Introduction

The existence for forest genetic resources is increasingly important to fulfill current and future needs, such as increasing productivity and anticipating pest and disease epidemic (Nirsatmanto 2020). Indeed, they are important in maintaining biodiversity, protecting soil, and adapting forests to climate change. As the effects caused by human activities and climate change, there is a high risk of predisposing forest genetic resources to decline and loss. Efforts are ongoing to resolve trade-offs inherent in the demand for timber and the demand for other goods and services (e.g., recreation) (Aggestam et al. 2020).

According to FAO report, Africa and Asia are two regions with the highest forest species diversity and almost half of the forest species reported by countries are threatened (Beardmore et al. 2014). The conservation of forest species is the most rational way to keep fauna and flora ecosystem in balance. In conserving forest genetic resources, we are concerned primarily with conserving genetic diversity within and among populations of a species (Rajora and Mosseler 2001). However, plant species differ enormously in biological traits and environmental requirements, making it unrealistic to apply a single system to all species (Peñas et al. 2016). Moreover, establishment and management of a collection with many individuals, especially for perennial woody plants, need spacious areas, skillful labor, and expertise in plant identification which may be a high-cost operation (Brown 1989; Hintum et al. 2000). Commonly, it is not feasible to maintain the entire gene pool; therefore, so-called core collections are established (De Beukelaer et al. 2018). A core collection is a subset of accessions which were selected by eliminating closely related samples while still capturing the genetic diversity of the original set of accessions (Frankel 1984).

For tree species with a wide natural distribution range, the purpose of constructing a core collection is to protect the diversity of germplasm resources for high-quality characteristics that are easily identified in the wild (e.g., growth traits) while for tree species with a limited natural range, the purpose is to protect the diversity of germplasm resources (Wang et al. 2021). Many various core collection selection methods have been proposed, and each of them differs in defining the selection criteria for a respective core collection establishment (Kelblerová et al. 2022). Originally, phenotypic data containing both morphological and agronomical traits were used to create core collections, whereas nowadays, molecular markers as neutral tools for measuring genetic variation have become the tool of choice (Soleimani et al. 2020). They have been used for creation of core collection in various crops such as Triticum aestivum (Balfourier et al. 2007) (Pascual et al. 2020), Glycine max (L.) Merr. (Jeong et al. 2019), Juglans regia L. (Mahmoodi et al. 2019), Olea europaea L. (El Bakkali et al. 2013), Capsicum annuum L. (Zewdie et al. 2004), Setaria italica (L.) Beauv. (Upadhyaya et al. 2008), Saccharum officinarum L. (Balakrishnan et al. 2000), and Oryza sativa L. (Zhang et al. 2011). For most forest tree species, there are only few core germplasm collections based on genotypic or phenotypic traits or both, such as Picea abies (Kelblerová et al. 2022), Pinus yunnanensis (Wang et al. 2021), Eucalyptus cloeziana (Lv et al. 2020), Ceratonia siliqua L. (Di Guardo et al. 2019), Dalbergia odorifera (Liu et al. 2019), Quercus suber L. (Laakili et al. 2018), Cunninghamia lanceolata (Duan et al. 2017), Argania spinosa (Mouhaddab et al. 2016), and Cryptomeria japonica (Miyamoto et al. 2015).

Using molecular markers, for creating a core collection reflecting directly the variation at the genetic level, is preferred because each entry in the core is chosen on the basis of its allelic potential, irrespectively from its phenotypic expression (Mahmoodi et al. 2019). Indeed, the cost, time, and effort required for phenotypic characterization, especially in a woody perennial species collection, are much greater than required for an assessment using molecular tools (El Bakkali et al. 2013). In light of this situation, few studies were devoted to the genetic diversity and the structure of T. articulata using inter–simple sequence repeat (ISSR) markers (Makkaoui et al. 2020; Sánchez-Gómez et al. 2013) and amplified fragment length polymorphism (AFLP) and SSR markers (García-Castaño et al. 2021). Among different marker systems, ISSR are mostly dominant genetic markers that also include the AFLP, random amplified polymorphic DNA (RAPD) markers, and their derivatives (Ng and Tan 2015). The method provides highly reproducible results and generates abundant polymorphisms (Machkour-M’Rabet et al. 2009) and can rapidly differentiate the closely related individuals (Panahi and Ghorbanzadeh Neghab 2013). These characteristics allow ISSR markers to be suitable for developing core collection for T. articulata.

Two main methods have been used to assist the establishment of core collections: the M strategy (for maximization) and PowerCore. The M strategy or maximizing method is based on using genetic markers to sample the core collection while maximizing allele richness at each marker locus (Schoen and Brown 1993); it is worth noting that the M strategy not only determines the number of accessions that should come from different groups but also identifies the accessions that should be included (Hintum et al. 2000). PowerCore on the other hand selects the accessions with higher diversity representing the total coverage of marker alleles and trait states present in the entire collection (Kim et al. 2007) (Belaj et al. 2012). To the best of our knowledge, there are no studies focused on constructing a core collection for Tetraclinis articulata based on molecular marker data and maximizing method.

Tetraclinis articulata (Vahl) Masters consists of a monotypic Mediterranean genus of the Cupressaceae family, located in northern Africa and southern Spain (Rourke 1991). In Morocco, the extent of T. articulata reaches 80% of the total tetraclinaie area (HCEFLCD 2004). Like other woody species, T. articulata is a multifunctional conifer offering ecological and economical functions: high yield, high pest resistance, and a very desirable wood (Fidah et al. 2015), appreciated for its natural beauty and its quality for marquetry (El Bouhtoury-Charrier et al. 2009). The chemical composition of essential oil extracted from T. articulata showed antimicrobial activity against Staphylococcus aureus and Micrococcus luteus (Bourkhiss et al. 2007). At present, Arar tree is in an elevated risk of losing habitat area and genetic material from its gene pool caused by fires, invasive species, storms, and droughts combined with greater competition for land with other sectors, including agriculture and urbanization. Accordingly, T. articulata requires particular attention for conservation throughout in situ and ex situ conservation, germplasm banks, and core collections. In light of this situation, only two studies were devoted to study the genetic diversity of T. articulata using ISSR markers (Makkaoui et al. 2020; Sánchez-Gómez et al. 2013).

In order to construct a core subset of T. articulata by understanding the genetic diversity pattern and structure, 15 ISSR markers are used to assess the genetic diversity of 129 germplasms of T. articulata originated from ten wild populations across the entire distribution area in Morocco (Makkaoui et al. 2020). Analyzing differences in genetic diversity of the core collection and through the entire population is a useful approach for testing the effectiveness of the core collection (Wang et al. 2011). Hence, choosing the suitable evaluation measurement depends on the purpose of the core collection. Core collections can be created with two general purposes in mind: (1) maximizing the total genetic diversity in a core (2) maximizing the representativeness of the genetic diversity of the whole collection in a core collection (Marita et al. 2000). A variety of measures have been used to evaluate the core collection (Wang et al. 2021). Genetic diversity parameters, such as total alleles, Na, Ne, H, I, and PCoA, are frequently used to evaluate the genetic diversity of the core collection (Li et al. 2006; Wang et al. 2011). Other criteria are also used to evaluate and validate the representativeness of the core collection such as the analysis of molecular variance (AMOVA) and the unweighted pair group method with arithmetic (UPGMA) based on genetic distances. In this context, the main objective of this study was to provide a methodical strategy for the construction of a core collection for T. articulata germplasm resources and evaluate their genetic diversity and structuring based on variations as revealed by ISSR markers.

Material and Methods

Sampling and Assessing Suitability DNA for ISSR Amplification

The initial collection is composed of 129 samples from ten natural forests across the entire distribution area in Morocco (Makkaoui et al. 2020). Using a commercial kit (DNA mini kit, Bioline, USA), genomic DNA was extracted from 129 lyophilized and ground leaves (50-mg powder), according to the manufacturer’s instructions. Purified DNA quality was verified by spectrophotometry (ND-2000, NanoDrop, USA) at 260 and 280 nm and by electrophoresis on 1% agarose gel. The DNA concentration was adjusted to a 50 ng/μL and stored at − 20 °C for ISSR amplification.

A total of 51 primers were initially screened. Fifteen of them, which generated discernable and bright bands, were employed for the analysis of 129 samples (Makkaoui et al. 2020). The ISSR amplifications were carried out in a total volume of 25 μL containing 2.5 mM MgCl2, 2 mM dNTP, 4 μM of primer, 1 unit Taq DNA polymerase, and 100 ng of total DNA. The reaction mixture was subjected to amplification in GenAmp® thermal cycler (Applied Biosystems, CA, USA) following the program: initial denaturation for 7 min at 94 °C, followed by 39 cycles of 30 s at 94 °C, 45-s annealing temperature 2 min at 72 °C, and 7 min final extension at 72 °C. A control was run by replacing genomic DNA with ddH2O. Amplified products were electrophoresed in 2.8% agarose gel in 1X TAE buffer at 120 V for 2 h and then stained with ethidium bromide for 20 min. Gels were visualized under ultraviolet and photographed with Enduro™ GDS (Labnet, USA). The size of amplified fragments was estimated using 1-kb ladder molecular size standards (Bioline, USA) (Makkaoui et al. 2020).

Allelic Richness Distribution and Structure of the Full Collection

ISSR amplified fragments were scored for band presence (1) or absence (0), and a binary qualitative data matrix was constructed as previously described (Makkaoui et al. 2020). The allelic richness—also known as the neutral variability—and the polymorphism of the 10 populations of Articulata were mapped using the DIVA-GIS software (Hijmans et al. 2001); it was also used to map frequently alleles, locally common alleles, rare alleles, and private alleles (Fig. 1).

Fig. 1
figure 1

Spatial variation of different genetic parameters, represented at a resolution of 30-s grid cells and circular areas of 10 km

Full size image

To explore spatial genetic structure of T. articulata in Morocco, the spatial principal component analysis (sPCA) (Fig. 2) was performed using all loci.

Fig. 2
figure 2

Scores of sampled trees as projected on the first ordination axis of the biplot of a spatial analysis of principal components

Full size image

The ten original germplasm resources were grouped into two major regions, according to geographical proximity, for further analysis:

  1. 1.

    North and East regions: Oued Laou, Dar ben Karrich, Chefchaouen, Al Hoceima, and Oujda

  2. 2.

    Central and South regions: Oulmes, Fes-Boulemane, Beni-Mellal, Essaouira, and Marrakech

Core Collection for T. articulata

Core Collection Construction

In the present study, the maximizing strategy implemented in PowerCore software (Kim et al. 2007) is used to construct the core subset. The M strategy (maximization or marker allele richness) maximizes the number of alleles at each marker by eliminating redundant accessions (Ndjiondjop et al. 2017; Schoen and Brown 1993). PowerCore selects from the whole collection the accessions with the highest diversity through a modified heuristic algorithm (Mahmoodi et al. 2019). Thus, the size of the final core collection is not known a priori and it depends on the levels of variability and redundancy present in the collection (Belaj et al. 2012).

Genetic Diversity Analysis

Using GenAlEx v6.5 software and assuming the Hardy–Weinberg equilibrium, the genetic diversity parameters (molecular traits) (the observed number of alleles (Na), the effective number of alleles (Ne), the Shannon information index (I), the expected heterozygosity (He), the observed heterozygosity (uHe), the percentage of polymorphic loci (%P), and allelic frequencies were calculated for the primary collection and the core collection GenAlEx v6.5 was also used to describe the AMOVA, the genetic differentiation among and within populations using 999 permutations, and principal coordinate analysis (PCoA). The mean t-test and variance F-test, respectively, were employed to detect whether the differences of the molecular traits and the AMOVA between the core subset and the primary collection were significant.

Genetic Structure of the Core Collection

A dendrogram of the genetic relationships among the core collection germplasms was constructed based on the unweighted pair group method with arithmetic mean (UPGMA). The genetic structure of the core collection was performed using STRUCTURE software version 2.1 (Pritchard et al. 2000) and STRUCTURE Harvester (Girardot et al. 2016), based on model testing for k-values ranging from 1 to 14 with 20 independent runs per k value. For each run, the burn-in was set to 50.000 followed by 100.000 MCMC (Markov Chain Monte Carlo) iterations. Then, the number of clusters ΔK was estimated (Evanno et al. 2005). The core collection will permit to have a representative of the whole genetic diversity of T. articulata in Morocco.

Results

Allelic Richness and Polymorphism Distributions of the Whole Collection

The software DIVA GIS was used to map the polymorphism and the allelic richness of the entire collection. In general, an elevated level of polymorphism exceeding 50% was detected. The allelic richness is a straightforward measure of genetic diversity that is commonly used in studies based on molecular markers and aim at selecting populations for conservation (Petit et al. 1998). The high number of alleles present in Chefchaouen (242) and Al Hoceima shows the lowest allelic richness (172) (Fig. 1). In the whole collection, four populations retain both locally common and rare alleles, respectively: Essaouira, Fes-Boulemane, Chefchaouen, and Dar ben Karrich. Thus, rare alleles are essential in conservation because they represent unique evolutionary products and could provide the species with advantageous properties to cope with eventual environmental shifts (Peñas et al. 2016); they do not have a satisfactory representation in our study (0 to 2 alleles). The sPCA is a multivariate method devoted to the identification of spatial genetic patterns (Jombart et al. 2008). In our case, the sPCA clustered the collection in five groups: Oujda, Al Hoceima; Essaouira, Beni-Mellal, DB Karrich; Chefchaouen, Oulmes, Fes-Boulemane; and Marrakech, Oued Laou (Fig. 2).

Genetic Diversity of Core Collection

Selected Core Collection Individuals

Based on the results of the PowerCore analysis using a heuristic algorithm, 271 bands were obtained using 15 ISSR markers (Fig. 3), for 129 T. articulata germplasm and can be represented in a collection of 13 germplasms. The heuristic method is used as part of a global procedure that guarantees to find the optimum solution for a problem (Martí and Reinelt 2011). The 13 germplasms belong to eight provenances and are grouped on four regions, with the maximum genetic diversity, constituting the most representative core collection of T. articulata in Morocco, noticing that germplasms from Beni-Mellal and Oued Laou are not represented in the core collection. The allelic richness varied from 13 (TF7) to 115 (TC4) (Table 1) and the core collection sampling proportion is 10.07% (Table 2).

Table 1 List of core collection germplasms
Full size table
Fig. 3
figure 3

ISSR amplification pattern of selected T. articulata germplasms using the primer BTH1 (A). ISSR amplification pattern using primer BTH11 for germplasms from the entire collection (B)

Full size image
Table 2 Genetic diversity per locus in core collection and entire collection
Full size table

The Genetic Diversity of the Core Collection and the Whole Collection

The estimates of genetic diversity (molecular traits) for the core collection and the whole collection are shown in Table 3. There were no significant differences in the mean of the effective number of alleles (Ne), the Shannon information index (I), and the expected heterozygosity (He). Moreover, the values of Ne, I, and He presented by the core collection are higher than those presented by the entire collection (1.398, 0.395, and 0.253 and 1.363, 0.382, and 0.238, respectively). The t-test (P < 0.005) of the observed number of alleles exhibited significant differences between the core collection and the entire collection. Na is lower in the core collection than this presented by the primary collection (1.398 and 1.712, respectively). The variance F-test (P < 0.005) showed no significant differences in the mean genetic diversity parameters (Na, Ne, I, He), between the core subset and the primary collection.

Table 3 Results of mean t-test and variance F-test between the primary collection and core subset of T. articulata based on molecular traits
Full size table

Genetic Structure of the Core Collection

Complementary measures are utilized to describe the genetic structure of 13 selected germplasms of T. articulata. Based on Nei’s genetic distances. The PCoA revealed three main clusters: TA1-TA17-TF7-TO2-TO3, TC15-TF2-TOM6-TM2-TM6, and TC4-TDBK3-TS10 (Fig. 4). STRUCTURE software version 2.1 (Pritchard et al. 2000) including the Bayesian method was also used to determine the structure and genetic relations among 13 germplasms of T. articulata, identified via STRUCTURE Harvester (Girardot et al. 2016). The K-value was used to estimate the number of clusters of the selected collection. The results indicated that ΔK reached a sharp peak when K = 2 (Fig. 5), suggesting the presence of two main groups in the core collection (Fig. 4). In addition, there was a small peak observed at k = 9 (Fig. 5), which indicates another informative population structure (Fig. 5). Hence, the core collection individuals were classified into two populations: TA1, TA17, TF7, TO2, and TO3 and TC5, TC15, TDBK3, TF2, TOM6, TM2, TM6, and TS10. To further determine the relationships among the core collection, UPGMA clustering was conducted using Nei’s unbiased genetic distance matrix. The dendrogram clearly shows two groups (A and B). Within group A, three individuals (TS10, TDBK3, TF1) clustered in subgroup (A1) and five individuals (TM1, TM2, TC1, TC2, TOM6) clustered in subgroup (A2). Within group B, four individuals (TF2, TO1, TO2, TA2) formed a subgroup while TA1 formed another subgroup (Fig. 6). Based on genetic distances, there was a similarity between PCoA and UPGMA in discriminating T. articulata germplasms.

Fig. 4
figure 4

Principal coordinate analysis of Tetraclinis articulata populations of ISSR based on pair-wise Nei’s (1973) genetic distances

Full size image
Fig. 5
figure 5

Results of STRUCTURE analysis for 13 T. articulata germplasms based on ISSR markers. a Estimation of population using lnP(D)-derived delta K with cluster number (K) ranged from one to 14. b Genetic STRUCTURE of 13 germplasms when k = 2. c Genetic STRUCTURE of 13 germplasms when k = 9

Full size image
Fig. 6
figure 6

UPGMA dendrogram of core collection (13 germplasms) and entire collection (129 germplasms) of T. articulata constructed based on genetic similarity coefficients of ISSR data in the core collection and entire collection, respectively

Full size image

Correlation Between the Core Collection and the Whole Collection

The number of amplification fragments produced by the 15 ISSR primers in the core collection and the full collection is shown in Table 2. The amount of polymorphism rate in the core collection and the whole collection shows differences, 98.52% and 85.61%, respectively. However, no significant difference through mean F-test (P < 0.005) but significant difference is observed by the mean T-test (P < 0.005) (Table 3). The band frequencies per locus in the core collection are lower from those in the entire collection suggesting that all band types are presented in the core collection. The Shannon index (I), the observed heterozygosity (He), and the expected heterozygosity (uHe) are calculated respectively for both core and entire collections. Of 15 ISSR primers, the observed heterozygosity (He) and the expected heterozygosity (uHe) for BTH 6 primer were lower in the core collection, accounting 6% of all the ISSR primers used. The AMOVA carried out for the core subset and the entire collection is summarized in Table 5; the variation among population is 4%, 9% respectively; and the variation within population is 96% and 91% respectively, showing a difference in the variation among population (Table 4). By the mean of T-test (P < 0.005) and F-test (P < 0.005), there are no significant differences between the core subset and the primary collection (Table 5).

Table 4 AMOVA of T. articulata core collection populations (P < 0.001)
Full size table
Table 5 Comparison of AMOVA of T. articulata between the core collection and the entire collection populations (P < 0.001)
Full size table

Discussion

Areas of Moderate and High Level of Allelic Richness

Measures of genetic diversity based on allelic richness are considered very important in the field of genetic conservation, and marker-assisted methods for maximizing the number of alleles conserved have been shown to be effective (Schoen and Brown 1993). Besides, the allelic richness consists on estimating the number of alleles expected in samples of specified size, using the rarefaction method applied in ecology (Foulley and Ollivier 2006). In previous investigation, populations of T. articulata show low level of inbreeding (Gst = 0.13) which is confirmed by the high level of heterozygosity (0.221); this one plays an important role in the response to environmental changes (Slatkin and Barton 1989) and indicate that the species did not endure drastic genetic bottlenecks (Makkaoui et al. 2020). Inbreeding may occur because of reduction and fragmentation of natural stands and cultivated areas, increasing the risk of allele loss, which eventually leads to genetic erosion (Lowe et al. 2005). On the other side, the Moroccan collection of T. articulata shows differences in allele distribution; populations of Essaouira, Fes-Boulemane, and Marrakech hold the highest number of both locally common and rare alleles. The priority for conservation should be given to populations that retain locally common alleles; these alleles occur in high frequency in a limited area and can indicate the presence of genotypes adapted to specific environments (Frankel et al. 1995).

Evaluation of the Core Collection of T. articulata Germplasm

One of the major goals of establishing a core collection is to collect a representative germplasm of the complete collection for in situ and ex situ conservations. Our study is aimed at selecting individuals representing the whole collection of T. articulata; this type of core collection called CC-I is especially suitable, for situations requiring an overview of the genetic diversity of the accessions of the whole collection (Odong et al. 2013). Thus, the present core collection is created based on 15 ISSR markers and assisted by maximizing strategy (M) implemented in PowerCore software; it allows all alleles to be captured in a minimum number of accessions (Kim et al. 2007). This is the first such attempt in the Moroccan forestry. The analysis results in a core collection with 13 germplasms forming eight gene pools of T. articulata (Table 1). Oued Laou and Beni-Mellal regions were not included in the core collection because they do not contain a useful significant fraction useful for conservation strategy. The selected individuals have most of the genetic diversity and richness; also, they captured all the alleles (100%) existing in the entire collection (Table 2). The M strategy is the most appropriate approach for selecting entries with the most diverse alleles and eliminating redundancy (Franco et al. 2006). The percentage of germplasms selected by the algorithm is 10.07%, which is of reasonable size as previous studies proposed 5–20% core sizes (Thachuk et al. 2009; Hintum et al. 2000). However, there is no perfect proportion or fixed size for all core collections, as different species have different research needs because of the breadth and complexity of germplasm resources (Wang and Li 2022).The small size of our core subset may be explained by the high redundancy and the presence of clonal origin (El Bakkali et al. 2013) or by the mode of reproduction (Noirot et al. 2003; Hintum et al. 2002). Various criteria were used to evaluate and validate the relevance of the established core collection. The evaluation of a core collection with the same data set that was used to create it ignores this simple but very important point (Odong et al. 2013). This is quite important, especially in the case of molecular marker data where the key assumption is that by maximizing diversity in a given set of marker loci, the diversity of genes of interest will also be maximized (Odong et al. 2013). By means of T-test and F-test showing no significant difference between the core subset and the entire collection, the selected collection could represent the diversity of the original germplasm. Hence, the number of effective alleles, Shannon index, and the observed heterozygosity of the selected collection were higher than those in the entire collection, an expected result, since genetic diversity increases with the elimination of genetically similar (or redundant) accessions (Agrama et al. 2009). The high variance within population revealed by the AMOVA in the core subset and the entire collection may explain the post-fire resprouting capacity of T. articulata (Hadjaj-Aoul et al. 2009; Terras et al. 2008), allowing the persistence of the ancestral gene pool and increasing the generation time of this species (Premoli and Steinke 2008). In other hand, based on the DNA amplification of ISSR marker system, the genetic distance between the selected germplasms of distinct geographical regions was revealed. The clustering methods UPGMA and PCoA are used to convert the large and complicated data set into the simplest graphical representations (Pandian et al. 2018). PCoA and UPGMA clustered the 13 germplasms according to their genetic distance and genetic similarity. The clustering of the selected germplasms did not only reflect their geographic locations or their climatic zones but also reflect the interaction of different processes (Sánchez-Gómez et al. 2013). Each cluster contains germplasms from locations with maritime and continental effects, i.e., Essaouira, Dar Ben Karrich and Fes-Boulemane; Oujda, Al Hoceima, and Fes-Boulemane. Moreover, the same group contains geographically distant individuals, i.e., Chefchaouen and Marrakech, proving the extent of genetic relationship between the genotype of the same species (Pandian et al. 2018). The present distribution of the selected individuals of T. articulata is probably occurred by human activities (trade and deliberate introductions for its wood) or naturally wind-mediated pollen or seed dispersal (García-Castaño et al. 2021). The STRUCTURE program revealed another pattern of the selected collection, suggesting groups of geographically close individuals, i.e., Al Hoceima, Oujda, and Fes-Boulemane; Chefchaouen, Oulmès, and Marrakech. It was noted that the germplasms belonging to the north (Al Hoceima, Oujda) may have an Iberian origin (Sánchez-Gómez et al. 2013). Therefore, the three grouping methods used in this study (UPGMA, PCoA, and STRUCTURE) did not manage to significantly group the selected core subset in identical clusters. From another point of view, the weak spatial genetic pattern and the low number of fossil records in the southern shore of the Mediterranean are major impediments for phylogeographical reconstructions in T. articulata (Anderson et al. 2009). Different hypotheses could explain the complex structure of T. articulata; i.e., fossil evidence indicates that Tetraclinis Masters spread in the past from North America to Europe (Kvaček et al. 2000) or a genetic affinity between European and African populations (Kvaček et al. 2000).

Conclusion

To the best of our knowledge, the present study is the first to establish a core collection of T. articulata based on ISSR molecular markers and the M strategy implemented in PowerCore software. The synergy of these approaches revealed an important potential to locate germplasms with high genetic diversity that could serve accelerating breeding programs and exploring genetic diversity for conservation of T. articulata germplasms.

However, one type of data may not be sufficient to accurately assess the genetic variability of a germplasm and to develop effective core collections that capture the maximum genetic diversity of a germplasm (Yibo et al. 2010). Therefore, both molecular marker and morphological data should be considered when constructing a core collection (Wang et al. 2011). Moreover, core collection based on medicinal traits may also give insights about the representation of Arar tree diversity.