Vol.: (0123456789) Euphytica (2025) 221:90 https://doi.org/10.1007/s10681-025-03498-4 RESEARCH Genome‑wide association study for salinity tolerance in the African rice, Oryza glaberrima Nafeti Titus Mheni · Newton Kilasi · Atugonza Bilaro · Marie‑Noelle Ndjiondjop · Shalabh Dixit · Abdelbagi M. Ismail · Susan Nchimbi Msolla Received: 5 December 2024 / Accepted: 11 March 2025 © The Author(s), under exclusive licence to Springer Nature B.V. 2025 Abstract  Oryza sativa, L., and Oryza glaberima, Steud, are the two most-grown rice species, making rice the second most-produced crop globally. While O. sativa is widely cultivated due to its high yield and marketability, O. glaberrima possesses valuable genetic traits for stress tolerance, including resist- ance to drought, flooding, and salinity. Genome-wide association studies (GWAS) have gained more popu- larity in O. sativa than in O. glaberrima, providing information on biological mechanisms underlying key agronomic traits. The current study aimed to find the essential genes for salinity tolerance in O. gla- berrima through marker-trait associations (MTAs) for traits related to salinity. Using the Yoshida nutri- ent solution, this study evaluated a previously devel- oped association mapping panel of 335 O. glaber- rima accessions under screen house conditions. The association mapping panel was genotyped using 9990 single nucleotide polymorphism (SNP) markers. Using 6103 polymorphic SNP markers, a GWAS was conducted to detect genomic regions associated with salinity tolerance. 34 MTAs were identified using the mixed linear model approach, representing 11 genomic regions on chromosomes 1, 3, 4, 5, 6, 7, 8, 9, 10, 11, and 12, except on chromosome 2, highlighting different significant loci contributing to salinity stress tolerance. Among the tested accessions, 21 geno- types were categorized as tolerant to moderately tol- erant based on the standard evaluation system score, representing promising materials for breeding pro- grams. The identified SNPs, genotypes, and genomic regions are valuable resources for understanding the potential genes and developing functional markers for salinity tolerance screening in African rice. This study underscores the potential of O. glaberrima as a genetic resource for improving rice productivity in salt-affected areas, thereby contributing to sustainable rice production. Keywords  Rice · O. glaberrima · Salinity · Genome-wide association study · Marker trait association · SNP N. T. Mheni (*) · N. Kilasi · S. N. Msolla  Department of Crop Science and Horticulture, Sokoine University of Agriculture, P.O. Box 3005, Morogoro, Tanzania e-mail: mheni22@yahoo.co.uk N. T. Mheni  Tanzania Agricultural Research Institute (TARI), Selian Centre, P.O. Box 6024, Arusha, Tanzania A. Bilaro  Tanzania Agricultural Research Institute (TARI), P.O. Box 1571, Dodoma, Tanzania M.-N. Ndjiondjop  AfricaRice Centre, Abidjan, Côte d’Ivoire S. Dixit · A. M. Ismail  International Rice Research Institute, DAPO Box 7777, 1301 Metro Manila, Philippines http://crossmark.crossref.org/dialog/?doi=10.1007/s10681-025-03498-4&domain=pdf Euphytica (2025) 221:90 90   Page 2 of 18 Vol:. (1234567890) Introduction Oryza sativa, L. and Oryza glaberima, Steud. are the two most widely grown rice species making it the second most produced crop globally (Breseghello and Coelho 2013; Dagallier et  al. 2021). Rice is among the cereals cultivated globally due to its agronomical and nutritional importance, providing food to more than half of the global population (Mohidem et  al. 2022; Wambugu et al. 2021). O. sativa is thought to have originated in Asia, while O. glaberrima origi- nated in Africa, especially in western Africa (Wam- bugu et al. 2021). Asian rice is more cultivated than African rice because of several advantages, including high yield and marketability (Adam et al. 2023; Brit- wum and Demont 2021). African rice is believed to have essential genes for tolerance to various stresses, including resistance to several diseases and tolerance to drought, flooding, and salinity (Sikirou et al. 2018; Wambugu et al. 2013; Wang et al. 2014). Between 1961 and 2019, the world rice productiv- ity increased from 1.86 to 4.66  tons/ha; however, in Africa, rice productivity remained among the low- est, with production of approximately 2 t/ha (Rahman and Zhang 2022). There are several causes for these low yields in Africa, including abiotic stresses such as salinity, nutrient deficiencies, toxicities, drought, and biotic stresses like pests and diseases. Salt stress, in particular, causes low crop yields (Hussain et  al. 2017) in dryland and coastal areas as it inhibits plant growth, hence decreasing crop productivity (Munns and Tester 2008). Salinity and abiotic stress greatly reduce rice yield in salt-affected areas, and due to the adversities caused by climate change, salt-affected areas are progressively increasing (Islam et al. 2021). While certain wild rice varieties exhibit relatively high tolerance levels, most elite cultivars are sensitive to salinity. This creates the need to explore variation in African rice to identify genes responsible for desir- able traits like drought tolerance, salinity tolerance, and higher yield, enabling their integration into new rice cultivars (Wambugu et al. 2019). Rice is a salt-sensitive crop; when plants are exposed to saline conditions, their growth and devel- opment are affected by excess salts (Hussain et  al. 2017). To overcome the effect of salinity, it is essen- tial to understand various processes such as metabo- lism, gene expression under salt stress conditions, and related physiological responses (Hasanuzzaman and Fujita 2022). Rice breeding programs should focus on developing varieties with tolerance to stresses, including salinity, to enhance resilience in affected environments. Since salinity stress can affect rice at various growth stages, from seedling establishment to grain filling, breeding efforts should prioritize stage- specific tolerance mechanisms (Sing et  al. 2021). Developing salt-tolerant cultivars will help mitigate yield losses in salt-affected regions where rice culti- vation is viable (Nayyeripasand et al. 2021). Genetic resources are significant for identifying genetic variation related to different traits of inter- est to plant scientists (Salgotra and Chauhan 2023). For instance, genome-wide association analysis has been applied in the study of various traits in several plant species such as soybean, cotton, and many other cereal crops including wheat, maize, barley, and rice (Sahito et al. 2024; Wang et al. 2021). The growing availability of genomic resources for various rice spe- cies has provided valuable genetic information and tools for developing new varieties with enhanced stress adaptability, including salinity tolerance (Song et al. 2018). The use of association analysis provides the ability to analyze many traits concurrently as a suitable alternative to the traditional quantitative trait analysis procedures that rely on biparental popula- tions to analyze quantitative traits (Liu et  al. 2016; Cui et  al. 2017). The association analysis technique enhances our ability to dissect and understand the genetic basis underlying various phenotypic traits (Xu et al. 2017). GWAS is based on statistical approaches for determining genotype–phenotype association and, as such, helps provide additional information to sci- entists by discovering chromosome regions together with the markers related to the phenotypes (Xu et al. 2017). Genome-wide association analysis has the advantage that, to determine the genes linked to the phenotypes, the whole genome of the plant or any organism is examined through a detailed mapping resolution, and the process is faster because there is no necessity to develop a mapping population (Rahim et  al. 2018). Also, technological improve- ment has dramatically enhanced the scale and speed with which the analysis can be conducted, hence promoting the use of association analysis to study quantitative traits of economic importance (Varsh- ney et  al. 2009). With traditional quantitative trait loci (QTL) analysis, the process involves developing Euphytica (2025) 221:90 Page 3 of 18  90 Vol.: (0123456789) a population and defining molecular markers associ- ated with the traits of interest (Hasna et al. 2021). The use of association analysis in O. sativa has gained more popularity, and it has been providing more helpful information on the biological mecha- nisms underlying key agronomic characters (Song et al. 2018; Wang et al. 2021). Numerous genome- wide association studies conducted in O. sativa have provided valuable insights when exploring the genes underlying salinity tolerance in rice (Ju et al. 2022). For example, Zhang et al. (2020) conducted a GWAS whereby seven QTLs linked to salinity tol- erance on chromosomes 1, 2, 5, and 9 were detected as the most significantly associated. Also, an asso- ciation analysis approach detected eight significant QTLs for the early seedling growth stage salt toler- ance (Li et al. 2019). Additionally, 4 QTLs related to stress tolerance were discovered through the GWAS approach (Kim and Kim 2023). While these studies may not be specific to O. glaberrima, they demonstrate the potential of GWAS for detecting the essential genes linked to different traits in rice. However, O. glaberrima is known to be less extensively studied through GWAS, with only a few studies conducted so far (Ntakirutimana et al. 2023). A small number of these studies focused on traits related to salinity tolerance (Meyer et al. 2016) and other traits such as panicle morphological traits, days to flowering, and rice yellow mottle virus (RYMV) (Cubry et al. 2020). Despite GWAS stud- ies receiving less research attention in O. glaber- rima, this method has proven crucial in uncovering the important chromosome regions related to traits essential for improving rice breeding programs in different areas (Yang et  al. 2018). To improve rice tolerance to salt stress, rice breeding programs are required to find important QTLs related to rice- sensitive growth stages and molecular markers for marker-assisted selection (Walia et al. 2005). Also, due to its enormous genetic potential for resilience to biotic and abiotic stress, O. glaberrima could be used as a genetic resource for rice improvement to increase and sustain productivity in salt-affected areas, as well as other abiotic stresses (Wambugu et  al. 2013). The primary purpose of this research was to examine the quantitative trait loci related to salinity stress tolerance in O. glaberrima by con- ducting marker-trait associations (MTAs) for traits related to salinity tolerance. Materials and methods Experimental materials and growing conditions The study utilized 335 O. glaberrima accessions obtained from the Africa Rice Centre (AfricaRice, Côte d’Ivoire). To assess salinity tolerance, we used known standard checks with well-documented salin- ity responses to serve as a reference for comparison instead of traditional controls. Both tolerant and sen- sitive check cultivars were included, FL478 and Pok- kali as tolerant checks and IRRI 154 and IR29 as sensitive checks. The O. glaberrima association map- ping panel was screened for salinity tolerance using a hydroponic solution following the protocol described by Yoshida et  al. (1976). The experiment was con- ducted at Sokoine University of Agriculture (SUA) from November 2022 to February 2023. The experi- ment was performed twice to gather data for various parameters. The hydroponic solution (Yoshida solu- tion) was used to evaluate rice genotypes for seedling stage salinity tolerance screening in a screen house. The average temperature during the experiments was 27.7 °C, and the average relative humidity was 62.6%. The experiment was conducted under screen house conditions to minimize the environmental factors that could largely influence the performance of the geno- types other than salt treatment. Seeding of the experimental materials The seeds were prepared initially by drying at 50 °C in a conventional oven to prevent the seeds from delayed germination and poor seedling establishment. After breaking seed dormancy, the seeds were washed with distilled water. To initiate the germination, seeds were moistened by soaking them in petri dishes cov- ered with filter papers for two days. Seeds were indi- vidually placed onto the seedling floats immersed in the distilled water for further development. The materials were arranged in an alpha lattice design and replicated in two replications. Seedlings were evalu- ated using the modified IRRI standard protocol (Gre- gorio et  al. 1997) for seedling-stage salinity screen- ing. Once the seedlings were established (three days later), a Yoshida nutrient solution was substituted for the distilled water. Salt stress treatment was imposed 7 days after seeding, starting with an electrical con- ductivity (EC) of 6 dS/m using sodium chloride salt. Euphytica (2025) 221:90 90   Page 4 of 18 Vol:. (1234567890) After three days, the solution’s electrical conductiv- ity (EC) was increased from 6 to 12 dS/m, while the pH of the growth media was thoroughly kept at 5.0 every day, with the nutrient solution being replaced once a week. The entries were assessed for tolerance to salinity 21 days following the treatment using the IRRI standard evaluation system (SES; IRRI 2013). Evaluation of salt tolerance To assess the salinity tolerance of the O. glaberrima accessions, several traits were evaluated, namely standard evaluation system (SES) scores for salt injury, root length (RL; cm), shoot length (SL; cm), shoot dry weight (SDW; g), root dry weight (RDW; g), and leaf samples were collected for determina- tion of Na+ and K+ concentration (mg/mL) and shoot Na+/K+ ratio. Shoot length was assessed by measur- ing plants from the base of the plant to the top of the longest leaf, while data for root length were collected by measuring the roots from the base of the root to the tip of the longest root. For the seedling salt injury score, the seedlings after 14 and 21 days of exposure to salt stress were evaluated using SES scores, with 1 as highly tolerant, 3 as tolerant, 5 as moderately toler- ant, 7 as sensitive, and 9 as highly sensitive (Table 1). Determination of shoot Na+ and K+ ion concentrations The shoot Na+ and K+ ion concentrations were determined using the samples collected from the greenhouse experiment. At 21  days after saliniza- tion, three seedlings were collected per replication and thoroughly washed 3 times with distilled water. The seedlings were then dried in an oven at 70 °C for three days. Dried shoots were weighed and chopped into small pieces, transferred into 10  ml falcon tubes, and digested by dissolving into 0.1 N acetic acid (C2H4O2) solution. Digestion was done by heating the mixture for two hours using a water bath at 90 °C. After heating, the solution underwent filtration using Whatman filter papers arranged in glass funnels. To estimate ion concentration, the filtrate was poured into flasks and diluted tenfold with deionized water; the solution was then used to determine sodium and potassium concentrations using a flame spectrophotometer (Sherwood Model 420). Genotyping A mini-core collection comprising 350 O. glaber- rima accessions, created from an initial pool of 3,130 accessions, was genotyped using Single- Nucleotide Polymorphisms (SNPs), and 335 acces- sions were selected for the current study. Deoxyri- bonucleic Acid (DNA) was extracted following the previously outlined protocol by Ndjiondjop et  al. (2022) for the genotyping process. Approximately 5 mg of fresh leaf tissue was collected from a sin- gle 3- to 4-week-old seedling and placed in 1.1 mL microtubes (Bioquote Limited). The samples were dried at 57  °C in a Binder FD53 E2 drying oven (Akribis Scientific Limited) with micronic sealing mats (NBS Scientific) covering the tubes and then sent to DArT Pty Ltd (http://​www.​diver​sitya​rrays.​ com/) for genotyping. The accessions used in the current study were genotyped by 9990 SNPs via the Diversity Arrays Technology, DArTseq-based genotyping by sequencing (GBS) technology (Kil- ian et al. 2012). Table 1   Modified Standard Evaluation Score (SES) of visual salt injury at seedling stage (IRRI 2013) adapted from Gregorio et al. (1977) SES Description Tolerance 1 No symptoms of salt injury, with no visual signs of salt stress, only the old leaves show white tips Very high 3 Slight symptoms of salt injury, especially at the leaf tips, and some older leaves become whitish, partially High 5 Moderate symptoms of salt injury, retarded growth; most old leaves are severely injured, and few young leaves elongating Moderate 7 Severe symptoms of salt injury, poor growth; most leaves dried; only a few young leaves still green Susceptible 9 Severe symptoms, almost all plants are dead or dying Highly susceptible http://www.diversityarrays.com/ http://www.diversityarrays.com/ Euphytica (2025) 221:90 Page 5 of 18  90 Vol.: (0123456789) SNP data filtering SNP data filtering was done using TASSEL (Trait Analysis by aSSociation, Evolution, and Linkage) 5.2.94 software (Bradbury et  al. 2007) to clean the SNP data and make them suitable for downstream genome-wide association studies (GWAS). SNP data with unknown chromosome positions were removed from the dataset. The 9065 polymorphic SNPs spread across the 12 chromosomes were filtered based on the following criteria: (1) minor allele frequency (MAF) ≥ 0.05, (2) missing data ≤ 20%, (3) maximum heterozygosity set to 0.2. After filtering, 6103 SNP markers were retained for genome-wide association analysis. Data analysis Statistical analysis of phenotypic traits The analysis was conducted on the traits associated with salinity tolerance using R software version 4.3.2, developed by the R core team (2023). To perform the association analysis, phenotypic data of traits related to salinity stress tolerance were averaged as mean values from two replications. Table 2 provides an overview of each trait’s mean, standard deviation, coefficient of variation, and minimum and maximum values. R package Corrplot (Wei and Simko 2021) was utilized to compute the correlations between the measured traits, and the results are presented in (Fig. 1). Estimation of traits heritability The mean of traits was used to estimate heritability. For each trait, broad-sense heritability was deter- mined, as the ratio between the genetic variance and total phenotypic variance, and the heritability in the broad sense was computed as where: H represents the heritability while the geno- typic component, genotype-by-environment inter- action, and error variance are denoted as�2 g ,�2 ge ,�2 e respectively. Component e represents the number of environments, and r the replicates within each envi- ronment, as described (Fehr 1987). Genome‑wide association analysis Association analysis was performed for all measured traits using genomic association and prediction inte- grated tool (GAPIT), a package created and operated in R environment (Lipka et al. 2012) using the phe- notypic and genotypic (SNP) data sets. To perform the analysis, we used the 335 African rice acces- sions with phenotypic data and the associated geno- typic data and excluded the accession lacking associ- ated genotype data. The analysis used a linear mixed model (Zhang et  al. 2010), as implemented in the GAPIT R package (Lipka et al. 2012). To correct for population structure, the kinship matrix (K) containing the random effects was com- puted automatically using the Van Raden method, and the first three (3) principal components (PCs) of a PCA of genomic data were used as the Q matrix. The MLM analysis of the equation used by GAPIT: Y = Xβ + Zu + e. where Y stands for the vector of observed phe- notypes, β is an unknown vector comprising fixed effects, including the genetic marker, popula- tion structure (Q), and the intercept; u denotes an unknown vector of random additive genetic effects H = � 2 g �2 g + �2 ge e + �2 e er Table 2   Number of accessions with different SES scores after subjecting to salt stress for 21 days Highly tolerant Tolerant Moderate tolerant Susceptible Highly suscepti- ble SES score 1–2 3–4 5–6 7–8 9 No of accessions 0 4 17 148 166 Tolerant checks 0 2 0 0 0 Susceptible checks 0 0 0 0 2 Euphytica (2025) 221:90 90   Page 6 of 18 Vol:. (1234567890) contributed by multiple background QTLs for indi- viduals or lines; X and Z are the known design matrices; and e represents the unobserved vector of residuals. Manhattan plots were generated using the CMplot package in R version 3.5.3 to visualize the significant MTAs for each trait (Yin et  al. 2021). The p values were plotted as − log10(p) to cre- ate the Manhattan plots. Linkage disequilibrium (LD) analysis between significant (p < 0.001) SNPs was conducted using the Tassel 5.2.94 software (Bradbury et  al. 2007) to determine whether the significant SNPs are in strong LD (high r2) and thus likely tagging the same locus, rather than independ- ent associations. Variances explained by significant SNP marker Utilizing various variance components from the genome-wide association study results, the proportion of variance (PVE) explained by the markers was esti- mated for each SNP. The methodology for doing the estimation was based on the approach outlined (Kumar et al. 2021) and as described in the following equation. pVE(SNP) 2 × ( beta2 ) ×MAF × (1 −MAF) 2 × ( beta2 ) ×MAFX(1 −MAF) + ( ( SED(beta))2 ) × 2 × N ×MAF × (1 −MAF) Fig. 1   Boxplot showing mean performance of the geno- types for root dry weight (R DW), shoot dry weight (SDW), root length (RL), shoot length (SL), standard evaluation score (SES), and Na/K ratio. Black horizontal lines in the middle of the boxes are the median values of a respective trait Euphytica (2025) 221:90 Page 7 of 18  90 Vol.: (0123456789) where N represents the panel’s sample size, beta denotes the effect of the SNP markers, SE (beta) denotes the standard error of the SNP effects, and MAF denotes the minor allele frequency for the SNP markers of interest. Results Response to salinity tolerance in African rice accessions We assessed the tolerance to salinity of O. glaberrima accessions using the hydroponic system with an electri- cal conductivity (EC) of 12 dS/m using sodium chlo- ride salt (NaCl). Eight traits related to salt tolerance were assessed. The O. glaberrima genotypes showed a wide range of variability in salt tolerance. Based on the standard evaluation system (SES) scores, some accessions were identified as the most tolerant geno- types, though the tolerant checks (FL478 and Pokkali) were still the most tolerant. Several moderately toler- ant accessions were also identified (Table 2), with SES scores ranging between 5 and 6. Generally, the O. gla- berrima genotypes showed a wide range of variability in salt tolerance from highly susceptible to tolerant. Table 3 summarizes the means, standard deviation (SDs), minimum, maximum values, and coefficients of variation (CV) of the eight traits related to salin- ity stress tolerance. Broad sense heritability (H2) was generally high for Na+ ions concentration, standard evaluation scores (SES), and root length. In contrast, shoot length and shoot dry weight had relatively high heritability. Nevertheless, traits like root dry weight, K+ ions concentration, and Na+/K+ ratio in the shoot showed low heritability. Phenotype variability and correlation analysis among the traits We performed boxplot analysis for root dry weight (RDW), shoot dry weight (SDW), shoot length (SL), Root Length (RL), standard evaluation score (SES), and Na/K ratio. The results suggest that most geno- types exhibit values near the mean, with a few outli- ers in each trait (Fig. 1). The SES values are skewed toward the higher end, with most genotypes cluster- ing around a high mean value. The lower whisker is shorter, and some outliers are below the main distri- bution, suggesting that while most genotypes exhib- ited significant salt stress symptoms, a few performed better under salinity conditions. Generally, the pres- ence of multiple outliers, both high and low, for some of these traits indicates some genetic variation for salinity tolerance. Correlation analyses for all rice traits related to salinity tolerance, including SL, RL, SDW, RDW, Na, SES scores, and Na/K ratio (Fig. 2). The results indi- cated significant positive correlations existed among SL, RL, SDW, and RDW. Moreover, the Na/K ratio significantly correlated with Na+ ion concentration. Further, standard evaluation scores (SES) demon- strated a substantial association between Na+ (0.48) concentration and the Na+/K+ ratio (0.44). Table 3   Statistical data and heritability estimate of 8 salt-tolerant associated traits Trait Minimum Maximum Mean Standard deviation (SD) Coefficient of variation (CV, %) H2 SIS (SES) 4.25 9.00 8.5 0.84 1.00 0.69 RL (cm) 13.93 33.18 25.2 3.06 12.00 0.66 SL (cm) 2.33 7.2 3.79 4.79 12.6 0.58 SDW (cm) 3.40 1.95 1.04 2.89 28.0 0.55 RDW (cm) 1.47 7.13 3.82 9.98 26.0 0.41 Na+ 32.7 93.8 60.4 11.1 18.0 0.79 K+ 12.1 28.3 19.2 2.201 11.0 0.24 Na/K ratio 1.65 6.6 3.43 0.83 24.0 0.35 Euphytica (2025) 221:90 90   Page 8 of 18 Vol:. (1234567890) Fig. 2   Trait correlation analysis: blue color = posi- tive correlations, red color = negative correla- tions, and white color = cor- relations not significantly different from 0 as repre- sented by white boxes. The legend color on the right side shows the correlation coefficient and the matching colors. SL stands for shoot length; SIS; salt injury score (SES) stands for standard evaluation system; RL stands for root length; RDW for root dry weight; SDW for shoot dry weight; NaKratio for Na + /K + ion ratio; K for Potassium and Na is for Sodium Table 4   Number of markers and significant markers (p < 0.05) for each trait Chromosome Number of markers Number of significant markers SES SL RL SDW RDW N a K Na/K ratio 1 641 49 38 35 24 36 26 44 35 2 654 42 25 26 29 33 37 28 32 3 699 30 44 24 38 30 27 26 27 4 502 26 23 21 14 30 27 21 26 5 444 23 11 19 17 29 19 23 26 6 536 23 32 22 25 20 13 21 21 7 539 24 37 26 15 23 35 24 29 8 394 11 24 28 21 20 12 13 16 9 358 28 19 20 26 17 18 22 17 10 365 19 20 21 34 23 49 28 19 11 513 16 18 24 22 22 17 34 28 12 458 19 26 19 25 24 33 23 22 Total 6103 310 317 285 290 307 313 307 298 Euphytica (2025) 221:90 Page 9 of 18  90 Vol.: (0123456789) Association analysis for rice seedling stage traits related to salinity tolerance Markers associated with salinity tolerance In this study, the significance threshold for marker- trait associations (MTAs) was set at P = 0.001 to mini- mize the risk of false-positive results. The association analysis results between the SNP markers and salt- tolerant-related traits have been presented in Manhat- tan plots (Fig. 2a–f). Table 4 presents the number of markers associated with each chromosome and signif- icant markers for each trait and chromosome. The results indicated that chromosome 3 has the highest SNPs associated with the salinity tolerance traits, while chromosome 9 has the lowest number. Shoot length was significantly associated with 317 SNPs, the highest number compared to other char- acteristics tagging all 12 chromosomes. At the same time, root length (RL) had the lowest number of SNPs, significantly associated with salinity tolerance at different chromosomes. Table 5   SNPs significantly (P value < 0.001) associated with different traits related to salinity tolerance Key: Chr = Chromosome, Pos = position, Na/K = Na/K ratio Trait Best SNP Markers Chr Pos P value MAF Effect Log10 SL 3,054,771|F|0–40:T > C-40:T > C 3 18.95 0.0006 0.05 − 2.08 3.22 SL 5,403,201|F|0–20:A > C-20:A > C 6 3.21 0.0007 0.06 − 1.87 3.18 SL 3,993,239|F|0–29:G > T-29:G > T 1 29.12 0.0002 0.09 − 2.72 3.65 SL 3,444,079|F|0–41:A > G-41:A > G 5 18.93 0.0005 0.10 − 1.46 3.30 SL 9,746,845|F|0–13:A > G-13:A > G 7 16.77 0.0001 0.16 1.32 3.86 SL 3,994,492|F|0–54:A > G-54:A > G 12 16.93 0.0005 0.23 − 1.15 3.33 SES 5,141,121|F|0–25:T > C-25:T > C 6 17.01 0.0009 0.09 − 0.30 3.04 SES 5,399,172|F|0–60:A > G-60:A > G 4 1.11 0.0003 0.21 0.21 3.48 SES 5,392,986|F|0–23:T > C-23:T > C 7 8.59 0.0005 0.24 0.19 3.34 SES 3,450,778|F|0–62:A > G-62:A > G 7 17.06 0.0001 0.42 0.19 3.96 SDW 3,453,106|F|0–10:A > G-10:A > G 11 7.90 0.0004 0.10 9.47 3.38 SDW 3,061,050|F|0–38:G > A-38:G > A 10 12.46 0.0006 0.15 − 8.74 3.22 SDW 9,757,915|F|0–9:T > A-9:T > A 7 17.01 0.0009 0.20 − 7.08 3.07 SDW 5,395,205|F|0–24:A > G-24:A > G 9 8.64 0.0001 0.22 − 7.89 3.86 SDW 3,053,456|F|0–14:A > G-14:A > G 11 14.34 0.0004 0.49 6.85 3.45 RL 21,590,207|F|0–65:A > T-65:A > T 4 14.82 0.0005 0.10 1.18 3.32 RL 3,996,281|F|0–29:T > C-29:T > C 7 19.30 0.0004 0.12 1.16 3.42 RL 19,319,924|F|0–55:C > T-55:C > T 7 19.81 0.0008 0.12 1.11 3.08 RL 3,439,519|F|0–61:A > G-61:A > G 7 19.51 0.0004 0.16 1.20 3.44 RL 3,450,440|F|0–16:G > A-16:G > A 7 19.52 0.0004 0.16 1.16 3.41 RL 3,771,779|F|0–23:A > T-23:A > T 8 3.04 0.0007 0.17 − 1.12 3.19 RL 9,753,558|F|0–66:A > G-66:A > G 10 16.79 0.0008 0.37 0.63 3.07 RL 3,751,944|F|0–37:C > A-37:C > A 1 22.06 0.0010 0.41 0.63 3.02 RL 3,994,983|F|0–8:T > C-8:T > C 8 0.60 0.0008 0.49 − 0.65 3.10 RDW 5,406,485|F|0–65:G > T-65:G > T 12 9.94 0.0007 0.14 5.86 3.13 RDW 3,057,760|F|0–41:C > T-41:C > T 12 15.22 0.0001 0.24 2.81 4.05 RDW 9,746,538|F|0–14:T > C-14:T > C 1 10.75 0.0006 0.26 − 2.32 3.21 RDW 5,382,049|F|0–56:T > C-56:T > C 10 0.01 0.0008 0.31 2.45 3.08 Na/K ratio 9,759,917|F|0–22:G > A-22:G > A 1 5.02 0.0007 0.06 − 0.33 3.15 Na/K ratio 5,383,035|F|0–11:C > A-11:C > A 11 5.80 0.0009 0.10 − 0.25 3.03 Na/K ratio 3,438,412|F|0–21:C > A-21:C > A 11 4.50 0.0007 0.13 − 0.25 3.15 Na/K ratio 9,754,340|F|0–45:T > G-45:T > G 10 9.99 0.0010 0.14 − 0.42 3.02 Na/K ratio 5,375,716|F|0–49:A > G-49:A > G 4 7.74 0.0008 0.34 − 0.20 3.11 Na/K ratio 3,440,687|F|0–32:C > G-32:C > G 4 7.70 0.0006 0.35 − 0.19 3.19 Euphytica (2025) 221:90 90   Page 10 of 18 Vol:. (1234567890) Marker trait significant associations This study has identified 34 significant marker-trait associations (MTAs) related to salinity tolerance traits (Table  5). To determine marker-trait associa- tions (MTAs), (− log10 (p value) ≥ 3.0) was used. All the measured traits had significant MTAs, and the distribution of the associated SNPs varied consider- ably within the genome. For instance, chromosome 7 harbored the highest number of MATs by having 8 SNP markers associated with different traits. Only chromosome 2 recorded none of the MTAs for the measured traits. The significant SNP markers were selected mainly based on − log10 (p value) ≥ 3.0 and by considering the p values (< 0.001) as highly sig- nificantly associated SNP for each trait. Phenotypic variance explained by  highly significant SNP markers  Phenotypic variance (%PVE) for the highly significant SNP markers was determined, as summarized in Table 6, together with the %PVE, the minor allele frequency (MAF), and the allele effect values for the highly significant SNPs. Standard evaluation scores (SES) and Na/K ratio  The standard evaluation scores (SES) indicated that three chromosome regions were linked to four (4) SNP markers located on chromosomes 4, 6, and 7, which were significantly associated with the SES (Fig. 3e). The markers explained between 3.23 and 4.37% of the phenotypic differences for standard evaluation scores (Table 6). The SNP marker with stronger associations was observed at 17.06 Mbp on chromosome 7 of the analyzed genome (Table 5). For the sodium to potas- sium (Na+/K+) ratio, four chromosomes were linked to six (6) SNP markers located on chromosomes 1, 4, 10, and 11, shown in (Table 5, Fig. 3f). The pheno- typic differences expressed by the SNPs ranged from 3.21% to 3.43% for variation in Na+/k+ ratio (Table 6). The best SNP was observed on chromosome 4 at 7.70 Mbp, which explained about 3.43% of the phenotypic variation for the trait (Table  5). The pairwise link- age disequilibrium (LD) analysis revealed that the LD between SNPs at Chromosome 4 at 7.74 Mb and Chromosome 4 at 7.70 Mb is very high (R2 = 0.86), implying that the alleles at these loci are inherited most of the time. Shoot and  root measurements  Shoot length (SL) notable association was identified on six (6) sections of chromosomes linked to 6 different SNP on chro- mosomes 1, 3, 5, 6, 7, and 12 (Fig. 3a). The signifi- cant SNPs linked to shoot length ranged from 3.41% to 4.25%, of the observed variation in phenotype dif- ferences for shoot length (Table 6). The best associ- ated SNP was observed on chromosome 7 at position 16.77 Mbp (Table 5). For root length (RL), we identi- fied nine highly significant SNPs linked to chromo- somes 1, 4, 7, 8, and 10 (Table 6, Fig. 3b). The SNPs explained 0.321 to 3.73% of the differences in root length. The best SNP was observed on chromosome 7 at 19.51  Mbp (Table  5). Also, there were several MTAs on chromosome 7 related to root length how- ever through the LD analysis, only two SNP markers had moderate LD, Chromosome 7 (19.8Mbp) and Chromosome 7 (19.3Mbp) (R2 = 0.40). Table 6   The number of MTAs for each trait, minor allele frequency, the allelic effect, and the phenotypic variance explained by the SNPs Trait The number of highly significant SNPs MAF Allelic effect %PVE SES 4 0.085–0.418 (− 0.303) to 0.209 3.23–4.37 SL 6 0.054–0.232 (− 2.722) to 1.317 3.41–4.25 RL 9 0.096–0.4102 (− 1.120) to 1.195 3.21–3.73 SDW 5 0.105–0.491 (− 8.7431) to 9.473 3.32–4.24 RDW 4 0.139–0.305 (− 2.316) to 5.857 3.28–4.49 Na/k Ratio 6 0.061–0.338 (− 0.419) to (− 0.198) 3.20–3.43 Euphytica (2025) 221:90 Page 11 of 18  90 Vol.: (0123456789) Fig. 3   Manhattan plots showing MTAs for SNPs associated with salinity tolerance traits. a shoot length (SL), b root length (RL), c root dry weight (RDW), d shoot dry weight (SDW), e standard evaluation score (SES), and f Na/K ratio. The horizontal line shows each trait’s significance threshold, p < 0.001. X and Y axes are the position of SNPs across the 12 chromosomes of rice and the p values on a log scale at each locus, respectively. Quantile– quantile (Q–Q) plots g–l showing estimated − log10 (p) values versus observed values for respective traits Euphytica (2025) 221:90 90   Page 12 of 18 Vol:. (1234567890) Shoot and root dry weight measurements  The results for shoot dry weight indicated that the highly signifi- cant SNPs linked to this trait were found on chromo- somes 7, 9, 10, and 11 (Table 5, Fig. 3d). The SNP markers described between 3.27–4.25% of the pheno- type differences for shoot dry weight (Table 6). Among the highly significant associated SNPs, the strongest SNPs were positioned at 8.64 Mbp on chromosome 9 of the genome (Table  5). Significant SNP mark- ers for root dry weight were found on chromosomes 1, 10, and 12 (Table 5, Fig. 3c). The SNPs described between 3.29 and 4.49% of the phenotype variation for root dry weight (Table 6). Among highly significant SNPs, the best SNP was located on chromosome 12 at 15.22 Mbp (Table 5). Discussion Breeding rice for adaptation to salinity stress has faced limitations due to the limited availability of salinity-tolerant rice accessions (Qin et al. 2020). Pre- vious studies have shown that only a few rice cultivars capable of withstanding salinity stress are used by dif- ferent breeding programs, primarily those originating from the O. sativa subspecies. This study used 335 accessions of O. glaberrima and 6,103 SNP markers to perform a genome-wide association study to find essential chromosome segments and the associated SNP markers related to salinity stress tolerance. Numerous indicators measure salinity tolerance in crop plants, including rice. However, the direct tech- nique quantifies Na+ and K+ ions concentration and their Na+/K+ ratio values (Yuan et al. 2022) in plant tissues. In addition, some useful and commonly used indicators, including relative dry weight, standard Fig. 3   (continued) Euphytica (2025) 221:90 Page 13 of 18  90 Vol.: (0123456789) evaluation scores, relative root and shoot length, and the overall plant biomass, are among them (Javed et al. 2011). However, the standard evaluation scores (SES) for salt injury serve as a comprehensive param- eter for plant salinity stress response, offering a pri- mary clear insight into the tolerance level exhibited by a specific genotype (Amoah et  al. 2020). Thus, understanding the damaging impacts of salinity on rice growth and yield while concurrently explor- ing various genotypes’ capabilities for salinity toler- ance could help to identify significant loci for use in breeding. The boxplots presented show the distribution of key traits related to salinity tolerance in rice. The var- iability observed in traits across genotypes suggests that some rice lines exhibit higher salinity tolerance (lower SES, higher biomass, and better ion regula- tion) while others are highly sensitive. Also, the fit- ness-related parameters, such as dry weights, root and shoot lengths, had heritability estimates varied from 41 to 66%, indicating relatively high heritability for these traits. High heritability shows that a substantial portion of the differences in phenotype was attribut- able to variation in inheritance, thereby presenting the prospects of improving the traits through selec- tive breeding. When the evaluated parameters show a wide range of heritability estimates, from low to high, this suggests that the difference in appearance of the traits was due to the impacts of salinity (Bhadru et al. 2012). Some reported heritability values for fitness- related traits usually ranged from 10 to 20% in natural populations (Visscher et al. 2008). The standard evaluation scores (SES) for salt injury showed positive correlations to Na+ ion con- centration and Na/K ratio while displaying an inverse relationship with root and shoot lengths. The correla- tion between the Na /K ratio and SES was particularly outstanding, with the ratio showing a stronger cor- relation to Na+ ion concentrations than K+ ion con- centrations. This suggests that the overall response of the plant shoot to salinity is primarily determined via sodium absorption and subsequently transfer to the shoots, as emphasized by Pires et al. (2015). A simi- lar trend has been previously reported by Leon et al. (2015) and Okeshkumar et al. (2023). The negative correlations between SES and various rice growth parameters emphasize that sodium ion accumulation damages plant tissues for plants grown in salt-prone environments. Meanwhile, the positive and significant correlations observed in parameters such as root and shoot dry weights and root and shoot lengths collectively indicate the importance of these traits in adaptation to seedling stage salinity stress. The standard evaluation scores for salt injury, Na+, and K+ ions concentrations have been established to be reliable measures for evaluating rice’s ability to withstand salt stress (Babu et  al. 2014; Krishnamur- thy et al. 2014). This system provides a standard indi- cator, providing valuable insights into rice’s capacity to tolerate saline environments. These results align with some of the reported findings that the correlation between growth-related traits implies the presence of genetic variations for seedling stage salinity tolerance (Bimpong et al. 2014; Rahim et al. 2018). This study has identified 34 MTAs for salinity stress tolerance in the O. glabberima population. Although the phenotypic variances explained by the identified MTAs are low, they align with the poly- genic inheritance commonly reported for salinity tolerance in rice. This means that most of the trait’s salinity tolerance is contributed collectively by many minor-effect loci, and environmental factors may have further influenced the observed phenotypes. This is because salinity tolerance has been reported to be a complex trait, influenced by multiple quantitative trait loci distributed across the genome, each contribut- ing variably to phenotypic expression (Munns et  al. 2006). The SNPs linked to the Na+/K+ ratio provided a range of variation for phenotype differences between 3.20 and 3.43% for the observed difference in this trait. The lowest values within the given range signify that the SNP markers have a minimal impact on the phenotype, while SNPs at the higher end contribute slightly more but still describe a small percentage of the overall variation. Only one significant SNP marker was detected on chromosome 1 at 5.02 Mbs, on the same chromosome as the previously identified candidate gene (HKT1;5) for Na+ exclusion detected on chromosome 1, having a role in Na⁺ exclusion for enhanced salinity tolerance (Platten et  al. 2023). Also, a well-known QTL (Saltol) governs salinity tol- erance at the seeding stage in O. sativa (Asian rice) was reported on the same chromosome. The Saltol QTL is well-known for conferring salinity tolerance by regulating key physiological processes, such as ion homeostasis, particularly sodium and potassium bal- ance, under salt-stress conditions (Sing et  al. 2018). Euphytica (2025) 221:90 90   Page 14 of 18 Vol:. (1234567890) Mazumder et  al. (2024) evaluated interspecific RILs derived from O. glaberrima and found potential novel QTLs for seedling salinity tolerance, distinct from the known Saltol QTL in O. sativa. Other SNPs associ- ated with the Na+/K+ ratio were identified on other chromosomes, which could indicate that other genetic regions contribute to adaptation to salinity beyond what is accounted for by the major QTL (Saltol) (Krishnamurthy et al. 2020). Furthermore, the SNP markers associated with standard evaluation scores (SES) for salt injury explained 3.23–4.37% of the phenotypic variation. The small proportion of variation explained by these significant markers suggests that they account for only a limited fraction of the phenotype differ- ences within this population. As opposed to the current study, previous studies, such as Asif et  al. (2022), reported QTLs for salt injury scores (SIS- SES) with a phenotypic variance explained (PVE) between 22 and 34%, although these findings were based on populations developed from Indica-Japon- ica genetic backgrounds, which differ from the genetic composition of the current study. The lower PVE observed in this study may indicate the poly- genic nature of salinity tolerance within this popula- tion, as supported by previous studies that identified multiple loci contributing to this trait. Notably, the identified marker-trait associations (MTAs) for SES scores were located outside the Saltol QTL region, suggesting the involvement of additional genes in salinity tolerance (Bizimana et  al. 2017; Bimpong et al. 2014). The shoot length (SL) results revealed signifi- cant associations in three distinct genomic regions, with the variability explained by these SNP mark- ers ranging from 3.41 to 4.25%. The explained vari- ance suggests that these loci are minor- rather than major-effect QTLs. Previous studies have reported QTLs with more significant effects on shoot length; however, these studies were conducted in Indica- Japonica populations with slightly different genetic backgrounds from the current study. For instance, Takehisa et  al. (2004) identified 12 QTLs for shoot length in O. sativa, explaining between 12 and 30% of the phenotypic variation. Similarly, Jahan et  al. (2020) reported QTLs with PVE ranging from 6.8 to 18.20%. All these findings provide an insight into the genetic variation of shoot length in rice. Root length (RL) associated SNP markers collec- tively described the modest proportion of variation in phenotype ranging between 3.21 and − 3.73%. While these may appear small compared to other traits, they still highlight the genetic base influencing variation for this trait. In some studies, some QTLs associ- ated with root length (RL) explained high phenotypic variation between 5.9 and 23.7% for this trait (Jahan et  al. 2020). Root dry weight (RDW) SNP markers collectively explained 3.28–4.49% of the phenotype variation. This suggests that genetic factors within the specific genomic regions play a crucial role in determining variation for root dry weight. Shoot dry weight (SDW), another important trait related to salinity tolerance, the related SNP markers explained between 3.32 and 4.24%, which is a low variation in contrast to some of the earlier findings reported in O. sativa by Jahan et al. (2020), which showed variation caused shoot dry weight extending between 11.2 and 16.3%. This study provides valuable insights into SNP associations with salt-tolerance traits; however, some limitations should be noted. The sample size of 335 individuals, though sufficient for initial GWAS map- ping, may reduce the ability to detect loci with small to moderate effects. Also, low marker density may lead to failure in identifying critical regions. This can be improved through increased sample size and high-density SNP arrays for greater statistical power to improve result robustness (Wray et al. 2013). Addi- tionally, to enhance the robustness and reliability of the identified MTAs, further analysis employing alter- native GWAS methods such as Multi-Locus Mixed Model (MLMM) (Segura et  al. 2012), Compressed Mixed Linear Model (CMLM) (Zhang et  al. 2010), and Bayesian-information and Linkage-disequilib- rium Iteratively Nested Keyway (BLINK) (Huang et  al. 2019). These models have demonstrated bet- ter sensitivity and power to detect associations while accounting for population structure and relatedness. Thus, most identified SNPs are linked to shoot- related traits, emphasizing their significance for seed- ling-stage salinity adaptation in O. glaberrima. More- over, our findings indicated that certain SNP markers were situated within the known genomic regions with QTLs associated with salinity tolerance identified before in O. sativa (Koyama et  al. 2001; Lin et  al. 2004; Takehisa et  al. 2004; Thomson et  al. 2010; Jahan et al. 2020). However, several newly identified Euphytica (2025) 221:90 Page 15 of 18  90 Vol.: (0123456789) SNPs were found in distinct genomic regions, poten- tially harboring novel sources of salinity tolerance QTLs not previously documented, especially for the O. glaberrima population. This novelty could be attributed to the utilization of rice accessions possess- ing unique sources of salinity tolerance. These find- ings provide us with a framework for understanding the potential of the new chromosome regions and the SNP markers linked to salinity stress tolerance in African rice. Plant breeders could benefit from the newly identified SNP markers for marker-assisted selection for rice crop development. Conclusion and recommendations This study used 335 accessions of O. glaberrima and 6103 SNP markers to perform a genome-wide association study. Twenty-one genotypes cat- egorized as tolerant to moderate tolerant represent promising materials for salinity tolerance breeding. In total, 34 MTAs were detected for measured traits by the mixed linear model (MLM). The results from this research work can, therefore, be utilized for rice crop development. However, validating MTAs iden- tified in this study is critical for confirming their utility in breeding programs. This can be achieved through independent GWAS in more extensive and more diverse populations, fine mapping to narrow down candidate regions, and functional studies to characterize the biological roles of identified loci. Additionally, incorporating advanced GWAS methods such as BLINK or MLMM in future anal- yses will enhance the robustness and reliability of these findings. Only a few studies have been con- ducted on salinity tolerance in O. glaberrima com- pared to O. sativa, and those few studies did not provide specific candidate genes for salinity toler- ance. Therefore, future studies should explore QTLs within similar genetic backgrounds to validate these findings further. Thus, the identified SNPs and chromosome regions in the current study could be valuable resources for further study of the candidate genes and develop markers for salinity tolerance screening in African rice. Acknowledgements  The support from the climate-smart African rice research project for financial support through DANIDA is highly appreciated. Tanzania Agricultural Research Institute (TARI) is highly acknowledged and recog- nized for granting study leave to the first author. Author contributions  NTM designed the experiment and worked on data collection, analysis, and drafting of the manu- script, NK guided on experimental design and worked on the manuscript, SN-M, worked to revise the manuscript, SD and AI guided phenotyping, data analysis, and manuscript revi- sion, AB assisted on experimental design and manuscript revi- sion and M-NN germplasm collection and genotyping of the accessions. Funding  The Climate-Smart African Rice Research Project, DANIDA, Grant No. 19-3-KU supported this research work. Data availability  The datasets for the current study are avail- able from the corresponding author upon request. Declarations  Conflict of interest  The authors declare no competing inter- ests. References Adam H, Gutiérrez A, Couderc M, Sabot F, Ntakirutimana F, Serret J (2023) Genomic introgressions from African rice (Oryza glaberrima ) in Asian rice (O sativa) lead to the identification of key QTLs for panicle architec- ture. BMC Genomics 24:587. https://​doi.​org/​10.​1186/​ s12864-​023-​09695-6 Amoah A, Akromah R, Kena W, Manneh B, Dieng I, Bim- pong I (2020) Mapping QTLs for tolerance to salt stress at the early seedling stage in rice (Oryza sativa L) using a newly identified donor ‘Madina Koyo.’ Euphytica 216:156. https://​doi.​org/​10.​1007/​s10681-​020-​02689-5 Asif S, Kim E, Jang Y, Jan R, Kim N, Asaf S, Farooq M, Kim K (2022) Identification of the OsCML4 Gene in Rice related to salt stress using QTL analysis. Plants 11:2467. https://​doi.​org/​10.​3390/​plant​s1119​2467 Babu NN, Vinod KK, Krishnan SG, Bhowmick PK, Vanaja T, Krishnamurthy TSL, Nagarajan M, Singh NK, Prabhu KV, Singh AK (2014) Marker-based haplotype diversity of Saltol QTL in relation to seedling stage salinity toler- ance in selected genotypes of rice. Indian J Genet Plant Breed 74:16–25. https://​doi.​org/​10.​5958/j.​0975-​6906.​ 74.1.​003 Bhadru D, Rao VT, Mohan YC, Bharathi D (2012) Genetic variability and diversity studies in yield and its compo- nent traits in rice (Oryza sativa L.). SABRAO J Breed Genet 44:129–137 Bimpong KI, Baboucarr M, Diop B, Ghislain K, Sow A, Amoah NK, Gregorio G, Singh RK, Ortiz R, Wop- ereis W (2014) New quantitative trait loci for enhanc- ing adaptation to salinity in rice from Hasawi, a Saudi landrace into three African cultivars at the reproductive stage. Euphytica 200:45–60. https://​doi.​org/​10.​1007/​ s10681-​014-​1134-0 https://doi.org/10.1186/s12864-023-09695-6 https://doi.org/10.1186/s12864-023-09695-6 https://doi.org/10.1007/s10681-020-02689-5 https://doi.org/10.3390/plants11192467 https://doi.org/10.5958/j.0975-6906.74.1.003 https://doi.org/10.5958/j.0975-6906.74.1.003 https://doi.org/10.1007/s10681-014-1134-0 https://doi.org/10.1007/s10681-014-1134-0 Euphytica (2025) 221:90 90   Page 16 of 18 Vol:. (1234567890) Bizimana JB, Yan J, Yin X (2017) Mapping QTLs using a novel source of salinity tolerance from Hasawi and their interaction with environments in rice. Rice 10(1): 47. https://​doi.​org/​10.​1186/​s12284-​017-​0186-x Bradbury PJ, Zhang Z, Kroon DE, Casstevens TM, Ramdoss Y, Buckler ES (2007) TASSEL: software for association mapping of complex traits in diverse samples. Bioin- formatics 23:2633–2635. https://​doi.​org/​10.​1093/​bioin​ forma​tics/​btm308 Breseghello F, Coelho ASG (2013) Traditional and modern plant breeding methods with examples in rice (Oryza sativa L.). J Agric Food Chem 61:8277–8286. https://​ doi.​org/​10.​1021/​jf305​531j Britwum K, Demont M (2021) Tailoring rice varieties to consumer preferences induced by cultural and colonial heritage: Lessons from New Rice for Africa (NERICA) in The Gambia. Outlook Agric 50:305–314. https://​doi.​ org/​10.​1177/​00307​27021​10197​58 Cubry P, Pidon H, Ta KN, Tranchant-dubreuil C, Thuillet A, Holzinger M, Adam H, Kam H, Chrestin H, Ghesquière A, François O, Sabot F, Vigouroux Y, Albar L, Jouannic S (2020) Genome-wide association study pinpoints key agronomic QTLs in African rice Oryza glaberrima. Rice 13:66. https://​doi.​org/​10.​1186/​s12284-​020-​00424-1 Cui Y, Wang Q, Yin X (2017) Molecular mechanisms of salin- ity tolerance in rice. The Crop Journal 5(5): 471–481. Dagallier B, Kagoshima A, Komoto N, Kearns P (2021) Revised consensus document on the biology of rice (Oryza sativa L.). OECD Ser Harmon Regul Oversight Biotechnol 70:1–89 Fehr WR (1987) Principles of cultivar development. Theory and technique, vol 1. Collier Macmillan Publisher, Lon- don, pp 95–105 Gregorio G, Senadhira D, Mendoza R (1997) Screening rice for salinity tolerance. IRRI discussion paper series no. 22 Hasanuzzaman M, Fujita M (2022) Plant responses and toler- ance to salt stress: physical and molecular interventions. Int J Mol Sci 23:4810. https://​doi.​org/​10.​3390/​ijms2​30948​ 10 Hasna M K, Sultana S, Rahman MS (2021) Screening of Rice (Oryza sativa L.) Genotypes for Salinity Tolerance Based on Morphological and Physiological Traits. Plants 13(7): 1036. https://​doi.​org/​10.​3390/​plant​s1307​1036 Huang M, Liu X, Zhou Y, Summers RM, Zhang Z (2019) BLINK: a package for the next level of genome-wide association studies with both individuals and markers in the millions. GigaScience 8(2):giz006. https://​doi.​org/​10.​ 1093/​gigas​cience/​giz006 Hussain S, Zhang J, Zhong C, Zhu L, Cao X, Yu S, Bohr JA, Hu J, Jin Q (2017) Effects of salt stress on rice growth, development characteristics, and the regulating ways: a review. J Integr Agric 16:2357–2374. https://​doi.​org/​10.​ 1016/​S2095-​3119(16)​61608-8 IRRI (2013) Standardization evaluation system for rice. Inter- national Rice Research Institute, Philippines, p 34 Islam MA, Shelia V, Ludwig F, de Bruyn LL, Rahman MH, Hoogenboom G (2021) Bringing farmers’ perceptions into science and policy: Understanding salinity tolerance of rice in southwestern Bangladesh under climate change. Land Use Policy 101:105159. https://​doi.​org/​10.​1016/j.​ landu​sepol.​2020.​105159 Jahan N, Zhang Y, Lv Y, Song M, Zhao C, Hu H, Cui Y (2020) QTL analysis for rice salinity tolerance and fine map- ping of a candidate locus qSL7 for shoot length under salt stress. Plant Growth Regul 90:307–319. https://​doi.​org/​10.​ 1007/​s10725-​019-​00566-3 Javed MA, Huyop FZ, Wagiran A, Mohd F (2011) Identifica- tion of QTLs for morph-physiological traits related to salinity tolerance at seedling stage in Indica rice. Procedia Environ Sci 8:389–395. https://​doi.​org/​10.​1016/j.​proenv.​ 2011.​10.​061 Ju C, Ma X, Han B, Zhang W, Zhao Z, Geng L, Cui D, Han L (2022) Candidate gene discovery for salt tolerance in rice (Oryza sativa L.) at the germination stage based on genome-wide association study. Front Plant Sci 13:1010654. https://​doi.​org/​10.​3389/​fpls.​2022.​10106​54 Kilian A, Wenzl P, Huttner E, Carling J, Xia L, Blois H, Caig V, Heller-Uszynska K, Jaccoud D, Hopper C, Evers M (2012) Diversity arrays technology: a generic genome pro- filing technology on open platforms. Methods Mol Biol 888:67–89. https://​doi.​org/​10.​1007/​978-1-​61779-​870-2_5 Kim TH, Kim SM (2023) Identification of candidate genes for salt tolerance at the seedling stage using integrated genome-wide association study and transcriptome analy- sis in rice. Plants 12:1401. https://​doi.​org/​10.​3390/​plant​ s1206​1401 Koyama ML, Levesley A, Koebner RMD, Flowers TJ, Yeo AR (2001) Quantitative trait loci for component physiologi- cal traits determining salt tolerance in rice. Plant Physiol 125:406–422. https://​doi.​org/​10.​1104/​pp.​125.1.​406 Krishnamurthy SL, Sharma SK, Gautam RK, Kumar V (2014) Path and association analysis and stress indices for salinity tolerance traits in promising rice (Oryza sativa L.) geno- types. Cereal Res Commun 42:474–483. https://​doi.​org/​ 10.​1556/​CRC.​2013.​0067 Krishnamurthy SL, Pundir P, Warraich AS, Rathor S (2020) Introgressed Saltol QTL lines improves the salinity tol- erance in rice at seedling stage. Front Plant Sci 11:833. https://​doi.​org/​10.​3389/​fpls.​2020.​00833 Kumar A, Gupta C, Thomas J, Pereira A (2021) Genetic Dis- section of grain yield component traits under high night- time temperature stress in a rice diversity panel. Front Plant Sci 28:12–712167. https://​doi.​org/​10.​3389/​fpls.​ 2021.​712167 Leon TBD, Linscombe S, Gregorio G, Subudhi PK (2015) Genetic variation in Southern USA rice genotypes for seedling salinity tolerance. Panel Front Plant Sci 6:374. https://​doi.​org/​10.​3389/​fpls.​2015.​00374 Li N, Zheng H, Cui J, Wang J, Liu H, Sun J, Liu T, Zhao H, Lai Y, Zou D (2019) Genome-wide association study and candidate gene analysis of alkalinity tolerance in japonica rice germplasm at the seedling stage. Rice 12:24. https://​ doi.​org/​10.​1186/​s12284-​019-​0285-y Lin HX, Zhu MZ, Yano M, Gao JP, Liang ZW, Su WA, Hu XH, Ren ZH, Chao DY (2004) QTLs for Na+ and K+ uptake of the shoots and roots controlling rice salt toler- ance. Theor Appl Genet 108:253–260. https://​doi.​org/​10.​ 1007/​s00122-​003-​1421-y Lipka AE, Tian F, Wang Q, Peiffer J, Li M, Bradbury PJ, Gore MA, Buckler ES, Zhang Z (2012) GAPIT: genome https://doi.org/10.1186/s12284-017-0186-x https://doi.org/10.1093/bioinformatics/btm308 https://doi.org/10.1093/bioinformatics/btm308 https://doi.org/10.1021/jf305531j https://doi.org/10.1021/jf305531j https://doi.org/10.1177/00307270211019758 https://doi.org/10.1177/00307270211019758 https://doi.org/10.1186/s12284-020-00424-1 https://doi.org/10.3390/ijms23094810 https://doi.org/10.3390/ijms23094810 https://doi.org/10.3390/plants13071036 https://doi.org/10.1093/gigascience/giz006 https://doi.org/10.1093/gigascience/giz006 https://doi.org/10.1016/S2095-3119(16)61608-8 https://doi.org/10.1016/S2095-3119(16)61608-8 https://doi.org/10.1016/j.landusepol.2020.105159 https://doi.org/10.1016/j.landusepol.2020.105159 https://doi.org/10.1007/s10725-019-00566-3 https://doi.org/10.1007/s10725-019-00566-3 https://doi.org/10.1016/j.proenv.2011.10.061 https://doi.org/10.1016/j.proenv.2011.10.061 https://doi.org/10.3389/fpls.2022.1010654 https://doi.org/10.1007/978-1-61779-870-2_5 https://doi.org/10.3390/plants12061401 https://doi.org/10.3390/plants12061401 https://doi.org/10.1104/pp.125.1.406 https://doi.org/10.1556/CRC.2013.0067 https://doi.org/10.1556/CRC.2013.0067 https://doi.org/10.3389/fpls.2020.00833 https://doi.org/10.3389/fpls.2021.712167 https://doi.org/10.3389/fpls.2021.712167 https://doi.org/10.3389/fpls.2015.00374 https://doi.org/10.1186/s12284-019-0285-y https://doi.org/10.1186/s12284-019-0285-y https://doi.org/10.1007/s00122-003-1421-y https://doi.org/10.1007/s00122-003-1421-y Euphytica (2025) 221:90 Page 17 of 18  90 Vol.: (0123456789) association and prediction integrated tool. Bioinformat- ics 28:2397–2399. https://​doi.​org/​10.​1093/​bioin​forma​tics/​ bts444 Liu C, Mao B, Yuan D, Chu C, Duan M (2016) Salt tolerance in rice: Physiological responses and molecular mecha- nisms. The Crop Journal 4(3):126–134. https://​doi.​org/​10.​ 1016/j.​cj.​2021.​02.​010 Mazumder A, Rohilla M, Krishnan SG, Kole PC, Mondal TK (2024) Phenotypic assessment of RILs of an inter- specific biparental population (IR29/Oryzaglaberrima Steud) for seedling salinity tolerance and an assessment of the presence of novel QTL/haplotype in African rice (O. glaberrima S.) genome. Indian J Genet Plant Breed 84(4):644–651 Meyer RS, Choi JY, Sanches M, Plessis A, Flowers JM, Amas J, Dorph K, Barretto A, Gross B, Fuller DQ, Bimpong IK, Ndjiondjop MN, Hazzouri KM, Gregorio GB, Purugga- nan MD (2016) Domestication history and geographical adaptation inferred from a SNP map of African rice. Nat Genet 48:1083–1088. https://​doi.​org/​10.​1038/​ng.​3633 Mohidem NA, Hashim N, Shamsudin R, Che Man H (2022) Rice for food security: revisiting production, diversity, rice milling process, and nutrient content. Agriculture 12:741. https://​doi.​org/​10.​3390/​agric​ultur​e1206​0741 Munns R, Tester M (2008) Mechanisms of salinity tolerance. Annu Rev Plant Biol 59:651–681. https://​doi.​org/​10.​1146/​ annur​ev.​arpla​nt.​59.​032607.​092911 Munns R, James RA, Läuchli A (2006) Approaches to increas- ing the salt tolerance of wheat and other cereals. J Exp Bot 57(5):1025–1043. https://​doi.​org/​10.​1093/​jxb/​erj100 Nayyeripasand L, Garoosi GA, Ahmadikhah A (2021) Genome-wide association study (GWAS) to identify salt- tolerance QTLs carrying novel candidate genes in rice during early vegetative stage. Rice 14:9. https://​doi.​org/​ 10.​1186/​s12284-​020-​00433-0 Ndjiondjop MN, Gouda AC, Eizenga GC, Warburton ML, Kpeki SB, Wambugu PW, Gnikoua K, Tia DD, Bachabi F (2022) Genetic variation and population structure of Oryza sativa accessions in the AfricaRice collection and development of the AfricaRice O. sativa core collection. Crop Sci 63:724–739. https://​doi.​org/​10.​1002/​CSC2.​ 20898 Ntakirutimana F, Tranchant-Dubreuil C, Cubry P, Chougule K, Zhang J, Wing RA, Adam H, Lorieux M, Jouannic S (2023) Genome-wide association analysis identifies natu- ral allelic variants associated with panicle architecture variation in African rice, Oryza glaberrima Steud. G3 (Bethesda) 13(10):jkad174. https://​doi.​org/​10.​1093/​g3jou​ rnal/​jkad1​74 Okeshkumar BML, Krishnamurthy SLK, Athor SR, Arriach ASW, Inaykumar NMV, Ushyanthakumar BMD, Harma PCS (2023) Science direct morphophysiological diversity and haplotype analysis of Saltol QTL region in diverse rice landraces for salinity tolerance. Rice Sci 30:306–320. https://​doi.​org/​10.​1016/j.​rsci.​2023.​02.​001 Pires IS, Negrão S, Oliveira MM, Purugganan MD (2015) Comprehensive phenotypic analysis of rice (Oryza sativa). Physiol Plants 155:43–54. https://​doi.​org/​10.​1111/​ ppl.​12356 Platten JD, Egdane JA, Ismail AM (2023) Salinity tolerance, Na+ exclusion and allele mining of HKT1;5 in Oryza sativa and O. glaberrima: many sources, many genes, one mechanism. BMC Plant Biol 13:32. https://​doi.​org/​10.​ 1186/​1471-​2229-​13-​32 Qin H, Li Y, Huang R (2020) Advances and challenges in the breeding of salt tolerance rice. Int J Mol Sci 21:8385. https://​doi.​org/​10.​3390/​ijms2​12183​85 Rahim MS, Sharma H, Parveen A, Roy JK (2018) Trait map- ping approaches through association analysis in plants. In: Varshney R, Pandey M, Chitikineni A (eds) Plant genetics and molecular biology. Adv. Biochem. Eng. Biotechnol, vol 16, pp 483–108. https://​doi.​org/​10.​1007/​10_​2017_​50 Rahman ANM, Zhang J (2022) Trends in rice research: 2030 and beyond. Food Energy Sec 1:17. https://​doi.​org/​10.​ 1002/​fes3.​390 Sahito JH, Zhang H, Gishkori ZGN, Ma C, Wang Z, Ding D, Zhang X, Tang (2024) Advancements and prospects of genome-wide association studies (GWAS) in Maize. Int J Mol Sci 25:1918. https://​doi.​org/​10.​3390/​ijms2​50319​18 Salgotra RK, Chauhan BS (2023) Genetic diversity, con- servation, and utilization of plant genetic resources. Genes 14:174. https://​doi.​org/​10.​3390/​genes​14010​ 174017-​0186-x Segura V, Vilhjálmsson BJ, Platt A, Korte A, Seren Ü, Long Q, Nordborg M (2012) An efficient multi-locus mixed-model approach for genome-wide association studies in struc- tured populations. Nat Genet 44(7):825–830. https://​doi.​ org/​10.​1038/​ng.​2314 Sikirou M, Shittu A, Konaté KA, Maji AT, Ngaujah AS, Sanni KA, Ogunbayo SA, Akintayo I, Saito K, Dramé KN, Ahanchédé A, Venuprasad R (2018) Field crops research screening African rice (Oryza glaberrima) for tolerance to abiotic stresses: I. Fe toxicity field. Crops Res 220:3–9. https://​doi.​org/​10.​1016/j.​fcr.​2016.​04.​016 Singh RK, Kota S, Flowers TJ (2021) Salt tolerance in rice: seedling and reproductive stage QTL mapping come of age. Theor Appl Genet 134(11):3495–3533. https://​doi.​ org/​10.​1007/​s00122-​021-​03890-3 Singh VK, Singh BD, Kumar A, Maurya S, Krishnan SG, Vinod KK, Singh MP, Ellur RK, Bhowmick PK, Singh AK (2018) Marker-assisted introgression of Saltol QTL enhances seedling stage salt tolerance in the rice variety “Pusa Basmati 1.” Int J Genomics 2018:8319879. https://​ doi.​org/​10.​1155/​2018/​83198​79 Song S, Tian D, Zhang Z, Hu S, Yu J (2018) Rice genomics: over the past two decades and into the future. Genomics Proteom Bioinform 16:397–404. https://​doi.​org/​10.​1016/j.​ gpb.​2019.​01.​001 Takehisa H, Shimodate T, Fukuta Y, Ueda T, Yano M, Yamaya T, Kameya T, Sato T (2004) Identification of quantitative trait loci for plant growth of rice in paddy field flooded with salt water. Field Crops Res 89:85–95. https://​doi.​org/​ 10.​1016/j.​fcr.​2004.​01.​026 Thomson MJ, de Ocampo M, Egdane J et  al (2010) Charac- terizing the Saltol quantitative trait locus for salinity tol- erance in rice. Rice 3:48–160. https://​doi.​org/​10.​1007/​ s12284-​010-​9053-8 Varshney RK, Nayak SN, May GD, Jackson SA (2009) Next- generation sequencing technologies and their implica- tions for crop genetics and breeding. Trends Biotechnol 27:522–530. https://​doi.​org/​10.​1016/j.​tibte​ch.​2009.​05.​006 https://doi.org/10.1093/bioinformatics/bts444 https://doi.org/10.1093/bioinformatics/bts444 https://doi.org/10.1016/j.cj.2021.02.010 https://doi.org/10.1016/j.cj.2021.02.010 https://doi.org/10.1038/ng.3633 https://doi.org/10.3390/agriculture12060741 https://doi.org/10.1146/annurev.arplant.59.032607.092911 https://doi.org/10.1146/annurev.arplant.59.032607.092911 https://doi.org/10.1093/jxb/erj100 https://doi.org/10.1186/s12284-020-00433-0 https://doi.org/10.1186/s12284-020-00433-0 https://doi.org/10.1002/CSC2.20898 https://doi.org/10.1002/CSC2.20898 https://doi.org/10.1093/g3journal/jkad174 https://doi.org/10.1093/g3journal/jkad174 https://doi.org/10.1016/j.rsci.2023.02.001 https://doi.org/10.1111/ppl.12356 https://doi.org/10.1111/ppl.12356 https://doi.org/10.1186/1471-2229-13-32 https://doi.org/10.1186/1471-2229-13-32 https://doi.org/10.3390/ijms21218385 https://doi.org/10.1007/10_2017_50 https://doi.org/10.1002/fes3.390 https://doi.org/10.1002/fes3.390 https://doi.org/10.3390/ijms25031918 https://doi.org/10.3390/genes14010174017-0186-x https://doi.org/10.3390/genes14010174017-0186-x https://doi.org/10.1038/ng.2314 https://doi.org/10.1038/ng.2314 https://doi.org/10.1016/j.fcr.2016.04.016 https://doi.org/10.1007/s00122-021-03890-3 https://doi.org/10.1007/s00122-021-03890-3 https://doi.org/10.1155/2018/8319879 https://doi.org/10.1155/2018/8319879 https://doi.org/10.1016/j.gpb.2019.01.001 https://doi.org/10.1016/j.gpb.2019.01.001 https://doi.org/10.1016/j.fcr.2004.01.026 https://doi.org/10.1016/j.fcr.2004.01.026 https://doi.org/10.1007/s12284-010-9053-8 https://doi.org/10.1007/s12284-010-9053-8 https://doi.org/10.1016/j.tibtech.2009.05.006 Euphytica (2025) 221:90 90   Page 18 of 18 Vol:. (1234567890) Visscher PM, Hill WG, Wray NR (2008) Heritability in the genomics era–concepts and misconceptions. Nat Rev Genet 9:255–266. https://​doi.​org/​10.​1038/​nrg23​22 Walia H, Wilson C, Condamine P, Liu X, Ismail AM, Zeng L, Close TJ (2005) Comparative transcriptional profiling of two contrasting rice genotypes under salinity stress dur- ing the vegetative growth stage. Plant Physiology 139(2): 822–835. https://​doi.​org/​10.​1104/​pp.​105.​065961 Wambugu PW, Furtado A, Le Waters D, Nyamongo DO, Henry RJ (2013) Conservation and utilization of African Oryza genetic resources. Rice 6:1–13. https://​doi.​org/​10.​ 1186/​1939-​8433-6-​29 Wambugu PW, Ndjiondjop M-N, Henry R (2019) Advances in molecular genetics and genomics of African rice (Oryza glaberrima Steud). Plants 8(10):376. https://​doi.​org/​10.​ 3390/​plant​s8100​376 Wambugu PW, Ndjiondjop MN, Henry R (2021) Genet- ics and genomics of African rice (Oryza glaberrima Steud) domestication. Rice 14:6. https://​doi.​org/​10.​1186/​ s12284-​020-​00449-6 Wang M, Yu Y, Haberer G, Marri PR, Fan C, Goicoechea JL, Zuccolo A, Song X, Kudrna D, Ammiraju JSS, Cossu RM, Maldonado C, Chen J, Lee S, Sisneros N, Baynast K, De Golser W, Wissotski M, Kim W et al (2014) The genome sequence of African rice (Oryza glaberrima) and evidence for independent domestication. Nat Pub Grp 46:9. https://​doi.​org/​10.​1038/​ng.​3044 Wang A, Jiang Y, Shu X, Zha Z, Yin D, Liu Y, Zhang D, Xu D, Jiao C, Jia X, Ye X, Li S, Deng Q, Wang S, Zhu J, Liang Y, Zou T, Liu H, Wang L, …Zheng A, (2021) Genome- wide association study-based identification genes influ- encing agronomic traits in rice (Oryza sativa L.). Genom- ics 113:1396–1406. https://​doi.​org/​10.​1016/j.​ygeno.​2021.​ 03.​016 Wei T, Simko V (2021) R package ’corrplot’: visualization of a correlation matrix (version 0.92). https://​github.​com/​tai- yun/​corrp​lot Wray NR, Yang J, Hayes BJ, Price AL, Goddard ME, Viss- cher PM (2013) Pitfalls of predicting complex traits from SNPs. Nat Rev Genet 14(7):507–515. https://​doi.​org/​10.​ 1038/​nrg34​57 Xu Y, Li P, Yang Z, Xu C (2017) Genetic mapping of quantita- tive trait loci in crops. Crop J 5:175–184. https://​doi.​org/​ 10.​1016/j.​cj.​2016.​06.​003 Yang M, Lu K, Zhao F, Xie W, Ramakrishna P, Wang G, Du Q, Wang W, Dong H, Hu J, Ming L, Xing Y, Wang G, Xiao J (2018) Genome-wide association studies reveal the genetic basis of ionomic variation in rice. Plant Cell 30:2720–2740. https://​doi.​org/​10.​1105/​tpc.​18.​00375 Yin L, Zhang H, Tang Z, Xu J, Yin D, Zhang Z, Yuan X, Zhu M, Zhao S, Li X, Liu X (2021) rMVP: a memory- efficient, visualization-enhanced, and parallel-accelerated tool for genome-wide association study. Genom Proteom Bioinform Genom 19:619–628. https://​doi.​org/​10.​1016/j.​ gpb.​2020.​10.​007 Yoshida S, Forno DA, Cock JK, Gomez KA (1976) Labora- tory manual for physiological studies of rice. International Rice Research Institute, Los Baños, Philippines Yuan L, Zhang L, We X, Wang R, Li N, Chen G, Fan F, Huang S, Li J, Li S (2022) Quantitative trait locus mapping of salt tolerance in wild rice Oryza long staminate. Int J Mol Sci 23:2379. https://​doi.​org/​10.​3390/​ijms2​30423​79 Zhang Z, Ersoz E, Lai CQ, Todhunter RJ, Tiwari HK, Gore MA, Bradbury PJ, Yu J, Arnett DK, Ordovas JM, Buckler ES (2010) The mixed linear model approach adapted for genome-wide association studies. Nat Genet 42:355–360. https://​doi.​org/​10.​1038/​ng.​546 Publisher’s Note  Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law. https://doi.org/10.1038/nrg2322 https://doi.org/10.1104/pp.105.065961 https://doi.org/10.1186/1939-8433-6-29 https://doi.org/10.1186/1939-8433-6-29 https://doi.org/10.3390/plants8100376 https://doi.org/10.3390/plants8100376 https://doi.org/10.1186/s12284-020-00449-6 https://doi.org/10.1186/s12284-020-00449-6 https://doi.org/10.1038/ng.3044 https://doi.org/10.1016/j.ygeno.2021.03.016 https://doi.org/10.1016/j.ygeno.2021.03.016 https://github.com/taiyun/corrplot https://github.com/taiyun/corrplot https://doi.org/10.1038/nrg3457 https://doi.org/10.1038/nrg3457 https://doi.org/10.1016/j.cj.2016.06.003 https://doi.org/10.1016/j.cj.2016.06.003 https://doi.org/10.1105/tpc.18.00375 https://doi.org/10.1016/j.gpb.2020.10.007 https://doi.org/10.1016/j.gpb.2020.10.007 https://doi.org/10.3390/ijms23042379 https://doi.org/10.1038/ng.546 Genome-wide association study for salinity tolerance in the African rice, Oryza glaberrima Abstract Introduction Materials and methods Experimental materials and growing conditions Seeding of the experimental materials Evaluation of salt tolerance Determination of shoot Na+ and K+ ion concentrations Genotyping SNP data filtering Data analysis Statistical analysis of phenotypic traits Estimation of traits heritability Genome-wide association analysis Variances explained by significant SNP marker Results Response to salinity tolerance in African rice accessions Phenotype variability and correlation analysis among the traits Association analysis for rice seedling stage traits related to salinity tolerance Markers associated with salinity tolerance Marker trait significant associations Phenotypic variance explained by highly significant SNP markers Standard evaluation scores (SES) and NaK ratio Shoot and root measurements Shoot and root dry weight measurements Discussion Conclusion and recommendations Acknowledgements References