Image-based phenotyping of faba bean genetic resources for water deficit responses under controlled conditions

Image-based phenotyping of faba bean genetic resources for water deficit responses under controlled 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 Research Article Image-based phenotyping of faba bean genetic resources for water deficit responses under controlled conditions Sylvain Poque, Ulrika Carlson-Nilsson, Muhammad Omer, Anna Palmé, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9598396/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 12 You are reading this latest preprint version Abstract Faba bean ( Vicia faba L.) has great potential to contribute to sustainable agriculture and protein security globally but is known to be very sensitive to drought stress. Uncovering drought-resilient germplasm is critical for developing resilient cultivars and advancing our understanding of the mechanisms underlying stress adaptation. However, high-throughput plant phenotyping under stress conditions remain a major bottleneck in crop genetics and breeding programs. In this study, a multi-sensor indoor phenotyping platform was used to assess 44 faba bean genotypes under water deficit conditions. Standardized, monitored stress conditions were achieved by watering-by-weighing for drought onset, duration, and intensities allowing genotype-level comparisons. The genotypes showed a range of stress responses in growth and physiology, including traits such as plant height, biomass, water use efficiency (WUE), and chlorophyll fluorescence parameters. Digital biomass, derived from combined top- and side-view plant imaging, was strongly correlated with biological biomass at the experimental endpoint, validating its use as a non-destructive proxy for growth assessment in faba bean. Time-resolved generalized additive modelling further revealed genotype-specific differences in the timing and magnitude of water deficit response. Genotypes that maintained growth and WUE under water deficit conditions may serve as valuable pre-breeding materials for development of drought-adapted faba bean. drought stress plant phenomics legumes digital biomass water use efficiency Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Introduction Faba bean ( Vicia faba L.) is an important cool-season grain legume crop grown worldwide because of its high seed protein content and versatility in agricultural systems, contributing to food security and sustainable agriculture (Jensen et al. 2010 ; Khazaei and Vandenberg 2020 ). Nevertheless, faba bean is considered to be sensitive to drought stress (Khan et al. 2010 ) as the major environmental factor that negatively affects its growth, development, and yield potential. Global climate change has dramatically changed the frequency and patterns of rainfall over the last century. The decrease in precipitation due to infrequent rain events, combined with the predicted rise in atmospheric temperature, intensifies the frequency and severity of drought incidents, negatively affecting crop performance (Spinoni et al. 2018 ; Bevacqua et al. 2022 ). For example, in the Nordic region, early- to mid-summer droughts have become extremely likely due to climate change and cause significant crop yield losses (Peltonen-Sainio et al. 2021 ). Therefore, development of drought-adapted crop varieties is essential, particularly in drought-susceptible crops such as faba bean. Plant phenotyping, the process of systematic measurement and analysis of plant characteristics, is fundamental to understanding plant-environment interactions (Pieruschka and Schurr 2019 ; Visakh et al. 2024 ). Through phenotyping, drought-adapted germplasm can be identified to support the development of resilient cultivars and to advance understanding of stress adaptation mechanisms (Swarup et al. 2021 ). This requires systematic phenotyping of diverse genotypes under water-deficit conditions to identify those that exhibit improved production in water-limited environments (Guadarrama-Escobar et al. 2024 ). A number of faba bean genotypes have been screened for this purpose (e.g., Amede et al. 1999 ; Khazaei et al. 2013 ; Ali et al. 2016 ). However, these studies have been largely constrained by the lack of reliable and standardized phenotyping tools and have often been conducted using low-throughput approaches. High-Throughput Plant Phenotyping (HTPP) technologies have revolutionized plant breeding by enabling rapid, precise, and large-scale assessment of plant traits under controlled and field conditions. The HTPP platforms leverage advanced imaging technologies, such as visible light RedGreenBlue (RGB) imaging, thermal imaging, hyperspectral imaging, and chlorophyll fluorometry, to provide detailed insights into the physiological, morphological, and biochemical responses of plants to drought stress (Kim et al. 2021 ; Gill et al. 2022 ; Mansoor and Chung 2024 ; Sozzani et al. 2014 ). The HTPP facilitates the characterization of key traits related to shoot and root architecture (Nagel et al. 2012 ), and physiological adaptation to water stress (Shapiguzov et al. 2025 ). These platforms allow researchers to identify drought adapted germplasm and enhance genetic studies by integrating phenotyping and genotyping data (Roitch et al. 2022 ; Al-Tamimi et al. 2022 ; Anshori et al. 2023 ). Integrated image analysis and machine learning has recently enabled the assessment of large populations of faba bean under natural field conditions (Ji et al. 2024 ). Furthermore, digital phenotyping tools and protocols, for example to measure digital biomass, have been more developed and applied in cereals (Neumann et al. 2015 ; Dodig et al. 2021 ) than in grain legumes. Despite the potential of HTPP technologies, field-based phenotyping remains resource-intensive and is often challenged by the inherent variability in drought onset, duration, and intensity. To address these challenges, researchers have employed innovative approaches such as mobile field rainout shelters (Mohammadi et al. 2025; Balko et al. 2023 ). Outdoor and indoor HTPP approaches are complementary, as indoor facilities enable controlled, high-precision monitoring of plant traits under controlled environmental conditions, while outdoor platforms capture plant performance under field conditions. To better understand the response of faba bean above-ground to water deficit, we utilized an indoor HTPP phenotyping facility, which allowed automated watering and precise water deficit treatment. The National Plant Phenotyping Infrastructure (NaPPI), a state-of-the-art indoor facility at the University of Helsinki, Finland, was used to monitor plant stress responses using visible-light RGB cameras and chlorophyll fluorometry (Alexandersson et al. 2018 ; Pollari et al. 2022 ; Chovancek et al. 2025 ). The aims of this study were to (i) characterise phenotypic variation in a faba bean germplasm collection for growth dynamics, biological and digital biomass, water use efficiency (WUE), and chlorophyll fluorescence parameters under well watered and water deficit; (ii) assess the potential of image-derived digital biomass and digital WUE to function as non-destructive proxies for their biological counterparts in this species, and (iii) apply time-resolved generalized additive modelling to classify genotypes by the timing of growth responses to water deficit. Materials and methods Plant material Forty-four faba bean genotypes were used in this study. Information on genotype identification, origin and collection sites are presented in Table S1 . The material includes both large-seeded garden types and smaller seeded fodder types and originated from Nordic countries and northern Europe. Most of the genotypes represent landraces collected in Finland and Sweden (Leino 2023 ) and were obtained from the Nordic genebank (NordGen; Alnarp, Sweden). Genotypes ILB 938/2 and Mélodie/2 were added to the set as benchmarks for high water use efficiency and Aurora/2 as a cultivar with low water use efficiency (Khazaei et al. 2014 and 2018 ). Growing conditions The plants were grown in a plexiglass greenhouse within the Viikki Plant Growth Facilities (ViGOR), University of Helsinki, Finland. The growth conditions were set to a 16 h light and 8 h dark photoperiod, with day and night air temperatures set at 21 C and 16 C, respectively, and a target relative humidity (RH) of approximately 60%. A photosynthetic photon flux density of approximately 250–300 µmol.m⁻².s⁻¹ was maintained at the canopy level by high-pressure sodium lamps. Environmental conditions inside the greenhouse were logged every minute throughout the experiment. Across the experimental period, mean daytime values were 22.8°C, 46.1% RH; mean nighttime values were 18.9°C and 55.4% RH. The corresponding mean vapour pressure deficit (VPD) was 1.53 kPa during the day and 0.99 kPa at night, indicating moderate to occasionally high evaporative demand. All conditions were uniform across treatments and genotypes. Seeds of all 44 genotypes were sown in three batches in October 2024, in 5 L pots filled with peat-based potting mix (Karkea Ruukutusseos, W R8014, Kekkilä Oy, Vantaa, Finland). Before sowing, they were inoculated with Rhizobium leguminosarum biovar. viciae (Elomestari Oy, Tornio, Finland). The 264 pots were divided into three batches each containing a pot for the well watered treatment and for the water deficit treatment. To minimise potential positional effects within the greenhouse, pots were assigned to positions within each batch in a fully randomised order at the start of the experiment. Batches were then rotated weekly across greenhouse positions as a unit, ensuring that no batch remained in the same spatial position for more than one week throughout the experiment. Water deficit treatment During the first seven days after germination (DAG), all the pots were watered to reach field capacity to allow proper seed germination. Prior to the experiment, the water holding capacity (WHC) of the peat was determined by weighing equally soil filled pots, saturated with water and allowed to drain freely for 24 hours (equivalent to field capacity), representing 100% of the WHC. Then the same pots were completely oven-dried to 0% of the WHC and weighed. These established the relationship between pot weight and actual water content (%) used to monitor soil WHC% during the experiment (Figure S1 ). For the experiment the soil in each pot was weighed to ensure uniformity. This standardized approach minimized variation in soil volume, providing consistent water levels for all plants. The pots were covered with blue mat to limit evaporation and plant imaging. Automated watering by weighing allowed daily recording of pot weights, enabling calculation of both water loss between consecutive days (g) and the amount of water supplied. The plants were subsequently subjected to two distinct water regimes: 80% of the WHC as the control treatment and 30% of the WHC as the water deficit treatment, which was gradually achieved by 20 DAG. The 30% WHC level was selected based on preliminary trials in which 25% WHC caused excessive stress and rapid decline in several genotypes, whereas 30% WHC produced a stable, moderate drought level suitable for differentiating genotype responses without inducing premature plant failure. These watering levels were maintained for 10 days (Fig. 1 ). After 31 DAG, the water deficit treatment was intensified by reducing the water level to 20% of the WHC for 5 days, in order to amplify stress severity and capture a broader range of genotypic responses. By 35 DAG, WHC was adjusted back to 30% and maintained at this level until 39 DAG (Fig. 1 and Figure S1 ). Image-based phenotyping The phenotyping was performed at the National Plant Phenotyping Infrastructure (NaPPI, http://www.helsinki.fi/en/infrastructures/national-plant-phenotyping ). The PlantScreen™ Modular platform is an integrated system for automated plant imaging and management. Plants are set in individual transportation disks and moved between handling, imaging and watering stations. The feature of watering by weighing allows for management of water content of the soil (Chovancek et al. 2025 ). This phenotyping platform is equipped with imaging units, for top- and side-view visible light RGBs and top-view pulse amplitude modulated (PAM) chlorophyll fluorometer (FluorCam). Lighting conditions, plant positioning, and camera settings were fixed throughout the experiment. Visible-light digital imaging The RGB imaging unit is a light-isolated box equipped with a turning table with precise angle positioning, RGB cameras using GigE uEye UI-5480SE-C/M − 5 Megapixels QSXGA Camera with 1/2” CMOS Sensor (IDS, Germany) are supplemented with LED-based lighting source to ensure homogenous illumination of the imaged object. Top view RGB images of the plants were captured at resolution 2560 × 1920 pixels and camera height was automatically adjusted by a light curtain unit measuring plant height. For side view projections, line scan camera captured images at resolution of 2560 × 3476 pixels. Weekly RGB imaging was performed for all plants, with additional imaging of individual batches conducted separately within the same week, for more than five weeks. Top and side view RGB imaging were used to capture plant growth parameters such as the canopy area, side view area and height during the time series. The plant surface area was segmented using pixel colour thresholding of the RGB images and all growth parameters were automatically analysed by PlantScreen Data Analyzer v3.2.4.5 (PSI, Drásov, Czech R.). To facilitate top-view canopy data extraction, blue rubber mats were placed around each plant in each pot. Representative examples of segmented side-view and top-view images under both watering treatments are shown in Figure S2 . During the experiment plant height was determined using the extraction from the side-view images (RGB1) by counting the distance between the lowest and highest pixels. For RGB1, three photos were taken at each time point, with two 120º rotations of the plant between each imaging. Any morphological parameters associated with side views were calculated as the average of values derived from the three-angles (0º, 120º, and 240º). Chlorophyll fluorescence measurements A top-view Pulse Amplitude Modulated (PAM) chlorophyll a fluorometer (FluorCam FC-800MF, PSI, Czech Republic), equipped with a dark-adaptation tunnel and an imaging area of 800 × 800 mm, was used for chlorophyll fluorescence (ChlF) measurements (Tschiersch et al. 2017 ). The ChlF imaging was used to monitor plant physiological responses under well watered and water-deficit conditions by assessing parameters indicative of photosynthetic efficiency, including the quantum yield of PSII (QY) and stress-induced heat dissipation, measured as non-photochemical quenching (NPQ). Measurements were distributed across three batches over three consecutive days over a period of five weeks. The ChlF parameters were determined after a 30-min dark period using a quenching protocol developed with FluorCam 7.0 software (PSI, Drásov, Czech R.). The quenching protocol consisted of two phases: initial dark phase and actinic light phase. In the initial dark-adapted phase, after measuring the minimal fluorescence (Fo), a saturating pulse was applied to determine the maximum fluorescence (Fm). The variable fluorescence (Fv, calculated as Fm – Fo) was then divided by Fm to obtain the maximum PSII quantum yield (Fv/Fm). During the subsequent light-adapted phase, actinic light was applied, and saturating pulses were used to determine the maximum fluorescence under light (Fm′) and the steady-state fluorescence (Fs). The effective quantum yield of PSII at light steady state (QY_Lss) was calculated as (Fm′ – Fs) / Fm′, providing an estimate of the fraction of absorbed light energy used for photochemistry. Non-photochemical quenching at light steady state (NPQ_Lss) was calculated as (Fm – Fm′) / Fm′, providing a measure of the regulated dissipation of excess excitation energy as heat (Horton and Ruban 1992 ). Calculated parameters Image-based measurements were used to derive several traits that quantify plant responses to drought stress. To enable meaningful comparisons across genotypes, normalized differences were calculated for plant height and digital biomass. These represent the relative change, expressed as a percentage, between water deficit (WHC-30%) and well watered (WHC-80%) conditions, capturing the extent of growth reduction due to limited water availability. The normalized difference in plant height was calculated as: Normalized difference in plant height (%) = ((plant height of WHC-30% / plant height of WHC-80%) – 1) × 100 Digital biomass was calculated by combining pixel areas extracted from three side-view and one top-view image as an arbitrary volumetric measurement (Rahaman et al. 2017 ). Digital biomass is calculated as: Digital biomass =√ (average of three side view areas) × top area The normalized difference in digital biomass, recorded on the final day of the experiment, were calculated as: Normalized digital biomass (%) = ((digital biomass of WHC-30% / digital biomass of WHC-80%) – 1) × 100 Fresh weight and dry matter For the endpoint measurements at 39 DAG, the above-ground parts of the plants were collected into paper bags, and their fresh weights were recorded with a Mettler PM480 balance (GWB, Germany). After this, the plants were dried in a sample oven at 80°C for 72 h, and the dry matter content was recorded on the same balance after two weeks. These two endpoint measurements represent so-called biological biomass in our study. Water use efficiency (WUE) Daily watering by weighing allowed recording of the total amount of water given to each plant from 4 to 39 DAGs. We determined the biological WUE by dividing the endpoint dry matter produced (g) by the total amount of water (kg) supplied until 39 DAG. To establish non-destructive method for the same a digital WUE was calculated by dividing the digital biomass (volume) by the total amount of water (kg) provided until 37 DAG, i.e. , one day prior to imaging at 38 DAG. Data analysis Each RGB image underwent a fisheye correction and background exclusion. Data Analyzer v.3.4.17.3 (PSI, Drásov, Czech Republic) was used to manage the original images and data and their storage in the database. Canopy area, side-view areas, and height data were extracted from the green pixel measurements by MorphoAnalysis v.1.0.14.4 software (PSI, Drásov, Czech Republic). The ChlF data were analyzed via FluorCam 7 software (PSI.cz). The numerical data were processed via homemade pipelines via Python 3.13 ( https://www.python.org ). Two-way ANOVA was performed in JMP Pro 18, and all bar plots were generated using the Seaborn library (Waskom 2021 ). Pairwise differences among genotypes were assessed using Tukey's honest significant difference (HSD) post-hoc test, with compact letter displays (CLDs) generated via the split-then-absorb algorithm (Piepho 2004 ) implemented through the pairwise_tukeyhsd function in the Python statsmodels package (α = 0.05). In figures bars sharing the same letter are not significantly different from one another, while between-treatment differences for each accession were assessed using Welch's t-test; significance is indicated by brackets above each genotype pair. GAM modelling To assess the dynamic response of genotypes under water deficit, we applied generalized additive models (GAMs) with an autoregressive error structure. For each genotype and treatment conditions, a GAM was fitted using the mgcv package in R (Wood 2011 ), based on plant height and digital biomass data. The model residuals were inspected for temporal autocorrelation using the estimated lag-1 autocorrelation coefficient (ρ). When ρ exceeded 0.2, the model was refitted using mgcv::bam() including a first-order autoregressive [AR(1)] term to account for serial dependence across measurement days. This correction was applied to models for both plant height and digital biomass across all 44 genotypes, as all genotypes exceeded the ρ > 0.2 threshold in both traits (plant height: ρ = 0.458–0.775; digital biomass: ρ = 0.360–0.828). To quantify treatment effects, we used the plot_diff function from the itsadug package (van Rij et al. 2022 ) to estimate pointwise differences between WHC-30% and WHC-80% smooths across DAG, evaluated at 100 equally spaced time points. The difference was considered significant when its 95% confidence interval did not overlap zero. For plant height, the onset of divergence, defined as the first evaluated time point at which the lower bound of the 95% CI exceeded zero, was recorded as the drought impact day for each genotype, and genotypes were classified as early responders (divergence ≤ 20 DAG) or late responders (divergence > 20 DAG). The precision of this onset estimate is bounded by the evaluation grid resolution (~ 0.35 DAG per step), representing the maximum timing uncertainty for any genotype. Smooth response curves were visualized alongside observed means and standard deviations using ggplot2 (Wickham 2016 ). All analyses were conducted in R version 4.3.3 (R Core Team 2025 ). Results Plant height Water deficit treatment significantly reduced plant height for all the studied genotypes ( P < 0.001; Table S2 a), as visualized by the GAM-modelled trajectories in Figure S3 . The pointwise difference for plant height among genotypes in well watered and water deficit treatments, estimated from the fitted GAM smooths, are presented in Figure S4. Based on the first day of significant divergence, half of the genotypes showed an onset of treatment divergence before 20 DAG. In contrast, some genotypes such as Romfartuna, Lövånger, and Dalabona responded after 20 DAG to water deficit (Fig. 2 ). The results showed that water deficit treatment caused significant reduction in normalized plant height. Among the studied genotypes, Lövånger, Kärra, Romfartuna, Imatra me0101, and AP8308100101 had lowest reduction in normalized plant height under water deficit conditions (Table S2 b; Fig. 3 ). Digital biomass Water deficit treatment significantly reduced digital biomass for all the studied genotypes ( P < 0.001) (Table S2 a). For all genotypes the digital biomass was calculated as time series under both well watered and water deficit conditions (Figure S5). Among the studied genotypes, Romfartuna, and Primus had lowest reduction in digital biomass under water deficit conditions (Fig. 4 ). Biological WUE Under water deficit conditions, ILB 938/2 and Mélodie/2 (our reference genotypes for high WUE) had significantly higher biological WUE in both well watered and water deficit treatments than Aurora (reference genotype for low WUE) (Fig. 5 ; Table S2 b). Genotypes Lövånger, Fuego, and Suontakainen me0302 had highest biological WUE under water deficit conditions, although pairwise comparisons did not reveal statistically significant differences from all other genotypes (Table S2 b; Fig. 5 ). Digital WUE Similar to biological WUE, the digital WUE significantly varied across all faba bean genotypes (Table S2 b). Under water-deficit conditions, ILB 938/2 ranked among the genotypes with high digital WUE, while Aurora/2 is among those with the lowest value (Fig. 6 ). However, not all genotypes followed the same trend as observed for the biological WUE. Chlorophyll fluorescence imaging The ChlF measurements among all faba bean genotypes showed significant reduction in the effective quantum yield (QY_Lss) and an increase in non-photochemical quenching (NPQ_Lss) parameters, indicative of photosynthetic efficiency and stress induced heat dissipation, respectively (Fig. 7 ; Table S2 a and S2b). Relationships among the studied parameters The correlation matrix is presented in Fig. 8 showing the relationships among studied parameters. Both digital biomass and biological biomass were measured and correlated among the endpoint measurements, including fresh weight, dry matter, and other measured traits between the two water treatments. Digital biomass showed strong correlations with both dry matter ( r = 0.71) and fresh weight ( r = 0.66) under water deficit conditions, and even stronger correlations under well watered conditions ( r = 0.74). In contrast, plant height showed only a moderate correlation with biomass traits [fresh weight: r = 0.31–0.42 (water deficit and well watered conditions, respectively); dry matter: r = 0.47–0.53 (water deficit and well watered conditions, respectively)], indicating its limited reliability as a predictor of final biomass. Digital WUE showed only moderate correlation with biological WUE under water deficit conditions ( r = 0.56). The ChlF parameters showed overall weak correlations with the morphological traits under water deficit conditions. Discussion Phenotyping has long been a major bottleneck in effectively and precisely characterizing crop phenotypic diversity and environmental stress responses. It has been suggested that the HTPP platforms hold significant potential for accelerating breeding programs by enabling high-throughput quantification of variation of plant genetic resources essential for agricultural sustainability under climate change (Langstroff et al. 2022 ). Indoor HTPP systems, integrated with advanced imaging technologies, provide high-resolution, non-invasive monitoring of crop performance, allowing real-time data collection under stress. In recent years, HTPP platforms have enabled time-resolved screening of germplasm collections for stress responses. Thereby allowing for establishing more complex calculated traits. Here, we present a case study in which a relatively large faba bean germplasm collection was subjected to water deficit stress screened in a controlled environment, by using advanced imaging technologies. Building on multi-side imaging capabilities, we explored for the first time how digital biomass correlates with traditional biological measurements in faba bean. Our results revealed a strong positive correlation between the digital and biological biomass at the endpoint, validating that digital biomass can serve as a proxy for assessing the plant growth responses of faba bean during water deficit in controlled environments. Importantly, this relationship reflects phenotypic variation under imposed stress rather than intrinsic superiority of genotypes under field drought conditions. Digital biomass derived from digital top- and side-view images has been used to predict plant biological biomass evolution in barley and maize (e.g., Klukas et al. 2014 ; Chen et al. 2014 ; Neumann et al. 2015 ) and to monitor the effects of biostimulants on drought stress in tomatoes (e.g., Chovancek et al. 2025 ) and the salinity response in safflower (Thoday-Kennedy et al. 2021 ). WUE is one of the key traits in drought adaptation, as it reflects a plant’s ability to maintain productivity under limited water availability (Blum, 2009 ). High WUE indicates that a plant can assimilate more biomass per unit of water consumed, making it a valuable target for breeding crops better adapted to water-limited environments. WUE is a complex trait, and it is highly influenced by multiple physiological and morphological factors. An increase in WUE does not always translate directly into higher yield under field conditions (e.g., Hatfield and Dold 2019 ). In this study, we used ILB 938/2 and Mélodie/2 as known genotypes with high WUE and Aurora/2 as a poor WUE genotype (Khazaei et al. 2013 ; Khan et al. 2007 ; Mandour et al. 2023 ). Our results confirmed that these genotypes maintained their expected WUE also under controlled growing conditions. This highlights the ability of the phenotyping facility to screen a germplasm collection uniformly and precisely. The digital biomass and precise recording of the watering events during the entire experiment allowed us to calculate digital WUE but it was only moderately correlated with biological WUE under water deficit condition, indicating important limitations in its use as a proxy. This discrepancy likely reflects differences in tissue density that are not captured by image-derived traits, drought-induced changes in plant architecture affecting the relationship between projected area and biomass, and temporal mismatches between time-resolved digital measurements and endpoint-based biological biomass assessments. These results suggest that while digital WUE is useful for HTPP screening, it should be interpreted with caution and not considered a direct substitute for biological WUE. ChlF parameters showed weak correlations with growth-related traits, which may reflect a decoupling between photosynthetic performance and biomass accumulation under water deficit stress (Lawson and Vialet-Chabrand 2019 ). This may also be influenced by the use of discrete measurement time points, which do not fully capture the dynamic nature of photosynthetic responses to stress. ChlF parameters are highly sensitive to short-term environmental fluctuations and may therefore not directly translate into cumulative growth outcomes when assessed at limited time points (Murchie and Lawson 2013 ). Time-resolved analysis of ChlF parameters could provide more detailed insights into stress progression and genotype-specific responses, particularly when integrated with growth dynamics derived from imaging data. A large proportion of the genotypes used in this study are landraces recently collected from home gardens and small-scale farms in Finland and Sweden (Leino 2023 ) making them highly variable in the context of outcrossing rates. The HTPP facility enabled time-resolved classification of genotypes based on plant height trajectories under water deficit using GAMs and pointwise significance analysis. This temporal classification, applied to plant height data, revealed variation in drought sensitivity, with early responders showing rapid growth reduction shortly after water withdrawal, and late responders maintaining height growth until soil moisture dropped below critical levels. This temporal classification, applied to plant height trajectories, revealed variation in water stress sensitivity, with early responders showing rapid reduction in height growth shortly after water withdrawal, and late responders maintaining height increase until soil moisture dropped below critical levels. These patterns highlight differences in drought adaptation strategies among genotypes. Late water deficit responding landraces such as Lövånger and Romfartuna could potentially be used in breeding efforts toward more drought adapted faba beans. Both accessions originated from Sweden and are characterized by early maturity. The water deficit imposed in this study represents a gradual and moderate stress scenario. Therefore, the responses observed here reflect early-stage drought adaptation under controlled growing conditions. While this approach enables precise and reproducible phenotyping, it may not fully capture plant responses to more intense or prolonged drought stress typically experienced under field conditions. Our results allowed characterising faba bean genotypes that may be used as pre-breeding materials for breeding drought-adapted germplasm. Conclusions This study investigated the water deficit response of a faba bean collection using an automated indoor HTPP phenotyping facility under controlled environmental conditions. Our results, derived from optimized data analysis methodologies, demonstrate that high-throughput indoor phenotyping platforms can potentially screen germplasm collections for drought adaptation-related traits while also enabling the identification of additional traits such as WUE, digital biomass, and growth dynamics. These findings support the integration of HTPP into routine crop improvement pipelines, thereby accelerating the development of climate-resilient faba bean cultivars. Declarations Authorship contribution SP: Investigation, Formal analysis, Writing – review & editing. UC-N: Funding acquisition, Resources. MO: Investigation. KH: Writing – original draft, Writing – review & editing, Supervision, Investigation, Conceptualization, Resources. HK: Writing – original draft, Writing – review & editing, Supervision, Investigation, Conceptualization, Funding acquisition. AP: Funding acquisition, Resources. IMV: Funding acquisition, Resources. GP: Funding acquisition, Resources. MWL: Funding acquisition, Resources. All authors have commented and reviewed the final manuscript. Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Conflict of interest The authors declare no competing interests. Funding This project was funded by the Nordic Genetic Resource Center (NordGen), project ID 309203, and the Research Council of Finland, Academy projects, funding decision 363375 (Fabagen). Author Contribution SP: Investigation, Formal analysis, Writing – review & editing. UC-N: Funding acquisition, Resources. MO: Investigation. KH: Writing – original draft, Writing – review & editing, Supervision, Investigation, Conceptualization, Resources. HK: Writing – original draft, Writing – review & editing, Supervision, Investigation, Conceptualization, Funding acquisition. AP: Funding acquisition, Resources. IMV: Funding acquisition, Resources. GP: Funding acquisition, Resources. MWL: Funding acquisition, Resources. All authors have commented and reviewed the final manuscript. Acknowledgement We would like to thank Fereshteh Dehghani and Markku Tykkyläinen, technical assistants of the glasshouse of the University of Helsinki, for their kind assistance during the experiments. The project was initiated and carried out within the framework of NordGen’s grain legumes working group. We are also grateful to the staff of the seed laboratory at NordGen for preparing the main part of the seed material. Data Availability The data will be made available upon request. All imaging and numerical data generated in this study will be deposited in PHIS (Phenotyping Hybrid Information System, http://www.phis.inra.fr/), a platform designed for the organization, management, and sharing of plant phenotyping data. This approach aims to ensure highly structured data organization in alignment with the FAIR principles (Findable, Accessible, Interoperable, Reusable), thereby fully supporting open science practices. The faba bean genetic resources used in this study are available for order from the NordGen (https://nordic-baltic-genebanks.org). References Alexandersson E, Keinänen M, Chawade A, Himanen K (2018) Nordic research infrastructures for plant phenotyping. Agric Food Sci 27:7–16. https://doi.org/10.23986/afsci.68870 Ali MBM, Welna GC, Sallam A, Martsch R, Balko C, Gebser B, Sass O, Link W (2016) Association analyses to genetically improve drought and freezing tolerance of faba bean ( Vicia faba L.). Crop Sci 56:1036–1048. https://doi.org/10.2135/cropsci2015.08.0503 Al-Tamimi N, Langan P, Bernád V, Walsh J, Mangina E, Negrão S (2022) Capturing crop adaptation to abiotic stress using image-based technologies. Open Biol 12:210353. http://doi.org/10.1098/rsob.210353 Amede T, Kittlitz EV, Schubert S (1999) Differential drought responses of faba bean ( Vicia faba L.) inbred lines. J Agron Crop Sci 183:35–45. https://doi.org/10.1046/j.1439-037x.1999.00310.x Anshori MF, Dirpan A, Sitaresmi T, Rossi R, Farid M, Hairmansis A, Purwoko B, Suwarno WB, Nugraha Y (2023) An overview of image-based phenotyping as an adaptive 4.0 technology for studying plant abiotic stress: A bibliometric and literature review. Heliyon 9:e21650. https://doi.org/10.1016/j.heliyon.2023.e21650 Balko C, Torres AM, Gutierrez N (2023) Variability in drought stress response in a panel of 100 faba bean genotypes. Front Plant Sci 14:1236147. https://doi.org/10.3389/fpls.2023.1236147 Bevacqua E, Zappa G, Lehner F, Zscheischler J (2022) Precipitation trends determine future occurrences of compound hot–dry events. Nat Clim Change 12:350–355. https://doi.org/10.1038/s41558-022-01309-5 Blum A (2009) Effective use of water (EUW) and not water-use efficiency (WUE) is the target of crop yield improvement under drought stress. Field Crops Res 112: 119–123. https://doi.org/10.1016/j.fcr.2009.03.009 Chen D, Neumann K, Friedel K, Kilian B, Chen M, Altmann T, Klukas C (2014) Dissecting the phenotypic components of crop plant growth and drought responses based on high-throughput image analysis. Plant Cell 26:4636–6455. https://doi.org/10.1105/tpc.114.129601 Chovancek E, Poque S, Bayram E, Borhan E, Jokel M, Rantanen I-M, Haznedarogly B, Himanen K, Sirin S, Allahverdiyeva Y (2025) Stepwise processing of Chlorella sorokiniana confers plant biostimulant that reduces mineral fertilizer requirements. Bioresour Technol 418:131923. https://doi.org/10.1016/j.biortech.2024.131923 Dodig D, Božinović S, Nikolić A, Zorić M, Vančetović J, Ignjatović-Micić D, Delić N, Weigelt-Fischer K, Altmann T, Junker A (2021) Dynamics of maize vegetative growth and drought adaptability using image-based phenotyping under controlled conditions. Front Plant Sci 12:652116. https://doi.org/10.3389/fpls.2021.652116 Gill T, Gill SK, Saini DK, Chopra Y, de Koff JP, Sandhu KS (2022) A Comprehensive review of high throughput phenotyping and machine learning for plant stress phenotyping. Phenomics 2:156–183. https://doi.org/10.1007/s43657-022-00048-z Guadarrama-Escobar LM, Hunt J, Gurung A, Zarco-Tejada PJ, Shabala S, Camino C, Hernandez P, Pourkheirandish M (2024) Back to the future for drought tolerance. New Phytol 242:372–383. https://doi.org/10.1111/nph.19619 Hatfield JL, Dold C (2019) Water-use efficiency: Advances and challenges in a changing climate. Front Plant Sci 10:103. https://doi.org/10.3389/fpls.2019.00103 Horton P, Ruban AV (1992) Regulation of Photosystem II. Photosynthesis Res 34:375–385. https://doi.org/10.1007/BF00029812 Jensen ES, Peoples MB, Hauggaard-Nielsen H (2010) Faba bean in cropping systems. Field Crops Res 115:203–216. https://doi.org/10.1016/j.fcr.2009.10.008 Ji Y, Liu Z, Cui Y, Liu R, Chen Z, Zong X, Yang T (2024) Faba bean and pea harvest index estimations using aerial-based multimodal data and machine learning algorithms. Plant Physiol 194:1512–1526. https://doi.org/10.1093/plphys/kiad577 Khan HR, Link W, Hocking TJH, Stoddard FL (2007) Evaluation of physiological traits for improving drought tolerance in faba bean ( Vicia faba L.). Plant Soil 292:205–217. https://doi.org/10.1007/s11104-007-9217-5 Khan HR, Paull JG, Siddique KHM, Stoddard FL (2010) Faba bean breeding for drought-affected environments: A physiological and agronomic perspective. Field Crops Res 115:279–286. https://doi.org/10.1016/j.fcr.2009.09.003 Khazaei H, Link W, Street K, Stoddard FL (2018) ILB 938, a valuable faba bean ( Vicia faba L.) accession. Plant Genetic Resour 16:478–482. https://doi.org/10.1017/S1479262118000205 Khazaei H, O'Sullivan DM, Sillanpää MJ, Stoddard FL (2014) Use of synteny to identify candidate genes underlying QTL controlling stomatal traits in faba bean ( Vicia faba L.). Theor Appl Genet 127:2371–2385. https://doi.org/10.1007/s00122-014-2383-y Khazaei H, Street K, Bari A, Santanen A, Stoddard FL (2013) Do faba bean ( Vicia faba L.) accessions from environments with contrasting seasonal moisture availabilities differ in stomatal characteristics and related traits? Genet Resour Crop Evol 60:2343–2357. https://doi.org/10.1007/s10722-013-0002-4 Khazaei H, Vandenberg A (2020) Seed mineral composition and protein content of faba beans ( Vicia faba L.) with contrasting tannin contents. Agronomy 10:511. https://doi.org/10.3390/agronomy10040511 Kim M, Lee C, Hong S, Kim SL, Baek J-H, Kim K-H (2021) High-throughput phenotyping methods for breeding drought-tolerant crops. Int J of Mol Sci 22:8266. https://doi.org/10.3390/ijms22158266 Klukas C, Chen D, Pape J-M (2014) Integrated analysis platform: An open-source information system for high-throughput plant phenotyping. Plant Physiol 165:506–518. https://doi.org/10.1104/pp.113.233932 Langstroff A, Heuermann MC, Stahl A, Junker A (2022) Opportunities and limits of controlled-environment plant phenotyping for climate response traits. Theor Appl Genet 135:1–16. https://doi.org/10.1007/s00122-021-03892-1 Lawson T, Vialet-Chabrand S (2019) Speedy stomata, photosynthesis and plant water use efficiency. New Phytol 221:93–98. https://doi.org/10.1111/nph.15330 Leino MW (2023) Vicia faba plant genetic resources preserved in home gardens in Sweden. Legume perspectives 24:13–14. Mandour H, Khazaei H, Stoddard FL, Dodd IC (2023) Identifying physiological and genetic determinants of faba bean transpiration response to evaporative demand. Ann Bot 131:533–544. https://doi.org/10.1093/aob/mcad006 Mansoor S, Chung YS (2024) Functional phenotyping: Understanding the dynamic response of plants to drought stress. Current Plant Biol 38:100331. https://doi.org/10.1016/j.cpb.2024.100331 Mohammadi S, Uhlen AK, Lillemo M, Ergon Å, Shafiee S (2024) Enhancing phenotyping efficiency in faba bean breeding: integrating UAV imaging and machine learning. Precis Agric 25:1502–1528. https://doi.org/10.1007/s11119-024-10121-4 Murchie EH, Lawson T (2013) Chlorophyll fluorescence analysis: a guide to good practice and understanding some new applications. J Exp Bot 64:3983–3998. https://doi.org/10.1093/jxb/ert208 Nagel KA, Putz A, Gilmer F, Heinz K, Fischbach A, et al. (2012) GROWSCREEN-Rhizo is a novel phenotyping robot enabling simultaneous measurements of root and shoot growth for plants grown in soil-filled rhizotrons. Funct Plant Biol 39:891–904. https://doi.org/10.1071/FP12023 Neumann K, Klukas C, Friedel S, Rischbeck P, Chen D, Entzia, A, Stein N, Graner A, Kilian B (2015) Dissecting spatiotemporal biomass accumulation in barley under different water regimes using high-throughput image analysis. Plant Cell Environ 38:1980–1996. https://doi.org/10.1111/pce.12516 Peltonen-Sainio P, Juvonen J, Korhonen N, Parkkila P, Sorvali J, Gregow H (2021) Climate change, precipitation shifts and early summer drought: An irrigation tipping point for Finnish farmers? Climate Risk Manag 33:100334. https://doi.org/10.1016/j.crm.2021.100334 Piepho H-P (2004) An algorithm for a letter-based representation of all-pairwise comparisons. J Comput Graph Stat 13:456–466. https://doi.org/10.1198/1061860043515 Pieruschka R, Schurr U (2019) Plant phenotyping: Past, present, and future. Plant Phenom 2019:7507131. https://doi.org/10.34133/2019/7507131 Pollari M, Sipari N, Poque S, Himanen K, Mäkinen K (2022) Effects of poty-potexvirus synergism on growth, photosynthesis and metabolite status of Nicotiana benthamiana . Viruses 15:121. https://doi.org/10.3390/v15010121 R Core Team (2025) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. https://doi.org/10.32614/R Rahaman MM, Ahsan MA, Gillani Z, Chen M (2017) Digital biomass accumulation using high-throughput plant phenotype data analysis. J Integr Bioinform 14:20170028. https://doi.org/10.1515/jib-2017-0028 Roitch T, Himanen K, Chawade A, Jaakola L, Nehe A, Alexandersson E (2022) Functional phenomics for improved climate resilience in Nordic agriculture. J Exp Bot 73:5111–5127. https://doi.org/10.1093/jxb/erac246 Shapiguzov A, Punkkinen M, Laine T, Engström S, Aphalo PJ, Khazaei H (2025) Linking stomatal function with photosynthetic light reactions and stress response in faba bean. Environ Exp Bot 241:106290. https://doi.org/10.1016/j.envexpbot.2025.106290 Sozzani R, Busch W, Spalding EP, Benfey PN (2014) Advanced imaging techniques for the study of plant growth and development. Trends Plant Sci 19:304–310. https://doi.org/10.1016/j.tplants.2013.12.003 Spinoni J, Vogt JV, Naumann G, Barbosa P, Dosio A (2018) Will drought events become more frequent and severe in Europe? Int J of Climat 38:1718–1736. https://doi.org/10.1002/joc.5291 Swarup S, Cargill EJ, Crosby K, Flagel L, Kniskern J, Glenn KC (2021) Genetic diversity is indispensable for plant breeding to improve crops. Crop Sci 61:839–852. https://doi.org/10.1002/csc2.20377 Thoday-Kennedy E, Joshi S, Daetwyler HD, Hayden M, Hudson D, Spangenberg G, Kant S (2021) Digital phenotyping to delineate salinity response in safflower genotypes. Front Plant Sci 12:662498. https://doi.org/10.3389/fpls.2021.662498 Tschiersch H, Junker A, Meyer RC, Altmann T (2017) Establishment of integrated protocols for automated high throughput kinetic chlorophyll fluorescence analyses. Plant Methods 13:54. https://doi.org/10.1186/s13007-017-0204-4 van Rij J, Wieling M, Baayen R, van Rijn H (2022) itsadug: Interpreting time series and autocorrelated data using GAMMs. R package version 2.4.1. https://doi.org/10.32614/CRAN.package.itsadug Visakh R, Anand S, Reddy S, Jha U, Sah R, Beena R (2024) Precision phenotyping in crop science: from plant traits to gene discovery for climate-smart agriculture. Plant Breed https://doi.org/10.1111/pbr.13228 Waskom ML (2021) seaborn: statistical data visualization. Journal of Open Source Software 6:3021. https://doi.org/10.21105/joss.03021 Wickham H (2016) ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York, USA. R 550 package version 3.5.2. https://doi.org/10.32614/CRAN.package.ggpplot2 Wood SN (2011) Fast stable restricted maximum likelihood and marginal likelihood estimation of semiparametric generalized linear models. J Royal Statist Soc (B) 73:3–36. https://doi.org/10.1111/j.1467-9868.2010.00749.x Additional Declarations No competing interests reported. Supplementary Files TableS1.xlsx TableS2ab.xlsx SupportingFiguresGRACE.docx Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 19 May, 2026 Reviewers agreed at journal 12 May, 2026 Reviewers agreed at journal 12 May, 2026 Reviews received at journal 11 May, 2026 Reviewers agreed at journal 11 May, 2026 Reviewers agreed at journal 11 May, 2026 Reviewers agreed at journal 08 May, 2026 Reviewers agreed at journal 07 May, 2026 Reviewers invited by journal 07 May, 2026 Editor assigned by journal 07 May, 2026 Submission checks completed at journal 07 May, 2026 First submitted to journal 03 May, 2026 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-9598396","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":638832252,"identity":"15a3caa7-25ac-49f6-a3bf-f4fbc27cc375","order_by":0,"name":"Sylvain Poque","email":"","orcid":"","institution":"National Plant Phenotyping Infrastructure, Organismal and Evolutionary Research Programme, Helsinki Institute of Life Science, Biocenter Finland, University of Helsinki","correspondingAuthor":false,"prefix":"","firstName":"Sylvain","middleName":"","lastName":"Poque","suffix":""},{"id":638832253,"identity":"595655bd-eb62-42ac-ab46-0ad2df6720f1","order_by":1,"name":"Ulrika Carlson-Nilsson","email":"","orcid":"","institution":"Nordic Genetic Resource Center (NordGen)","correspondingAuthor":false,"prefix":"","firstName":"Ulrika","middleName":"","lastName":"Carlson-Nilsson","suffix":""},{"id":638832254,"identity":"be3ecc07-6864-49a0-8ef5-f1b65f52026c","order_by":2,"name":"Muhammad Omer","email":"","orcid":"","institution":"University of Helsinki","correspondingAuthor":false,"prefix":"","firstName":"Muhammad","middleName":"","lastName":"Omer","suffix":""},{"id":638832255,"identity":"847df96d-82de-4158-9f82-1f320093ce26","order_by":3,"name":"Anna Palmé","email":"","orcid":"","institution":"Nordic Genetic Resource Center (NordGen)","correspondingAuthor":false,"prefix":"","firstName":"Anna","middleName":"","lastName":"Palmé","suffix":""},{"id":638832256,"identity":"6f256b59-614d-4751-9229-931cd48cb6e4","order_by":4,"name":"Ingunn M. Vågen","email":"","orcid":"","institution":"Norwegian Institute of Bioeconomy Research (NIBIO)","correspondingAuthor":false,"prefix":"","firstName":"Ingunn","middleName":"M.","lastName":"Vågen","suffix":""},{"id":638832257,"identity":"a81ffaff-6ce6-4224-8ad1-2e39158c8ea8","order_by":5,"name":"Gert Poulsen","email":"","orcid":"","institution":"Frøsamlerne","correspondingAuthor":false,"prefix":"","firstName":"Gert","middleName":"","lastName":"Poulsen","suffix":""},{"id":638832258,"identity":"3f7839dc-d745-468e-9b1a-5d559f7de9cd","order_by":6,"name":"Matti W. Leino","email":"","orcid":"","institution":"Stockholm University","correspondingAuthor":false,"prefix":"","firstName":"Matti","middleName":"W.","lastName":"Leino","suffix":""},{"id":638832259,"identity":"0d43c5c5-b91a-44ae-b2cb-91959d369c82","order_by":7,"name":"Kristiina Himanen","email":"","orcid":"","institution":"National Plant Phenotyping Infrastructure, Organismal and Evolutionary Research Programme, Helsinki Institute of Life Science, Biocenter Finland, University of Helsinki","correspondingAuthor":false,"prefix":"","firstName":"Kristiina","middleName":"","lastName":"Himanen","suffix":""},{"id":638832263,"identity":"b28187f0-c30a-4090-a596-e6a4bb8c7d3a","order_by":8,"name":"Hamid Khazaei","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA60lEQVRIiWNgGAWjYBACAyA+wMOQkADmJVRYMEgA6Q8JxGs5IwHSwjiDkBYGuBbGNqgWfA4zZz+deOANQ1oef//hhw8ezpNInNnewNjwAI8Wy57cDQfnMOQUSxw4ZmyQuE0icTbPAcYGvA47kLvhMA9DRWLDwQYzCZCWeRIJ7A/wajn/FqJl/mH27z8S5wC1yD8gYMsNsC05iRuO8ZgxJDYAHSbBQEjLW6BfDNISN57hKZZIOCZhPLMnsRG/lvO5mz+8qUhOnHf++MaPP2psZGccP3yw8QceLVCNKDzGBoIaRsEoGAWjYBTgBwBjb1mC9/7bUgAAAABJRU5ErkJggg==","orcid":"","institution":"Natural Resources Institute Finland","correspondingAuthor":true,"prefix":"","firstName":"Hamid","middleName":"","lastName":"Khazaei","suffix":""}],"badges":[],"createdAt":"2026-05-03 08:10:08","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9598396/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9598396/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":109286934,"identity":"b6fc028c-e0fc-4b92-919a-02371981da04","added_by":"auto","created_at":"2026-05-15 02:37:46","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":132776,"visible":true,"origin":"","legend":"\u003cp\u003eWater stress treatment and measurements during the experiment. Two treatments were maintained from days after germination (DAG) 7 until harvest at DAG 39: a well-watered control (WHC-80%, upper track, blue) held continuously at 80% of the water-holding capacity (WHC), and water deficit treatment (lower track) underwent progressive water withholding from DAG 7, reaching 30% of the WHC at DAG 20 (gradient shading indicates the induction period), which was then maintained until DAG 31. A transient intensification to 20% of the WHC was imposed between DAG 31 and DAG 35 (dark red), after which soil water content was restored to 30% of the WHC until harvest. RGB canopy imaging was performed weekly (green triangles), chlorophyll fluorescence (ChlF) measurements on three consecutive days (DAG 33–35; purple squares), and plants harvested at DAG 39 (red star). Dashed vertical lines indicate key treatment transitions.\u003c/p\u003e","description":"","filename":"11.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9598396/v1/b26d0702663420afa8622ba2.jpg"},{"id":109296201,"identity":"dc32eb6d-fe25-4186-bd33-30dd128f834b","added_by":"auto","created_at":"2026-05-15 08:46:06","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":104260,"visible":true,"origin":"","legend":"\u003cp\u003eBar plot showing days after germination (DAG) of the water deficit impact on plant height. Significance of divergence in plant height between the two water holding capacity treatments was determined using pairwise comparisons of fitted smooths from generalized additive models. The dashed black line indicates 20 DAG, corresponding to the time when the water-deficit level was reached.\u003c/p\u003e","description":"","filename":"12.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9598396/v1/1f75fb822b4b5c7b893d0250.jpg"},{"id":109286937,"identity":"1b86792e-357d-410b-a840-9dae69709005","added_by":"auto","created_at":"2026-05-15 02:37:46","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":88583,"visible":true,"origin":"","legend":"\u003cp\u003eAverage normalized plant height (±SD) at the end point (38 DAG) among 44 faba bean genotypes under water deficit (WHC-30%). Bars sharing the same letter are not significantly different (Tukey's HSD, α = 0.05).\u003c/p\u003e","description":"","filename":"13.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9598396/v1/63d39c814b87feb811dd9dd5.jpg"},{"id":109297862,"identity":"5500f0d5-7c5d-4ad9-b9d0-8a2558f198b9","added_by":"auto","created_at":"2026-05-15 09:07:02","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":104253,"visible":true,"origin":"","legend":"\u003cp\u003eAverage normalized digital biomass (±SD) at the end point (38 DAG) among the 44 faba bean genotypes under water deficit (WHC-30%). Bars sharing the same letter are not significantly different (Tukey's HSD, α = 0.05).\u003c/p\u003e","description":"","filename":"14.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9598396/v1/f373f08f1d22bdb54a3d7165.jpg"},{"id":109286939,"identity":"3401b659-ad04-4119-96e5-e341b7fc7bf0","added_by":"auto","created_at":"2026-05-15 02:37:46","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":232550,"visible":true,"origin":"","legend":"\u003cp\u003eAverage biological water use efficiency (±SD) among 44 faba bean genotypes (\u003cstrong\u003eA\u003c/strong\u003e) under well watered (WHC-80%) and (\u003cstrong\u003eB\u003c/strong\u003e) water deficit (WHC-30%) conditions. Within each treatment, bars sharing the same letter are not significantly different (Tukey's HSD, α = 0.05).\u003c/p\u003e","description":"","filename":"15.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9598396/v1/23c6d81b0fcb265310065e09.jpg"},{"id":109296221,"identity":"868793ac-3d34-4381-9ed2-9a35d81e8852","added_by":"auto","created_at":"2026-05-15 08:46:16","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":245262,"visible":true,"origin":"","legend":"\u003cp\u003eAverage digital water use efficiency (±SD) calculated as digital biomass (volume) divided by total water consumed (kg) until 38 DAG, among 44 faba bean genotypes (\u003cstrong\u003eA\u003c/strong\u003e) under well watered (WHC-80%) and (\u003cstrong\u003eB\u003c/strong\u003e) water deficit (WHC-30%) conditions. Within each treatment, bars sharing the same letter are not significantly different (Tukey's HSD, α = 0.05).\u003c/p\u003e","description":"","filename":"16.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9598396/v1/38e911b19190a1829669a157.jpg"},{"id":109296560,"identity":"47fb6181-8cf3-46e4-9b93-019c40818f90","added_by":"auto","created_at":"2026-05-15 08:48:12","extension":"jpg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":269501,"visible":true,"origin":"","legend":"\u003cp\u003eChlorophyll fluorescence measured at 39 days after germination under well watered (blue) and water deficit (orange) conditions. (\u003cstrong\u003eA\u003c/strong\u003e) Effective quantum yield of PSII at light steady state (QY_Lss). (\u003cstrong\u003eB\u003c/strong\u003e) Non-photochemical quenching at light steady state (NPQ_Lss). Brackets above each accession indicate significant between-treatment differences (Welch's t-test: * \u003cem\u003eP\u003c/em\u003e\u0026lt;0.05, ** \u003cem\u003eP\u003c/em\u003e\u0026lt;0.01, *** \u003cem\u003eP\u003c/em\u003e\u0026lt;0.001; ns, not significant).\u003c/p\u003e","description":"","filename":"17.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9598396/v1/a2c290774b76024ff7e3cbd9.jpg"},{"id":109286941,"identity":"2792af12-e479-46cd-a405-7d166f764912","added_by":"auto","created_at":"2026-05-15 02:37:46","extension":"jpg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":269940,"visible":true,"origin":"","legend":"\u003cp\u003eCorrelation matrix of the studied parameters. The values are Pearson’s correlation coefficients (\u003cem\u003er\u003c/em\u003e) for traits under well watered (WHC-80%, top) and water-stress (WHC-30%, bottom) conditions. The average values of the measurements across genotypes and treatments are presented in Table S2b.\u003c/p\u003e","description":"","filename":"18.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9598396/v1/1ddf57be2249ebf2184f49ea.jpg"},{"id":109286905,"identity":"b45c05fa-7553-4ac3-b47b-f93086ba0411","added_by":"auto","created_at":"2026-05-15 02:37:35","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":216276,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9598396/v1/a927ad6b-f0dd-45ee-b7d8-8ab528dedd3f.pdf"},{"id":109286933,"identity":"2e01182e-76dd-48e1-8fe6-8aa63911fdc9","added_by":"auto","created_at":"2026-05-15 02:37:46","extension":"xlsx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":11378,"visible":true,"origin":"","legend":"","description":"","filename":"TableS1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9598396/v1/02e56741b02054db56625770.xlsx"},{"id":109296328,"identity":"6c760964-aaa1-42b3-8eea-97fe411f1aab","added_by":"auto","created_at":"2026-05-15 08:46:30","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":39876,"visible":true,"origin":"","legend":"","description":"","filename":"TableS2ab.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9598396/v1/960c4a82c96773705e4ccb0f.xlsx"},{"id":109286936,"identity":"e860301b-bdc9-45c4-a117-73ec2199008b","added_by":"auto","created_at":"2026-05-15 02:37:46","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":3515624,"visible":true,"origin":"","legend":"","description":"","filename":"SupportingFiguresGRACE.docx","url":"https://assets-eu.researchsquare.com/files/rs-9598396/v1/e19a6d8d1977c6ad2a330837.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Image-based phenotyping of faba bean genetic resources for water deficit responses under controlled conditions","fulltext":[{"header":"Introduction","content":"\u003cp\u003eFaba bean (\u003cem\u003eVicia faba\u003c/em\u003e L.) is an important cool-season grain legume crop grown worldwide because of its high seed protein content and versatility in agricultural systems, contributing to food security and sustainable agriculture (Jensen et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Khazaei and Vandenberg \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Nevertheless, faba bean is considered to be sensitive to drought stress (Khan et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2010\u003c/span\u003e) as the major environmental factor that negatively affects its growth, development, and yield potential. Global climate change has dramatically changed the frequency and patterns of rainfall over the last century. The decrease in precipitation due to infrequent rain events, combined with the predicted rise in atmospheric temperature, intensifies the frequency and severity of drought incidents, negatively affecting crop performance (Spinoni et al. \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Bevacqua et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). For example, in the Nordic region, early- to mid-summer droughts have become extremely likely due to climate change and cause significant crop yield losses (Peltonen-Sainio et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Therefore, development of drought-adapted crop varieties is essential, particularly in drought-susceptible crops such as faba bean.\u003c/p\u003e \u003cp\u003ePlant phenotyping, the process of systematic measurement and analysis of plant characteristics, is fundamental to understanding plant-environment interactions (Pieruschka and Schurr \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Visakh et al. \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Through phenotyping, drought-adapted germplasm can be identified to support the development of resilient cultivars and to advance understanding of stress adaptation mechanisms (Swarup et al. \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). This requires systematic phenotyping of diverse genotypes under water-deficit conditions to identify those that exhibit improved production in water-limited environments (Guadarrama-Escobar et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). A number of faba bean genotypes have been screened for this purpose (e.g., Amede et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Khazaei et al. \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Ali et al. \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). However, these studies have been largely constrained by the lack of reliable and standardized phenotyping tools and have often been conducted using low-throughput approaches.\u003c/p\u003e \u003cp\u003eHigh-Throughput Plant Phenotyping (HTPP) technologies have revolutionized plant breeding by enabling rapid, precise, and large-scale assessment of plant traits under controlled and field conditions. The HTPP platforms leverage advanced imaging technologies, such as visible light RedGreenBlue (RGB) imaging, thermal imaging, hyperspectral imaging, and chlorophyll fluorometry, to provide detailed insights into the physiological, morphological, and biochemical responses of plants to drought stress (Kim et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Gill et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Mansoor and Chung \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Sozzani et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). The HTPP facilitates the characterization of key traits related to shoot and root architecture (Nagel et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2012\u003c/span\u003e), and physiological adaptation to water stress (Shapiguzov et al. \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). These platforms allow researchers to identify drought adapted germplasm and enhance genetic studies by integrating phenotyping and genotyping data (Roitch et al. \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Al-Tamimi et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Anshori et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Integrated image analysis and machine learning has recently enabled the assessment of large populations of faba bean under natural field conditions (Ji et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Furthermore, digital phenotyping tools and protocols, for example to measure digital biomass, have been more developed and applied in cereals (Neumann et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Dodig et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) than in grain legumes.\u003c/p\u003e \u003cp\u003eDespite the potential of HTPP technologies, field-based phenotyping remains resource-intensive and is often challenged by the inherent variability in drought onset, duration, and intensity. To address these challenges, researchers have employed innovative approaches such as mobile field rainout shelters (Mohammadi et al. 2025; Balko et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Outdoor and indoor HTPP approaches are complementary, as indoor facilities enable controlled, high-precision monitoring of plant traits under controlled environmental conditions, while outdoor platforms capture plant performance under field conditions.\u003c/p\u003e \u003cp\u003eTo better understand the response of faba bean above-ground to water deficit, we utilized an indoor HTPP phenotyping facility, which allowed automated watering and precise water deficit treatment. The National Plant Phenotyping Infrastructure (NaPPI), a state-of-the-art indoor facility at the University of Helsinki, Finland, was used to monitor plant stress responses using visible-light RGB cameras and chlorophyll fluorometry (Alexandersson et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Pollari et al. \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Chovancek et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). The aims of this study were to (i) characterise phenotypic variation in a faba bean germplasm collection for growth dynamics, biological and digital biomass, water use efficiency (WUE), and chlorophyll fluorescence parameters under well watered and water deficit; (ii) assess the potential of image-derived digital biomass and digital WUE to function as non-destructive proxies for their biological counterparts in this species, and (iii) apply time-resolved generalized additive modelling to classify genotypes by the timing of growth responses to water deficit.\u003c/p\u003e"},{"header":"Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003ePlant material\u003c/h2\u003e \u003cp\u003eForty-four faba bean genotypes were used in this study. Information on genotype identification, origin and collection sites are presented in Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e. The material includes both large-seeded garden types and smaller seeded fodder types and originated from Nordic countries and northern Europe. Most of the genotypes represent landraces collected in Finland and Sweden (Leino \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and were obtained from the Nordic genebank (NordGen; Alnarp, Sweden). Genotypes ILB 938/2 and M\u0026eacute;lodie/2 were added to the set as benchmarks for high water use efficiency and Aurora/2 as a cultivar with low water use efficiency (Khazaei et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2014\u003c/span\u003e and \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eGrowing conditions\u003c/h3\u003e\n\u003cp\u003eThe plants were grown in a plexiglass greenhouse within the Viikki Plant Growth Facilities (ViGOR), University of Helsinki, Finland. The growth conditions were set to a 16 h light and 8 h dark photoperiod, with day and night air temperatures set at 21 C and 16 C, respectively, and a target relative humidity (RH) of approximately 60%. A photosynthetic photon flux density of approximately 250\u0026ndash;300 \u0026micro;mol.m⁻\u0026sup2;.s⁻\u0026sup1; was maintained at the canopy level by high-pressure sodium lamps. Environmental conditions inside the greenhouse were logged every minute throughout the experiment. Across the experimental period, mean daytime values were 22.8\u0026deg;C, 46.1% RH; mean nighttime values were 18.9\u0026deg;C and 55.4% RH. The corresponding mean vapour pressure deficit (VPD) was 1.53 kPa during the day and 0.99 kPa at night, indicating moderate to occasionally high evaporative demand. All conditions were uniform across treatments and genotypes. Seeds of all 44 genotypes were sown in three batches in October 2024, in 5 L pots filled with peat-based potting mix (Karkea Ruukutusseos, W R8014, Kekkil\u0026auml; Oy, Vantaa, Finland). Before sowing, they were inoculated with \u003cem\u003eRhizobium leguminosarum\u003c/em\u003e biovar. \u003cem\u003eviciae\u003c/em\u003e (Elomestari Oy, Tornio, Finland). The 264 pots were divided into three batches each containing a pot for the well watered treatment and for the water deficit treatment. To minimise potential positional effects within the greenhouse, pots were assigned to positions within each batch in a fully randomised order at the start of the experiment. Batches were then rotated weekly across greenhouse positions as a unit, ensuring that no batch remained in the same spatial position for more than one week throughout the experiment.\u003c/p\u003e\n\u003ch3\u003eWater deficit treatment\u003c/h3\u003e\n\u003cp\u003eDuring the first seven days after germination (DAG), all the pots were watered to reach field capacity to allow proper seed germination. Prior to the experiment, the water holding capacity (WHC) of the peat was determined by weighing equally soil filled pots, saturated with water and allowed to drain freely for 24 hours (equivalent to field capacity), representing 100% of the WHC. Then the same pots were completely oven-dried to 0% of the WHC and weighed. These established the relationship between pot weight and actual water content (%) used to monitor soil WHC% during the experiment (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). For the experiment the soil in each pot was weighed to ensure uniformity. This standardized approach minimized variation in soil volume, providing consistent water levels for all plants. The pots were covered with blue mat to limit evaporation and plant imaging. Automated watering by weighing allowed daily recording of pot weights, enabling calculation of both water loss between consecutive days (g) and the amount of water supplied. The plants were subsequently subjected to two distinct water regimes: 80% of the WHC as the control treatment and 30% of the WHC as the water deficit treatment, which was gradually achieved by 20 DAG. The 30% WHC level was selected based on preliminary trials in which 25% WHC caused excessive stress and rapid decline in several genotypes, whereas 30% WHC produced a stable, moderate drought level suitable for differentiating genotype responses without inducing premature plant failure. These watering levels were maintained for 10 days (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). After 31 DAG, the water deficit treatment was intensified by reducing the water level to 20% of the WHC for 5 days, in order to amplify stress severity and capture a broader range of genotypic responses. By 35 DAG, WHC was adjusted back to 30% and maintained at this level until 39 DAG (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e and Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003eImage-based phenotyping\u003c/h3\u003e\n\u003cp\u003eThe phenotyping was performed at the National Plant Phenotyping Infrastructure (NaPPI, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.helsinki.fi/en/infrastructures/national-plant-phenotyping\u003c/span\u003e\u003cspan address=\"http://www.helsinki.fi/en/infrastructures/national-plant-phenotyping\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The PlantScreen\u0026trade; Modular platform is an integrated system for automated plant imaging and management. Plants are set in individual transportation disks and moved between handling, imaging and watering stations. The feature of watering by weighing allows for management of water content of the soil (Chovancek et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). This phenotyping platform is equipped with imaging units, for top- and side-view visible light RGBs and top-view pulse amplitude modulated (PAM) chlorophyll fluorometer (FluorCam). Lighting conditions, plant positioning, and camera settings were fixed throughout the experiment.\u003c/p\u003e\n\u003ch3\u003eVisible-light digital imaging\u003c/h3\u003e\n\u003cp\u003eThe RGB imaging unit is a light-isolated box equipped with a turning table with precise angle positioning, RGB cameras using GigE uEye UI-5480SE-C/M\u0026thinsp;\u0026minus;\u0026thinsp;5 Megapixels QSXGA Camera with 1/2\u0026rdquo; CMOS Sensor (IDS, Germany) are supplemented with LED-based lighting source to ensure homogenous illumination of the imaged object. Top view RGB images of the plants were captured at resolution 2560 \u0026times; 1920 pixels and camera height was automatically adjusted by a light curtain unit measuring plant height. For side view projections, line scan camera captured images at resolution of 2560 \u0026times; 3476 pixels. Weekly RGB imaging was performed for all plants, with additional imaging of individual batches conducted separately within the same week, for more than five weeks. Top and side view RGB imaging were used to capture plant growth parameters such as the canopy area, side view area and height during the time series. The plant surface area was segmented using pixel colour thresholding of the RGB images and all growth parameters were automatically analysed by PlantScreen Data Analyzer v3.2.4.5 (PSI, Dr\u0026aacute;sov, Czech R.). To facilitate top-view canopy data extraction, blue rubber mats were placed around each plant in each pot. Representative examples of segmented side-view and top-view images under both watering treatments are shown in Figure \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e. During the experiment plant height was determined using the extraction from the side-view images (RGB1) by counting the distance between the lowest and highest pixels. For RGB1, three photos were taken at each time point, with two 120\u0026ordm; rotations of the plant between each imaging. Any morphological parameters associated with side views were calculated as the average of values derived from the three-angles (0\u0026ordm;, 120\u0026ordm;, and 240\u0026ordm;).\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eChlorophyll fluorescence measurements\u003c/h2\u003e \u003cp\u003eA top-view Pulse Amplitude Modulated (PAM) chlorophyll a fluorometer (FluorCam FC-800MF, PSI, Czech Republic), equipped with a dark-adaptation tunnel and an imaging area of 800 \u0026times; 800 mm, was used for chlorophyll fluorescence (ChlF) measurements (Tschiersch et al. \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). The ChlF imaging was used to monitor plant physiological responses under well watered and water-deficit conditions by assessing parameters indicative of photosynthetic efficiency, including the quantum yield of PSII (QY) and stress-induced heat dissipation, measured as non-photochemical quenching (NPQ). Measurements were distributed across three batches over three consecutive days over a period of five weeks.\u003c/p\u003e \u003cp\u003eThe ChlF parameters were determined after a 30-min dark period using a quenching protocol developed with FluorCam 7.0 software (PSI, Dr\u0026aacute;sov, Czech R.). The quenching protocol consisted of two phases: initial dark phase and actinic light phase. In the initial dark-adapted phase, after measuring the minimal fluorescence (Fo), a saturating pulse was applied to determine the maximum fluorescence (Fm). The variable fluorescence (Fv, calculated as Fm \u0026ndash; Fo) was then divided by Fm to obtain the maximum PSII quantum yield (Fv/Fm). During the subsequent light-adapted phase, actinic light was applied, and saturating pulses were used to determine the maximum fluorescence under light (Fm\u0026prime;) and the steady-state fluorescence (Fs). The effective quantum yield of PSII at light steady state (QY_Lss) was calculated as (Fm\u0026prime; \u0026ndash; Fs) / Fm\u0026prime;, providing an estimate of the fraction of absorbed light energy used for photochemistry. Non-photochemical quenching at light steady state (NPQ_Lss) was calculated as (Fm \u0026ndash; Fm\u0026prime;) / Fm\u0026prime;, providing a measure of the regulated dissipation of excess excitation energy as heat (Horton and Ruban \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e1992\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eCalculated parameters\u003c/h3\u003e\n\u003cp\u003eImage-based measurements were used to derive several traits that quantify plant responses to drought stress. To enable meaningful comparisons across genotypes, normalized differences were calculated for plant height and digital biomass. These represent the relative change, expressed as a percentage, between water deficit (WHC-30%) and well watered (WHC-80%) conditions, capturing the extent of growth reduction due to limited water availability.\u003c/p\u003e \u003cp\u003eThe normalized difference in plant height was calculated as:\u003c/p\u003e \u003cp\u003e \u003cem\u003eNormalized difference in plant height (%) = ((plant height of WHC-30% / plant height of WHC-80%) \u0026ndash; 1) \u0026times; 100\u003c/em\u003e \u003c/p\u003e \u003cp\u003eDigital biomass was calculated by combining pixel areas extracted from three side-view and one top-view image as an arbitrary volumetric measurement (Rahaman et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Digital biomass is calculated as:\u003c/p\u003e\n\u003ch3\u003eDigital biomass =√ (average of three side view areas) × top area\u003c/h3\u003e\n\u003cp\u003eThe normalized difference in digital biomass, recorded on the final day of the experiment, were calculated as:\u003c/p\u003e \u003cp\u003e \u003cem\u003eNormalized digital biomass (%) = ((digital biomass of WHC-30% / digital biomass of WHC-80%) \u0026ndash; 1) \u0026times; 100\u003c/em\u003e \u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eFresh weight and dry matter\u003c/h2\u003e \u003cp\u003eFor the endpoint measurements at 39 DAG, the above-ground parts of the plants were collected into paper bags, and their fresh weights were recorded with a Mettler PM480 balance (GWB, Germany). After this, the plants were dried in a sample oven at 80\u0026deg;C for 72 h, and the dry matter content was recorded on the same balance after two weeks. These two endpoint measurements represent so-called biological biomass in our study.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eWater use efficiency (WUE)\u003c/h2\u003e \u003cp\u003eDaily watering by weighing allowed recording of the total amount of water given to each plant from 4 to 39 DAGs. We determined the biological WUE by dividing the endpoint dry matter produced (g) by the total amount of water (kg) supplied until 39 DAG. To establish non-destructive method for the same a digital WUE was calculated by dividing the digital biomass (volume) by the total amount of water (kg) provided until 37 DAG, \u003cem\u003ei.e.\u003c/em\u003e, one day prior to imaging at 38 DAG.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eData analysis\u003c/h2\u003e \u003cp\u003eEach RGB image underwent a fisheye correction and background exclusion. Data Analyzer v.3.4.17.3 (PSI, Dr\u0026aacute;sov, Czech Republic) was used to manage the original images and data and their storage in the database. Canopy area, side-view areas, and height data were extracted from the green pixel measurements by MorphoAnalysis v.1.0.14.4 software (PSI, Dr\u0026aacute;sov, Czech Republic). The ChlF data were analyzed via FluorCam 7 software (PSI.cz). The numerical data were processed via homemade pipelines via Python 3.13 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.python.org\u003c/span\u003e\u003cspan address=\"https://www.python.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e).\u003c/span\u003e Two-way ANOVA was performed in JMP Pro 18, and all bar plots were generated using the Seaborn library (Waskom \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Pairwise differences among genotypes were assessed using Tukey's honest significant difference (HSD) post-hoc test, with compact letter displays (CLDs) generated via the split-then-absorb algorithm (Piepho \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2004\u003c/span\u003e) implemented through the pairwise_tukeyhsd function in the Python statsmodels package (α\u0026thinsp;=\u0026thinsp;0.05). In figures bars sharing the same letter are not significantly different from one another, while between-treatment differences for each accession were assessed using Welch's t-test; significance is indicated by brackets above each genotype pair.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eGAM modelling\u003c/h2\u003e \u003cp\u003eTo assess the dynamic response of genotypes under water deficit, we applied generalized additive models (GAMs) with an autoregressive error structure. For each genotype and treatment conditions, a GAM was fitted using the \u003cem\u003emgcv\u003c/em\u003e package in R (Wood \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2011\u003c/span\u003e), based on plant height and digital biomass data. The model residuals were inspected for temporal autocorrelation using the estimated lag-1 autocorrelation coefficient (ρ). When ρ exceeded 0.2, the model was refitted using mgcv::bam() including a first-order autoregressive [AR(1)] term to account for serial dependence across measurement days. This correction was applied to models for both plant height and digital biomass across all 44 genotypes, as all genotypes exceeded the ρ\u0026thinsp;\u0026gt;\u0026thinsp;0.2 threshold in both traits (plant height: ρ\u0026thinsp;=\u0026thinsp;0.458\u0026ndash;0.775; digital biomass: ρ\u0026thinsp;=\u0026thinsp;0.360\u0026ndash;0.828). To quantify treatment effects, we used the plot_diff function from the \u003cem\u003eitsadug\u003c/em\u003e package (van Rij et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) to estimate pointwise differences between WHC-30% and WHC-80% smooths across DAG, evaluated at 100 equally spaced time points. The difference was considered significant when its 95% confidence interval did not overlap zero. For plant height, the onset of divergence, defined as the first evaluated time point at which the lower bound of the 95% CI exceeded zero, was recorded as the drought impact day for each genotype, and genotypes were classified as early responders (divergence\u0026thinsp;\u0026le;\u0026thinsp;20 DAG) or late responders (divergence\u0026thinsp;\u0026gt;\u0026thinsp;20 DAG). The precision of this onset estimate is bounded by the evaluation grid resolution (~\u0026thinsp;0.35 DAG per step), representing the maximum timing uncertainty for any genotype. Smooth response curves were visualized alongside observed means and standard deviations using ggplot2 (Wickham \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). All analyses were conducted in R version 4.3.3 (R Core Team \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003ePlant height\u003c/h2\u003e \u003cp\u003eWater deficit treatment significantly reduced plant height for all the studied genotypes (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003ea), as visualized by the GAM-modelled trajectories in Figure \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e. The pointwise difference for plant height among genotypes in well watered and water deficit treatments, estimated from the fitted GAM smooths, are presented in Figure S4. Based on the first day of significant divergence, half of the genotypes showed an onset of treatment divergence before 20 DAG. In contrast, some genotypes such as Romfartuna, L\u0026ouml;v\u0026aring;nger, and Dalabona responded after 20 DAG to water deficit (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The results showed that water deficit treatment caused significant reduction in normalized plant height. Among the studied genotypes, L\u0026ouml;v\u0026aring;nger, K\u0026auml;rra, Romfartuna, Imatra me0101, and AP8308100101 had lowest reduction in normalized plant height under water deficit conditions (Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003eb; Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eDigital biomass\u003c/h2\u003e \u003cp\u003eWater deficit treatment significantly reduced digital biomass for all the studied genotypes (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001) (Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003ea). For all genotypes the digital biomass was calculated as time series under both well watered and water deficit conditions (Figure S5). Among the studied genotypes, Romfartuna, and Primus had lowest reduction in digital biomass under water deficit conditions (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eBiological WUE\u003c/h2\u003e \u003cp\u003eUnder water deficit conditions, ILB 938/2 and M\u0026eacute;lodie/2 (our reference genotypes for high WUE) had significantly higher biological WUE in both well watered and water deficit treatments than Aurora (reference genotype for low WUE) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e; Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003eb). Genotypes L\u0026ouml;v\u0026aring;nger, Fuego, and Suontakainen me0302 had highest biological WUE under water deficit conditions, although pairwise comparisons did not reveal statistically significant differences from all other genotypes (Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003eb; Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eDigital WUE\u003c/h2\u003e \u003cp\u003eSimilar to biological WUE, the digital WUE significantly varied across all faba bean genotypes (Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003eb). Under water-deficit conditions, ILB 938/2 ranked among the genotypes with high digital WUE, while Aurora/2 is among those with the lowest value (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). However, not all genotypes followed the same trend as observed for the biological WUE.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eChlorophyll fluorescence imaging\u003c/h2\u003e \u003cp\u003eThe ChlF measurements among all faba bean genotypes showed significant reduction in the effective quantum yield (QY_Lss) and an increase in non-photochemical quenching (NPQ_Lss) parameters, indicative of photosynthetic efficiency and stress induced heat dissipation, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e; Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003ea and S2b).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003eRelationships among the studied parameters\u003c/h2\u003e \u003cp\u003eThe correlation matrix is presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e showing the relationships among studied parameters. Both digital biomass and biological biomass were measured and correlated among the endpoint measurements, including fresh weight, dry matter, and other measured traits between the two water treatments. Digital biomass showed strong correlations with both dry matter (\u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.71) and fresh weight (\u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.66) under water deficit conditions, and even stronger correlations under well watered conditions (\u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.74). In contrast, plant height showed only a moderate correlation with biomass traits [fresh weight: \u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.31\u0026ndash;0.42 (water deficit and well watered conditions, respectively); dry matter: \u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.47\u0026ndash;0.53 (water deficit and well watered conditions, respectively)], indicating its limited reliability as a predictor of final biomass. Digital WUE showed only moderate correlation with biological WUE under water deficit conditions (\u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.56). The ChlF parameters showed overall weak correlations with the morphological traits under water deficit conditions.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003ePhenotyping has long been a major bottleneck in effectively and precisely characterizing crop phenotypic diversity and environmental stress responses. It has been suggested that the HTPP platforms hold significant potential for accelerating breeding programs by enabling high-throughput quantification of variation of plant genetic resources essential for agricultural sustainability under climate change (Langstroff et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Indoor HTPP systems, integrated with advanced imaging technologies, provide high-resolution, non-invasive monitoring of crop performance, allowing real-time data collection under stress. In recent years, HTPP platforms have enabled time-resolved screening of germplasm collections for stress responses. Thereby allowing for establishing more complex calculated traits.\u003c/p\u003e \u003cp\u003eHere, we present a case study in which a relatively large faba bean germplasm collection was subjected to water deficit stress screened in a controlled environment, by using advanced imaging technologies. Building on multi-side imaging capabilities, we explored for the first time how digital biomass correlates with traditional biological measurements in faba bean. Our results revealed a strong positive correlation between the digital and biological biomass at the endpoint, validating that digital biomass can serve as a proxy for assessing the plant growth responses of faba bean during water deficit in controlled environments. Importantly, this relationship reflects phenotypic variation under imposed stress rather than intrinsic superiority of genotypes under field drought conditions. Digital biomass derived from digital top- and side-view images has been used to predict plant biological biomass evolution in barley and maize (e.g., Klukas et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Chen et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Neumann et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) and to monitor the effects of biostimulants on drought stress in tomatoes (e.g., Chovancek et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) and the salinity response in safflower (Thoday-Kennedy et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eWUE is one of the key traits in drought adaptation, as it reflects a plant\u0026rsquo;s ability to maintain productivity under limited water availability (Blum, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). High WUE indicates that a plant can assimilate more biomass per unit of water consumed, making it a valuable target for breeding crops better adapted to water-limited environments. WUE is a complex trait, and it is highly influenced by multiple physiological and morphological factors. An increase in WUE does not always translate directly into higher yield under field conditions (e.g., Hatfield and Dold \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). In this study, we used ILB 938/2 and M\u0026eacute;lodie/2 as known genotypes with high WUE and Aurora/2 as a poor WUE genotype (Khazaei et al. \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Khan et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Mandour et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Our results confirmed that these genotypes maintained their expected WUE also under controlled growing conditions. This highlights the ability of the phenotyping facility to screen a germplasm collection uniformly and precisely. The digital biomass and precise recording of the watering events during the entire experiment allowed us to calculate digital WUE but it was only moderately correlated with biological WUE under water deficit condition, indicating important limitations in its use as a proxy. This discrepancy likely reflects differences in tissue density that are not captured by image-derived traits, drought-induced changes in plant architecture affecting the relationship between projected area and biomass, and temporal mismatches between time-resolved digital measurements and endpoint-based biological biomass assessments. These results suggest that while digital WUE is useful for HTPP screening, it should be interpreted with caution and not considered a direct substitute for biological WUE.\u003c/p\u003e \u003cp\u003eChlF parameters showed weak correlations with growth-related traits, which may reflect a decoupling between photosynthetic performance and biomass accumulation under water deficit stress (Lawson and Vialet-Chabrand \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). This may also be influenced by the use of discrete measurement time points, which do not fully capture the dynamic nature of photosynthetic responses to stress. ChlF parameters are highly sensitive to short-term environmental fluctuations and may therefore not directly translate into cumulative growth outcomes when assessed at limited time points (Murchie and Lawson \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). Time-resolved analysis of ChlF parameters could provide more detailed insights into stress progression and genotype-specific responses, particularly when integrated with growth dynamics derived from imaging data.\u003c/p\u003e \u003cp\u003eA large proportion of the genotypes used in this study are landraces recently collected from home gardens and small-scale farms in Finland and Sweden (Leino \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) making them highly variable in the context of outcrossing rates. The HTPP facility enabled time-resolved classification of genotypes based on plant height trajectories under water deficit using GAMs and pointwise significance analysis. This temporal classification, applied to plant height data, revealed variation in drought sensitivity, with early responders showing rapid growth reduction shortly after water withdrawal, and late responders maintaining height growth until soil moisture dropped below critical levels. This temporal classification, applied to plant height trajectories, revealed variation in water stress sensitivity, with early responders showing rapid reduction in height growth shortly after water withdrawal, and late responders maintaining height increase until soil moisture dropped below critical levels. These patterns highlight differences in drought adaptation strategies among genotypes. Late water deficit responding landraces such as L\u0026ouml;v\u0026aring;nger and Romfartuna could potentially be used in breeding efforts toward more drought adapted faba beans. Both accessions originated from Sweden and are characterized by early maturity. The water deficit imposed in this study represents a gradual and moderate stress scenario. Therefore, the responses observed here reflect early-stage drought adaptation under controlled growing conditions. While this approach enables precise and reproducible phenotyping, it may not fully capture plant responses to more intense or prolonged drought stress typically experienced under field conditions. Our results allowed characterising faba bean genotypes that may be used as pre-breeding materials for breeding drought-adapted germplasm.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eThis study investigated the water deficit response of a faba bean collection using an automated indoor HTPP phenotyping facility under controlled environmental conditions. Our results, derived from optimized data analysis methodologies, demonstrate that high-throughput indoor phenotyping platforms can potentially screen germplasm collections for drought adaptation-related traits while also enabling the identification of additional traits such as WUE, digital biomass, and growth dynamics. These findings support the integration of HTPP into routine crop improvement pipelines, thereby accelerating the development of climate-resilient faba bean cultivars.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eAuthorship contribution\u003c/h2\u003e \u003cp\u003eSP: Investigation, Formal analysis, Writing \u0026ndash; review \u0026amp; editing. UC-N: Funding acquisition, Resources. MO: Investigation. KH: Writing \u0026ndash; original draft, Writing \u0026ndash; review \u0026amp; editing, Supervision, Investigation, Conceptualization, Resources. HK: Writing \u0026ndash; original draft, Writing \u0026ndash; review \u0026amp; editing, Supervision, Investigation, Conceptualization, Funding acquisition. AP: Funding acquisition, Resources. IMV: Funding acquisition, Resources. GP: Funding acquisition, Resources. MWL: Funding acquisition, Resources. All authors have commented and reviewed the final manuscript.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e \u003cp\u003eNot applicable.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eConsent for publication\u003c/strong\u003e \u003cp\u003eNot applicable.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eConflict of interest\u003c/strong\u003e \u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eThis project was funded by the Nordic Genetic Resource Center (NordGen), project ID 309203, and the Research Council of Finland, Academy projects, funding decision 363375 (Fabagen).\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eSP: Investigation, Formal analysis, Writing \u0026ndash; review \u0026amp; editing. UC-N: Funding acquisition, Resources. MO: Investigation. KH: Writing \u0026ndash; original draft, Writing \u0026ndash; review \u0026amp; editing, Supervision, Investigation, Conceptualization, Resources. HK: Writing \u0026ndash; original draft, Writing \u0026ndash; review \u0026amp; editing, Supervision, Investigation, Conceptualization, Funding acquisition. AP: Funding acquisition, Resources. IMV: Funding acquisition, Resources. GP: Funding acquisition, Resources. MWL: Funding acquisition, Resources. All authors have commented and reviewed the final manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eWe would like to thank Fereshteh Dehghani and Markku Tykkyl\u0026auml;inen, technical assistants of the glasshouse of the University of Helsinki, for their kind assistance during the experiments. The project was initiated and carried out within the framework of NordGen\u0026rsquo;s grain legumes working group. We are also grateful to the staff of the seed laboratory at NordGen for preparing the main part of the seed material.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe data will be made available upon request. All imaging and numerical data generated in this study will be deposited in PHIS (Phenotyping Hybrid Information System, http://www.phis.inra.fr/), a platform designed for the organization, management, and sharing of plant phenotyping data. This approach aims to ensure highly structured data organization in alignment with the FAIR principles (Findable, Accessible, Interoperable, Reusable), thereby fully supporting open science practices. The faba bean genetic resources used in this study are available for order from the NordGen (https://nordic-baltic-genebanks.org).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAlexandersson E, Kein\u0026auml;nen M, Chawade A, Himanen K (2018) Nordic research infrastructures for plant phenotyping. Agric Food Sci 27:7\u0026ndash;16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.23986/afsci.68870\u003c/span\u003e\u003cspan address=\"10.23986/afsci.68870\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAli MBM, Welna GC, Sallam A, Martsch R, Balko C, Gebser B, Sass O, Link W (2016) Association analyses to genetically improve drought and freezing tolerance of faba bean (\u003cem\u003eVicia faba\u003c/em\u003e L.). Crop Sci 56:1036\u0026ndash;1048. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2135/cropsci2015.08.0503\u003c/span\u003e\u003cspan address=\"10.2135/cropsci2015.08.0503\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAl-Tamimi N, Langan P, Bern\u0026aacute;d V, Walsh J, Mangina E, Negr\u0026atilde;o S (2022) Capturing crop adaptation to abiotic stress using image-based technologies. Open Biol 12:210353. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://doi.org/10.1098/rsob.210353\u003c/span\u003e\u003cspan address=\"10.1098/rsob.210353\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAmede T, Kittlitz EV, Schubert S (1999) Differential drought responses of faba bean (\u003cem\u003eVicia faba\u003c/em\u003e L.) inbred lines. J Agron Crop Sci 183:35\u0026ndash;45. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1046/j.1439-037x.1999.00310.x\u003c/span\u003e\u003cspan address=\"10.1046/j.1439-037x.1999.00310.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnshori MF, Dirpan A, Sitaresmi T, Rossi R, Farid M, Hairmansis A, Purwoko B, Suwarno WB, Nugraha Y (2023) An overview of image-based phenotyping as an adaptive 4.0 technology for studying plant abiotic stress: A bibliometric and literature review. Heliyon 9:e21650. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.heliyon.2023.e21650\u003c/span\u003e\u003cspan address=\"10.1016/j.heliyon.2023.e21650\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBalko C, Torres AM, Gutierrez N (2023) Variability in drought stress response in a panel of 100 faba bean genotypes. Front Plant Sci 14:1236147. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fpls.2023.1236147\u003c/span\u003e\u003cspan address=\"10.3389/fpls.2023.1236147\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBevacqua E, Zappa G, Lehner F, Zscheischler J (2022) Precipitation trends determine future occurrences of compound hot\u0026ndash;dry events. Nat Clim Change 12:350\u0026ndash;355. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41558-022-01309-5\u003c/span\u003e\u003cspan address=\"10.1038/s41558-022-01309-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBlum A (2009) Effective use of water (EUW) and not water-use efficiency (WUE) is the target of crop yield improvement under drought stress. Field Crops Res 112: 119\u0026ndash;123. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.fcr.2009.03.009\u003c/span\u003e\u003cspan address=\"10.1016/j.fcr.2009.03.009\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen D, Neumann K, Friedel K, Kilian B, Chen M, Altmann T, Klukas C (2014) Dissecting the phenotypic components of crop plant growth and drought responses based on high-throughput image analysis. Plant Cell 26:4636\u0026ndash;6455. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1105/tpc.114.129601\u003c/span\u003e\u003cspan address=\"10.1105/tpc.114.129601\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChovancek E, Poque S, Bayram E, Borhan E, Jokel M, Rantanen I-M, Haznedarogly B, Himanen K, Sirin S, Allahverdiyeva Y (2025) Stepwise processing of \u003cem\u003eChlorella sorokiniana\u003c/em\u003e confers plant biostimulant that reduces mineral fertilizer requirements. Bioresour Technol 418:131923. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.biortech.2024.131923\u003c/span\u003e\u003cspan address=\"10.1016/j.biortech.2024.131923\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDodig D, Božinović S, Nikolić A, Zorić M, Vančetović J, Ignjatović-Micić D, Delić N, Weigelt-Fischer K, Altmann T, Junker A (2021) Dynamics of maize vegetative growth and drought adaptability using image-based phenotyping under controlled conditions. Front Plant Sci 12:652116. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fpls.2021.652116\u003c/span\u003e\u003cspan address=\"10.3389/fpls.2021.652116\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGill T, Gill SK, Saini DK, Chopra Y, de Koff JP, Sandhu KS (2022) A Comprehensive review of high throughput phenotyping and machine learning for plant stress phenotyping. Phenomics 2:156\u0026ndash;183. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s43657-022-00048-z\u003c/span\u003e\u003cspan address=\"10.1007/s43657-022-00048-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuadarrama-Escobar LM, Hunt J, Gurung A, Zarco-Tejada PJ, Shabala S, Camino C, Hernandez P, Pourkheirandish M (2024) Back to the future for drought tolerance. New Phytol 242:372\u0026ndash;383. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/nph.19619\u003c/span\u003e\u003cspan address=\"10.1111/nph.19619\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHatfield JL, Dold C (2019) Water-use efficiency: Advances and challenges in a changing climate. Front Plant Sci 10:103. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fpls.2019.00103\u003c/span\u003e\u003cspan address=\"10.3389/fpls.2019.00103\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHorton P, Ruban AV (1992) Regulation of Photosystem II. Photosynthesis Res 34:375\u0026ndash;385. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/BF00029812\u003c/span\u003e\u003cspan address=\"10.1007/BF00029812\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJensen ES, Peoples MB, Hauggaard-Nielsen H (2010) Faba bean in cropping systems. Field Crops Res 115:203\u0026ndash;216. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.fcr.2009.10.008\u003c/span\u003e\u003cspan address=\"10.1016/j.fcr.2009.10.008\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJi Y, Liu Z, Cui Y, Liu R, Chen Z, Zong X, Yang T (2024) Faba bean and pea harvest index estimations using aerial-based multimodal data and machine learning algorithms. Plant Physiol 194:1512\u0026ndash;1526. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/plphys/kiad577\u003c/span\u003e\u003cspan address=\"10.1093/plphys/kiad577\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhan HR, Link W, Hocking TJH, Stoddard FL (2007) Evaluation of physiological traits for improving drought tolerance in faba bean (\u003cem\u003eVicia faba\u003c/em\u003e L.). Plant Soil 292:205\u0026ndash;217. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11104-007-9217-5\u003c/span\u003e\u003cspan address=\"10.1007/s11104-007-9217-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhan HR, Paull JG, Siddique KHM, Stoddard FL (2010) Faba bean breeding for drought-affected environments: A physiological and agronomic perspective. Field Crops Res 115:279\u0026ndash;286. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.fcr.2009.09.003\u003c/span\u003e\u003cspan address=\"10.1016/j.fcr.2009.09.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhazaei H, Link W, Street K, Stoddard FL (2018) ILB 938, a valuable faba bean (\u003cem\u003eVicia faba\u003c/em\u003e L.) accession. Plant Genetic Resour 16:478\u0026ndash;482. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1017/S1479262118000205\u003c/span\u003e\u003cspan address=\"10.1017/S1479262118000205\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhazaei H, O'Sullivan DM, Sillanp\u0026auml;\u0026auml; MJ, Stoddard FL (2014) Use of synteny to identify candidate genes underlying QTL controlling stomatal traits in faba bean (\u003cem\u003eVicia faba\u003c/em\u003e L.). Theor Appl Genet 127:2371\u0026ndash;2385. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00122-014-2383-y\u003c/span\u003e\u003cspan address=\"10.1007/s00122-014-2383-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhazaei H, Street K, Bari A, Santanen A, Stoddard FL (2013) Do faba bean (\u003cem\u003eVicia faba\u003c/em\u003e L.) accessions from environments with contrasting seasonal moisture availabilities differ in stomatal characteristics and related traits? Genet Resour Crop Evol 60:2343\u0026ndash;2357. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10722-013-0002-4\u003c/span\u003e\u003cspan address=\"10.1007/s10722-013-0002-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhazaei H, Vandenberg A (2020) Seed mineral composition and protein content of faba beans (\u003cem\u003eVicia faba\u003c/em\u003e L.) with contrasting tannin contents. Agronomy 10:511. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/agronomy10040511\u003c/span\u003e\u003cspan address=\"10.3390/agronomy10040511\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim M, Lee C, Hong S, Kim SL, Baek J-H, Kim K-H (2021) High-throughput phenotyping methods for breeding drought-tolerant crops. Int J of Mol Sci 22:8266. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/ijms22158266\u003c/span\u003e\u003cspan address=\"10.3390/ijms22158266\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlukas C, Chen D, Pape J-M (2014) Integrated analysis platform: An open-source information system for high-throughput plant phenotyping. Plant Physiol 165:506\u0026ndash;518. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1104/pp.113.233932\u003c/span\u003e\u003cspan address=\"10.1104/pp.113.233932\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLangstroff A, Heuermann MC, Stahl A, Junker A (2022) Opportunities and limits of controlled-environment plant phenotyping for climate response traits. Theor Appl Genet 135:1\u0026ndash;16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00122-021-03892-1\u003c/span\u003e\u003cspan address=\"10.1007/s00122-021-03892-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLawson T, Vialet-Chabrand S (2019) Speedy stomata, photosynthesis and plant water use efficiency. New Phytol 221:93\u0026ndash;98. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/nph.15330\u003c/span\u003e\u003cspan address=\"10.1111/nph.15330\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeino MW (2023) \u003cem\u003eVicia faba\u003c/em\u003e plant genetic resources preserved in home gardens in Sweden. Legume perspectives 24:13\u0026ndash;14.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMandour H, Khazaei H, Stoddard FL, Dodd IC (2023) Identifying physiological and genetic determinants of faba bean transpiration response to evaporative demand. Ann Bot 131:533\u0026ndash;544. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/aob/mcad006\u003c/span\u003e\u003cspan address=\"10.1093/aob/mcad006\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMansoor S, Chung YS (2024) Functional phenotyping: Understanding the dynamic response of plants to drought stress. Current Plant Biol 38:100331. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.cpb.2024.100331\u003c/span\u003e\u003cspan address=\"10.1016/j.cpb.2024.100331\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMohammadi S, Uhlen AK, Lillemo M, Ergon \u0026Aring;, Shafiee S (2024) Enhancing phenotyping efficiency in faba bean breeding: integrating UAV imaging and machine learning. Precis Agric 25:1502\u0026ndash;1528. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11119-024-10121-4\u003c/span\u003e\u003cspan address=\"10.1007/s11119-024-10121-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMurchie EH, Lawson T (2013) Chlorophyll fluorescence analysis: a guide to good practice and understanding some new applications. J Exp Bot 64:3983\u0026ndash;3998. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/jxb/ert208\u003c/span\u003e\u003cspan address=\"10.1093/jxb/ert208\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNagel KA, Putz A, Gilmer F, Heinz K, Fischbach A, et al. (2012) GROWSCREEN-Rhizo is a novel phenotyping robot enabling simultaneous measurements of root and shoot growth for plants grown in soil-filled rhizotrons. Funct Plant Biol 39:891\u0026ndash;904. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1071/FP12023\u003c/span\u003e\u003cspan address=\"10.1071/FP12023\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNeumann K, Klukas C, Friedel S, Rischbeck P, Chen D, Entzia, A, Stein N, Graner A, Kilian B (2015) Dissecting spatiotemporal biomass accumulation in barley under different water regimes using high-throughput image analysis. Plant Cell Environ 38:1980\u0026ndash;1996. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/pce.12516\u003c/span\u003e\u003cspan address=\"10.1111/pce.12516\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePeltonen-Sainio P, Juvonen J, Korhonen N, Parkkila P, Sorvali J, Gregow H (2021) Climate change, precipitation shifts and early summer drought: An irrigation tipping point for Finnish farmers? Climate Risk Manag 33:100334. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.crm.2021.100334\u003c/span\u003e\u003cspan address=\"10.1016/j.crm.2021.100334\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePiepho H-P (2004) An algorithm for a letter-based representation of all-pairwise comparisons. J Comput Graph Stat 13:456\u0026ndash;466. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1198/1061860043515\u003c/span\u003e\u003cspan address=\"10.1198/1061860043515\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePieruschka R, Schurr U (2019) Plant phenotyping: Past, present, and future. Plant Phenom 2019:7507131. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.34133/2019/7507131\u003c/span\u003e\u003cspan address=\"10.34133/2019/7507131\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePollari M, Sipari N, Poque S, Himanen K, M\u0026auml;kinen K (2022) Effects of poty-potexvirus synergism on growth, photosynthesis and metabolite status of \u003cem\u003eNicotiana benthamiana\u003c/em\u003e. Viruses 15:121. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/v15010121\u003c/span\u003e\u003cspan address=\"10.3390/v15010121\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eR Core Team (2025) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. \u003cdiv class=\"ExternalRefDOI\"\u003ehttps://doi.org/10.32614/R\u003c/div\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRahaman MM, Ahsan MA, Gillani Z, Chen M (2017) Digital biomass accumulation using high-throughput plant phenotype data analysis. J Integr Bioinform 14:20170028. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1515/jib-2017-0028\u003c/span\u003e\u003cspan address=\"10.1515/jib-2017-0028\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRoitch T, Himanen K, Chawade A, Jaakola L, Nehe A, Alexandersson E (2022) Functional phenomics for improved climate resilience in Nordic agriculture. J Exp Bot 73:5111\u0026ndash;5127. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/jxb/erac246\u003c/span\u003e\u003cspan address=\"10.1093/jxb/erac246\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShapiguzov A, Punkkinen M, Laine T, Engstr\u0026ouml;m S, Aphalo PJ, Khazaei H (2025) Linking stomatal function with photosynthetic light reactions and stress response in faba bean. Environ Exp Bot 241:106290. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.envexpbot.2025.106290\u003c/span\u003e\u003cspan address=\"10.1016/j.envexpbot.2025.106290\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSozzani R, Busch W, Spalding EP, Benfey PN (2014) Advanced imaging techniques for the study of plant growth and development. Trends Plant Sci 19:304\u0026ndash;310. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.tplants.2013.12.003\u003c/span\u003e\u003cspan address=\"10.1016/j.tplants.2013.12.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSpinoni J, Vogt JV, Naumann G, Barbosa P, Dosio A (2018) Will drought events become more frequent and severe in Europe? Int J of Climat 38:1718\u0026ndash;1736. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/joc.5291\u003c/span\u003e\u003cspan address=\"10.1002/joc.5291\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSwarup S, Cargill EJ, Crosby K, Flagel L, Kniskern J, Glenn KC (2021) Genetic diversity is indispensable for plant breeding to improve crops. Crop Sci 61:839\u0026ndash;852. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/csc2.20377\u003c/span\u003e\u003cspan address=\"10.1002/csc2.20377\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThoday-Kennedy E, Joshi S, Daetwyler HD, Hayden M, Hudson D, Spangenberg G, Kant S (2021) Digital phenotyping to delineate salinity response in safflower genotypes. Front Plant Sci 12:662498. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fpls.2021.662498\u003c/span\u003e\u003cspan address=\"10.3389/fpls.2021.662498\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTschiersch H, Junker A, Meyer RC, Altmann T (2017) Establishment of integrated protocols for automated high throughput kinetic chlorophyll fluorescence analyses. Plant Methods 13:54. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s13007-017-0204-4\u003c/span\u003e\u003cspan address=\"10.1186/s13007-017-0204-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Rij J, Wieling M, Baayen R, van Rijn H (2022) itsadug: Interpreting time series and autocorrelated data using GAMMs. R package version 2.4.1. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.32614/CRAN.package.itsadug\u003c/span\u003e\u003cspan address=\"10.32614/CRAN.package.itsadug\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVisakh R, Anand S, Reddy S, Jha U, Sah R, Beena R (2024) Precision phenotyping in crop science: from plant traits to gene discovery for climate-smart agriculture. Plant Breed \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/pbr.13228\u003c/span\u003e\u003cspan address=\"10.1111/pbr.13228\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWaskom ML (2021) seaborn: statistical data visualization. Journal of Open Source Software 6:3021. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.21105/joss.03021\u003c/span\u003e\u003cspan address=\"10.21105/joss.03021\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWickham H (2016) ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York, USA. R 550 package version 3.5.2. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.32614/CRAN.package.ggpplot2\u003c/span\u003e\u003cspan address=\"10.32614/CRAN.package.ggpplot2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWood SN (2011) Fast stable restricted maximum likelihood and marginal likelihood estimation of semiparametric generalized linear models. J Royal Statist Soc (B) 73:3\u0026ndash;36. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1467-9868.2010.00749.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1467-9868.2010.00749.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":true,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"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":"genetic-resources-and-crop-evolution","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"gres","sideBox":"Learn more about [Genetic Resources and Crop Evolution](https://www.springer.com/journal/10722)","snPcode":"10722","submissionUrl":"https://submission.nature.com/new-submission/10722/3","title":"Genetic Resources and Crop Evolution","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"drought stress, plant phenomics, legumes, digital biomass, water use efficiency","lastPublishedDoi":"10.21203/rs.3.rs-9598396/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9598396/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eFaba bean (\u003cem\u003eVicia faba\u003c/em\u003e L.) has great potential to contribute to sustainable agriculture and protein security globally but is known to be very sensitive to drought stress. Uncovering drought-resilient germplasm is critical for developing resilient cultivars and advancing our understanding of the mechanisms underlying stress adaptation. However, high-throughput plant phenotyping under stress conditions remain a major bottleneck in crop genetics and breeding programs. In this study, a multi-sensor indoor phenotyping platform was used to assess 44 faba bean genotypes under water deficit conditions. Standardized, monitored stress conditions were achieved by watering-by-weighing for drought onset, duration, and intensities allowing genotype-level comparisons. The genotypes showed a range of stress responses in growth and physiology, including traits such as plant height, biomass, water use efficiency (WUE), and chlorophyll fluorescence parameters. Digital biomass, derived from combined top- and side-view plant imaging, was strongly correlated with biological biomass at the experimental endpoint, validating its use as a non-destructive proxy for growth assessment in faba bean. Time-resolved generalized additive modelling further revealed genotype-specific differences in the timing and magnitude of water deficit response. Genotypes that maintained growth and WUE under water deficit conditions may serve as valuable pre-breeding materials for development of drought-adapted faba bean.\u003c/p\u003e","manuscriptTitle":"Image-based phenotyping of faba bean genetic resources for water deficit responses under controlled conditions","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-15 02:37:31","doi":"10.21203/rs.3.rs-9598396/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2026-05-19T08:39:00+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"161931923954544013784503972253581645762","date":"2026-05-12T08:17:13+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"78910168492317252024285361226361639696","date":"2026-05-12T06:48:47+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-11T11:06:41+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"333821099518433746836803050242350155620","date":"2026-05-11T09:32:16+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"136105875045215289766305502025114826900","date":"2026-05-11T07:37:06+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"139824071998015162940116947069621516589","date":"2026-05-08T06:40:42+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"176245943529322571370754275473140263812","date":"2026-05-07T10:28:49+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-05-07T06:28:10+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-05-07T06:18:41+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-05-07T06:17:41+00:00","index":"","fulltext":""},{"type":"submitted","content":"Genetic Resources and Crop Evolution","date":"2026-05-03T07:51:58+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"genetic-resources-and-crop-evolution","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"gres","sideBox":"Learn more about [Genetic Resources and Crop Evolution](https://www.springer.com/journal/10722)","snPcode":"10722","submissionUrl":"https://submission.nature.com/new-submission/10722/3","title":"Genetic Resources and Crop Evolution","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"f308945e-fa3b-4ad8-946e-7eb595a0d525","owner":[],"postedDate":"May 15th, 2026","published":true,"recentEditorialEvents":[{"type":"editorInvitedReview","content":"","date":"2026-05-19T08:39:00+00:00","index":30,"fulltext":""},{"type":"reviewerAgreed","content":"161931923954544013784503972253581645762","date":"2026-05-12T08:17:13+00:00","index":29,"fulltext":""},{"type":"reviewerAgreed","content":"78910168492317252024285361226361639696","date":"2026-05-12T06:48:47+00:00","index":27,"fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-11T11:06:41+00:00","index":26,"fulltext":""},{"type":"reviewerAgreed","content":"333821099518433746836803050242350155620","date":"2026-05-11T09:32:16+00:00","index":25,"fulltext":""},{"type":"reviewerAgreed","content":"136105875045215289766305502025114826900","date":"2026-05-11T07:37:06+00:00","index":24,"fulltext":""},{"type":"reviewerAgreed","content":"139824071998015162940116947069621516589","date":"2026-05-08T06:40:42+00:00","index":23,"fulltext":""},{"type":"reviewerAgreed","content":"176245943529322571370754275473140263812","date":"2026-05-07T10:28:49+00:00","index":20,"fulltext":""},{"type":"reviewersInvited","content":"15","date":"2026-05-07T06:28:10+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-05-07T06:18:41+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-05-07T06:17:41+00:00","index":"","fulltext":""},{"type":"submitted","content":"Genetic Resources and Crop Evolution","date":"2026-05-03T07:51:58+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-15T02:37:31+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-15 02:37:31","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9598396","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9598396","identity":"rs-9598396","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}