Keywords

Until the mid-1960s, the genetic diversity of populations was accessed by morphological traits as sizes, shapes, or color patterns. These kinds of markers contributed to broaden our knowledge about population genetics. However, there are innumerous limitations to morphological traits; for instance, genetic variation could be overestimated because of phenotypic plasticity (Freeland 2005). Later, the genetic diversity was accessed by the sizes, shapes, and numbers of chromosomes, which was used to reconstruct the evolutionary history of Drosophila pseudoobscura by Sturtevant and Dobzhansky (1936). Chromosomal variation was studied between species and populations, but there was no consistent relationship between morphological and chromosomal variation (Rowe et al 2004). Hence, the development of molecular markers revolutionized this scenario, and genetic variation could be accessed from polymorphic proteins or DNA sequences.

A very large amount of information about different species could be performed, which allowed quantifying genetic diversity, population subdivision, gene flow, effective population size, breeding structure, inbreeding depression, natural selection, and genetic drift. All these studies are predominantly intraspecific (within a particular species). Nevertheless, molecular markers as DNA sequences are used in both systematics and phylogenetic studies which focus on the species level of classification (Rowe et al 2004). All this genetic knowledge has become extremely important for the effective conservation of many species.

The studies on the genus Baccharis accompanied the development of tools for genetics investigation. Genetic studies of Baccharis started in 1945 when chromosome numbers of 28 families of angiosperms were described by Bowden (1945) and included Baccharis pingraea, B. genistelloides, B. genistifolia, B. halimifolia, and B. phyteumoides. This report contained polar views of meiotic stages and mitotic metaphases of different species. The same number of chromosomes was registered for all these species (n = 9 or 2n = 18). On the other hand, reports on the chromosome number for species in the genus have increased dramatically (see Table 3.1), mainly to better understand the systematics and phylogeny of the group. The chromosome number of B. dracunculifolia (n = 9) was first described in 1970 using botanical materials collected in the state of Minas Gerais, Brazil (Coleman 1970). It was suggested that the ancestral chromosome number of the family Asteraceae is n = 9 (Solbrig et al. 1969; Solbrig 1977; Nesom 2000; Mota et al. 2016). While the majority of Baccharis species has a basic chromosome number of x = 9, some exceptions were detected as in B. latifolia (Turner et al. 1967; Powell and King 1969; Spooner et al. 1995), B. glutinosa (Ariza-Espinar 1974), and B. salicifolia (Solbrig et al. 1969), all of them with n = 18 or in B. nitida where n = 25 (Powell and King 1969). There is a description of n = 10 for B. tricuneata (Turner et al. 1967), but it is a questionable count since authors pointed that it is possible that a supernumerary or “B” chromosome was mistaken as a bivalent. In fact, B chromosome and some fragments were described in species of the genus as in B. decussata (Turner et al. 1967; Powell and King 1969), B. flabellata (Wulff et al. 1996), B. prunifolia (Powell and Cuatrecasas 1970, 1975), B. punctulata (Rozenblum et al. 1985), B. serrifolia (Anderson et al. 1974), and B. thesioides (Spellenberg and Ward 1988). However, as the genus Baccharis appears to exhibit chromosomal stability (Solbrig et al. 1969; Solbrig 1977), molecular markers are needed to provide information about genetic diversity within and between populations.

Table 3.1 Gametic (n) and somatic (2n) chromosome numbers in Baccharis
Full size table

The majority of the information on the molecular biology of the Baccharis genus was generated to clarify the taxonomic identity of this taxon. Despite being the fourth largest genus of the family Asteraceae and the most specious within the tribe Astereae (Heiden 2014), the evolutionary relationship of the genus is still in discussion. Zanowiak (1991) studied the systematic and phyletic relationships within the subtribe Baccharidinae. This chloroplast DNA study suggests that South American Conyza spp. should be included in the subtribe Baccharidinae, that the Baccharidinae consists of some species of Baccharis (published as Heterothalamus) and Archibaccharis clades, while another clade includes South American Exostigma notebellidiastrum (published as Conyza notebellidiastrum), Baccharis, and Baccharidastrum. In this same study, Zanowiak (1991) verified that B. neglecta and B. halimifolia hybridize, with B. neglecta being the maternal parent. Some novelties towards a phylogenetic infrageneric classification of Baccharis were published by Heiden and Pirani (2016) which includes names of new taxa, new combinations, and names at new rank for subgenera and sections of the genus. Later, Heiden et al. (2019), based mostly on phylogenetic grounds, proposed that Baccharis should comprise 440 species classified into 7 subgenera and 47 sections.

In the 2000s, Gomes et al. (2004), for the first time, conducted an intraspecific genetic study in this genus. These authors investigated the genetic variability in Baccharis concinna using randomly amplified polymorphic DNA (RAPD) markers. This species is a rare, dioecious, and threatened shrub, endemic to Serra do Espinhaço, Southeastern Brazil. The authors studied 335 individuals belonging to 6 populations along an altitudinal gradient. Despite the high genetic variability within populations of B. concinna, the populations studied were very similar, and genetic variability was not related to the altitudinal gradient. The authors argued that their findings could be explained by the B. concinna mating system. This shrub is pollinated and dispersed by wind, which may promote an intense gene flow among the studied species patches, independent of elevation. The authors also emphasized the absence of a physical barrier to gene flow by pollen and seed dispersal among the studied patches of individuals in the landscape.

The RAPD technique was developed in the 1990s (Welsh and Mcclelland 1990; Williams et al. 1990); it is quick and easily generated by PCR and requires no prior DNA sequence information. Although RAPD markers were commonly used for genetic diversity in plants (e.g., Wachira et al. 1995; Iqbal et al. 1997; Ram et al. 2008), they are subject to some limitations. Due to the dominance of the RAPD markers, it is not possible to distinguish between homozygotes (one copy of allele) and heterozygotes (two copies of allele) individuals. Furthermore, the RAPD markers do not allow the investigation of direct gene flow using paternity analysis. In addition, RAPD markers are of limited reproducibility because the segments of DNA are amplified by PCR using arbitrary primers that copy genome regions according to the annealing temperature of user selection. Nowadays, the use of other molecular markers is needed to investigate the genetic diversity of plant species with confidence.

Alternatively, the microsatellite markers, also known as simple sequence repeats (SSRs) and short tandem repeats (STR) (Jacob et al. 1991; Edwards et al. 1991), have been used in population and conservation genetics studies (Guichoux et al. 2011). The microsatellite markers are repeating motifs in tandem that are found at high frequency in most taxa genomes and exhibit high levels of polymorphism due to the high mutation rate that make them more informative than other molecular markers (i.e., single nucleotide polymorphism – SNP) (Bhargava and Fuentes 2010). The microsatellite markers are relatively uniformly distributed in the genomes of species, and due to their co-dominance, the distinction between homozygote and heterozygote individuals is possible. Traditionally, microsatellite development was slow, costly, and labor-intensive and required the construction of genomic libraries using recombinant DNA enriched for a few targeted SSR motifs. The repeating motifs can be mono-, di-, tri-, tetra-, penta-, and hexanucleotide repeats (Litt and Luty 1989; Zane et al. 2002).

Dinucleotide microsatellite repeats are commonly produced by the genomic library technique. However, dinucleotide repeats are prone to polymerase slippage during the PCR amplification (slipped-strand mispairing) and suffer from stutter bands (PCR products from the same fragment that are shorter by one or a few repeats) (Chambers and MacAvoy 2000). Unfortunately, this feature of dinucleotide repeats may lead to genotype scoring errors (Clarke et al. 2001; McDowell et al. 2002) making allele definition difficult (Levinson and Gutman 1987; Meldgaard and Morling 1997), especially for heterozygotes with adjacent alleles (Guichoux et al. 2011). Besides that, tri-, tetra-, and pentanucleotide loci perform better than dinucleotides because they are less prone to enzyme slippage (Edwards et al. 1991; Acharige et al. 2012).

The conservation of the sequence in the primer sites flanking the microsatellite loci and the stability of those sequences during evolution (Dayanandan et al. 1997; Ciampi et al. 2008; Feres et al. 2009) allows the use of SSR markers developed from one species to another. The transfer of polymorphic markers in plants is mainly successful within genera, and it has been successfully applied to the genetic analysis of tropical species (Zucchi et al. 2002; Cota et al. 2012; Moreira et al. 2012). Thus, in the last two decades, microsatellite markers have been used to assess gene flow at the population level and recent demographic events and aided in phylogenetic inferences (Braga et al. 2007; Ciampi et al. 2008; Moreira et al. 2008; Cruz et al. 2012; Muñoz-Pajares et al. 2017; Larranaga et al. 2017; Morris and Shaw 2018).

Despite the great number of species in the genus Baccharis, there are microsatellite markers developed for B. dracunculifolia only (Belini et al. 2016). A set of 17 markers was developed for B. dracunculifolia using a genomic microsatellite library (Belini et al. 2016), but out of 17, 12 are dinucleotide microsatellite that implies genotype scoring errors detailed above. Besides, six of them were monomorphic for three B. dracunculifolia populations (N = 315 individuals), which reinforced the need to advance in the development of new microsatellite markers for this species.

1 Development of Microsatellite Markers for Baccharis dracunculifolia Using NGS Technology

A good strategy to develop microsatellite markers with better performance is using next-generation sequencing (NGS) technology (Zalapa et al. 2012; Ambreen et al. 2015; Bonatelli et al. 2015; Hodel et al. 2016). The NGS allows the rapid and efficient development of microsatellite markers for non-model organisms for ecological and evolutionary studies. Moreover, the advent of NGS provided a cheaper and faster microsatellite development (Guichoux et al. 2011). We followed this approach to develop microsatellite markers for B. dracunculifolia, as described next.

A genomic library was built from 100 μg of one individual of B. dracunculifolia, in which DNA was extracted from leaves, through paired-end strategy that was sequenced using MiSeq® platform (Illumina©, San Diego, CA) to produce paired-end 250 base reads. A total of 21.4 million reads were obtained, and we used the Perl script PAL_FINDER_v0.02.04 (see Castoe et al. 2012) to identify 11,296 potentially amplifiable locus (PAL) (Table 3.2). We extracted reads that contained perfect dinucleotide, trinucleotide, tetranucleotide, pentanucleotide, and hexanucleotide tandem SSRs, totaling 7,277 PALs.

Table 3.2 Simple sequence repeat types in Baccharis dracunculifolia contigs sequences
Full size table

The SSR repeat motifs consisted of 24.08% dinucleotide, 58.55% trinucleotide, 6.95% tetranucleotide, 5.35% pentanucleotide, and 5.04% hexanucleotide repeat units (Fig. 3.1a). The AT/TA motif repeats were the most abundant dinucleotide SSR, accounting for 76.32% of all dinucleotide repeats (Fig. 3.1b). The trinucleotide AAT/TAA motif repeats were the most abundant type, accounting for 27.20% of all trinucleotide repeat motifs, while ATA/TAT and ATT/TAA repeats accounted for 17.10% and 14.90%, respectively (Fig. 3.1c).

Fig. 3.1
figure 1

Characteristics of SSR motifs identified in Baccharis dracunculifolia using NGS sequencing. (a) Frequency of di-, tri-, penta-, tetra-, and hexanucleotide SSR motif repeats. (b) Frequency of different dinucleotide SSR motifs. (c) Frequency of different trinucleotide SSR motifs

Full size image

Then, a manual filtering step was performed to select exclusively SSR loci with long and perfect repeats motifs since they tended to be more polymorphic (Zalapa et al. 2012). Thus, we chose tri-, tetra-, penta-, or hexanucleotide repeats present in long reads, larger than 274 bp, and obtained 1356 microsatellite loci candidates for microsatellite markers. To ensure that SSR loci chosen for B. dracunculifolia could follow as much as possible the stepwise mutation model used in coalescent-based methods to infer demographic events as proposed by Estoup et al. (2001), we selected just perfect motifs. Thus, 36 primer pairs flanking the SSR loci were designed using Primer3 software (Untergasser et al. 2012). Following, a set of 17 perfect microsatellite markers were chosen for amplification screening using 15 B. dracunculifolia individuals from the rupestrian grassland vegetation in Serra do Cipó, Brazil (Fig. 3.2). The PCR products were viewed on a polyacrylamide gel electrophoresis (PAGE) 6% and stained with silver nitrate (Sanguinetti et al. 1994).

Fig. 3.2
figure 2

Sites where individuals of Baccharis dracunculifolia were sampled in the rupestrian grassland vegetation in Serra do Cipó, Brazil

Full size image

A total of 12 microsatellite loci (Bdr6, Bdr7, Bdr9, Bdr11, Bdr13, Bdr20, Bdr21, Bdr22, Bdr25, Bdr26, Bdr31, Bdr34) produced clear amplicons with expected size in the acrylamide gel (Fig. 3.3). Then, we designed all these 12 primer pairs, and the forward primers were marked with 4 dyes: VIC®, 6-FAM™, PET®, and NED™.

Fig. 3.3
figure 3

An acrylamide gel stained with silver nitrate to verify polymorphism in Bdr6 microsatellite marker developed for Baccharis dracunculifolia before labeling them with fluorophores. The numbers above correspond to B. dracunculifolia individuals

Full size image

To assess the polymorphism and population genetic parameters with these microsatellite markers, we genotyped 60 individuals of Baccharis dracunculifolia from the Serra do Cipó region: 20 individuals between 760 and 839 m, 20 between 1026 and 1040 m, and 20 from 1348 to 1356 m altitudes (Fig. 3.2). The DNA was extracted from leaves using CTAB 2% protocol (Doyle and Doyle 1990). DNA purity and concentration were checked using NanoDrop 2000 Spectrophotometer (Thermo Fisher Scientific, Waltham, Massachusetts, USA).

All B. dracunculifolia individuals were genotyped with six microsatellite loci (Bdr6, Bdr7, Bdr22, Bdr25, Bdr26, Bdr31). PCR amplifications were performed in a 13 μL volume containing 10.0 μM of each primer, 1.5 μL of 5X special IVB PCR buffer, 1 unit of Taq DNA polymerase (Phoneutria, BR), 0.25 mM of each dNTP, 0.215–2.15 mM of MgCl2 (according to each primer – Table 3.3), and 10.0 ng of template DNA. DNA amplification was accomplished in a PCR system (Veriti™ 96-Well Thermal Cycler; Applied Biosystems, California, USA) under the following conditions: 94 °C for 5 min (one cycle); 94 °C for 1 min, 52–62 °C for 1 min (according to each primer – Table 3.3); 72 °C for 1 min (35 cycles); and 72 °C for 50 min. The PCR products were electrophoresed on an ABI Prism 3730 automated DNA sequencer (Fig. 3.4) (Applied Biosystems, California, USA) and were sized by comparison to a GeneScan 500 LIZ dye Size Standard (Applied Biosystems, California, USA). Fluorescent PCR products were automatically sized using Geneious 10.2.3 (Kearse et al. 2012).

Table 3.3 Characterization of 12 polymorphic microsatellite markers developed in Baccharis dracunculifolia
Full size table
Fig. 3.4
figure 4

Microsatellite profile of Bdr31 marker developed for Baccharis dracunculifolia detecting different heterozygote peaks

Full size image

The number of alleles (A) ranged from 2 to 8 per locus, and the average number of alleles in this population was 5.33 (Table 3.4). Despite the use of just 6 microsatellite markers, we found higher allelic richness than Belini et al. (2016) using 11 microsatellite markers based on 315 samples from 3 populations. Besides, Belini et al. (2016) developed six other microsatellite markers which were monomorphic in all these individuals. All these monomorphic markers were dinucleotide repeats, and the polymorphic markers were composed of six dinucleotide and five compound markers. Also, these authors used the traditional genomic microsatellite library to develop microsatellite markers. This result reinforces the better outcomes when using NGS to identify microsatellites that could enable the selection of tri-, tetra-, or pentanucleotide motif repeats to avoid “stutter” bands and genotype scoring errors because of dinucleotide repeat. All microsatellite markers developed by us have tri- or more motif repeats, and all of them are perfect microsatellite markers, which are considered more polymorphic. These characteristics highlight the potential of our markers.

Table 3.4 Characterization of 6 microsatellite loci based on a sample of 60 adult individuals of Baccharis dracunculifolia from Serra do Cipó, Brazil
Full size table

The observed heterozygosity (Ho) ranged from 0.231 to 0.789 per locus, and the average was 0.461. The expected heterozygosity (He) ranged from 0.408 to 0.807 per locus, and the average was 0.603. For most loci (except Bdr25), the observed heterozygosity was lower than expected under the Hardy-Weinberg equilibrium (HWE), with fixation indexes (Fis) significantly different from zero (Table 3.4). These HWE deviations may be the presence of null alleles or due to the low number of analyzed individuals, which must be insufficient to reveal all possible genotypic combinations (López-Márquez et al. 2016), hence leading to a possible underestimation of allele frequencies and heterozygosity (McInerney et al. 2011). In addition, the observed heterozygosity was lower than expected under HWE which may be due to excess homozygotes since all evaluated individuals belong to only one population and inbreeding can occur in this population.

The probability of excluding two individuals as related when they are not was 74.3% with the Bdr6 locus. However, the combined probability of paternity exclusion increased to 99.5% when the six loci were included in the analysis (QC = 0.995 – Table 3.4), indicating that this set of microsatellite markers is efficient in kinship and gene flow analyses. The probability of genetic identity (I) ranged from 0.187 to 0.934 per locus (Table 3.3), and a low combined probability of genetic identity (IC) was attained when the six loci were included in the analysis (IC = 5.1 × 10−05 – Table 3.3). Although some loci presented a significant excess of homozygotes, the higher combined probability of paternity exclusion and lower combined probability of genetic identity show that this battery of microsatellite markers is suitable for population genetic analyses.

2 Cross-Amplification in Baccharis

We sampled other two species, B. concinna and B. aphylla, to evaluate the transferability of microsatellite markers developed for B. dracunculifolia. All of these three species occur in sympatry in the rupestrian grasslands of Serra do Cipó. We collected leaves from eight individuals of B. concinna and nine individuals of B. aphylla and extracted their DNA using the CTAB 2% protocol (Doyle and Doyle 1990). DNA purity and concentration were checked using NanoDrop 2000 Spectrophotometer (Thermo Fisher Scientific, Waltham, Massachusetts, USA).

The cross-amplification analysis was realized with the six microsatellite markers characterized in B. dracunculifolia (see Table 3.3) under the same PCR conditions used to amplify B. dracunculifolia. The PCR fragments were viewed on a 6% polyacrylamide gel electrophoresis (PAGE) stained with silver nitrate (Sanguinetti et al. 1994). Five microsatellite markers that successfully amplified fragments were obtained both in B. concinna (Bdr6, Bdr7, Bdr25, Bdr26, Bdr31) (Fig. 3.5a) and in B. aphylla (Bdr6, Bdr7, Bdr22, Bdr25, Bdr31) (Fig. 3.5b).

Fig. 3.5
figure 5

Microsatellite profile of microsatellite markers developed for Baccharis dracunculifolia amplified in congeneric species, (a) Bdr6 profile in Baccharis concinna, PCR fragments were detected in individuals 1–5, (b) Bdr31 profile in Baccharis aphylla, PCR fragments were detected in individuals 3, 6, 7, and 9

Full size image

Modifications of the tested PCR conditions (mainly annealing temperature, DNA and MgCl2 concentration) may have increased this preliminary success of cross-species amplification in Baccharis. In addition, other microsatellites developed for B. dracunculifolia, but not yet characterized for this species, can be used in future cross-amplification in this genus. These microsatellite markers must provide new information about the population genetic structure of B. dracunculifolia and related species and may help elucidate more details on the evolutionary relationships in this genus. Besides, this new molecular tool may help in the management and conservation of B. dracunculifolia as well as other species in the genus. We argue that urgent genetic studies on the genus Baccharis are called for as the importance of its species in community assembly, ecosystem services, and potential invisibility of disturbed communities is increasing in the recent decade.