Elevated Root-Zone Dissolved Inorganic Carbon Alters Plant Nutrition of Lettuce and Pepper Grown Hydroponically and Aeroponically

: Enhancing root-zone (RZ) dissolved inorganic carbon (DIC) levels of plants grown hydroponically and aeroponically can increase biomass accumulation but may also alter plant nutrient uptake. These experiments investigated how bicarbonate (HCO 3 − ) added to a hydroponic nutrient solution and CO 2 gas added to an aeroponic system a ﬀ ected biomass and nutrient concentrations of lettuce and pepper plants. Applying high RZ HCO 3 − concentrations (20 mM) to lettuce plants grown hydroponically decreased foliar N, P, Cu, K, Mn and Zn concentrations, concurrent with decreased biomass accumulation (50% less than control plants). On the contrary, 1 mM RZ HCO 3 − promoted biomass accumulation (10% more than control plants), but this could not be attributed to higher tissue nutrient concentrations. While elevated RZ CO 2 did not alter biomass accumulation and nutrient concentrations in pepper grown aeroponically, it decreased foliar Mg and S concentrations in lettuce grown aeroponically even though nutrient contents (concentration x biomass) did not di ﬀ er between treatments, due to 22% more biomass than control plants. In addition, elevated RZ CO 2 enhanced N, P, Cu and Zn contents relative to control plants, indicating greater uptake of those elements. Nevertheless, there was no consistent relationship between plant growth promotion and altered plant nutrition, suggesting alternative mechanisms of growth regulation.


Introduction
Atmospheric CO 2 is the main form of inorganic carbon assimilated by terrestrial photosynthetic organisms. As water percolates through the soil, it becomes enriched with CO 2 from plant and microbial respiration. Some of this CO 2 can be dissolved in the soil, producing dissolved inorganic carbon (DIC), which is controlled by the partial pressure of CO 2 (pCO 2 ), pH and temperature [1]. The soil inorganic carbon comprises CO 2 in the gas phase, the liquid phase solution containing bicarbonate (HCO 3 − ) and CO 3 − and the carbonate in the solid phase. CO 2 is a weak acid that slowly dissolves basic minerals such as CaCO 3 + . When the CO 2 dissolves in water, at the same time, hydrolysis of HCO 3 − is converted into pedogenic carbonates with other salts, usually with Ca + and Mg 2+ [2]. Although it is well known that CO 2 is absorbed through the stomata in the leaf, many studies have observed that roots are able to take up dissolved inorganic carbon (DIC) contained in soils as well as gaseous CO 2 respired by the roots. Altering root-zone (RZ) DIC concentrations has both positive and negative impacts on plant growth. The effects of altered RZ DIC depend on the enrichment system, plant species, pH, air temperature, irradiance, mineral nutrition, abiotic stresses such as high irradiance or salinity, the duration of RZ DIC enrichment, DIC concentration applied and the RZ DIC concentration. Across with low bioavailability of plant nutrients, high Ca content and high alkalinity [9,10]. In calcareous soils with high pH (>7), the availability of Fe and other essential micronutrients such as Zn, Mn, and Cu are usually low due to their precipitation as oxides or carbonates [10]. On the other hand, Ca, Mg and K are less available and Al, Fe and Mn can cause plant toxicity in acidic soils [11]. In soilless culture systems, a well-prepared nutrient solution will provide every element to the plant in an optimum ratio. In general, increasing the concentration of HCO 3 − by 5, 10, 20 mM decreased K, NO 3 − , Mg, S, P and Fe uptake, but not Ca uptake, in sorghum and maize plants when nutrient solution pH was~8. Decreased uptake of nutrients was correlated with lower shoot and root biomass [12,13]. On the other hand, 5.68 mM HCO 3 − increased tomato biomass accumulation, leaf blade and root N content, K content in all tissues and Ca content in roots, shoot and leaf blades, although P content did not differ from the control [4]. While the variability of some elemental concentrations could be due to the different species, nutrient solutions and experimental design used, it seems that HCO 3 − application increases Ca in all cases. CaHCO 3 − formation is very common in calcareous soils because Ca precipitates with HCO 3 − , which could also occur in a hydroponic system with high pH and high [HCO 3 − ]. Other studies where RZ CO 2 gas was applied focused mainly on nitrogen metabolism with no major data related to other nutrient elements. Tomato plants grown for 60 days at elevated RZ CO 2 (2500, 5000 and 10,000 ppm) decreased root N, P, K, Ca and Mg concentrations after 60 days, compared to those exposed to ambient RZ CO 2 of 370 ppm [14]. Although tissue nutrient concentrations have been measured when applying high RZ HCO 3 − concentrations, little is known regarding the effects of elevated RZ CO 2 on nutrient concentrations. Since plant growth and development largely depend on the combination and concentration of mineral nutrients available in the RZ, this study aims to understand how HCO 3 − added to hydroponic nutrient solution or CO 2 gas into the RZ of lettuce and pepper plants grown aeroponically affects foliar and root tissue nutrient concentrations. Based on previous results, high (>10 mM) HCO 3 − concentrations in the hydroponic RZ are expected to decrease nutrient uptake by increasing nutrient solution pH (>7). On the contrary, since RZ enrichment with CO 2 is not expected to alter the pH of the nutrient solution, no effect on plant nutrient uptake should occur.

Materials and Methods
To determine the effect of HCO In the first two experiments, bicarbonate was applied in the form of NaHCO 3 at 0, 1, 5, 10 and 20 mM, and in a third experiment, bicarbonate was applied in the form of NaHCO 3 at 0, 1, and 20 mM.
The lids (43 × 33 cm) were modified with four 2.5 cm holes in each quadrant to hold four plants per box (4 × 0.14 m −2 ). Two boxes were used for each treatment and were completely randomized. In the middle of the lid, an additional hole was cut to accommodate a closed cell foam piece, through which a pipe with an external diameter of 6 mm was inserted. The end of the pipe outside the box was connected to an aquarium air pump (All Pond Solution Ltd, Middlesex, UK) which continuously supplied ambient air (Flow rate: 3.2 L min −1 ) to add O 2 to the nutrient solution as well as stirring it. The medium was changed every 3-4 days, and the pH was maintained at 6.4 (at which CO 2 and HCO 3 − concentrations are equivalent) by adjusting the pH via dropwise addition of 1 N HCl or NaOH once every day. Eight experiments were carried out between 2016 and 2018-four with lettuce and four with pepper-using an aeroponic growing system. Lettuce (Lactuca sativa cv. Consul) and pepper (Capsicum annuum (L.) cv. Bellboy F1) seeds were individually sown in 150-cell plug trays in 2 cm × 2 cm × 4 cm rockwool cubes (Growell, Ltd, UK). Pepper was germinated in the glasshouse and lettuce in the CE room. The glasshouse was naturally lit with automated supplementary lighting supplied by ten high-pressure sodium lamps (600 W Greenpower, Osram, St Helens, UK) when the PPFD was <400 µmol m −2 s −1 for a 12 h photoperiod (08:00 hrs to 20:00 hrs). Temperature ranged between 18 and 25 • C and relative humidity between 30% and 55%. Plants were transferred to the system at the 4-leaf stage in lettuce and at the 2-4-leaf stage in pepper. After transplanting, two different [CO 2 ]-400 and 1500 ppm-were applied into each bin. The system consisted of an enriched channel supplemented with CO 2 and a non-enriched channel supplied only with compressed air. The air from the enriched channel was completely mixed in a mixing box before entering the aeroponic system. The [CO 2 ] in the mixing box was monitored continuously using a CO 2 gas analyser (PP Systems, WMA-4). To prevent leakages, the lid was sealed with self-adhesive rubber foam around the rim. The air above the lid and at the shoot base was routinely sampled with a Li-Cor 6400 infra-red gas analyser, with no significant difference compared to the ambient air. The pH was controlled everyday between 5.8 and 6.3 with the necessary dropwise application of HCl or NaOH.

Plant Measurements
Shoot fresh weight was determined after 10 days of treatment, along with leaf area using a leaf area meter (Li-Cor Model 3100 Area Meter, Lincoln, NE, USA). Roots were collected, rinsed with dH 2 O and dried with absorbent paper. Both shoot and root material were then dried at 70 • C for 4 days to record dry weight and stored in airtight containers to provide samples for nutrient analysis.

Nutrient Analysis
For the bicarbonate experiment, the median four plants were taken from each of the 0, 1 and 20 mM NaHCO 3 − treatments and sent to NRM Technologies Ltd. (Bracknell, UK) for C013 Plant Foliar Suite Analysis, incorporating analysis for: total N, P, K, Mg, Ca, S, Cu, Mn, Zn, Fe and B. For the lettuce aeroponic experiment, macronutrients (Ca, K, Mg, Na, P and S) were analysed via acid microwave digestion followed by using an Inductively Coupled Plasma Optical Emission Spectrometer (ICP-OES). Nitric acid (HNO 3 ) was used to decompose all organic matter to CO 2 . Ball-milled (MM400, Retsch, Haan, Germany) oven dried leaf and root tissue (0.25 g) was weighed in acid-washed and rinsed reaction vessels. A volume of 5 mL of 100 % HNO 3 (Aristar grade) was added and left for 15 min in a fume hood until the initial reaction was finished. Vessels were sealed and weighed and then placed in the rotor in a MARS 6 microwave (CEM, Buckingham, UK). Vessels were heated to 200 • C over 15 min, and then held at 200 • C for another 15 min. After cooling down, vessels were weighed again to note weight loss. Samples and blank solutions were then diluted in two steps to first 20% HNO 3 and second to the final concentration of 2% HNO 3 by using MilliQ water. To analyse nutrients, an ICP-OES (iCAP 6300, Thermo Scientific, Massachusetts, USA) with axial view configuration was used. To validate the digestion, tomato and spinach leave samples with known nutrient concentrations were run and the recovery detected through the ICP-OES was used to calculate final sample concentration. The element reference standard solutions were prepared daily from 1000 mg L −1 stock solutions. The same protocol was used for leaves and roots of pepper plants grown under elevated and ambient RZ CO 2 .
In each experiment, shoot and root nutrient content was calculated multiplying tissue nutrient concentration by its dry mass. When nutrient concentrations of both tissues were not available, only shoot nutrient content was determined.

Total Nitrogen and NO 3 − Analysis
Total leaf and root N in % were analysed using an Elemental Analyser (VARIO-El elemental analyser). Oven-dried leaf and root tissue samples were wrapped in aluminium capsules and dropped into a furnace held at 905 • C onto CuO with a pulse of O 2 and a constant flow of Helium carrier gas. N was converted to gas (N 2 ) and a pure copper reduction unit after the furnace reduced any conversion of NO x to N 2 . N 2 was measured in a TCD (total dissolved carbon) detector positioned at the end of the elemental analyser and peak areas were compared to standards and amounts of N calculated. NO 3 − concentration in the leaf tissue was measured as previously described [15]. Oven-dried and ground samples (0.1 g) were suspended in 10 mL dH 2 O. The suspensions were incubated at 80 • C for 2 h. After mixing, the samples were centrifuged at 1300× g for 5 min and the supernatants were decanted and saved for analysis. Aliquots (0.1 mL) of the extracts were pipetted into 10 mL tubes and mixed thoroughly with 0.4 mL 5% (w/v) salicylic acid-sulphuric acid. After 20 min at room temperature, 9.5 mL of 8% (w/v) NaOH was added to raise the pH above 12. Samples were cooled to room temperature for 20 min and absorbance at 410 nm was determined in an Ultrospec 2100 Pro UV/visible spectrophotometer (GE Healthcare UK Ltd, Little Chalfont, UK). Standards containing 10 to 120 mg/L in a 0.1 mL aliquot were analysed with each sample.

Root-zone Bicarbonate Enrichment of Plants Grown Hydroponically
In lettuce, 1 mM and 5 mM HCO 3 − increased shoot fresh weight, shoot dry weight and leaf area by~10%, while root dry weight was significantly lower in control plants and those exposed to 1 mM HCO 3 − than other treatments (Table 1). Total plant dry weight (data not shown) was~10% higher under 1 and 5 mM HCO 3 − but 20% lower at 20 mM HCO 3 − . 33 In pepper, 1 mM HCO 3 − significantly increased shoot fresh weight, shoot dry weight, and leaf area by 10% compared to control plants. Furthermore, high HCO 3 − concentrations (>10 mM) decreased all these variables compared to control plants. However, root dry weight of control plants was significantly lower than those exposed to all bicarbonate concentrations ( Table 2). Since lettuce was more responsive to bicarbonate, nutrient analyses were performed just in lettuce.  59 In a replicate experiment, lettuce shoot dry weight, leaf area and shoot relative growth rate were 10%, 20% and 20% higher at 1 mM HCO 3 − and~50% lower at 20 mM HCO 3 − compared to control plants (Table 3) (Table 4).   Table 4. Total nutrient content of lettuce plants exposed to 0, 1 and 20 mM HCO3 -. Experiment 3.
Values are means ± S.E. of 4 replicates, with different letters within a column denoting significant differences between means (post-hoc LSD p < 0.05).

Root-zone Carbon Dioxide Enrichment of Plants Grown Aeroponically
In lettuce, four similar experiments showed that elevated RZ CO2 (1500 ppm) significantly increased dry shoot biomass by approximately ~ 20% compared to those grown with 400 ppm RZ CO2, regardless of the variety and location of the experiment (Table 5). On the other hand, elevated RZ CO2 did not significantly alter root dry weight in any of the experiments.   Values are means ± S.E. of 4 replicates, with different letters within a column denoting significant differences between means (post-hoc LSD p < 0.05).

Root-Zone Carbon Dioxide Enrichment of Plants Grown Aeroponically
In lettuce, four similar experiments showed that elevated RZ CO 2 (1500 ppm) significantly increased dry shoot biomass by approximately~20% compared to those grown with 400 ppm RZ CO 2 , regardless of the variety and location of the experiment (Table 5). On the other hand, elevated RZ CO 2 did not significantly alter root dry weight in any of the experiments. In pepper, applying 1500 ppm CO 2 to the root-zone of pepper grown aeroponically did not significantly change shoot dry weight, total leaf area and root dry weight, compared to control plants grown at~400 ppm RZ CO 2 . Experiment 11 showed lower shoot fresh/ dry weight, total leaf area and root dry weight compared to control plants (Table 6). In Experiment 7, shoot dry weight of lettuce grown aeroponically was 22% higher under elevated RZ CO 2. Elevated RZ CO 2 significantly decreased foliar Mg and S concentrations by approximately 24% and 15% respectively, tended to decrease Ca and K concentrations by 20% and 15% respectively, while P concentration did not change. In the roots, elevated RZ CO 2 tended to increase foliar Ca concentration by 10%. No major changes were found in the other macronutrients in either root or shoot tissues (Figure 2A,B). Regarding micronutrients, elevated RZ CO 2 tended to increase foliar Zn concentration by 15% and tended to decrease Mn and Fe concentrations by 10% ( Figure 2C,D). No other treatment differences were detected in micronutrient concentrations in either root or shoot tissues. Elevated RZ CO2 tended to decrease total shoot N concentration by 5%, but significantly increased root N concentration by 5%. Elevated RZ CO2 significantly increased root NO3concentrations and there was a greater proportion of NO3 -. Elevated RZ CO2 tended to decrease C/N ratio in both roots and leaves (Figure 3). With RZ CO2 enrichment, lettuce had higher shoot Zn (50%), Cu (22%) and P (35%) contents, and significantly higher N (25%) content. Root nutrient content did not significantly differ between treatments (Table 7). Elevated RZ CO 2 tended to decrease total shoot N concentration by 5%, but significantly increased root N concentration by 5%. Elevated RZ CO 2 significantly increased root NO 3 − concentrations and there was a greater proportion of NO 3 − . Elevated RZ CO 2 tended to decrease C/N ratio in both roots and leaves (Figure 3). With RZ CO 2 enrichment, lettuce had higher shoot Zn (50%), Cu (22%) and P (35%) contents, and significantly higher N (25%) content. Root nutrient content did not significantly differ between treatments (Table 7).    Pepper plants did not show any significant treatment differences in foliar macronutrient and micronutrient concentrations. In the roots, elevated RZ CO 2 tended to increase Zn and Fe concentrations by 12% and 15% respectively (Figure 4). Elevated RZ CO 2 had similar effects on leaf and root nitrogen concentration as in lettuce, but the differences were not significant. Shoot N concentration tended to decrease by 4%, while root N concentration tended to increase by 5%. C/N ratio was lower in the roots but higher in the leaves under RZ CO 2 ( Figure 5). Pepper plants did not show any significant treatment differences in foliar macronutrient and micronutrient concentrations. In the roots, elevated RZ CO2 tended to increase Zn and Fe concentrations by 12% and 15% respectively (Figure 4). Elevated RZ CO2 had similar effects on leaf and root nitrogen concentration as in lettuce, but the differences were not significant. Shoot N concentration tended to decrease by 4%, while root N concentration tended to increase by 5%. C/N ratio was lower in the roots but higher in the leaves under RZ CO2 ( Figure 5).

Discussion
Whether RZ CO2 enrichment regulates plant growth by altering plant nutrition was tested by growing two common horticultural species (lettuce and pepper) under commercially relevant production systems (hydroponics and aeroponics). Growth promotion of lettuce was correlated with enhanced nutrient concentrations following aeroponic CO2 enrichment, but could not be attributed to enhanced nutrition following hydroponics-enrichment with low (<5 mM) bicarbonate concentrations. Pepper plants showed limited changes in nutrient status independent of method of RZ CO2 enrichment. Even though enhanced crop nutrition is unlikely to account for the growthpromoting effects of RZ CO2 enrichment, it is important to consider likely physiological effects in specific cases where nutrient concentrations did change.
In general, high concentrations of bicarbonate (≥ 5 mM) in the rhizosphere decreased foliar K, Mg, S, P and Fe concentrations, while high leaf tissue Ca concentrations occurred when applying HCO3 -and in plants grown in calcareous soils at high pH [4,12,13]. In this study, applying 20 mM HCO3 -did not change Ca concentration, but K, P, N, Zn, Cu and Mn concentrations decreased remarkably, while B and Mg significantly increased compared to control and 1 mM HCO3 -. Moreover, Fe concentration at 20 mM HCO3 -was higher than at 1 mM HCO3 - (Figure 1). According to the adequate range of shoot nutrient concentrations in lettuce (Table A1), at 20 mM HCO3 -, only Zn was below the limits, while B was above the limits. Although P concentrations exceeded the recommended range, this was independent of the bicarbonate concentrations applied. Thus, HCO3enrichment of the nutrient solution should have caused minimal nutritional stress.
Zn influences many biological processes, including carbohydrate metabolism and cell proliferation, and is an integral component of some enzyme structures, such as carbonic-anhydrase, alcohol dehydrogenase, and glutamate dehydrogenase [16]. Although photosynthetic and growth responses to Zn deficiency differ between species, meaning that some plants adapt better to such conditions [9], Zn deficiency rapidly inhibited plant growth and development, and photosynthesis of many plants, including rice [17] and spinach [18]. Zn deficiency restricts growth of lettuce plants, with necrotic spots with a dark margin appearing later along leaf edges [19]. Although plants in this study showed no signs of necrosis, decreased Zn availability may have restricted shoot growth of plants grown at 20 mM HCO3 − . B is an essential micronutrient for higher plants and is required at different concentrations for optimal growth in different species. B plays an important role in some plant functions including

Discussion
Whether RZ CO 2 enrichment regulates plant growth by altering plant nutrition was tested by growing two common horticultural species (lettuce and pepper) under commercially relevant production systems (hydroponics and aeroponics). Growth promotion of lettuce was correlated with enhanced nutrient concentrations following aeroponic CO 2 enrichment, but could not be attributed to enhanced nutrition following hydroponics-enrichment with low (<5 mM) bicarbonate concentrations. Pepper plants showed limited changes in nutrient status independent of method of RZ CO 2 enrichment. Even though enhanced crop nutrition is unlikely to account for the growth-promoting effects of RZ CO 2 enrichment, it is important to consider likely physiological effects in specific cases where nutrient concentrations did change.
In general, high concentrations of bicarbonate (≥ 5 mM) in the rhizosphere decreased foliar K, Mg, S, P and Fe concentrations, while high leaf tissue Ca concentrations occurred when applying HCO 3 − and in plants grown in calcareous soils at high pH [4,12,13]. In this study, applying 20 mM HCO 3 − did not change Ca concentration, but K, P, N, Zn, Cu and Mn concentrations decreased remarkably, while B and Mg significantly increased compared to control and 1 mM HCO 3 − . Moreover, Fe concentration at 20 mM HCO 3 − was higher than at 1 mM HCO 3 − (Figure 1). According to the adequate range of shoot nutrient concentrations in lettuce (Table A1), at 20 mM HCO 3 − , only Zn was below the limits, while B was above the limits. Although P concentrations exceeded the recommended range, this was independent of the bicarbonate concentrations applied. Thus, HCO 3 − enrichment of the nutrient solution should have caused minimal nutritional stress. Zn influences many biological processes, including carbohydrate metabolism and cell proliferation, and is an integral component of some enzyme structures, such as carbonic-anhydrase, alcohol dehydrogenase, and glutamate dehydrogenase [16]. Although photosynthetic and growth responses to Zn deficiency differ between species, meaning that some plants adapt better to such conditions [9], Zn deficiency rapidly inhibited plant growth and development, and photosynthesis of many plants, including rice [17] and spinach [18]. Zn deficiency restricts growth of lettuce plants, with necrotic spots with a dark margin appearing later along leaf edges [19]. Although plants in this study showed no signs of necrosis, decreased Zn availability may have restricted shoot growth of plants grown at 20 mM HCO 3 − .
B is an essential micronutrient for higher plants and is required at different concentrations for optimal growth in different species. B plays an important role in some plant functions including metabolic pathways, sugar translocation, pollen germination, hormone action, root development, normal growth and functioning of the apical meristem, water translocation from roots to the upper part of the plant body and membrane structure and body [20,21]. Although B concentrations at 20 mM HCO 3 − did not reach toxic levels (60 mg kg −1 vs. >100 mg kg −1 ), foliar B accumulation can reduce crop yield [22]. Mg levels were higher in some studies at high (>5 mM) HCO 3 − concentrations in white lupin [23], peach rootstocks [24] and tomato [25], while in others, the Mg concentration remained unchanged or decreased such as in tomato, tobacco, maize and olive trees. Although Mg deficiency in soils can inhibit plant growth [26], high soil Mg concentrations do not damage the crop growth but might hinder K uptake [27], which may explain the decrease in K under 20 mM HCO 3 − . Mg deficiencies in agriculture are not easily recognized and therefore little research has been done regarding Mg nutrient metabolism in the plant [28]. Despite Mg concentrations being in optimal range under high HCO 3 − concentrations, deciphering the physiological meaning of the increase will need further studies comparing interaction between Mg and HCO 3 − response curves.
Although the plants were not showing any visual chlorosis symptoms, chlorotic leaves actually have higher Fe concentrations of Fe than green leaves [30,31]. This phenomenon, called the "chlorosis paradox", is caused by Fe precipitation as insoluble compounds in Fe-deficient leaves [32,33]. Longer periods of elevated RZ HCO 3-treatment will probably cause leaf chlorosis. In addition, the inhibition of leaf expansion may diminish the dilution of Fe within the leaves, resulting in higher Fe concentrations. Decreased foliar nutrient (N, P, K, Cu and Zn) concentrations at 1 mM HCO 3 − were likely a consequence of similar nutrient uptake but greater growth, with shoot weight increasing by 10% (Tables 1 and 3 and lettuce plants have different nutrient requirements for their growth and development, and therefore nitrogen metabolism likely differs between these species when exposed to bicarbonate. Despite pepper plants not showing any significant variability among macronutrients and micro nutrients under elevated RZ CO 2 , lettuce grown aeroponically showed significantly lower shoot Mg and S concentrations under high RZ CO 2 , although root macronutrient concentrations did not differ between RZ CO 2 treatments. Further, K, Ca, Fe and Mn tended to decrease, and Zn increase compared to control plants ( Figure 2). Under elevated aerial CO 2 , nutrient concentrations can decline if increased photosynthesis and carbohydrate production dilute other nutrient elements within a larger biomass [35]. Nevertheless, no consistent and significant changes in photosynthesis rates were found during the experiments (data not shown). In addition, biomass increased significantly at an average of 22% among four experiments under elevated RZ CO 2 (Table 5). Thus, alternative explanations must account for these changes in biomass and nutrient concentrations.
An early effect of Mg deficiency in plants is the disturbed partitioning of assimilates between roots and shoots because the supply of sink organs with photosynthetic products is impaired, and sugars accumulate in source leaves [27]. The most commonly known function of Mg in plants is probably its role as the central atom of the chlorophyll molecule in the light-absorbing complex of chloroplasts and its contribution to photosynthetic fixation of carbon dioxide. However, the Mg bond to chlorophyll makes up only a small part of the total Mg fraction. Depending on plant Mg status, between~20% and 35% of the element is localised in the chloroplast, with the remainder present in more mobile forms [28]. Because of its high phloem mobility, Mg can easily be translocated to actively growing parts of the plant where it is needed for chlorophyll formation, enzyme activation for protein biosynthesis, and phloem export of photosynthates to ensure vegetative and generative growth [27]. Although visual symptoms occur in older leaves, no visual changes were observed in our plants as the concentration was in the recommended range.
Despite only 0.1% of plant dry matter existing as sulphur, it is an essential macronutrient for protein structure and is fundamental in many compounds with critical catalytic and electrochemical functions. Higher plants acquire S predominantly in the form of anionic sulphate from the soil. In plastids, sulphate is reduced to sulphite which then combines with O-acetyl-Ser to form Cysteine [36]. Then, sulphur is converted either into methionine or directly incorporated into proteins and glutathione [37]. Despite Mg and S being in optimal ranges in both treatments at the time of harvest (Mg: 3.3-4.4 g kg −1 ; S: 3.5-4.2 g kg −1 ), a significant interaction between Mg and S might be occurring at elevated RZ CO 2 .
Although there were differences in nutrient concentration in plant tissues as discussed above, there were fewer differences in nutrient content between treatments. Shoot P, N, Cu and Zn contents were significantly higher under elevated RZ CO 2 compared to control plants while no significant differences were detected in root nutrient contents, probably because there was no difference in root dry biomass (Table 7).
Since elevated atmospheric CO 2 fundamentally alters source-sink relationships by increasing net photosynthesis and decreasing shoot N and water use, elevated RZ CO 2 was also expected to vary these relationships. Shoot-to-root communication of leaf N status is necessary to optimize carbohydrate allocation in roots among growth, N uptake and inorganic N assimilation. Coordination of N transport from root to shoot and of carbohydrate transport from shoot to root maintains an optimal C/N ratio for plant growth and development [38,39]. Foliar N concentrations were lower and root concentrations higher under elevated RZ CO 2 . C/N ratio was lower in roots and was similar in shoots compared to control plants (Figure 3). At low soil NO 3 − concentrations, root C/N ratios are high, and roots have enough carbohydrates to assimilate most of the NO 3 − they absorb and thus they deliver little NO 3 − to the shoot. At high soil NO 3 − concentrations, root C/N ratios are low and a greater proportion of absorbed NO 3 − remains unassimilated in the root and therefore is transported to the shoot [40,41].
The N data agree with previous studies under elevated RZ CO 2 , with initial stimulation of NO 3 − uptake but no overall effects after prolonged (15 days) RZ CO 2 application [7,42]. In addition, numerous studies have reported positive interactions between N and P, which increases P absorption and higher yields [43]. Although C, N and P ratios under elevated aerial CO 2 are not well understood [44], elevated RZ CO 2 might change the signalling mechanisms to regulate their concentrations. Zn uptake tends to display a linear pattern with its concentration in the nutrient solution or in the soils [45,46], with roots loading the shoot tissues via the xylem [47]. Zn translocation to the root xylem occurs via symplast and apoplast [47,48], but high Zn levels have also been detected in the phloem, indicating translocation through both xylem and phloem tissues [49,50]. Some studies reported that elevated atmospheric CO 2 enhanced phytoextraction of heavy metals (including Cu, Pb, Cd and Zn) [51,52]. Perhaps RZ CO 2 had the same effects in lettuce plants, increasing the uptake of Cu and Zn without reaching toxic levels.

HCO 3
− addition to hydroponic media promoted growth at 1-5 mM but must be used moderately as higher concentrations suppress growth. Higher HCO 3 − concentrations decreased N, P, Cu, K and Zn concentrations, perhaps because high solution pH decreased nutrient availability. On the contrary, growth promotion at 1 mM HCO 3 − could not be attributed to higher nutrient concentrations. While biomass and nutrient concentration did not vary between aeroponic RZ CO 2 treatments in pepper, elevated RZ CO 2 increased biomass and decreased foliar Mg and S concentrations in lettuce grown aeroponically even though shoot Mg and S content did not differ between treatments. Elevated RZ CO 2 increased root N concentrations but shoot N concentrations did not change. Higher shoot N, P and Zn content indicates greater uptake of those elements. Interactions between form of DIC and solution pH seem important in determining the nutritional impacts of root-zone CO 2 enrichment. Acknowledgments: Special thanks to Andrew Spence for his contribution to the RZ HCO 3 − nutrient analysis and to Catherine Wearing for technical assistance. We are grateful to AHDB for a PhD fellowship for the CP 143 project.

Conflicts of Interest:
The authors declare no conflict of interest. Reference sources: [53,54].