Genetic Diversity and Population Structure of Saffron (Crocus sativus L.) in Morocco Revealed by Sequence-Related Amplified Polymorphism Markers

Abstract

Saffron (Crocus sativus L.) is one of the most expensive spices in the world. Saffron, prized for its vibrant color, aroma, and taste, is essential in the food industry and traditional medicine. Its culinary uses, therapeutic benefits, and potential antioxidant, anti-inflammatory, and anticancer properties highlight its significant importance. Its genetic diversity has significant implications for cultivation and quality. In this study, genetic diversity among 76 saffron accessions, collected from 13 localities of Taliouin region of Morocco, were evaluated using sequence-related amplified polymorphism (SRAP) markers. A total of 63 polymorphic fragments were produced with an average of total number and polymorphic bands per primer were of 10.5 and 10.16, respectively. Most of the variations among the localities, revealed by the Analysis of Molecular Variance, originated from the within accessions differentiation (81%; p < 0.010). Cluster Analysis, Principal Coordinate Analysis (PCoA), and population structure confirmed the main groups and corroborated genetic homogeneity across accessions. In fact, close relationships were revealed between accessions from different locations, showing that there was no relationship between genetic divergence and geographical locality. This investigation represents a pivotal advance towards fostering sustainable development and bolstering the economic empowerment of the saffron farming communities in Morocco.

1. Introduction

Crocus sativus L., or saffron, is a semi-perennial, monocotyledonous geophyte from the Iridaceae family, comprising about 100 corm-bearing species, with 30–40 cultivated varieties [1,2,3,4,5,6]. Growing demand for saffron in medical and cosmetic industries drives its market growth, as it remains one of the most valuable cash crops, often referred to as “red gold” and the “king of spices” due to its high value and limited production [7]. This small, stemless autumnal herb features purple flowers and grass-like leaves [8]. Uniquely, C. sativus exists only as a cultivated species, being an autumn-flowering, male-sterile triploid (2n = 3x = 24), with an uncertain origin from diploid and polyploid Crocus species [9,10,11,12].

Flowering in Crocus sativus occurs in autumn, either before, during, or after leaf emergence, with corm replacement and propagation continuing until mid-spring, followed by real dormancy and summer pseudo-dormancy [13]. Traditionally propagated through corms, saffron yield strongly depends on corm size, with small corms weighing 9 g and large mother corms 10–14 g, producing 6–10 cormlets per season for the next cycle [14,15]. According to Abdusamat et al. [16], its biological cycle spans 220 days, beginning with autumnal vegetative growth after the first rains and concluding with corm formation. Highly adaptable to various soils and climates, saffron has been cultivated from the Mediterranean to the Near East since ancient times [9,17,18,19]. Iran dominates global production, supplying over 90% of the world’s saffron, followed by India, Spain, Morocco, Greece, and Italy [20,21,22].

Morocco is the fourth-largest producer globally, with 95% of its production concentrated in the Taliouine region, renowned for its ideal saffron-growing conditions. Cultivation in this area, and more recently in the Tazenakht area and surrounding regions, has been practiced for centuries [23,24]. Saffron fields are located at altitudes ranging from 800 to 1800 meters, benefiting from a warm, dry microclimate and well-drained sandy loam or calcareous clay soils, which are optimal for saffron production [25,26].

Moroccan saffron farming requires around 3 tons of corms per hectare, contributing to about 1.5% of global saffron production. Cultivation follows a meticulous method, with 2 × 2 m raised beds, 20 cm row spacing, and corms planted in bunches of two or three, 10–15 cm apart at a depth of 15 cm [27]. This setup, combined with favorable conditions, enhances yield and quality. Morocco produces an annual average of 6.5 tons, exporting 1.2 tons mainly to Spain and Switzerland. However, productivity remains low at 2–3.5 kg/ha compared to Italy’s 8.4 kg/ha. Traditional labor-intensive methods increase costs, as one kilogram of flowers yields only 72 g of fresh and 12 g of dried stigmas, highlighting the sector’s untapped potential for improvement [7].

The Moroccan saffron, renowned for its high quality, owes its prestige to the unique climate and soil of Taliouine and the expertise of local farmers, making it a valuable global commodity [24]. Commercial saffron comprises the dried red stigmas with a small portion of the yellowish style, widely used in the food industry and medicine. Its bioactive compounds—crocin (color), picrocrocin (flavor), and safranal (aroma)—enhance both culinary and therapeutic applications [28,29,30]. Crocetin exhibits anticancer properties, while safranal and picrocrocin contribute to antidepressant effects [31,32]. Additionally, C. sativus possesses antihypertensive, anticonvulsant, anti-inflammatory, and other medicinal benefits [33]. However, saffron cultivation faces sustainability threats due to poor soil fertility, disease susceptibility, labor-intensive harvesting, climate change impacts, and spice adulteration, all contributing to genetic erosion, particularly in Mediterranean regions [34,35,36,37].

Efforts to conserve and improve saffron cultivation include implementing better agricultural practices, agronomic and post-harvesting techniques, improving disease resistance through breeding programs, and ensuring fair trade practices to protect against adulteration [38,39,40,41]. Unfortunately, most of them have been hard to achieve or have resulted in an incomplete yield [42]. However, recent studies have highlighted the potential of innovative growing techniques to address these challenges. In fact, an experiment conducted on a certified organic farm in Sicily demonstrated that emergency irrigation during the pre-flowering stage, combined with organic nutrition protocols (amino acids and beneficial microorganisms such as Frankia spp. and Pochonia chlamydosporia), significantly increased both flower and stigma yields up to three times higher than the control treatment [43].

Characterization and preservation of C. sativus ecotypes remain essential to protect and promote the biodiversity of this crop [44,45,46]. Preserving genetic diversity involves maintaining traditional varieties while exploring new cultivars to enhance resilience against environmental changes [47,48,49,50]. Innovative approaches, such as those tested in Sicily [43], underscore the importance of combining genetic and agronomic strategies to sustainably improve saffron yields and quality.

Over the years, the methodologies used to study tools of genetic variation among C. sativus have evolved significantly. Initially, researchers relied on morphological and agronomic traits to differentiate and understand various accessions. However, with advancements in biotechnology, the focus has shifted towards molecular analysis, providing more precise and detailed insights into genetic diversity [37,51]. Thus, exploring its genetic diversity using molecular markers is critically important for germplasm conservation. On the other hand, molecular markers are essential to confirm the accession structure, identify unique genetic profiles, detect polymorphisms, and assess the levels and patterns of genetic variation within and between C. sativus accessions [8,52]. Several techniques based on DNA polymorphism analysis have been used worldwide to assess the genetic diversity and population structure of C. sativus accessions from different countries, such as amplified polymorphic DNA (RAPD) [9,44,53], inter simple sequence repeat (ISSR) [20,54], and amplified fragment length polymorphism (AFLP) [55,56]. For instance, ISSR and SSR markers have been effectively used to study the genetic diversity of Iranian saffron, revealing significant genetic variability, which is crucial for breeding programs aimed at improving disease resistance and stress tolerance [8,20,57,58,59]. Similarly, recent studies on Moroccan saffron employed 19 EST-SSR, 10 SSRg, and 10 ISSR markers to evaluate genetic diversity across 14 individuals from distinct regions. The findings from such studies underscore the critical role of molecular markers in supporting breeding programs and conservation efforts tailored to regional contexts [60].

Otherwise, sequence-related amplified polymorphism (SRAP) markers, developed and demonstrated by Haq [61] and Li and Quiro [62], have been used to assess the genetic diversity of 28 C. sativus accessions collected from different regions of Iran highlighting four groups of accessions, most of which are clustered in a unique and major group. In the latter study, the close relationships among genotypes were explained by a plausible narrow genetic background of the species due to the vegetative propagation and human selection of superior genotypes [8,63]. The application of molecular markers extends beyond conservation. They play a pivotal role in enhancing breeding programs by facilitating the selection of desirable traits and enabling the development of new cultivars with improved qualities [54]. This is particularly important in the context of challenges such as poor soil fertility, increased disease susceptibility, and the labor-intensive nature of saffron harvesting. By leveraging molecular markers, breeders can develop saffron varieties that are more resilient to environmental stresses and have higher yields [34,64]. In addition to improving breeding practices, molecular analysis aids in combating the issue of spice adulteration. By establishing a genetic fingerprint for authentic saffron, it becomes easier to verify the purity of saffron products in the market, thereby protecting both consumers and producers from fraud. This is critical for maintaining the economic value and reputation of saffron, particularly in countries where it is a major agricultural product [47,65,66]. Despite varying results regarding the genetic diversity of saffron reported in the literature, molecular markers are more valuable and reliable tools than morphological markers in assessing the saffron genetic variations. While morphological markers are relatively easy and inexpensive to use, their reliability is often compromised due to their susceptibility to environmental conditions and the developmental stages of the plant. These factors can introduce significant variability, making it difficult to obtain consistent and accurate assessments of genetic diversity [37,67].

The development of the saffron sector in Morocco is crucially dependent on increasing yields and quality, which directly enhance the incomes of farmers and smallholders. This economic improvement can significantly boost the livelihoods of rural communities engaged in saffron cultivation [37,68]. To achieve these goals, it is essential to protect and optimize the cultivation of local Crocus sativus ecotypes. A fundamental step in this process is gaining a thorough understanding of the population structure and genetic diversity of the saffron plants being cultivated [59,69,70]. Unfortunately, there is currently a lack of comprehensive information regarding the genetic variations within the Moroccan C. sativus germplasm. This knowledge gap poses a significant challenge to the efficient selection and cultivation of superior local germplasm, which is necessary for improving yield and quality. Without detailed genetic data, efforts to enhance the saffron sector through breeding and conservation programs are severely hampered [71,72].

In response to this critical need, the present study aims to assess the genetic diversity and population structure of C. sativus accessions from the Taliouine region, a key area for saffron cultivation in Morocco. To achieve this, we employed Sequence-Related Amplified Polymorphism (SRAP) markers, a powerful tool for detecting genetic variation and understanding the relationships between different accessions. SRAP markers are an efficient and straightforward marker technique system for genetic diversity characterization because they possess high reproducibility and discriminatory power, disclose many codominant markers, and target open reading frames [73,74]. Moreover, SRAP markers were initially developed for Brassica species to target open reading frames (ORFs) within genomes, facilitating the study of functional genetic diversity, as shown by Haq [61] and Li and Quiro [62]. In the context of saffron (Crocus sativus), SRAP markers have proven valuable for genetic studies. For instance, Babaei [8] utilized SRAP markers to analyze genetic diversity among saffron accessions from different regions of Iran, demonstrating the markers’ effectiveness in detecting polymorphisms within this species. Moreover, the successful application of SRAP markers in a wide range of plant species such as Coffee Populations [75], Tomato Landraces [76], Cumin Genotypes [77], and Quercus petraea (Matt.) Populations [78] indicates their versatility and reliability in genetic analyses beyond their original development for Brassica.

Using SRAP markers, this research seeks to fill the existing knowledge gap and provide valuable data on the genetic makeup of Moroccan saffron. The findings will not only contribute to the scientific understanding of C. sativus in Morocco but also support practical efforts to enhance the saffron sector. Improved genetic knowledge will enable the selection of high-performing ecotypes, leading to better crop management, increased yields, and ultimately, higher incomes for farmers and smallholders.

2. Materials and Methods

2.1. Plant Material

A total of 76 individuals of C. sativus were collected from natural fields of 13 localities in the Taliouin region (Figure 1). Between five and six individuals representing each locality were used for the molecular study (

Table 1. List of Crocus sativus accessions and the number of samples with their geographical origins in the Taliouin region: 13 accessions exhibiting similar morphological characteristics used for molecular analysis.

Localities	Number of Samples	Geographical Locality (Town)	Code	Latitude	Longitude	Altitude
1	6	Imin tlafoumlil	IMI-1	30.568818	−7.85346	1876
2	5	Taoulouzte	TAO-2	30.738507	−7.779522	1964
3	6	Aglimhakki	AGL-3	30.808924	−7.786903	2035
4	5	Tighboula	TIG-4	30.80522	−7.772141	2121
5	6	Dar Ifrane	DAR-5	30.232073	−7.796592	1832
6	6	Izourzen	IZO-6	30.233111	−7.794296	1780
7	6	Talaght	TAL-7	30.148356	−8.266525	1615
8	6	Azghar Nizri	AZG-8	30.147929	−8.289528	1650
9	6	Agzoule	AGR-9	30.154140	−8.266954	1620
10	6	Issoura	ISS-10	30.571637	−7.611294	1676
11	6	Guersafen	GUE-11	30.572882	−7.61065	1680
12	6	Idawtala	IDA-12	30.610221	−7.6117	1700
13	6	Bettale	BET-13	30.757580	−7.374315	1604

). Plant tissues were independently harvested; the leaves from these samples were lyophilized and stored for DNA extraction.

2.2. DNA Extraction

Genomic DNA was extracted from lyophilized young leaves according to the modified CTAB procedure described by Lassner et al. [79]. For each sample, an amount from 50 to 100 mg of powder leaves was suspended with 1 mL of CTAB buffer (1M Tris-HCl, 5 M NaCl, 0.5 M EDTA; 2% (w/v) CTAB, and 0.2% (v/v) β-mercaptoethanol). Samples were mixed and incubated at 65 °C for 60 min, then cooled on ice for 5 min. Subsequently, 0.8 ml of an equal volume of chloroform-isoamyl alcohol (24:1-v/v) was added to the template-mixing and centrifuged for 10 min at 12,000 rpm. DNA was precipitated overnight with 0.8 mL of isopropanol and then centrifuged at 12,000 rpm for 10 min. The extracted DNA was washed with ethanol (75%) and dissolved in 0.4 mL of TE buffer (Tris 10 mM pH 8 and EDTA 1 mM). The extracted DNA was evaluated using electrophoresis in agarose gel, then quantified using Nano micro-spectrophotometer (Xian Yima Opto-electrical Technology, Xi’an, Shaanxi, China) at 260 nm. Finally, DNA was stored at −20 °C for further analyses.

2.3. SRAP-PCR Analysis

A set of 30 SRAP primer combinations of five forward (ME) and six reverse (EM) [62] were tested across the 13 C. sativus accessions to determine the most robust and reproducible ones (ME1/EM1, ME1/EM2, ME1/EM3, ME1/EM4, ME1/EM5, ME1/EM6, ME2/EM1, ME2/EM2, ME2/EM3, ME2/EM4, ME2/EM5, ME2/EM6, ME3/EM1, ME3/EM2, ME3/EM3, ME3/EM4, ME3/EM5, ME3/EM6, ME4/EM1, ME4/EM2, ME4/EM3, ME4/EM4, ME4/EM5, ME4/EM6, ME5/EM1, ME5/EM2, ME5/EM3, ME5/EM4, ME5/EM5, and ME5/EM6) (

Table 2. List and sequence of SRAP primers tested across the 13 C. sativus accessions.

Forward		Reverse
Name	Sequence (5′-3′)	Name	Sequence (5′-3′)
ME1	TGAGTCCAAACCGGATA	EM1	GACTGCGTACGAATTAAT
ME2	TGAGTCCAAACCGGAGC	EM2	GACTGCGTACGAATTTGC
ME3	TGAGTCCAAACCGGAAT	EM3	GACTGCGTACGAATTGAC
ME4	TGAGTCCAAACCGGACC	EM4	GACTGCGTACGAATTTGA
ME5	TGAGTCCAAACCGGAAG	EM5	GACTGCGTACGAATTAAC
		EM6	GACTGCGTACGAATTGCA

). SRAP fragments were generated in a total volume of 20 µL, containing 25 ng of genomic DNA sample, 1U of Taq™ DNA polymerase (BIOLINE, London, UK), 0.5 µM of each forward and reverse primers, and 1X PCR buffer (BIOLINE, London, UK). Amplifications were carried out according to the method by Li and Quiros [62].

A PCR program was carried out using the DNA Thermocycler (Applied Biosystems thermocycler (9902), Foster City, CA, USA), which consisted of the two following stages. DNA amplifications were performed in 5 cycles of 94 °C for 1 min, 35 °C for 1 min, and 72 °C for 1 min for denaturing, annealing, and extension, respectively. Then, the annealing temperature was set to 50 °C for other 35 cycles.

PCR products were resolved by electrophores with a DNA marker (HyperLadder^TM 50 bp, BIOLINE, London, UK) using 8% (w/v) polyacrylamide gel at 12.5 v/cm for 1.5 h. Then, the gel was carefully stained in an ethidium bromide solution for 5 min. The DNA fragments were visualized and photographed using a Bio Doc-It™ imaging system UVP, LLC, Upland, CA, USA.

2.4. Statistical Analyses

The amplified DNA fragment for each primer combination was scored as present (1) or absent (0) of homologous bands to assemble a matrix of the different SRAP phenotypes. The data analysis was conducted using Gen AlEx ver. 6.5. The polymorphic Information Content value (PICv) was calculated for each primer combinations, based on the approach by Roldàn-Ruiz et al. [80], using the formula:

$$PI{C}_i=2\times f_i(1-f_i)$$

• PIC_i is the polymorphic information content of marker ‘i’,
• f_i is the frequency of the amplified allele (band present),
• 1 − f_i is the frequency of the null allele.

Using Gen AlEx ver. 6.5 software [81], genetic diversity within accessions was measured using the standard format by calculating the expected heterozygosity, percentage of polymorphic loci, and genetic structure.

The number of permutations for significant testing was set at 1000 for analysis. Analysis of Molecular Variance (AMOVA) based on PhiPT values was carried out using the same program to calculate F_ST analogue. Genetic relationship among accessions were further analyzed by the Principal Coordinate Analysis (PCoA) using the matrix of inter-individual Dice’s distance coefficients [82]. Cluster Analysis using Jaccard’s similarity coefficient was performed with the XLSTAT version 5.14 software, using the un-weighted pair-group method with the arithmetic average (UPGMA) method.

Finally, to infer accession structure, a Bayesian clustering approach described by [83] was executed using STRUCTURE software v.2.3.4. Ten independent replicates were executed for the clusters (K) from 2 to 13, under the admixture model with correlated allele frequencies. The analysis was run according to the STRUCTURE software manual recommendations. Also, each simulation implied 100,000 Monte-Carlo Markov Chain (MCMC) iterations after a burn-in period length of 50,000. For all datasets, the best fitting K value was identified using an online tool implementing the Evanno method [84] ΔK method. STRUCTURE results were graphically represented using the same software.

3. Results

3.1. Variation for SRAP Markers

The aim of the present study was to determine the genetic relationship and population structure among 76 Moroccan C. sativus accessions from the Taliouin region applying SRAP markers. In fact, in the initial trial, using the 30 SRAP primer combinations across the 13 localities of the Taliouin region, the primers’ banding patterns that were difficult to score and failed to amplify consistently in all accessions were excluded. Consequently, six primer combinations (ME2/EM1, ME1/EM3, ME2/EM4, ME1/EM2, ME1/EM5, and ME3/EM3) were selected and were able to produce intense bands in all samples (

Table 3. Total number, polymorphic bands, percentage of polymorphism loci, and PIC value of the six SRAP primer combinations with 76 Moroccan C. sativus.

SRAP Primers Combinations	Total Number of Bands	Number of Polymorphic Bands	Polymorphism Rate (%)	PICv
ME2/EM1	13	13	100	0.28
ME1/EM3	10	10	100	0.36
ME2/EM4	10	9	90	0.29
ME1/EM2	13	13	100	0.30
ME1/EM5	10	9	90	0.34
ME3/EM3	7	7	100	0.26
Average	10.5	10.6	96.66	0.30
Total	63	61	-	-

ME: forward of SRAP primers; EM: reverse of SRAP primers; PICv: Polymorphic Information Content values.

). A total of 63 reproducible fragments were identified (Table 3). Among them, 61 bands were polymorphic (96.66%), ranging in size from 50 bp to 2000 bp. The average number of total bands and polymorphic bands per primer were 10.5 and 10.16, respectively. The number of fragments detected by individual primer combinations ranged from 7 bands (ME3/EM3) to 13 bands (ME2/EM1 and ME1/EM2). Four of the SRAP primer combinations analyzed, ME2/EM1, ME1/EM3, ME1/EM2, and ME3/EM3, yielded 100% polymorphic bands. On the other hand, the minimum proportion of polymorphic bands (90%) was recorded with ME2/EM 4 and ME1/EM5 (Table 3).

PICv was calculated in order to measure the efficiency of polymorphic loci among the accessions. It ranged from 0.26 (ME3/EM3) to 0.36 (ME1/EM3), with an average of 0.30 (Table 3).

3.2. Genetic Diversity of Moroccan C. sativus

The SRAP marker analysis of the 13 C. sativus accessions revealed a high level of genetic variation with 96.82% of polymorphic bands. However, a variable genetic diversity within accessions was observed, with a variation ranging from 28.57 to 93.65%, and a mean percentage of polymorphic loci of 65.56%. The highest percentage was found in locality IMI-1, whereas the lowest value was that of locality TIG-4 (

Table 4. Genetic diversity indices of 76 Moroccan C. sativus revealed by SRAP markers.

Localities	Na (±SE)	Ne (±SE)	I (±SE)	He (±SE)	PPL (%)
IMI-1	1.921 (0.041)	1.663 (0.042)	0.538 (0.025)	0.369 (0.019)	93.65
TAO-2	1.159 (0.085)	1.216 (0.048)	0.174 (0.035)	0.118 (0.024)	31.75
AGL-3	1.810 (0.059)	1.589 (0.047)	0.481 (0.031)	0.329 (0.023)	84.13
TIG-4	0.968 (0.099)	1.252 (0.053)	0.186 (0.038)	0.131 (0.027)	28.57
DAR-5	1.810 (0.063)	1.620 (0.045)	0.501 (0.030)	0.345 (0.022)	85.71
IZO-6	1.635 (0.089)	1.630 (0.049)	0.485 (0.035)	0.340 (0.025)	76.19
TAL-7	1.540 (0.087)	1.472 (0.053)	0.379 (0.038)	0.261 (0.027)	65.08
AZG-8	1.429 (0.084)	1.403 (0.053)	0.319 (0.040)	0.222 (0.028)	52.38
AGR-9	1.730 (0.076)	1.639 (0.050)	0.494 (0.034)	0.344 (0.025)	80.95
ISS-10	1.619 (0.083)	1.513 (0.050)	0.419 (0.036)	0.288 (0.026)	71.43
GUE-11	1.190 (0.084)	1.246 (0.049)	0.197 (0.037)	0.135 (0.026)	33.33
IDA-12	1.460 (0.098)	1.447 (0.050)	0.368 (0.038)	0.252 (0.027)	63.49
BET-13	1.810 (0.063)	1.652 (0.046)	0.515 (0.030)	0.356 (0.022)	85.71
Average	1.545 (0.024)	1.488 (0.015)	0.389 (0.011)	0.268 (0.008)	65.57

Na: number of different alleles; Ne: number of effective alleles; I: Shannon’s diversity index; He: expected heterozygosity; and PPL: percentage of polymorphic loci.

). The diversity analysis within accessions using Shannon’s index (I) as well as the expected heterozygosity (He) were estimated for all accessions as a single accession. They were ranked in the following descending order of locality: TAO-2, TIG-4, GUE-11, AZG-8, BET-13, IDA-12, TAL-7, ISS-10, AGL-3, IZO-6, AGR-9, DAR-5, to locality IMI-1, with an average of 0.389 (I) and 0.268 (He), respectively (Table 4).

3.3. Genetic Structure of Global C. sativus Accessions Based on Geographic Origin

The genetic distance and identity were calculated by Nei’s method [85]. Genetic differentiation between C. sativus accessions was relatively high, and all accessions were also significantly different (p < 0.05). In fact,

Table 5. Estimates of differentiation between C. sativus accessions from the 13 localities of the Taliouin region.

IMI-1	TAO-2	AGL-3	TIG-4	DAR-5	IZO-6	TAL-7	AZG-8	AGR-9	ISS-10	GUE-11	IDA-12	BET-13
****	0.771	0.945	0.784	0.938	0.869	0.885	0.806	0.896	0.875	0.647	0.844	0.938	IMI-1
0.260	****	0.748	0.782	0.737	0.708	0.757	0.834	0.768	0.738	0.701	0.708	0.742	TAO-2
0.056	0.291	****	0.760	0.886	0.821	0.874	0.803	0.881	0.856	0.682	0.854	0.907	AGL-3
0.244	0.245	0.275	****	0.767	0.707	0.781	0.732	0.769	0.735	0.583	0.714	0.755	TIG-4
0.064	0.305	0.122	0.265	****	0.910	0.887	0.836	0.848	0.909	0.611	0.863	0.870	DAR-5
0.141	0.345	0.197	0.347	0.095	****	0.797	0.812	0.810	0.839	0.690	0.804	0.784	IZO-6
0.122	0.278	0.135	0.247	0.120	0.227	****	0.804	0.829	0.813	0.661	0.809	0.873	TAL-7
0.215	0.181	0.219	0.312	0.180	0.209	0.218	****	0.763	0.833	0.739	0.840	0.767	AZG-8
0.109	0.264	0.127	0.262	0.165	0.211	0.187	0.270	****	0.835	0.732	0.809	0.902	AGR-9
0.134	0.303	0.156	0.307	0.096	0.176	0.207	0.183	0.180	****	0.642	0.869	0.827	ISS-10
0.436	0.356	0.383	0.540	0.493	0.371	0.414	0.302	0.313	0.442	****	0.718	0.693	GUE-11
0.169	0.345	0.158	0.337	0.148	0.218	0.213	0.174	0.212	0.141	0.331	****	0.838	IDA-12
0.064	0.298	0.097	0.281	0.139	0.243	0.136	0.266	0.104	0.190	0.367	0.176	****	BET-13

Pairwise-comparison matrix of sites and Nei genetic identity values (above diagonal) and Nei genetic distance values (below diagonal).

presents the pairwise genetic differentiation among C. sativus (saffron) accessions collected from 13 localities in the Taliouine region. It uses Nei’s genetic identity values (above the diagonal) and Nei’s genetic distance values (below the diagonal). Each column and row represent a locality (the full names correspond to the codes of localities provided in Table 1). The values above the diagonal represent Nei’s genetic identity values, which range from 0 to 1. Higher values indicate a closer genetic similarity between two accessions. For example, the genetic identity between IMI-1 and GUE-11 was 0.647, indicating that these two accessions are relatively similar. Conversely, the values below the diagonal represent Nei’s genetic distance values, which quantify the genetic divergence between two accessions. Smaller values correspond to greater genetic similarity, while larger values reflect greater differentiation. For instance, the genetic distance between IMI-1 and TAO-2 was 0.260, suggesting they are genetically closer compared to IMI-1 and GUE-11, which had a genetic distance of 0.436.

The highest Nei’s genetic diversity average between accessions was estimated between TIG-4 and GUE-11 accessions (0.540), while IMI-1 and AGL-3 accessions revealed the lowest levels of Nei’s genetic diversity by 0.056. Considering genetic identity, the values ranged from 0.583 to 0.945, while genetic distance varied from 0.097 to 0.540 (Table 5). A pairwise comparison of accession using genetic identity and genetic distance among the 13 localities of the Taliouin region demonstrated that localities IMI-1 and AGL-3 had the highest genetic similarity (0.945), while localities TIG-4 and GUE-11 had the least similarity (0.583).

Analysis of Molecular Variance (AMOVA) conducted on 76 Moroccan C. sativus accessions performed with the SRAP matrix revealed that the total genetic variation was highly significant (p < 0.001), with 81% accounting for the variation within accessions and 19% for the variation among accessions (PhiPT = 0.194) (

Table 6. AMOVA of 76 Moroccan C. sativus accessions based on data from SRAP markers.

Source of Variation	Df	SS	MS	Est. Var.	%	PhiPT	p
Among Accessions	12	284.773	23.731	2.372	19	0.194	0.000
Within Accessions	63	621.767	9.869	9.869	81
Total	75	906.539		12.241	100

Df = degree of freedom. SS = sums of squares. MS = mean squares. Est. var = estimate of variance. % = percentage of total variation. p-value is based on 1000 permutation.

). The fixation index (PhiPT = 0.194) indicates a moderate level of genetic differentiation. The highly significant p-value (0.000) confirms the robustness of the findings, highlighting the importance of genetic variation within accessions for C. sativus in Morocco.

3.4. Cluster Analysis

The estimated Jaccard’s index, calculated with data from SRAP matrix and used to build an Unweighted Pair Group Method with Arithmetic Mean (UPGMA dendrogam) tree, varied from 0 to 0.99 (p = 0.001).

The dendrogram clustered all C. sativus accessions coming from the 13 localities in the Taliouine region of Morocco into four main groups; individuals’ codes referred to Table 1. The coefficient scale at the bottom quantifies the genetic similarity between individuals. A value close to 1 indicates high similarity, while values closer to 0 represent greater genetic divergence.

Based on the genetic similarity, group one (blue lines) included nine accessions (IMI-1, AGL-3, DAR-5, TAL-7, AZG-8, AGR-9, ISS-10, IDA-12, and BET-13). Cluster II (brown lines) encompassed two samples of IZO-6 locality. The third group (pink lines) comprised all locality accessions. The last cluster (green lines) also contained different samples of nine accessions (IMI-1, AGL-3, DAR-5, TAL-7, AZG-8, AGR-9, ISS-10, IDA-12, and BET-13) (Figure 2).

3.5. Principal Coordinates Analysis (PCoA)

PCoA of 76 C. sativus samples collected from 13 localities of Taliouin region was performed to explore and visualize the relationships among the different accessions based on Dice’s similarity matrix. The PCoA of locus SRAP variation explained 26.07%, indicating the major patterns of genetic differentiation among the samples, with the first component (PCo1) interpreting 15.99% and the second component (PCo2) translating 10.08% of the total genetic variation (Figure 3). Each point in the plot represents an individual C. sativus sample. In fact, the points are color-coded and shaped according to their accession of origin. For example, red diamonds represent IMI-1, green squares represent TAO-2, and blue triangles represent AGL-3, etc. Regarding sample clustering, samples from the same accession tended to group closely, indicating genetic similarity within accessions. Some samples overlapped or clustered near samples from different accessions, suggesting potential gene flow or shared genetic traits.

Independently to their origins, a biplot generated by PCoA showed a high positive correlation between C. sativus accessions, as shown in the UPGMA cluster (Figure 2). However, despite the overall reduced variability at the genomic level (Table 6), the distribution of some points represented single individuals along the first factor (Cood. 1).

3.6. Population Genetic Structure

The reconstruction of the population structure, using STRUCTURE software v.2.3.4 following the Bayesian approach, made it possible to identify the best K value. In fact, the Evanno method implemented with STRUCTURE HARVESTER revealed the highest delta K (Δkat) value, with K equal to six, therefore suggesting that the most probable structure of testing the 76 C. sativus samples is their distribution into six main accessions (Figure 4). However, no clear geographic structure was detected using this approach.

C. sativus individuals are represented by colored columns. The same color in different individuals indicates that they belong to the same genetic group. However, multiple colors in each vertical bar show admixed genetic constitutions of each sample, therefore indicating the posterior probability to belong to different genetic clusters.

4. Discussion

Despite the significant national and international importance of Moroccan saffron, the genetic variability within and among Moroccan saffron accessions remains poorly studied. This lack of knowledge hampers efforts to optimize saffron cultivation and improve yields and quality through targeted breeding programs. Addressing this knowledge gap is crucial for the sustainable development of the saffron industry in Morocco [86].

Molecular genetic markers are highly advantageous for defining genetic diversity in plant breeding due to their precision and reliability [87]. The aim of the present study was to determine the genetic relationships and population structure among 76 Moroccan C. sativus accessions from the Taliouin region applying Sequence-Related Amplified Polymorphism (SRAP) markers.

The use of SRAP markers, originally developed for Brassica species [62], has proven to be particularly effective due to their higher polymorphism, discrimination power (DP), and higher reproducibility (consistent repeatability), which means they can identify genetic differences between samples with great accuracy and simplicity. This makes them more reliable for certain applications compared to other marker systems such as AFLP, SSR, ISSR, and RAPD [20,88,89,90,91]. This high level of polymorphism and reliability makes SRAP markers ideal for the identification of a greater variety of landraces among the studied resources [92].

Saffron is highly genetically homogeneous due to its asexual reproduction, using SRAP markers to target both coding (functional) and adjacent non-coding regions of the genome, offering insights into genetic variability linked to phenotypic traits [62]. Also, these marker characteristics make them suitable for detecting even subtle genetic differences in clonally propagated crops like saffron. While these specific SRAP primers were not originally developed for saffron, their application has been validated in previous studies, such as that by Babaei [8], which highlighted their utility in detecting genetic variability in saffron accessions from Iran.

Exploring saffron genotypes based on agro-morphological characteristics and molecular data is a critical foundation for enhancing traits, i.e., quality and yield-related traits in arid and semi-arid regions. These traits serve as valuable indicators for assessing genetic diversity, understanding population structure, and informing future saffron breeding programs [93]. Regarding saffron diversity, SRAP markers generated an average polymorphism rate of 96.66%. Considerable polymorphisms have been observed in previous studies as well. For instance, Keify and Beiki [94] reported polymorphism rates of 54% using RAPD and SRAP markers, while Babaei et al. [8] reported a polymorphism rate of 43.88% in C. sativus populations sampled from an important center of saffron production in Iran using SRAP markers. These findings underscore the effectiveness of SRAP markers in genetic analysis due to their high level of polymorphism, which is crucial for identifying genetic diversity and population structure. In more recent studies, advances in molecular marker technologies have continued to demonstrate the importance of such tools in genetic research [95]. For example, molecular markers like SRAP, AFLP, SSR, and ISSR have been also increasingly employed to analyze genetic diversity and population structure in various crop species, including saffron [44,58]. These markers provide a robust framework for breeding programs aimed at improving crop resilience, yield, and quality [96].

Content (PIC) value is a significant metric (index) used to define the discriminating power of marker among genotypes [97] and to assess the efficiency of polymorphic loci in identifying genetic diversity [8,26,98]. The average PIC value recorded during the present research was 0.30, ranging from 0.26 (ME3/EM3) to 0.36 (ME1/EM3) using the approach of Roldàn-Ruiz et al. [80]. This result indicates that ME1/EM3 was the most efficient and informative primer for discriminating among saffron genotypes. The PIC values detected in the present study were higher than those reported in previous saffron studies that had similar, the same, or different genetic marker systems. For instance, to detect genetic diversity among Iranian saffron accessions, Babaei et al. [8] reported the mean PIC value to be 0.150. Furthermore, Mir et al. [58] used RAPD, SSR, and ISSR markers to determine the molecular characterization of saffron and found the average PIC values to be 0.03 and 0.018 for RAPD and ISSR markers, respectively. Indices’ differences among studies could be appropriate to the instructive level of selected primer. The differences in PIC values among various studies could be attributed to the specificity and resolution level of the selected primers and the marker systems employed. SRAP markers are designed to amplify open reading frames (ORFs), which are regions in the genome that contain functional genes coding for proteins. These regions are more conserved and directly associated with phenotypic traits or biological functions [62]. This specificity makes SRAP markers particularly useful for assessing functional genetic diversity, which is essential for breeding programs aimed at improving crop traits such as disease resistance, stress tolerance, and yield [44,58].

Shannon’s information index (I) and Nei’s genetic diversity (h) produced by the six SRAP primer combinations for all accessions as a single population were 0.389 and 0.232, respectively. The genetic diversity indexes detected in the present study were higher than the mean Shannon’s information index (0.152) by Babaei et al. [8] using SRAP markers and by Mir et al. [58] using iPBS-retrotransposon markers (0.16). This higher genetic diversity observed in the current study underscores the effectiveness of SRAP markers in revealing genetic variation in Moroccan saffron accessions. Understanding the genetic structure of saffron populations can inform conservation strategies. Preserving genetically diverse populations ensures the long-term sustainability of the saffron plant. Efforts should focus on maintaining and enhancing genetic diversity through the conservation of traditional varieties and the exploration of new cultivars [44].

The AMOVA analysis performed with the SRAP analysis revealed that the majority of the variation in Moroccan C. sativus occurred inside accessions (81%) rather than between them (19%), with PhiPT = 0.194. This result showed that genotypes from the same location had a considerable genetic variation, but relatively little genetic variation was observed among localities. Using other genetic analysis, the monomorphism of C. sativus was also detected by other authors [92]. According to our findings, the low variation between accessions could be attributed to several factors. One major factor is the mating systems of C. sativus that could contribute to our results across accessions. In fact, the vegetative multiplication of C. sativus tenders advantages in maintaining the genetic characteristics of the plant [1,99] and may contribute to the low degree of genetic improvement among accessions. Brighton [100] also confirmed that C. sativus displayed closely homogenous and biological traits worldwide, and slight morphological differences were observed. Furthermore, their findings were in accordance with authors’ studies reporting the absence of genomic differences of saffron DNA from five different locations in Europe and Israel, using the RAPD technique [9].

The genetic diversity among saffron accessions was clearly presented in the UPGMA dendrogram and PCoA. In fact, despite their geographical localities, close relationships were revealed between accessions from different locations, showing that there was no relationship between genetic divergence and geographical locality. Likewise, Rubio-Moraga et al. [20], who used combined datasets from the RAPD, ISSR, and microsatellite analyses, strongly suggested that C. sativus is a monomorphic species as identical clones, not only because of morphological characters but also at the molecular level. Furthermore, Babaei et al. [8] also noted close relationships between saffron from important Iranian production regions, suggesting the existence of vegetative propagation, human selection of superior genotypes, and subsistence of narrow genetic base of saffron. The use of advanced techniques such as high-throughput sequencing and genome-wide association studies (GWAS) has facilitated the identification of genetic loci linked to important traits, thereby offering opportunities for improving saffron cultivation and breeding [63,101]. These technologies have also contributed to better understanding the evolutionary history and domestication processes of saffron, further supporting the need for preserving genetic diversity within saffron populations.

The reconstruction of the population structure using the STRUCTURE software v.2.3.4, based on the Bayesian approach results, supported UPGMA dendrogram findings regarding saffron accessions. This further supports the fact that the 13 accessions are subsamples of a population collected in the different area. The biological significance of saffron Moroccan saffron accessions can be explained by the high degree of commonness in these accessions due to the sterile and triploid nature of saffron species. As a matter of fact, they vegetatively propagate by saffron corm, and not undergoing sexual reproduction allows the preservation of the plant’s genetic characteristics [99,102]. Nevertheless, several authors have reported that hierarchical analyses, such as those based on STRUCTURE or similar software, rely on strict assumptions, thus resulting in incorrect evaluations of the population’s structure [102]. These analyses often assume that populations adhere to the Hardy–Weinberg equilibrium and that loci are in linkage in equilibrium; however, these conditions are not always met [83,103].

The reliability of a good genetic molecular marker system lies in its affordability, rapidity, ease, considerable level of precision, polymorphic, and equal distribution across the whole genome. Furthermore, among the important marker characteristics is the ability to differentiate between genetic differences using a low amount of DNA and no previous knowledge of the genome [104]. In fact, SRAP markers tend to have higher discriminating power compared to RAPD and ISSR markers, which both target various regions (coding and non-coding) of the genome. The application of RAPD markers in molecular-assisted breeding has been restricted due to their disadvantages. In fact, RAPD markers amplify random genomic segments of DNA throughout the genome, primarily within non-coding regions, using short arbitrary oligonucleotide short primers. These non-coding regions are less conserved, often capturing genetic variation that does not correlate with functional traits, and RAPD also suffers from reproducibility issues, further limiting its reliability. The amplification reactions for RAPD markers require highly standardized experimental protocols because they are very sensitive to the reaction conditions [105]. Moreover, RAPD markers generally require DNA in relatively higher quality and at a higher molecular weight [106]. Similarly, ISSR markers target regions between microsatellites, commonly located in non-coding parts of the genome. While useful for assessing genetic diversity due to their polymorphic nature, ISSR markers lack direct association with functional traits, reducing their utility for detecting biologically significant genetic differences [96].

Considering our results, using RAPD markers on previous studies reported the presence of artefactual bands (false positives and false negatives) [89]. Also, RAPD markers were less powerful for the detection the genetic diversity according to Nei’s genetic diversity (h) and Shannon’s information index (I) average, as reported by iPBS-retrotransposon markers [107]. On the other hand, SRAP results seem more trustworthy, as they have the highest average discriminating power among the different systems’ asset to amplify regions of the genome with primers targeting ORFs. In this manner, the findings from the current study confirm that SRAP markers satisfy all these criteria for Crocus sativus.

5. Conclusions

Assessing crop genetic diversity with molecular markers is crucial for both breeding programs and phytogenetic resources conservation initiatives, as well as for an effective cultivar selection. The findings of the current study highlight the efficiency of SRAP markers in evaluating the genetic diversity and population structure of Moroccan saffron. Thus, the observed high intra-location variation aligns with previous research conducted in Italy, Iran, and Spain. As saffron contains nutraceutical properties, due to the presence of an antioxidant substances, such as carotenoids, phenolic compounds, flavonoids, and vitamins [31], our data should be considered as valuable information for genetic improvement and conservation of expensive resources of Moroccan C. sativus in breeding programs, especially the Moroccan location that exhibited low genetic diversity (TAU-2 and GUE-11). Although, saffron enhancement is a fundamental prerequisite for a sustainable development strategy and agricultural diversification opportunities for growers. These results underscore the significance of molecular markers in revealing genetic variability and support ongoing efforts to enhance saffron cultivation practices, ultimately contributing to the socioeconomic development of saffron farmers in the region.

Recent advancements in next-generation sequencing (NGS) technologies and high-throughput genotyping have opened up new avenues to explore the genetic architecture of saffron (Crocus sativus L.) at an unprecedented depth. These genomic tools offer complementary insights by enabling genome-wide analysis, which can uncover subtle variations that further refine our understanding of saffron genetics [108]. Furthermore, the transformative potential of omics technologies, particularly epigenetics, in advancing saffron research will play a significant role in modulating traits such as the flowering time, stress tolerance, and stigma quality, which are critical to saffron production.