Abstract Forest-dwelling taxa of the Brazilian Atlantic rainforest experienced range contractions during the glacial periods of the Quaternary. In contrast, species of drier environments, such as the restinga, may show different patterns. Here, we investigated a bromeliad from the southern Brazilian restinga, Aechmea kertesziae. We used nine nuclear microsatellite markers, two non-coding plastid sequences and controlled pollination experiments to examine the genetic structure and diversity across the geographical distribution of the species. Our results suggest demographic stability for A. kertesziae during the Quaternary, despite other restinga species showing demographic expansion. Aechmea kertesziae shows high genetic diversity, which is probably the result of its self-incompatibility, allied to the long-term persistence of the populations, constant population sizes and clonal reproduction. A strong phylogeographical break was identified with the plastid markers, recovering two main evolutionary lineages showing little gene flow between them. However, no geographical barrier could be associated with this deep Pliocene divergence. Moderate levels of genetic structure were detected with nuclear microsatellites, which was associated with a more recent habitat fragmentation process. Conservation efforts should prioritize the establishment of effective gene flow among populations. INTRODUCTION The restinga is a component of the Atlantic rainforest, characterized by dunes and sandy plains covered by herbaceous and shrubby vegetation under direct sunlight, stretching along the eastern coast of Brazil (Rocha et al., 2007). Most of the Brazilian human population is concentrated along the coast and consequently coastal habitats have been seriously fragmented. Having suffered high rates of habitat loss (Rocha et al., 2007; Marques, Silva & Liebsch, 2015; Cosendey, Rocha & Menezes, 2016), only 0.47% of the original area of restinga habitat type remain (Ribeiro et al., 2009). The flora of the restinga comprises a subset of the species found in neighbouring vegetation, which colonized these Quaternary habitats after the last glaciations of the Quaternary (Marques et al., 2015, and references therein). Some degree of endemism is observed in this environment, with the presence of species considered to be restricted to this habitat, e.g. Androtrichum trigynum (Spreng.) H.Pfeiff., Connarus nodosus Baker, Cryptanthus acaulis Beer, Kielmeyera albopunctata Saddi and Maytenus obtusifolia Mart., indicating that taxa from different vegetation types differentiated in the restinga over time, resulting in new species (Marques et al., 2015). Dry environments such as the restinga have been relatively neglected in studies aiming to understand the processes behind biodiversity, especially in subtropical South America (e.g. Pinheiro et al., 2011; Werneck, 2011; Brunes et al., 2015; Cardoso et al., 2015). The few phylogeographical studies conducted (Salgueiro et al., 2004; Pinheiro et al., 2011; Lopes et al., 2013; Cardoso et al., 2015; Turchetto-Zolet et al., 2016) have demonstrated that species from the restinga may not have been restricted to glacial refugia during the Pleistocene, as described for forest-dwelling species of the Atlantic rainforest (Turchetto-Zolet et al., 2013). Whereas forest-dwelling taxa suffered geographical fragmentation during the Quaternary glaciations (Turchetto-Zolet et al., 2013), most restinga species experienced range expansions, and show signs of recent demographical expansion and low population genetic structure (Pinheiro et al., 2011; Cardoso et al., 2015; Turchetto-Zolet et al., 2016). These contrasting patterns observed in the same biome, the Atlantic rainforest (e.g. Turchetto-Zolet et al., 2013, 2016; Cardoso et al., 2015), reflect the complexity of the phylogeography of this region and highlight the need for more studies to understand the origin and diversification of its biota (Turchetto-Zolet et al., 2013). Genetic structure and divergence among populations, which can ultimately lead to diversification, can arise from many factors. For plants these factors may include historical events, e.g. those imposed by the climatic oscillations of the Pleistocene and more recent anthropogenic disturbance that causes habitat fragmentation (Fahrig, 2003), which can impose barriers to gene flow (e.g. Federman et al., 2014). Moreover, autogamous species show higher genetic structure than outcrossing species, reflecting the effect of mating systems on the genetic composition of natural populations (Hamrick, 1982; Hamrick & Godt, 1996; Holsinger, 2000; Nybom, 2004; Glémin, Bazin & Charlesworth, 2006). Identifying the main causes of the levels of intraspecific genetic structure currently observed and of the mating system of a species is crucial for understanding the patterns and distribution of genetic diversity among and within its populations and ensuring its survival through effective conservation strategies. In highly endangered species from fragmented habitats, such as the restinga, these aspects take on a particular importance. Nuclear microsatellite (simple sequence repeats, SSRs) are one of the most useful molecular markers in population genetics, due to their high degree of polymorphism, co-dominant inheritance and ease of use (Nybom, Weising & Rotter, 2014). Due to their high mutation rates, microsatellites reflect a more recent pattern of pollen and seed dispersal compared to plastid DNA (Provan, Powell & Hollingsworth, 2001). Plastid markers are used to estimate genealogical histories in populations based on their non-recombinant nature, low mutation rates and haploid, generally maternal inheritance (Ennos, 1994; Avise, 2009). Therefore, the use of different kinds of molecular markers, biparentally (SSR) and uniparentally (plastid DNA) inherited, will provide the contemporaneous and historical aspects, respectively, that are behind the evolutionary history of a species (Scotti-Saintagne et al., 2013) and will help in understanding the main factors responsible for the levels of intraspecific genetic structure found. To contribute to the understanding of the diversification of plants from dry environments of subtropical South America we used as a model Aechmea kertesziae Reitz (Bromeliaceae). Bromeliads play an important role in dry environments because many species are able to hold great amounts of water in the tanks formed by the insertion of their leaves. The water accumulated in the tanks of bromeliads facilitates their maintenance in this kind of environment. It is also used by associated fauna inhabiting the tanks, such as insects, spiders and frogs, and helps in the establishment of other plant species (Benzing, 2000; Cogliatti-Carvalho et al., 2010, and references therein). Aechmea Ruiz & Pav. is one of the most diverse genera in subfamily Bromelioideae, comprising c. 280 species with a geographical distribution ranging from Mexico to Uruguay (Smith & Downs, 1979; Luther, 2012). Aechmea kertesziae is endemic to subtropical South America, and occurs preferentially in the restinga habitat of Santa Catarina state, Brazil, between 26° and 28°S (Reitz, 1983; Falkenberg, 1999; Goetze et al., 2016a). As seen for other restinga species (Pinheiro et al., 2011; Cardoso et al., 2015; Turchetto-Zolet et al., 2016), we hypothesized that A. kertesziae would show signs of recent demographic expansion and low historical population structure. We also expected to find high rates of recent genetic structure, mostly revealed by using nuclear molecular markers, and low genetic variability due to the high rates of habitat fragmentation encountered in coastal areas of Brazil. Here, we used nine nuclear SSRs, two non-coding plastid DNA regions and controlled pollination experiments to investigate the genetic diversity and structure of A. kertesziae populations. The aims of the study were: (1) to test if A. kertesziae underwent a persistent range during the climatic oscillations of the Pleistocene, and if it presents signs of recent demographic expansion, as observed for other restinga species; (2) to investigate the extent of genetic structure across the small range of the species; (3) to identify the contribution of the mating system of the species to the observed genetic diversity; and (4) to determine the conservation status of the species as well as recommend conservation strategies. MATERIAL AND METHODS Studied species and area Aechmea kertesziae is a tank-forming bromeliad that can reproduce clonally. It is pollinated mainly by insects, and bumble bees are the most frequent visitors (Reitz, 1983; M. V. Büttow, unpubl. data). As described for other Aechmea spp., its seeds are probably dispersed by birds (Fischer & Araujo, 1995; Lenzi, Matos & Orth, 2006). The development of mature fruits with viable seeds takes about 6 months from pollination (M. V. Büttow, unpubl. data). The development of some fruits without seeds is observed for this species, probably due to parthenocarpy, as discussed for other taxa of Aechmea (Lenzi et al., 2006; Büttow, 2012). Aechmea kertesziae was recently included in the Brazilian government’s list of endangered species (MMA, 2014); it belongs to the yellow-flowered Aechmea subgenus Ortgiesia Regel, which includes seven morphologically similar species (Goetze et al., 2017). Aechmea kertesziae occurs preferentially under shrubby vegetation in restinga, where it is mostly found as a terrestrial (growing at the sand surface) or rupicolous species, and less frequently as an epiphyte (Reitz, 1983). The area studied here, and where A. kertesziae is found, is highly impacted by tourism and most of the populations occur near famous beaches. According to herbarium records (speciesLink: http://www.splink.org.br, accessed 13 March 2017) and field expeditions, A. kertesziae is currently found in five main areas, which were sampled for this study (Table 1, Fig. 1A). In the ITA, CAM and BOM populations, individuals are found near the sea. In the FLO population, A. kertesziae is found within a conservation area, under secondary forest. After FLO, LAG is the least impacted area, and the individuals are found near a small village that has only limited access by car. Figure 1. View largeDownload slide Map showing the sampled populations of Aechmea kertesziae in southern Brazil and the haplotype network recovered. (A) Pie charts reflecting the frequency of occurrence of each haplotype in each population. Haplotype colours correspond to those shown in the key on the left. Population codes correspond to those in Table 1. (B) Median-joining network based on plastid DNA sequences of A. kertesziae. Each circle represents one haplotype; the diameter of the circles is proportional to each haplotype’s total frequency. More than one mutational step required to explain transitions among haplotypes is indicated by numbers along the network. Dotted lines indicate the two main groups found. Figure 1. View largeDownload slide Map showing the sampled populations of Aechmea kertesziae in southern Brazil and the haplotype network recovered. (A) Pie charts reflecting the frequency of occurrence of each haplotype in each population. Haplotype colours correspond to those shown in the key on the left. Population codes correspond to those in Table 1. (B) Median-joining network based on plastid DNA sequences of A. kertesziae. Each circle represents one haplotype; the diameter of the circles is proportional to each haplotype’s total frequency. More than one mutational step required to explain transitions among haplotypes is indicated by numbers along the network. Dotted lines indicate the two main groups found. Table 1. Details of sampling localities and sample sizes of Aechmea kertesziae. Population ID Voucher Habitat Latitude S Longitude W Elevation (m) Sample size Nuclear Plastid Itajaí ITA ICN 191153 restinga 26°55′ 48°38′ 19 22 15 Camboriú CAM FURB 28103 restinga 27°00′ 48°34′ 5 - 6 Bombinhas BOM CESP 62360 restinga 27°08′ 48°29′ 19 36 8 Florianópolis FLO UPBC 35253 secondary forest 27°31′ 48°30′ 139 20 8 Laguna LAG ICN 167498 restinga 28°30′ 48°45′ 11 25 15 Total 103 52 Population ID Voucher Habitat Latitude S Longitude W Elevation (m) Sample size Nuclear Plastid Itajaí ITA ICN 191153 restinga 26°55′ 48°38′ 19 22 15 Camboriú CAM FURB 28103 restinga 27°00′ 48°34′ 5 - 6 Bombinhas BOM CESP 62360 restinga 27°08′ 48°29′ 19 36 8 Florianópolis FLO UPBC 35253 secondary forest 27°31′ 48°30′ 139 20 8 Laguna LAG ICN 167498 restinga 28°30′ 48°45′ 11 25 15 Total 103 52 CESJ, Herbarium Leopoldo Kriger; FURB, Herbarium Dr. Roberto Miguel Klein; ICN, Herbarium of Instituto de Ciências Naturais; UPBC, Herbarium Departamento de Botânica, Universidade Federal do Paraná. View Large Population sampling and DNA isolation We collected 109 individuals from five populations of A. kertesziae across its current geographical distribution (Reitz, 1983; Goetze et al., 2016a; Table 1, Fig. 1A). The closest populations are ITA and CAM (c. 9 km apart), and the two most distant populations are ITA and LAG (c. 174 km apart). Distances between the populations are provided in Supporting Information, Table S1. Sample collection took place in 2010 and 2011 and, to avoid misidentifications, only reproductive individuals (with flowers or fruits) were sampled. Information on sampling localities and the number of individuals from each population used in DNA sequencing and SSR analyses is summarized in Table 1. To avoid repeated sampling of the same individual due to clonal reproduction, samples were collected at least 10 m apart. Young leaves were collected and stored in silica gel for drying. Total genomic DNA was extracted using the cetyltrimethylammonium bromide (CTAB) protocol (Doyle & Doyle, 1990). Microsatellite markers and genotyping assays Nine microsatellite loci were used to genotype 103 individuals from four populations: Ac01, Ac11, Ac25 and Ac55 (Goetze et al., 2013); Acom_71.3, Acom_78.4, Acom_82.8 and Acom_91.2 (Wörhmann & Weising, 2011); and Dd10 (Zanella et al., 2012a). The population named ‘CAM’ was not included in the microsatellite analysis because only six individuals were found. All PCR amplifications were performed in a Veriti 96-Well Thermal Cycler (Applied Biosystems, Foster City, CA, USA) following the conditions described by Goetze et al. (2013). Microsatellite alleles were resolved on an ABI 3500 DNA Analyzer Sequencer (Applied Biosystems) and sized against the GS500 LIZ molecular size standard (Applied Biosystems) using GeneMarker Demo version 1.97 (SoftGenetics, State College, PA, USA). Amplification and sequencing Two non-coding plastid DNA regions, rpl32-trnL and rps16-trnK, were amplified and sequenced for 52 individuals from all sampled populations (Table 1), using the primers described by Shaw et al. (2007). All PCRs were carried out as described by Goetze et al. (2016b). PCR amplifications were performed in a Veriti 96-Well Thermal Cycler (Applied Biosystems). PCR products were sequenced in both directions using the BigDye Kit (Applied Biosystems) at Macrogen Inc. (Seoul, Korea). All sequences have been deposited in GenBank under the following accession numbers: KY556987–KY557038 (rpl32-trnL) and KY557262–KY557313 (rps16-trnK). Analysis of microsatellite data To characterize the genetic diversity of A. kertesziae populations, the following population genetics indices were determined using the programs Fstat 184.108.40.206 (Goudet, 1995) and MSA 4.05 (Dieringer & Schlötterer, 2003): number of alleles (A), number of private alleles (AP), allelic richness (RS), observed (HO) and expected (HE) heterozygosities and inbreeding coefficient (FIS; Weir & Cockerham, 1984). For each population, departures from Hardy–Weinberg equilibrium (HWE) were identified using exact tests in Genepop on the web (Raymond & Rousset, 1995). To identify possible null alleles, we used the software Micro-Checker 2.2.3 (Van Oosterhout et al., 2004). Estimates of FST (Weir & Cockerham, 1984) and G′ST (Hedrick, 2005) were obtained to assess the genetic differentiation of populations using the software Fstat. Pairwise comparisons of FST between populations were estimated using the program Arlequin 220.127.116.11 (Excoffier & Lischer, 2010). Partitioning of genetic diversity within and among populations was examined by analysis of molecular variance (AMOVA; Excoffier, Smouse & Quattro, 1992) implemented in the software Arlequin with 10000 permutations. The correlation between geographical and genetic distance matrices (FST) was estimated to test the hypothesis that populations are differentiated because of isolation-by-distance (Wright, 1965). A standardized Mantel test was run using Genepop and the significance was assessed through a randomization test using 10000 Monte Carlo simulations. To investigate population structure, we performed a Bayesian analysis implemented in the software Structure version 2.3.4 (Pritchard, Stephens & Donnelly, 2000). We ran Structure for K = 1–6 with ten replicates each and a model based on admixture and correlated allelic frequencies, without taking into account information regarding sampling localities. Each run had 106 iterations with a burn-in of 250000. The best K value was determined by using the maximum value of ΔK (Evanno, Regnaut & Goudet, 2005), with Structure Harvester version 0.6.94 (Earl & von Holdt, 2012). To depict relationships between populations, a neighbor-joining (NJ) tree was constructed based on the proportion of shared alleles among populations (Bowcock et al., 1994). One thousand bootstrap replicates of the distance matrix were obtained in MSA, and the NJ trees were generated in PHYLIP 3.69 (Felsenstein, 2005). The software FigTree 1.4 was used to draw the tree (Rambaut, 2008). Each population was tested for recent population size reductions (i.e. genetic bottlenecks), using a heterozygosity excess test implemented in the software Bottleneck 1.2.02 (Piry, Luikart & Cornuet, 1999). The analysis was carried out using a two-phased mutation model (TPM), with 12% of variance and 95% of the stepwise mutation model in the TPM. Statistical significance was assessed by 10000 replicates using a one-tailed Wilcoxon signed-rank test. Analysis of plastid DNA sequences All sequences were checked using Chromas 2.32 (Technelysium, Helensvale, Australia) and aligned using the MUSCLE (Edgar, 2004) tool implemented in MEGA version 5.10 (Tamura et al., 2011). Mononucleotide repeat length variations were excluded due to ambiguous alignment. Indels of more than one base pair (bp) were coded as a single mutational event. The two plastid DNA sequences (rpl32-trnL and rps16-trnK) were concatenated for all analyses. To characterize genetic diversity, haplotype (h) and nucleotide (π) diversities (Nei, 1987), GC content and number of variable sites were estimated for each population using Arlequin. Haplotypes were identified using DnaSP 5.10.01 (Librado & Rozas, 2009), and the evolutionary relationships among them were estimated with Network 5 (http://www.fluxus-engineering.com/sharenet.htm, accessed 14 March 2017), using the median-joining method (Bandelt, Forster & Röhl, 1999). An AMOVA was conducted to assess the genetic differentiation among populations, using Arlequin under 10000 permutations. A hierarchical AMOVA was also conducted based on the results obtained with network analysis (see Results, Fig. 1B). Pairwise comparisons of ΦST (analogous to FST, preferentially used with sequence data) between populations were estimated using the program Arlequin with 10000 permutations. BAPS version 6 (Corander et al., 2008) was used to analyse the population genetic structure by clustering sampled individuals into groups. This analysis was carried out as described by Goetze et al. (2016b). Ten iterations of each K from 1 to 7 were conducted. To assess the demographic history of A. kertesziae, Tajima’s D (Tajima, 1989) and Fu’s Fs (Fu, 1997) neutrality tests were carried out using plastid DNA in Arlequin. Statistical significance was determined based on 10000 simulations. These analyses were run considering the species as a whole (all individuals) and grouped accordingly into groups I and II observed in the data (see Results, Fig. 1B). Additionally, changes in population size over time for the species as a whole and for groups I and II were estimated using Bayesian Skyline Plot analysis (BSP; Drummond et al., 2005), performed in Beast version 1.7.5 (Drummond et al., 2012). The following priors were applied: lognormal relaxed clock (uncorrelated) with a substitution rate previously estimated for plastid DNA in subfamily Bromelioideae (7.64 × 10–4 ± 4.5 × 10–6; D Silvestro, pers. commun.), and the HKY nucleotide substitution model. Markov chains were run for 50000000 steps, sampling every 1000 steps. BSP computation and convergence checks were performed in Tracer 1.5 (Rambaut et al., 2013). An effective sample size (ESS) > 200 was used as a threshold (Drummond & Rambaut, 2007). The time of plastid DNA haplotype divergence was estimated using a Bayesian approach implemented in the software Beast, using Aechmea nudicaulis Griseb. and Bromelia antiacantha Bertol. (GenBank accession numbers: B. antiacantha – KY557334 for rpl32-trnL and KY557335 for rps16-trnK; A. nudicaulis – MF737165 for rpl32-trnL and MF737166 for rps16-trnK) as an outgroup. Priors used included the birth–death model, the HKY nucleotide substitution model and the lognormal relaxed clock (uncorrelated). The same substitution rate for plastid DNA was used as in the BSP analysis. Markov chains were run for 10000000 steps, sampling every 1000 steps. The results were viewed in Tracer to check for convergence to a stationary distribution and for ESS > 200 (Drummond & Rambaut, 2007). TreeAnnotator 1.7.5, part of the Beast package, was used to summarize the trees, and statistical support for all branches was measured in Bayesian posterior probabilities (PP). The software FigTree was used to draw the tree. Controlled pollination experiments To determine the breeding system of A. kertesziae, hand pollination experiments were conducted in situ in population FLO, where the experiment could be carried out without interference. One day before the experiments, all flowers that would open the next day were covered with a paper bag to avoid uncontrolled pollination. Five treatments were conducted: manual cross-pollination – emasculated flowers were pollinated with a fresh pollen mixture collected from distinct plants and bagged (number of flowers manipulated = 19); open-pollination (control) – flowers were tagged and not manipulated (37); manual self-pollination – flowers were pollinated with their own pollen and bagged (18); spontaneous autogamy – flowers were bagged to exclude insect visits (15); and agamospermy emasculated flowers were bagged (12). After 6 months, fruit set and number of seeds per fruit were recorded. Treatments that produced fruits were analysed using a chi-squared (χ2) test in SPSS 19.0 (IBM, New York, NY, USA). RESULTS Genetic variation High levels of genetic variation were found in A. kertesziae populations genotyped at the nine nuclear microsatellite loci (Tables 2 and S2). The number of alleles per locus ranged from seven (Dd10) to 16 (Ac55), with a mean of 10.56 (Table S2). The number of alleles per population ranged from 57 to 73, and the allelic richness ranged from 5.68 to 7.22. The observed and expected heterozygosities per population ranged from 0.550 to 0.691 and from 0.661 to 0.747, respectively. The number of private alleles varied from three to nine per population. The inbreeding coefficients ranged from 0.075 to 0.221, and all populations departed significantly from HWE, with an excess of homozygotes (Table 2). Micro-Checker analysis detected the presence of null alleles at six loci in different populations. However, the estimated frequency of null alleles was not significant (P > 0.05), except for locus Ac55 in population BOM (data not shown). Table 2. Characterization of genetic variability in five populations of Aechmea kertesziae in restinga of southern Brazil. Population SSR Plastid DNA A AP RS HO HE FIS* Haplotypes h π ITA 57 3 5.92 0.592 0.702 0.191 1, 2, 3, 4 0.7048 0.000951 CAM – – – – – – 4, 5, 6 0.6000 0.001623 BOM 62 9 5.68 0.550 0.661 0.221 4, 5, 7 0.7500 0.000653 FLO 57 3 5.98 0.639 0.728 0.131 5, 8, 9 0.6786 0.000674 LAG 73 8 7.22 0.691 0.747 0.075 10, 11, 12, 13, 14 0.7810 0.001043 Population SSR Plastid DNA A AP RS HO HE FIS* Haplotypes h π ITA 57 3 5.92 0.592 0.702 0.191 1, 2, 3, 4 0.7048 0.000951 CAM – – – – – – 4, 5, 6 0.6000 0.001623 BOM 62 9 5.68 0.550 0.661 0.221 4, 5, 7 0.7500 0.000653 FLO 57 3 5.98 0.639 0.728 0.131 5, 8, 9 0.6786 0.000674 LAG 73 8 7.22 0.691 0.747 0.075 10, 11, 12, 13, 14 0.7810 0.001043 A, number of alleles; AP, number of private alleles; RS, allelic richness; HO, observed heterozygosity; HE, expected heterozygosity; FIS, inbreeding coefficient; h, haplotype diversity; π, nucleotide diversity. *All inbreeding coefficients (FIS) departed significantly from Hardy–Weinberg equilibrium (HWE) at P < 0.001. View Large Sequencing of the intergenic spacers rpl32-trnL and rps16-trnK generated fragments of 895 and 780 bp in length, respectively. After removing mononucleotide repeats and editing indels longer than 1 bp, the final dataset of concatenated plastid DNA spacers totalled 1645 bp, with a GC content of 27.8%. Twenty-four polymorphic sites were observed (ten transitions, eight transversions and six indels). Fourteen different haplotypes were found in the 52 individuals analysed. Haplotype diversity ranged from 0.6000 to 0.7810 and nucleotide diversity varied from 0.000653 to 0.001623. The number of haplotypes varied from three to five per population (Table 2). Haplotypes 4 and 5 were the most frequent, occurring in populations from group I (Fig. 1). Population structure Moderate levels of genetic differentiation across populations were found using microsatellite data, as indicated by FST (0.110) and G′ST (0.106). Pairwise FST values also revealed moderate genetic structure, ranging from 0.070 to 0.113; all values were statistically significant (P < 0.001; Table 3). AMOVA results indicated that most genetic variation resides within populations (90.85%, P < 0.001), and only 9.15% is found among populations. No correlation among genetic and geographical distances was detected by the Mantel test (r2 = 0.2152, P = 0.109), indicating the absence of isolation by distance among the collecting sites. The optimum K value determined by Evanno’s method, implemented in Structure Harvester, was K = 4 genetic groups, as shown in Figure S1. The four genetic groups observed correspond to the collection sites. Several individuals showed some degree of admixture (Fig. 2A), in line with the moderate levels of genetic differentiation (FST) recovered among populations. The NJ tree recovered the four populations with high to moderate bootstrap values (Fig. 2B). In this analysis, populations ITA and BOM are closely related to each other. Population LAG is closely related to ITA and BOM, and FLO is the most differentiated population. Table 3. Pairwise genetic divergence for Aechmea kertesziae populations based on nine microsatellite loci (FST; below diagonal) and plastid sequence data (ΦST; above diagonal) ITA CAM BOM FLO LAG ITA 0.368 0.283 0.541* 0.849* CAM – 0.144 0.169 0.794* BOM 0.070* – 0.419 0.847* FLO 0.090* – 0.103* 0.844* LAG 0.113* – 0.083* 0.100* ITA CAM BOM FLO LAG ITA 0.368 0.283 0.541* 0.849* CAM – 0.144 0.169 0.794* BOM 0.070* – 0.419 0.847* FLO 0.090* – 0.103* 0.844* LAG 0.113* – 0.083* 0.100* *All values were significant at P < 0.001. Dashes indicate that the population was not analysed for microsatellites. View Large Figure 2. View largeDownload slide Population genetic structure and relationship in Aechmea kertesziae. (A) Bayesian assignment analysis for the K = 4 populations model based on nine nuclear microsatellite loci inferred with Structure. (B) Neighbor-joining tree obtained from a distance matrix based on shared alleles among populations. Bootstrap values (> 70) are shown above the branches. Population codes correspond to those in Table 1. Figure 2. View largeDownload slide Population genetic structure and relationship in Aechmea kertesziae. (A) Bayesian assignment analysis for the K = 4 populations model based on nine nuclear microsatellite loci inferred with Structure. (B) Neighbor-joining tree obtained from a distance matrix based on shared alleles among populations. Bootstrap values (> 70) are shown above the branches. Population codes correspond to those in Table 1. The analysis of non-coding plastid DNA regions found a network with two main groups of haplotypes. Group I included the haplotypes found in populations ITA, CAM, BOM and FLO, and group II corresponded to the haplotypes from population LAG. Groups I and II were separated by six mutational steps and do not share haplotypes (Fig. 1B). High levels of genetic differentiation were detected among populations from group I (ITA, CAM, BOM and FLO) and LAG, based on pairwise ΦST estimates (Table 3). According to the results of the AMOVA, most of the genetic variation is due to differences among populations (75.48%, P < 0.001), with 24.21% of the variation residing within populations. Hierarchical AMOVA does not indicate differentiation between groups I and II (FCT = 0.755, P = 0.202). BAPS analysis revealed that the best K value for plastid DNA is four. Although BAPS results indicated K = 4, the majority of the individuals belonged to two groups, named II and IV. Most of the individuals from populations ITA and CAM belonged to cluster II, together with all individuals from BOM and FLO. Individuals from LAG belonged to a distinct and unique cluster, IV (Fig. S2). Demographic analyses and time of divergence No excess of heterozygosity was detected in the bottleneck analysis for any of the four populations investigated with microsatellite loci, suggesting no changes in population sizes. The results of Tajima’s D and Fu’s Fs neutrality tests were not significant, either for the species as a whole or for either of groups I and II, indicating demographic stability (Table S3). The BSP analysis for the species as a whole suggested a recent bottleneck event. However, this result should be interpreted with caution given the size of the estimated confidence limits, which do not indicate statistical significance (Fig. S3A). The BSP results for the two groups recovered by network analysis showed no significant changes in population sizes through time (Fig. S3B, C). The divergence of the plastid DNA haplotypes of A. kertesizeae started around 4 Mya (95% highest posterior density: 1.95–7.14 Mya). Two main clades were observed in the phylogenetic tree with strong statistical support: one formed by haplotypes from populations ITA, CAM, BOM and FLO (group I in the network analysis), and the other with haplotypes from the LAG population (group II in the network analysis). Although the crown age of diversification is around 4 Mya, most lineages of A. kertesziae probably started to diversify in the early Pleistocene at around 2.0–1.5 Mya (Fig. 3). Figure 3. View largeDownload slide Bayesian phylogenetic tree of plastid DNA haplotypes with posterior probabilities (> 0.7) shown below the branches, and ages indicated for selected nodes. The time scale is in millions of years ago (Mya). Figure 3. View largeDownload slide Bayesian phylogenetic tree of plastid DNA haplotypes with posterior probabilities (> 0.7) shown below the branches, and ages indicated for selected nodes. The time scale is in millions of years ago (Mya). Breeding system The hand pollination experiments showed that only the manual cross-pollination and the open-pollination treatments produced fruits and seeds (Table 4). The manual cross-pollination experiment showed that c. 57% of the manipulated flower developed into fruits with seeds, whereas in the open-pollination treatment fruit with seed production reached 78%. Seed production was higher in the manual cross-pollination experiment (98%) than in the open-pollination treatment (61%) (Table 4), probably due to avoidance of geitonogamy in the first experiment. These results suggest that A. kertesziae is self-incompatible and an obligate outcrosser. Table 4. Breeding system experiments in Aechmea kertesziae in southern Brazil Treatment Number of flowers used per treatment Fruit with seed production Seed production (mean ± SE) N % Manual cross-pollination 19 11 57.89 98.68 ± 24.47 b Open pollination 37 29 78.38 61.19 ± 11.52 a Manual self-pollination 18 0 0 0 Spontaneous autogamy 15 0 0 0 Agamospermy 12 0 0 0 Treatment Number of flowers used per treatment Fruit with seed production Seed production (mean ± SE) N % Manual cross-pollination 19 11 57.89 98.68 ± 24.47 b Open pollination 37 29 78.38 61.19 ± 11.52 a Manual self-pollination 18 0 0 0 Spontaneous autogamy 15 0 0 0 Agamospermy 12 0 0 0 Seed production according to treatments. Mean followed by different letters are statistically different according to the χ2 (α = 0.05) test. χ2 = 66.89, d.f. = 36, P = 0.001. View Large DISCUSSION Aechmea kertesziae is mostly found in the restinga of southern Brazil, an area that has received less research attention than the Atlantic rainforest. Both microsatellite and plastid DNA-based analyses revealed a high level of genetic diversity and demographic stability for A. kertesziae. The patterns of genetic structure found, however, were different for the two types of markers. The plastid DNA analysis detected an important phylogeographical break, with two main evolutionary lineages, whereas nuclear microsatellites showed moderate genetic differentiation with four main groups. These results suggest a historical pattern of vicariance, followed by a more recent structuring among the populations found in group I (ITA, BOM and FLO). Genetic structure and demographic history The results of this study suggest two main evolutionary lineages for A. kertesziae, which diverged c. 4 Mya, during the Pliocene. These two lineages do not share haplotypes, thus revealing a marked genetic structure (Figs 1 and 3, Table 3). We did not find any reasonable geographical barrier compatible with this deep Pliocenic divergence. In other studies focusing on the flora of the southern Brazilian restinga, genetic breaks were observed further south, around the city of Torres (Pinheiro et al., 2011; Turchetto-Zolet et al., 2016), in a region historically recognized as an important phytogeographical boundary (Rambo, 1950). However, A. kertesziae does not reach this region, and LAG is the southernmost population of the species (Reitz, 1983; Goetze et al., 2016a). For the ant Mycetophylax simplex, no genetic structure was observed along its geographical distribution in the restinga, i.e. the southernmost Brazilian state of Rio Grande do Sul to Rio de Janeiro (Cardoso et al., 2015). Therefore, the historical event responsible for the deep phylogeographical break identified for A. kertesziae remains to be discovered. Despite the availability of larger coastal areas for occupation by A. kertesziae during the regression of the ocean in the Quaternary, the species maintained demographic stability, as shown by the neutrality tests and BSP analysis (Table S3, Fig. S3). In contrast, Cardoso et al. (2015) found a gradual demographic expansion, which coincided with low sea levels during the Quaternary for the restinga ant M. simplex. Northern populations of the tree Eugenia uniflora L., which are associated with the restinga, showed moderate changes in effective population sizes, with signatures of recent demographic expansion (Turchetto-Zolet et al., 2016). A plausible explanation for this difference is that A. kertesziae is found sheltered under shrubby and herbaceous vegetation in the restinga (our personal observations). Therefore, if the species that serve as shelter were not able to expand their range during the glacial periods of the Pleistocene, this could explain why A. kertesziae remained demographically stable, whereas other species of the restinga underwent a population expansion. Glacial periods during the Pleistocene were characterized by drier and cooler conditions compared to interglacial times in subtropical South America. Therefore, species from dry environments, already used to dry conditions, experienced a suitable climate to expand their ranges (Behling & Negrelle, 2001; Behling, 2002). However, according to a recent review, species associated with open vegetation tended to expand, maintain or shrink their geographical distribution ranges during glacial cycles, demonstrating a more variable response to the climatic oscillations of the Pleistocene than forest-dependent taxa in South America (Turchetto-Zolet et al., 2013). The results found in this study, using A. kertesziae as a model, showed that this species was demographically stable during the Pleistocene, and does not show signs of recent demographic expansion, which is in line with other studies conducted in dry environments in South America but is the opposite pattern to what is observed for Atlantic rainforest-dwelling taxa. However, high historical genetic structure was found for our model species, which is not documented by other studies conducted in the restinga. The pattern of genetic structure seen in the only two restinga plants so far investigated is linked to these species transition from restinga to other types of vegetation: grassland for the orchid Epidendrum fulgens Brongn. (Pinheiro et al., 2011) and riparian forests for the tree E. uniflora (Turchetto-Zolet et al., 2016). In these studies, populations located inside the restinga showed low genetic structure, in contrast to the pattern observed for A. kertesziae. The high historical genetic structure found for A. kertesziae could indicate seed dispersion barriers, as the plastid genome is probable maternally inherited in Bromeliaceae, as shown for the genus Fosterella L.B.Sm. (Wagner et al., 2015). Birds, e.g. passerines including Chiroxiphia spp., Tachyphonus coronatus and Tangara spp., are described as seed dispersers of Aechmea (Fischer & Araujo, 1995; Lenzi et al., 2006), which could have faced barriers to maintenance of gene flow among northern populations of A. kertesziae and population LAG, or even in the establishment of new populations. Using SSRs, moderate genetic structure was observed among populations (Fig. 2A and Table 3), and in contrast to the results from the plastid genome (Figs 1B and S2, Table 3), gene flow between population LAG and the remaining populations of A. kertesziae appears to have been restored (Fig. 2A). However, the pairwise FST estimates indicate that the genetic structuring is both moderate and highly significant, a pattern that was not observed among populations of group I using plastid DNA (Fig. 1, Table 3). These results could indicate recent genetic structuring among all populations, and especially in group I. Because our analysis did not detect any isolation by distance (see Results), these moderate levels of genetic structure are probably not the result of the geographical distance between populations of A. kertesziae. Instead, this recent genetic structuring may be the result of anthropogenic actions causing fragmentation of the restinga habitat, in line with the high human population density along the Brazilian coast. The fragmentation of restinga vegetation might affect the movement of pollinators and dispersers of A. kertesziae, which consequently are not able to maintain gene flow between populations. The main pollinators of the species are bumble bees (M. V. Büttow, unpubl. data), which do not fly beyond 2500 m (Moure & Sakagami, 1962; Hagen, Wikelski & Kissling, 2011). Chiroxiphia spp. are among the taxa identified as seed dispersers for Aechmea (Lenzi et al., 2006) and a study conducted with C. caudata indicates that this species has a flight capacity of c. 130 m in open areas in fragmented landscape (Uezi, Metzger & Vielliard, 2005). Thus, considering the range distance that separates the populations of A. kertesziae (Table S1), the flight capacity of C. caudata, for example, may not ensure the connection of the populations of our model species. Therefore, habitat fragmentation might mean that dispersers and pollinators stay within each of the populations of A. kertesziae, rather than connecting them to ensure effective levels of gene flow. This scenario seems to be particularly important for the FLO population, which is located on an island and was found to be the most differentiated population in the NJ analysis (Fig. 2B). Considering the pairwise estimates of nuclear genetic divergence (FST) and plastid DNA (ΦST), higher levels of genetic structure were observed in the plastid genome (Table 3), which suggests that gene flow in A. kertesziae is more effective through pollen than seeds. This is a common pattern observed in plants in general (Petit et al., 2005), and for bromeliads (Barbará et al., 2008; Palma-Silva et al., 2009, 2011; Paggi et al., 2010), including other Aechmea spp. (Goetze et al., 2016b). Moreover, the variance observed in the levels of genetic structure with nuclear (biparentally) and plastid DNA (maternally) inherited markers can also be attributed to differences in effective population sizes, as plastid DNA is more strongly affected by demographic processes and genetic drift (Ennos, 1994; Petit & Excoffier, 2009), which can increase genetic divergence. High genetic diversity and self-incompatibility in Aechmea kertesziae Our results revealed that A. kertesziae has high genetic diversity (Table 3). SSR-derived levels of diversity are higher in A. kertesziae than in other species from the same subgenus (Goetze et al., 2013, 2015, 2016b), and diversity indices are similar to those of A. nudicaulis, another restinga species (Loh et al., 2015). Its levels of genetic diversity are also higher than those of species from other genera of Bromeliaceae (Zanella et al., 2012b; Lavor et al., 2014; Soares et al., in press). When the plastid genome is considered, A. kertesziae has a much higher diversity than A. calyculata (E.Morren) Baker (subgenus Ortgiesia). Using the same two non-coding plastid regions used here, the latter was found to have only five haplotypes (Goetze et al., 2016b), compared to the 14 for A. kertesziae. The haplotype diversity found in A. kertesziae was similar to the levels observed in other bromeliad species, such as Vriesea carinata Wawra and V. incurvata Gaudich. (Zanella et al., 2016). The difference is that the latter two species have a wide distribution, ranging from 19° to 29°S, and 22° to 29°S, respectively, in contrast to A. kertesziae, which is a restricted species (Reitz, 1983; Goetze et al., 2016a;,Zanella et al., 2016). The high levels of genetic diversity observed in A. kertesziae could be explained, at least in part, by its breeding system. As shown by our hand pollination experiments, A. kertesziae is an obligate outcrosser (Table 4). It is well documented that outcrossing species possess higher levels of genetic diversity than selfers (Hamrick & Godt, 1996; Nybom, 2004; Glémin et al., 2006). Nevertheless, the congeneric outcrossers A. caudata Lindm. and A. winkleri Reitz (Kamke et al., 2011; M. V. Büttow, unpubl. data), both from subgenus Ortgiesia, show lower genetic diversity than A. kertesziae (Goetze et al., 2013, 2015). This indicates that other factors may help to explain the diversity found in A. kertesziae, including long-term population persistence, constant population sizes and clonal reproduction. We found private haplotypes and SSR alleles in all populations (Table 2), suggesting long-term persistence of A. kertesziae at all localities sampled. Ancestral populations are often assumed to possess higher genetic diversity. Moreover, no signs of a bottleneck were detected in A. kertesziae (see Results), reflecting constant population sizes, thus slowing down the effects of genetic drift on decreasing genetic diversity levels (Bennett & Provan, 2008). Similar results were found for the restinga orchid E. fulgens, in which populations that did not show signs of a decrease in size also presented higher genetic diversity (Pinheiro et al., 2011). Aechmea kertesziae is able to reproduce clonally (M. V. Büttow, unpubl. data), an additional factor which might help explain the high levels of genetic diversity observed. The long life span of clonal plants promotes the overlapping of many generations (multiple copies of the same genotype), thus putting a brake on the erosion of genetic diversity through genetic drift (Orive, 1993; Young, Boyle & Brown, 1996). The maintenance of different genotypes through clonal reproduction has been suggested to lead to increased levels of genetic variation in other clonal bromeliads (Izquierdo & Piñero, 2000; Zanella et al., 2011; Ribeiro et al., 2013; Goetze et al., 2015; Loh et al., 2015). Therefore, the high levels of genetic diversity found in A. kertesziae may be caused by a combination of the long-term persistence of the species, constant population sizes, obligate outcrossing breeding system and clonal reproduction. The highest levels of genetic diversity across the entire range of A. kertesziae were found in the LAG population, for both SSR and plastid DNA markers. This population also had the lowest inbreeding coefficient (FIS), although all populations deviated from HWE (Table 3). Since A. kertesziae is an outcrosser (Table 4), other factors besides its breeding system may cause the excess of homozygotes found in all the populations investigated. Genetic structuring is one of the possible explanations for these results, which with genetic drift may promote the fixation of some alleles and the loss of others, increasing the frequencies of homozygotes. In addition, the excess of homozygotes in bromeliads is frequent, and was therefore postulated as a general pattern in Bromeliaceae by Lavor et al. (2014). The lineage that gave origin to the individuals of the LAG population is highly differentiated and was geographically isolated for a long time (Fig. 3). Individuals from this population are not morphologically different from the others, indicating an absence of phenotypic divergence associated with genetic isolation and unique adaptations (Moritz, 2002, and references therein). However, given the distinctiveness of this population, more studies should be carried out to better understand its patterns of genetic diversity. Conservation remarks Aechmea kertesziae is a characteristic element of the restinga vegetation in southern Brazil, especially on the coast of Santa Catarina state, and provides valuable storage of water in this dry environment. Although a high degree of genetic diversity was observed, we also found moderate levels of genetic structure. Since A. kertesziae preferentially inhabits the beach, an area greatly impacted by tourism, this genetic structure is probably the result of anthropogenic actions. Effective in situ conservation strategies should prioritize the enabling of gene flow between populations, especially because A. kertesziae is self-incompatible. Currently, only the FLO population is located within a conservation unit. However, given the high levels of genetic diversity and distinctiveness found for the LAG population, it should also be protected. Ex situ conservation actions might include the creation of a germplasm bank, based on collection of seeds and plants for all known populations. The recent inclusion of A. kertesziae in the official Brazilian list of threatened flora (MMA, 2014) holds hope for the development of conservation strategies targeted at this species. CONCLUSIONS Our study has revealed that A. kertesziae maintained constant population sizes during the climatic oscillations of the Pleistocene. However, a deep phylogeographical break was detected across the small geographical range of the species, which could not be linked to a geographical barrier. Our results also highlighted that the life history traits of A. kertesziae help maintain its high genetic diversity, despite the identification of moderate levels of genetic structure using SSR markers. Hence, the fragmentation and loss of habitats along the Brazilian coast may represent a threat to this species and to other species of the restinga vegetation. ACKNOWLEDGEMENTS We thank Christian R. Rohr, Rafael V. B. Moreira and Silvâneo for their help with sampling. We thank Andreia C. Turchetto-Zolet and Nelson J. R. Fagundes for their valuable suggestions on an earlier version of the manuscript. We are grateful to Dr Clarisse Palma-Silva and three anonymous reviewers for valuable comments and suggestions, which improved the manuscript. Finally, we thank IBAMA (Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais Renováveis) for processing of collection permits. This work was supported by Conselho Nacional de Desenvolvimento Científico e Tecnológico - CNPq (479413/2011–8); Fundação de Amparo à Pesquisa do Estado do Rio Grande do Sul - FAPERGS (10/0198-0 and 06/2010 – 1015348); and Programa de Pós-Graduação em Genética e Biologia Molecular – PPGBM-UFRGS. SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article at the publisher’s web-site: Figure S1. Magnitude of ΔK from Structure analysis of K (mean ± SD across ten replicates), calculated by following the ΔK method proposed by Evanno et al. (2005), for Aechmea kertesziae microsatellite data. Figure S2. Population genetic structure based on plastid DNA. Bayesian admixture proportions inferred with BAPS for individuals of Aechmea kertesziae for the K = 4 groups model. Figure S3. Bayesian skyline plot showing the fluctuations in effective population size over time. The dark line indicates the median estimate and the area between blue lines the 95% confidence interval. (A) For all individuals of Aechmea kertesziae. (B) For individuals of group I. (C) For individuals of group II. The time scale is in millions of years ago (Mya). Table S1. Distance (km) separating the populations of Aechmea kertesziae. Table S2. Characterization of nine microsatellite loci in four populations of Aechmea kertesziae. Table S3. Neutrality tests (D, FS) for each genetic group recovered by chloroplast DNA analysis and for all individuals of Aechmea kertesziae. REFERENCES Avise JC. 2009. Phylogeography: retrospect and prospect. Journal of Biogeography 36: 3– 15. Google Scholar CrossRef Search ADS Bandelt H-J, Forster P, Röhl A. 1999. Median-joining networks for inferring intraspecific phylogenies. Molecular Biology and Evolution 16: 37– 48. Google Scholar CrossRef Search ADS PubMed Barbará T, Lexer C, Martinelli G, Mayo S, Fay MF, Heuertz M. 2008. Within-population spatial genetic structure in four naturally fragmented species of a neotropical inselberg radiation, Alcantarea imperialis, A. geniculata, A. glaziouana and A. regina (Bromeliaceae). Heredity 101: 285– 296. Google Scholar CrossRef Search ADS PubMed Behling H, Negrelle RRB. 2001. Tropical rain forest and climate dynamics of the Atlantic lowland, southern Brazil, during the Late Quaternary. Quaternary Research 56: 383– 389. Google Scholar CrossRef Search ADS Behling H. 2002. South and southeast Brazilian grasslands during Late Quaternary times: a synthesis. Palaeogeography, Palaeoclimatology, Palaeoecology 177: 19– 27. Google Scholar CrossRef Search ADS Bennett KD, Provan J. 2008. What do we mean by ‘refugia’? Quaternary Science Reviews 27: 2449– 2455. Google Scholar CrossRef Search ADS Benzing DH. 2000. Bromeliaceae: profile of an adaptive radiation . New York: Cambridge University Press. Google Scholar CrossRef Search ADS Bowcock AM, Ruíz-Linares A, Tomfohrde J, Minch E, Kidd JR, Cavalli-Sforza LL. 1994. High resolution human evolutionary trees with polymorphic microsatellites. Nature 368: 455– 457. Google Scholar CrossRef Search ADS PubMed Brunes TO, Thomé MTC, Alexandrino J, Haddad CFB, Sequeira F. 2015. Ancient divergence and recent population expansion in a leaf frog endemic to the southern Brazilian Atlantic forest. Organisms Diversity & Evolution 15: 695– 710. Google Scholar CrossRef Search ADS Büttow MV. 2012. Estudo do sucesso reprodutivo, dos padrões de cruzamento e do fluxo de pólen em Aechmea winkleri, uma espécie endêmica do sul do Brasil . Unpublished D. Phil. Thesis, Universidade Federal do Rio Grande do Sul. Cardoso DC, Cristiano MP, Tavares MG, Schubart CD, Heinze J. 2015. Phylogeography of the sand dune ant Mycetophylax simplex along the Brazilian Atlantic Forest coast: remarkably low mtDNA diversity and shallow population structure. BMC Evolutionary Biology 15: 106. Google Scholar CrossRef Search ADS PubMed Cogliatti-Carvalho L, Rocha-Pessôa TC, Nunes-Freitas AF, Rocha CFD. 2010. Water volume stored in bromeliad tanks in Brazilian restinga habitats. Acta Botanica Brasilica 24: 84– 95. Google Scholar CrossRef Search ADS Corander J, Marttinen P, Sirén J, Tang J. 2008. Enhanced Bayesian modeling in BAPS software for learning genetic structures of populations. BMC Bioinformatics 9: 539. Google Scholar CrossRef Search ADS PubMed Cosendey BN, Rocha CFD, Menezes VA. 2016. Population density and conservation status of the teiid lizard Cnemidophorus littoralis, an endangered species endemic to the sandy coastal plains (restinga habitats) of Rio de Janeiro state, Brazil. Journal of Coastal Conservation 20: 97– 106. Google Scholar CrossRef Search ADS Dieringer D, Schlötterer C. 2003. Microsatellite analyser (MSA): a platform independent analysis tool for large microsatellite data sets. Molecular Ecology Notes 3: 167– 169. Google Scholar CrossRef Search ADS Doyle JJ, Doyle JL. 1990. Isolation of plant DNA from fresh tissue. Focus 12: 13– 15. Drummond AJ, Rambaut A. 2007. BEAST: Bayesian evolutionary analysis by sampling trees. BMC Evolutionary Biology 7: 214. Google Scholar CrossRef Search ADS PubMed Drummond AJ, Rambaut A, Shapiro B, Pybus OG. 2005. Bayesian coalescent inference of past population dynamics from molecular sequences. Molecular Biology and Evolution 22: 1185– 1192. Google Scholar CrossRef Search ADS PubMed Drummond AJ, Suchard MA, Xie D, Rambaut A. 2012. Bayesian phylogenetics with BEAUti and the BEAST 1.7. Molecular Biology and Evolution 29: 1969– 1973. Google Scholar CrossRef Search ADS PubMed Earl DA, von Holdt BM. 2012. Structure Harvester: a website and program for visualizing Structure output and implementing the Evanno method. Conservation Genetics Resources 4: 359– 361. Google Scholar CrossRef Search ADS Edgar RC. 2004. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Research 32: 1792– 1797. Google Scholar CrossRef Search ADS PubMed Ennos RA. 1994. Estimating the relative rates of pollen and seed migration among plant populations. Heredity 2: 250– 259. Google Scholar CrossRef Search ADS Evanno G, Regnaut S, Goudet J. 2005. Detecting the numbers of clusters of individuals using the software Structure: a simulation study. Molecular Ecology 14: 2611– 2620. Google Scholar CrossRef Search ADS PubMed Excoffier L, Lischer HEL. 2010. Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows. Molecular Ecology Resources 10: 564– 567. Google Scholar CrossRef Search ADS PubMed Excoffier L, Smouse PE, Quattro JM. 1992. Analysis of molecular variance inferred from metric distances among DNA haplotypes – application to human mitochondrial DNA restriction data. Genetics 131: 479– 491. Google Scholar PubMed Fahrig L. 2003. Effects of habitat fragmentation on biodiversity. Annual Review of Ecology, Evolution, and Systematics 34: 487– 515. Google Scholar CrossRef Search ADS Falkenberg DB. 1999. Aspects of the flora and secondary vegetation in the restinga from Santa Catarina State, South Brazil. Insula 28: 1– 30. Federman S, Hyseni C, Clement W, Oatham MP, Caccone A. 2014. Habitat fragmentation and the genetic structure of the Amazonian palm Mauritia flexuosa L.f. (Arecaceae) on the island of Trinidad. Conservation Genetics 15: 355– 362. Google Scholar CrossRef Search ADS Felsenstein J. 2005. PHYLIP (phylogeny inference package) version 3.6. Distributed by the author. Seattle: Department of Genome Sciences, University of Washington. Available at: http://evolution.genetics.washington.edu/phylip.html. Fischer EA, Araujo AC. 1995. Spatial organization of a bromeliad community in the Atlantic rainforest, south- eastern Brazil. Journal of Tropical Ecology 11: 559– 567. Google Scholar CrossRef Search ADS Fu YX. 1997. Statistical tests of neutrality of mutations against population growth, hitchhiking and background selection. Genetics 147: 915– 925. Google Scholar PubMed Glémin S, Bazin E, Charlesworth C. 2006. Impact of mating systems on patterns of sequence polymorphism in flowering plants. Proceedings of the Royal Society B 273: 3011– 3019. Google Scholar CrossRef Search ADS PubMed Goetze M, Büttow MV, Zanella CM, Paggi GM, Bruxel M, Pinheiro FP, Sampaio JAT, Palma-Silva C, Witt FC, Bered F. 2015. Genetic variation in Aechmea winkleri, a bromeliad from an inland Atlantic rainforest fragment in Southern Brazil. Biochemical Systematics and Ecology 58: 204– 210. Google Scholar CrossRef Search ADS Goetze M, Louzada RB, Wanderley MGL, Souza LM, Bered F, Palma-Silva C. 2013. Development of microsatellite markers for genetic diversity analysis of Aechmea caudata (Bromeliaceae) and cross-species amplification in other bromeliads. Biochemical Systematics and Ecology 48: 194– 198. Google Scholar CrossRef Search ADS Goetze M, Palma-Silva C, Zanella CM, Bered F. 2016b. East-to-west genetic structure in populations of Aechmea calyculata (Bromeliaceae) from the southern Atlantic rainforest of Brazil. Botanical Journal of the Linnean Society 181: 477– 490. Google Scholar CrossRef Search ADS Goetze M, Schulte K, Palma-Silva C, Zanella CM, Büttow MV, Capra F, Bered F. 2016a. Diversification of Bromelioideae (Bromeliaceae) in the Brazilian Atlantic rainforest: a case study in Aechmea subgenus Ortgiesia. Molecular Phylogenetics and Evolution 98: 346– 357. Google Scholar CrossRef Search ADS Goetze M, Zanella CM, Palma-Silva C, Büttow MV, Bered F. 2017. Incomplete lineage sorting and hybridization in the evolutionary history of closely related, endemic yellow-flowered Aechmea species of the subgenus Ortgiesia (Bromeliaceae). American Journal of Botany 104: 1073– 1087. Google Scholar CrossRef Search ADS PubMed Goudet J. 1995. Fstat (version 1.2): a computer program to calculate F-statistics. Journal of Heredity 86: 485– 486. Google Scholar CrossRef Search ADS Hagen M, Wikelski M, Kissling WD. 2011. Space use of bumble bees (Bombus spp.) revealed by radio-tracking. PLoS One 6: e19997. Google Scholar CrossRef Search ADS PubMed Hamrick JL. 1982. Plant population genetics and evolution. American Journal of Botany 69: 1685– 1693. Google Scholar CrossRef Search ADS Hamrick JL, Godt MJW. 1996. Effects of life history traits on genetic diversity in plant species. Philosophical Transactions of the Royal Society B 351: 1291– 1298. Google Scholar CrossRef Search ADS Hedrick P. 2005. A standardized genetic differentiation measure. Evolution 59: 1633– 1638. Google Scholar CrossRef Search ADS PubMed Holsinger KE. 2000. Reproductive systems and evolution in vascular plants. Proceedings of the National Academy of Sciences of the United States of America 97: 7037– 7042. Google Scholar CrossRef Search ADS PubMed Izquierdo LY, Piñero D. 2000. High genetic diversity in the only known population of Aechmea tuitensis (Bromeliaceae). Australian Journal of Botany 48: 645– 650. Google Scholar CrossRef Search ADS Kamke R, Schmid S, Zillikens A, Lopes BC, Steiner J. 2011. The importance of bees as pollinators in the short corolla bromeliad Aechmea caudata in southern Brazil. Flora 206: 749– 756. Google Scholar CrossRef Search ADS Lavor P, van der Berg C, Jacobi CM, Carmo FF, Versieux LM. 2014. Population genetics of the endemic and endangered Vriesea minarum (Bromeliaceae) in the Iron Quadrangle, Espinhaço Range, Brazil. American Journal of Botany 7: 1167– 1175. Google Scholar CrossRef Search ADS Lenzi M, Matos JZ, Orth AI. 2006. Variação morfológica e reprodutiva de Aechmea lindenii (E. Morren) Baker var: lindenii (Bromeliaceae). Acta Botanica Brasilica 20: 487– 500. Google Scholar CrossRef Search ADS Librado P, Rozas J. 2009. DnaSP v5: a software for comprehensive analysis of DNA polymorphism data. Bioinformatics 25: 1451– 1452. Google Scholar CrossRef Search ADS PubMed Loh R, Scarano FR, Alves-Ferreira M, Salgueiro F. 2015. Clonality strongly affects the spatial genetic structure of the nurse species Aechmea nudicaulis (L.) Griseb. (Bromeliaceae). Botanical Journal of the Linnean Society 178: 329– 341. Google Scholar CrossRef Search ADS Lopes CM, Ximenes SSF, Gava A, Freitas TRO. 2013. The role of chromosomal rearrangements and geographical barriers in the divergence of lineages in a South American subterranean rodent (Rodentia: Ctenomyidae: Ctenomys minutus). Heredity 111: 293– 305. Google Scholar CrossRef Search ADS PubMed Luther HE. 2012. An alphabetical list of bromeliad binomials , 13th edn. Holst BK, Rabinowitz L, eds. Sarasota: Marie Selby Botanical Gardens/Bromeliad Society International. Marques MCM, Silva SM, Liebsch D. 2015. Coastal plain forests in southern and southeastern Brazil: ecological drivers, floristic patterns and conservation status. Brazilian Journal of Botany 38: 1– 18. Google Scholar CrossRef Search ADS MMA – Ministério do Meio Ambiente. 2014. Normative Statement No. 443 , 17 December 2014. Moritz C. 2002. Strategies to protect biological diversity and the evolutionary processes that sustain it. Systematic Biology 51: 238– 254. Google Scholar CrossRef Search ADS PubMed Moure JS, Sakagami SF. 1962. As mamangabas sociais do Brasil (Bombus Latreille) (Hymenoptera, Apoidea). Studia Entomologica 5: 65– 194. Nei M. 1987. Molecular evolutionary genetics . New York: Columbia University Press. Nybom H. 2004. Comparison of different nuclear DNA markers for estimating intraspecific genetic diversity in plants. Molecular Ecology 13: 1143– 1155. Google Scholar CrossRef Search ADS PubMed Nybom H, Weising K, Rotter B. 2014. DNA fingerprinting in botany: past, present, future. Investigative Genetics 5: 1– 35. Google Scholar CrossRef Search ADS PubMed Orive ME. 1993. Effective population size in organisms with complex life-histories. Theoretical Population Biology 44: 316– 340. Google Scholar CrossRef Search ADS PubMed Paggi GM, Sampaio JAT, Bruxel M, Zanella CM, Goetze M, Büttow MV, Palma-Silva C, Bered F. 2010. Seed dispersal and population structure in Vriesea gigantea, a bromeliad from the Brazilian Atlantic Rainforest. Botanical Journal of the Linnean Society 164: 317– 325. Google Scholar CrossRef Search ADS Palma-Silva C, Lexer C, Paggi GM, Barbará T, Bered F, Bodanese-Zanettini MH. 2009. Range-wide patterns of nuclear and cloroplast DNA diversity in Vriesea gigantea (Bromeliaceae), a neotropical forest species. Heredity 103: 503– 512. Google Scholar CrossRef Search ADS PubMed Palma-Silva C, Wendt T, Pinheiro F, Barbará T, Fay MF, Cozzolino S, Lexer C. 2011. Sympatric bromeliad species (Pitcairnia spp.) facilitate tests of mechanisms involved in species cohesion and reproductive isolation in Neotropical inselbergs. Molecular Ecology 20: 3185– 3201. Google Scholar CrossRef Search ADS PubMed Petit RJ, Duminil J, Fineschi S, Hampe A, Salvivi D, Vendramin GG. 2005. Comparative organization of chloroplast, mitochondrial and nuclear diversity in plant populations. Molecular Ecology 14: 689– 701. Google Scholar CrossRef Search ADS PubMed Petit RJ, Excoffier L. 2009. Gene ﬂow and species delimitation. Trends in Ecology & Evolution 24: 386– 393. Google Scholar CrossRef Search ADS PubMed Pinheiro F, Barros F, Palma-Silva C, Fay MF, Lexer C, Cozzolino S. 2011. Phylogeography and genetic differentiation along the distributional range of the orchid Epidendrum fulgens: a Neotropical coastal species not restricted to glacial refugia. Journal of Biogeography 38: 1923– 1935. Google Scholar CrossRef Search ADS Piry S, Luikart G, Cornuet JM. 1999. BOTTLENECK: a computer program for detecting recent reductions in the effective population size using allele frequency data. Journal of Heredity 90: 502– 503. Google Scholar CrossRef Search ADS Pritchard JK, Stephens M, Donnelly P. 2000. Inference of population structure using multilocus genotype data. Genetics 155: 945– 959. Google Scholar PubMed Provan J, Powell W, Hollingsworth PM. 2001. Chloroplast microsatellites: new tools for studies in plants ecology and evolution. Trends in Ecology and Evolution 16: 142– 147. Google Scholar CrossRef Search ADS PubMed Rambaut A. 2008. FigTree v1.4: tree figure drawing tool. Available at: http://tree.bio.ed.ac.uk/software/figtree/. Rambaut A, Suchard MA, Xie D, Drummond AJ. 2013. Tracer v1.5. Available at: http://beast.bio.ed.ac.uk/Tracer. Rambo B. 1950. A Porta de Torres. Anais Botânicos do Herbário Barbosa Rodrigues 2: 125– 136. Raymond M, Rousset F. 1995. Genepop (version 1.2): population genetics software for exact tests and ecumenicism. Journal of Heredity 86: 248– 249. Google Scholar CrossRef Search ADS Reitz R. 1983. Bromeliáceas e a malária - bromélia endêmica . Itajaí: Flora Ilustrada Catarinense Herbário Barbosa Rodrigues. Ribeiro MC, Metzger JP, Martensen AC, Ponzoni FJ, Hirota MM. 2009. The Brazilian Atlantic Forest: how much is left, and how is the remaining forest distributed? Implications for conservation. Biological Conservation 142: 1141– 1153. Google Scholar CrossRef Search ADS Ribeiro PCC, Pinheiro LC, Domingues R, Forzza RC, Machado MA, Viccini LF. 2013. Genetic diversity of Vriesea cacuminis (Bromeliaceae): an endangered and endemic Brazilian species. Genetics and Molecular Research 12: 1934– 1943. Google Scholar CrossRef Search ADS PubMed Rocha CFD, Bergallo HG, Van Sluys M, Alves MAS, Jamel CE. 2007. The remnants of restinga habitats in the Brazilian Atlantic Forest of Rio de Janeiro state, Brazil: habitat loss and risk of disappearance. Brazilian Journal of Biology 67: 263– 273. Google Scholar CrossRef Search ADS Salgueiro F, Felix D, Caldas JF, Margis-Pinheiro M, Margis R. 2004. Even population differentiation for maternal and biparental gene markers in Eugenia uniflora, a widely distributed species from the Brazilian coastal Atlantic rain forest. Diversity and Distributions 10: 201– 210. Google Scholar CrossRef Search ADS Scotti-Saintagne C, Dick CW, Caron H, Vendramin GG, Troispoux V, Sire P, Casalis M, Buonamici A, Valencia R, Lemes MR, Gribel R, Scotti I. 2013. Amazon diversification and cross Andean dispersal of the widespread Neotropical tree species Jacaranda copaia (Bignoniaceae). Journal of Biogeography 40: 707– 719. Google Scholar CrossRef Search ADS Shaw J, Lickey EB, Schilling EE, Small RL. 2007. Comparison of whole chloroplast genome sequences to choose noncoding regions for phylogenetic studies in angiosperms: the tortoise and the hare III. American Journal of Botany 94: 275– 288. Google Scholar CrossRef Search ADS PubMed Smith LB, Downs RJ. 1979. Flora Neotropica, Monograph No. 14, Part 3, Bromelioideae (Bromeliaceae) . New York: Hafner Press. Soares LE, Goetze M, Zanella CM, Bered F. in press. Genetic diversity and population structure of Vriesea reitzii (Bromeliaceae), a species from the southern Brazilian Highlands. Genetics and Molecular Biology . Tajima F. 1989. Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123: 585– 595. Google Scholar PubMed Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S. 2011. MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Molecular Biology and Evolution 28: 2731– 2739. Google Scholar CrossRef Search ADS PubMed Turchetto-Zolet AC, Pinheiro F, Salgueiro F, Palma-Silva C. 2013. Phylogeographical patterns shed light on evolutionary process in South America. Molecular Ecology 22: 1193– 1213. Google Scholar CrossRef Search ADS PubMed Turchetto-Zolet AC, Salgueiro F, Turchetto C, Cruz F, Veto NM, Barros MJF, Segatto ALA, Freitas LB, Margis R. 2016. Phylogeography and ecological niche modelling in Eugenia uniflora (Myrtaceae) suggest distinct vegetational responses to climate change between the southern and the northern Atlantic Forest. Botanical Journal of the Linnean Society 182: 670– 688. Google Scholar CrossRef Search ADS Uezi A, Metzger JP, Vielliard JME. 2005. Effects of structural and functional connectivity and patch size on the abundance of seven Atlantic Forest bird species. Biological Conservation 123: 507– 519. Google Scholar CrossRef Search ADS Van Oosterhout C, Hutchinson WF, Wills DPM, Shipley P. 2004. Micro-Checker: software for identifying and correcting genotyping errors in microsatellite data. Molecular Ecology Notes 4: 535– 538. Google Scholar CrossRef Search ADS Wagner ND, Wöhrmann T, Öder V, Burmeister A, Weising K. 2015. Reproduction biology and chloroplast inheritance in Bromeliaceae: a case study in Fosterella (Pitcairnioideae). Plant Systematics and Evolution 301: 2231– 2246. Google Scholar CrossRef Search ADS Weir BS, Cockerham CC. 1984. Estimating F-statistics for the analysis of population structure. Evolution 38: 1358– 1370. Google Scholar PubMed Werneck FP. 2011. The diversification of eastern South American open vegetation biomes: Historical biogeography and perspectives. Quaternary Science Reviews 30: 1630– 1648. Google Scholar CrossRef Search ADS Wörhmann T, Weising K. 2011. In silico mining for simple sequence repeat loci in a pineapple expressed sequence tag database and cross-species ampliﬁcation of EST-SSR markers across Bromeliaceae. Theoretical and Applied Genetics 123: 635– 647. Google Scholar CrossRef Search ADS PubMed Wright S. 1965. The interpretation of population structure by F-statistics with special regards to system of mating. Evolution 19: 395– 420. Google Scholar CrossRef Search ADS Young A, Boyle T, Brown T. 1996. The population genetic consequences of habitat fragmentation for plants. Trends in Ecology and Evolution 11: 413– 418. Google Scholar CrossRef Search ADS PubMed Zanella CM, Bruxel M, Paggi GM, Goetze M, Büttow MV, Cidade FW, Bered F. 2011. Genetic structure and phenotypic variation in wild populations of the medicinal tetraploid species Bromelia antiacantha (Bromeliaceae). American Journal of Botany 98: 1511– 1519. Google Scholar CrossRef Search ADS PubMed Zanella CM, Janke A, Paggi GM, Goetze M, Reis MS, Bered F. 2012a. Microsatellites in the endangered species Dyckia distachya (Bromeliaceae) and cross-amplification in other bromeliads. International Journal of Molecular Science 13: 15859– 15866. Google Scholar CrossRef Search ADS Zanella CM, Janke A, Palma-Silva C, Katchuck-Santos E, Pinheiro FG, Paggi GM, Soares LES, Goetze M, Büttow MV, Bered F. 2012b. Genetics, evolution, and conservation of Bromeliaceae. Genetics and Molecular Biology 35: 1020– 1026. Google Scholar CrossRef Search ADS Zanella CM, Palma-Silva C, Goetze M, Bered F. 2016. Hybridization between two sister species of Bromeliaceae: Vriesea carinata and V. incurvata. Botanical Journal of the Linnean Society 181: 491– 504. Google Scholar CrossRef Search ADS © 2018 The Linnean Society of London, Botanical Journal of the Linnean Society
Botanical Journal of the Linnean Society – Oxford University Press
Published: Mar 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.
All for just $49/month
Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.
Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.
It’s easy to organize your research with our built-in tools.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera