Elevated Atmospheric CO 2 Concentration Inﬂuences the Rooting Habits of Winter-Wheat ( Triticum aestivum L.) Varieties

: The intensity and the frequency of extreme drought are increasing worldwide. An elevated atmospheric CO 2 concentration could counterbalance the negative impacts of water shortage; however, wheat genotypes show high variability in terms of CO 2 reactions. The development of the root system is a key parameter of abiotic stress resistance. In our study, biomass and grain production, as well as the root growth of three winter-wheat varieties were examined under optimum watering and simulated drought stress in a combination with ambient and elevated atmospheric CO 2 concentrations. The root growth was monitored by a CI-600 in situ root imager and the photos were analyzed by RootSnap software. As a result of the water shortage, the yield-related parameters decreased, but the most substantial yield reduction was ﬁrst detected in Mv Karizma. The water shortage inﬂuenced the depth of the intensive root development, while under water-limited conditions, the root formation occurred in the deeper soil layers. The most intensive root development was observed until the heading, and the maximum root length was recorded at the beginning of the heading. The period of root development took longer under elevated CO 2 concentration. The elevated CO 2 concentration induced an accelerated root development in almost every soil layer, but generally, the CO 2 fertilization induced in the root length of all genotypes and under each treatment.


Introduction
One of the most important drivers of climatic changes is the increasing atmospheric CO 2 concentration [1]. The rising temperature is a well-known worldwide phenomenon; however, the changes in precipitation vary greatly by region [2][3][4]. Still, it can be stated that the frequency and the intensity of drought have been increasing in the most important agricultural areas of the world, and this has resulted in reduced harvested yield and has caused uncertainties in this strategic sector [5]. Many studies have focused on the negative effects of a water-limited environment regarding winter-wheat productivity [2,5,6]. Apart from yield reduction, due to intensive drought, water sources can be utilized less efficiently [7,8]. Several recently published studies confirmed that the elevated CO 2 could more or less counterbalance the negative effects of water shortage through the intensification of photosynthesis in C4 plants [9,10] and the regulation of the stomata closure in C3 plants [11,12]. The majority of these studies focused on the aboveground biomass, especially on the harvested yield; however, it must be highlighted that the root system plays an important role in the water and nutrient uptake of the plants [13][14][15]. Rooting depth, as well as the structure of the root system and the depth of intensive root development, are the crucial factors that can influence plants' water uptake. Still, a well-developed root structure is not the sole determinant of drought tolerance. Roots are often more varied than shoots and are affected by changes in the climate, soil conditions, plant varieties and soil nutrient and water availability throughout the growing season [16]. The accurate examination of the belowground parts of a plant is more complicated than the analysis of the aboveground sary, the natural-light intensity was enhanced by artificial illumination to 500 μmol m -2s-1 at the beginning of the vegetation period, which was gradually increased to 700 μmol m -2s-1 . Thiovit Jet fungicide (Syngenta AG, Basel, Switzerland) (active ingredient: sulfur) and Karate 2.5 WG insecticide (Syngenta AG, Basel, Switzerland) (active ingredient: lambdacyhalothrin) were applied two times (BBCH 23 and 37) [32] following the distributor's recommendations by the dosage against powdery mildew and aphids, respectively. The atmospheric CO2 concentration in the control (ambient) chamber was maintained at ~400 ppm and the gas concentration was enhanced to 750 ppm in the other chamber by using a network of perforated pipes placed at a height of 0.5 m above the plants. The uniform distribution was achieved through ventilation.   was regulated by ventilating the greenhouse chambers' air [31]. Whenever it was necessary, the natural-light intensity was enhanced by artificial illumination to 500 μmol m -2s-1 at the beginning of the vegetation period, which was gradually increased to 700 μmol m -2s-1 . Thiovit Jet fungicide (Syngenta AG, Basel, Switzerland) (active ingredient: sulfur) and Karate 2.5 WG insecticide (Syngenta AG, Basel, Switzerland) (active ingredient: lambdacyhalothrin) were applied two times (BBCH 23 and 37) [32] following the distributor's recommendations by the dosage against powdery mildew and aphids, respectively. The atmospheric CO2 concentration in the control (ambient) chamber was maintained at ~400 ppm and the gas concentration was enhanced to 750 ppm in the other chamber by using a network of perforated pipes placed at a height of 0.5 m above the plants. The uniform distribution was achieved through ventilation.   The air temperature and the additional light intensity of the greenhouse chambers were automatically regulated. The air temperature was increased from the initial 10-12 • C to 24-26 • C over 16 weeks, while air humidity was maintained between 60% and 80% and was regulated by ventilating the greenhouse chambers' air [31]. Whenever it was necessary, the natural-light intensity was enhanced by artificial illumination to 500 µmol m −2 s −1 at the beginning of the vegetation period, which was gradually increased to 700 µmol m −2 s −1 . Thiovit Jet fungicide (Syngenta AG, Basel, Switzerland) (active ingredient: sulfur) and Karate 2.5 WG insecticide (Syngenta AG, Basel, Switzerland) (active ingredient: lambdacyhalothrin) were applied two times (BBCH 23 and 37) [32] following the distributor's recommendations by the dosage against powdery mildew and aphids, respectively. The atmospheric CO 2 concentration in the control (ambient) chamber was maintained at~400 ppm and the gas concentration was enhanced to 750 ppm in the other chamber by using a network of perforated pipes placed at a height of 0.5 m above the plants. The uniform distribution was achieved through ventilation.
The containers were separated into two parts of equal size (optimum watered and drought-stressed) by using water-insulating PVC foil (thickness 1 mm).
The water-holding capacity of the soil was determined by using the gravimetric method before starting the experiment and the control treatments were watered until optimum (60%) soil-water content [SWC]. The water content of the soil was monitored by 5 TA sensors (Decagon Devices Ltd., Pullman, WA, USA) at 3 depths (30, 60 and 90 cm). Water-stressed plants did not receive additional watering after the planting until the volumetric soil-water content dropped below 8-10% (average of the three depths). The plants in the drought-stress treatment were irrigated first at the BBCH 51 stage, 68 days after the planting. Afterwards, halved water doses compared to the control were applied to the stress-treated stands.

Measurements
Without overlapping, three transparent polycarbonate tubes were set up in the containers in a horizontal position, at 30, 60 and 90 cm soil depths (Figure 1a). The root development/turnover was monitored every two weeks in the same positions of the tubes by a CI-600 in situ root imager (CID-Bioscience Ltd., Camas, WA, USA). The plant phenophases were ranked according to the BBCH scale (Table 1) [32]. RootSnap software (CID-Bioscience Ltd., Camas, WA, USA) was applied for image processing and to determine the root length ( Figure 3). Simultaneously, soil temperature and soil-water content were monitored continuously in the three layers by 5 TE sensors and EM50 data loggers (Decagon Devices Ltd., Pullman, WA, USA). The containers were separated into two parts of equal size (optimum watered and drought-stressed) by using water-insulating PVC foil (thickness 1 mm).
The water-holding capacity of the soil was determined by using the gravimetric method before starting the experiment and the control treatments were watered until optimum (60%) soil-water content [SWC]. The water content of the soil was monitored by 5 TA sensors (Decagon Devices Ltd., Pullman, WA, USA) at 3 depths (30, 60 and 90 cm). Water-stressed plants did not receive additional watering after the planting until the volumetric soil-water content dropped below 8-10% (average of the three depths). The plants in the drought-stress treatment were irrigated first at the BBCH 51 stage, 68 days after the planting. Afterwards, halved water doses compared to the control were applied to the stress-treated stands.

Measurements
Without overlapping, three transparent polycarbonate tubes were set up in the containers in a horizontal position, at 30, 60 and 90 cm soil depths (Figure 1a). The root development/turnover was monitored every two weeks in the same positions of the tubes by a CI-600 in situ root imager (CID-Bioscience Ltd., Camas, WA, USA). The plant phenophases were ranked according to the BBCH scale (Table 1) [32]. Ripening, early dough RootSnap software (CID-Bioscience Ltd., Camas, WA, USA) was applied for image processing and to determine the root length ( Figure 3). Simultaneously, soil temperature and soil-water content were monitored continuously in the three layers by 5 TE sensors and EM50 data loggers (Decagon Devices Ltd., Pullman, WA, USA). thousand-kernel weight was measured by a digital balance (ME 1002E, Mettler-Toledo Ltd., Worthington, OH, USA), and the harvest index (Equation (1)) was calculated.
Harvest index = grain yield (g) total aboveground biomass (g) * 100 (1) Relative changes in the root length in response to elevated carbon dioxide levels were calculated as: where A is the root length at the 400 ppm CO 2 level, Ex is the root length at the 700 ppm CO 2 level.

Analysis
The experimental design (2 × 2 × 3) consisted of two CO 2 levels, two watering levels and three genotypes. The rooting habits of a closed plant stand were analyzed. The scanning was always carried out in three replications in the same position. The effects of the tested factors on yield parameters were determined by multi-way ANOVA and means were compared by Tukey's HSD test (p ≤ 0.05). The significant differences in the length parameters between the water treatments and CO 2 levels were evaluated by Student's t-test and the significant differences between the root lengths measured in the plant growth stages were analyzed by one-way ANOVA followed by Tukey's HSD test.

Effects of Water Shortage and Elevated Atmospheric CO 2 Concentration on Yield Parameters
The biomass (BM), grain yield (GY), thousand-kernel weight (TKW) and harvest index (HI) of the individual treatments are summarized in Table 2. The ANOVA shows that the effects of the watering levels, genotypes and the CO 2 concentrations significantly influenced the grain yield, but their interactions were not statistically significant. The effects of the watering and the genotypes were significant on the aboveground biomass, but the ANOVA shows that the interactions of the factors remained insignificant (Tables S1 and S2). Different CO 2 levels and irrigation regimes were assumed to have contradictory effects on the yield-related parameters. Water shortage resulted in decreased BM and GY both under ambient and elevated CO 2 concentrations. The ratio of the yield reduction did not differ between the two CO 2 concentrations. Compared to the well-watered plant stands, the biomass of Mv Pálma and Mv Karéj decreased by 8.02% and 10.03% under ambient CO 2 concentration and 11.2% and 10.45% under elevated CO 2 , respectively, in the drought-stress treatment (Table 2).
Generally, the simulated drought stress did not significantly affect the TKW and the HI, but the effects of the genotypes and watering levels were statistically significant in terms of both parameters (Tables S3 and S4). In the case of Mv Pálma, water shortage led to a 12.75% and 13.75% increase in TKW compared with the non-stressed controls under 400 ppm and 750 ppm CO 2 concentration, respectively. The highest yielding capacity was observed for Mv Karizma, and this genotype showed the lowest yield reduction under the simulated drought stress. In Mv Karizma, the elevated CO 2 concentration played a role in counterbalancing the negative effects of the water shortage. Taking the average of the three examined varieties, the CO 2 enrichment increased the biomass (by 4.57% and 4.97% under normal watering and drought stress, respectively) as well as the grain yield (by 8.1% and 8.96% under normal watering and drought stress, respectively) ( Table 2). Improved harvest indices (4.3% and 4.0% under normal watering and drought-stress conditions, respectively) could be determined as a result of the observed tendencies in biomass and grain weight, which would be highly favorable for nutrient-and water-utilization efficiency. In Mv Karéj, no differences were observed between the two water treatments in terms of CO 2 reactions. The interaction of water shortage and CO 2 fertilization showed opposite tendencies in the other two varieties. The increase in the yield parameters was more intense under the control water supply in Mv Pálma, while the positive effect of the elevated CO 2 compared with the control was more significant under drought-stress conditions in Mv Karizma ( Table 2).

Dynamics of the Root Development of Winter-Wheat Varieties under Optimum Watering and Drought-Stressed Conditions Grown at Ambient and Elevated CO 2 Concentrations
Under ambient CO 2 concentration, the root length of Mv Pálma continuously developed in the well-watered treatment after planting until it reached its maximum level at the BBCH 51 stage in the two upper soil layers (30 and 60 cm), while a faster root development was detected in the drought-stressed treatment when the root length did not significantly increase after the BBCH 29 and BBCH 37 stages at 30 and 60 cm, respectively. (Table 3). Significantly higher root length was observed at 30 cm between the BBCH 29 and 51 stages under drought-stressed conditions, but the root length did not significantly differ between the water treatment at 60 cm except at the BBCH 21 and 83 stages. At 90 cm, the root length of Mv Pálma reached its maximum (does not significantly increase further) at BBCH 69 under optimum irrigation and at BBCH 51 under drought-stressed conditions, and significantly higher root length was measured under drought stress between BBCH 37 and BBCH 83 than in the control treatment (Table 2). Under elevated CO 2 , the root length of Mv Pálma was consequently higher in the well-irrigated treatment at 30 cm between BBCH 21 and BBCH 77 than under drought stress ( Table 3). The root length did not significantly increase at 30 or 90 cm after the BBCH 37 stage, but at 60 cm, the root length did not significantly increase after the BBCH 29 stage under drought-stressed conditions.
Under ambient CO 2 concentration, the root length of Mv Karéj reached its maximum in the upper soil layer (30 cm) at the BBCH 29 stage in both water treatments, and the root length was significantly higher under optimum watering at BBCH 51 and BBCH 77 than in the drought-stressed treatment (Table 4). In the middle layer (60 cm), the root length did not significantly increase after BBCH 37 and BBCH 29 in the control and drought-stressed treatments, respectively, and a significant difference between the water treatments can be observed only at the BBCH 29 phenophase. In Mv Karéj, the maximum root length was observed at the BBCH 51 stage at 90 cm in both water treatments, but significantly higher   Under 750 ppm CO 2 , the root length of Mv Karéj reached its maximum at the BBCH 37 stage at each soil layer under optimum and drought-stressed conditions. No significant differences were observed between the water treatments at the 30 cm soil layer after the BBCH 21 stage, but the CO 2 enrichment resulted in a significant decrease in root length in each phenophase at 60 and 90 cm ( Table 4).
The root development of Mv Karizma showed the highest variability among the studied varieties under ambient CO 2 concentration in different soil layers and irrigation regimes. Intensive root development was observed from planting until the BBCH 29 stage at 30 cm, but the root length did not increase further during vegetation. Oppositely, a significant reduction in root length was determined from the BBCH 69 stage ( Table 5). The root length at 30 cm was significantly higher under optimum watering than under drought stress in each phenophase. At 60 cm, the root-length development was faster under drought-stress conditions than under optimum irrigation, and the measured data did not significantly increase after BBCH 37 and BBCH 51 under drought stress and optimum watering, respectively. The water shortage induced a significantly more developed root system in Mv Karizma for each phenophase at 60 cm ( Table 5). The root system developed intensively until the BBCH 51 stage at 90 cm under both watering regimes, then a root turnover can be observed in the BBCH 69 and 77 stages, but another intensive root formation was detected at the BBCH 83 stage. Significantly higher root-length values were measured in Mv Karizma under drought-stressed conditions at 90 cm than in the control treatment from BBCH 29 until BBCH 83, except BBCH 37 at 400 ppm CO 2 concentration. Under elevated CO 2 levels (750 ppm), a delay can be observed at 30 cm in the time that the root length takes to reach its maximum compared to the ambient treatment. The maximum values were observed at the BBCH 37 stage in both water treatments. The water supply had no significant effects on the root length of Mv Karizma during vegetation at 30 cm under 750 ppm CO 2 between the BBCH29 and BBCH77 stages. At 60 cm, the highest root length was observed at the end of vegetation under optimum watering, while the highest data were measured at the BBCH 37 and BBCH 51 stages under limited water supply. The water shortage induced a significant increase in root length in each phenophase of Mv Karizma under CO 2 enrichment. The trends in root development at 90 cm were similar to that at 60 cm, and the water shortage resulted in an increase in root length between the BBCH 37 and BBCH 77 stages (Table 5). Lowercase letters indicate significant differences between drought and control treatments within the same soil layer and CO 2 treatment (Student's t-test) (p < 0.05) and the numbers in superscripts indicate significant differences between phenophases (HSD test (p ≤ 0.05) (n = 3), n.r., no roots were observed.

CO 2 Reactions of Winter-Wheat Genotypes during Vegetation at Different Soil Layers under Well-Watered and Drought-Stressed Conditions
The responses of Mv Pálma to the elevated CO 2 showed a variability during the vegetation period and significant alterations were determined between the well-watered and drought-stressed treatments (Figure 4). The CO 2 response was positive and significant from the sowing until the BBCH 21 stage at 30 and 60 cm, and this tendency can be observed for 90 cm at the BBCH 29 stage. Generally, the CO 2 enrichment induced faster root development in Mv Pálma.
Positive and statistically significant CO 2 reactions were observed between the BBCH 37 and BBCH 83 growth stages under optimum watering at 90 cm, and between BBCH 51 and BBCH 69 at 30 cm (Figure 4). The CO 2 fertilization resulted in a significant decrease in the root length of Mv Pálma under drought-stressed conditions at the BBCH 37 stage, and at the later stages of vegetation in each soil layer except in the BBCH 51 stage at 60 cm ( Figure 4).
Overall, the CO 2 responses of Mv Karéj were negative in each phenophase, and the reactions were not influenced by the watering (Figure 5). The negative CO 2 responses were more intensive at the end of vegetation, which indicates that the CO 2 fertilization influenced the water balance of plants and induced faster root turnover. The unfavorable impacts of the CO 2 enrichment were more intensive under drought-stressed conditions on the root length, but this process could be a component of the survival strategy of this genotype ( Figure 5).
between drought and control treatments within the same soil layer and CO2 treatment (Student's t-test) (p < 0.05) and the numbers in superscripts indicate significant differences between phenophases (HSD test (p ≤ 0.05) (n = 3), n.r., no roots were observed.

CO2 Reactions of Winter-Wheat Genotypes during Vegetation at Different Soil Layers under Well-Watered and Drought-Stressed Conditions
The responses of Mv Pálma to the elevated CO2 showed a variability during the vegetation period and significant alterations were determined between the well-watered and drought-stressed treatments (Figure 4). The CO2 response was positive and significant from the sowing until the BBCH 21 stage at 30 and 60 cm, and this tendency can be observed for 90 cm at the BBCH 29 stage. Generally, the CO2 enrichment induced faster root development in Mv Pálma. Positive and statistically significant CO2 reactions were observed between the BBCH 37 and BBCH 83 growth stages under optimum watering at 90 cm, and between BBCH 51 and BBCH 69 at 30 cm ( Figure 4). The CO2 fertilization resulted in a significant decrease in the root length of Mv Pálma under drought-stressed conditions at the BBCH 37 stage, and at the later stages of vegetation in each soil layer except in the BBCH 51 stage at 60 cm ( Figure 4).
Overall, the CO2 responses of Mv Karéj were negative in each phenophase, and the reactions were not influenced by the watering (Figure 5). The negative CO2 responses were more intensive at the end of vegetation, which indicates that the CO2 fertilization influenced the water balance of plants and induced faster root turnover. The unfavorable impacts of the CO2 enrichment were more intensive under drought-stressed conditions on the root length, but this process could be a component of the survival strategy of this genotype ( Figure 5).
The CO2 reactions of Mv Karizma showed variability during vegetation, and this parameter was influenced significantly by the watering (Figure 6). Roots can be observed only at 30 cm at the BBCH 17 stage, and significant positive CO2 responses were observed in terms of the root development under drought-stressed conditions, while the opposite tendency was observed under optimum irrigation. The negative impacts of the CO2 enrichment were detected at 30 cm under optimum irrigation in the BBCH 21 stage, but significant positive responses to both watering levels were observed at 60 cm . The reactions of Mv Karizma to CO2 enrichment was consequently negative at 30 cm, and this trend was not influenced by the intensity of watering, while the CO2 reactions were tendentially negative between the BBCH29 and BBCH 83 stages under drought-stressed conditions and positive at 60 cm under controlled irrigation. Negative CO2 reactions were determined at 90 cm at both watering levels, but this tendency was significant at first when plants reached the BBCH 51 stage ( Figure 6).  The CO 2 reactions of Mv Karizma showed variability during vegetation, and this parameter was influenced significantly by the watering (Figure 6). Roots can be observed only at 30 cm at the BBCH 17 stage, and significant positive CO 2 responses were observed in terms of the root development under drought-stressed conditions, while the opposite tendency was observed under optimum irrigation. The negative impacts of the CO 2 enrichment were detected at 30 cm under optimum irrigation in the BBCH 21 stage, but significant positive responses to both watering levels were observed at 60 cm. The reactions of Mv Karizma to CO 2 enrichment was consequently negative at 30 cm, and this trend was not influenced by the intensity of watering, while the CO 2 reactions were tendentially negative between the BBCH29 and BBCH 83 stages under drought-stressed conditions and positive at 60 cm under controlled irrigation. Negative CO 2 reactions were determined at 90 cm at both watering levels, but this tendency was significant at first when plants reached the BBCH 51 stage (Figure 6).

Discussion
CO 2 molecules in the atmosphere are essential substrates of photosynthesis; therefore, in general, increasing the concentration of CO 2 leads to improved assimilation and crop productivity. Previously, the effects of elevated CO 2 have been widely investigated in various field crops, such as wheat [33,34], maize [35,36], rice [37], sorghum [35], etc. Primarily, these studies determined that CO 2 fertilization improved biomass and grain production, which is also confirmed by our experiment. It has already been correspondingly established that there are differences in the CO 2 reactions of cereal varieties, especially under stress conditions [37][38][39]. In our experiment, Mv Karizma showed the most intense responses to the increased CO 2 level. Previously conducted studies also highlighted that the increasing CO 2 concentration could partly counterbalance the negative impact of abiotic stresses, such as drought or heat [7,40]. In our experiment, the CO 2 response was variety dependent and the most favorable reactions were observed for Mv Karizma. The yield reduction under drought stress dropped by approximately one half under elevated CO 2 compared to ambient conditions. The physiological background of this result might be that the high CO 2 concentration modifies the intensity of the photosynthesis [39,41], decreases the stomatal conductance and increases the evaporation [42]. Based on this, it can be concluded that although the influence of elevated CO 2 on the aboveground parts of the plants is well-known, the development of the root system should also offset the increased water loss and nutrient uptake. There are differences in the rooting habits of wheat varieties, even under ambient conditions [43,44]; furthermore, the cultivars also give diverse responses to CO 2 [45,46]. Considering the above, it would be highly important to reveal how these effects interact under various irrigation regimes.
The novel aspect of our study is that the effects of the elevated CO 2 concentration were combined with the drought stress, and the differences in the CO 2 responses of the tested varieties were also determined.
Asseng et al. [47] described in a minirhizotron study that the fastest root-growth development was observed between the visible terminal spikelet stage (BBCH 51) and anthesis (BBCH 65), and the maximum root length was measured beyond 90 days after sowing. Based on our results, the most intensive root growth occurred in the vegetative phase, while the date of the highest root length was influenced by the atmospheric CO 2 as well. Under ambient conditions, the maximum root length could be observed at the BBCH 37 stage, but the elevated CO 2 resulted in a constant increase in the root length until the BBCH 51 phase. Uddin et al. [48] described that elevated CO 2 resulted in the intensive rootlength formation of wheat plants, especially in the upper soil layer. In our study, however, significant differences were observed between the examined varieties in terms of CO 2 reactions. Our study partly confirmed this trend: although accelerated root development was detected in each soil layer induced by the CO 2 fertilization, overall, the root length was reduced under elevated CO 2 concentration. Mitchell et al. [49] proposed that the potential reason for this phenomenon could be that the surplus assimilates availability for extra root growth under elevated CO 2 . Manschadi et al. [50] found in a wheat experiment, in root-observation chambers, that genotypes that were developed for dry conditions had longer root systems in deep soil layers. In our study, the intensive root formation in the deeper soil layers (60 and 90 cm) under drought-stressed conditions indicate the good adaptive capacity of the examined genotypes, especially that of Mv Pálma and Mv Karizma, to a water-limited environment.
In the root characteristics, no differences were detected in near-isogenic lines in the case of the reduced height genes (Rht1 and Rht2) in Western Australia [51], whereas in Argentina, an increase in the total root length and root weight was associated with reduced height [52]. In our experiment, Mv Karizma had the greatest root length, especially under dry conditions, in the deepest soil layer. This might be due to the presence of the lessefficient dwarfing gene (Rht2) in its genome. Mv Pálma had the less-developed root system, even under elevated CO 2 concentration, which could be in accordance with the presence of the Rht8 gene.

Conclusions
As a consequence of the water shortage, the depth of the intensive root development remained in the deeper soil layers and under elevated CO 2 concentration, the distribution of the root system was more homogeneous in the whole soil profile than under ambient conditions. The elevated CO 2 concentration induced an accelerated root formation, but considering the whole vegetation period, the CO 2 fertilization had a reducing effect on the root length.
The most intensive root development was detected in the vegetative stage of the plants. The maximum root length was observed at the beginning of the heading, and this period took longer under elevated CO 2 concentration. After the heading, the development of the new roots became slower and intensive root turnover was observed. Increased root formation was determined at the maturity stage in almost every treatment, which could be in association with the re-growth of the stubble under field conditions. The water shortage generally stimulated root growth, but the depth of the intensive root development proved to be variety dependent and this phenomenon was more intensive under ambient CO 2 concentration.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/su14063304/s1, Table S1: The effects of the tested factors and their interactions on the aboveground biomass, Table S2: The effects of the tested factors and their interactions on the grain yield, Table S3: The effects of the tested factors and their interactions on the thousand-kernel weight, Table S4: The effects of the tested factors and their interactions on the harvest index.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.