Plant material.

Seeds from two accessions of S. condensatum, two accessions of S. sanguineum, and 75 accessions of S. scoparium were obtained from the USDA-ARS NPGS (Table 1) (USDA, 2013). One seed per accession randomly selected, except for four accessions, was potted in commercial potting medium (Metro-Mix 360; Sun Gro Horticulture, Agawam, MA). The seeded cultivars PI 421552 [Cimmarron (10 individual plants)], PI 421553 [Aldous (10 individual plants)], PI 642388 [Prairie View (nine individual plants)], and PI 648369 [Pastura (nine individual plants)] were obtained from the USDA-ARS NPGS and used to examine diversity within plants of the same accession.

Table 1.

Schizachyrium accessions and cultivars that were examined for ploidy level and genetic variation.

View Table

Library construction, ion torrent sequencing, and microsatellite discovery.

DNA was extracted from the little bluestem cultivar The Blues using a Power Plant Pro DNA Isolation kit (MO Bio Laboratories, Carlsbad, CA). ‘The Blues’ was provided by M. Harrison. A genomic DNA library (400 bp read) was developed for ‘The Blues’ by following the manufacturer’s instructions using the Ion Xpress Plus gDNA Fragment Library Preparation kit (Life Technologies, Carlsbad, CA). After fragment preparation and prior sequencing, the library was quantified using the Ion Library Quantitation Kit (Life Technologies) to determine the appropriate library dilution. After library dilution, the library was enriched with Ion Sphere Particles using the Ion personal genome machine (PGM) Template OT2 400 Kit (Life Technologies) in preparation for sequencing. The library was sequenced on the Ion PGM System (Life Technologies) using the Ion PGM Sequencing 400 Kit (Life Technologies) and Ion 318 Chip v2 (Life Technologies). The sequencing reads were assembled into contigs using the default parameters in SeqMan NGen software (DNASTAR, Madison, WI). Contigs were exported in FASTA format for microsatellite discovery. The sequences were searched for microsatellites (di- to hexa-nucleotide motifs) and primers were designed using BatchPrimer3 (You et al., 2008). Primer design used the default settings except that minimum, optimal, and maximum primer melting temperatures were set to 55.0, 60.0, and 65.0 °C, respectively. The software program Censor was used to identify those contigs with repetitive elements (Kohany et al., 2006).

DNA extraction and PCR amplification of SSRs from the germplasm collection.

DNA was extracted from 78 Schizachyrium accessions or cultivars (Table 1) using a Power Plant Pro DNA Isolation kit (MO Bio). Concentration was determined on all samples using a spectrophotometer (NanoDrop 2000c; Thermo Fisher Scientific, Waltham, MA) and template quality was determined for a random subset of samples using a 1% agarose gel. Each primer was initially tested using DNA from the cultivar The Blues at 50 °C. Gradient polymerase chain reaction (PCR) (45 to 60 °C) was used for those SSR markers that did not amplify well at 50 °C. For the diversity study, DNA samples were aliquoted into strip tubes to form an array and the sample ‘The Blues’ was included twice on the array to ensure PCR amplification reliability. Each SSR marker was amplified from each template using a 10-μL reaction volume that contained 2 μL of 5× Clear GoTaq reaction buffer (Promega, Madison, WI), 1 μL of 25 mm MgCl₂, 0.8 μL of 2.5 mm dNTP mix, 0.5 μL of M13-tagged forward primer, 2.0 μL of reverse primer at 1 μM, 1.8 μL of 1 μM M13 primer (M13-TGTAAAACGACGGCCAGT) fluorescently labeled with the IRDye 800 CW fluorophore (Eurofins MWG Operon, Huntsville, AL), 0.04 μL of GoTaq DNA polymerase (Promega), 0.86 μL of sterile water, and 1 μL of 2.5 ng·μL⁻¹ DNA. Thermocycler conditions were: an initial denaturation at 94 °C for 3 min, 39 cycles of 94 °C for 30 s, 45 to 60 °C (Supplemental Table 1) for 1 min, 72 °C for 1 min 10 s, and a final elongation step at 72 °C for 10 min. The thermocycler used was a Gene Amp PCR System 9700 dual block (Applied Biosystems, Foster City, CA). The individual PCR products (2 μL) were combined with 5 μL of Blue Stop (LI-COR Biosciences, Lincoln, NE) and 0.35 μL of this mixture was loaded on a 6.5% v/v acrylamide gel using a 4300 DNA Analyzer (LI-COR Biosciences). Gel images were scored visually and coded as a “1” for presence of a band or “0” for absence of a band for each accession for each marker. The polymorphism information content (PIC) value and expected heterozygosity were calculated using an online PIC calculator (University of Liverpool, 2002).

Generation of SSR markers from Genbank sequences.

The National Center for Biotechnology (NCBI) nucleotide database was searched using the keywords “Schizachyrium,” “Andropogon,” and “Hyparrhenia” and 52, 67, and 40 nucleotide sequences were identified, respectively. Using the program SSRIT (Temnykh et al., 2001), 21, 17, and 11 sequences contained SSR repeats with a motif a pentameter in length or less with at least four repeats in Schizachyrium, Andropogon, and Hyparrhenia sequences. Assembly of these 49 sequences using Sequencher 4.10.1 (Gene Codes Corp., Ann Arbor, MI) generated eight singletons and 10 contigs. A single sequence was selected from eight contigs and eight singletons and primers were designed using Primer3 (Rozen and Skaletsky, 2002) to amplify the repeat.

Cloning of PCR amplicons.

Amplicons from three primer pairs (developed from the ion torrent sequencing), 148812, 149427, and 15116, that generated more than four fragments per accession were selected for sequencing. PCR reactions contained 12.4 μL of water, 4 μL of 5× Phusion HF buffer, 0.4 μL of 10 mm dNTPs, 1 μL of 10 μM forward primer, 1 μL of 10 μM reverse primer, 0.2 μL of DNA polymerase (Phusion; Thermo Fisher Scientific), and 1 μL of 100 ng·μL⁻¹ ‘The Blues’ DNA. PCR reactions were carried out in a thermocycler (PTC-100; MJ Research, Waltham, MA). After 30 s of denaturation at 98 °C, 30 cycles were performed with 10 s at 98 °C, 30 s at 50 °C, 1 min at 72 °C, and a final extension step of 10 min at 72 °C. One microliter of unpurified PCR products was ligated into pJET1.2/blunt Cloning Vector following the manufacturer’s instructions (Thermo Fisher Scientific). The ligation mixture was used to transform chemically competent Escherichia coli, Turbocells (Genlantis, San Diego, CA). Transformed E. coli cells were plated onto Luria Broth plate supplemented with 100 μg·mL⁻¹ of ampicillin. To confirm the presence of an insert in the resulting transformants, colony PCR was carried out using the pJET1.2 forward and reverse sequencing primers. Plasmids, containing inserts, were isolated from overnight E. coli cultures grown in Luria Broth with 100 μg·mL⁻¹ of ampicillin using a PureYield Plasmid Miniprep System (Promega). Seven independent clones generated with each primer pair were sequenced by Eurofins Genomics (Huntsville, AL).

Cluster analysis.

For the unweighted pair group method with arithmetic mean (UPGMA) cluster analysis, marker data were imported into NTSYSpc (Rohlf, 2008). Genetic similarity between each pair of accessions was calculated using the SIMQUAL module using the DICE coefficient of similarity (Nei and Li, 1979). A dendrogram was generated from the similarity matrix by using the UPGMA procedure in the SAHN module of NTSYSpc. The software program FreeTree (Hampl et al., 2001) was used to do the bootstrapping analysis and the number of repetitions was 1000. Only bootstrap values greater than 50% are shown (Figs. 1 and 2). For the Bayesian cluster analysis, groupings were analyzed using Structure software Version 2.3 (Pritchard et al., 2000) using a burn-in period of 10,000 and 50,000 Markov Chain Monte Carlo replications for all Schizachyrium samples and a burn-in period of 50,000 and 50,000 Markov Chain Monte Carlo replications for S. scoparium samples. The parameters were set to an admixture model with allele frequencies correlated. To determine the best K (number of populations), the L(K) procedure (Pritchard et al., 2000) was used by performing a batch job from K = 1 to K = 10 and the delta K procedure (Evanno et al., 2005) was used by performing a batch job from K = 1 to K = 10 with 20 iterations for all Schizachyrium samples and 11 iterations for the S. scoparium samples. The program Structure Harvester was used to calculate ΔK (Earl and vonHoldt, 2012).

View Full Size

Unweighted pair group method with arithmetic mean cluster analysis for all Schizachyrium samples.

Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 140, 1; 10.21273/JASHS.140.1.78

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Unweighted pair group method with arithmetic mean cluster analysis for Schizachyrium scoparium samples.

Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 140, 1; 10.21273/JASHS.140.1.78

• Download Figure
• Download figure as PowerPoint slide

Flow cytometry.

Schizachyrium samples were processed for analysis on a BD Accuri C6 flow cytometer (Becton Dickinson, Franklin Lakes, NJ) by first adding 1 mL of Tris-MgCl₂ nuclei extraction buffer (Pfosser et al., 1995) to 0.02 to 0.03 g of fresh leaf tissue using a modified protocol (Galbraith, 2009). An assumed tetraploid sample (2n = 4x = 40), accession PI 213875, was cochopped with each sample. Tissue was chopped using an Astra double-edged razor blade (Procter & Gamble Hungary, Budapest, Hungary) until fully macerated. The resulting slurry was then pipetted into a 50-μm CellTrics disposable filter (Partec, Munster, Germany) sitting on top of a plastic 12 × 75-mm culture tube. The filter was removed and 500 μL of Rnase/propidium iodide solution (BD Biosciences, San Jose, CA) was added to each filtered sample. Samples were then placed on ice for 15 min and then analyzed on the flow cytometer by gating the selected objects that exhibited a strong correlation between the FL2 and FL3 signals. At least 50,000 events were collected for each sample, the flow rate was set to custom (11 μL sample per minute), and the mean FL2 was recorded for the Gap 1 (G1) and G2 peaks. The flow cytometry procedure was repeated for samples producing wide peaks, where the cv of the mean FL2 data was greater than 5% or if samples showed a different ploidy than 4×. Variation in ploidy level was determined based on the G1, G2 sample peaks relative to the control G1, G2 peaks.

Genome sequencing.

Ion torrent PGM sequencing of ‘The Blues’ with a 318 chip generated a total of 681.37 Mb data and 2,654,634 quality reads in a single run. Sequences were deposited into the NCBI database under the BioProject accession PRJNA265209. The length of the reads ranged from 50 to 450 bp with an average of 256 bp. All reads were assembled into 8713 contigs with a mean length of 2063 bp. A total of 2813 SSRs was identified from the contigs and 32.3% of the contigs contained SSRs. Sequences were considered to contain SSRs if they contained a motif 2 to 6 bp in length and the SSR length was at least 12 bp. The most frequent types of SSR were trinucleotide motifs followed by tetranucleotide motifs (Table 2).

Table 2.

Summary of perfect microsatellite motif classes identified in contigs from ion torrent personal genome machine sequencing of a genomic DNA library of Schizachyrium scoparium ‘The Blues’.

View Table

Of the 8713 contigs developed, the first 337 contigs were examined to determine the frequency of repetitive elements. The majority of sequences [297 of 337 (88%)] had homology to repetitive elements. Long terminal repeat (LTR) retrotransposons, DNA transposons, non-LTR retrotransposons, retroviruses, and integrated viruses were identified. The most common repetitive elements were Gypsy LTR retrotransposons (338 elements), EnSpm/CACTA DNA transposons (119 elements), and Copia LTR retrotransposons (104 elements). The sequencing of grass genomes has revealed that the transposable element content in these genomes can be quite high, especially as the size of the genome increases [rice (Oryza sativa, 35% transposable elements), sorghum (Sorghum bicolor, 62%), maize (Zea mays, 85%) (Hua-Van et al., 2011)]. Low coverage sequencing using 454 technology in Aegilops cylindrica and A. geniculata [wild relatives of wheat (Triticum aestivum)] identified that more than 70% of the reads obtained were transposable elements, primarily LTR retrotransposons (Senerchia et al., 2013). Little bluestem has a genome size of ≈4303.2 Mbp (assuming 1 pg = 978 Mbp) (Bai et al., 2012), which is larger than maize at 2900 Mbp (Hua-Van et al., 2011) but smaller than wheat at 15,000 Mbp (Paterson et al., 2005). Its genome, from our low-coverage sequencing, appears rich with repetitive elements.

Use of existing resources.

There are currently 52 nucleotide sequences available in GenBank (Benson et al., 2005) for the genus Schizachyrium generated from phylogeny studies of the tribe Andropogoneae (Mathews et al., 2002; Teerawatananon et al., 2011). Of these 52 sequences, 21 contain SSR repeats and assembly of these sequences generated one contig and nine singletons. To increase the number of SSR markers from multiple loci, all of the nucleotide sequences from GenBank of the related genera Andropogon (67) and Hyparrhenia (40) (Mathews et al., 2002) were downloaded, screened for SSRs, and those sequences containing SSRs were assembled forming 10 contigs and eight singletons. Primers were developed from the contig and singleton sequences and 11 were of sufficient quality to be scored for all the Schizachyrium accessions (Supplemental Table 1; these sequences are represented by two letters followed by six numbers, e.g., JQ951654). Five of the 11 SSR markers from GenBank sequences were polymorphic among S. scoparium accessions.

Amplification of Schizachyrium accessions.

The number of SSR primer sets designed was 2812 (Supplemental Table 1). For many plant species, a higher degree of polymorphism has been seen in Class I microsatellites (20 bp or greater) than in Class II microsatellites (12 to 19 bp) (Pandey et al., 2013; Temnykh et al., 2001). Thus, 95 SSR primer pairs and five SSR primer pairs were selected with a repeat length of 20 bp or greater (Class I) or 18 bp (Class 2), respectively (Supplemental Table 1). Amplification of these primers with DNA from ‘The Blues’ (to determine annealing temperature) and from the Schizachyrium collection revealed that 82 primer pairs were of sufficient quality to be scored. The SSR markers that were scored (82 obtained from ion torrent sequencing and 10 from Genbank) contained 24 dimer repeats, 36 trimer repeats, three tetramer repeats, nine pentamer repeats, and 20 hexamer repeats (Supplemental Table 1). The number of scored fragments per primer pair ranged from 0 (monomorphic, single product of the same size) to 12, which was surprisingly low for a primarily outcrossing tetraploid species and 183 fragments were scored. The average number of scored fragments was 1.98. Furthermore, only 35 primer pairs generated polymorphism of sufficient quality (clear distinct fragments) to be scored among S. scoparium accessions (Supplemental Table 1). These 35 primer pairs generated 102 scored fragments. Class I microsatellites [26 polymorphic/78 that amplified (33%)] were less polymorphic than Class II microsatellites [six of six (100%)] for the S. scoparium accessions, although the number of Class II microsatellites used was small. Twenty-two primer pairs generated more than four fragments per accession and eight of these generated more than 10 fragments per accession. The generation of so many fragments, many at a high annealing temperature, indicates the primers are specifically amplifying multiple regions of the genome.

Three of the primer pairs (149427, 151116, and 148812) that generated more than four fragments per little bluestem accession were selected and fragments were amplified using ‘The Blues’ template. Seven clones from each primer pair were sequenced and assembled forming four contigs and two singletons. One sequence from each contig and each singleton sequence was used to search for repetitive elements using the software Censor (Kohany et al., 2006). Primer pair 148812 generated approximately six fragments per accession and all seven clones had no similarity to repetitive elements and formed a single contig. Primer pairs 149427 and 151116 generated more than 10 fragments per little bluestem accession and clones from both primer pairs generated three contigs (two of the three contigs contained sequences from both primer pairs) and two singletons. Four of the five sequences from these two primer pairs had similarity (79% to 88%) to an En/Spm-type DNA transposon from sorghum.

Examination of all little bluestem SSR sequences for repetitive elements.

Of the 117 sequences containing SSRs that were used to design primers, only nine (8%) did not have similarity to DNA transposons, retrotransposons, viruses, or satellite sequences when analyzed with Censor software. The most frequent fragments identified were the LTR retrotransposons Gypsy (177 fragments) and Copia (98 fragments) and the DNA transposon EnSpm (60 fragments). Most sequences contained homology to more than one superfamily of repetitive elements (67 of 117). The association between SSR repeats and transposons has been documented in many studies including plants. In rice, an association between (AT)n repeats and the Micropon family of MITEs exists (Temnykh et al., 2001). The authors noted that primers designed to amplify these repeats often generated multiple fragments. In barley (Hordeum vulgare), dinucleotide SSR repeats are associated with retrotransposons and other repetitive elements (Ramsay et al., 1999). In wheat, many of the SSR clones had similarities to LTR retrotransposons causing irregular amplification profiles such as smear products (Koike et al., 2006). In pea (Pisum sativum), variable SSR repeats were identified within the LTR of a pea Ty1/copia retrotransposon and the LTR retrotransposons were shown to contribute to the origin and distribution of microsatellite repeats (Smykal et al., 2009). Thus, the little bluestem sequences that contained SSR repeats, that we examined, are rich with repetitive elements (92%) and are much higher than those of previous reports in plants, which found between 24% and 54% of the microsatellites to be associated with other repetitive elements (Ramsay et al., 1999; Temnykh et al., 2001; Tero et al., 2006).

Genetic relationships among Schizachyrium samples.

Cluster analysis, using Structure and NTSYS, of the SSR data for all the Schizachyrium samples revealed two major groups. The S. condensatum samples formed one group (Wright’s measure of population subdivision, Fst = 0.7543) and the S. sanguimeum and S. scoparium samples formed another group (Fst = 0.5816), although there is clear separation between samples of S. sanguineum and S. scoparium (Fig. 1). Admixture was measured as 0.0250, which is close to zero, suggesting that individuals are from one population or another and no interbreeding is occurring (Evanno et al., 2005). Cluster analysis using the UPGMA procedure for the SSR data for only the S. scoparium samples revealed five groups but these groups were not strongly supported (bootstrapping less than 50%) (Fig. 2). Group I consisted of 13 accessions (the sample ‘The Blues’ was repeated twice to confirm scoring accuracy) primarily collected from Texas, Group II consisted of 58 accessions including all of the individual samples of the forage and conservation cultivars as well as the clonal samples ‘Carousel’ and ‘Blue Heaven’. Samples within this group were collected primarily from locations with latitudes between 34 and 43° N. Group III and V consisted each of a sample, PI216760 and PI216770, respectively, collected in 1953 by the Agricultural Experiment Station, College Station, TX. Group IV consisted of a single sample PI 666193, a coastal little bluestem (S. scoparium var. littoralis), of an unknown collection location. Coastal little bluestem grows on the coastal strip from southern New England to Texas and can be distinguished from inland little bluestem by its bent stems at the base of the plant (USDA, 2014).

Because the S. scoparium groupings were not well supported using the UPGMA procedure, cluster analysis of the S. scoparium SSR data was performed using the software program Structure Version 2.3, which uses a Bayesian algorithm. Four clusters using the ΔK method (Evanno et al., 2005) were identified. Structure Group 2 corresponded to Group I of the UPGMA grouping but with an additional member #14, which formed Group III in the UPGMA grouping (Fig. 3). The remaining three Structure groups correspond to the UPGMA Group II and the merging of the UPGMA groups with one member to these groups. UPGMA Groups III and V that merged with the larger groups in the Structure clustering were not assigned to these Structure groups at a high probability (membership coefficient = 0.56 and 0.46, respectively). Structure Group 1 consisted of only the ‘Cimarron’ accessions. Structure Group 3 corresponded to all the ‘Aldous’ accessions, the ‘Blaze’ accessions, three ‘Cimarron’ accessions, ‘Carousel’, and accessions from Kansas, Texas, and Wyoming. Structure Group 4 consisted of the ‘Prairie View’ and ‘Pastura’ accessions as well as ‘Blue Heaven’ and accessions from Ohio and New Mexico. Admixture was measured as 0.0577 indicating that little interbreeding between groups is occurring.

View Full Size

Bayesian cluster analysis using all Schizachyrium scoparium samples.

Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 140, 1; 10.21273/JASHS.140.1.78

• Download Figure
• Download figure as PowerPoint slide

Unequal sampling sizes can affect phylogenetic analysis (Nelson and Anderson, 2013). To determine if having an unequal number of individuals (four accessions had nine or 10 plants per accession, whereas the rest had only one plant per accession) affected the number of clusters, only one plant per accession was analyzed for the S. scoparium samples. Structure analysis revealed three groups. As to be expected, the group that consisted only of the ‘Cimarron’ accessions was no longer present in the new analysis. Structure Group A (Fig. 4) consisted of the cultivars Badlands, Pastura, Prairie View, Blaze, Carousel, Southlow Michigan, and Blue Heaven as well as accessions from Kansas, Ohio, Wyoming, Texas, and New Mexico. Structure Group B (Fig. 4) consisted of the cultivars Cimarron and Aldous, accessions from Kansas and Texas, and the coastal little bluestem accession. Structure Group C (Fig. 4) consisted of the samples that belonged to Group I of the UPGMA analysis. These samples included the cultivar The Blues and samples predominantly from Texas.

View Full Size

Bayesian cluster analysis using one plant per accession for all Schizachyrium scoparium samples.

Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 140, 1; 10.21273/JASHS.140.1.78

• Download Figure
• Download figure as PowerPoint slide

Future breeding for little bluestem should be done across these cluster groups to determine if they could serve as heterotic groups. Gabriel (2005) found that crossing sorghum lines within groups (groups were defined with molecular markers using UPGMA clustering) exhibited inferior heterotic expression for grain yield and other traits as compared with across group crosses. Similar results were found in tropical maize (Pinto et al., 2003).

Among the S. scoparium accessions, the lowest genetic similarity between accessions was 86% (Fig. 2). This high level of genetic similarity was a surprise given the high degree of morphological variation seen in this species and previous observations of genetic variation within populations (Fu et al., 2004; Huff et al., 1998). Previous research suggests that genetic diversity correlating with geographic traits among populations is minor compared with the genetic variation within populations (Fu et al., 2004). The high genetic similarity values among accessions could indicate a narrow genetic base. However, the SSR markers that we designed were from sequences that were rich with repetitive elements and may have increased the genetic similarity values, yet for those nine sequences in which the Censor software did not detect other repetitive elements, only seven amplified and three of seven (42%) markers were polymorphic among S. scoparium accessions. This rate of polymorphism is not much higher than for those sequences containing other repetitive elements and that were polymorphic among S. scoparium accessions, 32 of 86 (37%). Furthermore, examination of genetic diversity in natural little bluestem populations in Manitoba and Saskatchewan found the maximum interpopulation distance was 0.13 (87% similar) (Fu et al., 2004). Huff et al. (1998) observed a low level of variation among four little bluestem populations (5%) in comparison with other outcrossing species. Thus, the low level of interpopulation differentiation may be caused by a high level of genetic variability (Gonzalez-Hernandez et al., 2009), which is inconsistent with our SSR results.

Ecoclinal variation for traits such as maturity, height, leafiness, and disease resistance exists in little bluestem (Anderson and Aldous, 1938; Cornelius, 1947). Phan and Smith (2000) identified differences among ecotypes for culms per plant and seedheads per culm in 3-year-old spaced plants of collected little bluestem from Manitoba. Boe and Bortnem (2009) found strong phenotypic plasticity for biomass production and that genetic variation for phenotypic plasticity exists in the cultivar Camper. Furthermore, Springer (2012) used 19 of the little bluestem accessions this study genotyped and found variation for morphological traits such as canopy morphology, height and width of canopy, and leafiness of the culm. Thus, the maximum genetic distance of 14% found between S. scoparium accessions using SSR markers and phenotypic plasticity allows a wide range of morphological phenotypes.

Within-cultivar variability.

Examination of within-accession variability of four forage cultivars, Cimarron, Aldous, Prairie View, and Pastura, showed that all plants from these accessions clustered in UPGMA Group II (Fig. 2), but Structure analysis provided better separation (Fig. 3). From the Structure analysis, accessions from the cultivar Aldous clustered in Group 3 and accessions from the cultivars Prairie View and Pastura clustered in Group 4. Furthermore, from the Structure analysis, seven of the accessions from cultivar Cimarron clustered in Group 1 and three of Cimarron accessions clustered in Group 3 (Fig. 3). This suggests that greater diversity exists within the cultivar Cimarron than within the cultivars Aldous, Prairie View, and Pastura.

Ploidy variation.

Ploidy variation within Schizachyrium has been seen from 2n = 2x = 20 to 2n = 10x = 100 (Peichoto et al., 2011). Schizachyrium scoparium has been found to be 2n = 4x = 40 (Carman and Hatch, 1982); S. condensatum has been found to be 2n = 2x = 20 and 2n = 4x = 40 (Peichoto et al., 2011); and S. sanguineum has been found to be 2n = 2x = 20, 2n = 3x = 30, 2n = 4x = 40, 2n = 5x = 50, 2n = 6x = 60, 2n = 7x = 70, 2n = 8x = 80, and 2n = 10x = 100 (Peichoto et al., 2011). Ploidy variation was determined by flow cytometry on the 75 S. scoparium accessions obtained from the USDA and all plants had a ploidy level the same as the assumed tetraploid (2n = 4x = 40). Two accessions each of S. condensatum and S. sanguimeum were used as outgroups. The accessions of S. condensatum had the diploid cytotype (PI 203720, PI 203834), whereas S. sanguimeum had a tetraploid (PI 404629) and octoploid (PI 216115) cytotype (assuming our internal control was a tetraploid).

In summary, the USDA-ARS S. scoparium collection was evaluated for ploidy and genetic variation. The next-generation sequencing data for little bluestem is the first data set generated for this important horticultural grass species. Most of the sequences, from which microsatellites were developed, contained other repetitive elements. A set of 35 microsatellite markers was identified that was able to differentiate the S. scoparium accessions. UPGMA and Bayesian cluster analysis was used to analyze data from 102 fragments generated from the USDA-ARS S. scoparium collection to identify five and four groups, respectively. These microsatellite markers are the first sequence-specific markers designed for little bluestem and can serve as a resource for cultivar identification, pedigree analysis, breeding, etc. The 92 markers developed in this study may be useful in other Schizachyrium species as well.