Lettuce (Lactuca sativa L.), a leafy vegetable species highly valued for its nutritional richness, holds substantial economic importance and enjoys widespread global cultivation and consumption, with China being a prominent contributor to its production (Kim et al., 2016; Shin et al., 2021). However, due to its shallow taproot system, lettuce requires frequent irrigation. Even a brief period of drought stress can adversely affect the plant, causing the production of undersized leaves and ultimately resulting in low crop yield (Li et al., 2017). To fulfill the needs of a large population, conserving water is a crucial strategy for increasing the yield of leafy vegetables, and of lettuce in particular (Saed-Moucheshi et al., 2013). Implementing such an approach can serve as an effective measure to optimize human labor in the agricultural sector while simultaneously enhancing water use efficiency (Scardigno, 2020; Hou et al., 2023).

In numerous regions worldwide, agricultural producers have prioritized the selection of crop types that can withstand drought conditions (Sun et al., 2023). Among these, drought-tolerant cultivars have demonstrated their significance in mitigating water consumption and facilitating the objective of conserving water, particularly in areas prone to high temperatures and water scarcity (Ferguson, 2019; Neha et al., 2021). Several studies have revealed that the rearing of cultivars with robust drought tolerance potential can lead to substantial reductions in water usage and notable improvements in harvest yields (Jamalluddin et al., 2021; Zhang et al., 2023). Currently, the emphasis on screening drought-tolerant varieties predominantly revolves around agronomic and field crops such as wheat (Nevo and Chen, 2010; Abdolshahi et al., 2015; Guellim et al., 2020), barley (Nevo and Chen, 2010), maize (Ali et al., 2015), cotton (Ahmad, 2020; Imtiaz et al., 2023; Iqbal et al., 2023; Wedegaertner et al., 2023), rice (Marcelo et al., 2017; Asma et al., 2021), and millet (Choudhary et al., 2021). However, there is a notable absence of comprehensive evaluations and verifications based on indicator screening for drought-tolerant genotypes specifically in lettuce. This includes assessing morphological, physiological, and biochemical changes as indicators of drought tolerance in lettuce varieties.

Over the past two decades, numerous indicators and features have been proposed to identify drought-tolerant varieties (Cai et al., 2020). Typically, drought stress indicators based on yield loss in drought conditions have been utilized for this purpose (Mohi-Ud-din et al., 2021; Sun et al., 2023). Drought susceptibility is commonly measured by comparing yield reduction under normal and water-deficient environments (Saed-Moucheshi et al., 2013). However, due to the multifaceted nature of crop genotype tolerance and susceptibility, drought tolerance in plants encompasses multiple factors, and a single indicator cannot be utilized to fully or accurately evaluate this phenotype (Munns et al., 2010; Lonbani and Arzani, 2011). Moreover, relying solely on one indicator can lead to a biased and incomplete assessment (Aslam et al., 2023). To effectively identify drought-tolerant varieties, a comprehensive evaluation of morphological, physiological, and biochemical indicators is necessary. Additionally, many evaluation indicators for assessing drought tolerance have primarily focused on filed crops, which may not be suitable for screening lettuce genotypes due to variations in growth morphology, harvested parts, and water necessities.

The identification of indicator markers for drought tolerance in genotype screening plays a critical role in various stages of plant development (Levi et al., 2009; De Brito et al., 2011; Bo et al., 2017). Previous research has identified several indicators of drought tolerance in crops. These indicators include leaf water potential, net photosynthesis, water use efficiency, maximum quantum efficiency of PSII, proline and betaine content, SS, chlorophyll, MDA, antioxidant enzyme activity, leaf area, and yield output, among others (Yan et al., 2020). However, it is acknowledged that these indicators are numerous and complex. To effectively select drought-tolerant cultivars, it is important to identify key indicators and streamline the evaluation process. In this regard, we propose the use of the CDC, D-value, and WDC to accurate assessment and screening of cultivars under both normal and drought conditions (Xu et al., 2020; Bao et al., 2023). By transforming the original set of indicators into a concise collection of representative indicators using PCA, complex and large datasets can be simplified (Donde et al., 2019). Through multiple linear regression analysis, predictive models can be developed to assess water-deficit tolerance based on morphological and physiological variables (Quevedo et al., 2022). Therefore, it is essential to establish a concise set of indicators for evaluating drought stress or develop an effective model that combines various evaluation methods to rapidly determine the resistance of superior lettuce genotypes or cultivars to drought.

Reduced yield due to water deficiency is the primary outcome of current varieties’ lack of adaptability and represents a serious danger to the ability of crops to withstand drought (Kaur and Asthir, 2017). However, plants have numerous mechanisms, such as osmotic adjustments or osmoregulation via the buildup of proline, carbohydrates, and other substances, that help them grow and produce a high yield. Plants are protected from oxidative stress by additional mechanisms, such as an increased antioxidant activity system (SOD, POD, CAT), which neutralizes ROS (OH^-, O 2   . − and MDA) (Yang et al., 2021). Increased chlorophyll levels also improve photosynthesis because they allow for more efficient use of water through regulation of stomatal activity, which in turn causes a rise in CO₂ concentration (Parida et al., 2007; Niu et al., 2018). In addition, root development is essential for sustaining water and nutrient availability during periods of drought. There is deficit of studies that employ a multivariate approach to investigate the morphological, physiological, and biochemical characteristics of lettuce genotypes and identify marker variables associated with drought resistance. Therefore, it is crucial to discover easily measurable indicators of drought stress that can facilitate rapid evaluation and validation of elite germplasm or drought-tolerant cultivars. In this experiment, 42 distinct varieties of lettuce were cultivated under both well-watered conditions (with soil water content maintained at 75%-85% of field capacity) and water-limited conditions (35%-45% of field capacity). Through the use of PCA, D-value calculation, cluster analysis, correlation analysis, and membership functional value assessment, the drought tolerance of lettuce genotypes was systematically evaluated and categorized based on their morphological and physiological indicators. Furthermore, a stepwise regression-based prediction model was developed to verify the drought tolerance capacity and mechanisms of several genotypes. These findings provide valuable insights into the drought tolerance and susceptibility of lettuce genotypes and establish a foundation for screening indicators of drought tolerance for leafy vegetables.

Drought stress significantly affects plant morphology, physiology, and yield, leading to reduced overall performance (Claeys and Inzé, 2013). Previous studies have often focused on specific aspects such as morphology, photosynthesis, or physiology to assess drought tolerance. However, it is crucial to consider indicators related to both yield and morphological characteristics to comprehensively evaluate drought tolerance (Dolferus, 2014; Ndayiragije and Li, 2022). In evaluating drought tolerance in cotton, morphological and yield indicators have been commonly used (Sun et al., 2021). However, these indicators provide only a partial understanding of crop drought tolerance, neglecting the roles of photosynthesis, physiology, and biochemical attributes. Osmotic adjustment, including parameters such as MDA and proline content, is an important component of drought tolerance and should be incorporated into comprehensive assessments (Wei et al., 2009; Fang and Xiong, 2015). Therefore, when evaluating and verifying drought tolerance, it is necessary to integrate multiple indicators that encompass morphology, physiology, biochemical changes, and representative indexes. This comprehensive approach ensures a better understanding of drought tolerance screening and evaluation in plants.

In our study, we conducted field cultivation of 42 lettuce genotypes and evaluated 13 indicator variables associated with drought tolerance. These indicators encompassed morphological, physiological, and biochemical traits. Throughout the growth stages, ranging from seedling to harvest, we performed variance analysis on the drought tolerance coefficient for the different water treatments. Notably, significant differences were observed among the genotypes, particularly under drought stress conditions ( Table 2 ). These findings demonstrate the presence of substantial genetic diversity among the genotypes, making them representative samples for the region. Notably, drought stress had a significant impact on all the indicators examined (P<0.05), leading to decreased drought tolerance coefficients (<1) or increased drought tolerance coefficients (>1) ( Table 2 ). Furthermore, the CV values for most indicators were higher under water deficit conditions compared to the normal treatment. This observation suggests that the lettuce genotypes chosen for this study exhibit considerable diversity, effectively responding to drought stress and generating representative outcomes.

To account for inherent variations among cultivars, we utilized relative values in evaluating different pakchoi varieties under drought stress. However, drought tolerance is a complex trait influenced by multiple factors, and relying on individual or single-type indicators for assessment can lead to errors (Ndayiragije and Li, 2022; Xiong et al., 2022). Currently, there is no single indicator that can provide a complete and accurate evaluation of drought tolerance. Therefore, it is crucial to identify more comprehensive indicators and adopt suitable evaluation methods for plant assessments. Moreover, many indicators exhibit correlations with each other, resulting in overlapping responses as indicators of crop stress tolerance ( Figure 2 ). Consequently, the utilization of multivariate analysis methods becomes essential in evaluating and screening comprehensive indicators associated with drought tolerance. PCA has proven to be effective in reducing multiple variables to underlying factors, addressing missing data issues, and facilitating efficient grouping of drought-tolerant varieties (Wu and Bao, 2012; Maheswari et al., 2016). Through PCA, we successfully transformed the 13 individual indicators of lettuce genotypes under drought stress into four distinct independent comprehensive indexes. Notably, the cumulative contribution rate of the first four independent comprehensive indexes exceeded 94.83% ( Table 3 ), indicating that a significant portion of the data encompassed by the 13 indicators was effectively encompassed by these comprehensive indexes. Previous studies utilized the PCA method, and the Zhong R2016 and XinLuZao 45 cultivars, two highly drought-tolerant cotton genotypes, were identified among 104 cotton genotypes based on 19 drought-related indicators (Sun et al., 2021). The drought tolerance membership function value acts as a multivariate indicator that combines the drought tolerance coefficients of different indicators, providing a comprehensive representation of overall plant performance under drought stress. This indicator effectively captures and reflects the collective response of plants to drought conditions. By utilizing the principal component scores, we calculated the membership function values and subsequently determined the D-value by incorporating the respective weights. This approach enabled us to rank the lettuce genotypes based on their drought tolerance, with higher D-values indicating greater drought tolerance ( Table 4 ). Similar findings were demonstrated by (Liu et al., 2015), and the results showed that twenty-six of the 82 wheat addition lines that expressed high drought resistance were selected by membership function value based on 10 agronomic traits. A higher MFVD (membership functional value of drought resistance) was observed in the Agropyron elongatum 3E addition line, which was considered the most drought-resistant material. Previous studies have utilized the classification of onion cultivars based on waterlogging tolerance and wheat and maize varieties based on salt tolerance. In these studies, the cultivars or varieties were categorized into two groups using their respective characteristics. The classification was determined by evaluating the euclidean distances between the cultivars or varieties, which provided valuable insights into their comparative levels of tolerance (Huqe et al., 2021; Uzair et al., 2022). Likewise, the drought tolerance of cotton cultivars was classified using the membership function and D-value. This classification approach allowed for the grouping of cotton cultivars into distinct categories based on their individual levels of drought tolerance. This study observed significant differences in morphological, physiological, and biochemical characteristics among the lettuce genotypes, indicating the presence of abundant genetic diversity. PCA was employed to transform the 13 indexes of drought tolerance in lettuce varieties into 4 independent composite indicators according to previous studies. The D-values of various lettuce genotypes were then determined using the membership function. Combining CDC, D-values, and WDC analysis enhances the reliability and practicality of stress tolerance assessment in lettuce. Through stepwise regression analysis, it was determined that among the 13 studied indexes, six drought tolerance indexes (RSA, RV, BDW, SS, SP, and RWC) had significant effects on the drought tolerance of lettuce. These indexes can be considered primary indexes for the evaluation and screening of drought-tolerant lettuce genotypes in future studies. Moreover, a stepwise regression predictive model was developed to assess the drought tolerance of lettuce, represented by the equation Y= -0.399 + 0.520RSA + 0.302RWC + 0.181RV + 0.151BDW + 0.058SP + 0.016SS (R² = 0.985, P value< 0.01). By incorporating multiple indexes as predictors, this model offers a dependable approach for evaluating the drought tolerance of lettuce. Through hierarchical clustering analysis, the 42 lettuce genotypes were classified into 5 diverse categories based on their D-values: highly drought- resistant, moderately drought- resistant, drought- resistant, drought-sensitive, and highly drought-sensitive ( Figure 3 ). The results of the gray relational analysis further validate the accuracy of the regression analysis and enhance the scientific reliability and credibility of the identified indexes in assessing drought tolerance.

Comprehensive evaluation methods have proven highly effective in identifying drought-tolerant plant varieties. The evaluation process encompasses various morphological and physiological aspects that contribute to a plant’s ability to withstand drought. It is crucial to validate the classification of plants using comprehensive evaluation methods that consider multiple indicators. One of the key indicators for drought tolerance is stomatal behavior, including their patterning and morphology, which significantly influence water use efficiency. Stomata, found on the surfaces of leaves, exhibit a wide range of shapes, sizes, and numbers across different plant species (Liu et al., 2021). The variations in stomatal size and density can be attributed to genetic factors as well as plant growth under diverse environmental conditions. When subjected to drought stress, stomata close to reduce water loss through evapotranspiration, resulting in decreased stomatal density on both the upper and lower epidermis. Our results demonstrated that with an increase in drought stress, the stomata close, and their density in the upper and lower epidermis also decreases to prevent evapotranspiration. Similar results were observed in wheat crop cultivars that showed lower stomatal density under well-watered and water-stressed conditions, and the drought-sensitive cultivar had a nonsignificantly larger decrease under water-stressed conditions (Nyachiro et al., 2001). In the current study, the significant changes in stomatal traits of drought-resistant genotypes helped to improve drought resistance, such as the reduction in stomatal aperture and density under water deficit conditions. The stomatal aperture of drought-sensitive lettuce decreased more than that of drought-resistant genotypes, and the stomatal density of drought-sensitive lettuce increased more than that of drought-resistant genotypes ( Figure 6 ).

ROS play a crucial role in the metabolic pathways associated with plant drought tolerance. Under drought stress conditions, plants frequently exhibit heightened levels of ROS, including increased production of O 2 . ¯ and elevated H₂O₂ content. H₂O₂ production intensifies at the onset of drought stress against which CAT is synthesized. in response to which antioxidants are produced to scavenge the ROS. SOD catalyzes the dismutation of superoxide to molecular oxygen and H₂O₂, after which H₂O₂ is converted into water and oxygen inside the cytosol and chloroplasts, which protects the cell from the toxic effects of ROS (Sarker and Oba, 2018; Liu et al., 2022). The high activity of SOD, CAT, and POD inside the cell during drought stress reflects the ability of the genotypes to tolerate drought; therefore, the genotypes in which high concentrations of these antioxidants are generated were considered tolerant, and the genotypes that produced low concentrations were considered sensitive. For different drought genotypes of soybean, the antioxidant enzyme activities of SOD, CAT, and POD in drought-resistant genotypes exceeded those of sensitive genotypes under different treatment times and drought degrees (Zhou et al., 2022). Antioxidant enzymes are commonly utilized as physiological indicators to identify plant stress resistance (Sánchez-Rodríguez et al., 2010; Sallam et al., 2019). In this study, the O 2 . − and MDA contents increased in five genotypes under water deficit conditions. The O 2 . − production rates of Yidali151, Zidie, and Jingyanyidali showed no significant difference, and Ziya significantly increased by 82.6%, and the MDA contents of Yidali151 and Ziya increased by 19.1% and 97.8%, respectively. Under water deficit conditions, the SOD, POD and CAT activities increased 88.1%, 58.4%, and 84.2%, respectively, in Yidali151. In Ziya, SOD increased by 21.5%, and POD and CAT showed no significant difference between normal irrigation and water-deficit conditions.

Therefore, in the evaluation and analysis of water deficit tolerance, it is crucial to account for a comprehensive range of indicators encompassing morphological, physiological, and biochemical parameters. This comprehensive approach enhances the effectiveness of identifying and screening drought-tolerant pakchoi varieties while reducing the cost and time associated with phenotyping. However, further research is required to unravel the molecular mechanisms underlying drought tolerance in lettuce genotypes and to develop targeted breeding strategies for highly drought-tolerant genotypes.

In this study, the drought resistance of 42 lettuce genotypes was assessed under both drought-stressed and normal-irrigated conditions. Thirteen indexes related to morphology, physiology, biochemical characteristics, and osmoregulation were evaluated to determine the genotypes’ drought tolerance. Various analytical techniques, including PCA, membership function value analysis, multiple regression analysis, CDC, D value, and WDC analysis, and cluster analysis, were employed to effectively evaluate the drought tolerance of pakchoi varieties. The assessment of lettuce drought tolerance involved analyzing six indicators through stepwise regression based on the D value: RSA, RV, BDW, SS, SP, and RWC. A digital model was developed to evaluate the drought tolerance of lettuce, and representative varieties were selected from five groups to assess their water resistance ability. The tolerance levels of each variety were validated through measurements of stomatal conductance features, ROS, MDA levels, and antioxidant enzyme activity. This research contributes to the identification of drought-resistant lettuce genotypes and provides a theoretical basis for further investigations into the underlying mechanisms of lettuce drought resistance.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

JL, HG and SH designed the experiments. JL and KA drafted the initial manuscript. LW and WW conducted the experiments. BD and HX analyzed the data. GL, BG and XW revised the manuscript. All authors contributed to the manuscript and approved the final manuscript.