhizoboxes as Rapid Tools for the Study of Root Systems of Prunus Seedlings. Plants , 11 (16), 2081. https://doi.org/10.3390/plants11162081 Li, A., Zhu, L., Xu, W., Liu, L., & Teng, G. (2022). Recent advances in methods for in situ root phenotyping. PeerJ , 10 , e13638. https://doi.org/10.7717/peerj.13638 Li, L., Zhang, Q., & Huang, D. (2014). A review of imaging techniques for plant phenotyping. Sensors (Basel, Switzerland) , 14 (11), 20078–20111. https://doi.org/10.3390/s141120078 López-Serrano, L., Canet-Sanchis, G., Vuletin Selak, G., Penella, C., San Bautista, A., López-Galarza, S., & Calatayud, Á. (2019). Pepper Rootstock and Scion Physiological Responses Under Drought Stress. Frontiers in Plant Science , 10 , 38. https://doi.org/10.3389/fpls.2019.00038 Mašková, T., & Klimeš, A. (2020). The Effect of Rhizoboxes on Plant Growth and Root: Shoot Biomass Partitioning. Frontiers in Plant Science , 10 , 1693. https://doi.org/10.3389/fpls.2019.01693 McCoy, J. E., McHale, L. K., Kantar, M., Jardón-Barbolla, L., & Mercer, K. L. (2022). Environment of origin and domestication affect morphological, physiological, and agronomic response to water deficit in chile pepper (Capsicum sp.). PLOS ONE , 17 (6), e0260684. https://doi.org/10.1371/journal.pone.0260684 National Drought Mitigation Center. (n.d.). Types of drought. Retrieved October 21, 2023, from https://drought.unl.edu/Education/DroughtIn-depth/TypesofDrought.aspx Osmolovskaya, N., Shumilina, J., Kim, A., Didio, A., Grishina, T., Bilova, T., Keltsieva, O. A., Zhukov, V., Tikhonovich, I., Tarakhovskaya, E., Frolov, A., & Wessjohann, L. A. (2018). Methodology of Drought Stress Research: Experimental Setup and Physiological Characterization. International Journal of Molecular Sciences , 19 (12), 4089. https://doi.org/10.3390/ijms19124089 Paez-Garcia, A., Motes, C., Scheible, W.-R., Chen, R., Blancaflor, E., & Monteros, M. (2015). Root Traits and Phenotyping Strategies for Plant Improvement. Plants , 4 (2), 334–355. https://doi.org/10.3390/plants4020334 Passioura, J. B. (2006). Viewpoint: The perils of pot experiments. Functional Plant Biology , 33 (12), 1075. https://doi.org/10.1071/FP06223 Schmidt, J. E., Lowry, C., & Gaudin, A. C. M. (2018). An Optimized Rhizobox Protocol to Visualize Root Growth and Responsiveness to Localized Nutrients. Journal of Visualized Experiments , 140 , 58674. https://doi.org/10.3791/58674 Schneider, H. M., & Lynch, J. P. (2020). Should Root Plasticity Be a Crop Breeding Target? Frontiers in Plant Science , 11 , 546. https://doi.org/10.3389/fpls.2020.00546 Seethepalli, A., & York, L. M. (2020). RhizoVision Explorer—Interactive software for generalized root image analysis designed for everyone (2.0.2) [Computer software]. Zenodo. https://doi.org/10.5281/ZENODO.4095629 Seleiman, M. F., Al-Suhaibani, N., Ali, N., Akmal, M., Alotaibi, M., Refay, Y., Dindaroglu, T., Abdul-Wajid, H. H., & Battaglia, M. L. (2021). Drought Stress Impacts on Plants and Different Approaches to Alleviate Its Adverse Effects. Plants , 10 (2), 259. https://doi.org/10.3390/plants10020259 Smirnoff, N. (1993). The role of active oxygen in the response of plants to water deficit and desiccation. New Phytologist , 125 (1), 27–58. https://doi.org/10.1111/j.1469-8137.1993.tb03863.x Smith, A. G., Han, E., Petersen, J., Olsen, N. A. F., Giese, C., Athmann, M., Dresbøll, D. B., & Thorup‐Kristensen, K. (2022). R oot P ainter: Deep learning segmentation of biological images with corrective annotation. New Phytologist , 236 (2), 774–791. https://doi.org/10.1111/nph.18387 Sprunger, C. D., Culman, S. W., Robertson, G. P., & Snapp, S. S. (2018). Perennial grain on a Midwest Alfisol shows no sign of early soil carbon gain. Renewable Agriculture and Food Systems , 33 (4), 360–372. https://doi.org/10.1017/S1742170517000138 Sultan, S. E. (2000). Phenotypic plasticity for plant development, function and life history. Trends in Plant Science , 5 (12), 537–542. https://doi.org/10.1016/S1360-1385(00)017970 Takahashi, H., & Pradal, C. (2021). Root phenotyping: important and minimum information required for root modeling in crop plants. Breeding science , 71 (1), 109–116. https://doi.org/10.1270/jsbbs.20126 Uga, Y., Okuno, K., & Yano, M. (2011). Dro1, a major QTL involved in deep rooting of rice under upland field conditions. Journal of Experimental Botany , 62 (8), 2485–2494. https://doi.org/10.1093/jxb/erq429 USDA, National Agricultural Statistics Service. (2023). Vegetables 2022 summary (February 2023) (No. 81). Retrieved October 21, 2023, from https://downloads.usda.library.cornell.edu/usda-esmis/files/02870v86p/hq37x121v/4b29ck28c/vegean23.pdf Via, S., Gomulkiewicz, R., De Jong, G., Scheiner, S. M., Schlichting, C. D., & Van Tienderen, P. H. (1995). Adaptive phenotypic plasticity: Consensus and controversy. Trends in Ecology & Evolution , 10 (5), 212–217. https://doi.org/10.1016/S0169-5347(00)89061-8\ Wajhat-Un-Nisa, Sandhu, S., Ranjan, R. et al. Root plasticity: an effective selection technique for identification of drought tolerant maize (Zea mays L.) inbred lines. Sci Rep 13, 5501 (2023). https://doi.org/10.1038/s41598-023-31523-w Zhang, Y., Zhang, W., Cao, Q., Zheng, X., Yang, J., Xue, T., Sun, W., Du, X., Wang, L., Wang, J., Zhao, F., Xiang, F., & Li, S. (2022). WinRoots: A High-Throughput Cultivation and Phenotyping System for Plant Phenomics Studies Under Soil Stress. Frontiers in Plant Science , 12 , 794020. https://doi.org/10.3389/fpls.2021. Additional Declarations No competing interests reported. Supplementary Files chap2Supplementarymaterials.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7697278","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Method Article","associatedPublications":[],"authors":[{"id":519763989,"identity":"7b332953-57db-4fc3-acd2-f50eac4324b3","order_by":0,"name":"Chee Gang Ngui","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABG0lEQVRIiWNgGAWjYJCCAwwVNgkMDGwQHhs7QhyPljNpSFqYidDCwNh2GKGFgZAWc/41hgcY2M7nGRw/lviYp2KbPB8zA+PDn20Mcnw3ErBqsZzxxuAAA8/tYoMzaYeNec7cNmxjZmA25m1jMJbEocXgxhmgFonbiRtusLdJ87bdZgRqYZNmbGMAiuDTYnAOpKX9N++/2/YgLZJAh9Xj1HK+B6gl4QBQC9sxZt6G24kgLRJAhyUY4LSFreBAwoHkYskzacmSc47dTm5jZmw25jknYTjzzAMcthze/OHjP7s8vuPHDD+8qbltO7+9+eDDH2U28nzHsdvCIAEUh0kx8YApxgaQOHblIMB/AMFm/IFb3SgYBaNgFIxgAAAqA2Tw0RlY0AAAAABJRU5ErkJggg==","orcid":"","institution":"The Ohio State University","correspondingAuthor":true,"prefix":"","firstName":"Chee","middleName":"Gang","lastName":"Ngui","suffix":""},{"id":519763991,"identity":"2e9ac55e-1f0d-4d55-b906-1c3d20a361f6","order_by":1,"name":"Leah McHale","email":"","orcid":"","institution":"The Ohio State University","correspondingAuthor":false,"prefix":"","firstName":"Leah","middleName":"","lastName":"McHale","suffix":""},{"id":519763992,"identity":"c4af6c31-f87c-47ec-9ae4-6702fa5c7b29","order_by":2,"name":"Christine Sprunger","email":"","orcid":"","institution":"Michigan State University","correspondingAuthor":false,"prefix":"","firstName":"Christine","middleName":"","lastName":"Sprunger","suffix":""},{"id":519763993,"identity":"29e95333-8b12-43f8-a54f-439e3525dbf8","order_by":3,"name":"Kristin L. Mercer","email":"","orcid":"","institution":"The Ohio State University","correspondingAuthor":false,"prefix":"","firstName":"Kristin","middleName":"L.","lastName":"Mercer","suffix":""}],"badges":[],"createdAt":"2025-09-23 19:08:28","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7697278/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7697278/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":92845440,"identity":"c65a5cca-5903-45ea-956a-177e20803a1a","added_by":"auto","created_at":"2025-10-06 09:27:09","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8176,"visible":true,"origin":"","legend":"","description":"","filename":"26ea2069bfcd45afb5de687882b3b22c.json","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/644b978cbb4c03e4b7c70e87.json"},{"id":92845444,"identity":"b4a38bf6-ea96-4c0e-ae65-63ae889570e2","added_by":"auto","created_at":"2025-10-06 09:27:10","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4454573,"visible":true,"origin":"","legend":"","description":"","filename":"chap2Supplementarymaterials.docx","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/b7e94847f27ac293d486eb63.docx"},{"id":92845445,"identity":"933bd4fb-ac4b-45c6-bcb3-a1673ec8b507","added_by":"auto","created_at":"2025-10-06 09:27:10","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":136097,"visible":true,"origin":"","legend":"","description":"","filename":"26ea2069bfcd45afb5de687882b3b22c1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/22b8f61d955a45cff0bdef65.xml"},{"id":92845833,"identity":"3a9d4771-12d1-4d1b-a83c-5689fbe65405","added_by":"auto","created_at":"2025-10-06 09:35:10","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2528396,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/272b327e9b52a3c99acbc702.png"},{"id":92845453,"identity":"315d43bf-4bb8-4055-a577-0ff1bbbae915","added_by":"auto","created_at":"2025-10-06 09:27:10","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1195617,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/1c2714e79857a971b4a8da3f.png"},{"id":92845452,"identity":"012258e0-e5e2-42a7-83a5-1e4c19df0162","added_by":"auto","created_at":"2025-10-06 09:27:10","extension":"emf","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1252916,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.emf","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/7d9ebbc2b857cd5210467c3b.emf"},{"id":92845456,"identity":"41af3b21-b7ec-46dd-a273-3a5441c0f07d","added_by":"auto","created_at":"2025-10-06 09:27:10","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":224069,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/b9321fe53082783d9314fecd.png"},{"id":92845460,"identity":"38717189-22db-45b5-b208-84cea4e27880","added_by":"auto","created_at":"2025-10-06 09:27:10","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":155217,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/5ce67506a8ab3907150574f8.png"},{"id":92845835,"identity":"cdc3380f-157e-4bdb-8677-79fcbd3567a5","added_by":"auto","created_at":"2025-10-06 09:35:10","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":445913,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/6abb4ebd1a27303111da4071.png"},{"id":92845457,"identity":"350efa10-a5e8-4827-bf32-33142558f3d6","added_by":"auto","created_at":"2025-10-06 09:27:10","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":161701,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/abe262e503d3898b1feef574.png"},{"id":92845447,"identity":"0c48c4ac-54da-4515-8a1d-2fc7a1234408","added_by":"auto","created_at":"2025-10-06 09:27:10","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":14279,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/95ad47899036c0dfe10f6698.png"},{"id":92845449,"identity":"50d15bb3-8ca3-4e91-ad40-b7a8422ac00e","added_by":"auto","created_at":"2025-10-06 09:27:10","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":50847,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/1ddc55d783b16724f6c6509b.png"},{"id":92845454,"identity":"9ebedbe9-09b6-4475-9454-16921c4dade9","added_by":"auto","created_at":"2025-10-06 09:27:10","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":34586,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/734f2a2011b2e6f69613c0d1.png"},{"id":92845458,"identity":"e64efd2b-1293-4c29-8bf3-ab0e779b713b","added_by":"auto","created_at":"2025-10-06 09:27:10","extension":"xml","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":132171,"visible":true,"origin":"","legend":"","description":"","filename":"26ea2069bfcd45afb5de687882b3b22c1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/9d37621a120975853591e88f.xml"},{"id":92845455,"identity":"bfc6b5e2-4cc8-42b5-b83d-ef6bb8dd7dad","added_by":"auto","created_at":"2025-10-06 09:27:10","extension":"html","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":143081,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/f86737724045ef1c81d6262a.html"},{"id":92845439,"identity":"101ca3ea-6cbe-4492-a6df-3805526cf7d0","added_by":"auto","created_at":"2025-10-06 09:27:09","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":976920,"visible":true,"origin":"","legend":"\u003cp\u003e(a) and (b) rhizobox wrapped with a white and black poly film cover secured with a hook-and-loop fastener. (c) rhizobox opening area was covered with a poly film, leaving a small opening. (d) rhizboxes in the greenhouse. (e) Best time for transplanting. (f) the plant stages. (g) rhizobox holder with 45-degree inclination. (h) Large air gap in soil without using our method. (i) No air gap by filling up the soil with our method. (j) Conjoined roots mistaken for single, thicker root.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/2c5db4773fcaf88bfe9287b8.png"},{"id":92845830,"identity":"eedb7c95-78ce-46c6-830b-51ec50f1d7e0","added_by":"auto","created_at":"2025-10-06 09:35:09","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":495043,"visible":true,"origin":"","legend":"\u003cp\u003e(a) Workflow for in-box imaging phenotyping method. (b) Workflow for scanning washed root phenotyping method.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/737dd00afa07bcc4cee1e485.png"},{"id":92845448,"identity":"bf6601da-e68c-4a2d-80cf-3487aa933148","added_by":"auto","created_at":"2025-10-06 09:27:10","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":144911,"visible":true,"origin":"","legend":"\u003cp\u003eLog transformed least squares means of total root length from in-box imaging for individual day and irrigation treatment combinations (± se) from repeated measurement analysis. Means sharing a letter are not significantly different employing a Tukey’s HSD Test (P=0.05).\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/43b096dab6c6e8287ea02bfa.png"},{"id":92845831,"identity":"1d22f262-8ae0-4aea-8d0c-c0156e922337","added_by":"auto","created_at":"2025-10-06 09:35:10","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":193539,"visible":true,"origin":"","legend":"\u003cp\u003eLinear regression analysis for seven root traits with their regression equation from in-box imaging and scanning washed root with their regression equation. Y represented response variable and predictor denoted as X. The blue dots, line, and confidence interval show data from the well-watered treatment and the red from the water deficit treatment.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/48b4aaa07976ce0bb318bf75.png"},{"id":92845832,"identity":"9e188cf8-49c9-4fa0-bb44-04cfa4bf1a43","added_by":"auto","created_at":"2025-10-06 09:35:10","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":216342,"visible":true,"origin":"","legend":"\u003cp\u003eLinear regression analysis for total root length in predicting values from root phenotyping methods from root weight with their regression equation. Y represented response variable and predictor denoted as X. The blue dots, line, and confidence interval show data from the well-watered treatment, the red from the water deficit treatment and the black from both treatments combined.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/4fc3cf7fb49fb856aac0bc17.png"},{"id":94473573,"identity":"c2f26814-33d4-487a-ba28-ce0d04c716ef","added_by":"auto","created_at":"2025-10-27 15:44:52","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2861370,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/01b81891-c120-47d9-b6df-5b729d9c5858.pdf"},{"id":92845443,"identity":"8c063c82-8e24-4c6b-b93c-8ee011134df6","added_by":"auto","created_at":"2025-10-06 09:27:10","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":4454573,"visible":true,"origin":"","legend":"","description":"","filename":"chap2Supplementarymaterials.docx","url":"https://assets-eu.researchsquare.com/files/rs-7697278/v1/949550a9cf313b9abfe7394a.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"A comprehensive rhizobox pipeline for analyzing pepper root system architecture under well-watered and water deficit conditions","fulltext":[{"header":"Introduction","content":"\u003cp\u003eWhen experiencing drought, plant function can be severely disrupted, resulting in stunted growth, wilting, increased vulnerability to disease infection, and yield reduction (Seleiman et al., 2022). Monitoring belowground responses to drought could be critical for overall plan resilience. For instance, plant roots are the loci for plant-soil interactions, take up water and nutrients for plant development, and anchor the plant. Root system architecture (RSA) is the physical arrangement and traits of roots within the soil. RSA can be influenced by the allocation of water and nutrients, altering the development of lateral roots, elongation of root tips, and biomass allocation (Correa et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Phenotyping plant response under water deficit and identifying plant traits that contribute to sustaining productivity in water-limited environments are essential for genetic improvement of crop plants (Li., et al., 2014). Measuring RSA traits in plants grown under a range of conditions can help us evaluate genetic resources for important traits, such as drought tolerance. Yet research on improving plant root systems is relatively scarce compared to that of aboveground traits due to the difficulty of capturing RSA measurements (Takahashi \u0026amp; Pradal, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eRoot studies can be classified by their use of destructive and non-destructive phenotyping (Li et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Destructive phenotyping involves collecting the root system from the soil. The drawbacks include the time-consuming nature of the methods, the fact that you can only phenotype the root at a single timepoint. Non-destructive phenotyping allows for assessment of the same traits on the same plant over time (i.e., without destructive harvesting). In non-field-based root growing methods, soil media (like pots, rhizoboxes, or rhizotrons) or soil-less media (like agar, pouches, and hydroponic systems) can be used for both destructive and non-destructive phenotyping (Li et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Non-field-based root-growing methods can simplify root phenotyping; however, they may fail to replicate natural conditions, especially when small or irregularly shaped containers are used. Such setups often create uneven abiotic conditions, particularly in the deeper soil layers. (Maškov\u0026aacute; \u0026amp; Klimeš, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Li et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Paez-Garcia et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Kuijken et al., 2015). Moreover, most of these non-field-based methods, when used non-destructively, allow for the application of high throughput root phenotyping techniques and measurements of RSA over time (Li et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e\u003cp\u003ePhenotyping techniques for RSA can be separated into systems that allow for two-dimensional (2D) and three-dimensional (3D) views of the root system. The 2D techniques mainly use a camera or flatbed scanner to visualize the root system, either non-destructively in transparent containers or from destructively harvested, washed roots (Li et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The main advantage of 2D root phenotyping techniques is that the root imaging is affordable and simple to establish compared to 3D root phenotyping techniques, such as X-ray computed tomography, magnetic resonance imaging, electrical impedance tomography, neutron tomography, and ground penetrating radar, which are expensive. Moreover, it can be challenging to develop the technology in individual laboratories (Li et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eStudies employing rhizoboxes use soil as a medium and employ non-destructive 2D root phenotyping. The advantage of a rhizobox is that it allows us to observe the root growth in a soil environment over time by tracing or imaging roots through the transparent side of the box as the roots grow. One disadvantage is that rhizoboxes are constructed to have very little space between front and back acrylic panels, which can constrain natural root growth. Nevertheless, Maškov\u0026aacute; and Klimeš (\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) reported no significant difference between the total root biomass of 4-week-old plants grown in flat pots (rhizobox, 19.5cm height x 15 cm width x 1 cm inner diameter) and a more typical square pot (upper size 7 \u0026times; 7 cm, bottom size 5 \u0026times; 5 cm, height 8 cm); they did suggested that researchers be cautious of the limitations of rhizoboxes on natural root growth (Maškov\u0026aacute; and Klimeš, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). In the past, limitations of 2D root phenotyping with soil media in rhizoboxes included the lack of high-resolution root images and low accuracy of root segmentation (i.e., the process used to separate roots from the surrounding soil in the image). This was primarily due to the expense of high-quality cameras and the underdeveloped state of segmentation software (Li et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Such technical difficulties can now be overcome, with affordable cameras achieving a resolution of 9152 x 6944 pixels and high-quality root segmentation achieved using robust graphical user interface-based software such as RootPainter (Smith et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eRhizobox experiments under well-watered (WW) and water deficit (WD) conditions are few, but have been used to develop phenotyping methods with sugar beet, (Bodner et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2017\u003c/span\u003e ;Rambla et al, 2023) and to measure the responses of root traits, such as total root length, root dry weight, and root-to-shoot ratio, in soybeans grown in minirhizotrons (24.5 \u0026times; 24.5 \u0026times; 2.5 cm) (Dayoub et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Surprisingly, in this latter study, they found that there was no effect of WD (Dayoub et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Thus, we still lack a clear understanding of how WD manipulations in rhizoboxes may influence RSA traits via non-destructive methodology, especially in vegetable crops. By growing two distinct chile pepper accessions in relatively large rhizoboxes (53.5 cm (W) x 78.5 cm (H) x 1 cm (D)), under WW and WD conditions, we were able to non-destructively assess 2D RSA and compare those results to destructive assessments of 2D RSA at the end of the experiment. We had three objectives.\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003e1) To measure biological responses in RSA traits to our WW and WD treatments over the course of the experiment.\u003c/p\u003e\u003cp\u003e2) To determine the degree to which non-destructive in-box 2D imaging of RSA traits can predict and possibly replace 2D scans of destructively harvested and washed roots.\u003c/p\u003e\u003cp\u003e3) To assess the feasibility of utilizing traits, such as total root length (TRL) measured through non-destructive in-box 2D imaging methods to predict destructively harvested root biomass.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eWith these analyses, we aimed to provide guidance for researchers trying to improve and streamline their phenotyping pipeline without sacrificing data quality.\u003c/p\u003e"},{"header":"Method and materials","content":"\u003cp\u003e\u003cem\u003ePlant materials \u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eWe selected two distinct chile pepper accessions, Ca0045 and PI586666, as our model plants for the rhizobox study due to their differences in aboveground characteristics and origin location (McCoy et al., 2022). PI586666 is a commercial accession with a New Mexican pod type provided by the Plant Genetic Resources Conservation Unit of the USDA, Griffin, Georgia. Ca0045 is a Coste\u0026ntilde;o Rojo landrace accession collected from a plantation along the coast of the Mexican state of Oaxaca (personal communication, L. Jard\u0026oacute;n Barbolla). All the seeds for accessions were grown in the greenhouse, Columbus, Ohio. \u003c/p\u003e\n\u003cp\u003e\u003cem\u003eSeed sterilization, germination, transplantation, and growing conditions \u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eWe sterilized seeds with 0.825 % of sodium hydrochloride for ten minutes and rinsed with distilled water. We germinated the seeds in sanitized 11 cm x 11 cm square germination boxes with 75 g of autoclaved sand (Quikrete Co., Atlanta, GA, USA) and 15 ml of distilled water. We placed the seeds in the Conviron G30 germination chamber (Conviron, Winnipeg, Manitoba, Canada) for seven days. Germinated seeds were ready to transplant into rhizoboxes once the hypocotyl of the seedling turned green and the radicle reached 2 to 3 cm with the majority of the cotyledon part was enclosed by the seed coat (Figure 1e and f), then we transplanted the seedling into the soil-filled rhizobox. We found that earlier or later transplanting reduced the rate of seedling emergence from the soil. Normally, seedlings of both accessions emerged from the soil within 3 to 4 days. We applied 0.8 grams of slow-release 14 (nitrogen) -14 (phosphorus)-14 (potassium) fertilizer (Osmocote, Scotts Miracle-Gro, Marysville, OH, USA) to the soil surface after transplanting was completed. We grew the seedlings in rhizoboxes in the Howlett Greenhouse at the Ohio State University in Columbus, Ohio. The average and maximum temperature per run differed (see Supplementary Material 2), depending on the time of year and the use of shade cloth. For instance, in the summer (i.e., runs 4 and 5), the average and maximum temperature were 24\u0026deg;C and 36\u0026deg;C, respectively, so we placed shade cloth over the rhizoboxes to reduce direct heat. \u003c/p\u003e\n\u003cp\u003eThe rhizobox size was 53.5 cm (W) x 78.5 cm (H) x 1 cm (D), while the soil filling volume was approximately 47 cm (W) x 74 cm (H) x 1 cm (D). The description of building these rhizoboxes can be found in Supplementary Material 4. The rhizobox was supported by a customized white and black poly cover (Figure 1a). Each rhizobox was placed on a greenhouse bench with a custom PVC rhizobox holder based on the design suggested in Schmidt et al. (2018), but larger (Figure 1d and g). We kept the rhizoboxes at a 45\u0026deg; angle to maximize root visibility through the back panel (Bontpart et al. 2020). After transplanting, the exposed soil was covered by a white poly film to decrease evaporation (i.e., maintain plant available water in the soil) and allow us to achieve a long-term gradual reduction in water availability (Figure 1b) (Osmolovskaya et al., 2018). Once the seedling emerged, we opened a small area in the middle of the white poly film to allow it to grow toward the light. The experiment lasted for a duration ranging from 54 to 61 days.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eSoil preparation \u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eWe used a soil mixture of 70% topsoil and 30% dark grey mason sand (by volume) in the rhizoboxes. First, we sifted the topsoil (Mr. Mulch, Columbus, OH) through a 0.02mm test sieve to remove large particles\u0026mdash;a slow process (40 minutes for 5-gallons). Fine mason sand (smaller than 0.02mm in diameter) was most suitable since its dark grey color facilitated root segmentation. The sand and sifted soil were autoclaved separately in plastic bags to steam for an hour to kill weed seeds. The autoclave took 55 minutes to reach 200\u0026deg;F and the soil was exposed to a constant 200\u0026deg;F for 5 minutes (Bitarafan et al., 2022) and then air-dried in an open pan. \u003c/p\u003e\n\u003cp\u003e\u003cem\u003eExperimental design \u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eWe ran a factorial experiment assessing the effects of chile pepper accession and irrigation level in factorial combination. We performed multiple runs over time; within each run, we had one or multiple replicates separated into spatial blocks. Thus, in the entire experiment, the total number of runs was five, which were divided into seven blocks (Supplementary Material 2). Within each block, each rhizobox was an experimental unit since a rhizobox was the scale at which we applied one replicate of the factorial combinations of accession and irrigation treatment. The high levels of mortality in some runs and for some treatment combinations resulted in uneven representation across blocks. The irrigation application in runs four and five resulted in no plant mortality so, by the final runs, our methodology stabilized. However, due to lack of WD treatment representation across blocks, block 7 was assigned with WD. As described below, we included blocks in our analyses when using the ANOVA approach, but not when we used the regression approach. \u003c/p\u003e\n\u003cp\u003e\u003cem\u003eData Collection\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe data was collected from scanned washed roots collected at the end of the experiment, as well as in-box imaging root traits. The experiment was terminated 61 days after transplanting in run 1 but 55 days after transplanting in the other runs. For objective 1, repeated measures analyses of TRL captured by in-box imagining was analyzed from 25 days after transplanting to day 55. The image of the root was taken every Monday, Wednesday, and Friday. In the fifth run, only the WD treatment was applied since there was less representation of WD treatments in runs 3 and 4.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eManipulation of water and soil needed for the water deficit and well water treatments. \u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eIn this section, we incorporated Bodner et al. (2017), Passioura (2006) and Cassel \u0026amp; Nielsen (1986) methods to comprehensively measure soil and water requirements to implement WW and WD effects on the plants. \u003c/p\u003e\n\u003cp\u003eFirst, we determined the weight of the topsoil and mason sand combination by adding the air-dried topsoil into a beaker until it reached 350 ml and measured the weight. We did the same method for mason sand until it reached150 ml. Based on the weight, we knew the weight of 70% topsoil and 30% mason sand by volume. Next, we measured soil bulk density (Bodner et al., 2017). Subsequently, we determined field capacity (FC) and permanent wilting point (PWP) of our soil. To do this, we used a pressure plate device as described in Supplementary Material 1\u003cstrong\u003e.\u003c/strong\u003e Typically, FC is measured at a water potential of 33 kilopascals (kPa). However, in our study, we used a lower water content at a matrix potential of 8 kpa due to the height of our rhizobox (Passioura, 2006). Finally, we determined the plant available water (PAW) for the soil mixture for the WW and WD treatments (Bodner et al., 2017). The full procedure is described in Supplementary Material 1.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eSoil filling and irrigation treatments adjustment for well-watered and water deficit treatments \u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eTo fill the rhizobox with soil, we began by measuring the initial weight of the empty rhizobox on a floor scale. The first weight of soil was based on soil volume (47 cm \u0026times; 74 cm \u0026times; 1 cm) with a balance scale. We then mixed mason sand, topsoil and a specified volume of water in a bucket. The volume of water was based on the PAW of irrigation treatments. Adding water at this stage prevented the separation of mason sand and topsoil, ensuring a homogeneous mixture. Using a funnel, we carefully added the homogenized soil mixture into the upright rhizobox, gently swaying the rhizobox to ensure the soil settled uniformly and firmly. Due to bowing of the acrylic side panels, we added extra of soil mixture and water (the water added was based on the weight of the soil and treatment) to ensure a uniform height of 74 cm in each rhizobox. Therefore, the final weight of soil and water can vary slightly for each rhizobox. The process of adding the soil into each rhizobox took approximately 45 minutes. \u003c/p\u003e\n\u003cp\u003eWe used the concept of PAW to determine the exact amount of water required for each treatment, with slight adjustments to the percentage of PAW across different blocks or runs (Supplementary Material 3). This adjustment was necessary because young chile pepper seedlings in rhizoboxes showed high sensitivity to WD stress. After transplanting, we added water to the opening section of the rhizobox. In WW, the percentage of PAW for all the runs were increased after transplant, except run 1, which remained the same. The percentage of PAW in WW treatment was maintained until the experiment ended. In WD treatment, the PAW of all the runs were elevated after transplant except for run 1. Then, starting at day 18, we reduced the PAW percentage. Every 7 days, we reduced the PAW 10% until it reached 40%, which we then maintained until the experiment ended (Supplementary Material 3). Runs 1, 2, 3, 4, and 5 had 33%, 0%, 50%, 0% and 0% plant mortality, respectively (Supplementary Material 2). \u003c/p\u003e\n\u003cp\u003e\u003cem\u003eIn-box root imagining \u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eIn the root image-capturing process, we used the 64MP Hawkeye camera (Arducam, China) mounted with a Raspberry Pi 4 Model B (Raspberry PI, United Kingdom) used as a low-cost computer to produce high-quality root images with 9152 x 6944 pixels. The camera hardware and software setup were described in the user guide (https://docs.arducam.com/Raspberry-Pi-Camera/Native-camera/PiCamera2-User-Guide/). We took the root images at an image station we constructed to ensure all rhizoboxes at all time points were imaged in the same position, using the same light source and quantity (Refer to Supplementary Material 5 for image station structure.). We captured root images from both the front and back panels of the rhizoboxes; more roots were visible from the back panel, which faced down. \u003c/p\u003e\n\u003cp\u003eTo process the root images, we used two free software tools. The first was RootPainter (Smith et al., 2022), which was designed for root and soil segmentation and allowed us to obtain the image of the isolated root system from the rhizobox for further root analysis. We used 530 root images captured at various developmental stages of chile pepper in both irrigation treatments as a training data set to generate the final best model to extract the root image for further analysis. We used two methods to visualize the accuracy of root and soil segmentation. We checked the model\u0026rsquo;s accuracy using the S\u0026oslash;rensen\u0026ndash;Dice coefficient (Zou et al., 2004) in the metrics plot of RootPainter. We also compared the raw and segmented images by eye to determine the accuracy of the segmented image relative to the actual roots. We then used RhizoVision Explorer (Seethepalli and York, 2022) to analyze root images and quantifying RSA traits. Before conducting the analysis, we described the procedure for measuring the root image\u0026rsquo;s pixels per millimeter in Supplementary Material 7. \u003c/p\u003e\n\u003cp\u003eThe root traits extracted by RhizoVision Explorer from our in-box images and that we used for analyses related to objectives 1-3 were total root length (B_TRL, cm); root perimeter (B_PE, cm), cm; average root diameter (B_AVR); total root length for roots with 0 to 0.5 mm diameter (B_RD 0-0.5, cm); total root length for roots with 0.5 to 1 mm diameter (B_RD 0.5-1, cm); total root length for roots with 1 to 2 mm diameter (B_RD 1-2, cm) and total root length for roots with above 2 mm diameter (B_RD \u0026gt;2, cm). All the features in RhizoVision Explorer are described in the manual (Seethepalli \u0026amp; York 2020). Root traits such as B_TRL and B_PE from the front and back panels of each root sample were subsequently combined (e.g., Total root length in sample 1 = TRL, front panel + TRL, back pan\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eRoot washing and scanning washed roots\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAt the end of the experiment, we washed the roots using methods from Sprunger et al. (2018), losing few roots. First, we placed a 1 mm mesh mat on the drain to capture all the broken roots. We then laid the rhizobox on top of two bars and used a garden hose nozzle to wash away the soil (Figure 2b). Once the roots were clean, we removed the shoot biomass. We cut multiple of the secondary roots to separate the root system into sections to facilitate the scanning process. Then we placed the root sections in ziplock bags in the refrigerator and sprayed with deionized water to prevent them from drying out prior to scanning. To scan the roots, we submerged them in a water-filled tray to untangle the roots and scanned them using an EPSON Expression 10000 XL scanner (Epson, Japan) connected to the computer (Figure 2b). Winrhizo software (Regent Instruments Inc, Canada) was utilized for capturing the scanned root image. Prior to scanning, we configured the settings in Winrhizo, setting the resolution at 400 dot per inch (dpi) and defining the image acquisition parameters with a width 30 cm and a length 40 cm. Once the scanned root image acquired, it loaded to RhizoVision Explorer (Seethepalli \u0026amp; York 2020). \u003c/p\u003e\n\u003cp\u003eBefore conducting the root analysis, we performed various settings for image preprocessing and feature extraction. In the image preprocessing phase, we selected the \u0026ldquo;broken root\u0026rdquo; analysis mode since the scanned whole root system appeared in disconnected pieces. We used a resolution of 400 dpi and set the image thresholding level to 200. The \u0026ldquo;filter non-root object\u0026rdquo; was set to 0.5, we left all other boxes unticked. For feature extraction, we applied a root pruning threshold of 1 and set the roots into four different diameter ranges. It is important to note that the specific settings in RhizoVision Explorer may vary depending on factors such as the resolution of the root image, the density of the roots, and the presence of debris (soil, organic matter and unwanted materials). The scanned washed root traits were the following: total root length (S_TRL, cm); root perimeter (S_PE, cm), cm; average root diameter (S_AVR); total root length with 0 to 0.5 mm diameter (S_RD 0-0.5, cm); total root length with 0.5 to 1 mm diameter (S_RD 0.5-1, cm); total root length with 1 to 2 mm diameter (S_RD 1-2, cm) and total root length with above 2 mm diameter (S_RD \u0026gt;2, cm) . \u003c/p\u003e\n\u003cp\u003e\u003cem\u003eRoot biomass\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eOnce roots were scanned, we removed the excess water on the roots with a paper towel and measured the root fresh weight (RFW, g). Then, we obtained root dry weight (RDW, g) by placing them in an oven at 65\u0026deg;C for three days and reweighing (L\u0026oacute;pez-Serrano et al. 2019).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eStatistical analyses\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAll statistical analyses were performed in SAS software (version 9.4) (SAS Institute Inc., 2013). All general linear mixed model analyses were checked for the assumptions of normal distribution of residuals, constant variance of residuals, and independence of errors, as well as the presence of outliers. Similarly, for all linear regression analyses, we checked the assumptions of linearity between predictor and response variables, normal distribution of residuals, constant variance of residuals, independence of errors, the presence of outliers, and the influence of leverage (extreme values of predictor variables). \u003c/p\u003e\n\u003cp\u003eWe performed general linear model analyses for all three objectives. For objective one, we took a mixed model ANOVA approach using Proc MIXED. In this analysis, the fixed factors in our model were accession (Ca0045 and PI586666) and irrigation treatment (WW and WD), while the random factor was block. The response variables were B_TRL, S_TRL, RFW and RDW. The mean separation was done on least squares means using Tukey\u0026rsquo;s honestly significant difference (HSD) test with \u0026alpha;=0.05. Natural log transformation was performed on S_TRL and RFW to normalize residuals. No transformation was needed for B_TRL and RDW due to fit in the assumption of linear regression. Analyses for objective one also included repeated measurement analysis with an in-box imaging method using a general linear mixed model in Proc MIXED. In our general linear mixed model, we included the fixed factors of accession (Ca0045 and PI586666), irrigation treatment (WW and WD), and day (25, 27, 29, 32, 34, 36, 39, 41, 43, 46, 48, 50, 53 and 55 day), as well as their interactions. We also included block as a random factor. Since repeated measures analyses require appropriate variance covariance structures, we tested structures for our model and found the first order interdependence covariance structure resulted in the lowest AIC (Akaike\u0026apos;s information criterion) and BIC (Bayesian information criterion) and was significantly better than the simpler structure (Unstructured, compound symmetry, and heterogeneous compound symmetry) using a log likelihood test. Tukey\u0026rsquo;s honestly significant difference (HSD) test with \u0026alpha;=0.05 was utilized for mean separation. Natural log transformation was executed on B_TRL trait to normalize residuals. \u003c/p\u003e\n\u003cp\u003eFor the second objective, we took a regression approach running a general linear model using Proc REG to determine the degree to which RSA traits from the scanning washed root method could be well-predicted, and potentially replaced, by RSA traits from in-box imaging. Therefore, the predictor variable (X) was always an RSA trait from our in-box imaging and the response variable (Y) was an RSA trait from our scans of washed roots. The seven analyses were as follows: 1) predicting S_TRL from B_TRL; 2) predicting S_PE from B_PE; 3) S_AVRL from B_ AVRL; 4) predicting S_RD 0-0.5 from B_RD 0-0.5; 5) predicting S_RD 0.5-1 from B _RD 0.5-1; 6) predicting S_RD 1-2, cm from B_RD 1-2; and 7) predicting S_RD above 2 cm from B_RD above 2 cm. These analyses were performed separately on data from WW and WD treatments, as well as with both treatments combined (TC).\u003c/p\u003e\n\u003cp\u003eFor objective three, the goal was exploring the effectiveness of fresh and dry weight in predicting total root length in both scanning and in-box imaging. The response variable was either S_TRL or B_TRL. The predictor variable was either RFW or RDW. We performed four sets of analyses here: 1) predicting B_TRL from RFW ; 2) predicting B_TRL from RDW; 3) predicting S_TRL from RFW; and 4) predicting S_TRL. For objectives 2 and 3, analyses of WW and TC datasets for some traits better fit the assumptions of linear regression with natural log-log transformation. However, we present the analyses and interpretations using untransformed data because the R\u003csup\u003e2 \u003c/sup\u003evalues and p-values of the models showed little difference between the untransformed and natural log transformation datasets and interpretation of untransformed data is more straightforward.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eFor objective 1, we found a significant effect of the irrigation treatment on all the response variables analyzed (Table 1).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 1. F values and their significance for general linear mixed model results of end of experiment root system architecture traits. The numerator and denominator degree of freedom for each trait was 1 and 15, respectively. \u003csup\u003en\u0026nbsp;\u003c/sup\u003e No transformation.\u0026nbsp;\u003csup\u003ez \u0026nbsp;\u003c/sup\u003eNatural log transformation. Tukey\u0026rsquo;s HSD test with **** (p value \u0026le; 0.0001), *** (p value \u0026le; 0.001), ** (p value \u0026le; 0.01), * (p value \u0026le; 0.05), and ns (p value \u0026gt; 0.05) used to determine the significant effects. S_TRL cm (total root length for scanned washed roots), B_TRL cm (total root length, in box imaging), RFW g (root fresh weight) and RDW g (root dry weight).\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"526\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 114px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 116px;\"\u003e\n \u003cp\u003eScanned washed\u003cu\u003e\u0026nbsp;\u003c/u\u003eroots (cm)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003eIn-box image (cm)\u003c/p\u003e\n \u003cp\u003e\u003cu\u003e\u0026nbsp;\u003c/u\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 193px;\"\u003e\n \u003cp\u003eRoot biomass (g)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 114px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 116px;\"\u003e\n \u003cp\u003eS_TRL z\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 103px;\"\u003e\n \u003cp\u003eB_TRL n\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 111px;\"\u003e\n \u003cp\u003eRFW z\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 83px;\"\u003e\n \u003cp\u003eRDW n\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 114px;\"\u003e\n \u003cp\u003eAccession (A)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 116px;\"\u003e\n \u003cp\u003e0.91 (ns)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 103px;\"\u003e\n \u003cp\u003e1.06 (ns)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 111px;\"\u003e\n \u003cp\u003e0.16 (ns)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 83px;\"\u003e\n \u003cp\u003e0.22 (ns)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 114px;\"\u003e\n \u003cp\u003eIrrigation (I)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 116px;\"\u003e\n \u003cp\u003e60.62 (****)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 103px;\"\u003e\n \u003cp\u003e6.44 (*)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 111px;\"\u003e\n \u003cp\u003e31.32 (***)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 83px;\"\u003e\n \u003cp\u003e11.28 (**)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 114px;\"\u003e\n \u003cp\u003eA x I\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 116px;\"\u003e\n \u003cp\u003e1.55 (ns)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 103px;\"\u003e\n \u003cp\u003e0.83 (ns)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 111px;\"\u003e\n \u003cp\u003e0.09 (ns)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 83px;\"\u003e\n \u003cp\u003e0.23 (ns)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u0026nbsp;The back-transformed means for S_TRL (from scanned washed roots) were 6063 cm (WW) and 1619 cm (WD). In other words, TRL (from scanned washed roots) was 3.7 \u0026times; longer under the WW treatment than under the WD treatment (Table 2). Furthermore, the WW and WD treatments for B_TRL (from in-box imaging) were 2441 cm and 895 cm, respectively; B_TRL (In-box imaging) was 2.7 X longer in WW treatment than in WD treatment (Table 2). Additionally, the back-transformed means of RFW were 8.41 g (WW) and 3.06 g (WD), which RFW was 2.7 X greater under WW conditions than under WD conditions (Table 2.3). Likewise, the RDW were 0.5 g (WW) and 0.1 g (WD), or 5 X greater under WW conditions than under WD conditions (Table 2).\u003c/p\u003e\n\u003cp\u003eTable 2. Least square means (and s.e.) and mean separation for irrigation treatment effect at the end of experiment. \u003csup\u003en\u0026nbsp;\u003c/sup\u003eNo transformation.\u0026nbsp;\u003csup\u003ez\u0026nbsp;\u003c/sup\u003eNatural log transformation. WW (irrigation treatment effect for well-watered) and WD (irrigation treatment effect for water deficit). Superscript letters on least squares means used to indicate significant mean separation using Tukey\u0026rsquo;s HSD Test (P=0.05). The natural transformation means were back-transformed. Means sharing a letter are not significantly different. S_TRL cm (total root length for scanned washed roots), B_TRL cm (total root length, in box imaging), RFW g (root fresh weight) and RDW g (root dry weight).\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"552\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 122px;\"\u003e\n \u003cp\u003eScanned washed\u003cu\u003e\u0026nbsp;\u003c/u\u003eroots (cm)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003eIn-box image (cm)\u003c/p\u003e\n \u003cp\u003e\u003cu\u003e\u0026nbsp;\u003c/u\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 216px;\"\u003e\n \u003cp\u003eRoot biomass (g)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 122px;\"\u003e\n \u003cp\u003eS_TRL \u003csup\u003ez\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 138px;\"\u003e\n \u003cp\u003eB_TRL \u003csup\u003en\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 108px;\"\u003e\n \u003cp\u003eRFW \u003csup\u003ez\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 108px;\"\u003e\n \u003cp\u003eRDW \u003csup\u003en\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cu\u003eIrrigation\u003c/u\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 122px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 138px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 108px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 108px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003eWW\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 122px;\"\u003e\n \u003cp\u003e6063.2 (0.23) a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 138px;\"\u003e\n \u003cp\u003e2441.3 (551.07) a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 108px;\"\u003e\n \u003cp\u003e8.41 (0.2) a\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 108px;\"\u003e\n \u003cp\u003e0.51 (0.12) a\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003eWD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 122px;\"\u003e\n \u003cp\u003e1619.7 (0.22) b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 138px;\"\u003e\n \u003cp\u003e894.7 (499.06) b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 108px;\"\u003e\n \u003cp\u003e3.06 (0.19) b\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 108px;\"\u003e\n \u003cp\u003e0.10 (0.10) b\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp id=\"_Toc153650608\"\u003eOur repeated measurement analysis identified a significant effect of the day and irrigation treatment interaction (Table 3). In-box imaging showed that the B_TRL increased over time under WW and WD conditions; however, by day 50 and for the rest of the experiment, B_TRL in WD was lower than in the WW (Figure 3). Thus, in-box imaging was able to discern changes in an important RSA trait over time and identify when our experimental treatments caused its value to diverge.\u003c/p\u003e\n\u003cp\u003eTable 3. General linear mixed model results of repeated measures analysis for TRL measured by in-box imaging of chile pepper roots grown in rhizoboxes in the greenhouse at Ohio State University, Columbus, OH. The numerator degree of freedom (N DF), denominator degree of freedom (D DF), F value, and P value.\u0026nbsp;\u003csup\u003ez\u0026nbsp;\u003c/sup\u003e(natural log transformation). B_TRL, cm (total root length, in-box imaging).\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"4\" valign=\"top\" style=\"width: 258px;\"\u003e\n \u003cp\u003eB_TRL cm \u003csup\u003eZ\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 114px;\"\u003e\n \u003cp\u003e\u003cu\u003eEffects\u003c/u\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e\u003cu\u003eN df\u003c/u\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e\u003cu\u003eD df\u003c/u\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e\u003cu\u003eF value\u0026nbsp;\u003c/u\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 78px;\"\u003e\n \u003cp\u003e\u003cu\u003eP value\u003c/u\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 114px;\"\u003e\n \u003cp\u003eAccession (A)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e13.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e0.21 \u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 78px;\"\u003e\n \u003cp\u003ens\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 114px;\"\u003e\n \u003cp\u003eIrrigation (I)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e47.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e93.71\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 78px;\"\u003e\n \u003cp\u003e****\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 114px;\"\u003e\n \u003cp\u003eDay (D)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e14.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e0.15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 78px;\"\u003e\n \u003cp\u003ens\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 114px;\"\u003e\n \u003cp\u003eA x D\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e47.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e0.99\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 78px;\"\u003e\n \u003cp\u003ens\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 114px;\"\u003e\n \u003cp\u003eA x I\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e13.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e0 \u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 78px;\"\u003e\n \u003cp\u003ens\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 114px;\"\u003e\n \u003cp\u003eD x I\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e47.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e13.34\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 78px;\"\u003e\n \u003cp\u003e****\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 114px;\"\u003e\n \u003cp\u003eA x I x D\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e47.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e0.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 78px;\"\u003e\n \u003cp\u003ens\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp id=\"_Toc153650613\"\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe found that RSA traits measured with in-box images mostly predicted values from washed and scanned roots accurately. \u0026nbsp;Both treatments combined (TC) for all traits, all relationships were positive and significant (Supplementary Material 6). For instance, for total root length, a slope of 2.2 (\u0026plusmn; se = 0.09), means that for every 1 cm increase in B_TRL (in-box imaging), we found a 2.2 cm increase in S_TRL (washed root scan) (Supplementary Material 6a). Moreover, most overall analyses resulted in very high (\u0026gt;0.90) or high (\u0026gt;0.70) R\u003csup\u003e2\u003c/sup\u003e values, with total root length (0.96), perimeter length (0.94), and total root length (0-0.5mm diameter) (0.95) in the very high R\u003csup\u003e2\u003c/sup\u003e category (Supplementary Material 6a, b, and d) and total root length for 0.5-1 mm diameter (0.85), 1-2 mm diameter (0.72) and \u0026gt;2 mm diameter (0.82) in the high R\u003csup\u003e2\u003c/sup\u003e category (Supplementary Material 6e-g). Only the analysis of average root diameter had a low R\u003csup\u003e2\u003c/sup\u003e value (0.27) (Supplementary Material 6c). \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWhen we performed these regression analyses separately on plants grown in the WW and WD conditions (Figure 4), we found that the separate models did as good a job of predicting the data for TRL, perimeter, and total root length (0-0.5mm diameter) (Figure 4a, b, and d) as they had for the overall dataset (Supplementary Material 6a, b, and d). However, for the other traits, the model fit the data much better under WW than under WD (Figure 4c and e-g). \u0026nbsp;In particular, R\u003csup\u003e2\u003c/sup\u003e values in the WD analyses plummeted for average root diameter (0.09, Figure 4c), total root length for 1-2 mm diameter (0.35, Figure 4f), and \u0026gt;2 mm diameter (0.05, Figure 4g) and slopes for average root diameter and total root length for \u0026gt;2 mm diameter became non-significant under WD. Thus, predicting results of washed scanned roots from in-box imaging was less reliable under WD than under WW.\u003c/p\u003e\n\u003cp\u003eAll linear regression analyses predicting either fresh or dry root biomass from total root length (either from in-box imaging and scanning methods) showed significant positive relationships and very high (\u0026gt;0.9) or high (\u0026gt;0.7) R\u003csup\u003e2\u003c/sup\u003e values (Figure 5). However, as mentioned above, all R\u003csup\u003e2\u003c/sup\u003e values from WD analyses (0.61-0.72) were lower than those from the WW (0.78-0.9) and TC (0.84-0.92) analyses (Figure 5).\u0026nbsp;\u003c/p\u003e"},{"header":"Discussion","content":"\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003eThe effectiveness of irrigation on the morphological traits of chile peppers\u003c/h2\u003e\u003cp\u003eWhen discussing plant resource allocations, Bloom et al. (1985) implies that a plant can expand as long as the benefit of growth outweighs the cost, which means the cost of growing more biomass is relatively low. However, as resources (water, light or nutrients) become limited, the cost of producing more growth increases. As this relates to our results, the TRL (from in-box imaging and scanning methods) and root biomass were stunted under WD conditions. This finding confirmed that our imposed moisture treatments allowed us to assess the difference in biological responses under WW and WD treatments by the end of the experiment, but also over time. Sprunger et al. (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), observed that perennial intermediate wheatgrass increased root biomass as N inputs increased, supporting Bloom\u0026rsquo;s theory. It is important to note that the theoretical framework that Bloom et al. (1985), as well as the empirical support for that theory provided by Sprunger et al. (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) were based on field studies Our work with rhizoboxes, however different, also supports the same fundamental resource allocation patterns.\u003c/p\u003e\u003cp\u003eThese effects of WD on root growth may be primarily attributed to water limitation, which disrupts metabolism and physiological processes in plants, leading to the slowing of growth of new meristematic cells required for cell elongation. (Smirnoff, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e1993\u003c/span\u003e; Foyer et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e1994\u003c/span\u003e; Hirt \u0026amp; Shinozaki 2003; Wajhat et al., 2023). The exploration of WW and WD treatments in rhizobox experiments is relatively limited in existing research. However, Bodner et al. (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) conducted a rhizobox experiment examining WW and WD treatments that focused on method development. The study did not discuss the biological response of sugar beet root under WW and WD conditions. However, it did mention some physiological aspects, such as sugar beet cultivars with larger root surface areas having higher average stomatal conductance and keeping their stomata open longer.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\u003ch2\u003eScanning of washed root traits can be substituted with in-box root imaging\u003c/h2\u003e\u003cp\u003eAnalyzing scans of washed roots is considered the gold standard for accurately quantifying RSA in plant science. This experiment was the first instance of comparing results from scanned washed roots to those found through in-box root imaging. We found a strong relationship between TRL measured using in-box images and scans of washed roots overall and when analyzed by irrigation treatment. This means we can accurately predict TRL that would have been obtained from scanning washed roots based on the values calculated from in-box imaging. The same predictive ability is true for other traits. The values from in-box imaging of root perimeter, TRL for root diameters 0-0.5 mm, and TRL for root diameters 0.5-1 mm also predicted values from scanned washed roots well with R\u003csup\u003e2\u003c/sup\u003e of \u0026ge;\u0026thinsp;91%, \u0026ge;\u0026thinsp;85% and \u0026ge;\u0026thinsp;70%, respectively.\u003c/p\u003e\u003cp\u003eDespite this strong predictive relationship, the in-box imaging has some drawbacks in comparison to scanning washed roots. All root traits from in-box imaging underestimated the true measurements from scanned washed roots. Scanning washed roots allows for more accurate root trait measurement of the entire root system. By contrast, in-box imaging had limitations due to its two-dimensional nature. This method cannot fully capture all roots; as some remain hidden (Li et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Yet, our analysis indicates that the degree to which roots remain hidden is predictable. Moreover, to minimize the underestimation, we captured root traits from both the front and back panels when using in-box imaging. Thus, in-box imaging can be used instead of scanning of washed roots to reduce labor costs of RSA experiments.\u003c/p\u003e\u003cp\u003eFour distinct TRL root diameter classes showed the same pattern, with R\u003csup\u003e2\u003c/sup\u003e in WD analyses being lower than in WW and overall analyses. The high R\u003csup\u003e2\u003c/sup\u003e in TC was driven by the high R\u003csup\u003e2\u003c/sup\u003e in WW. In the WD analysis, the R\u003csup\u003e2\u003c/sup\u003e value was lower than that for WW for TRL for root diameters of 1\u0026ndash;2 mm and greater than 2 mm. Moreover, some predictions (e.g., for TRL of roots greater than 2 mm) were not significant at all. In contrast, the WW and TC analyses always yielded significant predictions. We also found when the TRL root diameter thickness increased, the R\u003csup\u003e2\u003c/sup\u003e values decreased. These might relate to the TRL root diameter distribution for in-box imaged and scanned washed roots. The TRL measurements from in-box imaging predominantly fell within the 0-0.5 mm and 0.5-1 mm diameter ranges. In contrast, the TRL from scanned washed roots was mainly within the 0.5-1 mm and 1\u0026ndash;2 mm diameter ranges.\u003c/p\u003e\u003cp\u003eIn addition to the average root diameter, our models performed well in predicting variation for the WW analyses. However, they were less effective for the WD and TC analyses due to low R\u0026sup2; values and a non-significant slope for WD. It is important to note the limitations of in-box imaging and scanned washed roots, which might contribute to inaccuracies in measuring average root diameter and TRL root diameter classification. For instance, by visualizing the root growth from the in-box imaging, there were multiple misrepresentations. First, if two separate roots grew close together or up against one another, RootPainter and RhizoVision Explorer sometimes recognized them as a single root with a thicker diameter, leading to inaccuracies in assigning TRL to different root diameter classes. Figure D.1 shows an example of where two separate roots joined together were misinterpreted as a single root with a thicker diameter. The average root diameter from in-box imaging was twice that of scanned washed roots (see Table). Second, when the plant is actively growing, a given length of root may have visible root hairs that can widen the root diameter estimated by RootPainter and RhizoVision Explorer. Those same roots, once washed, may scan as having a narrower diameter\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\u003ch2\u003eChallenges of irrigation treatments\u003c/h2\u003e\u003cp\u003eThe approach we used for imposing WW and WD treatments in our study was mostly based on the methods outlined in Bodner et al. (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), and Cassel \u0026amp; Nielsen (1986). Bodner et al. (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) measured the PAW under WW and WD conditions specifically for rhizoboxes. Cassel \u0026amp; Nielsen (1986) discussed using a pressure plate to determine the soil's field capacity and permanent wilting point. However, we made modifications to the PAW levels for our treatments over the course of the study. The rhizobox study in Bodner et al. (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) had 80% and 40% of PAW for their WW and WD treatments, respectively, and the study was performed in a controlled environment room which offered better control of environmental conditions than we had in the greenhouse. For instance, temperature and vapor pressure deficit values in the controlled environmental room were precisely managed to meet the needs of the WW and WD treatments. (Bodner et al. \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). By contrast, in our greenhouse, we were unable to maintain the temperature and vapor pressure deficit values recommended by Bodner et al. (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). In fact, the greenhouse temperature fluctuates. For instance, in run 3, the maximum temperature reached 34\u0026deg;C, with an average of 23.4\u0026deg;C. During our preliminary experiment, 18 days after transplanting, we decreased the PAW from 80% to 40% in the WD treatment, while maintaining the WW at 80% PAW. This resulted in a high mortality rate for the plants in WD treatments due to the high daytime temperatures in the greenhouse and the rapid drying of the upper soil layer led to a sudden onset of severe WD stress. To address this issue, we adjusted the PAW percentage by gradually decreasing it for the WD treatments in runs 1 to 5. Nevertheless, there were plant losses in runs 1 and 3 due to high daytime temperatures. However, no chili pepper plants died in run 2, likely because the rhizoboxes were positioned near the cooling pad. In runs 4 and 5 were conducted in summer with a maximum temperature of 36\u0026deg;C, we used shade cloth to reduce heat stress and adjusted the PAW levels. These changes eliminated plant mortality. By using shade cloth in high temperatures and starting with higher PAW levels before gradually decreasing them, we successfully established healthy seedlings before exposing them to WD stress.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\u003ch2\u003eThe feasibility of using both root phenotyping methods to predict the root biomass\u003c/h2\u003e\u003cp\u003eSetting up an image station and in-box image analysis pipeline, as well as scanning washed roots, can be labor intensive, so being able to accurately estimate root biomass from TRL would be beneficial. Additionally, no study has assessed whether both phenotyping methods can predict root biomass in a rhizobox. Based on our findings, using dry root weight gives a more accurate estimate of root biomass compared to using root fresh weight when predicting from TRL in root phenotyping studies. Similarly, for cell mass determination, measuring wet weight can be less precise due to liquid in the cell. However, dry weight, provides a more accurate measure by removing all the water content in the cell (Godbey. 2022). In a study on root growth for marigold and zinnia plants in clear cylindrical plexiglass tubes (7.6 cm diameter and 1.8 m tall) found that that there was a strong correlation between the total root length (visible roots) measured from a digital camera and the dry root weight (Judd et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Although the plant containers of rhizobox and cylindrical plexiglass tube had different designs. Our rhizobox study, along with the study by Judd et al. (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), demonstrated a strong relationship between total root length and root dry weight. Besides that, predicting root biomass from TRL measured using in-box imaging had lower R2 than scanning washed roots, possibly because some roots were hidden and unable to be captured by in-box imaging. It was unclear why the prediction of root biomass under WD had lower R2 than the WW condition. It may be due to the smaller root systems in WD. These smaller root systems were more prone to root loss during root washing and the hidden root from in-box imaging, leading to less accurate biomass predictions. In contrast, larger root systems might be less susceptible to root loss, resulting in higher R2 values in WW condition.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec19\" class=\"Section2\"\u003e\u003ch2\u003eBest practices for quantifying root traits via rhizoboxes\u003c/h2\u003e\u003cp\u003eOur use of acrylic panels in the rhizobox design introduced challenges during soil filling, impacting soil distribution. First, we explored the soiling-filling method introduced by Bodner et al. (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). The rhizobox used by Bodner et al. (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) was 100 cm high x 30 cm wide x 1 cm deep (internally). The back panel was made from 15 mm thick rigid grey PVC, and the front panel was composed of 6 mm thick hard mineral glass. This setup allowed them to effectively fill the rhizobox with soil by opening it and adding soil horizontally to the back panel. While our rhizobox had acrylic panels for specific reasons \u0026ndash; lower cost, lighter weight, and higher shatter resistance compared to glass -- it was less rigid than glass. We tested the soil filling method used by Bodner et al. (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) with our rhizobox (two 0.9-inch-thick acrylic panels); it resulted in the creation of large air gaps within the soil (Figure D.2). This was because it was difficult to achieve a completely flat soil surface during horizontal filling, especially considering that acrylic panels are less rigid compared to glass. This process appears to induce deformations in the acrylic panels when assembling the rhizobox, ultimately leading to the unintended formation of these air gaps within the soil. To address this problem, we implemented a soil-filling method as described in the methods section, which effectively eliminated air gaps within the soil.\" (Figure D.3). While the acrylic may have slightly deformed using our method (which can be seen in the variation in actual soil volume of 3100 to 3600 cm\u0026sup3; per box, rather than 3478 cm\u0026sup3;, it is impossible to completely avoid deformation due to use of the less rigid acrylic glass.\u003c/p\u003e\u003c/div\u003e"},{"header":"Conclusion","content":"\u003cp\u003eIn our rhizobox experiment, we successfully developed a comprehensive rhizobox pipeline capable of phenotyping the root architecture system of chile pepper plants, thus eliciting biological responses under both WW and WD treatments. We found that in-box imaging metrics can predict final scan metrics across WW and WD plants for many RSA traits. However, our study also uncovered some limitations. For instance, inbox imaging was not successful at determining average root diameter and TRL of root diameter above 2 mm in WD analysis. Moreover, dry root weight were effective predictors of total root length from the in-box imaging and scanning washed root methods in WW, WD and TC. By employing this affordable and user-friendly pipeline, researchers will be able to readily conduct root experiments using a rhizobox pipeline.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003e\u003cem\u003e2D\u003c/em\u003e: 2-dimensional\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e3D\u003c/em\u003e: 3-dimensional\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eANOVA\u003c/em\u003e: Analysis of variance\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eB_TRL\u003c/em\u003e: Total root length from root imaging\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eB_PE\u003c/em\u003e: Root perimeter from root imaging\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eB_AVR\u003c/em\u003e: Average root diameter from root imaging\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eB_RD 0-0.5\u003c/em\u003e: Total root length\u0026nbsp;for roots with 0 to 0.5 mm diameter from root imaging\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eB_RD 0.5-1\u003c/em\u003e: Total root length\u0026nbsp;for roots with 0.5 to 1 mm diameter from root imaging\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eB_RD 1-2\u003c/em\u003e: Total root length\u0026nbsp;for roots with 1 to 2 mm diameter from root imaging\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eB_RD \u0026gt;2\u003c/em\u003e: Total root length for roots with above 2 mm diameter from root imaging\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCa\u003c/em\u003e: Ca0045, Costeño Rojo landrace chile pepper accession\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCC\u003c/em\u003e: Equal number of accessions in both treatments per run\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ecm\u003c/em\u003e: Centimeter\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eD\u003c/em\u003e: Inner diameter between two panels\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ee.u\u003c/em\u003e: Experimental unit\u003c/p\u003e\n\u003cp\u003e\u003cem\u003edpi\u003c/em\u003e: dot per inch\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eFC\u003c/em\u003e: Field capacity\u003c/p\u003e\n\u003cp\u003eg: Gram\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eH\u003c/em\u003e: Height\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ekPa\u003c/em\u003e; Kilopascals\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePAW\u003c/em\u003e: plant available water\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePWP\u003c/em\u003e: Permanent wilting point\u003c/p\u003e\n\u003cp\u003ePI: PI 586666, commercial chile pepper accession\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eRDW\u003c/em\u003e:\u0026nbsp;Root dry weight\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eRFW\u003c/em\u003e:\u0026nbsp;Root fresh weight\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eRSA\u003c/em\u003e: Root system architecture\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eTC\u003c/em\u003e: Both treatments combined\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eTRL\u003c/em\u003e: Total root length\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eW\u003c/em\u003e: Width\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eWD\u003c/em\u003e: Water deficit treatment\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eWW\u003c/em\u003e: Well-watered treatment\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eX\u003c/em\u003e: Predictor variable\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eY\u003c/em\u003e: Response variable\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe would like to express our gratitude to J. McCoy, L. Connolly, H. Scheppler, E. Cigany, K. Fulcher, M. Lincicome and Vivian Bernau for their valuable assistance in carrying out the research and D. Francis for helpful guidance. We thank Mike Anderson and Jackie Rothgery for their assistance in preparing the greenhouse space and managing pest. We are especially grateful to our collaborator from Mexico, L. Jardón Barbolla, for providing the germplasm Ca0045 used in this study. We also acknowledge partial support for this research from the National Institute of Food and Agriculture (AFRI grant 20186701327555), the Center for Applied Plant Sciences at Ohio State University, and the Ohio Agricultural Research and Development Center (OARDC) to KLM and LKM (OSU HCS 25-XX).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBitarafan, Z., Kaczmarek-Derda, W., Berge, T. W., T\u0026oslash;rresen, K. S., \u0026amp; Fl\u0026oslash;istad, I. S. (2022). Soil steaming to disinfect barnyardgrass-infested soil masses. \u003cem\u003eWeed Technology\u003c/em\u003e, \u003cem\u003e36\u003c/em\u003e(1), 177\u0026ndash;185. https://doi.org/10.1017/wet.2021.107\u003c/li\u003e\n\u003cli\u003eBodner, G., Alsalem, M., Nakhforoosh, A., Arnold, T., \u0026amp; Leitner, D. (2017). RGB and Spectral Root Imaging for Plant Phenotyping and Physiological Research: Experimental Setup and Imaging Protocols. \u003cem\u003eJournal of Visualized Experiments\u003c/em\u003e, \u003cem\u003e126\u003c/em\u003e, 56251. https://doi.org/10.3791/56251\u003c/li\u003e\n\u003cli\u003eBodner, G., Nakhforoosh, A., \u0026amp; Kaul, H.-P. (2015). Management of crop water under drought: A review. \u003cem\u003eAgronomy for Sustainable Development\u003c/em\u003e, \u003cem\u003e35\u003c/em\u003e(2), 401\u0026ndash;442. https://doi.org/10.1007/s13593-015-0283-4\u003c/li\u003e\n\u003cli\u003eBucksch, A., Burridge, J., York, L. M., Das, A., Nord, E., Weitz, J. S., \u0026amp; Lynch, J. P. (2014). Image-Based High-Throughput Field Phenotyping of Crop Roots. \u003cem\u003ePLANT PHYSIOLOGY\u003c/em\u003e, \u003cem\u003e166\u003c/em\u003e(2), 470\u0026ndash;486. https://doi.org/10.1104/pp.114.243519\u003c/li\u003e\n\u003cli\u003eCassel, D. K., \u0026amp; Nielsen, D. R. (2018). Field Capacity and Available Water Capacity. In A. Klute (Ed.), \u003cem\u003eSSSA Book Series\u003c/em\u003e (pp. 901\u0026ndash;926). Soil Science Society of America, American Society of Agronomy. https://doi.org/10.2136/sssabookser5.1.2ed.c36\u003c/li\u003e\n\u003cli\u003eChen, X., Zhu, Y., Ding, Y., Pan, R., Shen, W., Yu, X., \u0026amp; Xiong, F. (2021). The relationship between characteristics of root morphology and grain filling in wheat under drought stress. \u003cem\u003ePeerJ\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e, e12015. https://doi.org/10.7717/peerj.12015\u003c/li\u003e\n\u003cli\u003eCorrea, J., Postma, J. A., Watt, M., \u0026amp; Wojciechowski, T. (2019). Soil compaction and the architectural plasticity of root systems. \u003cem\u003eJournal of Experimental Botany\u003c/em\u003e, \u003cem\u003e70\u003c/em\u003e(21), 6019\u0026ndash;6034. https://doi.org/10.1093/jxb/erz383\u003c/li\u003e\n\u003cli\u003eCrossett, R. N., Campbell, D. J., \u0026amp; Stewart, H. E. (1975). Compensatory growth in cereal root systems. \u003cem\u003ePlant and Soil\u003c/em\u003e, \u003cem\u003e42\u003c/em\u003e(3), 673\u0026ndash;683. https://doi.org/10.1007/BF00009951\u003c/li\u003e\n\u003cli\u003eDayoub, E., Lamichhane, J. R., Schoving, C., Debaeke, P., \u0026amp; Maury, P. (2021). Early-Stage Phenotyping of Root Traits Provides Insights into the Drought Tolerance Level of Soybean Cultivars. \u003cem\u003eAgronomy\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(1), 188. https://doi.org/10.3390/agronomy11010188\u003c/li\u003e\n\u003cli\u003eEl-Soda, M., Malosetti, M., Zwaan, B. J., Koornneef, M., \u0026amp; Aarts, M. G. M. (2014). Genotype \u0026times; environment interaction QTL mapping in plants: Lessons from Arabidopsis. \u003cem\u003eTrends in Plant Science\u003c/em\u003e, \u003cem\u003e19\u003c/em\u003e(6), 390\u0026ndash;398. https://doi.org/10.1016/j.tplants.2014.01.001\u003c/li\u003e\n\u003cli\u003eFoyer, C. H., Lelandais, M., \u0026amp; Kunert, K. J. (1994). Photooxidative stress in plants. \u003cem\u003ePhysiologia Plantarum\u003c/em\u003e, \u003cem\u003e92\u003c/em\u003e(4), 696\u0026ndash;717. https://doi.org/10.1111/j.1399-3054.1994.tb03042.x\u003c/li\u003e\n\u003cli\u003eJudd, L. A., Jackson, B. E., \u0026amp; Fonteno, W. C. (2015). Rhizometer: An Apparatus to Observe and Measure Root Growth and Its Effect on Container Substrate Physical Properties Over Time. \u003cem\u003eHortScience\u003c/em\u003e, \u003cem\u003e50\u003c/em\u003e(2), 288\u0026ndash;294. https://doi.org/10.21273/HORTSCI.50.2.288\u003c/li\u003e\n\u003cli\u003eHirt, H., \u0026amp; Shinozaki, K. (2003). Plant responses to abiotic stress (Vol. 4). Springer Science \u0026amp; Business Media.Kuijken, R. C. P., Van Eeuwijk, Fred. A., Marcelis, L. F. M., \u0026amp; Bouwmeester, H. J. (2015). Root phenotyping: From component trait in the lab to breeding: Table 1. \u003cem\u003eJournal of Experimental Botany\u003c/em\u003e, \u003cem\u003e66\u003c/em\u003e(18), 5389\u0026ndash;5401. https://doi.org/10.1093/jxb/erv239\u003c/li\u003e\n\u003cli\u003eK\u0026uuml;cke, M., Schmid, H., \u0026amp; Spiess, A. (1995). A comparison of four methods for measuring roots of field crops in three contrasting soils. \u003cem\u003ePlant and Soil\u003c/em\u003e, \u003cem\u003e172\u003c/em\u003e(1), 63\u0026ndash;71. https://doi.org/10.1007/BF00020860\u003c/li\u003e\n\u003cli\u003eLesmes-Vesga, R. A., Cano, L. M., Ritenour, M. A., Sarkhosh, A., Chaparro, J. X., \u0026amp; Rossi, L. (2022). Rhizoboxes as Rapid Tools for the Study of Root Systems of Prunus Seedlings. \u003cem\u003ePlants\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(16), 2081. https://doi.org/10.3390/plants11162081\u003c/li\u003e\n\u003cli\u003eLi, A., Zhu, L., Xu, W., Liu, L., \u0026amp; Teng, G. (2022). Recent advances in methods for \u003cem\u003ein situ\u003c/em\u003e root phenotyping. \u003cem\u003ePeerJ\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e, e13638. https://doi.org/10.7717/peerj.13638\u003c/li\u003e\n\u003cli\u003eLi, L., Zhang, Q., \u0026amp; Huang, D. (2014). A review of imaging techniques for plant phenotyping. \u003cem\u003eSensors (Basel, Switzerland)\u003c/em\u003e, \u003cem\u003e14\u003c/em\u003e(11), 20078\u0026ndash;20111. https://doi.org/10.3390/s141120078\u003c/li\u003e\n\u003cli\u003eL\u0026oacute;pez-Serrano, L., Canet-Sanchis, G., Vuletin Selak, G., Penella, C., San Bautista, A., L\u0026oacute;pez-Galarza, S., \u0026amp; Calatayud, \u0026Aacute;. (2019). Pepper Rootstock and Scion Physiological Responses Under Drought Stress. \u003cem\u003eFrontiers in Plant Science\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e, 38. https://doi.org/10.3389/fpls.2019.00038\u003c/li\u003e\n\u003cli\u003eMa\u0026scaron;kov\u0026aacute;, T., \u0026amp; Klime\u0026scaron;, A. (2020). The Effect of Rhizoboxes on Plant Growth and Root: Shoot Biomass Partitioning. \u003cem\u003eFrontiers in Plant Science\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e, 1693. https://doi.org/10.3389/fpls.2019.01693\u003c/li\u003e\n\u003cli\u003eMcCoy, J. E., McHale, L. K., Kantar, M., Jard\u0026oacute;n-Barbolla, L., \u0026amp; Mercer, K. L. (2022). Environment of origin and domestication affect morphological, physiological, and agronomic response to water deficit in chile pepper (Capsicum sp.). \u003cem\u003ePLOS ONE\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e(6), e0260684. https://doi.org/10.1371/journal.pone.0260684\u003c/li\u003e\n\u003cli\u003eNational Drought Mitigation Center. (n.d.). Types of drought. Retrieved October 21, 2023, from https://drought.unl.edu/Education/DroughtIn-depth/TypesofDrought.aspx\u003c/li\u003e\n\u003cli\u003eOsmolovskaya, N., Shumilina, J., Kim, A., Didio, A., Grishina, T., Bilova, T., Keltsieva, O. A., Zhukov, V., Tikhonovich, I., Tarakhovskaya, E., Frolov, A., \u0026amp; Wessjohann, L. A. (2018). Methodology of Drought Stress Research: Experimental Setup and Physiological Characterization. \u003cem\u003eInternational Journal of Molecular Sciences\u003c/em\u003e, \u003cem\u003e19\u003c/em\u003e(12), 4089. https://doi.org/10.3390/ijms19124089\u003c/li\u003e\n\u003cli\u003ePaez-Garcia, A., Motes, C., Scheible, W.-R., Chen, R., Blancaflor, E., \u0026amp; Monteros, M. (2015). Root Traits and Phenotyping Strategies for Plant Improvement. \u003cem\u003ePlants\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e(2), 334\u0026ndash;355. https://doi.org/10.3390/plants4020334\u003c/li\u003e\n\u003cli\u003ePassioura, J. B. (2006). Viewpoint: The perils of pot experiments. \u003cem\u003eFunctional Plant Biology\u003c/em\u003e, \u003cem\u003e33\u003c/em\u003e(12), 1075. https://doi.org/10.1071/FP06223\u003c/li\u003e\n\u003cli\u003eSchmidt, J. E., Lowry, C., \u0026amp; Gaudin, A. C. M. (2018). An Optimized Rhizobox Protocol to Visualize Root Growth and Responsiveness to Localized Nutrients. \u003cem\u003eJournal of Visualized Experiments\u003c/em\u003e, \u003cem\u003e140\u003c/em\u003e, 58674. https://doi.org/10.3791/58674\u003c/li\u003e\n\u003cli\u003eSchneider, H. M., \u0026amp; Lynch, J. P. (2020). Should Root Plasticity Be a Crop Breeding Target? \u003cem\u003eFrontiers in Plant Science\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e, 546. https://doi.org/10.3389/fpls.2020.00546\u003c/li\u003e\n\u003cli\u003eSeethepalli, A., \u0026amp; York, L. M. (2020). \u003cem\u003eRhizoVision Explorer\u0026mdash;Interactive software for generalized root image analysis designed for everyone\u003c/em\u003e (2.0.2) [Computer software]. Zenodo. https://doi.org/10.5281/ZENODO.4095629\u003c/li\u003e\n\u003cli\u003eSeleiman, M. F., Al-Suhaibani, N., Ali, N., Akmal, M., Alotaibi, M., Refay, Y., Dindaroglu, T., Abdul-Wajid, H. H., \u0026amp; Battaglia, M. L. (2021). Drought Stress Impacts on Plants and Different Approaches to Alleviate Its Adverse Effects. \u003cem\u003ePlants\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e(2), 259. https://doi.org/10.3390/plants10020259\u003c/li\u003e\n\u003cli\u003eSmirnoff, N. (1993). The role of active oxygen in the response of plants to water deficit and desiccation. \u003cem\u003eNew Phytologist\u003c/em\u003e, \u003cem\u003e125\u003c/em\u003e(1), 27\u0026ndash;58. https://doi.org/10.1111/j.1469-8137.1993.tb03863.x\u003c/li\u003e\n\u003cli\u003eSmith, A. G., Han, E., Petersen, J., Olsen, N. A. F., Giese, C., Athmann, M., Dresb\u0026oslash;ll, D. B., \u0026amp; Thorup‐Kristensen, K. (2022). R oot P ainter: Deep learning segmentation of biological images with corrective annotation. \u003cem\u003eNew Phytologist\u003c/em\u003e, \u003cem\u003e236\u003c/em\u003e(2), 774\u0026ndash;791. https://doi.org/10.1111/nph.18387 \u003c/li\u003e\n\u003cli\u003eSprunger, C. D., Culman, S. W., Robertson, G. P., \u0026amp; Snapp, S. S. (2018). Perennial grain on a Midwest Alfisol shows no sign of early soil carbon gain. \u003cem\u003eRenewable Agriculture and Food Systems\u003c/em\u003e, \u003cem\u003e33\u003c/em\u003e(4), 360\u0026ndash;372. https://doi.org/10.1017/S1742170517000138\u003c/li\u003e\n\u003cli\u003eSultan, S. E. (2000). Phenotypic plasticity for plant development, function and life history. \u003cem\u003eTrends in Plant Science\u003c/em\u003e, \u003cem\u003e5\u003c/em\u003e(12), 537\u0026ndash;542. https://doi.org/10.1016/S1360-1385(00)017970\u003c/li\u003e\n\u003cli\u003eTakahashi, H., \u0026amp; Pradal, C. (2021). Root phenotyping: important and minimum information required for root modeling in crop plants. \u003cem\u003eBreeding science\u003c/em\u003e, \u003cem\u003e71\u003c/em\u003e(1), 109\u0026ndash;116. https://doi.org/10.1270/jsbbs.20126\u003c/li\u003e\n\u003cli\u003eUga, Y., Okuno, K., \u0026amp; Yano, M. (2011). Dro1, a major QTL involved in deep rooting of rice under upland field conditions. \u003cem\u003eJournal of Experimental Botany\u003c/em\u003e, \u003cem\u003e62\u003c/em\u003e(8), 2485\u0026ndash;2494. https://doi.org/10.1093/jxb/erq429\u003c/li\u003e\n\u003cli\u003eUSDA, National Agricultural Statistics Service. (2023). Vegetables 2022 summary (February 2023) (No. 81). Retrieved October 21, 2023, from https://downloads.usda.library.cornell.edu/usda-esmis/files/02870v86p/hq37x121v/4b29ck28c/vegean23.pdf\u003c/li\u003e\n\u003cli\u003eVia, S., Gomulkiewicz, R., De Jong, G., Scheiner, S. M., Schlichting, C. D., \u0026amp; Van Tienderen, P. H. (1995). Adaptive phenotypic plasticity: Consensus and controversy. \u003cem\u003eTrends in Ecology \u0026amp; Evolution\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e(5), 212\u0026ndash;217. https://doi.org/10.1016/S0169-5347(00)89061-8\\\u003c/li\u003e\n\u003cli\u003eWajhat-Un-Nisa, Sandhu, S., Ranjan, R. et al. Root plasticity: an effective selection technique for identification of drought tolerant maize (Zea mays L.) inbred lines. Sci Rep 13, 5501 (2023). https://doi.org/10.1038/s41598-023-31523-w\u003c/li\u003e\n\u003cli\u003eZhang, Y., Zhang, W., Cao, Q., Zheng, X., Yang, J., Xue, T., Sun, W., Du, X., Wang, L., Wang, J., Zhao, F., Xiang, F., \u0026amp; Li, S. (2022). WinRoots: A High-Throughput Cultivation and Phenotyping System for Plant Phenomics Studies Under Soil Stress. \u003cem\u003eFrontiers in Plant Science\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e, 794020. https://doi.org/10.3389/fpls.2021. \u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-7697278/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7697278/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cem\u003eBackground\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eDrought stress can significantly impede plant productivity, adversely impacting crop yields. The root system is an important plant organ contributing to drought resistance mechanisms. Therefore, assessing root systems under drought stress conditions can provide insights to identify root traits associated with enhanced drought resistance. When seeking dense and high-quality root data, root phenotyping can be complex, costly, and time-consuming. The objectives of this study were to establish a method to grow chili pepper plants in a soil-based rhizobox container under water deficit conditions and compare two methods for collecting two-dimensional root trait data from their roots.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eMethod\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eWe grew two chile peppers (\u003cem\u003eCapsicum\u0026nbsp;annuum\u003c/em\u003e) accessions in soil-based rhizobox containers to analyze the responses of root architecture traits under well-watered and water-deficit conditions during the vegetative stage. The root traits were phenotyped using two different methods. The first method involved non-destructive in-box imaging of roots \u003cem\u003ein situ\u003c/em\u003e through acrylic glass while the plant grew. The second method involved scanning destructively harvested and washed roots—the gold standard for root measurements. For the first method, we developed a pipeline for rhizobox studies to demonstrate the response of root system architecture to water deficit over time and assessed the quality of non-destructive in-box imaging methods as compared to scans of destructively harvested and washed roots. We used a relatively large rhizobox (53.34 cm in width x 78.73 cm in height) into which we established and maintained well-watered and water deficit conditions based on the field capacity and permanent wilting point of the soil (Bodner et al. 2017; Cassel \u0026amp; Nielsen 1986).Our in-box root imaging pipeline captures high-resolution root images with an affordable camera that can achieve a maximum resolution of 9152 x 6944 pixels, as well as high-quality root segmentation using a robust graphical user interface-based software called RootPainter (Smith et al. 2022).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eResults\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eRoot growth decreased under water deficit compared to well-watered conditions. There were strong positive relationships between total root length using the washed scanned method and the in-box imaging method. The same was observed for root perimeter and most of the total root length distinct root diameter classes, but not for average root diameter. Some of these relationships weakened under water deficit conditions. In addition, we also found a strong relationship between root biomass and total root length using both phenotyping methods.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eConclusion\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eOverall, we developed a rhizobox pipeline for phenotyping the root system architecture of chile pepper plants under both well-watered and water-deficit conditions. We showed that measurements taken via non-destructive in-box imaging strongly predict those taken directly on washed scanned roots, with the added benefit of allowing repeated measurements over time.\u003c/p\u003e","manuscriptTitle":"A comprehensive rhizobox pipeline for analyzing pepper root system architecture under well-watered and water deficit conditions","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-06 09:27:05","doi":"10.21203/rs.3.rs-7697278/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"a716521c-6393-4544-8275-9d3b411cf98e","owner":[],"postedDate":"October 6th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-10-27T14:28:33+00:00","versionOfRecord":[],"versionCreatedAt":"2025-10-06 09:27:05","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7697278","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7697278","identity":"rs-7697278","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}