Gas exchange, growth, and production of cotton genotypes under water deficit in phenological stages

ABSTRACT Cotton cultivation in the Brazilian Northeast region faces water scarcity problems caused by the irregularity of the rainy season, leading to losses in yield. In this context, the objective of the present study was to evaluate the gas exchange, growth, and production of colored-fiber cotton genotypes under water stress, varying the water deficit management strategies in the different phenological stages of the plant. The study was carried out in the experimental area of the Federal University of Campina Grande, located in the municipality of Pombal, Paraíba, Brazil. A randomized block design was used, in a 3 × 7 factorial scheme, corresponding to three colored cotton genotypes (‘BRS Rubi’, ‘BRS Jade’, and ‘BRS Safira’) and application of water deficit (40% of actual evapotranspiration - ETr) management strategies in seven phenological stages of the crop. The ‘BRS Jade’ genotype is the most suitable for cultivation under water deficit conditions with 40% of the actual evapotranspiration. Colored-fiber cotton cultivation under water deficit in the flowering stage caused a reduction in physiological variables and growth. Water deficit during the vegetative and yield formation stages promoted lower losses in the production of seed cotton and total seed weight of the genotypes ‘BRS Rubi’, ‘BRS Jade’, and ‘BRS Safira’.


INTRODUCTION
Water scarcity is a global problem resulting in restrictions on agricultural production, a common characteristic in arid and semi-arid areas, causing reductions of more than 50% in crop yield worldwide (NDAMANI; WATANABE, 2015;BABOEV et al., 2017).Currently, cotton production areas are characterized by the occurrence of long periods of drought, causing considerable yield losses (CORDÃO SOBRINHO et al., 2015;LIMA et al., 2018).
Cotton cultivation in northeastern Brazil has stood out as one of the agricultural activities of great value for Brazilian agribusiness, with the production of 7,089,939 tons, accounting for 23.29% of the national production, of which the state of Paraíba produced a total of 3,596 tons of herbaceous cotton (IBGE, 2020).The areas are mostly cultivated under rainfed conditions, i.e., depending on rains that have irregular distribution, sometimes with prolonged periods of drought, and rains concentrated in a few months during the year, limiting crop development and production (NOBRE et al., 2012;SOARES et al., 2015).
In general, water deficit can affect the organs of developing plants, L. A. A. SOARES et al. especially during the flowering and fruiting stages, which negatively affects the morphological characteristics of the plant and its production components (ZONTA et al., 2017;ARAÚJO et al., 2019;SOARES et al., 2020).Water deficit can inhibit cell and leaf expansion, stem elongation, and leaf area index (LIU et al., 2017).Leaf area and reduction in the production of new leaves are also associated with a decrease in leaf water potential and stomatal closure (QI; TORII, 2018).Thus, maximizing water use efficiency at all scales is crucial in agricultural systems, especially during drought periods, when water for irrigation is scarce.Deficit irrigation is used to increase the water use efficiency of the crop, decreasing the water depth or the number and frequency of irrigation events (SAMPATHKUMAR et al., 2013).However, the efficiency of deficit irrigation depends mainly on the phenology of the crop and effects related to the moment, duration, physiological state of the crop, irrigation water quality, genotype, and stress severity (GARCÍA-TEJERO et al., 2012;CHAI et al., 2016).
In this context, genetic diversity and deficit irrigation can reduce water consumption in irrigation, increasing water use efficiency to ensure sustainable agricultural production under intermittent drought events (ABID et al., 2018;CORDÃO et al., 2018).Therefore, it is necessary to adopt appropriate agronomic strategies for agricultural production under water deficit in each developmental stage and recovery of plants in the phenological stages following water stress, as well as the consequences of the cumulative effect of water deficit in the phenological stages (ZONTA et al., 2015).
Thus, this study has the hypothesis that the identification of a cotton genotype tolerant to water deficit in a given stage through the effects on gas exchange, growth, and production will enable the increase of production capacity profitability.The effects of water deficit on plant physiological functions are dependent on genotype, stage, and period of exposure to stress.Therefore, the identification of the stage(s) in which the cotton crop is tolerant or sensitive is of paramount importance to ensure the sustainability of irrigated cultivation under semi-arid conditions and contribute to the generation of employment and income.
In this context, the objective of this study was to evaluate the gas exchange, growth, and production of colored cotton genotypes under water deficit in the different phenological stages.

MATERIAL AND METHODS
The experiment was carried out under field conditions at the Center of Science and Agri-Food Technology -CCTA of the Federal University of Campina Grande -UFCG, located in the municipality of Pombal, Paraiba, Brazil, at the geographic coordinates 6º47'20" S latitude and 37º48'01" W longitude, at an altitude of 194 m.The meteorological data collected during the experimental period are presented in Figure 1.Three colored cotton genotypes (G) (G1 -'BRS Rubi'; G2 -'BRS Jade' and G3 -'BRS Safira') and seven irrigation management strategies, under two water conditions, corresponding to irrigation with 100% of Actual Evapotranspiration -ETr (full irrigation) and irrigation with 40% of ETr (water deficit) were evaluated, as described in Table 1.The experimental design used was in randomized blocks, in a 3 × 7 factorial scheme, whose factors resulted in 21 treatments with three replicates and three plants per plot, totaling 189 experimental units.Irrigation strategies with water deficit can be characterized as alternative cultivation aiming at maximizing production and water use efficiency under conditions of water scarcity.
Rev. Caatinga, Mossoró, v. 36, n. 1, p. 145 -157, jan. -mar., 2023 147 The evaluated phenological stages of cotton were: vegetative -beginning with the appearance of the first pair of true leaves and end of stress with the opening of the 1 st flower; flowering -from the appearance of the 1 st flower extending up to the opening of the 1 st boll; yield formationfrom the opening of the 1 st boll up to the final harvest of the bolls in each experimental unit.
The plants were grown in plastic containers (pots) with 20 L capacity (35 cm high × 31 cm upper diameter × 20 cm lower diameter), with a geotextile at the base to avoid loss of soil material and filled with a 3-cm layer of crushed stone number 0 and 24.5 kg of a Neossolo Regolítico Eutrófico (Psamment) of sandy loam texture (0-0.30m depth) from an agricultural area in the municipality of Pombal-PB.Soil characteristics were determined according to Teixeira et al. (2017) before sowing and are described in Table 2.At the base of each pot, a transparent tube was connected to facilitate drainage, coupled to a 2.0 L container where the drainage water was collected to determine the water consumption of the plants through the difference between volume of water applied and drained.Ca 2+ and Mg 2+ extracted with 1 M KCl at pH 7.0; Na + and K + extracted with 1 M NH 4 OAc; P -Mehlich-1 extractant; pH sp -pH of saturated paste; and EC se -electrical conductivity of saturation extract, index 1 and 2 correspond to soil moisture at field capacity (FC) and permanent wilting point (PWP).
For the three cotton genotypes, sowing was performed using five seeds per pot at 2 cm depth and distributed equidistantly.After emergence, thinning was done, leaving the most vigorous plant.Soil moisture was maintained at the level corresponding to field capacity (FC) in all experimental units until the emergence of the first true leaf, when (after 20 days) the treatments began to be applied.
Irrigations were performed daily at 5 p.m., applying in each pot the volume of water corresponding to each treatment (40 or 100% ETr), determined by the water balance: volume applied minus the volume drained in the previous irrigation, plus a leaching fraction of 20% in plants irrigated with 100%  ETr, every seven days.The volume of water applied in each water deficit management strategy was determined based on the consumption of plants under 100% ETr, by the drainage lysimetry (BERNARDO et al., 2019).For the treatment with 40% ETr, the obtained value of ETr was multiplied by the percentage of evapotranspiration (0.40).At 75 DAS, gas exchanges were determined based on the CO 2 assimilation rate -A (μmol CO 2 m -2 s -1 ), stomatal conductance -gs (mol m -2 s -1 ), intercellular CO 2 concentration -Ci (μmol CO 2 m -2 s -1 ), transpiration -E (mmol H 2 O m -2 s -1 ), and these data were then used to calculate the instantaneous water use efficiency -WUEi (A/E) and instantaneous carboxylation efficiency -CEi (A/Ci).Gas exchange determinations were performed with an infrared gas analyzer -IRGA (Model LCpro -SD, from ADC BioScientific, UK).The measurements were performed between 7:00 and 10:00 a.m. on the third fully expanded and photosynthetically active leaf, counted from the apical bud, conducted under natural conditions of air temperature, CO 2 concentration, and using an artificial radiation source of 1,200 μmol m -2 s -1 , established through the curve of photosynthetic response to light (FERNANDES et al., 2021;ROQUE et al., 2022).
In the same period, the following growth variables were determined: number of leaves, plant height (PH), with a ruler graduated in millimeters, from the base of the plant to the apical meristem; stem diameter (SD), with a digital caliper 5 cm from the soil; and leaf area (LA), obtained by measuring the length of the leaves, using the methodology proposed by Grimes and Carter (1969), and calculated using Equation 1: Where: LA -leaf area of each cotton leaf (cm 2 ); x -leaf midrib length (cm); and LA total -total leaf area of the plant (cm 2 ) In the evaluations performed at 75 DAS, only four irrigation management strategies were considered in the statistical analysis ( T1-A and  T4-A 2 B 2 C 1 ), because in this evaluation, the deficit treatments (T5, T6, and T7) had not yet begun.At the end of the crop cycle, at 120 DAS, the plants were collected, separated into leaves, stems, and roots, placed in paper bags, and dried in a forced air circulation oven, maintained at 65 ºC until reaching constant weight.Subsequently, the material was weighed on a 0.001 g precision scale to obtain the biomass of the different parts, and the sum of leaves + stem resulted in shoot dry mass (SDM), thus allowing the calculation of root/shoot ratio (R/S).The production components were also quantified: seed cotton weight and total seed weight analyzed according to the methodology of Embrapa Cotton.The bolls were harvested per plant as they reached the harvest point, characterized by their opening along with drying of the bracts.
The data were subjected to the distribution normality test (Shapiro-Wilk test) and then evaluated by analysis of variance by the F test.In cases of significant effect, the Scott-Knott means comparison test was performed (p≤0.05) for water deficit management strategies, and the Tukey test (p≤0.05)was performed for the cotton genotypes (FERREIRA, 2019).

RESULTS AND DISCUSSION
There were significant effects of the interaction between water deficit management strategies and genotypes on all physiological variables of cotton at 75 DAS (Table 3).All variables except CO 2 assimilation rate were affected by water deficit management strategies, and genotypes did not affect only transpiration and instantaneous carboxylation efficiency (Table 3).
Table 3. Summary of the analysis of variance for stomatal conductance (gs), transpiration (E), internal CO 2 concentration (Ci), CO 2 assimilation rate (A), instantaneous water use efficiency (WUEi), and instantaneous carboxylation efficiency (CEi) of the colored-fiber cotton genotypes (G), as a function of different water deficit management strategies (MS), at 75 days after sowing.The values of stomatal conductance of the genotypes 'BRS Rubi' and 'BRS Jade' were similar, as a function of water deficit strategies in the different phenological stages (Figure 2A).However, the stomatal conductance of the genotype 'BRS Safira' decreased significantly with a water deficit in the flowering stage (T3) and successively in the  Due to the partial closure of the stomata, there were reductions of 27.21 and 37.23% in the leaf transpiration of the genotype 'BRS Safira' for plants subjected to water deficit according to treatments T2 (A 2 B 1 C 1 ) and T3 (A 1 B 2 C 1 ), compared to plants under full irrigation (100% ETr -T1) (Figure 2B).Among the cotton genotypes, there was a significant difference between the water deficit management strategies only when plants were irrigated with 40% ETr in the vegetative and flowering stages (T2 and T3), with the highest leaf transpiration in the 'BRS Jade' genotype (Figure 2B).Stomatal closure reduced the water loss by the stomata in the genotype 'BRS Safira', which explains the lowest transpiration rate observed in the flowering stage under 40% ETr.
Stomatal closure is probably the most important factor in controlling carbon metabolism, but the relative role of stomatal limitation in photosynthesis depends on the intensity of water deficit.Under severe stress, photosynthesis can be more controlled by the capacity of chloroplast to fix CO 2 (such as RuBisCO activity) than by the increase in diffusive resistance (DEEBA et al., 2012).Lower water loss through the stomata allows survival under conditions of water scarcity.However, the mechanism of water saving through stomatal closure may compromise photosynthesis, limiting plant growth (NEGIN; MOSHELION, 2016;MUDO et al., 2020).
The internal CO 2 concentration (Figure 3A) varied, following a reverse trend observed for the other variables as a function of water deficit, i.e., in the genotypes 'BRS Rubi', and 'BRS Jade', in which stomatal conductance and transpiration increased, the values of Ci decreased, as a consequence of the CO 2 flow for the synthesis of organic compounds.The highest mean among genotypes was obtained in 'BRS Jade' under water deficit during the vegetative stage, with 160 μmol CO 2 mol -1 in the T2-A 2 B 1 C 1 strategy (Figure 3A).Although decreases in stomatal conductance simultaneously reduce CO 2 concentration, it is maintained or even increased in the substomatal chamber, indicating nonstomatal limitations to photosynthesis (MAFAKHERI et al., 2010).The water deficit management strategies imposed on the CO 2 assimilation rate did not differ, with differences only among genotypes when water deficit was applied in the vegetative stage, T2 (A 2 B 1 C 1 ), where the lowest CO 2 assimilation rate was observed in the genotype 'BRS Jade' when irrigated with 40% ETr in the vegetative stage (Figure 3B).The reduction in the CO 2 assimilation rate verified in the genotype 'BRS Jade' under water deficit conditions may be related to the greater variation in the values of internal CO 2 concentration, compared to plants irrigated with 100% ETr.The decrease in A in cotton plants under water deficit conditions may be caused by the action of non-stomatal factors, such as the restriction of CO 2 diffusion in mesophyll cells and a decline in RuBisCO affinity with CO 2 (LUO, ZHANG; ZHANG, 2016), due to several coordinated events, such as stomatal closure and reduced activity of photosynthetic enzymes (DEEBA et al., 2012).
In the genotypes 'BRS Rubi' and 'BRS Safira' under water stress, the highest values of instantaneous water use efficiency, 8.32 and 9.29 [(μmol m -2 s -1 ) (mmol H 2 O m -2 s -1 ) -1 ], were obtained when the water deficit was applied in the flowering stage, T3 (A 1 B 2 C 1 ), for 'BRS Rubi' and in the vegetative stage, T2 (A 2 B 1 C 1 ), for 'BRS Safira', resulting in the respective increments of 42.87 and 39.65% compared to plants irrigated with 100% ETr -T1 (A 1 B 1 C 1 ).However, among genotypes, WUEi was lower in 'BRS Jade', with an average of 4.55 [(μmol m -2 s -1 ) (mmol H 2 O m -2 s -1 ) -1 ] in the vegetative stage while in the genotypes 'BRS Rubi' and 'BRS Safira' higher means, 7.42 and 9.29 [(μmol m -2 s -1 ) (mmol H 2 O m -2 s -1 ) -1 ], were observed in plants under water deficit in this same stage (Figures 4A).Similar results were reported by Sabir et al. (2020), in Ziziphus jujuba, observing a decrease in stomatal conductance caused an increase in the instantaneous water use efficiency under water stress.This stomatal response in plants subjected to water deficit may be related to the production of abscisic acid rather than water limitations, which may result in sustainable leaf water potential under water stress conditions (PAPACEK; CHRISTMANN;GRILL, 2019;YANG et al., 2019).
There was a significant effect of the interaction (p≤0.01) between the factors strategies of water deficit management and genotypes only on plant height (Table 4).Water deficit management strategies significantly influenced plant height, leaf area, and the number of leaves, while genotypes significantly affected leaf area and stem diameter (p≤0.05).Rev. Caatinga, Mossoró, v. 36, n. 1, p. 145 -157, jan. -mar., 2023 151 At 75 DAS (final period of the flowering stage), the management strategies with a water deficit in the vegetative and flowering stage, either isolated or combined ( T2-A LA of 12.58,13.89,and 20.22%,respectively,in comparison to plants irrigated with 100% ETr (Figure 5A).It was observed that the reduction of LA is not mainly due to water deficit but also to the senescence and/or abscission of adult leaves (Figure 5C), verified by reductions of 23.95, 15.12, and 22.59% in the number of leaves in plants subjected to the T2, T3, and T4 strategies, respectively, compared to the management without water deficit.Reduction in LA can be considered an important strategy to protect and/or acclimate plants to water deficit, eventually reducing water losses through transpiration and maintaining a high water potential in cells (LIU et al., 2017).
The genotype 'BRS Rubi' obtained the highest mean leaf area (2521.75cm 2 ), regardless of the water deficit management strategy, surpassing the genotype 'BRS Safira' by 12.50% (Figure 5B).The mean stem diameter of the genotype 'BRS Jade' was statistically higher than that of 'BRS Rubi', regardless of the strategies of water deficit management (Figure 5D).Thus, it can be observed that the genotypes 'BRS Rubi' and 'BRS Jade' showed higher growth in the leaf area; another important fact to point out in the leaf area is its direct relationship to the photosynthetic process through the interception of light energy.As the photosynthesis rate depends on leaf area, the faster the plant reaches a greater leaf area index and the longer it remains in photosynthetic activity, the higher the crop yield (SANQUETTA et al., 2014).Table 4. Summary of the analysis of variance for leaf area (LA), number of leaves (NL), stem diameter (SD), and plant height (PH) of the colored-fiber cotton genotypes (G), as a function of different water deficit irrigation management strategies (MS), at 75 days after sowing.Among the cotton genotypes as a function of the different strategies of water deficit management at 75 DAS, compared to plants without stress, it is observed that water stress reduced the values of height by 17.08, 22.16, and 30.79% (BRS Rubi) and by 20.40, 13.66, and 25.39% (BRS Safira) in plants under the strategies A 2 B 1 C 1 (T2), A 1 B 2 C 1 (T3) and A 2 B 2 C 1 (T4), respectively (Figure 6).The maximum reduction in plant height induced by water deficit was observed in 'BRS Rubi', with a decrease of 20.22 cm when water stress was applied successively in the vegetative and flowering stages, although this genotype had the highest height under the 100% ETr condition (Figure 6).When plants of the genotype 'BRS Jade' were subjected to different strategies of water deficit management, these strategies did not cause a significant difference in plant height, which was already observed for the CO 2 assimilation rate, reinforcing the hypothesis of a resumption in plant growth after the application of water stress (Figure 6).
The reduction in growth under water deficit conditions may be associated with reductions in the physiological variables of plants, such as CO 2 assimilation rate, instantaneous water use efficiency and instantaneous carboxylation efficiency of cotton genotypes.In line with these results, previous studies have also reported a significant reduction in the growth and physiological traits of cotton related to water stress (SAHITO et al., 2015;BOZOROV et al., 2018).
The water deficit management strategies and coloredfiber cotton genotypes significantly influenced (p≤0.01)shoot dry mass, seed cotton weight, and total seed weight, in addition to root/shoot ratio was only influenced by the water deficit management strategies (Table 5).Regarding the interaction between the factors, there was no significant effect (p>0.05) for the variables analyzed at 120 DAS (Table 5).Rev. Caatinga, Mossoró, v. 36, n. 1, p. 145 -157, jan. -mar., 2023 153 The water deficit management strategies significantly influenced shoot dry mass (Figure 7A) at 120 DAS, when plants subjected to the strategy A 1 B 1 C 1 (T1) were statistically superior to those cultivated under the strategies A 1 B 2 C 1 (T3), A 2 B 2 C 1 (T4), and A 2 B 1 C 2 (T6), which showed reductions in SDM of 18.60, 25.07, and 15.11%, respectively.Differently from the results observed when water stress was applied according to the strategies A 2 B 1 C 1 (T2), A 1 B 1 C 2 (T5), and A 1 B 2 C 2 (T7), there were no significant differences in SDM during the crop cycle compared to the management with 100% ETr (Figure 7A).The shoot dry mass of plants of the genotype 'BRS Rubi' was statistically higher than that of 'BRS Jade', regardless of water deficit management strategies (Figure 7B).The decrease in shoot dry mass production of cotton plants under water deficit occurs due to the adaptive physiological mechanisms developed by plants, such as the acceleration of leaf senescence, so the restrictions observed in the phenological stages (T3, T4, and T6) of cotton are correlated with the reductions in leaf area and, on the other hand, to the surface exposed to water losses through transpiration (ANDRADE; ABREU, 2007).For the root/shoot ratio (R/S), there were increments of 27.36 and 23.55% in plants subjected to the water deficit management strategies A 1 B 2 C 1 (T3) and A 1 B 2 C 2 (T4) compared to those cultivated under full irrigation throughout the cycle, respectively (Figure 7C).It is important to highlight that there were no significant differences among the T1, T2, T5, T6, and T7 treatments.Water deficit in the successive vegetative and flowering stages and the flowering stage reduced shoot dry mass accumulation and consequently increased the root/shoot ratio of cotton plants.Cotton response to water stress involves osmotic adjustment, the elasticity of the photochemical apparatus, and control of stomatal conductance; on the other hand, the nature of root growth and development is related to a greater uptake of water and nutrients, and the synthesis of plant hormones, organic acids, and amino acids (RAMAMOORTHY et al., 2017).
Regarding seed cotton weight -SCW (Figure 8A), plants under a deficit of 40% ETr in the vegetative stage (A 2 B 1 C 1 -T2) and yield formation stage (A 1 B 1 C 2 -T5) showed mean production of 154.46 and 144.46 g per plant, respectively, standing out as an indication of acclimatization to water stress, and the SCW values obtained were compatible with those of plants without water stress (158.69 g per plant).Plants irrigated with 40% ETr in the same stages also obtained the highest total seed weight (TSW), not differing from plants irrigated with 100% ETr throughout the cycle, with TSW of 104.65 and 94.78 g per plant in the strategies A 2 B 1 C 1 (T2) and A 1 B 1 C 2 (T5), respectively (Figure 8C).This may be related to the positive regulatory mechanism, which can compensate for the effect of water stress through the action of antioxidant enzymes, synthesis of organic solutes, and greater growth of the root system (NIU et al., 2018).Rev. Caatinga, Mossoró, v. 36, n. 1, p. 145 -157, jan. -mar., 2023 155 The results obtained for seed cotton weight and total seed weight are consequences of other variables already discussed, such as gas exchange and biomass accumulation, denoting greater resistance to water deficit during the vegetative and yield formation stages of cotton.Zonta et al. (2017), while evaluating cotton cultivars subjected to water deficit in different stages of the cycle, also found that water deficit imposed during the initial growth stages and after the opening of the first boll does not compromise cotton yield.According to Kumar;Sharma;Kumar (2008), tolerant plants usually stand out with higher dry mass accumulation and/or yield under water deficit than under other environmental conditions.For seed cotton weight and total seed weight as a function of colored-fiber cotton genotypes (Figures 8B and  D), it was verified that 'BRS Jade' stood out with the highest production, reaching values of 166.03 and 106.16 g per plant, respectively, which are 37.21 and 17.81% higher than the values of 'BRS Rubi' and 28.88 and 10.80% higher than the values of 'BRS Safira', respectively.It was found that the changes in gas exchange, explicitly related to the greater growth in stem diameter, contributed to higher SCW and TSW in 'BRS Jade', which is justified by the inherent genetic constitution of each genotype.According to information from Embrapa (2006), there are colored-fiber cotton genotypes with peculiar agronomic characteristics, which make them important for breeding programs.However, the genetic bases for the tolerance of crops is not fully understood due to the complexity of stress conditions and are influenced by multiple genes and environmental factors with small and varied effects (ABDELRAHEEM et al., 2019).

1Figure 1 .
Figure 1.Climatic data of maximum and minimum temperature (ºC), rainfall (mm), and relative humidity of air (%) during the experimental period.
Rev. Caatinga,Mossoró, v. 36, n. 1, p. 145 -157, jan.-mar., 2023  149    vegetative and flowering stages (T4) (Figure2A).The stomatal conductance in the different phenological stages, regardless of genotypes, was classified as follows: T1(A 1 B 1 C 1 )>T2(A 2 B 1 C 1 ) >T4(A 1 B 2 C 1 )>T3(A 2 B 2 C 1 ).Compared to T1, the mean values of gs in the T2, T4, and T3 strategies decreased by 4.51, 27.06, and 50.37%, respectively.Differences between genotypes were observed in gs when the plants were subjected to water deficit in the vegetative/ flowering (A 2 B 2 C 1 ) and flowering (A 1 B 2 C 1 ) stages, with reductions in gs of 38.30 and 37.72% in 'BRS Safira', when compared to the genotypes 'BRS Rubi' and 'BRS Jade', respectively.In general, plants under low water availability in the soil close their stomata to reduce water loss and to maintain turgor in their tissues (HAWORTH et al., 2018; QI; TORII, 2018), being dependent on the degree of water saturation of stomatal cells, with possible inhibition in transpiration flow if water deficit is high in the plant (SUASSUNA et al., 2014).In addition, stomatal closure causes restriction of normal CO 2 flow towards the carboxylation site (LIMA et al., 2020). 1 Means followed by the same lowercase letter indicate there is no significant difference between water deficit strategies (Scott-Knott, p > 0.05) for the same genotype, and means followed by the same uppercase letter for the same strategy indicate that the genotypes do not differ by Tukey test, p > 0.05.Vertical bars represent the standard error of the mean (n = 3).Water deficit management strategies 1 -plants under full irrigation throughout the cycle (1-120 days after sowing -DAS); 2 -water stress in the vegetative stage (20 -60 DAS); 3 -water stress in the flowering stage (61 -75 DAS); 4 -water stress in the vegetative and flowering stages (20 -60 and 61 -75 DAS).

Figure 2 .
Figure2.Follow-up of the interaction between genotypes and water deficit management strategies for stomatal conductance -gs (A) and transpiration -E (B), of colored-fiber cotton genotypes, at 75 days after sowing.
L. A. A. SOARES et al.

1
Means followed by the same lowercase letter indicate there is no significant difference between water deficit strategies (Scott-Knott, p > 0.05) for the same genotype, and means followed by the same uppercase letter for the same strategy indicate that the genotypes do not differ by Tukey test, p > 0.05.Bars represent the standard error of the mean (n = 3).Water deficit management strategies 1-plants under full irrigation throughout the cycle (1-120 days after sowing -DAS); 2-water stress in the vegetative stage (20 -60 DAS); 3-water stress in the flowering stage (61 -75 DAS); 4-water stress in the vegetative and flowering stages (20 -60/ 61 -75 DAS).

Figure 3 .
Figure3.Follow-up of the interaction between genotypes and water deficit management strategies for internal CO 2 concentration -Ci (A) and CO 2 assimilation rate -A (B), of colored-fiber cotton genotypes, at 75 days after sowing.

1
Means followed by the same lowercase letter indicate there is no significant difference between water deficit strategies (Scott-Knott, p > 0.05) for the same genotype, and means followed by the same uppercase letter for the same strategy indicate that the genotypes do not differ by Tukey test, p > 0.05.Bars represent the standard error of the mean (n = 3).Water deficit management strategies 1-plants under full irrigation throughout the cycle (1-120 days after sowing -DAS); 2-water stress in the vegetative stage (20 -60 DAS); 3-water stress in the flowering stage (61 -75 DAS); 4-water stress in the vegetative and flowering stages (20 -60/ 61 -75 DAS).

Figure 4 .
Figure 4. Follow-up of the interaction between genotypes and water deficit management strategies for the instantaneous water use efficiency (WUEi) (A) and instantaneous carboxylation efficiency (CEi) (B), respectively, of colored-fiber cotton genotypes, at 75 days after sowing.

1
Means followed by the same lowercase letter indicate there is no significant difference between water deficit strategies (Scott-Knott, p > 0.05), and means followed by the same uppercase letter indicate that the genotypes do not differ from each other by Tukey test, p > 0.05.Bars represent the standard error of the mean (n = 3 for strategies and n = 3 for genotypes).Water deficit management strategies 1-plants under full irrigation throughout the cycle (1-120 days after sowing -DAS); 2-water stress in the vegetative stage (20 -60 DAS); 3-water stress in the flowering stage (61 -75 DAS); 4-water stress in the vegetative and flowering stages (20 -60/ 61 -75 DAS).

Figure 5 .
Figure 5. Leaf area and number of leaves per plant as a function of water deficit management strategies (A and C) and leaf area per plant and stem diameter of colored-fiber cotton genotypes (B and D), respectively, at 75 days after sowing.

1
Means followed by the same lowercase letter indicate there is no significant difference between water deficit strategies (Scott-Knott, p > 0.05) for the same genotype, and means followed by the same uppercase letter for the same strategy indicate that the genotypes do not differ by Tukey test, p > 0.05.Bars represent the standard error of the mean (n = 3).Water deficit management strategies 1-plants under full irrigation throughout the cycle (1-120 days after sowing -DAS); 2 -water stress in the vegetative stage (20 -60 DAS); 3-water stress in the flowering stage (61 -75 DAS); 4-water stress in the vegetative and flowering stages (20 -60/ 61 -75 DAS).

Figure 6 .
Figure6.Follow-up of the interaction between genotypes and water deficit management strategies for the plant height of colored-fiber cotton genotypes, at 75 days after sowing.

1
Means followed by the same lowercase letter indicate there is no significant difference between water deficit strategies (Scott-Knott, p > 0.05), and means followed by the same uppercase letter indicate that the genotypes do not differ from each other by Tukey test, p > 0.05.Bars represent the standard error of the mean (n = 3 for strategies and n = 3 for genotypes).Water deficit management strategies 1-plants under full irrigation throughout the cycle (1-120 days after sowing -DAS); 2-water stress in the vegetative stage (20 -60 DAS); 3-water stress in the flowering stage (61 -75 DAS); 4-water stress in the vegetative and flowering stages (20 -60/ 61 -75 DAS); 5 -water stress in the yield formation stage (76 -120 DAS); 6-water stress in the vegetative and yield formation stages (20 -60/76 -120 DAS); 7-water stress in the flowering and yield formation stages (61 -75/76 -120 DAS).

Figure 7 .
Figure 7. Shoot dry mass (SDM) -A and root/shoot ratio (R/S) -C of colored-fiber cotton genotypes as a function of water deficit management strategies and SDM as a function of genotypes (B), at 120 days after sowing.

1
Means followed by the same lowercase letter indicate there is no significant difference between water deficit strategies (Scott-Knott, p > 0.05), and means followed by the same uppercase letter indicate that the genotypes do not differ from each other by Tukey test, p > 0.05.Bars represent the standard error of the mean (n = 3 for strategies and n = 3 for genotypes).Water deficit management strategies 1-plants under full irrigation throughout the cycle (1-120 days after sowing -DAS); 2-water stress in the vegetative stage (20 -60 DAS); 3-water stress in the flowering stage (61 -75 DAS); 4-water stress in the vegetative and flowering stages (20 -60/ 61 -75 DAS); 5 -water stress in the yield formation stage (76 -120 DAS); 6-water stress in the vegetative and yield formation stages (20 -60/76 -120 DAS); 7-water stress in the flowering and yield formation stages (61 -75/76 -120 DAS).

Figure 8 .
Figure 8. Seed cotton weight (SCW) and total seed weight (TSW) as a function of water deficit management strategies (A and C) and coloredfiber cotton genotypes (B and D), respectively, at 120 days after sowing.
: not significant and significant at p ≤ 0.01 and p ≤ 0.05; respectively, by the 'F' test.# Up to the date of measurement (75 DAS) only 4 out of the 7 management strategies had been applied to water deficit.
* L. A. A. SOARES et al.

Table 5 .
Summary of the analysis of variance for shoot dry mass (SDM), root/shoot ratio (R/S), seed cotton weight (SCW), and total seed weight (TSW) of the colored-fiber cotton genotypes (G), as a function of different water deficit management strategies (MS), at 120 days after sowing.L. A. A. SOARES et al.