Di ﬀ erential Response to Water Deﬁcit in Chili Pepper ( Capsicum annuum L.) Growing in Two Types of Soil Under Di ﬀ erent Irrigation Regimes

: Chili pepper ( Capsicum spp.) is one of the most economically important horticultural crops in the world; its production for the food and pharmaceutical industries has been increasing worldwide. The economic importance of this crop is due, in part, to the nutraceutical properties derived from its secondary metabolism. Drought is the main environmental factor that a ﬀ ects crop production. Nevertheless, studies involving water deﬁcit have considered short-term responses to sharp water deﬁcit rather than long-term acclimation processes through moderate and gradually increasing water deﬁcits, which omitted the dynamics and proﬁle of the secondary metabolism that are part of the plant’s defence system against this stress factor. The present study aimed to identify the di ﬀ erent mechanisms that chili pepper plants use to cope with drought stress using a progressive decrease and increase of water availability, conditions that commonly occur for crops in open ﬁelds. Four treatments were applied as follows: gradual water deﬁcit (GWD), initial waterlogging with gradual water deﬁcit (IWGD), sudden water deﬁcit with gradual recovery (SWDR), and no deﬁcit of water (NDW). These conditions should represent a more real situation similar to that faced by plants in the agricultural environment. In order to evaluate the response mechanisms associated with these water deﬁcits, changes in phenological variables, proline accumulation, and the gene expression of phenylalanine ammonia-lyase ( PAL ), chalcone synthase ( CHS ), peroxidase ( POD ), and superoxide dismutase ( SOD ) were measured in chili pepper plants growing on land under di ﬀ erent irrigation regimes in two contrasting soil types in areas where chili pepper plants are cultivated in central Mexico. The variables evaluated showed a di ﬀ erentiated response of the mechanisms in plants growing under di ﬀ erent levels of water deﬁcit. Given the di ﬀ erential response observed for the gene expression and morphological and biochemical variables studied in chili pepper plants against di ﬀ erent water regimes, in this work, this may have implications for more e ﬃ cient use of water in crops with high nutraceutical content, in addition to prospects for using products derived from secondary metabolism in the pharmaceutical industry. We studied the impacts of these conditions on morphological variables, proline accumulation, and the expression of phenylalanine ammonia-lyase (PAL ), chalcone synthase ( CHS ), peroxidase ( POD ), and superoxide dismutase ( SOD ) genes. These factors were used as indicators of the possible mechanisms used by chili pepper plants grow on land under di ﬀ erent irrigation regimes.


Introduction
The genus Capsicum spp. (chili pepper) includes 25 to 30 species, of which the following five are cultivated: C. annuum L., C. frutescens L., C. chinense Jaqch, C. baccatum, and C. pubescens. These five

Establishment of the Experiment
The plant material used for the present study was Capsicum annuum L. cv. ROMEL F-1 from the CALORO ® seed house, lot JR01107, and purity 99%. It was treated with Thiram and had uniform relative maturity. In the greenhouse, the chili pepper plants were handled in culture bags. The experimental design used was split plots as follows: A whole plot with two levels of soil type, sandy soil and clay soil, and a strip plot with four water stress regimes, i.e., gradual water deficit (GWD) ( Figure 1A), initial waterlogging with gradual water deficit (IWGD) ( Figure 1B), sudden water deficit with gradual recovery (SWDR) ( Figure 1C), and no deficit of water (NDW) ( Figure 1D). The experimental unit consisted of three plants and six repetitions. A randomized block design was utilized.
Agriculture 2020, 10, x FOR PEER REVIEW 3 of 15 Clay soil samples were taken at the coordinates 20°35′44ʺ N, 100°49′37ʺ W. Twenty representative samples were collected using the standardized methods for the collection of soil samples. Sandy soil was collected from the coordinates 18°19′47,622ʺ N, 97°10′23,498′ W. Twenty representative samples were collected. The characteristics of the soil are shown below in Table 1.

Establishment of the Experiment
The plant material used for the present study was Capsicum annuum L. cv. ROMEL F-1 from the CALORO ® seed house, lot JR01107, and purity 99%. It was treated with Thiram and had uniform relative maturity. In the greenhouse, the chili pepper plants were handled in culture bags. The experimental design used was split plots as follows: A whole plot with two levels of soil type, sandy soil and clay soil, and a strip plot with four water stress regimes, i.e., gradual water deficit (GWD) ( Figure 1A), initial waterlogging with gradual water deficit (IWGD) ( Figure 1B), sudden water deficit with gradual recovery (SWDR) ( Figure 1C), and no deficit of water (NDW) ( Figure 1D). The experimental unit consisted of three plants and six repetitions. A randomized block design was utilized.  (NDW). The Y-axis represents the percentage of the volume of water that corresponds to the field capacity and the X-axis represents the days during which the irrigation volume was applied. . The Y-axis represents the percentage of the volume of water that corresponds to the field capacity and the X-axis represents the days during which the irrigation volume was applied.

Calculation of Stress Type
To calculate the water available (AD) to the irrigation sheet, the field capacity (CC) formula for irrigation was used based on the texture (sandy (AR), clay (AC)) and the organic matter (MO) content according to Silva et al. (1) [17]. For calculation of the permanent wilting point (PMP), the formula described by Silva et al. (2) [17] was used. The available water was calculated with the following Formula (3) for every ten centimeters of soil, and the apparent density (DAP) was estimated in g/cm 3 using the soil texture class table from the Soil Survey Staff USDA. The soil was considered to be a single horizon; the irrigation sheet was calculated for 30 cm with a volume of soil for the culture bags of 0.00340 m 3 . The characteristics of the soil from which the amount of water used to irrigate the plants was calculated are shown in Table 1, and the chemical characteristics are shown in Table 2. (1)

Irrigation System
The stress conditions were induced by 4 l/h drippers Rainbirt ® (Rainbird CP075, Asuza, CA, USA), and the precision irrigation control system was managed by a SmartReley Schneider ® model zelio ® (Schneider Electric Automation GmbH, Lahr, Germany) and an independent solenoid valve system for each stress level and soil type, as shown in Figure 2. To calculate the water available (AD) to the irrigation sheet, the field capacity (CC) formula for irrigation was used based on the texture (sandy (AR), clay (AC)) and the organic matter (MO) content according to Silva et al. (1) [17]. For calculation of the permanent wilting point (PMP), the formula described by Silva et al. (2) [17] was used. The available water was calculated with the following Formula (3) for every ten centimeters of soil, and the apparent density (DAP) was estimated in g/cm 3 using the soil texture class table from the Soil Survey Staff USDA. The soil was considered to be a single horizon; the irrigation sheet was calculated for 30 cm with a volume of soil for the culture bags of 0.00340 m 3 . The characteristics of the soil from which the amount of water used to irrigate the plants was calculated are shown in Table 1, and the chemical characteristics are shown in Table 2

Irrigation System
The stress conditions were induced by 4 l/h drippers Rainbirt ® (Rainbird CP075, Asuza, CA, USA), and the precision irrigation control system was managed by a SmartReley Schneider ® model zelio ® (Schneider Electric Automation GmbH, Lahr, Germany) and an independent solenoid valve system for each stress level and soil type, as shown in Figure 2.

Measurement of the Morphological Variables
The plant height was measured at the beginning of the application of the water stress treatments, and then again at the end of each stress treatment. For the determination of this characteristic, a measuring tape was used. The tape was placed at the base of the main stem at the height of the root ball and was extended to the apex of growth. The stem diameter was measured at the base of the stem; the readings were measured at the beginning of the application of treatments, and then again at the end of each stress treatment using a digital Vernier caliper with a resolution of 0.01 mm. The numbers of fruits and leaves were quantified at the end of the application of the treatments.

Determination of Proline Concentration in Capsicum annuum L. Leaves
Proline determination was performed according to the method proposed by Bates et al. [18] with some modifications. First, the leaf samples were ground with liquid nitrogen. A quantity of 0.05 g of plant material was homogenized in 1 mL of 3% aqueous sulfosalicylic acid, and this was centrifuged at 12,000 rpm for 10 min. The aqueous phase was reacted with ninhydrin acid and glacial acetic acid in a 1:1:1 ratio. The reaction was incubated at 100 • C for 1 h, and then the reaction was stopped in an ice bath. The reaction mixture was extracted with toluene by vigorously mixing for 15 to 20 s. Finally, samples were read at an absorbance of 520 nm. The standard curve was developed at concentrations of 50, 40, 30, 20, 10, 5, and 0 g/mL of L-proline. The results are reported as µmoles proline/g of fresh weight material.

RNA Extraction and cDNA Synthesis
The samples were ground with liquid nitrogen. Total RNA was extracted from pepper leaves using Trizol reagent, according to the manufacturer s instructions. Subsequently, the ZYMO RESEARCH DIRECT ZOL RNA MINIPREP (Zymo Research, Irvine, CA, USA) kit was purified with RNA according to the manufacturer s specifications. Concentration, purity, and quality were determined using an RNA (concentration, quality, and purity) Thermo Scientific™ AquaMate 8000 UV-Vis Spectrophotometer (Thermo Fisher Scientific, Waltham, MA, U.S.). cDNA synthesis was performed with the Thermo Scientific™ First Strand cDNA Synthesis Kit MAXIMA FIRST STRAND (Thermo Fisher Scientific, Waltham, MA, USA) for RT-qPCR according to the manufacturer's instructions.

RT-qPCR Conditions and Analysis
The qRT-PCR analysis was carried out on 96-well plates with the CFX96 Touch Deep Well™ Real-Time PCR Detection System (Bio-Rad). The primers used in this study were those previously reported and used in a study in Chile by Rodríguez-Calzada et al. [14] and included the following: phenylalanine ammonia-lyase (PAL; AF081215) forward 5 -att cgc gct gca act aag at−3 and reverse 5 -cac cgt gta agg cct tgt tt-3 ; chalcone synthase (CHS; FJ705842.1) forward 5 -tcg acc ctc agt caa acg ac-3 and reverse 5 -tgg gcc acg gaa agt aac tg-3 ; superoxide dismutase (MnSOD; AF036936.2) forward 5 -ctc tgc cat aga cac caa ctt-3 and reverse 5 -cca agt tcg gtc ctt taa taa-3 ; peroxidase (POD; FJ596178.1) forward 5 -gca gca ttc ctc ctc cta ct-3 and reverse 5 -att tct ttg cct tgt tgt tg-3 ; and beta-tubulin (β-TUB; EF495259.1) forward 5 -gag ggt gag tga gca gtt c-3 and reverse 5 -ctt cat cgt cat ctg ctg tc-3 . These were all provided by Eurofins, Germany. To the reaction mixture, 5 µL of the iQ SYBR Green Supermix (Bio-Rad), 2 µL of cDNA, and the reverse primers (tubulin 0.15 µL, POD 0.15 µL, PAL 0.15 µL, CHS 0.9 µL, and MnSOD 0.15 µL) and forward primers (tubulin 0.3 µL, POD 0.15 µL, PAL 0.3 µL, CHS 0.9 µL, and MnSOD 0.15 µL) were added to reach a final concentration of 10 µL. The cycling conditions for the amplification of tubulin, PAL, and POD were 5 min at 94 • C for 40 cycles (1 min at 94 • C followed by 1 min at 55 • C). Finally, melting curve heating of the amplicon was performed through a step of 31 s at 65 • C and an increase of 0.5 • C to 95 • C. In the case of CHS and MnSOD, the alignment temperature changed with a temperature of 58 • C. Samples of each treatment were analyzed in triplicate using SyberGreen mix as the fluorophore to quantify by RT-qPCR (BIO-RAD ® C1000), and the results were evaluated using the 2 −∆∆ CT method [19].

Statistical Analysis
For all experiments, one-way ANOVA was performed, and for multiple mean comparisons, Tukey s Test with p = 0.05 was applied. Statistical analysis and graphics were performed using JMP ® 13.2.1 software (SAS Institute Inc.).

Effect of Soils on Morphological and Biochemical Variables of Capsicum annuum L.
The results of this section correspond to the analysis of variance of the experimental design of divided plots. Because an interaction effect was detected between the type of soil and the treatments of irrigation regimes, the results are presented independently because they also show significant effects. In order to evaluate the impacts that sandy and clay soils have on the performance of C. annuum, some morphological variables of chili plants grown in these soils were measured. Plant height, stem diameter, leaf number, and the number of fruit values were measured as indicators of plant performance. ANOVA analysis and a comparison of means by using Tukey s test for the morphological variables of chili grown in the two types of soil are shown in Figure 3. There were highly significant differences in the statistical average estimator for height (20. In order to determine whether proline is involved in any mechanisms of plant defence in the two types of soil, its concentration in the leaves was measured. The statistical average estimator of proline accumulation in the leaves of the plants grown in clay soil was 1.082 µmol/g of fresh matter as compared with 0.573 µmol/g of fresh matter in sandy soil, showing a significant difference (Figure In order to determine whether proline is involved in any mechanisms of plant defence in the two types of soil, its concentration in the leaves was measured. The statistical average estimator of proline accumulation in the leaves of the plants grown in clay soil was 1.082 µmol/g of fresh matter as compared with 0.573 µmol/g of fresh matter in sandy soil, showing a significant difference (Figure 4). The results suggest that plants grown in the clay soil turn on osmoregulation as a defence mechanism.
Comparison of means ± S.E. by using Tukey′s test for the morphological variables that were assessed in sandy and clay soils. In order to determine whether proline is involved in any mechanisms of plant defence in the two types of soil, its concentration in the leaves was measured. The statistical average estimator of proline accumulation in the leaves of the plants grown in clay soil was 1.082 µmol/g of fresh matter as compared with 0.573 µmol/g of fresh matter in sandy soil, showing a significant difference ( Figure  4). The results suggest that plants grown in the clay soil turn on osmoregulation as a defence mechanism. In the case of gene expression analysis, there was a non-significant difference.

Effects of the Type of Drought Treatment on Some Morphological and Biochemical Variables of Capsicum annuum L.
The effects of different water application regimes (treatments) on plants grown in sandy and clay soils were evaluated ( Figure 5). The water stress treatments evaluated can be seen in Figure 1. In the case of gene expression analysis, there was a non-significant difference.

Effects of the Type of Drought Treatment on Some Morphological and Biochemical Variables of Capsicum annuum L.
The effects of different water application regimes (treatments) on plants grown in sandy and clay soils were evaluated ( Figure 5). The water stress treatments evaluated can be seen in Figure 1. Note that the plants exposed to the GWD treatment were more stressed than those exposed to other treatments with an irrigation volume of 20% of field capacity maintained for seven days at the end of the trial period ( Figure 1A). In the case of the IWGD treatment, the stress was started at field capacity followed by seven days at 120%, irrigation at 80%, and finally, a decrease of 20% every eight days until the treatment was concluded at 40% ( Figure 1B). The SWDR treatment started with seven days with field capacity irrigation and then seven days of irrigation at 20% of the field capacity. Then, every week, it increased until field capacity was reached ( Figure 1C). Regarding treatment NDW, this was considered to be a control because field capacity was maintained throughout the experiment ( Figure 1D).
In Figure 5, it can be seen in all cases that the statistical average estimator values for the morphological variables in the NDW treatment, corresponding to field capacity, were as follows: plant height, 21.954 cm ( Figure 4); stem diameter, 6.23 mm ( Figure 5B); number of leaves, 44.064 ( Figure 5C); and number of fruit, 2.516 ( Figure 5D). These values were significantly higher than those measured in the rest of the treatments. Thus, the influence of hydric stress on plant behavior can be observed in the morphological characteristics.
In Figure 5, it can be seen in all cases that the statistical average estimator values for the morphological variables in the NDW treatment, corresponding to field capacity, were as follows: plant height, 21.954 cm ( Figure 4); stem diameter, 6.23 mm ( Figure 5B); number of leaves, 44.064 ( Figure 5C); and number of fruit, 2.516 ( Figure 5D). These values were significantly higher than those measured in the rest of the treatments. Thus, the influence of hydric stress on plant behavior can be observed in the morphological characteristics. In Figure 6, for the GWD treatment, it can be seen that the relative expression levels of the POD ( Figure 6B) and SOD ( Figure 6D) genes were significantly higher, 5.529-fold and 2.632-fold, respectively, as compared with the other treatments. On the one hand, in the IWGD treatment, PAL ( Figure 6A) expression was significantly higher (21.579-fold) as compared with the rest of the treatments. On the other hand, in SWDR, the proline concentration ( Figure 6C) was statistically higher (1.335 µmol/g) than the rest of the other treatments. The NDW treatment did not stand out in terms of any of the mechanisms that can be inferred from proline accumulation and gene expression. In Figure 6, for the GWD treatment, it can be seen that the relative expression levels of the POD ( Figure 6B) and SOD ( Figure 6D) genes were significantly higher, 5.529-fold and 2.632-fold, respectively, as compared with the other treatments. On the one hand, in the IWGD treatment, PAL ( Figure 6A) expression was significantly higher (21.579-fold) as compared with the rest of the treatments. On the other hand, in SWDR, the proline concentration ( Figure 6C) was statistically higher (1.335 µmol/g) than the rest of the other treatments. The NDW treatment did not stand out in terms of any of the mechanisms that can be inferred from proline accumulation and gene expression. To determine the effect that the type of soil has on the performance of the plants under different types of water stress (Figure 1), their performance in clay soil was measured (Table 1). The estimators of the mean for the morphological variables for the NDW treatment were significantly higher for height, 21.573 cm ( Figure 7A); stem diameter, 6.212 mm ( Figure 7B); number of leaves, 36.6 ( Figure  7C); and number of fruit, 2 ( Figure 7D). To determine the effect that the type of soil has on the performance of the plants under different types of water stress (Figure 1), their performance in clay soil was measured (Table 1). The estimators of the mean for the morphological variables for the NDW treatment were significantly higher for height, 21.573 cm ( Figure 7A); stem diameter, 6.212 mm ( Figure 7B); number of leaves, 36.6 ( Figure 7C); and number of fruit, 2 ( Figure 7D).
In the case of biochemical variables in the IWGD and SWDR treatments, highly significant accumulations of proline (1.55 and 1.65 µmol/g, respectively) were observed ( Figure 8A). In terms of gene expression, the GWD treatment showed the highest expression of SOD 3.09-fold ( Figure 8B). The stress-response variables showed statistically lower values in the plants treated with the different types of water stress and clay soil; however, only two defence indicators were statistically significant, i.e., proline and SOD gene expression. In the case of biochemical variables in the IWGD and SWDR treatments, highly significant accumulations of proline (1.55 and 1.65 µmol/g, respectively) were observed ( Figure 8A). In terms of gene expression, the GWD treatment showed the highest expression of SOD 3.09-fold ( Figure 8B). The stress-response variables showed statistically lower values in the plants treated with the different types of water stress and clay soil; however, only two defence indicators were statistically significant, i.e., proline and SOD gene expression.   In the case of biochemical variables in the IWGD and SWDR treatments, highly significant accumulations of proline (1.55 and 1.65 µmol/g, respectively) were observed ( Figure 8A). In terms of gene expression, the GWD treatment showed the highest expression of SOD 3.09-fold ( Figure 8B). The stress-response variables showed statistically lower values in the plants treated with the different types of water stress and clay soil; however, only two defence indicators were statistically significant, i.e., proline and SOD gene expression.  The influence of sandy soil (Table 1) on the morphological ( Figure 9) and biochemical ( Figure 10) variables of chili pepper plants was evaluated. The morphological variables for the NDW treatment were significantly higher for stem diameter, 6.246 mm ( Figure 9A); number of leaves, 51.06 ( Figure 9B); and number of fruit, 2.93 ( Figure 9C). The influence of sandy soil (Table 1) on the morphological ( Figure 9) and biochemical ( Figure  10) variables of chili pepper plants was evaluated. The morphological variables for the NDW treatment were significantly higher for stem diameter, 6.246 mm ( Figure 9A); number of leaves, 51.06 ( Figure 9B); and number of fruit, 2.93 ( Figure 9C). In the case of biochemical variables, the IWGD treatment showed highly significant gene expression of PAL, 17.88-fold ( Figure 10A) as compared with the other treatments; the GWD treatment showed highly significant gene expression of POD, 8.73-fold ( Figure 10B) and SOD, 2.17fold ( Figure 10B). The SWDR treatment displayed highly significant accumulation of proline, 1.11 µmol/g ( Figure 10C). The stress-response variables showed statistically significant values in the plants treated with different types of water stress and sandy soil, showing that four stress-response indicators were turned on. In the case of biochemical variables, the IWGD treatment showed highly significant gene expression of PAL, 17.88-fold ( Figure 10A) as compared with the other treatments; the GWD treatment showed highly significant gene expression of POD, 8.73-fold ( Figure 10B) and SOD, 2.17-fold ( Figure 10B). The SWDR treatment displayed highly significant accumulation of proline, 1.11 µmol/g ( Figure 10C). The stress-response variables showed statistically significant values in the plants treated with different types of water stress and sandy soil, showing that four stress-response indicators were turned on.

Discussion
The results of the four morphological variables (height, stem diameter, number of leaves, and number of fruit) evaluated as indicators to estimate the stress response in chili pepper plants cultivated under different irrigation regimes representing different drought conditions showed significantly higher values for plants grown in sandy as compared with clay soil. More specifically, our results highlight the importance of the soil type as a component of the environment in agriculture. Concerning the four variables, they influence the survival potential for chili pepper plants exposed to water deficit. The results are consistent with previous reports concerning the role of the soil in phenotypic expression and, in this case, in the plants′ response to drought stress [20]. Proline is an osmolyte that plays a role in the osmoregulation mechanism during the response of plants to drought stress, although the precise details of this mechanism are unknown [21]. Proline accumulation has been reported to occur in chili plants exposed to a wide range of environmental stressors, including drought and salinity [22]. In this context, the results observed in Figure 3 suggest that the significantly higher accumulation of proline in plants grown in clay soil was due to the higher sodium content of the soil (Table 2), which led to frequent use of osmoregulation as a mechanism to deal with stress. Thus, a greater effort was required by the plants, giving the plants grown in the sandy soil a disadvantage. According to Kahlaoui et al. [23], this could explain the difference in the observed survival potential. As previously noted, no significant differences were detected regarding the average gene expression of PAL, POD, CHS, and SOD when the data obtained from plants grown in sandy and clay soils were compared. This situation suggests that in both types of soils, the conditions required for growth of the chili plants are available, allowing the genes to be expressed in a similar way.

Discussion
The results of the four morphological variables (height, stem diameter, number of leaves, and number of fruit) evaluated as indicators to estimate the stress response in chili pepper plants cultivated under different irrigation regimes representing different drought conditions showed significantly higher values for plants grown in sandy as compared with clay soil. More specifically, our results highlight the importance of the soil type as a component of the environment in agriculture. Concerning the four variables, they influence the survival potential for chili pepper plants exposed to water deficit. The results are consistent with previous reports concerning the role of the soil in phenotypic expression and, in this case, in the plants response to drought stress [20]. Proline is an osmolyte that plays a role in the osmoregulation mechanism during the response of plants to drought stress, although the precise details of this mechanism are unknown [21]. Proline accumulation has been reported to occur in chili plants exposed to a wide range of environmental stressors, including drought and salinity [22]. In this context, the results observed in Figure 3 suggest that the significantly higher accumulation of proline in plants grown in clay soil was due to the higher sodium content of the soil (Table 2), which led to frequent use of osmoregulation as a mechanism to deal with stress. Thus, a greater effort was required by the plants, giving the plants grown in the sandy soil a disadvantage. According to Kahlaoui et al. [23], this could explain the difference in the observed survival potential. As previously noted, no significant differences were detected regarding the average gene expression of PAL, POD, CHS, and SOD when the data obtained from plants grown in sandy and clay soils were compared. This situation suggests that in both types of soils, the conditions required for growth of the chili plants are available, allowing the genes to be expressed in a similar way.
The results shown in Figures 5 and 6 indicate that, in each treatment, representing different types of drought conditions, the mechanisms used by the plant to cope with stress differed. It should be noted that, although the estimators of the means of the morphological variables, i.e., the plant height, stem diameter, the number of leaves, and the number of fruit were statistically equal in the GWD, IWGD, and SWDR treatments ( Figure 5), the data suggest that each one uses a different mechanism to cope with drought stress. As shown in Figure 6, in the GWD treatment, there was significantly greater gene expression activity for POD and SOD, which suggests that the plants probably used increased enzyme-type antioxidant activity [24]. To cope with long-term stresses, plants have evolved to counteract the effects of reactive oxygen species (ROS) with antioxidants and antioxidant enzymes that maintain the balance of ROS within the cell [25]. The antioxidant enzymes POD, and SOD play key roles in the detoxification of cellular ROS [26]. Recent evidence indicates that reactive oxygen species (ROS) and reactive nitrogen species (RNS) are among the main intracellular signal transducers that sustain autophagy [27]. Our results reinforce the notion of autophagy as a probable mechanism of the response to drought stress [28,29]. The plants grown under IWGD stress differed from those grown under the other treatments because greater PAL gene expression was induced ( Figure 6A). Thus, analogously to reports in tomato, our results suggest that chili pepper plants probably use a mechanism that differentially involves the synthesis of phenols and capsaicinoids [30]. It should be mentioned that in the IWGD treatment, water availability was decreased over time, contrary to in the SWDR treatment, which began with a low amount of water that was subsequently increased. However, plants grown under these two treatments showed greater amounts of proline, which indicates their accumulation on the day of harvest; consequently, the same result was obtained despite the phase of plant recovery or dehydration. Plants that were grown in this drought regime stand out slightly by their accumulation of proline ( Figure 6D). Proline is a molecule that has various functions in the plant related to the tolerance of both biotic and abiotic stress. Under drought conditions, proline helps the plant to resist dehydration through its function as an osmolyte [31]. Therefore, our data suggest that plants grown under the conditions of the SWDR treatment had a greater tendency to use osmoregulation to treat this type of stress. The NDW treatment, which was maintained at field capacity ( Figure 1D), was always significantly different than the other treatments in terms of the four morphological variables that were evaluated as indicators to estimate the stress response, and among these, the behaviors were statistically the same ( Figure 5). However, the results observed in the treatments with different types of drought stress applied showed significant differences in the proline concentration and the gene expression of PAL, POD, CHS, and SOD ( Figure 6). This suggests differentiated use of the mechanisms used to cope with drought in the two types of soil. On the one hand, the results of the estimators of the morphological variables in clay ( Figure 7) and sandy soil ( Figure 9) were similar to those shown in Figure 5. On the other hand, in sandy soil, significant differences in proline accumulation and the gene expression of POD, SOD, PAL, and CHS were observed in each type of water stress ( Figure 10). In the case of clay soil, there are two differentiated mechanisms, the GWD treatment showed gene expression of SOD, while IWGD and SWDR accumulated proline for osmoregulation activation (Figure 8).
The results obtained can serve as a basis for the design of irrigation systems aimed at saving water, an increasingly relevant perspective in agriculture. The above would be possible by taking advantage of the knowledge provided in this work regarding the expression of the genes that allow the plant to fight against stress due to lack of water.
In addition, it could be fundamental for designing irrigation strategies that allow an increase in the beneficial bioactive substances for the consumer, either by the increase of the antioxidant capacity related to the activity of the enzymes in the oxidation processes or by means of the synthesis of bioactive like phenols or flavonoids, which are substances that are dedicated to the defence of the plant against stress factors and with potential in the pharmaceutical industry.

Conclusions
Chili pepper plants reacts differentially to drought stress depending on the way that the plant experiences water deficit. This differential response involves several mechanisms among which are osmotic regulation, the use of secondary metabolism, and redox regulation. Moreover, under our experimental conditions, the results also suggest that the mechanism used by chili pepper cultivar when facing strong water stress is oxidative regulation through activation of SOD gene expression, and thus the enzyme superoxide dismutase (SOD) plays a major role.