Effect of Plant Growth-Promoting Bacteria Azospirillum brasilense on the Physiology of Radish (Raphanus sativus L.) under Waterlogging Stress

Stress due to waterlogging is considered an abiotic factor that negatively affects crop production, which, together with the excessive fertilization of crops, reduces cost-effectiveness and generates the need to create sustainable alternatives economically and environmentally. The effect of inoculation with Azospirillum brasilense on the physiology of the Raphanus sativus var. Crimson Giant subjected to waterlogging, was evaluated. Stomatal conductance, chlorophyll concentration and chlorophyll a fluorescence were analyzed to establish this effect, corroborating the beneficial effect of inoculation with A. brasilense in radish under waterlogging stress. The stomatal conductance of inoculated and waterlogged treatments presented the same values as the control plants, and photosystem II efficiency was favored in inoculated and waterlogged treatments (0.6 Fv/Fm) compared to non-inoculated and waterlogged treatments (0.3 Fv/Fm). The results suggested that this increased efficiency was due to the preservation of photosynthetic pigments in the tissues, allowing the preservation of stomatal conductance and a reduction in the amount of energy dissipated in the form of heat (fluorescence) due to inoculation with A. brasilense. Therefore, plant growth-promoting bacteria are responsible for activating and improving some physiological mechanisms of the plant.


Introduction
Radish (Raphanus sativus L.) is a plant of the Brassicaceae family, with important nutritional properties due to its high vitamin and mineral contents [1]. This plant has phytochemicals, such as glucosinolate, myrosinase and isothiocyanate, alkaloids and nitrogen compounds, coumarins (hydroxycoumarin aesculetin and scopoletin), organic acids, phenolic compounds and sulfur compounds, among others. Radish has biological activities, including the production of antimicrobial, antiviral, antitumor, antioxidant [2,3], antihyperlipidemic [4], and antihyperglycemic agents with the possibility of reducing diabetes and associated consequences [5]. Furthermore, radish is a fast-growing plant that can be harvested 25 to 35 days after sowing, a situation that makes it very attractive for small and medium vegetable producers [6,7]. The high precocity of this species is also an important aspect from an experimental point of view, as it allows researchers to obtain results of its physiological behavior in a short time.
Soil waterlogging has long been identified as major abiotic stress, and the constraints it imposes on roots have a marked effect on plant growth and development due to a decrease in oxygen in the soil by floods and an excess of rainfall [8][9][10]. Plants are known to wilt more rapidly in waterlogged soil due to decreased hydraulic conductance in the

Materials and Methods
The study was carried out at the facilities of the Biology Research Laboratory of Universidad El Bosque (INBIBO, Bogota, Colombia) with the radish seedlings var. Crimson Giant (R. sativus L.), with 14 days of germination and was inoculated at the time of transplantation with A. brasilense Cd DMS 1843 immobilized in alginate beads [19]. Radish var. Crimson Giant Crimson Giant has a harvest maturity of approximately 30 days and can reach a height of 10-13 cm. The root crop has a round upper part, is very firm and crisp in texture, with a red color and a diameter between 3.5 to 4 cm. Plants were transplanted into 14 oz. pots with Forza ® Mix (Fercon, Cali, Colombia) substrate consisting mainly of Canadian peat and fertilized every 3 days until day 15 with a Nutriponic ® (Walco S.A.S, Bogota, Colombia) solution designed to supply the essential elements. All treatments were maintained in a greenhouse at a temperature of 8 • C (night) to 18 • C (day), 75% relativity humidity, and 12/12 h photoperiod.

Experimental Design
A completely randomized experimental design with 4 treatments and 11 replications was used. Each experimental unit consisted of one plant for a total of 44 plants. Four treatments were established by randomization as follows (Table 1):

Physiological Measurements
Periodic data collection was carried out every 3 days, from days 9 to 21 after the start of the treatments. Stomatal conductance (gs) was measured on fully extended leaves from the middle of the plant, and readings were made between 9 to 11 a.m. using a leaf porometer (model SC-1; Decagon Devices, Pullman, WA, USA). In these same leaves, the chlorophyll index was measured using a Fieldscout CM1000 chlorofilometer (Spectrum, Collinsville, IL, USA), and fluorescence parameters were measured using a PAM fluorometer (Walz, Effeltrich, Germany) under controlled light exposure conditions and with a light pulse modulated to less than 50 µmol m −2 s −1 . Previously, leaves were kept in the dark for 20 min. Using Wincontrol-3 software (Walz), the ratio of the maximum efficiency of the PSII (Fv/Fm), electron transport rate (ETR), photochemical quenching (Pq), and non-photochemical quenching (NPq) were recorded.

Growth Measurements
At the end of the experiment (21 days after the start of waterlogging), the plants were harvested in 5 randomly selected experimental units of each treatment. The following parameters were measured: the equatorial diameter of the tuberous root (mm) was measured using a vernier caliper, the number of leaves was directly counted, the leaf area (cm 2 ) was measured using a CID Bio-Science meter (Cl-202; Camas, WA, USA), leaves were put on the equipment display and then the scan was performed through the equipment sensor, and the fresh and dry weight of the plants were measured using a 0.0001 g precision balance.
To calculate the dry weight, the samples were packed in paper bags and dried in an oven (Pinzuar, Madrid, Colombia) at 80 • C until constant weight.

Statistical Analysis
To evaluate the effect of the treatments, the results were subjected to an analysis of variance (ANOVA) when the assumptions of normality (Shapiro-Wilks test) and variance homogeneity (Levene's test) were met. The statistical differences of the means were analyzed according to Tukey's test (p < 0.05) with SPSS version 19 (IBM SPSS Statistics, New York, NY, USA).

Stomatal Conductance
Stomatal conductance showed no significant differences in radish on days 9 and 12 after waterlogging for any treatment. From day 12 to the end of the experiment (21 days after waterlogging), there were statistical differences between treatments. The inoculated plants (with and without waterlogging) showed the highest stomatal conductance from day 15. NIN+W plants had the lowest values and showed a continuous decrease in this parameter (71.7 ± 7.3 mmol H 2 O m −2 s −1 on day 21) (Figure 1).

Statistical Analysis
To evaluate the effect of the treatments, the results were subjected to an analysis of variance (ANOVA) when the assumptions of normality (Shapiro-Wilks test) and variance homogeneity (Levene's test) were met. The statistical differences of the means were analyzed according to Tukey's test (p < 0.05) with SPSS version 19 (IBM SPSS Statistics, New York, NY, USA).

Stomatal Conductance
Stomatal conductance showed no significant differences in radish on days 9 and 12 after waterlogging for any treatment. From day 12 to the end of the experiment (21 days after waterlogging), there were statistical differences between treatments. The inoculated plants (with and without waterlogging) showed the highest stomatal conductance from day 15. NIN+W plants had the lowest values and showed a continuous decrease in this parameter (71.7 ± 7.3 mmol H2O m −2 s −1 on day 21) (Figure 1).

Chlorophyll Index
There were statistical differences in most of the samples except on day 12, where all treatments showed the same chlorophyll index. NIN+W plants always showed the lowest chlorophyll values. On day 21, the chlorophyll index was 85.8 ± 26.8. In contrast, the plants of the other treatments showed significantly higher values, which also increased from day 15 ( Figure 2).

Maximum Efficiency of PSII
Significant differences were observed on all sampling days. Inoculated plants (with and without waterlogging) presented similar values to control plants throughout the study. In all cases, these plants presented higher values than the plants that were waterlogged but not inoculated, which had very low values from day 15 ( Figure 3).

Figure 2.
Chlorophyll index in R. sativus under waterlogged conditions and inoculated with A. brasilense. Averages followed by different letters at each sampling point show statistical differences according to Tukey's test (p < 0.05). Vertical bars in each average indicate standard error (n = 11).

Maximum Efficiency of PSII
Significant differences were observed on all sampling days. Inoculated plants (with and without waterlogging) presented similar values to control plants throughout the study. In all cases, these plants presented higher values than the plants that were waterlogged but not inoculated, which had very low values from day 15 ( Figure 3).

Photochemical Quenching (Pq)
Pq showed significant differences after 15 days of waterlogging. In waterlogged plants (even those inoculated), this parameter decreased significantly relative to non-waterlogged plants, with even lower values in NIN+W plants ( Figure 4).

Photochemical Quenching (Pq)
Pq showed significant differences after 15 days of waterlogging. In waterlogged plants (even those inoculated), this parameter decreased significantly relative to non-waterlogged plants, with even lower values in NIN+W plants ( Figure 4).  Differences were evident only on days 9 and 12 after waterlogging. In these samples, plants under waterlogging stress showed the highest values on days 9 and 12, especially those that were not inoculated. After day 12, NPq decreased and showed statistically similar values to the other treatments ( Figure 5).

Non-Photochemical Quenching (NPq)
Differences were evident only on days 9 and 12 after waterlogging. In these samples, plants under waterlogging stress showed the highest values on days 9 and 12, especially those that were not inoculated. After day 12, NPq decreased and showed statistically similar values to the other treatments ( Figure 5). test (p < 0.05). Vertical bars in each average indicate standard error (n = 11).
3.1.5. Non-Photochemical Quenching (NPq) Differences were evident only on days 9 and 12 after waterlogging. In these samples, plants under waterlogging stress showed the highest values on days 9 and 12, especially those that were not inoculated. After day 12, NPq decreased and showed statistically similar values to the other treatments ( Figure 5).

Electron Transport Rate (ETR)
It was evident that the highest response of this parameter was found in non-waterlogged plants with and without inoculation, but the effect was significant only after 12 days. On day 15, treatment with waterlogging and inoculation presented an ETR similar to non-waterlogging treatments ( Figure 6).

Electron Transport Rate (ETR)
It was evident that the highest response of this parameter was found in non-waterlogged plants with and without inoculation, but the effect was significant only after 12 days. On day 15, treatment with waterlogging and inoculation presented an ETR similar to nonwaterlogging treatments ( Figure 6).

Growth Parameters
Inoculated plants (with and without waterlogging) and plants without waterlogging showed the highest root diameter, number of leaves, and leaf area. The opposite response was evident in NIN+NW plants (Table 2). In fresh and dry weight measurements, nonwaterlogged plants (with and without inoculation) showed the highest weight, whereas NIN+W plants produced the lowest weight. It is important to mention that IN+W plants showed an increased weight ( Table 2).

Growth Parameters
Inoculated plants (with and without waterlogging) and plants without waterlogging showed the highest root diameter, number of leaves, and leaf area. The opposite response was evident in NIN+NW plants ( Table 2). In fresh and dry weight measurements, nonwaterlogged plants (with and without inoculation) showed the highest weight, whereas NIN+W plants produced the lowest weight. It is important to mention that IN+W plants showed an increased weight ( Table 2).

Stomatal Conductance
Waterlogged radish decreased stomatal conductance. It is known that the closure of stomata occurs in response to the stressful factor of waterlogging; thus, the plant manages to reduce the loss of water by transpiration, generating the reduction in stomatal conductance [25]. In Vigna radiata, a decrease of up to 62% occurred during the watering period [26].
Radish inoculated with A. brasilense presented greater stomatal conductance. This means that bacteria allowed flooded plants to generate mechanisms to cope with the stressful period, which gave them an advantage compared to those not inoculated (1). This advantage may also explain why inoculation mitigated the effect of waterlogging on growth (Table 2). Cohen et al. [27] experimentally demonstrated that Azospirillum possesses the necessary biochemical machinery to produce abscisic acid, thus providing the opportunity to overcome the stress period [28]. However, this would not explain why radish showed greater stomatal conductance when IN+W, and it becomes a subject for further investigation.
In a similar study conducted on rice, inoculation with Azospirillum did not produce differences in stomatal conductance of the plant by itself. Still, when inoculation was combined with irrigation of the plants, the physiological variables of rice increased up to 35% compared to those that were not inoculated [29].

Chlorophyll Index
The negative effect of waterlogging on this parameter was also reported in Physalis peruviana L., where the chlorophyll concentration decreased by about 30% [30]. The physiological imbalance generated by waterlogging reduced the synthesis of chlorophyll and accelerated its degradation. In melon, the decrease in photosynthetic rate, as a consequence of the waterlogging period, resulted in the degradation of chlorophylls [31]. In corn, a decrease of up to 70% in the percentage of chlorophyll was recorded after the start of waterlogging [32].
In contrast, Kumar et al. [33] argued that inoculation with A. brasilense significantly increased the photosynthetic pigments of the plants; after the physiological response to stress conditions, the additive effects of the inoculation allowed for an increase in phytohormones involved in the production of these pigments. Similar results were reported by Davaran-Hagh et al. [34], who found an 8% increase in the content of chlorophyll in corn after inoculation.
It was found that stomatal conductance and chlorophyll index in waterlogged plants decrease [35], confirming that the opening of stomata can be a limiting factor of photosynthesis when the plants are subjected to stressful factors, such as waterlogging [36], and explains the relationship between leaf conductance and the chlorophyll index. Accordingly, the idea that inoculation with A. brasilense allows radish to cope with the waterlogging condition by keeping the stomata open and thus avoiding photodamage conditions and retaining the chlorophyll concentration in their tissues can be corroborated, even with similar values to the control plants ( Figure 2). It is also possible to think that the result is due to an increase in aminocyclopropane-1-carboxylic acid deaminase produced by A. brasilense that would reduce the production of ethylene and the damage caused by waterlogging, as reported in previous studies for PGPB [12].

Maximum Efficiency of PSII
NIN+W radish showed very low values (even <0.6, Figure 3), which indicated the serious damage produced by this stress on the photo phase of the photosynthesis process [37]. As in normal conditions, the maximum efficiency of PSII ranged from 0.7 to 0.8 [38]. Nauman et al. [39] argued that the maximum efficiency of PSII was drastically decreased after several days of waterlogging when damage to the leaves of the plants was already evident. This type of stress would cause damage to PSII of the flooded radish, possibly due to the stomatal closure presented, which was evidenced by a lower stomatal conductance. This, in turn, was due to the decoupling with the photo phase that leads to the generation of reactive oxygen species (ROS) that damage the photosynthetic machinery and also because, in waterlogged plants, the levels of ethylene in the leaves are increased, which induces the degradation of chlorophylls and other biomolecules and even stimulates defoliation to avoid water loss.
In Distylium chinense, waterlogging produced a decrease in the maximum efficiency of PSII and was dependent on the waterlogging time [40]. Flórez-Velasco et al. [25] and Sánchez-Reinoso et al. [41] also reported a decrease in the maximum efficiency of PSII in Solanum quitoense plants subjected to waterlogging. In tomato, the values of the maximum efficiency of PSII were similar to those in the present study; as in radish, tomato that maintained normal levels of water in the soil had optimal levels for this variable compared to flooded plants [42].
Liu et al. [40] argued that a decrease in the maximum efficiency of PSII implied a photoinhibition effect of waterlogging on the tissues; however, when the values remained relatively constant, they indicated that the stressor did not affect the photochemical reactions of PSII. Accordingly, inoculation with A. brasilense represented a significant advantage to flooded radish, allowing it to cope with the stress-generating event and retain the maximum efficiency of PSII. In a study by Heidari and Golpayagani [43] on basil, the recorded maximum efficiency values for plants inoculated with PGPB organisms were always higher than those not inoculated.

Photochemical Quenching (Pq)
The photochemical dissipation value quantifies the amount of energy captured by the open PSII reaction centers and is used in the photochemical transport of electrons [34]. The average of this parameter in healthy plants is between 0.8 and 1 [44]. The values of waterlogged radish were below this range, even showing a reduction of about 50%, compared to the values of the plants not subjected to stress, indicating the serious damage caused by waterlogging on radish. In this regard, Liu et al. [40] recorded the gradual decrease in Pq as the waterlogging time in D. chinense increased until it reached 60%.
IN+W plants showed advantages of inoculation ( Figure 4). In the beginning, the plants coped with stress for a longer time. After day 12, the levels decreased significantly. However, it allowed plants to retain an adequate number of open PSII reaction centers compared to flooded and non-inoculated plants, thus ensuring that the efficiency of the photochemical pathway is increased.

Non-Photochemical Quenching (NPq)
Waterlogging generates stomatal closure in radish; as a consequence, the photo phase is affected. To avoid photoinhibition, the plant activates different mechanisms, and one of them is to dissipate excess energy in the form of heat. This process is activated to avoid excessive light absorption and photoinhibition in plants with some degree of stress [45]. NPq values indicate that the excess radiant energy is being dissipated in the form of heat in PSII; this may be related to the formation of zeaxanthin in the tissues [46].
Higher NPq values were evidenced in flooded plants up to 12 days after watering. However, there was no mitigation effect by A. brasilensis ( Figure 5), indicating that waterlogging initially affects this parameter, but it is not through this mechanism that A. brasilensis generates acclimatization to waterlogging in radish.

Electron Transport Rate (ETR)
This parameter was affected in waterlogged radish 12 days after stress onset (Figure 6), indicating the negative effect of this stress on the photo phase. In a study carried out on beans, the ETR decreased by 29% after the watering period. This reduction was associated with the damage to PSII, which resulted from ROS accumulation [47]. In D. chinense, the ETR was reduced by 60% after waterlogging. Additionally, the authors reported that the flow of electrons and the efficiency to transform energy were affected by waterlogging [40]. However, inoculation with A. brasilense made it possible to cope with the reduction in the ETR in waterlogged radish for up to 15 days, allowing a greater flow of electrons to the PSII to probably guarantee higher production rates of NADPH. This positive effect is related to the fact that A. brasilensis avoids damage to PSII (as mentioned above) and, in general, to the photosynthetic apparatus, possibly explained by the fact that it mitigates the effect of waterlogging and prevents stomatal closure. In general, the results demonstrate the improvement in physiological processes after inoculation when the plants are subjected to stressful conditions [48].

Tuberous Root Diameter
Waterlogging negatively affected the diameter of the tuberous root (Table 2), indicating the harmful effect of this condition on the growth of this organ [10]. Various studies demonstrated the increase in the concentration of phytohormones after inoculation with Azospirillum, not only improving the physiology of the plant when conditions are favorable but also acting when the environment is unfavorable [2], as what possibly occurred with radish under waterlogged conditions. Inoculation with the genus Azospirillum and in general, with PGPB, implies an improvement in the yield of the plant due to the increased production of photosynthates [49], which are then distributed to the different parts of the plant to favor an increase in size, and in this case, of the tuberous root the organ that can be harvested in radish. The greater root diameter can also be attributed to the ability of the bacteria to generate phytohormones, such as indole-3-acetic acid and gibberellins, which are involved in root growth and development [50]. In this regard, Méndez-Gómez et al. [51] claimed that Azospirillum reduces the effects of stressors, such as waterlogging, thanks to its ability to synthesize phytohormones at extreme pH.

Number of Leaves
In similar studies of seedling waterlogging, the number of leaves is reduced in the same way as it occurs in radish seedlings. In Populus angustifolia seedlings, waterlogging generated a gradual decrease in the number of leaves according to the total time of exposure to the stressor factor [52]. In tomato, waterlogging produced plants with a lower number of leaves, and the number of dead leaves during the study was twice as high in flooded plants [53]. In previous studies, inoculation with Azospirillum generated similar results in the number of leaves of plants. In seedlings of Theobroma cacao, inoculation generated an average of five more leaves per plant [54]. In flooded plants, the number of leaves was directly affected (Table 2), but A. brasilense allowed inoculated radish to cope with the waterlogging condition. Flooded soils are known to decrease the redox potential and the amount of oxygen present, affecting the availability of nutrients for plants [55]. As a consequence, plants need to activate a series of physiological and morphological mechanisms that allow them to cope with the main stressor event, the prolonged replacement of the gas phase present in the soil with a liquid phase, which generates the loss of nutrients and minerals due to root washing [56]. However, the genus Azospirillum can fix atmospheric nitrogen under microaerobic oxygen conditions. Méndez-Gómez et al. [51] argued that the contribution of nitrogen by Azospirillum was estimated at 10 kg N/ha/year, supporting the hypothesis that inoculation confers an advantage when it comes to increasing the resistance of plants, such as radish to stressful conditions.

Leaf Area
In waterlogged radish, a significant reduction was recorded in this parameter (Table 2), corroborating what was exposed by Moreno and Fischer [57], who argued that the leaf area is significantly reduced after waterlogging, as the roots reduce the transport of water and negatively induce cell turgor, affecting leaf development. Furthermore, waterlogging generates an energy crisis in the plant due to the limitation of a number of metabolic routes; therefore, the reduction in leaf area is a consequence of the energy savings that the plant makes to maintain the water balance.
However, inoculation with A. brasilense gave radish the ability to withstand the stressful condition and maintain leaf area due to the increased production of phytohormones [51]. It was found that A. brasilense increases the total area of the leaves, increasing the area for gas exchange; for example, in cocoa, A. brasilense promoted a significant increase in leaf area [54].

Fresh and Dry Weight
These growth parameters were negatively affected by waterlogging, but this negative effect was mitigated by the presence of A. brasilense ( Table 2). The growth of waterlogged plants is directly influenced by the lack of oxygen in the rhizosphere; as a consequence of anoxia, there is a reduction in the biomass of plants due to a decrease in gas exchange in the leaves, which in turn decreases the rate of photosynthesis [47]. Reduced dry weight was also found in P. peruviana plants [58]. It is also known that the significant reduction in the weight of the plants can be attributed to the low availability of nitrogen in the soil once it is flooded, in addition to the physiological dysfunctionalities derived from stomatal closure and photosynthetic limitations [59].
However, inoculation with A. brasilense in radish showed a positive effect in waterlogged and non-waterlogged plants, indicating the important role of this bacterium in the growth of radish. Veresoglou and Menexes [60] argued that inoculation with Azospirillum increases the production of dry biomass in plants by up to 33%, as indole acetic acid synthesized by the bacteria is used by the plant. In similar studies, inoculation with Azospirillum increases the yield of the plants; in cocoa, the dry weight of inoculated plants increased by 44% compared to non-inoculated ones [54]. In corn, the fresh weight of inoculated plants was 30% greater than that of non-inoculated plants. This increase was also attributed to the ability of A. brasilense to fix atmospheric nitrogen [34].
In summary, this study demonstrates that physiological and growth parameters of radish under waterlogging stress are improved when inoculated with A. brasilense. These results contribute to the studies of bacterial inoculants for agricultural purposes in which stress biotic conditions are present, such as waterlogging.