Identification of a robust QTL hotspot across environments governing drought tolerance in lentil (Lens culinaris Medikus)

Abstract

Background

Drought stress significantly hinders the growth and productivity of many crops, including lentil. Hence, improving our understanding of the genetic and molecular bases underlying drought response is essential for mitigating yield losses under water-limited conditions. The primary objective of this study was to identify quantitative trait loci (QTLs) associated with various morpho-physiological traits using genotyping-by-sequencing (GBS) and to identify putative candidate genes contributing to drought tolerance in lentil.

Result

A F₇ recombinant inbred line (RIL) population was developed to map QTLs for morpho-physiological traits including root and shoot length, fresh and dry weight of roots and shoots, drought score, seedling survivability, relative water content, proline, superoxide dismutase (SOD), glutathione peroxidase (GPX) enzyme activities and seed yield/plant under controlled (hydroponic) and rainout shelter conditions over multiple drought environments during 2021–2024. A ‘hotspot QTL’ cluster consisting of six QTLs was detected for drought tolerance related traits viz. seed yield/plant, fresh weight of roots and shoots and chlorophyll content on linkage group 7. These QTLs detected under four environments were found to be strongly linked with drought tolerance and explained 48.83% phenotypic variance on a cumulative basis. Furthermore, the hotspot harboured several drought responsive functionally annotated genes.

Conclusion

The present study provides valuable insights into QTLs and SNPs associated with drought tolerance in lentil. The identification of a robust QTL hotspot region, together with key candidate genes, offers promising targets for fine mapping and development of marker assisted breeding strategies aimed at improving drought tolerance in lentil.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12864-026-12711-0.

Background

Lentil is among the major dietary legumes, cultivated over 5.50 million hectares (m ha) of land globally with an annual production of 6.65 million tonnes [1]. It is grown worldwide under rainfed conditions [2]. In Asia lentil is harvested on an area of 2.92 million hectares and in India, it is grown on 1.74 million hectares [3]. Here mostly it is cultivated under rainfed condition, mainly in central and eastern parts of the country and in north plain zone under restricted irrigation condition. Drought stress is a major constraint affecting crop plants, including lentil as it limits growth, yield and overall productivity by disrupting key physiological and biochemical processes [4, 5]. The seedling and reproductive stages of lentil are sensitive and moderately sensitive to drought stress, respectively [6–9]. Drought stress adversely effects relative water content, membrane stability index, proline accumulation, and antioxidant activities while intensifying oxidative stress [5, 9], ultimately resulting in substantial reduction in biomass and seed yield [10]. Previous studies have reported that the yield reduction under drought stress in lentil can range between 38 and 74% [11]. Therefore, the development of drought-tolerant cultivars is essential to minimize yield losses under low rainfall conditions. Plant tolerance to drought stress is often linked with accumulation of osmoprotectants such as proline, glycine betaine, soluble carbohydrates, antioxidants, which help maintain cellular integrity and function [12, 13, 8]. These biochemical and physiological traits are regulated by multiple genes operating at molecular, cellular and whole plant levels, and collectively contribute to the maintenance of water and ion homeostasis, thereby protecting plants from wilting under drought stress [5].

Various morpho-physiological traits including root length, shoot length, fresh and dry weight of roots and shoots, yield and yield components, relative water content, membrane stability index, chlorophyll content, proline accumulation, and antioxidant activities, have been widely used to screen drought tolerance in crop plants such as lentil [5, 14], cowpea [15], and wheat [16]. However, the genetic basis of many of these traits remains poorly characterized in lentil. Previous studies have suggested both simple inheritance based on seedling survival [9, 17] and complex genetic control of drought tolerance [18]. Several attempts have been made to map QTLs for drought [9], heat [19] and salinity tolerances [20] in lentil using simple sequence repeat markers (SSRs). However, early linkage maps constructed using low-density SSR markers often resulted in large confidence intervals [21], and QTLs identified from such maps may carry undesirable linkage drag during marker assisted selection (MAS).

To overcome these limitations, genotyping-by-sequencing (GBS) has emerged as a robust and cost-effective approach for high-throughput SNP discovery and genotyping across plant genomes [22]. GBS reduces genome complexity through restriction enzyme digestion enabling the rapid development of dense genetic maps and facilitating QTL identification for complex traits in mapping populations such as recombinant inbred lines (RILs). In legumes, GBS has been successfully applied to construct high-resolution linkage maps and identify loci linked to abiotic stress tolerance in chickpea, pea, and lentil [23–25]. However, to date, no studies have reported the identification of hotspot QTLs for drought tolerance in lentil using GBS-based RIL populations evaluated across multiple drought stress environments.

Therefore, this research was aimed to: (i) evaluate a RIL population of lentil at both seedling and reproductive phases under hydroponic and field conditions, (ii) identify and map hotspot QTLs associated with key morpho-physiological traits under drought stress, (iii) identify putative candidate genes underlying these QTLs based on functional annotation and tight genetic linkage. The identification of these genomic regions and associated markers is expected to facilitate the development of drought-resilient lentil cultivars through marker-assisted breeding.

Materials and methods

Plant materials

A population of 162 F₇ recombinant inbred lines (RILs) was developed using single-seed decent from a cross between JL-3 (drought-sensitive genotype) and PDL-2 (drought-tolerant genotype) [5, 14]. The RILs were phenotyped under multiple drought stress conditions. The schematic flowchart of the experimental methodology is presented in Fig. 1 and details of phenotyping procedures are described below:

Fig. 1.

Schematic flow chart depicting phenotyping and genotyping for drought tolerance in lentil. RIL: Recombinant inbred lines, QTL: Quantitative trait loci

Phenotyping of RILs at seedling stage under hydroponic condition

The 162 F₇-RILs along with parents (JL-3 and PDL-2) were grown and screened for drought tolerance at the National Phytotron facility (NPF), Indian Council of Agricultural Research-Indian Agricultural Research Institute (ICAR-IARI), New Delhi, India. The air temperature was maintained at 18/25°C (± 2 °C) day/night, with ~ 45% of relative humidity.

Drought stress tolerance was evaluated under hydroponic condition following the protocol described by Singh et al. [26, 5, 9]. Seeds were surface sterilized with 1% sodium hypochlorite for 30 min and placed on moist filter paper for germination. After one week, uniform seedlings were transferred to a hydroponic nutrient medium containing KNO₃ (0.5 mM), Ca (NO₃)₂.4H₂0 (0.5 mM), MgSO₄.7H₂O (0.2 mM), KH₂PO₄ (0.1 mM), KCl (50 µM), H₃BO₃ (46 µM), Fe- EDTA (20 µM), MnCl₂.4H₂O (2 µM), ZnSO₄.7H₂O (1µM), CuSO₄.5H₂ O (0.3 µM) and NaMoO₄.2H₂ O (0.5µM).

Six -days post-transplantation, drought stress was imposed by exposing the seedlings’ roots to air for 5 h daily over six consecutive days, while, control plants remained continuously submerged in the nutrient solution throughout the experimental period. The pH of the hydroponic solution was maintained at 6.5 using 1 M HCl or 1 M KOH. The solution was continuously aerated with an aquarium air pump and replaced after every 2 days.

After 6-days of treatment, all RILs along with parents were scored for drought tolerance using 0 to 4 scale as described by Singh et al. [26, 5, 9] (Fig. 2). The RIL population was also evaluated based on the reduction in morphological traits including root and shoot lengths & their fresh and dry weights. Seedling survivability and drought score were recorded after 6 d of drought stress, with seedlings surviving without wilting classified as tolerant (scores 0–1), and non-surviving seedlings classified as sensitive (score 4) [9].

Fig. 2.

Phenotypic response to drought stress at seedling stage under hydroponic conditions: (a) Seeds germinated on filter paper under controlled conditions (b) Seedlings grown up to 15 days under control conditions, (c) seedlings exposed to air for 5 hours to impose drought stress (d) recovery response of tolerant and sensitive recombinant inbred lines (RILs) under normal conditions

In addition, physiological traits such as relative water contents (RWC), membrane stability index (MSI), chlorophyll content, proline content, and antioxidant enzyme activities were measured after 3 days of drought stress to assess drought tolerance. These analyses were performed following standard protocols described by Singh et al. [5, 9]. Shoots and roots were oven-dried at 80 °C for 72 h to determine dry weights. The experiment was conducted under a completely randomized design with three replications.

Phenotyping of RILs at reproductive stage for drought stress using rain-out shelter

The 162 F₇ RILs along with parental genotypes (JL-3 and PDL-2) were also evaluated under a rain-out shelter condition at ICAR-IARI, New Delhi. The soil at the experimental site was classified as sandy clay loam at a depth of 0–15 cm. Soil pH, electrical conductivity, and organic carbon content were recorded as 7.6, 0.35 dsm^-1, and 4.3 g/kg, respectively. Particle size distribution comprised 58% sand, 25% silt, and 22% clay. Soil pH and electrical conductivity were determined using a soil–water suspension following standard procedures [27, 28]. Soil organic carbon was estimated using the Walkley and Black [29] method, while particle-size distribution was analyzed using the hydrometer method [30].

Soil moisture content was assessed using the gravimetric method. At the time of sowing, soil moisture content at 0–15 cm depth was 24.8%. During the vegetative, flowering and maturity stages, soil moisture content at 0–15 cm depth ranged from 21.2 to 22.2%, 19.5 to 20.2% and 16.2% to 17.2%, respectively across environments E1-E3 under rainout shelter conditions. At the reproductive stage, soil moisture content at 15–30 cm depth ranged from 21.5 to 21.8% (E1-E3). The mean temperature during the cropping season (November to April, 2020-23) ranged from 17.9 to 30.9 °C.

No irrigation was provided till crop maturity and the RILs were completely grown entirely on the available soil moisture under drought stress conditions (Fig. 3). For comparison, two irrigations were applied from sowing to maturity under normal condition. Reduction in seed yield/plant (g) under drought stress was recorded during 2020-21 (E1), 2021-22 (E2) and 2022-23 (E3), which were designated as environments 1, 2 and 3, respectively. The percentage reduction in seed yield was calculated using the formula: (Seed yield under stress-Seed yield under control)/Seed yield under control x 100. Seed yield per plant was recorded from ten randomly selected plants after 120 days under drought stress and 140 days under normal conditions. The RIL population along with the parents were evaluated using randomized block design (RBD) with three replications, following recommended agronomic practices.

Fig. 3.

Phenotypic response of recombinant inbred lines (RILs) and parental genotypes to drought stress at reproductive stage under rainout shelter condition

Genomic DNA isolation and genotyping

Genomic DNA from the parental genotypes (JL-3 and PDL-2) and 162 RILs were digested with EcoRI and MseI using ddRAD approach to construct libraries for GBS. Sequencing quality was assessed with FastQC [31] and MultiQC [32], by examining base quality distribution, GC content, and adapter contamination. All samples met the quality criteria with more than 95% of bases exceeding a Q20 score. Adapter trimming and removal of low-quality reads were performed using Fastp [33], retaining only fragments containing both restriction enzyme cut sites. Paired-end sequencing of 164-plex libraries generated 1,127.9 million raw reads (149.99 Gb), of which 1,120.4 million high-quality reads were retained after quality filtering.

Clean reads were aligned to the indexed lentil reference genome (https://knowpulse.usask.ca/bio_data/2690904) using the BWA-MEM algorithm, and the resulting alignment files were processed with SAMtools [34]. Single nucleotide polymorphisms (SNPs) were identified using the reference based pipeline of Stacks v2.53 [35] with default parameters. A total of 743,116 biallelic SNPs were initially detected, of which 59,918 were retained after parental filtering. Subsequent filtering based on minor allele frequency (MAF) ≥ 0.05 and marker presence in atleast 60% of the RIL population reduced the dataset to 47,927 SNPs. Non-polymorphic markers and markers deviating significantly from expected segregation ratios were excluded, resulting in a final set of 3,120 high-quality SNPs distributed across seven linkage groups (Supplementary Fig. 1). On average, 99.48% of reads were successfully mapped to the reference genome.

Linkage map and QTL mapping

A genetic linkage map using 162 RILs was constructed by integrating phenotypic and genotypic data, with genetic distances estimated using the Kosambi mapping function [36]. QTLs associated with drought tolerance were identified using composite interval mapping (ICIM) implemented in QTL IcIMapping v4.2.53 [37]. Significance thresholds were determined from 1,000 permutation test with a minimum logarithm of odd (LOD) threshold ≥ 2.5. QTLs were visualized using MapChart v2.32 [38], and the key parameters including chromosomal position, confidence interval, flanking markers, LOD, phenotypic variance explained (PVE), additive effect, and physical size were recorded.

Putative candidate gene prediction

The Lens culinaris v1 reference genome was explored using the JBrowse tool available on Phytozome v13 to identify putative genes underlying the significant QTL regions [39]. Functional annotations of these genes were retrieved from Phytozome v13 (https://phytozome-next.jgi.doe.g.ov/) to gain insights into their potential roles in regulating the target traits. Genes located within the defined QTL confidence intervals were subsequently designated as candidate associated genes (CAGs).

Statistical analysis, and software used

Descriptive statistics, frequency distributions, and overall correlations among morpho-physiological traits under drought stress conditions were analysed using the Statistical Package for Social Sciences (SPSS Statistics, version 19, IBM, USA).

Results

Phenotypic evaluation under drought stress conditions

The F₇ intraspecific RIL population comprising 162 lines, along with the parental genotypes, were evaluated under drought stress condition. Descriptive statistics including mean, range, standard deviation (SD), coefficient of variation (cv), mean sum of squares (MSS), skewness, and kurtosis, were calculated for each phenotype in both RILs and parents. Analysis of variance revealed statistically significant differences between the parental genotypes for all the morpho-physiological traits under drought stress, except for superoxidase dismutase (SOD), for which the difference was not significant.

Substantial variation was observed among the RILs for all evaluated morpho-physiological traits including root and shoot length, fresh and dry weight of root and shoots, relative water content, membrane stability index, chlorophyll, SOD, glutathione peroxidase (GPX) and seed yield/plant (Table 1). Several RILs exhibited trait values that exceeded or were lower than those of both the parents indicating transgressive segregation and a continuous phenotypic distribution within the population. Skewness and kurtosis values were less than 1.0 for most traits, suggesting polygenic control (Table 1).

Table 1.

Descriptive statistical analysis and analysis of variance of RILs along with parents for various traits under drought stress during 2020–2023

Environment	Trait	RIL population							Parent
		Mean	Range	SD	CV	MSS	Skewness	Kurtosis	JL-3 (Mean)	PDL-2 (Mean)	MSS
E-1	Seed yield (% reduction)	56.54	11.1–82.1	15.74	0.28	683.29**	-0.127	-1.168	69.67	48.95	644.35**
E-2	Seed yield (% reduction)	55.15	13.4–81.3	15.78	0.28	674.43**	-0.294	-0.845	65.95	46.57	563.55**
E-3	Seed yield (% reduction)	55.32	14.6–80.8	14.33	0.26	562.65**	-0.346	-0.840	66.90	45.16	709.20**
E-4	Root length (%reduction)	42.96	25.9–61.5	10.87	0.25	346.02**	0.08	-1.73	53.20	32.67	632.427**
	Shoot length (%reduction)	54.26	26.7–77.9	16.86	0.31	848.05**	0.0	-1.76	72.23	42.30	1344.07**
	Fresh weight of root (% reduction)	66.03	41.6–82.6	9.97	0.15	289.09**	-0.08	-1.38	79.40	61.27	493.227**
	Fresh weight of shoot (%reduction)	44.93	25.4–66.5	14.09	0.31	589.84**	0.12	-1.78	60.80	32.20	2053.5**
	Dry weight of root (% reduction)	42.67	19.8–67.8	17.83	0.42	951.93**	0.11	-1.89	64.17	23.50	2480.67**
	Dry weight of shoot (%reduction)	52.46	24.6–84.5	22.38	0.43	1498.17**	0.13	-1.91	79.83	31.23	1587.63**
	Drought Score	1.99	0–4	1.93	0.97	11.10**	0.05	-1.95	4.00	0.33	3542.94**
	Seedling survivability (%)	52.36	0-100	49.64	0.95	7420.09**	-0.10	-1.99	0.00	100.00	16.67**
	RWC (% reduction)	45.94	26.5–66.7	13.18	0.29	512.12**	0.11	-1.82	62.13	35.33	15,000**
	MSI (% reduction)	39.32	14.5–65.9	19.12	0.49	1094.9**	0.06	-1.87	58.53	17.40	1077.36**
	Chlorophyll (% reduction)	42.68	20.8–64.8	14.76	0.35	646.79**	0.13	-1.85	56.60	28.37	2537.93**
	Proline (% increase)	25.09	14.8–37.8	5.51	0.22	87.73**	-0.10	-1.49	16.83	27.87	1195.68**
	SOD (% increase)	26.86	12.6–43.6	10.24	0.38	310.12**	-0.02	-1.68	12.63	38.90	4.33 NS
	GPX (% increase)	40.56	18.5–63.9	15.67	0.39	727.08**	0.15	-1.83	25.93	58.47	1123.40**

SD Standard deviation, CV Coefficient of variation, MSS Mean sum of squares, NS Non significant, RWC Relative water content, MSI Membrane stability index, SOD Superoxide dismutase, GPX Glutathione peroxidase

^**Significant values

Correlation analysis showed that all the major traits were significantly correlated at both seedling and reproductive stages under drought stress conditions (Table 2). The relationship between the two phenotyping approaches i.e. hydroponic screening (E4) and rainout shelter evaluations (E1-E3)—was also examined, and the hydroponic screening showed a significant correlation with seed yield under rainout shelter condition (Table 2). Based on combined morpho-physiological performance across drought stress environments, 15 promising RILs were identified. Among these, two lines (RIL-57 and RIL-98) consistently exhibited superior performance across multi-environment drought stress trials (Supplementary Tables 1 and 2).

Table 2.

Correlation analysis among traits under drought stress condition

Env.	Traits	SY	SY	SY	RL	SL	FWR	FWS	DWR	DWS	DS	SS	RWC	MSI	CHL	Proline	SOD	GPX
E1	SY	1
E2	SY	0.96**	1
E3	SY	0.92**	0.96**	1
E4	RL	0.90**	0.89**	0.87**	1
	SL	0.91**	0.89**	0.87**	0.97**	1
	FWR	0.89**	0.86**	0.84**	0.93**	0.96**	1
	FWS	0.92**	0.91**	0.89**	0.97**	0.98**	0.95**	1
	DWR	0.92**	0.90**	0.90**	0.97**	0.97**	0.94**	0.98**	1
	DWS	0.92**	0.9**	0.89**	0.97**	0.97**	0.94**	0.98**	0.99**	1
	DS	0.92**	0.90**	0.90**	0.98**	0.98**	0.94**	0.98**	0.99**	0.99**	1
	SS	-0.92**	-0.91**	-0.90**	-0.98**	-0.98**	-0.94**	-0.98**	-0.99**	-0.99**	-0.99**	1
	RWC	0.90**	0.88**	0.88**	0.96**	0.96**	0.94**	0.97**	0.97**	0.97**	0.98**	-0.98**	1
	MSI	0.89**	0.88**	0.89**	0.95**	0.93**	0.88**	0.95**	0.98**	0.97**	0.98**	-0.98**	0.95**	1
	CHL	0.91**	0.9**	0.90**	0.97**	0.96**	0.91**	0.96**	0.98**	0.98**	0.99**	-0.99**	0.97**	0.99**	1
	Proline	-0.9**	-0.87**	-0.85**	-0.92**	-0.94**	-0.93**	-0.94**	-0.93**	-0.94**	-0.94**	0.94**	-0.94**	-0.89**	-0.91**	1
	SOD	-0.88**	-0.87**	-0.88**	-0.92**	-0.91**	-0.86**	-0.93**	-0.96**	-0.95**	-0.95**	0.95**	-0.92**	-0.97**	-0.95**	0.88**	1
	GPX	-0.83**	-0.83**	-0.83**	-0.86**	-0.87**	-0.78**	-0.86**	-0.89**	-0.88**	-0.89**	0.90**	-0.85**	-0.93**	-0.90**	0.79**	0.93**	1

Env. Environment, SY Seed yield (% reduction), RL Root length (% reduction), SL Shoot length (% reduction), FWR Fresh weight of root (% reduction), FWS Fresh weight of shoot (% reduction), DWR Dry weight of root (% reduction), DWS Dry weight of shoot (% reduction), DS Drought score, SS Seed survivability (%), RWC Relative water content (% reduction), MSI Membrane stability index (% reduction), CHL Chlorophyll (% reduction), Proline (% increase), SOD Superoxide dismutase (% increase ), GPX Glutathione peroxidase (% increase)

** Significant values

Generation of ddRAD sequencing data computational processing and linkage mapping

A total of 1,127.87 million reads (149.9967Gb) were generated form 162 F₇RILs and the two parental genotypes. Raw sequencing data were processed to remove adapter sequences and other technical contaminants, resulting in 1120.40 million high quality clean reads used for downstream analyses. Reads passing quality controls were alligned to the indexed lentil reference genome (https://knowpulse.usask.ca/bio_data/2690904) using the BWA-MEM algorithm, with an average mapping rate of 99.48%.

Initially, 743,116 biallelic SNPs were identified. After parental genotype filtering, 59,918 SNPs were retained. Further filtering based on marker presence in at least 60% of the population and a MAF of 5% reduced the dataset to 47,927 SNPs. Markers showing segregation distortion or lacking polymorphism among progenies were excluded, resulting in a final set of 3,120 high quality SNPs mapped across 7 linkage groups (LG) of lentil (Fig. 4a). The resulting genetic map spanned 1045.134 cM with an average inter-marker distance of 0.335 cM (Supplementary Fig. 1).

Fig. 4.

Chromosome-wise single nucleotide polymorphism (SNP) distribution (a) and logarithm of odd (LOD) profiles across the lentil genome under various environments (b). Graphs depict genetic position on x-axis and LOD score on y-axis for the identified quantitative trait loci (QTLs). LG: Linkage group, qSY: QTL for seed yield, qFWR: QTL for fresh weight of root, qFWS: QTL for fresh weight of shoot, qCHL: QTL for chlorophyll content, Env: Environments

Mapping QTLs for drought tolerance

Quantitative trait loci (QTL) analysis was performed for multiple morpho-physiological traits and seed yield associated with drought tolerance across different environments. Six QTLs identified in this study shared the same left (Lcu. 2RBY.Chr7:13113170) and right (Lcu. 2RBY.Chr7:476136786) confidence intervals and were therefore considered to constitute a hotspot region on LG 7.

A major QTL influencing seed yield per plant was detected in Environment 1, while minor QTLs for the same trait were identified in Environments 2 and 3. In Environment 4, three minor QTLs associated with chlorophyll content, fresh weight of root and fresh weight of shoot were detected. The LOD values and PVE of these hot spot QTLs ranged from 2.54 to 4.07 and from 7.28% to 10.82%, respectively (Table 3).

Table 3.

Summary of identified Quantitative trait loci (QTLs) associated with various traits under drought stress

Trait Name	Environment	LG	Name of QTL	Position	Left Marker	Right Marker	LOD	PVE(%)	Add	Left CI	Right CI
Seed yield/plant	1	7	qSY7.1_Env1	59	Lcu.2RBY.Chr7:13113170	Lcu.2RBY.Chr7:476136786	4.0753	10.8243	-5.6744	58.5	59.5
Seed yield/plant	2	7	qSY7.1_Env2	59	Lcu.2RBY.Chr7:13113170	Lcu.2RBY.Chr7:476136786	3.1865	8.5844	-4.9596	58.5	59.5
Seed Yield/plant	3	7	qSY7.1_Env3	59	Lcu.2RBY.Chr7:13113170	Lcu.2RBY.Chr7:476136786	2.5428	7.0038	-4.0423	58.5	59.5
Fresh weight of root	4	7	qFWR7.1_Env4	59	Lcu.2RBY.Chr7:13113170	Lcu.2RBY.Chr7:476136786	2.8908	7.8065	-3.0448	58.5	60.5
Fresh weight of shoot	4	7	qFWS7.1_Env4	59	Lcu.2RBY.Chr7:13113170	Lcu.2RBY.Chr7:476136786	2.6919	7.3294	-4.2099	58.5	60.5
Chlorophyll content	4	7	qCHL7.1_Env4	59	Lcu.2RBY.Chr7:13113170	Lcu.2RBY.Chr7:476136786	2.5933	7.2831	-4.3231	58.5	60.5

Negative additive effects indicate that alleles from PDL-2 enhance corresponding trait values

LG Linkage group, LOD Logarithm of odds, PVE Phenotypic variance explained by the QTL, CI Confidence interval

Co-localization of the identified QTLs

The physical positions of all the identified QTLs were examined to assess their genomic distribution and potential co-localization. This analysis revealed that drought-tolerance associated QTLs were not dispersed across the genome, but instead overlapped, sharing nearly identical genomic coordinates. Co-localization was confirmed based on overlapping confidence intervals and the close proximity of flanking SNP markers.

Using integrated genotypic and phenotypic data, six significant QTLs were identified. Notably, all six QTLs mapped to LG 7, and occupied the same genomic region, thereby constituting a distinct QTL hotspot (Fig. 5). Within this hotspot, QTLs associated with seed yield, root fresh weight, chlorophyll content and shoot fresh weight—namely qSY7.1_Env2, qSY7.1_Env3, qFWR7.1_Env4, qCHL7.1_Env4 and qFWS7.1_Env4—co-localized with the primary QTL qSY7.1_Env1 on LG 7. These QTLs spanned a physical interval from Lcu. 2RBY.Chr7: 13,113,170 to Lcu. 2RBY.Chr7: 476,136,786 (Table 3).

Fig. 5.

Hotspot quantitative trait loci (QTLs) governing multiple drought tolerance traits on linkage group (LG) 7 of Lens culinaris. Genetic distances (cM) are shown on the left side of the LG. Single nucleotide polymorphism (SNP) marker names are displayed on the right, and QTLs are illustrated as colored bars positioned to the right of the LGs. qCHL: QTL for chlorophyll content, qSY: QTL for seed yield, qFWR: QTL for fresh weight of root, qFWS: QTL for fresh weight of shoot, Chr: Chromosome, Env: Environments

Identification of candidate genes for each QTL

Gene loci located within the physical intervals of each QTL were identified and the number of genes corresponding to each QTL are listed in Supplementary Table 3. As all the identified QTLs overlapped within the same genomic region, they shared an identical set of candidate genes totalling to 5,502 genes.

Functional annotation of these candidate genes revealed their involvement in diverse biological categories, including transporters, transcription factors, stress-responsive proteins, hormone synthesis as well as signalling pathways and cell wall modification. Notably, several candidate genes with well-established roles in abiotic stress responses were identified in the hotspot region, such as early responsive to dehydration protein, dehydration-responsive element-binding protein, drought induced protein type, late embryogenesis abundant protein (LEA), ABC transporters, ABA 8’-hydroxylase and MYB transcription factors, all of which have been previously implicated in drought tolerance. Key candidate genes with annotations directly related to drought stress tolerance are enlisted in Table 4.

Table 4.

Details of key candidate genes identified witin QTL region and their functional annotation

S.No.	Gene ID	Linkage group	Physical position	Length (bp)	Annotation
1.	Lcu.2RBY.7g051830	7	Lcu.2RBY.Chr7:408225454.408226605 reverse	1151	DHHC palmitoyltransferase (DHHC)
2.	Lcu.2RBY.7g052040	7	Lcu.2RBY.Chr7:409390014.409391284 reverse	1270	MYB Transcription Factor
3.	Lcu.2RBY.7g052590	7	Lcu.2RBY.Chr7:414136722.414138833 reverse	2111	Basic helix loop helix (bhlh) DNA-binding family protein
4.	Lcu.2RBY.7g054470	7	Lcu.2RBY.Chr7:428534399.428535569 reverse	1170	WD40 repeat-containing protein
5.	Lcu.2RBY.7g057770	7	Lcu.2RBY.Chr7:452878911.452882183 reverse	3272	Subtilisin-like protease
6.	Lcu.2RBY.7g028760	7	Lcu.2RBY.Chr7:182209268.182211533 forward	2265	AP2-like ethylene-responsive transcription factor
7.	Lcu.2RBY.7g010740	7	Lcu.2RBY.Chr7:35154345.35156217 reverse	1872	WRKY transcription factor 22 (WRKY22)
8.	Lcu.2RBY.7g057370	7	Lcu.2RBY.Chr7:450444912.450445790 reverse	878	Dehydration-responsive element-binding protein 1E-related
9.	Lcu.2RBY.7g055060	7	Lcu.2RBY.Chr7:432341328.432342130 forward	802	Early responsive to dehydration 15 protein
7.	Lcu.2RBY.7g017800	7	Lcu.2RBY.Chr7:92430796.92431346 forward	550	Early-responsive to dehydration stress protein (ERD4)
10.	Lcu.2RBY.7g019680	7	Lcu.2RBY.Chr7:106798423.106802311 forward	3888	Dehydration-responsive protein RD22
11.	Lcu.2RBY.7g052940	7	Lcu.2RBY.Chr7:417255888.417256466 reverse	578	Late embryogenesis abundant protein (LEA)
12.	Lcu.2RBY.7g026380	7	Lcu.2RBY.Chr7:160445260.160448568 reverse	3380	Abscisic acid 8’-hydroxylase 1-related
13.	Lcu.2RBY.7g041480	7	Lcu.2RBY.Chr7:309179222.309183756 reverse	4534	ABA 8’-hydroxylase
14.	Lcu.2RBY.7g061570	7	Lcu.2RBY.Chr7:475564617.475573962 forward	9345	ATP-binding cassette transporter subfamily A
15.	Lcu.2RBY.7g061590	7	Lcu.2RBY.Chr7:475939786.475946942 forward	7156	ATP-binding cassette transporter subfamily A
16.	Lcu.2RBY.7g061600	7	Lcu.2RBY.Chr7:476051818.476058889 forward	7071	ATP-binding cassette transporter subfamily A
17.	Lcu.2RBY.7g031940.1	7	Lcu.2RBY.Chr7:213412784.213417585 forward	4801	ABC transporter F
18.	Lcu.2RBY.7g031970.1	7	Lcu.2RBY.Chr7:213685007.213688269 forward	3262	ABC transporter F
19.	Lcu.2RBY.7g064250	7	Lcu.2RBY.Chr7:488725149.488730651 forward	5502	Palmitoyl-acyl carrier protein thioesterase
20.	Lcu.2RBY.7g049760	7	Lcu.2RBY.Chr7:386291422.386295070 forward	3648	Beta-1,4-xylosyltransferase IRX14-related
21.	Lcu.2RBY.7g042720	7	Lcu.2RBY.Chr7:321381488.321383226 reverse	1738	Lactoylglutathione lyase glyoxalase I
22.	Lcu.2RBY.7g046220	7	Lcu.2RBY.Chr7:352279944.352282451 reverse	2507	Glycoside hydrolase, family 35

Bp: Base pairs

Discussion

Drought is one of the most severe abiotic stresses limiting global crop productivity including lentil crop. Developing drought-tolerant cultivars, remains one of the most effective strategies to mitigate yield losses; however, this requires precise phenotyping, robust identification of breeding material and reliable mapping of QTLs associated with drought tolerance. Previous studies have demonstrated that hydroponic screening provides a controlled and reproducible platform for minimizing environmental heterogeneity and ensures precise nutrient availability [26, 5, 9]. Nevertheless, since the ultimate deployment of drought-tolerant genotypes occurs under field settings, validation under fluctuating environments is essential. Therefore, a combined strategy involving initial hydroponic screening followed by multi-environment field evaluation represents a scientifically robust approach for accurate assessment of drought tolerance [5, 9]. While QTLs for drought tolerance have been extensively mapped in crops such as chickpea [40], wheat [41], and mustard [42], genetic improvement of lentil for drought tolerance has received comparatively less attention. Therefore, the current study was undertaken to address this considerable gap.

The RIL population was evaluated across 4 environments over three cropping seasons (2020-21, 2021-22, 2022-23). Significant differences were observed among the parental genotypes and within RILs for most morpho-physiological traits under drought stress, except SOD in the parental genotypes (Table 1). Based on combined morpho-physiological and yield-related performance, 15 promising RILs were identified, among which two lines (RIL-98, RIL-57) consistently exhibited superior performance (except GPX) across multiple drought stress environments (Supplementary Tables 1 and 2). These lines represent valuable genetic resources that can be exploited in future breeding programs, particularly through marker-assisted backcrossing. The relevance of traits such as seedling survival, drought score, root length, shoot length, fresh and dry weights of root and shoot, chlorophyll content, membrane stability index, proline, antioxidant activities and seed yield in improving drought tolerance has been well documented in lentil [5, 9], and other crops like chickpea [43], and wheat [44], further supporting the trait selection strategy adopted in this study.

Correlation analysis revealed strong and significant associations among most morph-physiological traits and seed yield at both seedling and reproductive stages (Table 2). These relationships suggest that several of these traits could serve as effective selection criteria under both hydroponic and field conditions. Similar correlations between physiological traits and yield under drought stress have been reported earlier in lentil [26, 5, 9], reinforcing their relevance for genetic dissection and QTL mapping.

The use of GBS approach enabled generation of a large number of high-quality SNP markers uniformly distributed across the lentil genome. This aligns with previous reports where GBS has proven effective for genetic diversity analysis, linkage map construction, and QTL mapping in lentil [45, 24]. The use of RIL population, which is highly homozygous and genetically stable, further enhanced mapping resolution and reliability. High-throughput GBS-based approach have been successfully used to identify drought tolerance QTLs in crops such as chickpea [40], and wheat [21]. In the present study, ddRAD-based GBS facilitated the development of a high-density genetic map and identification of drought associated-QTLs evaluated across both seedling (hydroponic) and reproductive (field) stages. To our knowledge, this represents the first comprehensive effort to identify of drought tolerance-related QTLs in lentil across multiple growth stages and environments.

A total of 6 QTLs linked with 4 drought-related traits were identified. Among these, three QTLs governing seed yield/plant were consistently detected under rainout shelter conditions with qSY7.1_Env1 exhibited the highest LOD value (4.07) and explained the maximum phenotypic variance (10.82%). Although the phenotypic variance explained by the individual QTLs was modest, this is expected for a complex trait such as drought tolerance, which is governed by multiple genes, and strongly influenced by genotype x environment interactions. Similar low to moderate PVE values have been reported for drought-related QTLs in various crops [46, 47]. Importantly, the repeated detection of seed yield QTLs across three field environments indicates relatively stable genetic control under rainout shelter condition. In contrast, QTLs for fresh weight of root, fresh weight of shoot and chlorophyll content were detected only under hydroponic condition (E4), highlighting the strong environmental sensitivity and polygenic nature of these traits. Further, environment-specific differences were quite evident. Under rain-out shelter conditions (E1–E3), QTLs associated with seed yield per plant were consistently detected across three environments, indicating relatively stable genetic control of yield under field-based drought stress. In contrast, under controlled hydroponic conditions (E4), QTLs were detected primarily for morpho-physiological traits such as fresh weight of root, fresh weight of shoot, and chlorophyll content, reflecting early-stage physiological responses to imposed stress. These differences likely arise from the contrasting nature of drought stress in the two systems, with hydroponic screening capturing acute, short-term physiological responses, while rain-out shelter evaluations reflecting integrated, season-long stress effects on yield.

Notably, all six QTLs co-localized within the same genomic interval on LG 7 forming a robust QTL hotspot. This clustering suggests either pleiotropic effects or tight linkage among genes controlling multiple drought-responsive traits. The relatively narrow interval of the hotspot further strengthens its relevance for fine mapping and breeding application (Figs. 4b and 5; Table 3). Similar clustering has been reported in legumes such as chickpea and soybean, where genomic regions controlling yield related traits also harbour stress-responsive genes [48, 49]. All identified QTLs exhibited negative additive effects, indicating that the favourable alleles were contributed by the drought tolerant parent, PDL-2, which was used as the male parent. This observation is consistent with the interpretation that negative additive effects reflect trait-enhancing alleles derived from the male parent [50]. Mining of the hotspot region revealed a large number of candidate genes (Supplementary Table 3), several of which have known roles in drought stress responses. These include transcription factors and stress responsive proteins involved in seed development, dehydration tolerance, chlorophyll maintenance and cell wall modifications (Table 4). Notable examples include DHHC-type zinc finger family protein (Lcu.2RBY.7g051830) associated with plant structure and yield [51]. MYB transcription factors (Lcu.2RBY.7g052040) involved in hormonal signaling and abiotic stress tolerance [52], bHLH (basic helix-loop-helix) transcription factor (Lcu.2RBY.7g052590) and WD40/YVTN repeat-like-containing domain (Lcu.2RBY.7g054470) implicated in seed development [53], and WRKY transcription factor (Lcu.2RBY.7g010740) known to regulate responses to high temperatures and water scarcity [54]. Several dehydration responsive genes, including Lcu.2RBY.7g057370 and Lcu.2RBY.7g019680 annotated as Dehydration-responsive element-binding protein 1E-related and Dehydration-responsive protein RD22, respectively were also identified, which are known to enhance drought tolerance through abscisic acid (ABA) dependent and independent pathways [55]. Additionally, the late embryogenesis abundant (LEA) (Lcu.2RBY.7g052940) proteins, Lcu.2RBY.7g049760 annotated as Beta-1,4-xylosyltransferase IRX14-related protein involved in cell wall biosynthesis and Lcu.2RBY.7g046220 for glycoside hydrolase family proteins involved in drought tolerance [56–58], further underscore the biological relevance of the hotspot region.

In summary, this study provides valuable insights into the genetic architecture of drought tolerance in lentil by integrating multi-environment phenotyping with high-density GBS-based genotyping. The contrasting responses observed under hydroponic and rain-out shelter condition emphasize the importance of genotype x environment interactions in drought tolerance. The identification of a robust QTL-hotspot on LG 7, along with several biologically relevant candidate genes, offers promising targets for fine mapping. Functional validation of these candidate genes will further elucidate the molecular mechanism underlying drought tolerance and will facilitate the development of drought-resilient lentil cultivars. Importantly, co-localization of multiple drought-related traits within a single genomic region provides a practical opportunity for simultaneous trait improvement through marker assisted selection.

Abbreviations

QTL

Quantitative trait locus

SNP

Single nucleotide polymorphism

GBS

Genotyping-by-sequencing

ddRAD-seq

Double digestion restriction-site-associated DNA sequencing

RIL

Recombinant inbred line

CAPS

Cleaved amplified polymorphic sequence

SSR

Simple sequence repeat

MAS

Marker assisted selection

Authors’ contributions

DS conceived the idea and coordinated this project. DS was involved in developing and phenotyping of RILs. DS and KG were involved in genotyping and construction of linkage maps, QTL and statistical analyses. DS, AT, KG, JT and MP contributed to writing of the manuscript. All the authors reviewed and gave their consent to the final version of the manuscript.

Funding

This research was financially supported by a grant from the Council of Scientific and Industrial Research (CSIR), Government of India [No. 38 (1483)/19/EMR-II]. The funding organizations were not involved in the design of the study, data analysis, interpretation, or the manuscript writing.

Data availability

The VCF (Variant Call Format) of RIL mapping population and its parents has been submitted in Figshare (10.6084/m9.figshare.30345205.v1).

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.