Water use efficiency (WUE) and productivity of promising quinoa (Chenopodium quinoa Wild.) genotypes grown under three water regimes

Quinoa is one of the most nutritious grains and currently has attention due to its adaptation to a wide range of environments and abiotic stresses. This study was conducted under field conditions of an arid agroecosystem to evaluate the response physiological and yield of nineteen elite quinoa genotypes grown under three water regimes (95, 65, and 35% Field capacity) using a drip system in sandy soil. The experiment design was a split-plot in a randomized complete block design during the 2019/20 and 2020/21 growing seasons. The results showed significant differences among evaluated genotypes, water treatments, and their interaction. Fluorescence chlorophyll components were sensitive to water stress and strongly decreased at low soil moisture. Fluorescence (Fo) was the most correlated with seed yield and water use efficiency (WUE) under both full irrigation and drought stress. This may be used for improving yield and WUE in breeding programs. The optimum WUE achieved from moderate irrigation (65% FC), indicated the importance of detecting water requirements. The seed production ranges between 3.8 and 2.2 t ha under full irrigation, and it decreases to reach 62.8% under water regimes. The most suitable genotypes for growing under full irrigation were ‘V9’, ‘Apelawa’, ‘30TES’, and ‘27GR’ which produced 3.8, 3.7, 3.5, and 3.5 t ha, respectively. The highest seed yield under stream drought (1.4 t ha) was produced by ‘Ames 10334’ and ‘QU629-99’. However, the genotype ‘Apelawa’ could outperform under different moisture conditions. It produced 1.2 t ha under stream drought thus, recommended to cultivate it, especially in zones where precipitation fluctuates.


Introduction Introduction Introduction Introduction
The global interest in quinoa (Chenopodium quinoa Wild.) is rapidly increasing due to its nutritional qualities and its ability to grow in marginal environments. It could be one of the crops destined to offer food security in the 21 st century (Alandia et al., 2021). Quinoa has been cultivated for millennia under low rainfall conditions and in different agroclimatic zones. It is well adapted to various environments and has physiological 2 and morphological strategies (Bazile, 2016). This wide adaptability is due to the differentiation of a diversity of ecotypes originating in contrasting agro-environments (Zurita-Silva et al., 2015). Furthermore, quinoa can be a good candidate for cropping systems in marginal agricultural areas (Alandia et al., 2021). It is a potential crop in the consideration of future climate change challenges (Zurita-Silva et al., 2015). Quinoa is a very drought-resistant crop (Winkel et al., 2002) as well as it displays various adaptive mechanisms to water regimes, from physiological to morphological adaptations that led to a range of responses to stress; avoidance, resistance, and tolerance. Quinoa copes with water regimes by modifying or changing vital physiological processes, such as photosynthesis, water relations, and hormone metabolism (Zurita-Silva et al., 2015). The grain has high nutritional value due to its high protein content, with all essential amino acids and a wide range of vitamins and minerals. High levels of polyphenols and flavonoids are beneficial for human health (Repo-Carrasco-Valencia et al., 2010). However, it also contains saponins and tannins used in the pharmaceutical and cosmetic industries (Escribano et al., 2017).
Water deficit is one of the most challenging abiotic stress factors in arid and semi-dry regions, furthermore, water demand is increasing with increased population and global temperature (Huang et al., 2015;Altun et al., 2016;Hafez and Seleiman, 2017;Seleiman et al., 2019;Eid et al., 2020;Roy et al., 2021). Changing climate prolongs the duration of water deficiency, causing a reduction in crop production and decreasing food security in the world (Krannich et al., 2015;Ding et al., 2021;Kheir et al., 2021;Battaglia et al., 2022). Agricultural activity is the primary consumer of freshwater; thus, the selection and cultivation of tolerant drought crops is an approach to increase crop production levels under water deficit conditions. Water is a critical component of plant metabolism and yield production (Babur et al., 2021). Select adapted genotypes to stressful edaphic climatic conditions to resist periods of water shortage while maintaining high productivity for each situation is of great importance in breeding programs (Präger et al., 2018). Improving water use efficiency (WUE) should be an essential target in breeding programs, particularly in arid and semi-arid land (Elshayb et al., 2022;Seleiman et al., 2023;Al-Selwey et al., 2023). However, the complexity of this trait and the difficulty of its phenotyping has prevented many breeders from attempting to select WUE directly (Araus et al., 2002). WUE is a measure of the plant biomass produced per water used unit; it is defined in physiological concept, by Hatfield and Dold (2019) as the amount of carbon assimilated as biomass or grain produced per unit of water used by the crop; thus, the response of WUE is directly related to the physiological processes can control the gradients of CO2 and H2O. Basso and Ritchie (2018) illustrated that the productivity of maize grain could be increased with no difference in water use due to increased WUE.  found that the highest values of WUE, grain yield, and dry matter were in quinoa genotypes under moderate water irrigation. Also, Geerts et al. (2008) suggested that deficit irrigation can contribute to better agricultural planning due to better control of the phrenological development of quinoa and did not cause lower yields and resulted in equal or higher WUE than the wet fields. Analysis of chlorophyll fluorescence is appreciated as a widely used technique in researching photosynthetic efficiency in plants (Genty et al., 1989). PSII (Fv/Fm) parameter is the most often employed parameter for maximizing the quantum yield (Maxwell and Johnson, 2000). Fghire et al. (2015) suggested the use PSII to assess water stress among quinoa genotypes; however, they mentioned that the Pindex (Performance index) was more sensitive to water stress. Pindex was used to quantify the effects of environmental factors on photosynthesis in several studies (Tsimilli-Michael and Strasser, 2008).
The objective of this study was to identify the adapted and highly productive quinoa genotypes grown under different water regimes (95, 65 and 35%) using a drip system in sandy soil as well as to assess the relationship between the yield and physiological traits.

Materials, treatments, and experimental design
The field experiments were carried out at Education Agricultural Farm, King Saud University (KSU) (24 43' N, 46 37' E), Al Riyadh, Saudi Arabia, during the winter seasons of 2019/20 and 2020/21. The experiment field received sand for 0.4 m depth while irrigation was by dripper web and controlled by ECH2O soil moisture sensor and date recorded by ECHO Em50 (Decagon Devices, Inc.). The weather of the Al Riyadh area corresponds to thermal mild, with a dry sub-humid humidity regime, the average maximum temperature during January was 19.9 and 21.6 °C, and the average minimum temperature was 8.3 and 9.5 °C, in both seasons, respectively. There were dry weather conditions in the field confirmed by the cumulative precipitation curve (Figure 1). The experiment field received only 15.4 and 14.2 mm of precipitation in the first and second seasons, respectively. Nineteen quinoa genotypes were selected from the quinoa breeding program of KSU (EL-Harty et al., 2021). The name and origin of these genotypes are presented in Table 1.   The field experiments were split plotted in a randomized complete block design with three replications. Three water treatments in the main plot corresponded to the level of available water (AW) in the root zone at 95%, 65%, and 35% of field capacity. The drip irrigation system was used and treatments were applied after the emergence of the quinoa plants. Each lateral drip line was 30.0 m long at a distance of 0.5 m from adjacent lines with a 0.10 m emitter spacing. The drippers had a discharge rate of 4 L h −1 . The nineteen quinoa genotypes were randomly assigned to the sub-plots and the experiment plot contents were 3 m long and 3 rows (plot size = 4.5 m 2 ). Quinoa seeds were sown alongside the drippers at the end of November in both seasons. Phosphorus [100 kg ha −1 in the form of single furrow-banded calcium superphosphate (15.5% P2O5)] was applied basally during the soil preparation for planting. Nitrogen [160 kg ha −1 as ammonium nitrate (33.5% N)] and potassium [50 kg ha −1 as potassium sulfate (50.0% K2O)] fertilizers were applied through the fertigation system in five equal doses, respectively, from the time of planting to 70 days after planting.

Measurements
The behavior and performance of the photosynthetic apparatus at quinoa leaf in the second season were determined by recording the chlorophyll fluorescence using Hansatech Handy PEA (Hansatech Instruments, UK).
A pulse of saturating red light (peak at 650 nm, 3000 mmol m -2 s -1 ) was applied after dark adaptation for 30 min. This leads to an increase in chlorophyll fluorescence increases from minimal fluorescence (Fo), when all reaction centers are open, to maximal fluorescence (Fm), when all reaction centers are closed. The Fm (maximum fluorescence) is measured during the first saturation pulse after the adaptation of the plant to darkness. The maximum quantum yield was obtained (Fv/Fm), where Fv is the difference between the maximum and minimum fluorescence after adapting the plant to darkness.
The Fv/Fm ratio estimates the maximum portion of the absorbed light that is used in photosystem II (PSII) reaction centers. That represents the maximum potential of quantum efficiency of the PSII.
The young leaves were used to determine relative water content (RWC%) where the leaf was weighted (FW) and then floated on distilled water for 4 h at 22 °C under dim light. The turgid weight (TW) was determined after blotting, and the dry weight (DW) was measured after the samples had been dried for 24 h at 80 °C.

RWC=
At the maturing stage, data of the number of days from sowing to 95% maturing, plant height (cm), and number of branches/plants were recorded. All experiment plot plants were harvested together and seed yield per hectare was estimated as the ratio of the plot to the hectare. WUE was calculated as the ratio of seed yield per hectare (kg) to water applied for each treatment as follows: WUE = seed yield(kg)/water applied(m 3 ).
The stress tolerance index (STI) was calculated using the following formula according to Fernandez, 1992: where Ys = grain yield of a test genotype under drought-stressed conditions; Yp = grain yield of a test genotype under non-stressed conditions, and Xp = mean yield of test genotypes under non-stressed conditions.

Statistical analysis
Statistical analysis was done for each season separately and after confirmation of errors homogeneity for the two seasons, then combined analysis of the two seasons was applied. Means of treatments and genotypes were compared by Tukey , s test at a probability level of 5%. with the Turkey test using statistical analysis of MSTATC software. Correlation coefficients were estimated among studies traits according to (Steel et al., 1997).

Physiological traits
The fluorescence parameters provide information about a plant's ability to tolerate stress and how this could damage the photosynthetic apparatus (Maxwell and Johnson, 2000). Fluorescence chlorophyll components were sensitive and decreased with increased water regimes; the percentage of decline was 10.7, for Fo, 6.5, for Fm, and 3.7, Fv/Fm while Pindex was the highly sensitive parameter to water regimes with 35.4% decline ( Table 2). All genotypes had the same trend except for a few genotypes that increased under T2 and T3; however, these differences were insignificant in most cases. Furthermore, concerning the Fo parameter, the best five ranks under fully irrigated treatment (T1) and the average of the three treatments were occupied by the same genotypes ('94R', '27GR', '54ALC', 'Chadmo', and 'V9') with different ranks between them. The results indicated that these genotypes are less affected by a shortage of irrigation water. Krause and Weis (1991) concluded that the quinoa plant could obtain stable parameters under a lack of water conditions. Values of maximal fluorescence (Fm) were high with a wide range (1839 for 'QU629-99' to 1074 for 'Q 29' under T1). The evaluated genotypes had changed their superior due to changes in available moisture; genotypes 'QU629-99', 'Kaslaea', and 'Pison' were higher under T1, while genotypes '94R', '30TES', and 'Kaslaea' recorded the highest values under T3. However, 'Kaslaea' and 'QU629-99' resaved their superior and took high rank at an average of all treatments. In contrast, a narrow range for Fv/Fm ranged from 0.773 for 'V9' to 0.547 for 'QU629-99' under T1, with an average of 0.700; this range became wider under water regimes; from 0.757 for 'Tundri' to 0.463 for '20TES' under T3. However, drought treatments (T2 and T3) had a more negligible effect on the average of Fv/Fm with a decrease of 5.2 and 3.7%, respectively, compared with full irrigation treatment. Winkel et al., 2002 found that ontogenic variation was moderate to high for physiological parameters except for Fv/Fm. It may be due to Fv/Fm only reflecting changes in the values of Fo and Fm (Fghire et al., 2015). These results are similar to Yang et al. (2016) on the quinoa variety 'Titicaca' which reduced FV/FM when plants were exposed to drought. Valdivia-Cea et al. (2021) presented differences among irrigation treatments for Fv/Fm values from 0.74 to 0.82. The Pindex was used to quantify the effects of environmental factors on photosynthesis in several studies (Guan et al., 2022)). Pindex values had a wide range recorded among genotypes under irrigated treatments. The quinoa genotypes 'NSSL 86628' and '30TES' had the highest values under T1 and T3 combined with genotypes 'V9' under T1 and 'Tundri' under T3. Fghire et al. (2015) explained that the Pindex was much more sensitive to water regimes, maybe due to its response to changes in the fast fluorescence rise kinetics among the two fluorescence extremes (Fo and Fm).
These results, in agreement with Fghire et al. (2015) found that water regimes induced a significant decrease in Fv/Fm and Pindex. Physiological analyses revealed different responses of the genotypes to irrigation treatments, together with symptoms of water regimes; Da-Silva et al. (2021) reinforce the importance of detailed studies of the relationship between productivity, physiology, and water use. While Winkel et al. (2002) suggested using these parameters in assessing water stress in quinoa; contrast, Hamli et al. (2015) showed differences in the rank of durum genotypes according to these physiological traits; and proposed that these physiological parameters could not be recommended to replace grain yield field testing. Also, Valdivia-Cea et al. (2021) found that the environments play a main role in physiological parameters and estimated no significant differences among quinoa genotypes in season and other seasons. Also, Killi and Haworth (2017) reported that the indicators of the photosynthetic capacity were affected by water stress but retained photosynthetic capacity and photosystem II performance. El-Harty EH et al. (2023). Not Bot Horti Agrobo 51 (2) T1  T1 T1  T1  T2  T2 T2  T2  T3  T3 T3  T3  Mean  Mean Mean  Mean  T1  T1 T1  T1  T2  T2 T2  T2  T3  T3 T3  T3  Mean  Mean Mean  Mean  T1  T1 T1  T1  T2  T2 T2  T2  T3  T3 T3  T3  Mean  Mean Mean  Mean  T1  T1 T1  T1  T2  T2 T2  T2  T3  T3 T3

Mean performance; seed yield components
Based on the combined analysis of variance results, there were statistically significant differences among genotypes, irrigation treatments, and seasons and their interactions for all traits under the study. Seed yield and its components were decreased with decreased available irrigation water; seed yield was the most affected with 24.2 and 62.1% reduction under T2 and T3. The evaluated genotypes matured after 113.5 days; the earliest genotype was 'Q12' (102.7 days) and the latest one was 'QU629-99' (126.7 days) under full irrigation treatment (T1); however, both genotypes were among the earliest genotypes under water regimes, with mean values of 94.0 and 92.0 days, respectively (Table 3).
The tallest genotypes under T1 were '27GR' (160.0 cm), and 'Apelawa' (150 cm) had the highest depression under extreme drought (T3) with 53.1 and 55.0%. The average plant height was 128.3 cm under full irrigation treatment and reduced to 105.2 and 83.0 under T2 and T3. All genotypes had the same depression except 'Kaslaea', and 'NSSL 86628' had insignificant change under T2. High variation was shown among quinoa genotypes for seed yield and its components under Saudi Arabia conditions (EL-Harty et al., 2021). Lin and Chao (2021) found that the plant height of all quinoa genotypes was affected under extreme drought, but under mild water regimes, the plant height of tolerant genotypes did not change compared with full irrigation treatment. The average number of branches per plant was high (11.3) under T1 and decreased by 22.1 and 48.8% due to water stress treatments.  T1  T1 T1  T1  T2  T2 T2  T2  T3  T3 T3  T3  Mean  Mean Mean  Mean  T1  T1 T1  T1  T2  T2 T2  T2  T3  T3 T3  T3  Mean  Mean Mean  Mean  T1  T1 T1  T1  T2  T2 T2  T2  T3  T3 T3 Table 4 presents the RWC, WUE, SY, and STI of the evaluated genotypes under the three irrigation treatments. The highest RWC was estimated in the genotypes 'V29', 'V9', 'Ames 10334', and 'Chadmo' which maintain their superiority during all conditions, in contrast to other genotypes that had negative responses to drought. This result, in agreement with Jensen et al. (2000) estimated higher RWC under fully watered conditions, then decreased during severe soil drying. González et al. (2009) reported that high RWC under fully watered treatment was not significantly changed than under drought conditions.
The average of WUE under full irrigation treatments (T1) was 6.7 and reduced to 5.3 kg per m 3 due to water regimes (T3); however, most of the evaluated quinoa genotypes showed more water use efficiency under moderate irrigation conditions (T2) with average 7.0 kg per m 3 . The highest values of WUE under full irrigation were 8.3, 8.1, and 7.9 in the genotypes 'V9', 'Apelawa', and '27GR', while these genotypes decreased their efficiency under T2; the Bolivian genotype 'Apelawa' increased its efficiency to 8.4 and occupied the first rank with the genotypes '94R' (8.3), 'Chadmo' (7.8) this genotype had not deserved under both T1 and T3 treatments emphasizing that it is distinguished only under these while the genotypes 'Apelawa' and '27GR' were distinguished under both T1 and T2 the genotype '94R' was among the best genotypes under both T2 and T3 treatments. These genotypes recorded the highest average of WUE for the three treatments besides 'Ames 10334' and 'QU629-99' which had the highest WUE under extreme drought conditions. These indicated water use efficiency (WUE), and selected particular genotypes for each environment should be an essential target in breeding programs and importance increases in an arid and semi-arid land. Select adapted genotypes to stressful edaphic climatic conditions to resist periods of water shortage while maintaining high productivity for each situation is of great importance in breeding programs (Präger et al., 2018). The results of WUE were lower than the results of Da-Silva et al. (2021) found that quinoa WUE reaches up to 9.7 kg per m 3 for high production genotype under optimum conditions. But it is higher than irrigation WUE and total water (irrigation+ rainfall) WUE estimates, 3.6 and 2.3 kg per m 3 , respectively presented by ; but their low WUE may be due to the low production of their genotypes. The optimum WUE in bread wheat; was 12.4 kg per m 3 (Verma et al., 2021); but higher than the estimated WUE in chickpeas and lentils under Mediterranean environments; 3.4 and 3.2 kg per m 3 , respectively (Zhang et al., 2000). The low WUE in this study may be due to being grown in sandy soil without organic matter. These results, similar to the results of Naggar et al. (2017) showed increased water use efficiency (WUE) and reduction in quinoa yield under mild moisture.
The seed production of evaluated genotypes ranges between 3.8 and 2.2 t ha -1 for both US improved genotypes ('V9' and '20TES', respectively) under full irrigation, and it decreases to reach 62.8% under water regimes (T3) with a range between 1.3 to 0.6 t ha -1 . The highest seed yield was obtained by 'V9' (3.8 t ha -1 ), 'Apelawa' (3.7 t ha -1 ), '27GR' (3.5 t ha -1 ), '30TES' (3.5 t ha -1 ), and '94R' (3.4 t ha -1 ) under fully irrigation treatment. However, the reduction in their production was dramatically under both moderate (T2) and drought (T3) treatments except 'Apelawa' and '94R' which reduced only 24 & 68% and 21 & 62% compeering with the control treatment. The lowest reduction (9% under T2) was estimated in '20TES' and 'SU-5' which came at the bottom of the list of productivity under optimal conditions (T1). Nine genotypes occupied high rank under water regimes i.e., 'QU629-99', 'Ames 10334', '94R', '54ALC', 'Apelawa', 'Q12', 'Q18', 'V9', and 'SU-5' whit  under full irrigation and decreased combined with water irrigation rate. Stress tolerance index (STI) and the average seed yield over the three irrigation treatments indicated the superiority of three genotypes, i.e., '94R', 'Apelawa', and 'Ames 10334', with seed yield 2.5, 2.5, and 2.4 t ha -1 and STI was 0.5 for the three genotypes. Production of quinoa under full irrigation in Tunisia was between 2.1 and 0.3 t ha -1 . The variability in quinoa production in Egypt was significant and between 2.7 and 1.7 t ha -1 among early maturing genotypes (Naggar et al., 2017). 9 3 ) ) ) ) Seed yield (t) Seed yield (t) Seed yield (t) Seed yield (t)  STI  STI STI  STI  T1  T1 T1  T1  T2  T2 T2  T2  T3  T3 T3  T3  Mean  Mean Mean  Mean  T1  T1 T1  T1  T2  T2 T2  T2  T3  T3 T3  T3  Mean  Mean Mean  Mean  T1  T1 T1  T1  T2  T2 T2  T2  T3  T3 T3  A highly positive significant correlation between seed yield and all of the no. of branches, STI, and WUE under both full irrigation and drought (T1 and T3) conditions was observed, in addition to plant height under T3 only (Table 5). Among the physiological parameters, only Fo positively correlated with no. of branches, seed yield, STI, and WUE under full and extreme drought treatment. Also, Pindex was significantly correlated with seed yield and WUE under normal conditions (0.54 and 0.51, respectively). Physiological traits associated with WUE and/or seed yield could be used in quinoa breeding programs, particularly under drought conditions. This finding is similar to the results of other authors: Saddiq et al. (2021); Akram et al. (2021) and Valdivia-Cea et al. (2021) who documented a direct relationship between seed yield and WUE and leaf water content.  recorded a correlation between yield and WUE under three water stress.

Conclusions Conclusions Conclusions
Physiological parameters were decreased with increased water regimes, but Pindex was the most responsible. Fo was the most related to seed yield and showed a positively correlation with no. of branches, seed yield, STI, and WUE under full irrigation (T1) and under extreme drought treatment (T3). Also, Pindex showed a positive correlation with seed yield and WUE under normal conditions. These traits could be used as selection criteria in breeding programs to select high-seed yield genotypes under different water regimes. Mild drought (65%FC) increased WUE; indicted the importance of assessing the water requirements of each quinoa genotype. Quinoa can grow under drought-stress conditions with an acceptable seed yield. The stress tolerance index (STI) and the average seed yield of the three irrigation treatments indicated the superiority of genotypes '94R', 'Apelawa', and 'Ames 10334', with mean values of 2.5, 2.5, and 2.4 t ha -1 . These genotypes had high water use efficiency. However, seed yield potential was high in the case of full irrigation, with mean values of 3.8, 3.7, 3.5, and 3.5 t ha -1 for the genotypes 'V9', 'Apelawa', '27GR', and '30TES'. The genotype 'Apelawa' was among the best genotypes under each water treatment and due to its high WUE under all water regimes, it was recommended to be released as a new cultivar.