Seeds of 72 accessions of Trifolium were procured from the Western Australian and Southern Australian gene banks (part of the Australian Pastures Genebank-APG); however, three of them did not germinate (Table 1). Seed were sown in plastic pots in glass house condition at Wagga Wagga Agricultural Institute, Wagga Wagga, NSW Australia maintained at 18˚C ± 4˚C (Figure 1). In order to validate the genetic identity of accessions, passport information on growth habit, shape of cotyledonary and the first true leaf; and number of leaflets at the seedling stage, was recorded at seedling stage. The plants were cut; leaving 5 cm growth from the base (soil surface) after 45 and 65 days of sowing to observe the regeneration potential.

Fifteen days after germination, 2 to 3 g of young leaf tissue was collected from at least 10 plants from each accession and frozen in liquid nitrogen. DNA was extracted following the phenol-chloroform extraction method [7] .

This work was accomplished at the Indian Grassland and Fodder Research Institute, Jhansi, India. A library enrichment protocol [8] based on the PCR amplification of genomic DNA with 5’ anchored degenerate microsatellite primers KKVRVRV(GA)10, KKVRVRV(GGT)5, KKVRVRV(CA)10, KKVRVRV(AAT)6, KKVRVRV(GTG)6, KKVRVRV(GACA)5, and KKVRVRV(CAA)6 (where K = G/T, V = G/C/A, R = G/A), was used to amplify T. alexandrinum (cv. Wardan) genomic DNA following [9] . Positive recombinant clones were used for plasmid DNA isolation following [7] . DNA sequencing (ABI 3700) of the clones was done using the Big Dye Terminator reaction kit (Applied Biosystems, USA).

Primers flanking SSR motifs were designed using the PRIMER 3 software [10] having 18 - 24 nucleotides, Tm = 50˚C - 62˚C with an optimal of 56˚C, 100 - 400 bp PCR product, and an optimal 40% G + C content. Primers were designated as IGFRI-SSR 1 to 15 and details are provided in Table 2. These primer-pairs, in addition to those developed previously from T. repens [11] were used for genetic diversity study. SSR primer sequences were synthesized from Sigma-Aldrich (Australia); the forward primers were tailed with M13 sequence (19 bp long) and labeled with fluorescent dyes following [12] . PCR amplifications were performed under the conditions described previously ( [11] . SSR polymorphism was analyzed using Beckman Coulter CEQ^TM 8000 Genetic Analysis System according to [12] . Fragment analysis was done using three fluorescent labeled dyes i.e. D2, D3 and D4 with size standard 400 and SSR alleles scored on CEQ capillary electrophoresis.

Fragments amplified with SSR primer-pairs were scored into binary format (“1” for presence and “0” for absence). Genetic similarity, based on allelic data, was

^aProstrate and slow growing; ^bMorphologically close to T. salmoneum; ^cNo germination; Gp: Germplasm; cv: Cultivar.

estimated following [13] . Dendogram was prepared following SAHN clustering based on unweighted pair-group method with arithmetic average (UPGMA) method using the NTSYS PC software (http://www.exetersoftware.com/cat/ntsyspc/ntsyspc.html). The clusters in the dendogram were identified by drawing phenetic line at 0.5 similarity coefficient and considering the dedogram topologies, i.e. groups joining the other major group coinciding with the cutting points at root level. Polymorphic information content (PIC) was calculated as per [14] .

Amplicons generated through the five degenerate primers having (GA)10, (CT)10, (GTG)6, (GACA)5 and (CAA)6 repeats, were cloned and 89 colonies obtained. Of the 59 positive clones, 46 had inserts containing one or more SSR repeat motifs, suggesting that our approach for library construction was effective in isolation of SSR. Redundant clones visualized after BLAST analysis were removed. Although sequence analysis revealed all recombinants to possess terminal microsatellite repeats, to avoid any base pair degeneracy in degenerate primers, synthesis of forward and reverse primers was done; which helped in getting better cross-species reaction. The SSR motifs comprised of mononucleotide (T), dinucleotide (CA), tetranucleotides and compound repeats (Table 2). In all, 15 SSR primer pairs (IGFRI-SSR1 to 15), including 5 SSRs having compound repeats, were designed (Table 2). Degenerate primers are largely designed based on the common repeat sequences observed in plant genome and have been reported

*N - A/C/G/T, bp = base pair.

working with different legume and grass species [9] . Thus, the results demonstrate the utility of degenerate 5’ anchored primers as simple, fast and cost effective as reported earlier also by [9] .

Of the 15 IGFRI-SSR primer pairs (Table 2), derived from T. alexandrinum in this study, seven amplified the fragments of expected sizes in Egyptian clover. We further analyzed 7 polymorphic SSR markers to reveal allelic diversity in 69 accessions of the different Trifolium species. A total of 20 alleles were detected by the seven IGFRI-SSR markers (Table 3). The size of SSR alleles varied from

bp: base pair.

133 to 262 bp among T. alexandrinum accessions as well as across the different species. The number of alleles/SSR ranged from two to seven, with an average of 2.85 allele/SSR locus. The number of alleles/SSR among T. alexandrinum accessions ranged from one to seven with 2.71 alleles per SSR locus. PIC values observed with these primers ranged from 0.99 to 0.68. Primer IGFRI-SSR 8 was species specific and polymorphic with T. alexandrinum. IGFRI-SSR 14 amplified fragments with T. alexandrinum and the two off type (prostrate and slow growing) T. alexandrinum accessions 15,892 and 33,747 (Figure 1(E) & Figure 1(H)). The remaining five IGFRI-SSRs amplified fragments with other species also. Alleles of IGFRI-SSR 11, 13 and 15 were represented among most of the accessions under study, thus, indicating common ancestral contribution of the alleles. T. michelianum was found to possess the most common fragments of IGFRI-SSR 11 only (Table 3), hence, showed high degree of dissimilarity with other accessions. These results suggest that SSR markers, derived from T. alexandrinum, are suitable to effectively discriminate different Trifolium species.

To validate the usefulness of T. repens primers and to assess molecular diversity among Trifolium species, we analyzed second set of 15 SSR markers developed from T. repens [11] . Thirteen out of these 15 primers amplified fragments with T. alexandrinum; all 15 amplified with arrow clover; 10 with T. salmoneum, 9 with T. subterraneum and 13 with T. michelianum. A total of 109 alleles were scored (Table 3) which was attributed to the diverse genetic background of the accessions under the present study. More than 0.9 PIC values were noticed for 8 out of 15 primer pairs and the minimum PIC value of 0.82 was noticed with primer A01C10. Most of these primers amplified fragments with other species also. A02H09 and A02H03 primers did not amplify the DNA from T. alexandrinum, T. salmoneum, T. subterraneum and the morphologically T. salmoneum look-alike accessions (Table 1), thus suggesting a genetic relatedness among them. A few primers produced 6 to 10 alleles among T. alexandrinum accessions (Table 3) which was attributed to diverse genetic base of the T. alexandrinum accessions and that the accessions from different place of collection might be having allelic contribution from different Trifolium species. It is also possible that the primers could bind at many locations due to pleiotropism or non-specificity of the primers. Molecular studies on two ecotypes of Berseem as done by [15] also found that QTLs for different traits were mapped in the same region on map, thus, indicating common genetic control and possibility of pleiotropism and/or tight linkage of different polygenes or QTLs. The extent of genetic diversity in the present set of germplasm was larger than that assessed by morphological and isozyme studies among Indian collection of T. alexandrinum germplasm [7] [16] . This can be attributed to factors such as the diverse source of the germplasm, the efficacy of SSR primers and possible different pathway of origin of these germplasm lines.

In all, the two sets of primers proved to be effective in revealing the intra and the interspecies diversity. Additionally, high polymorphism exhibited among T. alexandrinum accessions indicated suitability of these SSRs for further allelic study of the species.

In order to investigate the genetic relationship among accessions of T. alexandrinum and related species, we generated dendogram using clustering method. Twelve distinct clusters were formed containing a varying number of accessions (Figure 2). T. alexandrinum accession No 594 formed independent cluster “A” (Figure 2), although this accession was morphologically similar to other Egyptian clover accession. Cluster “B”, the largest cluster was divided in two sub-clusters “B1” and “B2” containing 33 and 17 accessions respectively. The majority of T. alexandrinum accessions together with T. salmoneum, T. subterraneum and the three T. salmoneum type accessions (75,011, 77,743, 77,740) grouped in this cluster. Cluster “C”, “D” and “E” contained three, two and three T. alexandrinum accessions respectively. Although, T. alexandrinum accessions represented different geographical places, no trend of clustering based on place

of collection was established. This might be because of movement of germplasm and its intermixing with native germplasm. Two prostrate and slow growing T. alexandrinum accessions 32,668 and 15,892 formed independent clusters “F” and “G” respectively. These two accessions were having no morphological similarity with T. alexandrinum. T. alexandrinum accessions, possessing morphological similarity with T. salmoneum, 68,947 and 75,014 formed cluster “H”; accession 36,369 formed independent cluster “I” whereas accession 33,747 formed cluster “J” with an T. alexandrinum cv Elite II. Cluster “F”, “G” and “H” grouped together before joining the major cluster. Similarly, cluster “J” and “I” also grouped together before joining the major cluster. These T. alexandrinum accessions which differed in morphology were placed closely in the dendogram. Cluster “K” possessed the two species T. vesiculosum and T. repens whereas cluster “L” was represented with T. michelianum only.

The genus Trifolium has been divided in eight sections [1] . In the present study, clustering of T. subterraneum belonging to section Trichocephalum of the genus, with T. alexandrinum and T. salmoneum of section Trifolium, indicated that the three species possessed relatedness which was congruent with earlier reports [3] [16] [17] [18] [19] . In fact, Lotoidea is the largest section and has served as source taxa for the evolution of many other sections [1] . This is also supported with development of fertile intersection interspecific crosses [10] [20] [21] . Thus, this study reaffirms the high relatedness of the species as well as possibility of polyphyletic origin of T. alexandrinum.

Origin and ancestry of T. alexandrinum has remained controversial. T. berytheum [22] [23] , T. echinatum [24] , T. apertum [18] and T. salmoneum [23] were considered as its progenitor. In the present study, clustering of T. salmoneum and the T. salmoneum look-a-like accessions with T. alexandrinum accessions also indicated close affinity between the two species and that T. salmoneum might have contributed in evolution of T. alexandrinum. Clustering of T. alexandrinum Elite II, an Australian cultivar, with two T. salmoneum type accessions also suggested its ancestry with the former. In fact, the accessions mentioned as “T. salmoneum type” were phenotypically closer to T. salmoneum and possessed round cotyledonary leaves similar to T. salmoneum. Thus, the molecular data based similarity was congruent with phenotype of the plants.

Boundaries among different Trifolium species are extremely difficult to define because of the range of diversity caused by primary polymorphism [25] . Hence, identification of species is an important aspect in Trifolium genetic resource study. Occurrence of a few plants, either with serrated margin first leaf or bifoliate/trifoliate first leaf among accessions 668, 45,313, 669, 35,688, 73,308 and 77,740 of T. alexandrinum, were considered as admixtures which occurs as samples are collected from the field. However, the diversity as result of admixture can also be conserved because the basic objective of any gene bank is to conserve genetic diversity. Occasional occurrence of plants with a few multifoliate leaves in accessions 673, 6,3256 and 33,875 was in line with such reports in many Trifolium species. The genetics of the trait is not known, although pentafoliate trait was recently fixed in T. alexandrinum [26] .

Thus, this study developed a set of primer-pairs which have shown to amplify alleles from T. alexandrinum, T. salmoneum, T. vesiculosum, T. michelianum, T. repens and T. subterraneum, suggesting that this newly developed resource is useful for assessing molecular diversity among accessions of at least six Trifolium species. The seven SSR markers from T. alexandrinum genomic resource and 15 from T. repens genomic resource were able to distinguish different accessions of T. alexandrinum and the species under study. Allelic dissimilarity among morphologically similar accessions (accession 594, a typical T. alexandrinum, forming independent cluster) and clustering together of morphologically distinct accession (Elite II and accession Nos. 33,747 and 36,369) shows the efficiency of SSR primers in the study. High PIC value of SSR proves that these markers were suitable to differentiate among accessions. Cluster analysis indicated suitability of SSR markers for genome analysis. The study could successfully establish the larger diversity of the T. alexandrinum gene pool as superior resource for further breeding and enriching the SSR markers repertoire for further genetic study. The study also established the relatedness of T. subterraneum and T. salmoneum accession with T. alexandrinum, indicating their role in the evolution of T. alexandrinum.