Morphological, physiological, and biochemical responses of onion (Allium cepa L.) breeding lines to single and combined salt and drought stresses

Abiotic stresses deteriorate plant growth resulting in devastating yield losses. Salt stress solely cause ionic toxicity and disturbed homeostasis, whereas combined salt and drought stress has more pronounced effects on plants. This study aimed to screen 32 Turkish onion breeding lines and commercial cultivars based on their morpho-physiological and biochemical responses after exposure to drought, salt, and salt + drought stresses at the bulbification stage under greenhouse conditions. Physiological characteristics, such as gaseous exchange traits, chlorophyll index, leaf temperature, and morphological traits including the number of leaves, length, and diameter of leaf were measured during stress treatment, and yield response of the onions was quantified by measuring bulb length, bulb diameter, and bulb weight after harvest. Proline and malondialdehyde accumulation was estimated for the biochemical effect of stress on onion genotypes. All genotypes responded differentially to the applied single and combined stresses. Overall results revealed that in the breeding lines K25, U6, U17 and commercial cultivar K58, the bulb weight (41.71–47.93 g) was significantly (p ≤ 0.05) lower, therefore they were grouped as sensitive across all stresses; whereas in the breeding lines K41, U47, U49 and commercial cultivar K52, the bulb weight (96.75–106.31 g) was significantly (p ≤ 0.05) higher among all the tested breeding lines and commercial cultivars and therefore found to be the most stable upon stress. These resilient genotypes can be used as breeding material for future abiotic stress studies.


Introduction
Onion is a cross-pollinated, biennial crop that belongs to the Alliaceae family (Gökçe 2018;Ghodke et al. 2020). It is a robust crop that can grow under a wide range of climatic conditions, but its full potential can be observed under mild climatic conditions (Mubarak and Hamdan 2018). This bulbous vegetable is grown and consumed globally (Havey and Ghavami 2018). Depending on the area of cultivation, onion is considered as a biannual or perennial vegetable. It was originally domesticated from the region including South−Western Pakistan, Iran, and Afghanistan (Brewster 1994). Among vegetables, onion ranks to screen short-day and long-day onion breeding lines. According to our knowledge, this is the first study evaluating the response of onion to combined salt + drought stress. It was hypothesized that based on the results obtained 32 short-and long-day onion breeding lines all the observed physio-biochemical attributes can be used as a criterion for the rapid selection with a higher tolerance to applied stresses.

Stress treatment
Salt stress treatment was initiated at onion bulb formation stage and carried out as described in our earlier study Chaudhry et al. (2020). In short, salt stress was applied with a 3-day interval with increasing concentrations of 75 mM, 100 mM, 125 mM, and three times 150 mM saline water (NaCl) in a total of 6 watering to the soil saturation capacity in a 20-day period. Combined salt and drought stress was carried out by applying limited irrigation (70% of the soil saturation capacity) with the same salt concentrations used in salt stress application with 6 watering in 20 days. The drought stress application was carried out by stopping irrigation in onion plants belonging to the drought stress group at bulb formation stage for 20 days (Supplementary Fig. 1). After completion of stress treatments, drought-stressed plants were wellwatered, and salinity-stressed plants were washed with normal irrigation water to drain excessive salt. Afterward, plants were allowed to grow until they showed bending of the neck for a harvest of onion bulbs with regular watering to soil saturation capacity. Before harvesting, leaf samples were collected from 3rd or 4th fresh leaves in 3 replications for biochemical analyses, directly frozen in liquid nitrogen and stored − 20 °C.

Physiological measurements
Physiological traits were measured from all four treatment groups (control, salt stress, salt + drought stress, and drought stress) by selecting 3 random plants from each group. Measurements were taken in triplicates from the 3rd or 4th leaf of the plants which were properly exposed to the light.

Gaseous exchange traits
Photosynthetic rate, transpiration rate, and stomatal conductance measurements from all treatment groups were taken by constant light intensity of photosynthesis device (1500 μmol m −2 s −1 ), CO 2 amount (400 μmol) and airflow (500 μmol s −1 ) and, when photosynthesis gained steady state, the measurement was recorded using a portable gas exchange system LiCor 6400 XT (Li-Cor Biosciences, USA). All measurements were taken in 3 replicates and mean values were used for data analysis.

Chlorophyll index
Leaf chlorophyll index was measured in 3 replicates by using SPAD 502 Chlorophyll-Meter (Soil Plant Analysis Development; Minolta, Japan).

Leaf temperature
Leaf temperature was also measured in 3 replicates from each pot by using Infrared Thermometer (IRT) device (MASTECH BM380).

Morphological and yield traits
Measurements were performed by selecting 3 plants from each pot. Length of the leaf (cm) and diameter of a leaf (cm) were measured by measuring tape, whereas number of leaves was counted from each pot. Onion bulbs were harvested to measure bulb diameter (cm), bulb length (cm), and weight of the bulbs (g) was measured by precise weight balance. The leafy green part of onion plants was collected for fresh weight determination. Accordingly, after 24 h of drying at 180 °C, samples were weighed again, and the biomass change was calculated by measuring the difference between the fresh and dry weight of the leaves.

Biochemical measurements
Proline content was measured on the last day of stress treatments by following the procedure of Bates et al. (1973) with minor modifications. Accordingly, 100 mg leaf sample was ground in 1 ml 75% ethanol at room temperature overnight. A 100 µl of the upper phase of the sample was collected after centrifugation at maximum speed for 20 min, then transferred to a new tube, and 900 µl of fresh ninhydrin solution was added (1% ninhydrin, 60% acetic acid, 40% pure water). Samples were then kept at 100 °C for 1 h and immediately cooled at room temperature. After the addition of 3 ml of toluene, they were again kept at room temperature for 24 h. The absorbance of the resulting solution was then measured by spectroscopy at 520 nm and the amount of proline in the sample was determined with the help of the standard curve prepared using pure proline.
Malondialdehyde (MDA) contents were measured as suggested by Heath and Packer (1968). The leaf sample (300 mg) was collected and ground in 0.1% 3 ml (w/v) trichloroacetic acid (TCA) solution. Then this sample was centrifuged at 10,000 g for 10 min. 1.5 ml of the obtained upper phase was collected and 1.5 ml of 20% (w/v) TCA solution containing 0.5% (w/v) TBA was added. After 29 Page 4 of 12 Vol:. (1234567890) the mixture was incubated at 90 °C for 20 min, the tubes were kept on ice to stop the reaction. Samples were then centrifuged at 10,000 g for 5 min. The absorbance of the resultant product was measured at 532 nm and 600 nm. The amount of MDA was calculated using the coefficient value of 155 mM −1 cm −1 . In the absorbance measurement, 20% TCA solution containing 0.5% TBA was used.

Statistical analyses
Treatment data set comprised of stress environments for 32 diverse genotypes analyzed by Analysis of Variance (ANOVA). The least significant difference (LSD) was used at p ≤ 0.05 for the comparison of treatment means by Statistical package Statistix 8.1 (Tallahassee Florida, USA) (Steel et al. 1997). Pearson correlation and principal components analysis was done by using Origin 2020. The mean bulb weight of the onion was subjected to analyze the genotype by environment interaction. Heritability (H2) was estimated by using the equation H2 = VG/VP, where VG is genotypic variance and VP is phenotypic variance (Toker 2004). Stability analysis such as regression coefficient (bi) was done by STABILITYSOFT (Pour-Aboughadareh et al. 2019).

Analysis of variance for the observed traits
Analysis of variance evaluated in response to salt, drought, and combined salt + drought stresses showed highly significant (p ≤ 0.05) differences in the morpho-physiological and biochemical traits (Table 1). Mean performance of all the traits showed significant (p ≤ 0.05) differences that would be helpful for the selection of tolerant and sensitive genotypes. The genotype and its interaction with the environment were also significant that highlighted contrasting behavior of all the tested genotypes under applied stresses. The genotypes subjected to specific stresses showed differential inhibition of growth suggesting that some of the genotypes were sensitive to salt stress, while others showed sensitivity to drought stress conditions.
In response to drought stress, a significant positive correlation was observed between photosynthesis and transpiration rate (r = 0.50), stomatal   (Fig. 1b).

Principal component analysis (PCA)
Principal component analysis was performed for all the observed traits in response to three stress environments. The genotypes were grouped into four different groups, blue ellipse shows a tolerant group of genotypes, green ellipse grouped sensitive genotypes, red ellipse grouped moderate-tolerant genotypes, and black ellipse grouped moderate-sensitive genotypes (Fig. 2). The PCA biplot of the first two components (PC1 and PC2) for the salt stress explained 67.3% variance for the tested traits in 32 onion genotypes (Fig. 2a). The PCA resulted in the distinctive separation of the genotypes based on the performance under salt stress conditions. The genotypes K18, K41, K51, K52, U12, U47, and U49 exhibited tolerance to the salt stress, as they showed higher gaseous exchange traits (photosynthetic rate, stomatal conductance, and transpiration rate), yield attributes (bulb weight, bulb diameter, and bulb length) also aggregated at near to the gaseous exchange traits that suggested their tolerance to salt stress. The genotypes K25, K28, K35, K58, U6, U17, U24, and U33 showed higher leaf temperature that accumulated higher malondialdehyde and proline contents due to oxidative burst forming a group of sensitive genotypes to salt stress (Fig. 2a).
The PCA biplot of the first two components (PC1 and PC2) for the drought stress explained 66.8% variance for the observed traits (Fig. 2b). The genotypes K35, K41, K52, K59, U12, U47, and U49 were tolerant to water deficit conditions as they showed better morphological growth with higher leaf length, number of leaves, and diameter of leaf, additionally they also showed higher gaseous exchange and yield traits. The genotypes K20, K25, U6, U10, U11, U17, and U31 were grouped as sensitive genotypes as they showed higher leaf temperature and differential accumulation of malondialdehyde and proline contents under drought stress conditions (Fig. 2b).
The PCA biplot of the first two components (PC1 and PC2) for the salt + drought stress explained 71.5% variance. The PCA exhibited the close association of different observed traits as shown by the loading byplots (Fig. 2c). The genotypes K35, K37, K39, K41, K52, U12, U47, and U49 formed a group of tolerant genotypes. Moreover, a smaller angle between different morpho-physiological attributes classified the genotypes to be tolerant. These genotypes were also plotted closer along the line direction of the vector. The genotypes K58, U6, U16, U17, U31, U33, and U31 were grouped as a sensitive genotypes with exposure to salt + drought stress conditions (Fig. 2c).

Stability analysis
AMMI biplot showed visual interpretation of interrelationship among 32 genotypes and three environments. Mean bulb weight is plotted against IPC1 (Interactive principal component) as shown in Fig. 3. The displacement along abscissa describes the additive (main) effects, while interactive effects can be explained by displacements along the ordinate. If a genotype or an environment has an IPC1 score of nearly zero, it has small interaction effects and considered stable. All three environments were diverse. Drought stress was observed to be the most stable environment as compared to SS and SD, since its IPC1 score is close to zero, indicating small interaction effects. Genotypes (U49, K41, U47, K52, U12) with high mean bulb weight combined with IPC1 score close to zero clusters on the right quadrat of AMMI biplot (Table 2). These genotypes are considered stable among all the environments.

Broad sense heritability
The bulb weight is the most desired trait for abiotic stress breeding; therefore, it was considered to analyze the broad sense of heritability. Genotypic variance and phenotypic variance were calculated to be 231.6 and 306.7, respectively. Broad sense heritability was found to be 0.75, which means that 75% variability in a trait was caused by genetic differences among genotypes. The trait was less influenced by the environment and allowed the selection of genotypes for higher bulb weight (Table 3).

Discussion
Abiotic stresses are the leading obstacles resulting in retarded growth, development and yield losses of the onion bulb. Currently, very little information is available regarding the behavior of onions in response to salt or drought stress conditions (Chaudhry et al. 2020;Chaudhry et al. 2021a, b;Gedam et al. 2021). To the best of our knowledge, none of these studies has observed the influence of combined salt + drought influence on onion morpho-physiological and biochemical attributes. Therefore, the current study was conducted for the screening of 32 short-day and longday onion genotypes to varying stress conditions. The onion susceptibility to salt, drought, and combined salt + drought stress varied with the intensity of stresses. It was monitored that all the 32 onion genotypes manifested disruptions in observed traits with a grouping of sensitive and tolerant genotypes to stress environments.
All the 32 onion genotypes in this study acclimatized to salt, drought and combined salt + drought stresses showed contrasting behavior. The breeding lines K41, U47, U49 and commercial cultivar K52 showed higher gaseous exchange traits in response to stress conditions that suggested their inherent ability to maintain photosynthetic activity, in accordance with our earlier studies (Chaudhry et al. 2020). The breeding lines K18, K20, K25, and commercial cultivar K51 showed the lowest stomatal conductance, which means they were the most sensitive cultivars. A previous study indicated that with the fluctuation in environmental conditions the stomatal conductance decreases significantly (Sánchez-Virosta et al. 2021). Drought stress negatively influences photosynthesis activity and photosynthetic pigments, which results in decreased photosynthetic rate in plants. In salt and salt + drought stress, chlorophyll content (SPAD) was seen to be positively correlated with photosynthetic rate, stomatal conductance rate and transpiration rate. Higher chlorophyll content facilitates stable photosynthetic activity and enables plants to maintain their requirements under stress conditions; thereby, in this study the breeding lines K41, U12, U47, U49, and commercial cultivar K52 showed better performance in comparison to other genotypes.
Decreased morphological traits are attributed to reduced cell expansion and cell growth which may lead to decreased plant height under stress conditions. Other studies also supported our results with decreased morphological traits, as drought and salt stress negatively affected the plant growth of onion (Hanci and Cebeci 2015). The breeding lines K25, U6, U17, and commercial cultivar K58 showed a decline in morphological growth and yield traits, therefore grouped as sensitive genotypes. Results of this study are in accordance with Naghavi et al. (2013) that also reported decreased morphological growth functioning of maize. A study related to leaf attributes of soybean under drought stress rendered that leaf traits were significantly decreased under water scarcity (Chowdhury et al. 2016;Barrios et al. 2005).
In our study, high variability in the response by different genotypes for leaf temperature, proline, and MDA was observed. The tolerant breeding lines K41, U12, U47, U49 and commercial cultivar K52 performed better under stress conditions as they accumulated the least MDA and proline contents. In contrast, the breeding lines K25, U6, U17, and commercial cultivar K58 were designated as salt-sensitive cultivars as they exhibited higher temperature, MDA, decreased proline contents and poor yield under stress. In our earlier study, we observed similar results where a disruption in membrane stability due to MDA accumulation (Chaudhry et al. 2021a, b). Since proline is a key player in plant tolerance that accumulates to alleviate oxidative stress in plant tissues, higher proline accumulation in these genotypes suggested that they were under a higher influence of oxidative stress.
The bulb characteristics are the most desired attributes for screening of onion genotypes. Differential response of onion genotypes was observed in   (Chaudhry et al. 2020). Additionally, a recent study conducted by Gedam et al. (2021) also screened the onion genotypes based on yield reduction. In this study, the reduction in bulb characteristics is attributed to a decrease in gaseous exchange traits with damage to chlorophyll contents, which resulted in an oxidative burst. As sensitive genotypes excelled in accumulating MDA and proline contents that suggested these genotypes were under the higher influence of the applied stresses. The results of earlier studies showed a reduction in onion bulb yield due to restrained uptake of nutrients, likely reduction in photosynthetic rate with the closure of stomata that supports the results of this study (Chaudhry et al. 2020;Zheng et al. 2013;Wakchaure et al. 2018).
Correlation analysis further strengthened that tested traits resulted in a higher bulb weight of the onion. All the morpho-physiological traits except leaf temperature were significantly positively correlated with the bulb weight. Contrarily biochemical traits showed a strong negative correlation under stress regimes (Fig. 1). Therefore, the genotypes that exhibited improved gaseous exchange traits, less damage to chlorophyll contents, a greater number of leaves, and the least stunted growth were tolerant. The genotypes that showed higher membrane damage (MDA) and greater proline contents, on the other hand, were sensitive with a weak tolerance mechanisms to stress conditions. The PCA analysis revealed the grouping of the onion genotypes as tolerant, sensitive, moderatetolerant, and moderate-sensitive (Fig. 2). The biplot clearly grouped the onion genotypes into tolerant and sensitive genotypes based on their performance in response to salt, drought, and salt + drought stress conditions. The morpho-physiological traits accumulated in positive quadrat and close to each other that supported that tolerance behavior was due to their overall performance of these genotypes to each stress condition, whereas proline, MDA, and leaf temperature was in negative quadrat that resulted in oxidative stress in sensitive genotypes.
A biplot created between PC1 and PC2 showed a clear pattern of grouping genotypes along the vector line. The outstanding performance of genotypes for a specific trait was plotted closer to the vector line. Genotypes and traits that lie far away from the origin have better breeding potential than the other genotypes. The current study classified the breeding lines K25, U6, U17, and commercial cultivar K58 as sensitive to all stress conditions. Contrarily the breeding lines U12, U47, U49, and commercial cultivar K52 exhibited tolerance mechanisms and were classified as tolerant to all stress conditions. Similar findings were reported by (Gedam et al. 2021). Our results regarding PCA analysis were further corroborated by results reported by Al-Ashkar et al. (2019). Furthermore, it was also reported by Grzesiak et al. (2019) that tolerant genotypes were plotted in the positive quadrat near to stress promising traits.
AMMI stability analysis validated that all three stress environments were diverse and suggested differential genotypes response to the exposed stress environments. The onion breeding lines K41, U12, U47, U49, and commercial cultivar K52 clustered in the right quadrat were considered as stable genotypes to applied stress conditions. These genotypes had desired traits that can be used by the breeders for breeding of high-yielding onion varieties under varying environments.
In our study, the significant differences among tested onion genotypes indicated the presence of genetic variation for the measured bulb weight trait. The improvement in onion towards stress tolerance depends on genetic variability for the determination of stress tolerance. We quantified heritability that provided information about the magnitude of genetic and environmental variation (Roy and Shil 2020). It is also a potent tool to predict phenotypic reliability that contributes to breeding values. We observed higher heritability for bulb weight that suggested that it was due to genetic differences rather than the influence of environment on genotypes. Thereby it can be exploited as consistent criteria for screening of different genotypes to abiotic stress tolerance.

Conclusion
The current study explored the potential of 32 onion breeding lines and commercial cultivars to contrasting stress conditions. The genotypes were screened based on their morpho-physiological and biochemical changes. Traits analyzed in this study are the major selection criterion for the screening of stresstolerant and susceptible cultivars and can be used for the screening of large onion germplasm. The breeding lines K41, U47, U49, and commercial cultivar K52, were found to be stable across different stress conditions. These genotypes showed desirable traits to be used as breeding material for abiotic stress breeding.
Authors contribution ZNÖG and AFG conceptualized the idea and designed the study. UKC and MDJ performed the experiment, collected the data, analyzed the data, and wrote the initial draft. ZNÖG and AFG edited and reviewed the draft of manuscript. All the authors, read and approved the final draft.

Conflict of interest
The authors declare no conflict of interest.