Morphological, Physiological, and Biochemical Responses of Zinnia to Drought Stress

: Bedding plants in the nursery phase are often subject to drought stress because of the small volume of the containers and the hydraulic conductivity of organic substrates used. To analyse the morphological, physiological, and enzymatic responses of zinnia ( Zinnia elegans L.) plants at different irrigation levels, four treatments were performed: irrigated at 100% (100% ﬁeld capacity, FC); light deﬁcit irrigation (75% FC), medium deﬁcit irrigation (50% FC), and severe deﬁcit irrigation (25% FC). The growth of zinnia was signiﬁcantly inﬂuenced by drought stress treatments. Different morphological parameters (dry biomass, leaf number, root to shoot ratio (R/S)) were modiﬁed only in the more severe drought stress treatment (25% FC). The stomata density increased in 50% FC and 25% FC, while the stomata size was reduced in 25% FC. The net photosynthesis, stomatal conductance, and transpiration were reduced in 50% FC and 25% FC. The relative water content (RWC) was reduced in 25% FC. Severe drought stress (25% FC) increased proline content up to seven-fold. Catalase (CAT), peroxidase (GPX), and superoxide dismutase (SOD) activity signiﬁcantly increased in 50% FC and 25% FC. Principal component analysis (PCA) showed that the morphological and physiological parameters were mostly associated with the 100% FC and 75% FC treatments of the biplot, whereas the stomata density, R/S ratio, and antioxidant enzymes (GPX, CAT) were associated with 50% FC, and proline and DPPH were associated with 25% FC, respectively.


Introduction
Bedding plants play a relevant role in public green areas and private gardens. These plants can suffer from drought stress because they are not always properly watered, especially when grown in pots or show small root systems [1]. Bedding plants are in fact at greater risk of undergoing drought stress during the nursery phase because they are cultivated in small pots that can limit root growth, making the plants subject to greater levels of drought stress. Furthermore, the hydraulic conductivity of the substrates, which are often organic, used in the production of bedding plants decreases rapidly with small changes in the substrate water content [2], making the extraction of water very difficult for plants when the water content in the substrate is low. However, limited research exists regarding the physiological mechanisms that allow bedding plants to tolerate drought stress.
Drought stress results in damage to the plant's physiological and biochemical processes and represents one of the most relevant environmental factors that impair plant growth and performance [3]. Plants may exhibit numerous drought stress response mechanisms at the morphological and physiological levels [4,5]. At the level of the whole plant, some species increase their root biomass to enhance water uptake [6] and hence maintain the water state of the plant and ensure photosynthesis in drought conditions. In the nursery stage of bedding plant cultivation, this acclimatization response may not be possible because both root growth and available water are limited by the small pots. will enhance the frequency and severity of drought [31]. Therefore, in a changing climate, studying the main physiological limits to productivity in drought conditions will be crucial for enhancing yield stability.
Because the increased drought frequency strongly negatively affects plant growth and development [32], analyzing the effects of water deficit on plants is relevant to hypothesise the influence of future climate changes on the growth of a particular plant species [33]. Among bedding plants, studying the response to drought stress in various species and/or cultivars is strategic to individuate genotypes able to improve landscape performance and expand the use of these plants in drought areas [34].
The aim of this research was to determine the morphological, physiological, and enzymatic responses of zinnia plants at different irrigation levels and to evaluate the response to different intensities of drought and hence the possibility to reduce the water quantity used in the nursery phase.
Plants were grouped into three repetitions of nine plants per treatment and irrigated every day for 30 days. Four treatments were performed: irrigated at 100% (100% field capacity, FC), light deficit irrigation (75% FC), irrigated at 75% from the 100% FC treatment, medium deficit irrigation (50% FC), irrigated at 50% from the 100% FC treatment and severe deficit irrigation (25% FC), and irrigated at 25% from the 100% FC treatment. Water loss was determined following the methodology of Toscano et al. [35] through the gravimetric method; during the experimental period, the differences in weights (weight after irrigation, weight when drainage stopped, and weight before reirrigating) were calculated.
The treatments started when the plants showed four leaves and ended after 30 days when approximately 50% of plants showed the beginning of inflorescence emergence.
To determine the maximum water-holding capacity of the substrate, pots were mixed and submerged in water to 50% of their height, and the substrate was left to imbibe overnight. To avoid water evaporation, aluminium foil was placed on the upper surfaces of the containers. The next day, the containers were removed from the water bath and left to drain until reaching a constant weight. The weight of each container was then determined and considered as the weight at volumetric water content. Then, the substrate was dried in a thermo-ventilated oven at 105 • C until reaching a constant weight to determine the dry weight and calculate the volumetric water content. The difference between the fresh and dry weight was calculated, and the volumetric water content was determined (75%) and used as the substrate's container capacity [36].
The mean air temperature, relative humidity, and global radiation were determined on a data logger CR1000 (Campbell Scientific Ltd., Loughborough, UK) during the experimental periods. The minimum, maximum, and mean temperatures were 16.9, 54.6, and 26.2 • C, respectively. The mean relative humidity (RH) was 58.8%.

Biomass and Leaf Area
After the experimental period ended, for six plants per treatment, the roots were separated from the substrate with tap water, and the aerial parts were divided into stems and leaves. Drying the weighed samples in a thermo-ventilated oven at 70 • C until reaching constant weight allowed the determination of dry biomass. The total leaf area was measured using a leaf area meter (Delta-T Devices Ltd., Cambridge, UK).

Stomata Characteristics
Unfolded and mature leaves (4 leaves per treatment and for each repetition) were detached from the plants and used for stomata characteristics. A microscope (Nikon Eclipse E200, Japan) was utilised for determining the number and size of the stomata. On each slide, along a diagonal transect of the peel, four stomata for three leaves for each repetition were measured for pore lengths at 40×. Stomatal size (S) was defined by guard cell length and width.

Leaf Gas Exchange, Chlorophyll a Fluorescence, and RWC
At the end of the trial, the gas exchange in six plants per treatment (two plants for each repetition and three leaves per plant) was measured with a CO 2 /H 2 O infrared gas analyser (LCi, ADC Bioscientific Ltd., Hoddesdon, UK). The reliefs were carried out in the morning (from 09:00 to 13:00). For each drought stress treatment, the net photosynthetic rate (A N ), stomatal conductance (gs), transpiration rate (E), and water use efficiency (WUE) were determined.
The chlorophyll a fluorescence (Fv/Fm) was recorded in the same leaves using a modulated chlorophyll fluorimeter OS1-FL (Opti-Sciences Corporation, Tyngsboro, MA, USA). The leaf was dark-adapted using cuvette clips for 15 min (Opti-Sciences Corporation, Tyngsboro, MA, USA). The chlorophyll a fluorescence was reported as the Fv/Fm ratio, where Fm = the maximum fluorescence and Fv = the variable fluorescence.
The relative water content (RWC) was evaluated on fully opened leaves. For each replicate, 30 discs 10 mm in diameter were taken, and the fresh weights (FW) were determined. Then, the samples were immersed in distilled water for 24 h and re-weighed to measure the turgid weight (TW). Subsequently, the samples were dried at 75 • C for 24 h to measure the dry weights (DW). The RWC was expressed according to the formula: RWC% = (FW − DW/TW − DW) * 100

Determination of Chl and Carotenoid Content
After ending the experiment, chlorophyll a (Chl a), chlorophyll b (Chl b), total chlorophyll, and carotenoids were determined. For the extraction, 100 mg of fresh material was extracted with 5 mL of 99% methanol and incubated in the dark for 24

Estimation of Proline Content
Proline content was determined according to Ahmad et al. [38] using L-proline as the standard. Leaf samples (1 g) were homogenised in 5 mL of 3% aqueous sulfosalicylic acid and centrifuged at 14,000× g for 15 min (Neya 10R, REMI, Mumbai, India). The homogenate (2 mL) was added at the same quantity of acetic acid and ninhydrin, mixed and incubated for 1 h at 100 • C. Then, the reaction was stopped in an ice bath, and the supernatant was extracted with 4 mL of toluene. The absorbance of the extract was read at 525 nm (7315 Spectrophotometer, Jenway, Staffordshire, UK).

Estimation of MDA Content
Malondialdehyde (MDA) content was determined according to Li et al. [39]. Leaf samples (0.5 g) were extracted in 1.5 mL of 5% trichloroacetic acid (w/v). The homogenate was centrifuged at 5000× g for 10 min, and then the extract was diluted to 10 mL. A quantity of 2 mL of the diluted extract was homogenised with the same quantity of 0.67% thiobarbituric acid. The mixture was incubated at 95 • C for 30 min and then centrifuged at 5000× g for 10 min. The MDA content was calculated using the following formula: C (µmol/L) = 6.45 × (A 532 − A 600 ) − 0.56 × A 450 .
The catalase activity (CAT; EC 1.11.1.6) was evaluated according to Aguilera et al. [41] with modifications; 20 µL of the extract was homogenised to 830 µL potassium phosphate buffer (50 mM, pH 7). The reaction started with the addition of 150 µL of H 2 O 2 , and the decrease was monitored at 240 nm for 2 min. The unit of CAT was expressed as units mg −1 protein.
The glutathione peroxidase activity (GPX) was determined according to Ruley et al. [42]. The same amount of extract and 17 mM H 2 O 2 was homogenised with 2% guaiacol to obtain a final volume of 1 mL. The increase in absorbance was monitored at 510 nm for 3 min. The activity of GPX was defined as units mg −1 protein.
The superoxide dismutase activity (SOD; EC 1.15.1.1) was evaluated following Giannopolitis and Ries [43]. The SOD activity was read at 560 nm; the unit of SOD was defined as the amount of enzyme required to cause 50% inhibition of the reduction of NBT. The unit of SOD was expressed as units mg −1 protein.

2,2-Diphenyl-1-picrylhydrazyl (DPPH) Radical Scavenging Activity
The scavenging activity against the DPPH radical was evaluated using DPPH. A quantity of 1 g of fresh weight was homogenised with 1.5 mL of 80% methanol, sonicated for 30 min, and centrifuged for 10 min at 5 • C and 5000× g. Subsequently, 0.01 mL of the supernatant was homogenised with 1.4 mL of DPPH solution (150 µM) and incubated for 30 min in the dark. Later, the samples were read at 517 nm. The DPPH activity was expressed as Trolox equivalent antioxidant activity (mg TE g −1 ).

Statistical Analysis
The trial was conducted as a randomised complete design with three replicates. The statistical analyses were conducted with CoStat version 6.311 (CoHort Sofware, Monterey, CA, USA); one-way ANOVA was adopted. The differences between the means were determined using Tukey's test (p < 0.05). The data presented in the figures are the means ± standard error (SE) (Graphpad 7.0). The principal component loading plot and scores of PCA were performed using Minitab 16, LLC. Figure 1 shows the trend of evapotranspiration (L day −1 ) in the 100% FC treatment during the experimental period. The amounts of water manually added to each pot were 1.88, 1.41, 0.92 and 0.47 L, respectively, for 100% FC, 75% FC, 50% FC and 25% FC. The electrical conductivity of the water was 0.005 dS m −1 .

Biomass and Leaf Area
The growth of zinnia was significantly modified by drought stress treatments. Plant height was reduced because of the different irrigation regimes (Table 1). By increasing the level of drought stress from control to 25% FC, the height was reduced by ~17% and ~38%, respectively, for 50% FC and 25% FC (p < 0.001). No significant differences were observed in 75% FC compared with unstressed plants (Table 1). A similar trend was observed for the stem diameter, with a reduction in more stressed treatments by ~23 and ~37%, respectively, for 50% FC and 25% FC (p < 0.001). Leaf number was significantly reduced under 25% FC (by 39%) compared with the other treatments (p < 0.001) ( Table 1); the total leaf area showed a decrease from the moderate deficit irrigation with a reduction of 21% in 50% FC and by 50% in 25% FC. (p < 0.001). No significant differences were observed in 75% FC (Table 1).

Biomass and Leaf Area
The growth of zinnia was significantly modified by drought stress treatments. Plant height was reduced because of the different irrigation regimes (Table 1). By increasing the level of drought stress from control to 25% FC, the height was reduced by~17% and~38%, respectively, for 50% FC and 25% FC (p < 0.001). No significant differences were observed in 75% FC compared with unstressed plants (Table 1). A similar trend was observed for the stem diameter, with a reduction in more stressed treatments by~23 and~37%, respectively, for 50% FC and 25% FC (p < 0.001). Leaf number was significantly reduced under 25% FC (by 39%) compared with the other treatments (p < 0.001) ( Table 1); the total leaf area showed a decrease from the moderate deficit irrigation with a reduction of 21% in 50% FC and by 50% in 25% FC. (p < 0.001). No significant differences were observed in 75% FC (Table 1). Table 1. Effects of irrigation treatments on plant height (cm), stem diameter (mm), leaf number (n.), total leaf area (cm 2 ), leaf, stem, root fresh weight (g plant −1 ), total dry biomass (g plant −1 ), and root/shoot ratio (R/S) of potted zinnia plants at the end of the experimental period. Plants were irrigated every day. Four treatments were performed: irrigated at 100% (100% field capacity, FC); light deficit irrigation (75% FC), irrigated at 75% from the 100% FC treatment; medium deficit irrigation (50% FC), irrigated at 50% from the 100% FC treatment and severe deficit irrigation (25% FC), and irrigated at 25% from the 100% FC treatment.
The values are the means ± standard error (SE). The statistical analysis was one-way ANOVA; * significant at p < 0.05; ** significant at p < 0.01; *** significant at p < 0.001. The values in the same column followed by the same letter are not significantly different at p < 0.05 (Tukey's test).
A similar trend was detected for the leaf and stem fresh weight, with a reduction of 28% (leaf) and 15% (stem) and 56% (leaf) and 44% (stem), respectively, for 50% FC and 25% FC (p < 0.001 and p < 0.01). The light deficit treatments (75% FC) showed the same trend as unstressed plants. Root fresh biomass was only modified in 25% FC, with a reduction of 54% compared with the other treatments (p < 0.001) ( Table 1). In addition, the the total dry biomass decreased only in 25% FC (by~48%) (p < 0.01) compared with the other treatments ( Table 1). The root-to-shoot ratio increased in plants grown under moderate deficit irrigation treatments (50% FC) (p < 0.05) ( Table 1).
Stomata density showed significant differences among the irrigation treatments, with an increase in the more severe drought stress (50% FC and 25% FC) ( Table 2). The drought stress influenced the stomata size only in 25% FC. The latter showed a significant reduction of 31% compared with 100% FC and 75% FC. The 50% FC treatment did not show a significant change among treatments (Table 2 and Figure 2). Table 2. Effects of irrigation treatments on stomata density (n mm −2 ) and size (µm) of potted zinnia plants at the end of the experimental period. Plants were irrigated every day. Four treatments were performed: irrigated at 100% (100% FC); light deficit irrigation (75% FC), irrigated at 75% from the 100% FC treatment; medium deficit irrigation (50% FC), irrigated at 50% from the 100% FC treatment and severe deficit irrigation (25% FC), and irrigated at 25% from the 100% FC treatment.

Drought Stress
The values are the means ± standard error (S.E). The statistical analysis was one-way ANOVA; * significant at p < 0.05; ** significant at p < 0.01. The values in the same column followed by the same letter are not significantly different at p < 0.05 (Tukey's test).
25% FC (p < 0.001 and p < 0.01). The light deficit treatments (75% FC) showed the same trend as unstressed plants. Root fresh biomass was only modified in 25% FC, with a reduction of 54% compared with the other treatments (p < 0.001) ( Table 1). In addition, the the total dry biomass decreased only in 25% FC (by ~48%) (p < 0.01) compared with the other treatments ( Table 1). The root-to-shoot ratio increased in plants grown under moderate deficit irrigation treatments (50% FC) (p < 0.05) ( Table 1). Stomata density showed significant differences among the irrigation treatments, with an increase in the more severe drought stress (50% FC and 25% FC). (Table 2). The drought stress influenced the stomata size only in 25% FC. The latter showed a significant reduction of 31% compared with 100% FC and 75% FC. The 50% FC treatment did not show a significant change among treatments (Table 2 and Figure 2). Table 2. Effects of irrigation treatments on stomata density (n mm −2 ) and size (μm) of potted zinnia plants at the end of the experimental period. Plants were irrigated every day. Four treatments were performed: irrigated at 100% (100% FC); light deficit irrigation (75% FC), irrigated at 75% from the 100% FC treatment; medium deficit irrigation (50% FC), irrigated at 50% from the 100% FC treatment and severe deficit irrigation (25% FC), and irrigated at 25% from the 100% FC treatment.  Significant effects of drought stress treatments for gas exchange were observed in zinnia plants (Figure 3). The net photosynthesis, stomatal conductance, and transpiration rate showed similar behaviours. Significant differences for AN were observed in the 50% Significant effects of drought stress treatments for gas exchange were observed in zinnia plants (Figure 3). The net photosynthesis, stomatal conductance, and transpiration rate showed similar behaviours. Significant differences for A N were observed in the 50% FC and 25% FC treatments compared with the control plants and 75% FC. In particular, the plants irrigated at 50% FC and 25% FC showed a reduction of~30% and 66% compared with control plants (Figure 3a). A similar trend was observed for the stomatal conductance, with a reduction of~51% and 80%, respectively, for 50% FC and 25% FC (Figure 3b) compared with the control and 75% FC. The transpiration rate showed a significant reduction only in the severe drought stress treatment, with a reduction of 61% compared with 100% FC and the other stress treatments. (Figure 3c). Water use efficiency (WUE) showed significant differences among the more stressed treatment (25% FC) and the other treatments with an increase of~45% (Figure 3d).

Drought Stress
with control plants (Figure 3a). A similar trend was observed for the stomatal conductance, with a reduction of ~51% and 80%, respectively, for 50% FC and 25% FC (Figure 3b) compared with the control and 75% FC. The transpiration rate showed a significant reduction only in the severe drought stress treatment, with a reduction of 61% compared with 100% FC and the other stress treatments. (Figure 3c). Water use efficiency (WUE) showed significant differences among the more stressed treatment (25% FC) and the other treatments with an increase of ~45% (Figure 3d). Significant differences for Chl a and b, total Chl, and carotenoids were observed in the 25% FC treatment. In particular, the plants irrigated at 25% FC showed reductions of ~36%, 39%, 37%, and 40%, respectively, for Chl a and b, total Chl, and carotenoids compared with control plants (Figure 4a-d). Among the other treatments, no significant differences were observed (Figure 4a-d). Significant differences for Chl a and b, total Chl, and carotenoids were observed in the 25% FC treatment. In particular, the plants irrigated at 25% FC showed reductions of~36%, 39%, 37%, and 40%, respectively, for Chl a and b, total Chl, and carotenoids compared with control plants (Figure 4a-d). Among the other treatments, no significant differences were observed (Figure 4a-d).
The RWC under light and moderate drought stress (75% FC and 50% FC) did not show significant differences compared with control plants and the other treatments. In the condition of severe drought stress (25% FC), a significant reduction of~32% was observed (Table 3).
Horticulturae 2021, 7, x FOR PEER REVIEW 9 of 18 FC and 25% FC treatments compared with the control plants and 75% FC. In particular, the plants irrigated at 50% FC and 25% FC showed a reduction of ~30% and 66% compared with control plants (Figure 3a). A similar trend was observed for the stomatal conductance, with a reduction of ~51% and 80%, respectively, for 50% FC and 25% FC ( Figure 3b) compared with the control and 75% FC. The transpiration rate showed a significant reduction only in the severe drought stress treatment, with a reduction of 61% compared with 100% FC and the other stress treatments. (Figure 3c). Water use efficiency (WUE) showed significant differences among the more stressed treatment (25% FC) and the other treatments with an increase of ~45% (Figure 3d). Significant differences for Chl a and b, total Chl, and carotenoids were observed in the 25% FC treatment. In particular, the plants irrigated at 25% FC showed reductions of ~36%, 39%, 37%, and 40%, respectively, for Chl a and b, total Chl, and carotenoids compared with control plants (Figure 4a-d). Among the other treatments, no significant differences were observed (Figure 4a-d).   The RWC under light and moderate drought stress (75% FC and 50% FC) did not show significant differences compared with control plants and the other treatments. In the condition of severe drought stress (25% FC), a significant reduction of ~32% was observed (Table 3). Table 3. Effects of irrigation treatments on relative water content (RWC, %) and chlorophyll a fluorescence (Fv/Fm) of potted zinnia plants at the end of the experimental period. Plants were irrigated every day. Four treatments were performed: irrigated at 100% (100% FC); light deficit irrigation (75% FC), irrigated at 75% from the 100% FC treatment; medium deficit irrigation (50% FC), irrigated at 50% from the 100% FC treatment and severe deficit irrigation (25% FC), irrigated at 25% from the 100% FC treatment. The values are the means ± standard error (SE). The statistical analysis was one-way ANOVA; * significant at p < 0.05; ** significant at p < 0.01. The values in the same column followed by the same letter are not significantly different at p < 0.05 (Tukey's test).

Drought Stress RWC (%) Fv/Fm
In zinnia plants, a reduction in the maximum quantum efficiency (Fv/Fm) was observed only in severe drought stress conditions (10% FC), with a value of 0.68 (Table 3).
The amount of leaf proline content increased in 25% FC. Severe drought stress (25% FC) increased proline content up to seven-fold compared with the control plants ( Figure  5a). Among the other treatments, no significant differences were observed (Figure 5a).
The MDA content did not show a significant change among treatments (Figure 5b).  In zinnia plants, a reduction in the maximum quantum efficiency (Fv/Fm) was observed only in severe drought stress conditions (10% FC), with a value of 0.68 ( Table 3).
The amount of leaf proline content increased in 25% FC. Severe drought stress (25% FC) increased proline content up to seven-fold compared with the control plants (Figure 5a). Among the other treatments, no significant differences were observed (Figure 5a).
The MDA content did not show a significant change among treatments (Figure 5b). The results showed that drought stress significantly affected enzyme activity ( Table 4). The CAT activity significantly increased in moderate and severe drought stress (50% FC and 25% FC). The increment was 41% for CAT and 33% for GPX. In addition, SOD activity showed a similar trend: an increase of 42% in 50% FC and 25% FC was observed ( Table 4).
The results showed that drought stress significantly affected DPPH activity (p ≤ 0.01). DPPH activity significantly increased with severe drought stress treatment. The increment from control to 25% FC was~23%, while among the other treatments, no significant differences were observed (Table 4).
In order to visualise congruence among 100% FC, 75% FC, 50% FC, and 25% FC plants based on all of the morphological, physiological, and biochemical variables, the whole dataset was subjected to a principal component analysis (PCA; Figure 6). The PCA showed that the morphological and physiological parameters were mostly associated with the 100% FC and 75% FC treatments of the biplot, whereas the stomata density, R/S ratio, and antioxidant enzymes (GPX, CAT) were associated with the 50% FC treatment, and proline and DPPH were associated with the 25% FC treatment (Figure 6).
In zinnia plants, a reduction in the maximum quantum efficiency (Fv/Fm) was observed only in severe drought stress conditions (10% FC), with a value of 0.68 ( Table 3).
The amount of leaf proline content increased in 25% FC. Severe drought stress (25% FC) increased proline content up to seven-fold compared with the control plants ( Figure  5a). Among the other treatments, no significant differences were observed (Figure 5a).
The MDA content did not show a significant change among treatments (Figure 5b). Mean values ± standard error (S.E) (n = 6). Different letters indicate significant differences among the treatments as determined by Tukey's test. ns: no significant. Table 4. Effects of irrigation treatments on catalase (CAT), peroxidase (GPX), and superoxide dismutase (SOD) activity of potted zinnia plants at the end of the experimental period. Plants were irrigated every day. Four treatments were performed: irrigated at 100% (100% FC), light deficit irrigation (75% FC), irrigated at 75% from the 100% FC treatment; medium deficit irrigation (50% FC), irrigated at 50% from the 100% FC treatment and severe deficit irrigation (25% FC), irrigated at 25% from the 100% FC treatment.

Drought Stress
Enzyme Activity

Discussion
Drought tolerance is the capacity of plants to continue to be functional at lower tissue water potentials. Drought stress determines considerable changes in the physiological and biochemical activity of plants, including photosynthesis, respiration, transpiration, hormone metabolism, and enzyme activity [45]. The mechanisms of drought tolerance involve the maintenance of turgor (by the accumulation of solutes) and/or desiccation

Discussion
Drought tolerance is the capacity of plants to continue to be functional at lower tissue water potentials. Drought stress determines considerable changes in the physiological and biochemical activity of plants, including photosynthesis, respiration, transpiration, hormone metabolism, and enzyme activity [45]. The mechanisms of drought tolerance involve the maintenance of turgor (by the accumulation of solutes) and/or desiccation tolerance (by protoplasmic resistance) [46]. In urban areas, it is important to select species for planting that are able to tolerate water shortages without losing their aesthetic appearance [47]. Water availability is one of the principal factors that limit bedding and landscaping plant cultivation, particularly for seasonal and annual garden flowers with shallow roots [1]. In the nursery phase, the generalised use of containers, often of small volume, determines root restriction effects [35].
Because of the showiness of flowering herbaceous species, it is difficult to convince people not to use them in urban landscaping. These plants are particularly sensitive to water deficiency. Annual plants are vulnerable to increasing temperatures and decreasing rainfall, as they must complete their life cycle in a single season, and the persistence of the population is highly dependent on the reproductive capacity of a season [48]. As for woody plants, herbaceous plants also adopt different physiological and biochemical mechanisms to overcome drought stress. One of the first responses to drought stress is a reduction in plant growth. In our study, the drought stress treatments significantly reduced the height, shoot and root dry biomass, leaf number, and leaf area in the more stressed treatments. A moderate reduction in the amount of water applied to container-grown zinnias did not reduce the morpho-biometric characteristics. Previous studies have shown that dry biomass reduced significantly with increasing drought stress treatments in Trachyspermum ammi L. [49] and Ocimum basilicum L. [50]. At the end of the drought treatment, the plant biomass reduction could be linked to a reduction in cell elongation and expansion; lower water absorption led to a reduction in the production of metabolites to maintain normal cellular activity [51].
A lower water amount can have positive effects on growth, so light drought stress can be a useful water-saving strategy. In our study, the plants grown at 50% FC showed no reduction in growth parameters. In a study on parsley, Borges et al. [52] demonstrated that water reduction significantly modified the development of the aerial parts. According to Zulfiqar et al. [53], a reduction in biomass is a good strategy in Tagetes erecta to improve water uptake under drought stress in the long term. Drought stress influences the aerial parts of the plants more than the roots; furthermore, the growth of the aerial parts of the plant decreases earlier than the roots, causing an increased root-to-shoot ratio. This ratio is not always modified by drought stress. Sánchez-Blanco et al. [54] found that the R/S in Pistacia lentiscus and Phillyrea angustifolia was not influenced under water stress treatments. In our study, except for 50% FC, the same trend was observed. Eziz et al. [55], in a meta-analysis to explain the patterns of plant biomass allocation related to drought stress, found that roots of woody plants were more resistant than those of herbaceous plants to drought. Furthermore, drought stress in herbaceous plants had a more negative influence on the leaf mass fraction of woody plants. In our study, leaf fresh biomass reduced as water stress increased. A similar response was observed for the total leaf area and leaf number. The reduction of the leaf area is due to a reduction in the leaf number [36] or leaf size (unit leaf area) [56]. Thus, plants decrease water limitation by reducing the transpiration area. This is a typical mechanism of stress avoidance used by plants to overcome water stress conditions [57].
The gas exchange rates are most affected by water shortage. The principal site of stress in the plants is the photosynthetic apparatus, which is very sensitive to drought; as a consequence, the photosynthetic activity is reduced because of stomatal closure and complex non-gassing effects [58]. In the present study, A N , gs, and E significantly decreased in the leaves of zinnia under moderate and severe drought stress, and A N decreased substantially in 25% FC, indicating that the reduction in A N in the first stage might be determined by the stomatal closure inhibition of CO 2 uptake. Drake et al. [59] suggested that many small stomata lead to greater stomatal resistance and stomatal control, as well as a rapid response in water stress conditions. In our study, in the more stressed treatments (50% FC and 25% FC), an increase in stomata density and a decrease in size were observed. With the reduction of transpiration, the water use efficiency generally increases. Different studies have reported that stressed plants are more able to utilise the energy obtained by photosynthesis because of higher WUE [57]. This is in accordance with our results, which showed that the most stressed plants activated this mechanism by increasing the WUE values. Another typical avoidance strategy during drought stress conditions is a change in the size and density of the stomata [57].
A high RWC during drought stress is a key mechanism for maintaining the metabolic activity in plants; it is a well-known mechanism for inducing drought tolerance in breeding activity [60]. Babaei et al. [61], in a study on Tagetes minuta, noted that the low RWC values in this species during drought showing a low recovery capacity. A higher RWC is generally maintained by plants under moderate drought stress in relation to plants under severe drought stress. Furthermore, the growth parameters were more negatively influenced by the imposed water stress than they were in zinnia with higher RWC values.
Chlorophylls are responsible for the correct functioning of the photosynthetic apparatus because they are essential pigments of the superior assimilating tissues of plants. Under salt or water stress, the pigment photosynthetic (chlorophyll a + b) content can be indicative of stressful conditions [62]. Severe drought stress can also impair the concentrations of photosynthetic pigments [63]. A decrease in chlorophyll content under severe water stress has been noted in different species, including Catharanthus roseus [64], Helianthus annuus [65], Tagetes erecta [66], and Viola x witthrockiana [67]. In our study, the effects of the severe drought stress (25% FC), were highlighted in the leaves by a significant lowering of the chlorophyll a and b, total chlorophyll, and carotenoid content. A physiological adaptation mechanism of plants under drought stress conditions could be the maintenance of high chlorophyll content, even if this is correlated with the biological characteristics of the plant [68]. Indeed, Lu et al. [69] observed that increasing the chlorophyll content was a more effective strategy to avoid photooxidation damage and ROS induction by drought. Plants under drought stress typically accumulate osmotic compounds, such as proline, and change phytohormones [18]. Proline has been found to enhance cell turgor, maintain cell osmotic adjustment, and defend cells during dehydration [67]. In our study, it was noted that proline concentration was higher in zinnia plants under severe drought conditions compared with control plants. These results agree with the results of other studies [70] and underline the relevance of proline as a protective element in the stress response [71]. In addition, Oraee et al. [72] showed that the proline concentration was higher in plants under drought treatments compared with well-watered plants. In drought stress conditions, proline accumulation was inversely proportional to the water status of plants [73]; this was confirmed in our study, in which the highest values of proline in 25% FC corresponded to the lowest values of RWC. This suggests the contribution of this solute in plant osmotic adjustment. Similarly, in Indian grain soybean cultivars exposed to mild water stress, an increased proline accumulation corresponded to a smaller reduction in the relative water content and shoot and root fresh/dry weight [74].
Plants, when synthesizing antioxidants and increasing the activity of antioxidative enzymes, activate a protective mechanism against abiotic stress [75]. Antioxidative enzymes that control ROS level in cells-superoxide dismutase (SOD), catalase (CAT), and peroxidase (POX)-are commonly produced in response to drought. Numerous studies have found that drought-tolerant plants have strong scavenging systems that allow them to maintain low levels of ROS and reduce membrane lipid peroxidation during stress [22].
Among the antioxidative enzymes that act in scavenging ROS species, APX, SOD, and GPX have a key role. Ascorbate peroxidase (APX) acts directly on the H 2 O 2 molecules and reduces them to water. SOD conducts the dismutation reaction by reducing the O −2 molecule to H 2 O 2 ; APX and GPX convert H 2 O 2 to water, thus assuring its removal [76]. In this study, plants under 50% FC and 25% FC showed a significant increment in antioxidant activity at the end of the trial. CAT, SOD, and GPX activity was significantly enhanced at moderate and severe drought stress (50% FC and 25% FC). A similar trend was reported by Amiri et al. [77]. In a study by Tian et al. [78], the results indicated that the increase in enzyme activity in marigold was a tolerant response to drought treatment, while this self-regulation level was lower with the improvement of water deficit. The decrease in CAT activity leads to the accumulation of H 2 O 2 and an enhancement in lipid peroxidation, thus increasing MDA and causing damage in plants [78]. Recent studies have observed that proline may also play a role as an antioxidant and not only as an osmotic protectant. Thus, the accumulation of proline plays a key role under drought stress as an antioxidant or through stabilizing macromolecules during water stress [79]. The enhancement of DPPH activity observed in the more stressed plants is in line with other studies [61,80]. The total antioxidant activity is the combined results of all enzymatic and non-enzymatic antioxidant activity in plants subjected to abiotic stresses. Tolerant plants generally show a higher antioxidant content to defend against oxidative stress by keeping high antioxidant enzyme and antioxidant molecule activity and contents under stress conditions [81].
Through the general analysis of the data, we observed an interesting result that the activity of antioxidant enzymes is strongly linked to the increase of secondary metabolites; this is similar to the findings of Catola et al. [82] that secondary metabolites work as substrates for enzymes involved in enzymatic antioxidant reactions [51]. The PCA analysis reported in the present study could help to better understand the influence of drought stress on morphological, physiological, and biochemical characteristics. The PCA results demonstrated the boundary between the drought treatments and the control group was clear. Correlation analysis demonstrated that there were complex and close relationships in the more stressed treatments (50% FC and 25% FC) between the antioxidant enzyme activity, DPPH, proline accumulation, and stomata density, which cooperated to tolerate drought stress. The sensitivity of zinnia plants to drought stress is different, so the response time is also different.

Conclusions
The changes in growth parameters, stomata characteristics, gas exchange, osmotic regulators, and antioxidant activity were measured to study the response mechanism of Z. elegans to drought stress. During drought exposure in the nursery phase, zinnia showed adaptive changes to water limitation. At the physiological level, zinnia responded to the drought stress by reducing the RWC and biomass and increasing the levels of osmotic regulators (proline) and antioxidant activity. In severe deficit irrigation, the strategies adopted by the plants were not able to resist drought stress. With light deficit irrigation, the plants could perform as well as fully irrigated plants. In medium deficit irrigation, the mechanisms were not always are suitable to overcome drought stress.  Data Availability Statement: Main data are contained within the article; further data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.