1. Introduction
Tunisia has a large and wealthy forage and pasture biodiversity. According to Abdelguerfi and Abdelguerfi-Laouar [
1], there are more than 960 pasture legumes species, 336 of which are Mediterranean endemic. However, native and endemic Tunisian pasture plants are endangered due to many factors among which overgrazing [
2] and natural habitat loss into introduction of intensive farming and which are the most important ones [
3,
4]. Thereby, conservation and valorisation of pasture and forage genetic resources have become key tasks, especially in Tunisian Dorsal regions where pastoralism is an important genetic loss drivers. However, pasture genetic resources in Tunisia are still less developed, despite the fact that they can enrich the Gene Bank and include many varieties in breeding [
5].
Lucerne (
Medicago sativa L.) is also known as queen forage alfalfa [
6] thanks to its ecological attributes by avoiding erosion phenomena [
3], and its wide agronomic assets by way of its protein and nitrogenous matter contents [
7,
8]. Perennial species of
Medicago genus have an essential role in the economic sustainability of crop-livestock systems [
9]. Nevertheless, alfalfa productivity is limited by some abiotic stress such as salinity [
10,
11].
Medicago tunetana (Murb.) A.W. Hill is a perennial C3 species native to Algeria and Tunisia thriving in calcareous mountains of Western North and Midwest regions of Tunisia [
12]. Tunisian alfalfa has many interested eco-physiological assets such as drought tolerance, winter hardiness and calcareous soil tolerance; this genetic resource has been unstudied so far and threatened to disappear. Furthermore, the efficient management of this pasture plant species in its native regions of the Tunisian ridge may contribute in the agro-pastoral systems development. The plant breeding for the fodder crops like alfalfa is of a high importance especially for the farming system sustainability [
9]. For an optimal valorisation, it is necessary to conserve, evaluate and estimate the genetic and morphological variation of
M. tunetana and to determine the relationship within this endemic species and cultivated alfalfa. In fact, there is an ambiguity in
M. tunetana taxonomy; it is considered as a subspecies of
M. sativa [
13,
14,
15]. However, and according to other authors,
M. tunetana is classified as an independent species [
16,
17,
18].
Molecular characterization has become the most efficient and reliable tool for studies of the genetic diversity between populations and species as well thanks to its environment independency [
19,
20]. In fact, it is not influenced by the stage of plant development facilitating genetic resources management within Gene Bank [
21]. Molecular markers are complementary to morphological and biochemical markers [
22]. Several studies have been realized on the genetic diversity of
Medicago genotypes using different markers such as AFLP [
23,
24], RAPD [
25,
26], RFLP [
27,
28] and SSR [
29,
30,
31,
32]. The microsatellites as polymorphic DNA markers have been widely employed in molecular studies thanks to their facility and efficiency for population genetic analysis [
33,
34,
35] while, using SSR markers for genetic variability within perennial
Medicago genotypes had not been very developed. Diwan et al. [
36] are the first authors to demonstrate the use of SSRs to characterize genetic diversity and to analyse the genetic relationships among
Medicago genotypes as well. Cholastova and Knotova [
30] indicated that the genetic diversity estimation of
Medicago genotypes could be obtained even with only three SSR markers. Furthermore, 107 SSRs identified in the EST database of
Medicago truncatula Gaaertn. was mapped also in
M. sativa [
37].
In order to study the M. tunetana collection, an ecological and edaphic characterization of different prospected sites was realized and followed by morphological and molecular characterization for seven accessions of M. tunetana and one variety of M. sativa used as a reference. For the molecular characterization, seven microsatellites markers were used in order to estimate the genetic diversity among M. tunetana local genotypes and to analyse the genetic relationship between M. tunetana and M. sativa.
3. Discussion
M. tunetana had been found in three new prospected sites of the high mountains of Tunisian Dorsal namely Bargou (MT1), El Ayoun (MT7) and Thala (MT5) (
Figure 1). These sites were not mentioned in Tunisian flora [
43]. It must be noted that annual precipitation is ranging between 522 mm for Bargou and 299–327 mm for the others dorsal sites for the last twenty-seven year period (
Table 1). For annual average temperature, Makthar presented the lowest value of 19 °C, while the highest value was recorded by the Southern dorsal regions with an average of 20.5 °C [
4].
M. tunetana is native to Western North Tunisia and Algeria. It has many important ecological and agronomical interests including abiotic stress tolerance (cold tolerance, rhizome production and calcareous soil tolerance). These results are consistent with those obtained by El Makki-Ben Brahim et al. [
4] who mentioned that the dorsal regions are characterized by a calcareous soil and those of Ferchichi and Rouz [
43] who demonstrated that
M. tunetana genotypes are tolerant to limestone excess in soil. Conservation and management of this rare pastoral genetic resource is an essential preoccupation in order to contribute to the grassland amelioration of the mountains of Western North and Western Centre regions of Tunisia.
In this study, we investigated the genetic diversity of seven wild accessions of
M. tunetana and local variety of
M. sativa as reference using seven SSR markers. We obtained 54 different alleles for a sample of 20 plants per accession (
Table 3). While Li et al. [
20] have obtained just 22 different alleles in
M. sativa subsp.
falcata populations using the same number of SSR markers for a sample size of 25 plants. Whereas Falahati et al. [
29] obtained 68 alleles using eight SSR markers for a sample size of only 10 plants for each accession. Therefore, Julier et al. [
44] noted that a sample size of 40 plants is more moderate for tetraploid alfalfa genetic diversity. In fact, these results are consistent with those of Andru [
45] who reported that the number of observed alleles is highly dependent on the size of the studied samples. A sample size lower than ten plants gives rise to the loss of even non-rare alleles [
44]. The studied
M. tunetana genotypes show a large genetic variation with two different groups and two subgroups for only seven accessions (
Figure 3) which confirms the hypothesis of tetraploidy for this species. The high number of alleles per locus may be explained by the high heterozygosity and allogamy of heterogenous and allogamous genotypes of
M. sativa [
29]. This result concords with those obtained by Heyn [
46] who reported that all perennial species of
Medicago genus are auto-tetraploid. Whereas it differs with that obtained by Abdelkefi et al. [
47] who mentioned that
M. tunetana. is diploid and each gene is presented by only two copies. All forage species are frequently allogamous which makes their genetic card study more complex and difficult to carry out [
48]. This allogamy is one of the most important causes of genetic erosion as well as of initial allelic variation loss [
49] and obtaining new characters. The observed allogamy in perennial
Medicago species may be explained by the gametophytic self-incompatibility which makes getting a pure line so difficult or even impossible [
48].
The molecular study for seven
M. tunetana accessions and local variety of
M. sativa using seven SSR markers had showed a large genetic diversity within
Medicago genotypes. Two different groups of
Medicago accessions were formed based on the generated dendrogram at 57% of similarity. Gardon et al. [
50] showed high degree of variability among
M. sativa genotypes with six SSR markers. Cholastova and Knotova [
30] demonstrated that the estimation of genetic diversity of
Medicago genotypes could be obtained even with only three SSR markers. Based on obtained results, used SSR markers were as efficient to determine genetic variation among the studied
M. tunetana accessions. This work is the first genetic diversity study of
M. tunetana genotypes; the number of SSRs used (
Table 3) as well as the broad genetic base of
M. tunetana could be considered not enough. Therefore, this work must be completed by other molecular study using more SSR markers in order to obtain more information about genetic variability of
M. tunetana and karyotype analysis.
Exsitu conservation of
M. tunetana accessions as a
Pastoritum or seeds in Gene Bank contributes to this genetic resource conservation as well as in the local alfalfa breeding programs.
M. tunetana is an allogamous plant species needing an optimal sample size more than 20 individuals in order to obtain reproducible bands to differentiate between accessions within an active collection. A population sample size lower than 20 individuals could increase the effect of genetic drift which generates loss of biodiversity. Sample size to be conserved in Gene Bank must be surrounding 200 seeds per accession which will allow preserving genetic diversity [
51].
4. Material and Methods
4.1. Origin of Plant Material Collection
In order to collect
M. tunetana genotypes, many prospecting and collecting missions were carried out by visiting sites reported by Pottier-Alapetite [
15] during five years (2015–2020) in the mountains of Western North and Western Centre regions of Tunisia that are located in three different Tunisian governorates (Siliana, Kef and Kasserine) (
Table 1).These regions are characterized by a continental climate and generally roughly with an average annual rainfall ranging between 220 mm and 550 mm [
4].
4.2. Sample Size
Six wild accessions of M. tunetana were collected for this work. For the morphological study, eight plants from each accession were randomly selected and used. For the genetic analysis, a local variety of M. sativa was added to be used as reference. Twenty plants per accession were randomly selected, and green healthy leaves from each plant were chosen for DNA extraction. This study was carried out in the molecular laboratory of National Gene Bank of Tunisia.
4.3. Ecological Analyses
The ecological characterization of each prospected site was determined in order to evaluate the most favourable environmental factors (climatic, soil and altitude) for M. tunetana genotypes development. In fact, the annual precipitation and number of rainy days for each prospected site were obtained for twenty-seven years (1992 to 2019) from the Ministry of Agriculture, Hydraulic Resources and Fisheries of Tunisia. Therefore, three representative soil samples from each site were taken and had been submitted to four physical analyses (clay%, fine silt %, coarse silt % and sand %) and five chemical analyses (pH, organic matter %, total and active limestone % and potassium content).
4.4. Morphological Traits
M. tunetana accessions variability was examined according to description given by Pottier-Alapetite [
15] for its taxonomic classification. The morphological characterization was determined based on 48 samples and six
M. tunetana genotypes. Eleven quantitative traits relating to leaves, pods and seeds were measured: leaflet length (LL), leaflet width (LW), petiole length (PL), petiolule length (PeL), stipule length (StL), pod length (PoL), pod width (PoW), number of pod turns (NPT), seed length (SL), seed width (SW) and seed thickness (ST). Data were measured using electronic digital calliper (0–150 mm) with an accuracy of 0.01 mm.
4.5. Molecular Analysis
Genomic DNA was extracted from fresh leaf tissue plants of
M. tunetana according to the cetyltri-methyl-ammonium bromide (CTAB) technique of [
52] with slight modifications. Plant material was ground up in liquid nitrogen, resuspended in 1 mL of CTAB 2 × buffer and incubated at 60 °C for 30 min. After chloroform-isoamylic (24:1) extraction, the aqueous phase was collected and the nucleic acid was precipitated with cold isopropanol, then washed with 70% ethanol and resuspended in 1 × TE buffer after being air-dried. Genetic diversity of different accessions of
M. tunetana was evaluated using seven SSR markers identifying QTL genes responsible of the plant height which are positively correlated with forage yield potential in the model species
M. truncatula [
37]. These microsatellites are saved in the databases of EST (expressed sequence tag) and selected by Julier et al. [
37] as an efficient marker for molecular characterization of autotetraploid and allogamous plant’s
Medicago genus (
Table 6). The PCR was performed in a final volume of 25 µL containing 100 ng of genomic DNA, 0.5 µM of each primer pair, 2.5 µL of 1 × PCR buffer, 1.5 mM of MgCl
2, 0.2 mM dNTPs and 0.2 µL of Taq DNA polymerase. DNA amplification was performed in a Bio Rad C1000 thermocycler programmed with 35 cycles: 1 cycle for initial Pre-Denaturation at 94 °C for 4 min; 35 cycles consist of denaturation at 94 °C for 1 min, Melting at Tm °C for 1 min and extension at 72 °C for 2 min;1 cycle for final extension at 72 °C for 3 min. PCR products (12 µL per sample) were separated in 3% agarose gel at 100 volts constant power for 90 min and then visualized with BET fluorescence in UV table.
4.6. Statistical Analysis
A two-way analysis of variance (ANOVA) was realized for soil analyses and quantitative morphological traits to evaluate difference between means which genotype was the first factor (MT1, MT2, MT3, MT4, MT5, MT6, MT7). Results were statistically significant at p < 0.05 for Tukey test using the statistical software SPSS statistics version 22.0 (IBM, Armonk, NY, USA).
Morphological, ecological and edaphic traits were examined by a canonical correspondence analysis (CCA) in order to determine the relationship between the analysed traits and the similarity among accessions of
M. tunetana using PAST software, version 4.03 [
53]. The CCA was carried out to the data matrix (11 morphological traits, 6 ecological traits, 10 edaphic traits and 6 genotypes).
For the molecular traits, the bands visualized on the gel were identified to determine whether they were a single allele or just auto-amplification to be discarded. Faint and unreliable bands were not considered for the analysis. Amplified fragments were scored as 1 or 0 for bands presence or absence, respectively. Sizes amplification bands were estimated using DNA Leader 100 pb marker. Genetic distance between accessions was calculated according to Nei [
42]. Unweighted pair group method using arithmetic average UPGMA dendrogram were drawn using the SAHN clustering of NTSYSpc 2.10, based on Genetic Distances.