Journal of Applied Botany and Food Quality 92, 343 - 354 (2019), DOI:10.5073/JABFQ.2019.092.046 1Department of Beverage Research, Geisenheim University, Geisenheim, Germany 2Department of Vegetable Crops, Geisenheim University, Geisenheim, Germany 3Department of Biometry and Population Genetics, Justus Liebig University, Giessen, Germany 4Leibniz Institute of Vegetable and Ornamental Crops, Großbeeren, Germany Chemical composition of field grown radish (Raphanus sativus L. var. sativus) as influenced by season and moderately reduced water supply Christine Schlering1, 2*, Helmut Dietrich1, Matthias Frisch3, Monika Schreiner4, Ralf Schweiggert1, Frank Will1, Jana Zinkernagel2 (Submitted: June 25, 2019; Accepted: October 25, 2019) * Corresponding author Summary Seasonal variations in water availability as increasingly provoked by climate change pose severe challenges for vegetable production, particularly for crops requiring reliable and high water supply for achieving satisfactory quality. In contrast to most previous studies applying severe water deficits, we examined the effects of moderate water deficits on the chemical composition of red radish tubers dur- ing three consecutive years with variable climatic conditions. Radish were cultivated in open field, applying two different water supply treatments and following a randomized block design comprising four sets of six plots each. The resulting water reductions of 3-20% led to a significant increase of dry matter-based myo-inositol levels, where- as those of selected minerals and trace elements, phenolics and gluco- sinolates decreased. Anthocyanin levels remained unchanged. Fresh- matter related levels of most analytes increased upon reduced water treatments due to higher dry matter contents. While pigment levels in radish remained unchanged, mild water deficit affected other quality- related parameters such as pungency-related glucosinolates. Keywords: Vegetables, Polyols, Glucosinolates, Anthocyanins, Drought, Water Deficit Introduction Plant responses to changing environmental conditions are complex and vary between cultivation season and year. Abiotic stress factors such as water deficiency induce several changes in various physio- logical plant processes (Prasad et al. 2008), generally related to negative effects on plant growth, yield and quality. Future scenarios suggest that increasing global surface temperatures will very likely lead to higher incidences of precipitation-free periods such as for example in Germany (schmidt and Zinkernagel, 2017) or South and West European countries (nikulin et al., 2011), particularly be- cause rainfall patterns are shifting upon temperature changes. For example, 2018 proved to be the warmest and one of the driest years in Germany since weather records began. As a result of climate change, growing conditions undergo changes and inadequate water supply might become an increasingly important agricultural challenge for fruit and vegetable production, and drought-stress tolerant cultivars might be desired. Climate change conditions including uncommon temporal limitations or excess in water supply, increased regional mean temperatures, and enhanced radiation intensity induces nu- merous physiological reactions in plants that could change their che- mical composition of valuable phytochemicals, clearly influencing the quality of their harvest products (Wang and Frei, 2011). While the effects of considerable drought stress events on plant physiolo- gy have been investigated in detail (taiZ and Zeiger, 2010; akinci and dorothy, 2012), the impact of a mild reduced water supply on the quality of vegetables has only been insufficiently studied so far. Moreover, even less is known about the effects of moderate water shortage on the plant’s chemical composition in open field cultivation, caused by precipitation shifts from summer to winter as expected for many parts of Europe in the context of climate change (Bindi and olesen, 2011). Depending on its duration and intensity, reduced water supply will affect vegetable crop production in different ways and levels from physiology to crop quality. While perennial crops such as grapevine and fruit trees seem to be well adapted for mild water deficits, vegetable crops – particularly freshly marketed pro- ducts such as spinach as well as tomato and cucumber – appear to be more susceptible to water shortage (costa et al., 2007; nishihara et al., 2001). The drought-induced responses modulate large parts of the plant metabolism and the resulting effects are highly crop- specific (Wang and Frei, 2011) and often hardly predictable. Root vegetables might be affected in a particular way, especially if water deficits occurring during tuber or root initiation (daryanto et al., 2016), since drought increases the rigidity of the soil and, thus, might affect growth of the produce. For instance, red radish (Raphanus sativus L. var. sativus) might be particularly affected due to their short tuber growth periods. Red radish belongs to the botanical fa- mily Brassicaceae and occurs in many different color variations and shapes (PaPetti et al., 2014). Economically important cultivars often include those with red round hypocotyl tubers, which are eaten pre- dominantly as raw food and are being harvested at a very early stage of development long before flowering. They provide a range of nutri- ents as well as quality-related and bioactive compounds, which also have been associated with physiological benefits for human health (manchali et al., 2012). For instance, they were reported to contain considerable amounts of vitamin C (28 mg/100 g fresh weight), glu- cosinolates, and phenolic compounds like anthocyanins (schreiner et al., 2002; souci et al., 2008). Glucosinolates are important for the often desired mild, moderate or strong pungency of radish, addi- tionally their breakdown products being associated with health promoting such as anti-carcerogenic effects (hirata et al., 2018). Besides their importance for the color and consumer acceptance of radish, anthocyanins have been related to health-promoting pro- perties due to their high antioxidant capacity (yao et al., 2004). Furthermore, radish is rich in dry matter, soluble sugars, proteins, and minerals (souci et al., 2008). Moderate water reduction may support or affect the accumulation of quality-related and health promoting compounds in red radish. Earlier studies have shown that inadequate water supply can lead to higher concentrations of primary and se- condary metabolites in vegetables other than red radish (koyama et al., 2012; schreiner et al., 2009; Zhang et al., 2008). Moreover, varying climate conditions also contribute to alterations in the che- mical composition of the product, as shown for bioactive compounds 344 C. Schlering, H. Dietrich, M. Frisch, M. Schreiner, R. Schweiggert, F. Will, J. Zinkernagel in different Brassica vegetables (aires et al., 2011). Therefore, we sought to study the effect of a mild reduced water supply on quality relevant morphological and physio-chemical traits of the red radish tubers in open field cultivation during three successive years inclu- ding four cultivation sets, thus resembling more practically relevant conditions as compared to experiments in controlled environments like greenhouses or growth chambers. Material and methods Chemicals All reagents and solvents used were at least of analytical or HPLC quality unless specified differently. Folin-Ciocalteu’s phenol rea- gent and sodium carbonate were purchased from Merck (Darmstadt, Germany); L-(+)-ascorbic acid from Carl Roth (Karlsruhe, Ger- many); D-(+)-catechin-hydrate, myo-inositol (> 99.5%, HPLC) and anhydrous glycerol from Sigma Aldrich (Steinheim, Germany). Pelargonidin-3-O-glucoside obtained from Extrasynthese (Genay Cedex, France). L-aspartic acid (Ph. Eur., USP) was received from AppliChem (Darmstadt, Germany). Plant material and experimental setup Red radish cv. Celesta F1 (Enza Zaden, Dannstadt-Schauernheim, Germany) were grown from seeds sown in the field. Cultivation experiments were conducted under open field conditions on a san- dy loam in a plot installation at Geisenheim University, Germany (49°59´ N, 7°58´ E). A total of four cultivation sets were grown du- ring the years 2015 (1 set), 2016 (2 sets) and 2017 (1 set) as shown in Tab. 1. Each set consisted of six circular plots (marked by A, C, F, H, M and P) with a diameter of 11.9 m. Each plot was subdivided into four quarters (subplots) for the simultaneous cultivation of three different vegetable crops with an annual crop rotation. Each subplot quarter was again divided into three further segments for conduc- ting the different irrigation treatments following a randomized block design, while the outer segment of each subplot was excluded from harvest (Fig. 1). The irrigation levels remained assigned to the sub- plots during the full duration of the experiments, irrespective of the annual crop rotation. The size of each segment was at least 6.5 m², respectively. Cultivation and water supply The red radish seeds were sown in rows with a spacing of ca. 0.125 m between the rows and a sowing distance of ca. 0.033 m in the rows. Fertilization was carried out uniformly with calcium ammonium nitrate according to commercial standard specifications for radish cultivation (109 kg/N ha-1) based on mineralized N (NO3--N) in 0-15 cm soil depth. Crop protection was applied equally to all plots, whereby application depended on the growing set. The insecticide Karate Zeon (Syngenta, Maintal, Germany) was used against flea beetles (Psylliodes spp.) and Goldor Bait (BASF, Ludwigshafen, Germany) against wireworms (Elateridae) once per cultivation set. The removal of weeds was conducted manually. Water supply was carried out as follows: Drip irrigation was initiated when the soil moisture tension fell below -20 kPa in 10 cm depth and the well- irrigated segments (CTR) were provided with 100% water supply (6.2 L/m2 per irrigation), as controlled by a tensiometer with elec- tronic pressure sensor (Tensio-Technik, Bambach GbR, Geisenheim, Deutschland). Segments assigned to reduced water supply (RWS) were provided with 63% (2015, 2016-2), 68% (2016-1) and 51% (2017), respectively, of the aforementioned water amounts by irriga- tion. Since natural precipitation dominated the moisture tension, the total amount of irrigation events and, consequently, the total amount of supplied water was highly variable from set to set. As a result, the extent to which water supply had been reduced was variable, amounting to a reduction of 10% (2015), 15% (2016-1), 20% (2016-2) and 3% (2017), respectively, when considering both irrigation and precipitation (Tab. 1). Leaf water potential was measured two days before harvest (2016-1 and 2016-2) or just before harvest (2017) by using a Scholander PWSC 3000 (Soil Moisture Equipment Corp., USA). To ensure uniform germination and seedling growth, the soil of both the full and reduced water supply treatments was kept mode- rately moist by receiving the same, sufficiently high water amounts during initial stage. Depending on the growing period, this initial stage lasted for 9 days (2015), 7 days (2016-1, 2016-2) and 12 days (2017). Tab. 1 summarizes further detailed information on cultivati- on data and water supply. Harvest A total of 120 plants per segment were harvested randomly when tuber diameter reached 25-30 mm, representing the commercially targeted maturation stage for harvest. To avoid border effects, outer rows were omitted from harvest. Specifically, harvest dates were 26, 23, 23 and 33 days after sowing for the cultivation sets 2015, 2016- 1, 2016-2 and 2017, respectively (Tab. 1). After harvest and shoot removal with a knife, tubers were cleaned twice manually with fresh tap water to remove adhering soil particles. Then, radish tubers were frozen at -80 °C until analyses. Climatic data Local climate data were supplied by the local weather station, which was located at 100 meters distance to the experimental field site in Geisenheim. Climatic parameters for each cultivation period are summarized in Tab. 1. Based on detailed weather situations of dai- ly temperature and air humidity profile (supporting information, Fig. A.1 and A.4), the different growing periods were characterized Fig. 1: Circular structure of the open field experimental setup. Each quarter of the circle represents a plot divided into three subplots. Red radish was grown in one of the four plots. Subplots differed in the irrigation regimes. The outer subplot (marked by dot pattern) was excluded from the harvest. Chemical composition of field grown radish 345 and assigned as follow: 2015 (warm & moist), 2016-1 and 2016-2 (warm & dry) and 2017 (cold & moist). Strictly speaking, the cultiva- tion periods 2015 and 2017 were characterized by strong variations of global radiation and air humidity (supporting information, Fig. A.2 and A.3), whereas the conditions in 2016 were much more constant. Sample preparation For all analyses, except for the determination of ascorbic acid and anthocyanins, 25 radish tubers of each sample were lyophilized (BETA 2-8 LDplus, Martin Christ Gefriertrocknungsanlagen GmbH, Osterode am Harz, Germany) prior to grinding with a laboratory mill (IKA M 20; IKA-Werke, Staufen, Germany). For the analyses of anthocyanins, the colored external tissues of 15 radish tubers were carefully separated with a vegetable peeler, frozen immediately at -80 °C, and ground into a fine powder under liquid nitrogen with the aforementioned laboratory mill. Dry matter content was deter- mined from freeze-dried material. D-Glucose, D-fructose, titratable acidity, L-malic acid and fumaric acid as well as inorganic anions like nitrate, phosphate, sulfate and chloride were determined from aqueous extracts prepared from the lyophilized powder. For this pur- pose, an aliquot of 8-10 g of lyophilized radish powder was thorough- ly homogenized with 500 mL ultrapure water at room temperature for approx. 10 s at high level (setting 2) using a stainless steel food blender (Waring Blender, Waring Commercial, Torrington, CT 06790, USA). After transferring the extract into an 800 mL beaker using 150 mL ultrapure water, extraction was continued for 10 min under continuous magnetic stirring, followed by a single ultrasound- assisted extraction step in an ultrasound water bath for another 5 min. After centrifuging for 5 min at 4596 × g to separate liquid and solid phases, the supernatants were collected, filtered, and stored at -25 °C until analyses. Extraction procedures for all other target analytes are given below. Chemical analysis Unless otherwise noted, IFU-methods (International Fruit Juice Union, Paris) were used for determination of routine parameters in aqueous extracts, such as sugars, total acidity, organic acids and in- organic anions. Sugars and polyols D-glucose and D-fructose were determined with enzymatic kits (R-Biopharm, Darmstadt, Germany) using a Konelab 20 Xti analyzer (ThermoFisher, Dreieich, Germany). Determination of polyols was carried out in duplicates as follows. An aliquot of 400 mg of lyo- philized powdered radish was extracted twice with 12 mL ultrapure water in a shaking water bath for 1 h at 45 °C. After centrifuging for 10 min at 12.857 × g, supernatants were collected, made up to a volume of 25 mL and filtered through a 0.2 μm membrane (PES filter; VWR, Darmstadt, Germany). Extracts were injected (10 μL) into a HPAEC- PAD system (Dionex/ThermoFisher BIOLC system, ICS 5000+, ICS 3000 SP, Thermo Fisher Scientific, Dreieich, Germany) equip- ped with an anion-exchange Carbopac MA 1 column (250 × 4 mm, Thermo Fisher Scientific, Dreieich, Germany) and a guard column of the same material, both operated at 15 °C. Separation was achieved by isocratic elution with 500 mmol/L NaOH at a flow rate of 0.5 mL/ min. Quantitation was conducted with linear external calibrations of myo-inositol and glycerol. Organic acids Titratable acidity, calculated as citric acid, was measured potentio- metrically after titration to pH 8.1 with 0.3 M NaOH (Titroline al- pha, Schott, Mainz, Germany). L-malic acid was determined using enzymatic kits (R-Biopharm, Darmstadt, Germany). Fumaric acid was determined on a Summit HPLC-UV system (Dionex, Idstein, Germany) using a serial column connection of a Luna C18 250 × 4.6 mm and a Rezex Fast Fruit 100 × 7.8 mm (Phenomenex, Aschaffenburg, Germany). An isocratic elution profile with diluted aqueous phosphoric acid (1 mL H3PO4 (0.6%)/100 mL) and a flow rate of 0.7 mL/min was used. Quantitation was carried out by UV detection at 230 nm using an external calibration. Determination of ascorbic acid was conducted by iodometric titration. For this, an aliquot of 100 g of fresh frozen radish were homogenized in 200 g of oxalic acid (1%, w/v) using a Waring Blender. After centrifuging at 4596 × g for 10 min at 4 °C, the supernatant was acidified with aqueous sulfuric acid (10%) prior to ascorbic acid determination by automatic potentiometric titration (Schott-Titrator, Titrinoline Alpha Plus; software, Titrisoft) with 1/128 mol L-1 iodide-iodate solution. Results were expressed as ascorbic acid in mg g-1 fresh weight. Inorganic anions Nitrate, sulfate and phosphate were analyzed by ion chromato- graphy on an ICS 2100-system on an IonPac AS 17 column using a KOH-gradient (4-45 mmol/L in 20 min, flow rate 0.5 mL/min) and suppressed conductivity detection (Dionex, Idstein, Germany). Tab. 1: Cultivation and climate data during the different experimental periods. DAS: days after sowing. Irrigation (mm): total irrigation amount including irrigation during initial stage. Total water amount (%) RWS: water amount (%) of the reduced variants including watering after sowing. Growing season 2015 2016-1 2016-2 2017 Sowing date (Year-month-day) 2015-08-06 2016-06-29 2016-08-29 2017-04-19 Beginning of water supply differentiation (DAS) 15 7 7 28 Harvest date (DAS) 26 23 23 33 Temperature sum (°C) 565.1 488.1 483.6 408.9 Daily mean air temperature (°C) 21.7 20.3 20.2 12.0 Mean relative air humidity (%) 67.1 83.8 84.2 68.4 Global radiation sum (MJ/m²) 444.1 480.5 372.8 590.9 Daily mean global radiation (MJ/m²) 16.4 20.0 15.5 17.4 Wind speed sum (m/s) at height of 2 m 38.4 32.1 27.4 43.8 Evaporation sum (mm) 92.4 86.6 61.1 94.0 Precipitation sum (mm) 21.2 9.2 12.8 62.9 Number of differentiated irrigation events 2 6 5 1 Irrigation (mm) 93 56 43 48 Total water amount (mm) incl. precipitation 125 65 56 111 Total water amount (%) RWS 90% 85% 80% 97% 346 C. Schlering, H. Dietrich, M. Frisch, M. Schreiner, R. Schweiggert, F. Will, J. Zinkernagel Quantitation was based on linear calibrations with a multi-ion re- ference solution (Bernd Kraft, Duisburg, Germany). Chloride was analyzed by potentiometric titration with an AgCl-electrode using a Schott Titrator TitroLine alpha plus combined with a sample changer TW alpha plus and a piston burette Titronic universal (Schott, Mainz, Germany). Total carbon and nitrogen Elemental analyses of radish samples for carbon and nitrogen were carried out in duplicate by the Dumas combustion method (Vario MAX CNS, Elementar Analysensysteme GmbH, Langenselbold, Germany), combusting 300 mg of freeze-dried radish powder at 950 °C. Aspartic acid was used as reference substance. Minerals and trace elements For minerals and trace elements analyses, the principle of the Kjeldahl digestion with the Gerhardt Turbotherm rapid digestion unit (C. Gerhardt GmbH & Co. KG, Königswinter, Germany) was applied as follows: An aliquot of 250 mg of freeze-dried powdered plant material was mixed with 10 mL of a specifically designed di- gestion solution (420 mL 18 M H2SO4, 330 ml aqueous 30% H2O2 p.a., 0.48 g Se (black powdered), 14.0 g lithium sulfate (suprapur)). After digestion at 100 °C for 100 min, the extract was made up to 50 mL with ultrapure water. Elemental analyses for P, K, Ca, Mg, Mn, Zn and Cu were executed with an ICP-OES (SPECTRO ARCOS, SPECTRO Analytical Instruments, Kleve, Germany). Total phenols An aliquot of 500 mg of freeze-dried powdered radish was extracted twice with 12 mL of 80% aqueous methanol under ultrasonication for 30 min and consecutive centrifugation for 10 min at 12.857 × g. The combined supernatants were made up to 25 mL with extraction solvent and stored at -28 °C until spectrophotometric analyses with the Folin-Ciocalteu reagent and a linear (+) catechin calibration as described previously (singleton and rossi, 1965). Anthocyanins Separation, identification and quantitation of single anthocyanins in the radish skin was carried out by HPLC-PDA-MS using the proce- dure described by Will and dietrich (2006). In brief, an aliquot of 5 g of powdered fresh radish skin was extracted twice with 12 mL of aqueous 85% methanol acidified with 1% formic acid (v/v), using a cooled ultrasonic bath for 30 min in the dark. After centrifuga- tion, supernatants were collected and combined, made up to a vol- ume of 25 mL with extraction solvent and filtered through a 0.2 μm membrane (Chromafil O-20/25 PTFE syringe filter; Macherey- Nagel, Düren, Germany) prior to injection. Chromatographic and mass spectrometric parameters, settings and equipment was used as reported earlier (Will and dietrich, 2006). Injection volume was 4 μL and flow rate 250 μL/min at 40 °C. Anthocyanins were identified by comparison of retention times, UV/Vis absorption maxima, and mass spectra to those of authentic reference substances or literature data (Jing et al., 2012; Wu and Prior, 2005). Quantitation was car- ried out at 520 nm using an external calibration with pelargonidin- 3-O-glucoside. Glucosinolates Identification and quantitation of glucosinolates were determined ac- cording to Wiesner et al. (Wiesner et al., 2013). Statistical analysis Compositional data was analyzed with regard to both dry and fresh matter. Dry matter-related data should allow evaluations irrespective of different water contents in the plant material due to reduced ir- rigation treatments, while fresh matter-related data should represent the nutritional composition of the harvested edible plant material. Data were analyzed by fitting a linear mixed-effect model using the lmer-function within the lme4-package (Bates et al., 2015) of the statistical software R (R Core Team, 2016). The evaluation of the single cultivation sets based on a model with respect to the random plot-effects (eq. 1), while the total dataset was evaluated with respect to the interaction of plot and year (eq. 2): y = water supply + plot (eq. 1) y = water supply + plot:year (eq. 2), where y represents the analyzed parameter, water supply (control, re- duced) was a fixed factor and plot as well as the interaction plot:year were random factors. Comparisons of means derived from differ- ent treatments were considered significantly different if p-values < 0.10. Pairwise comparison of least-squares means was carried out with lsmeans-package (lenth, 2016) to estimate fixed-effects of the treatment as well as random effects for the plot and season. Significances of random effects were calculated by the lmerTest- package (kuZnetsova et al., 2016). The results for the treatment (control, reduced) were evaluated on basis of adjusted data generated from the model mentioned above. Principal Component Analysis (PCA) was carried out by using the R-packages FactoMineR (le et al., 2008) and factoextra (kassamBara and mundt, 2017). To avoid an overweight of single anthocyanins and glucosinolates, only total amounts were considered for Principal Component Analysis (PCA). Three missing values were calculated by the missMDA- package (Josse and husson, 2016). Results and discussion Evaluation of cultivation sets and season The entire data set was visualized by PCA after correcting for ran- dom plot-effects (cf. eq.1) by the least-square means method. The corresponding loading plot revealed a strong differentiation of the cultivation sets by their chemical composition and points out the seasonal effects on the quality of the produce. The first two principal components (PCs) accounted for 66.4% of the total variance in case of dry matter evaluation (Fig. 1) and 62.1% in relation to fresh matter evaluation (data not shown). Consequently, the chemical composition was strongly dependent on the season with all its individual climatic conditions, which was already shown to affect selected substances in Brassica varieties (aires et al., 2011; schreiner, 2005). Cultivation sets from 2015 and 2017 partly showed an overlap, while both sets from 2016 were clearly separated within the PCA loading plot (Fig. 1). The total water amounts were higher in 2015 (125 mm) and 2017 (111 mm) compared to 2016-1 (65 mm) and 2016-2 (56 mm), respectively. These years clusters also differed in mean relative air humidity, which were 67.1% (2015) and 68.4% (2017) as compared to the datasets 2016-1 (83.3%) and 2016-2 (84.2%) (Tab. 1). The re- sulting higher leaf-air vapor pressure difference in 2015 and 2017, respectively, might have altered parameters such as transpiration and tissue temperature, which in turn may affect ion uptake, carbon as- similation and water transport (grantZ, 1990). Nevertheless, both sets from 2016 were clearly separated within the PCA loading plot (Fig. 1). Effects of reduced water supply on the chemical composition of radish biomass When comparing chemical parameters across the full dataset by Chemical composition of field grown radish 347 PCA, the data points of the well-irrigated radish samples were main- ly clustered in sectors with positive PC2, while those of the less irri- gated radish samples were mainly clustered in sectors with negative PC2 (Fig. 2A). Most of the individuals were separated by the second principal component PC2 (Dim2), which accounted for 11.4% of the total variance. Accordingly, the first two principal components (Dim1+Dim2) accounted for 38.2% of the total variance in the data- set. The corresponding variables are represented in the loading plot (Fig. 2B). The most important variables with high contribution to Dim2 were contents of anions like phosphate and chloride as well as content of phosphorous (P), carbon (C), malic acid and total phe- nols. Dim1 was mainly influenced by contents of minerals like Ca, Mg, N, K, P and Zn as well as nitrate and anthocyanins. Apart from these multivariate results, univariate statistical evaluation shown in Tab. 2 was carried out to support the multivariate approach. While the content of malic acid, phosphate, P and Mn as well as total phe- nols and glucosinolates decreased in samples derived from reduced water supply, the content of polyols like inositol and glycerol as well as Cu increased in the same group. In brief, except for polyols and Cu, all other targeted constituents remained constant or were dimi- nished under reduced water supply (Tab. 2). When subjecting the fresh matter-related data to PCA, the distinction of both treatments was substantially clearer (Fig. 3A) than when using dry matter-rela- ted data (Fig. 2A). The second principal component (Dim2) explain- ing 17% of the total variance. Similarly to the dry matter-related evaluation, the first two principal components accounted for 47.3% of the total variance. However, discrimination of both treatments was mainly based on primary metabolites like sugars and polyols, total carbon, total phenols, ascorbic acid as well as sulfate and copper (Fig. 2A-B). Univariate evaluation (Tab. 2) confirmed these results. Total glucosinolates decreased under reduced water supply. In con- trast, contents of many other constituents such as polyols showed a significant increase. This is possibly being related to a concomitantly significantly increase in dry matter content (Tab. 2), which was also shown for carrot tap roots subjected to drought stress just prior to harvest (sørensen et al., 1997). Sugars and polyols The dry matter-related content of sugars such as glucose and fructose was similar irrespective of the water supply applied (0.1072 < p < 0.2932 and 0.1992 < p < 0.6536, resp., Tab. 2). Similarly, differences in fresh matter-related contents of glucose and fructose were insignificant, except for the set 2016-1, where glucose and fruc- tose were significantly higher under reduced water supply cultivation conditions (p = 0.0129 and p = 0.058, respectively). In this latter set (2016-1), the overall glucose and fructose levels (1.40-1.48 and 0.99- 1.03 g/100g FW) were higher than in all other sets (1.11-1.34 and 0.74-0.89 g/100 g FW). As shown in Tab. 1, the set of 2016-1 was grown under comparably warm and high radiation conditions, thus possibly evoking highest sugar accumulation as a result of enhanced photosynthetic rates. In agreement with our results, schreiner et al., (2002) found that contents of glucose and, in particular, fructose as metabolites of photosynthesis were highest in radish grown at high mean light intensities and high mean temperatures. By analogy to glucose and fructose, highest levels of polyols were also found in the warm and well-irradiated growing period 2016-1 (Tab. 2). However, in contrast to glucose and fructose, both dry and fresh matter-related concentrations of polyols like glycerol and myo- inositol were significantly increased in 2016 and 2015 upon reduced water supply (Tab. 2 and 3), except for the set 2017 grown under cold Fig. 2: Principal Component Analysis of the original, non-adjusted data- set including all radish cultivation sets (2015 (red circles), 2016-1 (green triangles), 2016-2 (blue squares) and 2017 (violet crosses)). Score plot represents the individual samples of each cultivation set (separated by color) and plot (marked by the letters A, C, F, H, M and P) based on dry matter-related chemical analyses. Fig. 3: Multivariate evaluation by PCA (PC1+PC2) of the total radish dataset including all cultivation sets (2015, 2016-1, 2016-2 and 2017) based on dry matter-analyses. Score plot (A) represents all individual plot samples (marked by the letters A, C, F, H, M and P) classified according control CTR (blue circles) and reduced water supply RWS (red triangles). The corresponding loading plot (B) shows the related variables determined by the chemi- cal analyses. 348 C. Schlering, H. Dietrich, M. Frisch, M. Schreiner, R. Schweiggert, F. Will, J. Zinkernagel (daily mean air temp.: 12 °C) and comparably well-watered (97% relative to control, Tab. 1) conditions. Supporting the hypothesis that water supply might have played a role, the reduced water supply in the set of 2015 was at 90% relative to full supply, already reaching a marginal statistical significance (p = 0.1303). Those of 2016-1 and 2016-2 were at a more intensely reduced water supply, i.e. at 85% and 80% rel. to full water supply, consequently yielding statistically significant differences (Tab. 2 and 3). Our data also shows that the climatic conditions in the set of 2016-1 resulted in an increase of total polyols, mainly that of inositol (Tab. 2). Polyols have been reported to act as compatible solutes in the osmotic adjustment and as osmo- protectants in plants, being accumulated to osmotically significant levels without adverse effects on the plant’s metabolism (nuccio et al., 1999). In this context, dietrich et al. (2007) found strong in- creases of sorbitol in pear and apple juice exposed to drought stress. In agreement with our findings, strong increases of myo-inositol were observed in drought-stressed pea (Pisum sativum L.) in both the field experiment and under greenhouse conditions (charlton et al., 2008). In our study, we show for the first time that the accumulation of myo-inositol and glycerol in radish might be elicited at already comparably small reductions in water supply, i.e., at ≤ 10% of the full water supply as indicated in our study. Organic acids The organic acids studied in our experiment were hardly affected by moderate water reduction, except for malic acid, which was ne- gatively affected in 2015, but not in 2016-1, 2016-2 and 2017. The reason for these observations remains unclear. Similarly, ascorbic acid levels remained unchanged by mildly reducing water supply, suggesting that the nutritional value with regard to vitamin C re- mained unaffected irrespective of the water supply. However, by ana- logy to glucose, ascorbic acid levels were highest in the sets 2016-1 and 2016-2 grown in the warm and well-irradiated conditions, possib- ly being associated as glucose represents the biosynthetic precursor of ascorbic acid. In contrast to mildly reducing water supply, stronger drought stress was earlier shown to lead to an increase of ascorbic acid in fresh matter of hydroponically grown leafy vegetables like lettuce, spinach and turnip leaf (koyama et al., 2012). In agreement with these results, we also found higher ascorbic acid levels in radish under mildly reduced water supply with statistical significance (p = 0.0891 and p = 0.0517, respectively) for the set 2016-1 as well as the total dataset. Anions Nitrate (NO3-) contents were only insignificantly influenced by re- duced water supply (Tab. 2 and 3, Fig. 2A-B). However, NO3- con- tents were substantially lower in 2016-1 (11.1-13.8 mg/g DM) and 2017 (17.95-18.96 mg/g DM) in contrast to other cultivation sets (26.5-32.8 mg/g DM) as shown in Tab. 1. High radiation and evapo- transpiration rates might have led to balanced conditions between absorption and assimilation of NO3-, resulting in lower NO3--levels. However, frequent stress-related conditions for the respective radish plants just before harvest might have affected NO3--levels in the set of 2016-1 (Tab. 1). In agreement with this hypothesis, sørensen et al. (1997) found lowered nitrate levels in carrots (Daucus carota L.) even when drought stress occurred just prior to harvest. In contrast to 2016-1, where irrigation was finished two days prior to harvest, ir- rigation in 2016-2 was finished 6 days before harvest due to adequate soil humidity, probably having led to more continuous NO3- uptakes in contrast to 2016-1. Further study is needed though. Although the dry matter-related content of phosphate (PO43-) was significantly decreased by water supply reduction when considering the overall dataset, the significant decline of PO43- was only observed in 2016-2, when water reduction was highest, i.e., to 80% of full water supply. Since phosphate mobility in the soil depends on mois- ture (taiZ and Zeiger, 2010), reduced soil moisture might have diminished the uptake into the tubers (da silva et al., 2010). Similarly to the other anions, sulfate (SO42-) and chloride (Cl-) levels were hardly affected by reduced water supply, except for the set 2017, where a significant increase in Cl- from 0.89 to 0.96 mg/100 g DM (p = 0.0375) was observed (Tab. 2). While the levels of SO42- and Cl- were highest in 2016-1 and 2016-2, when irrigation events were more frequent compared to the other sets, the mild water deficit evoked in our study appeared to be insufficient to significantly affect sulfate and chloride uptake. Carbon Water supply reductions did not influence dry matter-related total carbon (C) as shown by multivariate and univariate analyses. Fresh matter-related total carbon showed to be significantly higher in samp- les grown under reduced water supply (1.75-1.95 g/100 g FM) than in control samples (1.67-1.91 g/100 g FM). These findings correlate well to significantly higher contents of dry matter in radish grown under reduced water supply. In agreement, multivariate evaluation confirmed a considerable contribution of C to the separation of both variants on fresh matter basis (Fig. 3A-B). Fig. 4: Multivariate evaluation by PCA (PC1+PC2) of the total radish dataset including all cultivation sets (2015, 2016-1, 2016-2 and 2017) based on fresh matter-analyses. Score plot (A) represents all individual plot samples (marked by the letters A, C, F, H, M and P) classified according control CTR (blue circles) and reduced water supply RWS. The corresponding loading plot (B) shows the related variables determined by the chemical analyses. Chemical composition of field grown radish 349 Nitrogen Mild reductions in water supply had no significant impact on dry matter-related total nitrogen (N) in radish (Tab. 2, Fig. 2A-B), while it was associated with an apparently significant increase of fresh matter-related nitrogen contents in the warm and moist year 2015 and the cold and moist year 2017 (Tab. 3). In these years, the total nitrogen levels were substantially higher (94.5-104.0 mg/100 g FW) than in 2016 (72.1-81.6 mg/100 g FW). Because soil N availability is negatively affected by drought events and this strongly influences nutrient absorption and uptake by plants, it might be assumed that N mobility had been temporarily restricted due to insufficient total water amounts in 2016 (56-65 mm) in contrast to 2015 (125 mm) and 2017 (111 mm). In addition, lower air humidity in 2015 (67.1%) and 2017 (68.4%), in contrast to 2016 (83.3 to 84.2%), might have led to higher N levels. gisleröd et al. (1987) found decreased levels of selected macronutrients in greenhouse plants when exposed to high- er air humidity. Further studies showed that negative effects on the plants’ nitrogen level are alleviated with drying-rewetting cycles (he and diJkstra, 2014), possibly explaining the absent treatment effect in 2016-1 and 2016-2, although effects were expected due to several short-time drought events. Potassium By analogy to our observations on nitrogen levels, water supply limi- tations exerted only insignificant effects on dry matter-related con- tents of potassium according to both PCA analysis and univariate statistical evaluation (Fig. 2A-B, Tab. 2). In contrast, significantly increased fresh matter-related contents were observed in the sets of 2015 and 2017 with reduced water supply (Tab. 3), possibly being a result of generally increased dry matter, which in turn resulted from reduced moisture content. Since potassium plays multiple im- portant functional roles in plant cells (taiZ and Zeiger, 2010), its level should be strictly regulated by plant metabolism. In agreement, aFZal et al. (2014) also found only insignificant effects of drought stress on dry matter-related K+ concentrations in leaves and roots of mungbean (Vigna radiata (L.) Wilczek). In our study, radish showed highest absolute K+ levels in 2017 (232.6-255.6 mg/100 g FW; 51.5- 53.2 mg/g DW) when precipitation was most frequent and abundant, as compared to the sets of other years (198.6-222.3 mg/100 g FW; 39.0-47.4 mg/100 g DW). Under these conditions in 2017 (cold and moist), however, K+ levels notably increased upon water supply limi- tation (p = 0.0405, fresh matter; p = 0.0711, dry matter). Phosphorous The content of phosphorous (P) in radish dry matter was significantly decreased by moderately reduced water supply. By analogy to PO43-, the decline of P was highly significant for the total dataset as well as for the single cultivation sets 2015 and 2016-2 (Tab. 2). Multivariate analyses by PCA support this result (Fig. 2A-B). greenWay et al. (1969) found that even a small decrease in water potential reduces P uptake by tomato plants (Lycopersicon esculentum Mill.), while water uptake and transport were generally lower in gradually treated plants in contrast to plants suddenly exposed to low water potential. However, P uptake was much less for a considerable time even if plants were exposed to stress for only one hour. Further studies in agreement with these and our findings have been reported (sardans and Peñuelas, 2004; sáncheZ-rodrígueZ et al., 2010). Calcium Reduced water supply led to lower calcium (Ca2+) contents in rad- ish dry matter, being significant in 2016-1 (p = 0.0266) and 2016-2 (p = 0.0907), but not in the cultivation sets of 2015 and 2017. Contrary to K+ or N, absolute Ca2+ content was higher in reducibly watered samples in 2017, but water reduction was very low (Tab. 1). In gene- ral, Ca2+ acts as important constituent of the middle lamella of the cell wall and it is required in numerous metabolic processes (taiZ and Zeiger, 2010) including responses to abiotic stress like drought (hu and schmidhalter, 2005; knight et al., 1997). However, sáncheZ-rodrígueZ et al. (2010) found no differences in the Ca2+ accumulation of tomato leaves when plants were treated by moderate drought stress, although its uptake was significantly affected. hu and schmidhalter (2005) concluded that even if its uptake is reduced, the accumulation of Ca2+ declines very slightly. In low-transpiring organs like tubers or fruits, which are supplied predominantly via the phloem, its concentration is considerably lower than in leaves (marschner, 1986). In agreement, we did not observe any consis- tent effect of reduced waters supply on the Ca2+ content in red radish. Magnesium Similarly to other minerals, effects on magnesium (Mg2+) with re- gard to reduced waters supply on radish were apparently insignificant as shown by multivariate analyses (Fig. 2A-B). However, univariate evaluation demonstrated an increase in the Mg2+ content due to re- duced water supply in 2016-1 (p = 0.0946) (Tab. 2). In agreement, PuluPol et al. (1996) found only insignificant effects of deficit irri- gation on the dry weight-based levels of magnesium in tomato fruits grown in a glasshouse. However, fresh weight-based Mg2+ contents have been found to be higher in fruits of deficit irrigated plants than in fruits of well-watered plants (PuluPol et al., 1996), which also applies to radish fresh matter grown under moist conditions in 2015 (p = 0.0642) and 2017 (p = 0.0387) (Tab. 3). Micronutrients (Fe, Zn, Mn, Cu) Dry matter-related iron (Fe) contents were reduced in 2015 (p = 0.0846), whereas zinc (Zn) and copper (Cu) were increased in 2017 (p = 0.0297 and p = 0.0523, respectively) by slight water reduction (Tab. 2). These findings might be related to the involvement of these micronutrients in photosynthesis and their catalytic role in the anti- oxidant systems of plants (hänsch and mendel, 2009). Evaluation of the total dataset indicated a significant dry matter-related decrease of manganese (Mn) and a significant increase in Cu upon water supply limitation. Similar observations have been reported earlier (sáncheZ-rodrígueZ et al., 2010). PCA analyses supported the contribution of these micronutrients to the variants’ differentiation (Fig. 2B). Because micronutrient transport through the soil and towards the root is governed by diffusion, diminished levels were expected when limiting water supply. Considering the total fresh matter-related dataset (Tab. 3), significantly higher contents of Zn and Cu were found for samples with limited water supply in 2017. Our results showed generally reduced contents of Fe in 2016-1 and 2017 (Tab. 2), when global radiation was highest. The reduced con- tents of Fe in the set of 2016-1 are in agreement with the trend found for most macronutrients. However, this results are contradictory to Zancan et al. (2006), who found iron accumulation in roots and leaves of maize provoked by high UV-B radiation. Phenolic compounds Reduced water supply led to significant reductions (p = 0.0523) in the dry matter-related content of total phenols (Tab. 2), while fresh matter-related contents remained unchanged (Tab. 3). Previous stu- dies reported variable effects of drought stress on levels of polyphe- nols including anthocyanins in the respective plants (roBBins et al., 2005). In agreement with literature, the quantitatively most abundant antho- cyanins in radish skin were acylated derivatives, namely pelargo- 350 C. Schlering, H. Dietrich, M. Frisch, M. Schreiner, R. Schweiggert, F. Will, J. Zinkernagel Ta b. 2 : In flu en ce o f r ed uc ed w at er s up pl y on th e dr y m at te r ( D M )- re la te d le ve ls o f c on st itu en ts o f r ad is h tu be rs fr om d iff er en t c ul tiv at io n se ts (2 01 5, 2 01 6- 1, 2 01 6- 2 an d 20 17 ). L in ea r m ix ed m od el , t -t es t, p- va lu es : < 0. 1, < 0 .0 5 (* ), < 0. 01 (* *) a nd < 0 .0 01 (* ** ), nd = n ot d et er m in ed . C T R : C on tr ol tr ea tm en t w ith fu ll w at er s up pl y. R W S: R ed uc ed w at er s up pl y tr ea tm en t. R ad is h 20 15 R ad is h 20 16 -1 R ad is h 20 16 -2 R ad is h 20 17 R ad is h to ta l C T R R W S p- va lu e C T R R W S p- va lu e C T R R W S p- va lu e C T R R W S p- va lu e C T R R W S p- va lu e L ea f w at er p ot en tia l Ψ [- M Pa ] nd nd n d 0. 15 0. 21 0. 01 23 * * 0. 16 0. 24 0. 00 02 * ** 0. 07 0. 09 0. 00 06 * ** nd nd n d D ry m at te r [ % ] 4. 79 4. 97 0. 03 37 * 5. 00 5. 17 0. 08 86 4. 64 4. 71 0. 00 35 * * 4. 61 4. 75 0. 17 36 4. 76 4. 90 0. 00 01 * ** Su ga rs a nd P ol yo ls G lu co se [m g/ g] 25 4. 5 24 4. 6 0. 29 32 28 1. 1 28 6. 5 0. 15 77 27 2. 9 28 3. 7 0. 10 72 24 7. 06 23 9. 92 0. 18 46 26 9. 5 27 1. 6 0. 95 50 Fr uc to se [m g/ g] 18 4. 3 17 8. 9 0. 34 52 19 7. 3 19 9. 2 0. 63 87 16 0. 2 16 2. 3 0. 65 36 16 8. 39 16 2. 01 0. 19 92 18 0. 6 18 0. 1 0. 39 63 Po ly ol s to ta l [ m g/ g] 2. 07 2. 18 0. 13 03 3. 66 4. 40 0. 00 09 * ** 1. 71 1. 85 0. 00 72 * * 1. 97 2. 01 0. 50 62 2. 48 2. 81 0. 00 09 * ** In os ito l [ m g/ g] 1. 75 1. 85 0. 14 67 3. 15 3. 81 0. 00 19 * * 1. 37 1. 51 0. 00 71 * * 1. 41 1. 43 0. 60 12 2. 09 2. 39 0. 00 10 * * G ly ce ro l [ m g/ g] 0. 32 0. 34 0. 27 22 0. 51 0. 60 0. 00 21 * * 0. 35 0. 34 0. 45 49 0. 56 0. 58 0. 55 21 0. 39 0. 42 0. 01 97 * O rg an ic a ci ds To ta l a ci di ty [m g/ g] 15 .0 14 .6 0. 31 11 11 .2 10 .1 0. 37 01 12 .3 11 .8 0. 67 68 10 .6 0 10 .0 9 0. 08 16 12 .8 12 .2 0. 11 09 M al ic a ci d [m g/ g] 36 .1 34 .7 0. 04 54 * 30 .5 29 .7 0. 14 87 34 .9 34 .3 0. 18 65 35 .6 6 35 .7 0 0. 96 80 33 .8 32 .9 0. 02 35 * Fu m ar ic a ci d [m g/ g] 3. 39 3. 49 0. 49 69 4. 86 4. 63 0. 29 60 4. 81 6. 39 0. 31 58 2. 52 2. 70 0. 81 72 4. 35 4. 34 0. 87 18 A sc or bi c ac id [m g/ g] 5. 64 5. 47 0. 29 69 5. 80 6. 03 0. 23 39 6. 45 6. 45 0. 98 72 5. 21 5. 30 0. 30 32 5. 96 5. 98 0. 63 62 A ni on s N itr at e [m g/ g] 32 .8 32 .3 0. 82 00 13 .8 11 .1 0. 10 77 27 .8 26 .5 0. 54 18 17 .9 5 18 .9 6 0. 45 86 24 .7 7 23 .2 9 0. 35 40 Ph os ph at e [m g/ g] 11 .4 8 11 .1 6 0. 49 93 12 .9 0 12 .1 2 0. 16 68 11 .1 4 10 .1 7 0. 02 11 * 10 .0 4 9. 89 0. 66 74 11 .8 4 11 .1 5 0. 00 92 * * Su lf at e [m g/ g] 2. 52 2. 75 0. 52 39 3. 99 4. 13 0. 23 28 4. 00 4. 05 0. 79 45 3. 47 3. 32 0. 58 19 3. 51 3. 64 0. 56 80 C hl or id e [m g/ g] 0. 88 0. 98 0. 53 87 1. 90 1. 92 0. 88 35 1. 79 1. 27 0. 14 02 0. 89 0. 96 0. 03 75 * 1. 52 1. 38 0. 39 48 E le m en ts C ar bo n [m g/ g] 37 0. 7 36 9. 9 0. 54 56 38 2. 9 38 2. 6 0. 78 38 37 2. 8 37 1. 7 0. 65 34 36 8. 53 36 7. 43 0. 49 87 37 5. 5 37 4. 8 0. 29 69 M ac ro nu tr ie nt s Po ta ss iu m [m g/ g] 44 .9 5 44 .7 6 0. 79 44 40 .0 2 38 .9 5 0. 27 18 47 .3 5 46 .0 5 0. 41 33 51 .5 4 53 .1 8 0. 07 11 44 .1 0 43 .2 6 0. 66 44 N itr og en [m g/ g] 20 .2 4 20 .3 6 0. 64 06 14 .6 9 14 .1 4 0. 14 97 17 .6 2 16 .7 2 0. 10 60 20 .9 3 21 .6 0 0. 18 40 17 .5 2 17 .0 7 0. 48 93 Ph os ph or ou s [m g/ g] 4. 15 4. 03 0. 02 41 * 4. 29 4. 17 0. 32 20 4. 67 4. 35 0. 00 81 * * 4. 51 4. 48 0. 65 85 4. 37 4. 18 0. 00 56 * * C al ci um [m g/ g] 3. 57 3. 59 0. 72 81 3. 01 2. 86 0. 02 66 * 3. 54 3. 22 0. 09 07 2. 91 3. 07 0. 08 24 3. 37 3. 22 0. 22 59 M ag ne si um [m g/ g] 1. 30 1. 33 0. 35 91 0. 97 1. 03 0. 09 46 1. 26 1. 19 0. 21 25 1. 29 1. 35 0. 17 56 1. 20 1. 16 0. 58 56 M ic ro nu tr ie nt s Ir on [μ g/ g] 90 .2 7 76 .8 5 0. 08 46 68 .5 7 70 .2 9 0. 70 39 94 .7 0 97 .0 0 0. 79 62 84 .0 2 70 .7 6 0. 28 25 84 .8 4 81 .3 8 0. 18 73 Z in c [μ g/ g] 22 .6 5 22 .3 2 0. 43 37 20 .2 7 20 .1 4 0. 85 11 23 .4 3 23 .2 5 0. 67 63 22 .1 1 23 .1 1 0. 02 97 * 22 .1 2 21 .9 0 0. 71 66 M an ga ne se [μ g/ g] 11 .1 9 10 .8 7 0. 21 30 9. 45 9. 47 0. 93 89 10 .8 5 10 .5 1 0. 29 25 9. 91 8. 98 0. 13 12 10 .5 2 10 .2 8 0. 03 55 * C op pe r [ μg /g ] 4. 29 5. 39 0. 19 07 5. 83 6. 11 0. 64 88 5. 68 5. 77 0. 62 23 5. 21 5. 86 0. 05 23 5. 27 5. 76 0. 03 97 * P he no lic c om po un ds To ta l p he no ls [m g/ g] 11 .0 10 .8 0. 43 05 11 .2 11 .0 0. 08 49 11 .5 9 11 .1 6 0. 14 87 10 .1 7 10 .1 7 0. 97 50 11 .2 5 10 .9 9 0. 05 23 To ta l a nt ho cy an in s [m g/ g pe el ] 20 .3 6 21 .3 2 0. 15 70 26 .3 2 23 .9 6 0. 09 09 33 .9 9 31 .5 1 0. 10 40 29 .1 9 29 .5 3 0. 38 14 26 .8 9 25 .6 0 0. 12 17 G lu co si no la te s To ta l g lu co si no la te s [μ m ol /g ] 11 .9 7 12 .3 0 0. 79 59 28 .5 2 21 .4 9 0. 01 59 * * 26 .3 2 23 .6 7 0. 43 96 16 .0 6 16 .6 0 0. 05 28 22 .2 7 19 .1 6 0. 06 70 A lip ha tic g lu co si no la te s [ μm ol /g ] 11 .2 2 11 .5 3 0. 80 26 27 .8 8 21 .0 0 0. 01 53 * * 25 .5 5 23 .0 9 0. 46 27 15 .5 1 16 .0 1 0. 05 23 21 .5 5 18 .5 4 0. 06 91 In do lic g lu co si no la te s [μ m ol /g ] 0. 75 0. 79 0. 70 86 0. 64 0. 49 0. 20 62 0. 76 0. 58 0. 02 60 * 0. 55 0. 59 0. 24 19 0. 72 0. 62 0. 08 87 Chemical composition of field grown radish 351 Ta b. 3 : In flu en ce o f r ed uc ed w at er s up pl y on th e fr es h m at te r ( FM )- re la te d le ve ls o f c on st itu en ts o f r ad is h tu be rs fr om d iff er en t c ul tiv at io n se ts (2 01 5, 2 01 6- 1, 2 01 6- 2 an d 20 17 ). L in ea r m ix ed m od el , t -t es t, p- va lu es : < 0. 1, < 0 .0 5 (* ), < 0. 01 (* *) a nd < 0 .0 01 (* ** ), nd = n ot d et er m in ed . C T R : C on tr ol tr ea tm en t w ith fu ll w at er s up pl y. R W S: R ed uc ed w at er s up pl y tr ea tm en t. R ad is h 20 15 R ad is h 20 16 -1 R ad is h 20 16 -2 R ad is h 20 17 R ad is h to ta l C T R R W S p- va lu e C T R R W S p- va lu e C T R R W S p- va lu e C T R R W S p- va lu e C T R R W S p- va lu e L ea f w at er p ot en tia l Ψ [- M Pa ] nd nd n d 0. 15 0. 21 0. 01 23 * 0. 16 0. 24 0. 00 02 * ** 0. 10 0. 15 0. 00 06 * ** nd nd n d D ry m at te r [ % ] 4. 79 4. 97 0. 03 37 * 4. 64 4. 71 0. 08 86 4. 78 4. 98 0. 00 35 * * 4. 61 4. 75 0. 17 36 4. 76 4. 90 0. 00 01 * ** Su ga rs a nd P ol yo ls G lu co se [g /1 00 g ] 1. 22 1. 22 0. 97 32 1. 40 1. 48 0. 01 29 * 1. 27 1. 34 0. 07 34 1. 11 1. 15 0. 43 01 1. 25 1. 30 0. 03 15 * Fr uc to se [g /1 00 g ] 0. 88 0. 89 0. 86 12 0. 99 1. 03 0. 05 80 0. 74 0. 76 0. 36 32 0. 76 0. 78 0. 61 91 0. 84 0. 87 0. 10 60 Po ly ol s to ta l [ m g/ 10 0 g] 9. 93 10 .8 4 0. 06 40 18 .2 6 22 .4 7 0. 00 12 * * 36 .9 3 39 .2 3 0. 04 94 * 42 .2 4 42 .7 9 0. 72 67 26 .8 4 28 .8 3 0. 00 11 * * In os ito l [ m g/ 10 0 g] 8. 39 9. 18 0. 06 81 15 .7 0 19 .4 4 0. 00 19 * * 29 .4 6 32 .0 2 0. 02 99 * 30 .1 9 30 .4 5 0. 82 05 20 .9 3 22 .7 7 0. 00 06 * ** G ly ce ro l [ m g/ 10 0 g] 1. 55 1. 65 0. 10 56 2. 56 3. 03 0. 00 17 * * 7. 47 7. 21 0. 26 60 12 .0 5 12 .3 4 0. 63 87 5. 91 6. 06 0. 34 52 O rg an ic a ci ds To ta l a ci di ty [m g/ 10 0 g] 71 .6 7 73 .3 3 0. 56 23 55 .0 0 56 .6 7 0. 79 26 56 .6 7 53 .3 3 0. 36 32 48 .3 3 48 .3 3 1. 00 00 58 .3 3 57 .5 0 0. 66 43 M al ic a ci d [m g/ 10 0 g] 17 1. 66 17 5. 00 0. 36 32 15 3. 33 15 3. 33 1. 00 00 16 3. 33 16 0. 00 0. 17 47 16 0. 00 17 0. 00 0. 20 31 16 2. 09 16 4. 58 0. 26 56 Fu m ar ic a ci d [m g/ 10 0 g] 16 .2 0 16 .8 3 0. 36 15 24 .7 3 23 .8 9 0. 33 58 23 .5 9 23 .6 3 0. 94 88 10 .7 3 12 .3 0 0. 59 45 18 .8 1 19 .1 6 0. 64 67 A sc or bi c ac id [m g/ 10 0 g] 26 .9 8 27 .1 7 0. 77 96 28 .9 5 30 .9 8 0. 08 91 29 .9 4 30 .3 5 0. 60 99 24 .4 3 24 .8 6 0. 38 60 27 .5 7 28 .3 4 0. 05 17 A ni on s N itr at e [m g/ 10 0 g] 15 6. 61 16 0. 59 0. 66 72 68 .8 5 57 .1 7 0. 15 67 12 9. 17 12 4. 49 0. 62 69 80 .4 6 90 .9 4 0. 08 69 10 8. 77 10 8. 30 0. 91 65 Ph os ph at e [m g/ 10 0 g] 54 .9 2 55 .2 8 0. 89 17 64 .3 4 62 .5 6 0. 53 15 51 .6 9 47 .9 2 0. 06 03 45 .1 7 47 .5 8 0. 12 62 54 .0 3 53 .3 4 0. 53 12 Su lf at e [m g/ 10 0 g] 12 .0 4 13 .5 8 0. 39 44 19 .9 7 21 .3 4 0. 06 93 18 .5 5 19 .0 7 0. 56 22 15 .5 7 15 .7 3 0. 89 95 16 .5 3 17 .4 3 0. 11 53 C hl or id e [m g/ 10 0 g] 4. 23 4. 84 0. 39 83 9. 52 9. 63 0. 86 13 8. 32 5. 88 0. 14 62 4. 03 4. 53 0. 06 75 6. 53 6. 22 0. 52 44 E le m en ts C ar bo n [g /1 00 g ] 1. 77 1. 84 0. 03 92 * 1. 91 1. 95 0. 02 07 * 1. 73 1. 75 0. 19 62 1. 67 1. 77 0. 17 70 1. 77 1. 83 0. 00 46 * * M ac ro nu tr ie nt s Po ta ss iu m [m g/ 10 0 g] 21 5. 16 22 2. 30 0. 01 55 * 19 9. 39 19 8. 61 0. 86 30 21 9. 88 21 6. 72 0. 67 84 23 2. 64 25 5. 56 0. 04 05 * 21 6. 77 22 3. 30 0. 07 59 N itr og en [m g/ 10 0 g] 96 .7 7 10 1. 06 0. 00 88 * * 73 .1 8 72 .1 3 0. 52 82 81 .8 5 78 .6 9 0. 25 46 94 .5 2 10 4. 01 0. 04 69 * 86 .5 8 88 .9 7 0. 13 11 Ph os ph or ou s [m g/ 10 0 g] 19 .8 8 20 .0 3 0. 37 80 21 .3 7 21 .2 4 0. 80 66 21 .6 9 20 .5 1 0. 05 34 20 .2 1 21 .6 9 0. 13 23 20 .7 9 20 .8 7 0. 80 06 C al ci um [m g/ 10 0 g] 17 .0 4 17 .8 1 0. 11 88 15 .0 1 14 .5 7 0. 12 92 16 .4 2 15 .1 5 0. 16 31 13 .1 5 14 .8 0 0. 01 69 * 15 .4 1 15 .5 8 0. 59 58 M ag ne si um [m g/ 10 0 g] 6. 20 6. 59 0. 06 42 5. 14 4. 96 0. 19 69 5. 85 5. 58 0. 32 90 5. 84 6. 50 0. 03 87 * 5. 76 5. 91 0. 24 40 M ic ro nu tr ie nt s Ir on [μ g/ 10 0 g] 43 3. 98 38 2. 79 0. 17 48 34 2. 23 35 8. 24 0. 45 20 45 1. 42 45 6. 74 0. 88 85 37 7. 74 34 1. 80 0. 50 64 38 4. 89 40 2. 16 0. 39 76 Z in c [μ g/ 10 0 g] 10 8. 22 11 0. 80 0. 27 98 10 1. 03 10 2. 71 0. 61 31 10 8. 78 10 9. 39 0. 78 59 99 .9 9 11 1. 04 0. 01 99 * 10 4. 50 10 8. 49 0. 01 62 * M an ga ne se [μ g/ 10 0 g] 53 .5 6 54 .0 2 0. 71 45 47 .0 9 48 .2 8 0. 36 52 50 .3 7 49 .4 2 0. 43 95 44 .7 5 43 .2 6 0. 64 70 48 .9 6 48 .7 4 0. 82 79 C op pe r [ μg /1 00 g ] 20 .5 0 26 .7 3 0. 14 36 29 .0 9 31 .1 1 0. 49 15 26 .3 5 27 .1 9 0. 32 22 23 .5 4 28 .1 4 0. 01 52 * 24 .8 7 28 .2 9 0. 00 94 * * P he no lic c om po un ds To ta l p he no ls [m g/ g] 53 .3 4 54 .4 1 0. 42 28 55 .8 7 56 .6 6 0. 42 63 57 .9 3 57 .6 3 0. 82 91 46 .0 1 49 .0 8 0. 20 09 53 .2 9 54 .4 5 0. 13 64 To ta l a nt ho cy an in s [m g/ g pe el ] 12 8. 35 13 9. 14 0. 03 49 * 13 1. 17 12 2. 13 0. 15 53 15 8. 04 14 8. 32 0. 19 93 13 6. 99 13 8. 30 0. 35 81 13 8. 64 13 6. 97 0. 57 03 G lu co si no la te s To ta l g lu co si no la te s [μ m ol /g ] 57 .0 7 61 .0 4 0. 54 04 14 2. 23 10 9. 68 0. 02 53 * 12 2. 07 11 1. 04 0. 47 24 74 .1 9 76 .6 3 0. 06 35 98 .8 9 89 .6 0 0. 10 67 A lip ha tic g lu co sin ol at es [μ m ol /g ] 53 .4 8 57 .1 8 0. 55 45 13 9. 05 10 7. 19 0. 02 45 * 11 8. 53 10 8. 31 0. 49 68 72 .7 5 74 .9 6 0. 08 01 95 .9 5 86 .9 1 0. 10 76 In do lic g lu co si no la te s [μ m ol /g ] 3. 59 3. 86 0. 39 01 3. 19 2. 50 0. 24 58 3. 54 2. 72 0. 02 99 * 1. 44 1. 67 0. 06 28 2. 94 2. 69 0. 20 19 352 C. Schlering, H. Dietrich, M. Frisch, M. Schreiner, R. Schweiggert, F. Will, J. Zinkernagel nidin-3-O-( p-coumaroyl-)-diglucoside-5-O-(malonyl-)-glucoside and pelargonidin-3-O-(feruloyl)-diglucoside-5-O-(malonyl)-gluco- side, accounting for approx. 80% of the total anthocyanins (Wu and Prior, 2005) (data not shown). Limited water supply led to only in- significant effects on total anthocyanins in radish. Univariate analy- sis indicated a dry matter-related decline of total anthocyanins in radish skin upon limited water supply in 2016 (Tab. 2), whereas a fresh matter-related increase was found in 2015. In agreement, mul- tivariate analyses did not indicate a strong contribution of anthocya- nins to the differentiation of samples grown under full or reduced water supply (Fig. 2A-B). Interestingly, both fresh and dry matter- related contents of anthocyanins were highest in 2016-2 followed by 2017 (Tab. 2 and 3). In the context of seasonal effects on radish tuber color, schreiner et al. (2002) found most intense red coloration at high photosynthetic photon flux density, hypothesizing its produc- tion to be dependent on light intensity, quality and duration (maZZa and miniati, 1993). In addition, an earlier study revealed that rad- ish gained more distinctive red shades at lower mean temperatures (schreiner et al., 2002). In our study, lowest mean temperature and highest global radiation occurred in 2017, which also showed high amounts of anthocyanins apart from the set 2016-2 (Tab. 2 and 3). In brief, mild water supply reduction did not significantly affect antho- cyanin contents, representing a most important commercial quality trait due to its direct correlation with the tubers’ color. Glucosinolates Reduced water supply led to decreased dry matter-related contents of total, aliphatic and indolyl glucosinolates. Most abundant glucosi- nolates in radish were aliphatic derivatives such as glucoraphasatin (4-methylthio-3-butenyl glucosinolate) and glucoraphenin (4-methyl- sulfinyl-3-butenyl glucosinolate). In earlier studies, drought stress has been shown to more likely increase glucosinolate concentration in Brassica species such as Brassica carinata (schreiner et al., 2009), B. rapa ssp. rapifera L. (Zhang et al., 2008) and B. napus L. (Jensen et al., 1996). However, the intensity of drought appears to be a de- cisive impact factor in glucosinolate increase, because, under minor drought, glucosinolate concentrations did not increase in Brassica napus plants (Jensen et al., 1996) and decreased in Brassica olera- cea var. italica. (roBBins et al., 2005) Thus, roBBins et al. (2005) also observed decreasing concentrations of mainly indolyl-glucosi- nolates in broccoli (B. oleracea var. italica) induced by a slightly reduced water supply. Our results showed that total glucosinolates concentration was highest in 2016 (21.49-28.52 μmol/g DM) as com- pared to the other years (11.97-16.60 μmol/g DM). In 2016, precipita- tion was lowest and, thus, irrigation events were frequently released, but total water amounts were considerately lower compared to 2015 and 2017. Zhang et al. (2008) also proved significant interactions of growing season and water supply for turnip, which underlines the seasonal-dependent effects on glucosinolates. Conclusion In conclusion, slightly reduced water supply situations, as expected in the future due to climate change, were shown to alter the chemical composition of red radish grown in open field cultivation. The effects expectedly appeared to vary in dependence of the respective envi- ronmental conditions. Dry matter contents were mainly enhanced by moderate water reduction and the same was found for polyols, which seem to be highly sensitive to slight water lowering. Being important for color and pungency of the produce, anthocyanin and glucosino- late levels were shown to be only marginally affected. In brief, our results indicate that red radish plants are able to cope with moderate- ly reduced water supply, only marginally affecting the overall quality of its commercialized tubers. However, limitations of water supply in areas of vegetable production may increase by the on-going climate change and thus exceed those exposed in our simulation, ultimately aggravating the effects observed in our study and thus still impacting crop quality. Funding sources The LOEWE excellence cluster FACE2FACE from the Hessian State Ministry of Higher Education, Research and the Arts (Wiesbaden, Germany) funded this study. Supplementary data Information on the daily mean weather conditions during the diffe- rent cultivation periods 2015, 2016-1, 2016-2 and 2017 is supplied in Fig A.1-4 as supporting information. References aFZal, a., gulZar, i., shahBaZ, m., ashraF, m., 2014: Water deficit- induced regulation of growth, gas exchange, chlorophyll fluorescence, inorganic nutrient accumulation and antioxidative defense mechanism in mungbean (Vigna radiata (L.) Wilczek). J. Appl. Bot. Food Qual. 87, 147- 156. DOI: 10.5073/JABFQ.2014.087.022 aires, a., Fernandes, c., carvalho, r., Bennett, r.n., saavedra, m.J., rosa, e.a.s., 2011: Seasonal effects on bioactive compounds and anti- oxidant capacity of six economically important brassica vegetables. Molecules 16, 6816-6832. DOI: 10.3390/molecules16086816 akinci, s., dorothy, l.m., 2012: Plant Water-Stress Response Mechanisms. In: Rahman, I.M.M., Hasegawa, H. (eds.), Water Stress, 15-42. InTech. Bates, d., maechler, m., Bolker, B., Walker, s., 2015: Fitting Linear Mixed-Effects Models Using {lme4}. J. Stat. Softw. 67, 1-48. DOI: 10.18637/jss.v067.i01 Bindi, m., olesen, J.e., 2011: The responses of agriculture in Europe to climate change. Reg. Environ. Change 11, 151-158. DOI: 10.1007/s10113-010-0173-x charlton, a.J., donarski, J., harrison, m., Jones, s., godWard, J., oehlschlager, s., arques, J., amBrose, m., chinoy, c., mullineaux, P., domoney, c., 2008: Responses of the pea (Pisum sativum L.) leaf metabolome to drought stress assessed by nuclear magnetic resonance spectroscopy. Metabolomics 4. DOI: 10.1007/s11306-008-0128-0 costa, J.m., ortuño, m.F., chaves, m.m., 2007: Deficit irrigation as a strategy to save water. Physiology and potential application to horticul- ture. J. Integr. Plant Biol. 49, 1421-1434. DOI: 10.1111/j.1672-9072.2007.00556.x da silva, e.c., nogueira, r.J.m.c., silva, m., alBuquerque, m., 2010: Drought Stress and Plant Nutrition. Plant Stress 5. daryanto, s., Wang, l., Jacinthe, P.-a., 2016: Drought effects on root and tuber production. A meta-analysis. Agr. Water Manage. 176, 122-131. DOI: 10.1016/j.agwat.2016.05.019 dietrich, h., kruger-steden, e., PatZ, c.d., Will, F., rheinBerger, a., hoPF, i., 2007: Increase of sorbitol in pear and apple juice by water stress, a consequence of climatic change? Fruit Processing 6, 348-355. gisleröd, h.r., selmer-olsen, a.r., mortensen, l.m., 1987: The effect of air humidity on nutrient uptake of some greenhouse plants. Plant Soil 102, 193-196. DOI: 10.1007/BF02370702 grantZ, d.a., 1990: Plant response to atmospheric humidity. Plant Cell Environ. 13, 667-679. DOI: 10.1111/j.1365-3040.1990.tb01082.x greenWay, h., hughes, P.g., klePPer, B., 1969: Effects of water deficit on phosphorus nutrition of tomato plants. Physiol. Plantarum 22, 199-207. DOI: 10.1111/j.1399-3054.1969.tb07856.x hänsch, r., mendel, r.r., 2009: Physiological functions of mineral micro- nutrients (Cu, Zn, Mn, Fe, Ni, Mo, B, Cl). Curr. Opin. Plant Biol. 12, 259-266. DOI: 10.1016/j.pbi.2009.05.006 he, m., diJkstra, F.a., 2014: Drought effect on plant nitrogen and phospho- rus. A meta-analysis. New Phytol. 204, 924-931. DOI: 10.1111/nph.12952 http://dx.doi.org/10.5073/JABFQ.2014.087.022 http://dx.doi.org/10.3390/molecules16086816 http://dx.doi.org/10.18637/jss.v067.i01 http://dx.doi.org/10.1007/s10113-010-0173-x http://dx.doi.org/10.1007/s11306-008-0128-0 http://dx.doi.org/10.1111/j.1672-9072.2007.00556.x http://dx.doi.org/10.1016/j.agwat.2016.05.019 http://dx.doi.org/10.1007/BF02370702 http://dx.doi.org/10.1111/j.1365-3040.1990.tb01082.x http://dx.doi.org/10.1111/j.1399-3054.1969.tb07856.x http://dx.doi.org/10.1016/j.pbi.2009.05.006 http://dx.doi.org/10.1111/nph.12952 Chemical composition of field grown radish 353 hirata, t., cho, y.-m., suZuki, i., toyoda, t., akagi, J.-i., nakamura, y., numaZaWa, s., ogaWa, k., 2018: 4-Methylthio-3-butenyl isothiocyanate mediates nuclear factor (erythroid-derived 2)-like 2 activation by regulat- ing reactive oxygen species production in human esophageal epithelial cells. Food Chem. Toxicol. 111, 295-301. DOI: 10.1016/j.fct.2017.11.022 hu, y., schmidhalter, u., 2005: Drought and salinity. A comparison of their effects on mineral nutrition of plants. J. Plant Nutr. Soil Sci. 168, 541-549. DOI: 10.1002/jpln.200420516 Jensen, c.r., mogensen, v.o., mortensen, g., Fieldsend, J.k., milFord, g.F.J., andersen, m.n., thage, J.h., 1996: Seed glucosinolate, oil and protein contents of field-grown rape (Brassica napus L.) affected by soil drying and evaporative demand. Field Crop Res. 47, 93-105. DOI: 10.1016/0378-4290(96)00026-3 Jing, P., Zhao, s.-J., ruan, s.-y., xie, Z.-h., dong, y., yu, l., 2012: Anthocyanin and glucosinolate occurrences in the roots of Chinese red radish (Raphanus sativus L.), and their stability to heat and pH. Food Chem. 133, 1569-1576. DOI: 10.1016/j.foodchem.2012.02.051 Josse, J., husson, F., 2016: missMDA: A package for handling missing values in multivariate data analysis. J. Stat. Softw. 70, 1-31. DOI: 10.18637/jss.v070.i01 kassamBara, a., mundt, F., 2017: factoextra: Extract and visualize the re- sults of multivariate data analyses. R package version 1.0.4. https://CRAN.R-project.org/package=factoextra. knight, h., treWavas, a.J., knight, m.r., 1997: Calcium signalling in Arabidopsis thaliana responding to drought and salinity. Plant J. 12, 1067-1078. DOI: 10.1046/j.1365-313X.1997.12051067.x koyama, r., itoh, h., kimura syuJi, morioka, a., uno, y., 2012: Augmentation of antioxidant constituents by drought stress to roots in leafy vegetables. HortTechnology 22. kuZnetsova, a., BrockhoFF, P.B., christensen, r.h.B., 2016: lmerTest: Tests in linear mixed effects models. R package version 2.0-33. https://CRAN.R-project.org/package=lmerTest. le, s., Josse, J., husson, F., 2008: FactoMineR: An R package for multi- variate analysis. J. Stat. Softw. 25, 1-18. DOI: 10.18637/jss.v025.i01 lenth, r.v., 2016: Least-Squares Means: The R Package lsmeans. J. Stat. Softw. 69, 1-33. DOI: 10.18637/jss.v069.i01 manchali, s., chidamBara murthy, k.n., Patil, B.s., 2012: Crucial facts about health benefits of popular cruciferous vegetables. J. Funct. Foods 4, 94-106. DOI: 10.1016/j.jff.2011.08.004 marschner, h., 1986: Mineral nutrition of higher plants. Academic Press, London. maZZa, g., miniati, e., 1993: Anthocyanins in fruits, vegetables, and grains. CRC Press, Inc. nikulin, g., kJellström, e., hansson, u.l.F., strandBerg, g., ullerstig, a., 2011: Evaluation and future projections of temperature, precipitation and wind extremes over Europe in an ensemble of regional climate simu- lations. Tellus A 63, 41-55. DOI: 10.1111/j.1600-0870.2010.00466.x nishihara, e., inoue, m., kondo, k., takahashi, k., nakata, n., 2001: Spinach yield and nutritional quality affected by controlled soil water matric head. Agr. Water Manage. 51, 217-229. DOI: 10.1016/S0378-3774(01)00123-8 nuccio, m.l., rhodest, d., mcneil, s.d., hanson, a.d., 1999: Metabolic engineering of plants for osmotic stress resistance. Curr. Opin. Plant Biol. 2, 128-134. DOI: 10.1016/S1369-5266(99)80026-0 PaPetti, a., milanese, c., Zanchi, c., gaZZani, g., 2014: HPLC–DAD– ESI/MSn characterization of environmentally friendly polyphenolic ex- tract from Raphanus sativus L. var. “Cherry Belle” skin and stability of its red components. Food Res. Int. 65, 238-246. DOI: 10.1016/j.foodres.2014.04.046 Prasad, P.v.v., staggenBorg, s.a., ristic, Z., ahuJa, l.r., reddy, v.r., saseendran, s.a., yu, q., 2008: Impacts of drought and/or heat stress on physiological, developmental, growth, and yield processes of crop plants. In: Ahuja, L., Reddy, V.R., Saseendran, S.A., Yu, Q. (eds.), Response of crops to limited water. Understanding and modeling water stress effects on plant growth processes, 301-355. American Society of Agronomy; Crop Science Society of America; Soil Science Society of America, Madison, Wisconsin. PuluPol, l.u., BehBoudian, m.h., Fisher, k.J., 1996: Growth, yield, and postharvest attributes of glasshouse tomatoes produced under deficit irrigation. HortScience 31, 926-929. r core team, 2016: R: A language and environment for statistical comput- ing. R Foundation for Statistical Computing, Vienna, Austria. roBBins, r., keck, a., Banuelos, g., Finley, J., 2005: Cultivation condi- tions and selenium fertilization alter the phenolic profile, glucosinolate, and sulforaphane content of broccoli. J. Med. Food 8. DOI: 10.1089/jmf.2005.8.204 sáncheZ-rodrígueZ, e., del mar ruBio-Wilhelmi, m., cervilla, l.m., Blasco, B., rios, J.J., leyva, r., romero, l., ruiZ, J.m., 2010: Study of the ionome and uptake fluxes in cherry tomato plants under moderate water stress conditions. Plant Soil 335, 339-347. DOI: 10.1007/s11104-010-0422-2 sardans, J., Peñuelas, J., 2004: Increasing drought decreases phosphorus availability in an evergreen Mediterranean forest. Plant Soil 267, 367- 377. DOI: 10.1007/s11104-005-0172-8 schmidt, n., Zinkernagel, J., 2017: Model and growth stage based varia- bility of the irrigation demand of onion crops with predicted climate change. Water 9, 693. DOI: 10.3390/w9090693 schreiner, m., 2005: Vegetable crop management strategies to increase the quantity of phytochemicals. Eur. J. Nutr. 44, 85-94. DOI: 10.1007/s00394-004-0498-7 schreiner, m., Beyene, B., krumBein, a., stütZel, h., 2009: Ontogenetic changes of 2-propenyl and 3-indolylmethyl glucosinolates in Brassica carinata leaves as affected by water supply. J. Agric. Food Chem. 57, 7259-7263. DOI: 10.1021/jf901076h schreiner, m., huyskens-keil, s., Peters, P., schonhoF, i., krumBein, a., Widell, s., 2002: Seasonal climate effects on root colour and com- pounds of red radish. J. Sci. Food Agric. 82, 1325-1333. DOI: 10.1002/jsfa.1189 singleton, v.l., rossi, J.a., 1965: Colorimetry of total phenolics with phos- phomolybdic-phosphotungstic acid reagents. Am. J. Enol. Vitic. 16, 144. sørensen, J.n., Jørgensen, u., kühn, B.F., 1997: Drought effects on the mar- ketable and nutritional quality of carrots. J. Sci. Food Agric. 74, 379-391. DOI: 10.1002/(SICI)1097-0010(199707)74:3<379::AID-JSFA814>3.0.CO;2-Y souci, s.W., Fachmann, W., kraut, h., kirchhoFF, e., 2008: Food com- position and nutrition tables. Die Zusammensetzung der Lebensmittel, Nährwert-Tabellen. La composition des aliments, Tableaux des valeurs nutritives. Medpharm Scientific Publ, Stuttgart. taiZ, l., Zeiger, e., 2010: Plant Physiology. Fifth Edition. Das Original mit Übersetzungshilfen. Ainauer Associates Inc., Sunderland, Massachusetts USA. Wang, y., Frei, m., 2011: Stressed food – The impact of abiotic environmen- tal stresses on crop quality. Agr. Ecosyst. Environ. 141, 271-286. DOI: 10.1016/j.agee.2011.03.017 Wiesner, m., Zrenner, r., krumBein, a., glatt, h., schreiner, m., 2013: Genotypic variation of the glucosinolate profile in pak choi (Brassica rapa ssp. chinensis). J. Agric. Food Chem. 61, 1943-1953. DOI: 10.1021/jf303970k Will, F., dietrich, h., 2006: Optimised processing technique for colour- and cloud-stable plum juices and stability of bioactive substances. Eur. Food Res. Technol. 223, 419-425. DOI: 10.1007/s00217-005-0222-8 Wu, x., Prior, r.l., 2005: Identification and characterization of anthocya- nins by high-performance liquid chromatography-electrospray ioniza- tion-tandem mass spectrometry in common foods in the United States: vegetables, nuts, and grains. J. Agric. Food Chem. 53, 3101-3113. DOI: 10.1021/jf0478861 yao, l.h., Jiang, y.m., shi, J., tomás-BarBerán, F.a., datta, n., singanusong, r., chen, s.s., 2004: Flavonoids in food and their health benefits. Plant Food Hum. Nutr. 59, 113-122. DOI: 10.1007/s11130-004-0049-7 Zancan, s., cesco, s., ghisi, r., 2006: Effect of UV-B radiation on iron http://dx.doi.org/10.1016/j.fct.2017.11.022 http://dx.doi.org/10.1002/jpln.200420516 http://dx.doi.org/10.1016/0378-4290(96)00026-3 http://dx.doi.org/10.1016/j.foodchem.2012.02.051 http://dx.doi.org/10.18637/jss.v070.i01 http://dx.doi.org/10.1046/j.1365-313X.1997.12051067.x http://dx.doi.org/10.18637/jss.v025.i01 http://dx.doi.org/10.18637/jss.v069.i01 http://dx.doi.org/10.1016/j.jff.2011.08.004 http://dx.doi.org/10.1111/j.1600-0870.2010.00466.x http://dx.doi.org/10.1016/S0378-3774(01)00123-8 http://dx.doi.org/10.1016/S1369-5266(99)80026-0 http://dx.doi.org/10.1016/j.foodres.2014.04.046 http://dx.doi.org/10.1089/jmf.2005.8.204 http://dx.doi.org/10.1007/s11104-010-0422-2 http://dx.doi.org/10.1007/s11104-005-0172-8 http://dx.doi.org/10.3390/w9090693 http://dx.doi.org/10.1007/s00394-004-0498-7 http://dx.doi.org/10.1021/jf901076h http://dx.doi.org/10.1002/jsfa.1189 http://dx.doi.org/10.1002/(SICI)1097-0010(199707)74:3<379::AID-JSFA814>3.0.CO;2-Y http://dx.doi.org/10.1016/j.agee.2011.03.017 http://dx.doi.org/10.1021/jf303970k http://dx.doi.org/10.1007/s00217-005-0222-8 http://dx.doi.org/10.1021/jf0478861 http://dx.doi.org/10.1007/s11130-004-0049-7 354 C. Schlering, H. Dietrich, M. Frisch, M. Schreiner, R. Schweiggert, F. Will, J. Zinkernagel content and distribution in maize plants. Environ. Exp. Bot. 55, 266-272. DOI: 10.1016/j.envexpbot.2004.11.004 Zhang, h., schonhoF, i., krumBein, a., guteZeit, B., li, l., stütZel, h., schreiner, m., 2008: Water supply and growing season influence gluco- sinolate concentration and composition in turnip root (Brassica rapa ssp. rapifera L.). J. Plant Nutr. Soil Sci. 171, 255-265. DOI: 10.1002/jpln.200700079 ORCID Christine Schlering https://orcid.org/0000-0003-3605-9875 Helmut Dietrich https://orcid.org/0000-0002-0335-247X Matthias Frisch https://orcid.org/0000-0003-4234-8972 Monika Schreiner https://orcid.org/0000-0002-5629-4429 Ralf Schweiggert https://orcid.org/0000-0003-0546-1335 Frank Will https://orcid.org/0000-0003-0369-3087 Jana Zinkernagel https://orcid.org/0000-0001-6817-091X Address of the corresponding author: Christine Schlering, Department of Soil Science and Plant Nutrition, Depart- ment of Microbiology and Biochemistry, Geisenheim University, Germany E-mail: Christine.Schlering@hs-gm.de © The Author(s) 2019. This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creative- commons.org/licenses/by/4.0/deed.en). http://dx.doi.org/10.1016/j.envexpbot.2004.11.004 http://dx.doi.org/10.1002/jpln.200700079 Supplementary material I Supplementary material Fig. A.1-4: Daily mean weather conditions during the different cultivation periods 2015, 2016-1, 2016-2 and 2017: (1) Mean air temperature [°C], (2) Global radiation sum [W/m²], (3) Daily mean relative humidity [%] and (4) Daily mean evapotranspiration [mm]; DAS = days after sowing. 1