Skip to main content
NCBI home page
Annals of Botany logoLink to Annals of Botany
. 2012 May 28;110(1):11–21. doi: 10.1093/aob/mcs096

Is Drosera meristocaulis a pygmy sundew? Evidence of a long-distance dispersal between Western Australia and northern South America

F Rivadavia 1, V F O de Miranda 2, G Hoogenstrijd 3, F Pinheiro 4, G Heubl 5, A Fleischmann 5,*
PMCID: PMC3380593  PMID: 22641141

Abstract

Background and aims

South America and Oceania possess numerous floristic similarities, often confirmed by morphological and molecular data. The carnivorous Drosera meristocaulis (Droseraceae), endemic to the Neblina highlands of northern South America, was known to share morphological characters with the pygmy sundews of Drosera sect. Bryastrum, which are endemic to Australia and New Zealand. The inclusion of D. meristocaulis in a molecular phylogenetic analysis may clarify its systematic position and offer an opportunity to investigate character evolution in Droseraceae and phylogeographic patterns between South America and Oceania.

Methods Drosera meristocaulis

was included in a molecular phylogenetic analysis of Droseraceae, using nuclear internal transcribed spacer (ITS) and plastid rbcL and rps16 sequence data. Pollen of D. meristocaulis was studied using light microscopy and scanning electron microscopy techniques, and the karyotype was inferred from root tip meristem.

Key Results

The phylogenetic inferences (maximum parsimony, maximum likelihood and Bayesian approaches) substantiate with high statistical support the inclusion of sect. Meristocaulis and its single species, D. meristocaulis, within the Australian Drosera clade, sister to a group comprising species of sect. Bryastrum. A chromosome number of 2n = approx. 32–36 supports the phylogenetic position within the Australian clade. The undivided styles, conspicuous large setuous stipules, a cryptocotylar (hypogaeous) germination pattern and pollen tetrads with aperture of intermediate type 7–8 are key morphological traits shared between D. meristocaulis and pygmy sundews of sect. Bryastrum from Australia and New Zealand.

Conclusions

The multidisciplinary approach adopted in this study (using morphological, palynological, cytotaxonomic and molecular phylogenetic data) enabled us to elucidate the relationships of the thus far unplaced taxon D. meristocaulis. Long-distance dispersal between southwestern Oceania and northern South America is the most likely scenario to explain the phylogeographic pattern revealed.

Keywords: Droseraceae, Drosera sect. Bryastrum, America–Oceania disjunction, carnivorous plants, ITS, rbcL, rps16, phylogeny, pollen morphology, germination pattern, chromosome numbers

INTRODUCTION

The carnivorous plants known as sundews of the genus Drosera (Droseraceae) comprise nearly 200 species spread worldwide, mostly in the Southern Hemisphere and especially in southwestern Australia (Diels, 1906; Schlauer, 2007; McPherson, 2010). Species of the most distinctive groups of Drosera, known as the pygmy sundews – because of their usually diminutive size – are all endemic to the southwestern tip of Western Australia, except for D. pygmaea which is also found in southeastern Australia and New Zealand (Lowrie, 1989).

The pygmy sundews make up sect. Bryastrum (following the sectional classification of Seine and Barthlott, 1994), consisting of approx. 50 species (Lowrie, 1989, 1998; Lowrie and Carlquist, 1992; Lowrie and Conran, 2007; Mann, 2007), and are characterized not only by their relatively diminutive size, but also by large translucent papery stipules which are arranged as a dense stipule bud in the centre of the rosette, three to five undivided styles, long fibrous roots and their unique capability to reproduce vegetatively by small leaf-derived propagules known as gemmae. The gemmae are modified leaves, which are chlorophyllous and rich in starch (Goebel, 1908; Karlsson and Pate, 1992). Recent molecular phylogenetic data (Rivadavia et al., 2003) showed the pygmy sundews to be a well supported monophyletic group, which is part of a large clade containing mostly Australian species, and sister to a clade including mostly taxa native to the New World and southern Africa.

Botanical expeditions in the 1950s to the isolated highlands known as the Neblina massif on the Brazilian–Venezuelan border in the Amazonas lowlands of northern South America resulted in the description of numerous endemic species, including Drosera meristocaulis (Maguire and Wurdack, 1957) (Fig. 1). Because this species has only three undivided styles, a unique character among New World Drosera taxa, a monotypic sect. Meristocaulis was created for this taxon (Maguire and Wurdack, 1957; Seine and Barthlott, 1994), which was raised to subgeneric level by Schlauer (1996). Other conspicuous characters of this taxon include long stems up to 40 cm in length and nearly sessile flowers nested among the leaves and stipules (Fig. 1). Nonetheless, D. meristocaulis also presents characteristics reminiscent of pygmy sundews, such as diminutive leaves, large translucent papery stipules and long fibrous roots. The extreme isolation of the remote Neblina massif kept D. meristocaulis from being studied in more depth, thus heightening scientists' curiosity about the relationship of this species to other members of Drosera. Maguire and Wurdack (1957) were well aware of the similarities of D. meristocaulis to the pygmy Drosera of sect. Bryastrum from Australia (Fig. 1). Due to the undivided styles, however, they supposed a close relationship to the single South American member of sect. Thelocalyx, D. sessilifolia. Duno de Stefano (1995) studied the pollen morphology of D. meristocaulis for the first time and proposed a close relationship of sect. Meristocaulis with sect. Drosera.

Fig. 1.

Fig. 1.

Open in a new tab

Drosera meristocaulis (A, C) from the Neblina massif in the Amazon and the Western Australian pygmy sundew Drosera gibsonii (B, D) show a remarkable similarity in overall habit and in flower morphology.

The Neblina massif is a huge sandstone formation reaching nearly 3000 m above sea level and is covered in part by low vegetation (‘Neblinaria scrub’; Brewer-Carías, 1988; Huber, 1995) composed of species not found in the hot surrounding lowlands. Several expeditions to Neblina and other mountains of the Guayana Highlands (known as tepuis) documented an impressive number of endemic taxa and contributed to the idea of a diverse and unique flora with a high degree of endemism (Steyermark, 1979). In an attempt to explain this unique flora, the idea of ‘Lost Worlds’ was created, postulating that the origin of local biota was relictual as a result of a long history of evolution in isolation on the mountain summits (Rull, 2004). On the other hand, the ‘Vertical Displacement’ hypothesis assumes the lack of total geographical isolation among tepui summits, with extensive valleys and gentle slopes possibly being important paths connecting highlands with lowlands, thus providing hypothetical migrational pathways (Huber, 1988; Rull, 2004).

Long-distance dispersal (LDD) was accepted and rejected many times as a good theory to explain floristic similarities among continents since Darwin's experiments (1859). Besides the fact that LDD was accepted as a natural process that occurred on recent volcanic islands (Carlquist, 1966, 2010), the plate tectonics theory provided vicariance explanations for many cases of disjunctions (de Queiroz, 2005). Molecular clock techniques have revealed that many plant lineages have a recent origin, with radiation events occurring after continental splits (Givnish and Renner, 2004; Muñoz et al., 2004; Sytsma et al., 2004; Dick et al., 2007). Now many dispersion routes are corroborated by multiple taxa in the Southern Hemisphere (de Queiroz, 2005), and LDD can explain the disjunction patterns of many groups.

In the present study, a multidisciplinary investigation was carried out in order to clarify the phylogenetic position of D. meristocaulis in Droseraceae and to test the hypothesis of a putative common ancestry with species from sect. Bryastrum. The pattern of seed germination, pollen morphology, chromosome counts and a molecular approach based on nuclear and plastid DNA sequences were investigated.

MATERIALS AND METHODS

Seed germination

Seeds of Drosera meristocaulis and D. capillaris were obtained from a commercial carnivorous plant seed source (A. Lowrie, Duncraig, Australia) and were sown on pure peat and on milled long fibre sphagnum in a greenhouse, and kept moist at 20–25 °C.

Chromosome counts

Root tips of greenhouse-grown seedlings were used for karyotype analysis. In addition, in vitro raised plants of D. meristocaulis were obtained from a commercial nursery (bestcarnivorousplants.com). For mitotic chromosome counts, root tips of in vitro and ex vitro plants were collected and pre-treated with 0·002 m 8-hydroxyquinoline for 3 h to achieve mitotic arrest, and then fixed in ethanol:acetic acid (3:1) and stored at 4 °C. Fixed root tips were hydrolysed in 2 m hydrochloric acid at 60 °C for 10 min, and then enzymatically macerated with 5 % cellulase (Roth, Germany) at 37 °C for 20 min. Root tips were rinsed with distilled water, squashed on glass slides and the prepared root tip meristems were orcein stained (Orcein: Roth, Germany). Chromosome counts were made using a light microscope (Leitz, Germany), and slides were documented photographically using a digital camera (Nikon D5000, Germany).

Pollen analysis

Dried anthers were taken from herbarium specimens of D. meristocaulis deposited in SPF (voucher F. Rivadavia et al. 1881). The anthers were soaked in 10 % KOH overnight and then prepared by acetolysis following Erdtman (1960). After a final washing step, the acetolysed pollen grains were stored in acetone for light microscopy (LM) and scanning electron microscopy (SEM) analysis. Photomicrographs of pollen grains in LM were obtained with a video camera (Olympus) connected to a PC. SEM analyses were made using acetolysed pollen grains, which were washed in pure water at several steps to remove residual acetone, and then put on lightstub carbon plates. The samples were gold coated in a vacuum at 36 mA for 2 min using an SCD 050 sputter coater (BAL-TEC, Liechtenstein) and analysed with a 438VP scanning electron microscope (LEO, Germany).

Plant material and DNA extraction

Voucher specimens of D. meristocaulis were deposited at the University of São Paulo Herbarium SPF (F. Rivadavia et al. 1881). DNA from dried leaves was extracted using the cetyltrimethylammonium bromide (CTAB) buffer protocol (Doyle and Doyle, 1987). Genomic DNA of species of sect. Bryastrum and of Drosera glanduligera, Drosera regia and the outgroup taxon Dionaea muscipula (Droseraceae) (see Table 1) was extracted from fresh leaf tissue of greenhouse-grown plants from the private collection of A. Fleischmann, using a NucleoSpin® Plant Kit (Macherey-Nagel, Düren, Germany), following the manufacturer's protocol (Macherey-Nagel, 2007). Voucher specimens are listed in Table 1.

Table 1.

List of the Drosera species and outgroup taxa used for the combined phylogenetic analysis, including voucher data and GenBank accession numbers of the sequence data generated for this study

Species Source Distribution GenBank number rbcL GenBank number ITS GenBank number rps16
D. meristocaulis Neblina, F.Rivadavia et al. 1881 (SPF) Neblina massif, Brazil–Venezuela border JN388035 JN388038 JN388044
D. glanduligera cult. Fleischmann (M; photo voucher) SW Australia AB072511* JN388039 JN388045
D. barbigera cult. Fleischmann (M; photo voucher) SW Australia JQ712489 JQ712490 JQ712488
D. nitidula cult. Fleischmann (M; photo voucher) SW Australia JN388036 JN388040 JN388046
D. scorpioides cult. Fleischmann (M; photo voucher) SW Australia AB072509* JN388041 JN388047
D. occidentalis cult. Fleischmann (M; photo voucher) SW Australia AB072506* JN388042 JN388048
D. paradoxa cult. Fleischmann (M; photo voucher) Northern Australia D. petiolaris: L01913 JN388043 JN388049
D. ordensis cult. Fleischmann (M; photo voucher) Northern Australia JN388037 JN388075 JN388050
D. regia V.F.O. de Miranda 218 (HUMC) South Africa AB072566* JN388077 JN388051
Dionaea muscipula V.F.O. de Miranda 208 (HUMC) SE USA AB072558* JN388078 JN388052
Open in a new tab

Photographic vouchers are given as Supplementary Data, available online.

* Sequences published in Rivadavia et al. (2003).

PCR conditions/DNA amplification and sequencing

Amplification of the plastid molecular marker rbcL was performed using the primers and protocol of Hasebe et al. (1994). The rps16 intron was amplified and sequenced using the primers rpsF and rps2R and the protocol of Oxelman et al. (1997). The nuclear internal transcribed spacer (ITS) region was amplified using the PCR primers Leu1 (Walker and Sytsma, 2007) and ITS4 (White et al., 1990), following the PCR protocol published in White et al. (1990). ITS amplification of D. muscipula and D. regia followed the protocol of Miranda et al. (2010).

PCR-amplified sequences were purified using a GFX™ PCR DNA and Gel Purification Kit (GE Healthcare, USA). Both strands of the spacer region were sequenced by the dideoxy chain terminator method in a thermal cycler (GeneAmp® PCR System 9700, Applied Biosystems, Foster City, CA, USA). The sequencing reactions were performed in a total volume of 10 µL containing 30–50 ng of DNA, 5 µm of each primer, 2 µL of the ABI Prism BigDye Terminator v3·1 cycle sequencing ready reaction kit (Applied Biosystems) and 1 µL of 5× Sequencing Buffer (Applied Biosystems). The thermal cycling parameters were as follows: one cycle of 4 min at 94 °C, 40 cycles at 94 °C for 40 s, 52 °C for 40 s and 72 °C for 1 min. Electrophoresis and fluorescence detection were carried out on an ABI Prism 3100 Genetic Analyzer (Applied Biosystems).

Phylogenetic reconstruction

The sequences were aligned using ClustalW 1·4 (Thompson et al., 1994) followed by manual examination using BioEdit (Hall, 1999). Some of the ITS and rbcL sequences used here were obtained from previous studies (Rivadavia et al., 2003; V. Miranda et al., unpubl. res.) and are available from NCBI GenBank (accession numbers for all nucleotide sequences are listed in Table 1). Indels were treated as missing data. As a strategy of tree rooting, several taxa were initially employed as outgroups, most of them representatives from various families of Caryophyllales known to be closely related to Droseraceae (i.e. Ancistrocladaceae, Dioncophyllaceae and Nepenthaceae). Nevertheless most of these sequences resulted in pairwise similarity <75 %, compared with the sequences of the Drosera ingroup, a scenario that could increase noise in the analyses. Therefore, we chose to employ only the monotypic Dionaea as an outgroup in all the analyses, because of the higher values of pairwise similarities gained. The phylogenetic analyses were performed for each individual matrix (ITS, rps16 and rbcL) and as combined matrices (ITS + rps16 + rbcL). An additional analysis was carried out with the combined ITS + rps16 data set, because of an incongruent position of D. meristocaulis compared with the topology of the rbcL data set. An analysis with a more complete rbcL data set of Drosera spp. was also performed (Table 2; all rbcL sequence data for Drosera from Rivadavia et al., 2003 from GenBank). Further outgroup taxa were added to this rbcL analysis, based on sequences available in GenBank: Armeria bottendorfensis, Limonium sinense (Plumbaginaceae), Drosophyllum lusitanicum (Drosophyllaceae), and Polygonum capitatum and Rheum delavayi (Polygonaceae) (Tables 1 and 2).

Table 2.

List of the Drosera species and outgroup taxa additionally used for the enlarged rbcL data set from GenBank

Species GenBank number
Aldrovanda vesiculosa L. AB072550
Armeria bottendorfensis A.Schulz Z97640
Drosera adelae F.Muell. AY096107
D. alba E.Phillips AB072515
D. aliciae Raym.-Hamet AB072516
D. anglica Huds. AB072517
D. arcturi Hook. AB072512
D. ascendens A.St.-Hil. AB072542
D. brevifolia Pursh AB072519
D. burkeana Planch. AB072520
D. burmannii Vahl L01908
D. caduca Lowrie AB072510
D. capensis L. L01909
D. capillaris Poir. AB072521
D. chrysolepis Taub. AB072522
D. cistiflora L. AB072523
D.collinsiae N.E.Br. in Burtt Davy AB072524
D. cuneifolia L.f. AB072525
D. omissa Diels (as D. ericksoniae N.Marchant) AB072507
D. felix Steyerm. & L.B.Smith AB072527
D. filiformis Raf. L01911
D. gigantea Lindl. L19528
D. graminifolia A.St.-Hil. AB072528
D. graomogolensis T.R.S.Silva AB072529
D. hamiltonii C.R.P.Andrews AB072921
D. hilaris Cham. & Schlechtd. AB072530
D. hirtella A.St.-Hil. AB072531
D. indica L. L19529
D. macrantha Endl. subsp. planchonii N.G.Marchant AB072549
D. longiscapa Debbert (as D. madagascariensis DC) AB072533
D. montana A.St.-Hil. AB072534
D. natalensis Diels AB072537
D. pauciflora Banks ex DC. AB072552
D. peltata Thunb. L01912
D. pygmaea DC. AB072505
D. rotundifolia L. AB072538
D. schwackei (Diels) Rivadavia AB072535
D. sessilifolia A.St.-Hil. AB072551
D. spatulata Labill. L19530
D. stenopetala Hook.f. AB072539
D. stolonifera Endl. L19531
D. trinervia Spreng. AB072548
D. tomentosa A.St.-Hil. AB072536
D. uniflora Willd. AB072540
D. villosa A.St.-Hil. AB072541
Drosophyllum lusitanicum Link L01907
Limonium sinense Kuntze FJ872106
Polygonum capitatum Korth. ex Meisn. HM850243
Rheum delavayi Franch. FJ872104
Open in a new tab

Maximum parsimony

Phylogenetic analysis based on maximum parsimony (MP) of the sequence data was performed using PAUP* version 4b10 (Swofford, 2002). The phylogenetic trees were obtained by heuristic search through random addition with 5000 replications. The branch swapping followed the tree bisection–reconnection (tbr) algorithm. The robustness of the inferred trees was evaluated using decay indices (Bremer, 1988) and bootstrap resampling (Felsenstein, 1985) through 2000 replicates (pseudomatrices) with 40 heuristic search replicates and random taxon addition. Decay indices were calculated using TNT version 1·1 (Goloboff et al., 2008) and only absolute values ≤50 were considered.

Maximum likelihood and Bayesian analyses

The likelihood ratio test as implemented in ModelTest version 3·7 (Posada and Crandall, 1998), with the help of MrMTgui version 1·0 (P. Nuin, GNU General Public License), was employed to determinate the best-fit model of DNA substitution for each data set (individual and combined data sets) under the Akaike information criterion (AIC; Akaike, 1974) to estimate the parameters. We used maximum likelihood (ML) and a Bayesian framework (BA) with Metropolis-coupled Markov chain Monte Carlo (MCMCMC; Geyer, 1991) inference to estimate the phylogenetic hypotheses to each data set. The ML analyses were run in PAUP* version 4b10, using individual models, and estimated parameters to each matrix and clade support were calculated with 2000 replicates (with 40 heuristic search replicates and random addition). MCMCMC analyses were performed in MrBayes version 3·1·2 (Huelsenbeck and Ronquist, 2001; Ronquist and Huelsenbeck, 2003) for each data set with 9 × 106 generations sampled every 100 generations, using the default parameters. For each analysis, four separate runs were carried out starting from random trees. The sample points prior to reaching stationarity were discarded as burn-in. The posterior probabilities (PPs) for each clade obtained from individual analyses were compared for congruence and combined for evaluating a 50 % majority-rule consensus tree.

RESULTS

Germination pattern

Seed germination occurred after approx. 3–4 weeks at 20–25 °C. Drosera meristocaulis exhibits a cryptocotylar (hypogaeous) germination pattern, with the cotyledons remaining in the testa (Fig. 5).

Fig. 5.

Fig. 5.

Open in a new tab

Comparison of germination patterns of two different South American Drosera species. (A) Cryptocotyly in Drosera meristocaulis. (B) Phanerocotyly in D. capillaris. Abbreviations: c, cotelydons (hidden inside the testa in D. meristocaulis); ec, epicotyl; hc, hypocotyl; lr, lateral root; pl, primary leaf; pr, primary root; sc, seed coat (testa). Scale bars = 1 mm.

Pollen morphology

Drosera meristocaulis has pollen tetrads with the intermediate aperture type 7–8, following the terminology of Takahashi and Sohma (1982) (Fig. 6). The size measurements are based on our own LM observations and SEM micrographs, and on Duno de Stefano (1995): tetrahedral or frequently tetragonal tetrad, 90–130 µm in diameter (confirming Duno de Stefano, 1995), exine spicate, pollen inoperculate, aperture: one single central pore per grain (aperture type 7–8), with approx. 5–8 large channel openings with a thick exinous wall surrounding one proximal central pore, radial plaits poorly developed. Single grain 35–43 µm in diameter (45–55 µm by Duno de Stefano, 1995). Channel openings approx. 10 × 5 (–10) μm, standing alternate or opposite to those of adjoining grains. Spines up to 4 µm long, density of the spines 1·0–1·5 µm−2 (confirming Duno de Stefano, 1995), spinules absent (confirming Duno de Stefano, 1995).

Fig. 6.

Fig. 6.

Open in a new tab

Pollen of Drosera meristocaulis. (A) LM photograph, (B) SEM photograph. Abbreviations: co, channel opening; cp, central pore; rp, radial plait. Scale bars = 10 µm.

Chromosome counts

In total, ten meristematic root tips were prepared, and numerous counts were made. However, due to the small chromosome size, and overall small size of the meristematic root cells of D. meristocaulis of about 10 µm in diameter, an evaluation of the exact karyotype was not possible. The chromosome counts for D. meristocaulis revealed numbers of 32, 34 and 36 with equal frequency of occurrence. Therefore, a karyotype of 2n = approx. 32–36 is given for D. meristocaulis here.

Molecular data

All three markers used in this study revealed D. meristocaulis in the Australian Drosera clade (sensu Rivadavia et al., 2003), although the exact phylogenetic position differs between rbcL and the other two markers (Fig. 4; see Supplementary Data for sequence alignments). The plastid marker rbcL shows D. meristocaulis nested within the pygmy Drosera clade, as sister to the two sister pairs D. occidentalis and D. nitidula, and D. barbigera and D. scorpioides (Fig. 4). In both the ITS and rps16 data sets, and the combined phylogenetic reconstruction using all three markers, D. meristocaulis is revealed as sister to the pygmy clade (Fig. 3), and the two are sister to the petiolaris clade (represented by D. paradoxa, D. petiolaris and D. ordensis in the present study). All nodes get high statistical support, both in the trees resulting from each single marker data set (see Fig. 2 for rbcL) and in the combined tree (Fig. 3).

Fig. 4.

Fig. 4.

Open in a new tab

Incongruence of the phylogenetic position of Drosera meristocaulis between ITS and rps16 topology (left) and rbcL (right). Identical phylogenetic positions are indicated by dashed lines; a slash in the dashed line is for taxa equivalents used in the different data sets (see also Table 1).

Fig. 3.

Fig. 3.

Open in a new tab

Single most parsimonious tree (1331 steps) of the combined ITS, rps16 and rbcL data sets. Numbers above branches show MP bootstrap (left), ML bootstrap (middle) and Bayesian posterior probability (right); numbers below branches are decay index values. Branch lengths represent genetic distance based on the scale at the bottom.

Fig. 2.

Fig. 2.

Open in a new tab

Strict consensus from 576 most parsimonious trees (754 steps) of Droseraceae from Bayesian analysis of the rbcL data set. Numbers above branches show bootstrap (left), and Bayesian posterior probability (right) support values; numbers below branches are decay index values. The position of Drosera meristocaulis is highlighted. Taxonomic groups and clades are indicated, following Rivadavia et al. (2003).

DISCUSSION

Recently collected material of D. meristocaulis has revealed surprising new data supporting the placement of this species in the Australian Drosera clade, at the base of sect. Bryastrum. The most outstanding pattern recovered by the joint investigation of different biological aspects (seed germination pattern, pollen morphology, chromosome count and molecular phylogenetics) was the strong affinity of D. meristocaulis for the species from sect. Bryastrum.

Phylogenetic significance of germination pattern

Cryptocotylar germination in Droseraceae was thus far exclusively known from taxa belonging to a phylogenetic clade containing predominantly Australian Drosera spp., including the pygmy sundews of sect. Bryastrum (Conran et al., 1997, 2007). Thus, D. meristocaulis is the only New World Drosera species with this type of germination (Fig. 5), with all other species showing phanerocotyly (see D. capillaris in Fig. 5). Cryptocotylar germination in small-seeded plants like Droseraceae is rare (Clifford, 1984) and is usually associated with fluctuating ecological conditions and therefore interpreted as an adaptation to long-term seed dormancy which requires induced germination. Cryptocotyly has only evolved once in Drosera; it is a synapomorphy for the Australian clade (sensu Rivadavia et al., 2003), but was lost in the monotypic sect. Phycopsis consisting of D. binata. Thus, it is most likely that this germination pattern evolved among Drosera in Australia as the continent moved northwards and became drier (Yesson and Culham, 2006), as an adaptation to the seasonal Mediterranean climate with a pronounced dry season, occasional summer fires and cool moist winters. The seed remains dormant until germination is triggered by changing seasonal conditions, an ecological strategy followed by a range of Australian plants, including numerous Drosera spp. (Bell et al., 1993).

Although the summits of the Neblina massif are usually regarded as stable, wet tropical Amazonian habitats, D. meristocaulis occurs on the drier northern plateaus of these highlands, from where occasional fires have been reported (Givnish et al., 1986; McPherson, 2006). At least a few endemic plants from this area seem to present morphological adaptations to avoid fire damage (Givnish et al., 1986; Judziewicz, 1998). Seasonal droughts and wildfires are conditions reminiscent of the habitats occupied by Drosera sect. Bryastrum in Oceania, which may explain why cryptocotyly is maintained in D. meristocaulis.

Phylogenetic significance of pollen morphology

Further morphological similarities between D. meristocaulis and members of sect. Bryastrum can be found in pollen. Duno de Stefano (1995) observed one single central pore as an aperture in pollen tetrads of D. meristocaulis and therefore assigned it to aperture type 7, which is confined to sect. Drosera (Takahashi and Sohma, 1982). However, he did not recognize that a single proximal pore is also found in aperture type 8 and the intermediate type 7–8 (Takahashi and Sohma, 1982).

The pollen tetrad of D. meristocaulis shares common features of aperture type (one proximal central pore in each pollen grain) and pollen structure (radial channel plaits poorly developed and channel openings surrounded by a thick exinous wall) with pollen known as type 8 or intermediate type 7–8, respectively (Takahashi and Sohma, 1982). These two pollen types are confined to species of the Australian Drosera clade (sensu Rivadavia et al., 2003), except for D. glanduligera of the monotypic sect. Coelophylla, which exhibits a unique pollen tetrad of type 5 (Takahashi and Sohma, 1982). The ornamentation of the exinous wall of D. meristocaulis pollen is also distinct from that of all other South American Drosera spp. (Duno de Stefano, 1995), as it has no spinules and few rather large spines. This ornamentation is commonly found in Australian Drosera spp., especially in members of sect. Bryastrum and sect. Lasiocephala (Takahashi and Sohma, 1982).

Phylogenetic significance of leaf trichome characters

The leaves of members of the Bryastrum clade (including sect. Lasiocephala and sect. Bryastrum, sensu Seine and Barthlott, 1994) are characterized by the presence of biseriate sessile trichomes (‘microglands’) with elongated basal cells, which represent a synapomorphic character for this monophyletic group. These trichomes were called ‘Rorella-type glands’ by Seine and Barthlott (1993) and classified as ‘type 4 and 5 glands’ by Länger et al. (1995). Drosera meristocaulis has type T2 biseriate and T11–12 multiseriate sessile trichomes (Seine and Barthlott, 1993; Länger et al., 1995). Conran et al. (2007) stated that the trichome patterns found in D. meristocaulis are ambiguous, as they can be observed in members of both the Drosera and the Bryastrum clade (sensu Rivadavia et al., 2003), and that only in combination with the germination pattern could the phylogenetic position of sect. Meristocaulis be verified. However, the stout, short yellow gland-like trichomes on the adaxial and abaxial petiole surface of D. meristocaulis (Seine and Barthlott, 1993) do also occur in some pygmy Drosera spp. (e.g. Drosera nitidula and related species, A. Fleischmann, pers. obs.). These trichomes have a four-celled peduncle and a glandular head consisting of about 20 cells. This type of glandular trichome produces a sub-cuticular yellow secretion and occurs on the leaf surface and also on the emergences (Fig. 7).

Fig. 7.

Fig. 7.

Open in a new tab

Yellow glandular trichomes on a leaf of Drosera meristocaulis. Scale bar = 1 mm. Photograph by Daniel Olschewski, with kind permission.

Members of sect. Bryastrum all share a special, eight-celled biseriate type of microgland, so-called ‘Rorella-trichomes’ (Seine and Barthlott, 1993), which are usually found on the abaxial surface of the petiole and lamina. Seine and Barthlott (1993) did not observe these Rorella-trichomes in the specimens of D. meristocaulis they studied, and we did not detect them in our study material. The absence of Rorella-trichomes is a morphological character that supports the phylogenetic position of D. meristocaulis as sister to sect. Bryastrum, not as a member of this section, and rejection of subgenus Meristocaulis sensu Schlauer (1996).

Phylogenetic significance of karyology

Chromosome numbers in Drosera range from 2n = 6 to 2n = 80, and are in strong phylogenetic accordance with the clades revealed by Rivadavia et al. (2003). The Australian clade exhibits the greatest variability of karyotypes and forms extensive aneuploid and polyploid series, with relatively low chromosome numbers, ranging from 2n = 6 to 2n = 40, resulting from base numbers x = 3, 5, 6, 7, 8, 9, 10, 11, 13, 14 and 23 (Kondo and Lavarack, 1984; Sheikh and Kondo, 1995; Rivadavia et al., 2003; Rivadavia, 2005; Lowrie and Conran, 2007). In contrast, all New World species of Drosera (belonging to sect. Drosera and sect. Thelocalyx) form a homogeneous group, with relatively conserved chromosome numbers of 2n = 20 or 40 (i.e. polyploid series of the base number x = 10), suggesting at least two independent colonization events of a diploid and tetraploid group (Rivadavia et al., 2003; Rivadavia, 2005). Although representing an approximate range, the newly inferred karyotype of 2n = approx. 32–36 for D. meristocaulis contrasts with the polyploid series found in all other South American species, but fits the aneuploid series found in Drosera spp. of the Australian clade.

Karyology can be a useful character in Drosera taxonomy for both species delimitation and infrageneric classification. An example for the latter was shown with the proposal to remove the enigmatic northern Australian D. banksii from sect. Ergaleium and to place it in sect. Lasiocephala (Kondo and Lavarack, 1984). This suggestion was later confirmed by further morphological (Seine and Barthlott, 1994) and molecular phylogenetic data (A. Fleischmann, unpubl. res.), which revealed that this species grouped with sect. Lasiocephala.

CONCLUSIONS

Molecular phylogenetic data and morphological characters, including germination pattern, pollen anatomy, karyotype and leaf trichome characters, support the placement of D. meristocaulis in the Australian clade in a monotypic section (sect. Meristocaulis), as sister to sect. Bryastrum (Figs 3 and 4), or even in this section in the case of the rbcL data set (Fig. 2). In contrast to the Drosera of sect. Bryastrum, D. meristocaulis does not reproduce asexually by gemmae. It is possible that the ancestors of D. meristocaulis lost the ability to produce gemmae after reaching South America, but it is also probable that gemmae production evolved in pygmy sundews after this lineage split from D. meristocaulis. Gemmae are a synapomorphy of the pygmy sundews and are likely to have evolved as an adaptation to a seasonal climate as the Australian continent became drier (Yesson and Culham, 2006). The fact that gemmae are found in all species of sect. Bryastrum suggests that it is not only a successful means of asexual reproduction, but also an essential ecological survival strategy in the Mediterranean climate of southwestern Australia. The production of gemmae requires a considerable allocation of resources (Karlsson and Pate, 1992) and is possibly an important mechanism for rapid clonal colonization of seasonally available habitats.

Cryptocotylar germination may represent an adaptation to fire, a common phenomenon in both regions (Oceania and the Neblina massif), playing an important role in the maintenance of morphological similarities between D. meristocaulis and species of sect. Bryastrum.

Any explanation for the presence on the Neblina massif of a plant species descended from an Australian lineage is sure to be at least controversial. A recent study estimated that sect. Bryastrum began its diversification about 13–12 Mya (Yesson and Culham, 2006), and therefore contradicts a Gondwanan origin for D. meristocaulis. Despite the lack of information on a dispersal route from Australia to northern South America, the evidence that this did in fact occur cannot be rejected. As a vector for this rare LDD event, birds or wind seem most conceivable, although no avian migratory pathways from Australia to northern South America have been reported (Lomolino et al., 2006). An Australian to temperate South America disjunction is also known from a few plant families (Thorne, 1972), including Winteraceae. A strikingly similar biogeographic pattern is found in the three earliest branching members of Loranthaceae (showy mistletoes), namely the terrestrial monotypic genera Nuytsia floribunda from south-western Western Australia, Atkinsonia ligustrina from eastern Australia and Gaiadendron from Central and South America (also occurring on Neblina) (Vidal-Russell and Nickrent, 2008). In the case of Loranthaceae, a Gondawanan origin is assumed, which would explain the biogeography of the three taxa, which are successive sister taxa to all remaining Loranthaceae (Vidal-Russell and Nickrent, 2008).

Recent LDD from Australia (or southeast Asia) to South America has previously been proposed for the species sister pair Drosera burmannii and D. sessilifolia of sect. Thelocalyx (Rivadavia et al., 2003). In accordance with phylogenetic and other evidence presented above, D. meristocaulis is most probably also descended from an LDD event from Australia to South America, and is probably not a supposed palaeoendemic species that descended from pygmy sundew-like plants previously widespread in Gondwana, and which also led to extant sect. Bryastrum.

SUPPLEMENTARY DATA

Supplementary data are available online at www.aob.oxfordjournals.org and consist of photographic vouchers for Drosera barbigera, D. glanduligera, D. nitidula, D. occidentalis, D. ordensis, D. paradoxa and D. scorpioides, and sequence alignments (.txt files) for rbcL, ITS and rps16.

Supplementary Data
supp_110_1_11__index.html (1.4KB, html)

ACKNOWLEDGEMENTS

We thank Allen Lowrie for providing plant material for germination studies and for useful discussions on Australian pygmy Drosera species; Ivan Snyder and Matt Hochberg for helping with the seed germination experiments; Kamil Pasek for sending in vitro material of Drosera meristocaulis; Susanne Renner for helpful input on long-distance dispersal events; Daniel Olschewski, Bochum, for providing photos of the glands from cultivated plants; Tanja Ernst, Munich, and M. V. A. Sluys, São Paulo, for laboratory support; and Eva Facher, Munich, for assistance with SEM photography. Michael Fay and two anonymous reviewers are thanked for helpful comments on the manuscript.

LITERATURE CITED

  1. Akaike H. Information theory and an extension of the maximum likelihood principle. In: Petrov BN, Csaki F, editors. Second International Symposium on Information Theory. Budapest: Akademiai Kiado; 1973. pp. 267–281. [Google Scholar]
  2. Akaike H. A new look at the statistical model identification. IEEE Transactions on Automatic Control. 1974;19:716–723. [Google Scholar]
  3. Bell DT, Plummer JA, Taylor SK. Seed germination ecology in southwestern Western Australia. Botanical Review. 1993;59:1491–1495. [Google Scholar]
  4. Bremer K. The limits of amino acid sequence data in angiosperm phylogenetic reconstruction. Evolution. 1988;42:795–803. doi: 10.1111/j.1558-5646.1988.tb02497.x. [DOI] [PubMed] [Google Scholar]
  5. Brewer-Carías C, editor. Cerro de la Neblina: resultados de la Expedición 1983–1987. Caracas: FUDECI; 1988. [Google Scholar]
  6. Carlquist S. The biota of long-distance dispersal 1 – principles of dispersal and evolution. Quarterly Review of Biology. 1966;41:247–270. doi: 10.1086/405054. [DOI] [PubMed] [Google Scholar]
  7. Carlquist S. Darwin on island plants. Botanical Journal of the Linnean Society. 2010;162:S4–S9. [Google Scholar]
  8. Clifford HT. Cryptocotyly in Australian dicotyledons. Flora Malesiana Bulletin. 1984;37:49–53. [Google Scholar]
  9. Conran JG, Jaudzems G, Hallam ND. Droseraceae germination patterns and their taxonomic significance. Botanical Journal of the Linnean Society. 1997;123:211–223. [Google Scholar]
  10. Conran JG, Jaudzems G, Hallam ND. Droseraceae gland and germination patterns revisted: support for recent molecular phylogenetic studies. Carnivorous Plant Newsletter. 2007;36:14–20. [Google Scholar]
  11. Darwin C. On the origin of species. London: John Murray; 1859. [Google Scholar]
  12. De Queiroz A. The resurrection of oceanic dispersal in historical biogeography. Trends in Ecology and Evolution. 2005;20:68–73. doi: 10.1016/j.tree.2004.11.006. [DOI] [PubMed] [Google Scholar]
  13. Dick CW, Bermingham E, Lemes MR, Gribel R. Extreme long-distance dispersal of the lowland tropical rainforest tree Ceiba pentandra L. (Malvaceae) in Africa and the Neotropics. Molecular Ecology. 2007;16:3039–3049. doi: 10.1111/j.1365-294X.2007.03341.x. [DOI] [PubMed] [Google Scholar]
  14. Diels L. Droseraceae. In: Engler A, editor. Das Pflanzenreich IV, 112. Leipzig: Verlag Wilhelm Engelmann; 1906. pp. 1–136. [Google Scholar]
  15. Doyle JA, Doyle JL. A rapid isolation procedure for small quantities of fresh leaf tissue. Phytochemistry Bulletin, Botanical Society of America. 1987;19:11–15. [Google Scholar]
  16. Duno de Stefano R. El genero Drosera (Droseraceae) en Venezuela. Acta Botánica Venezuelica. 1995;18:67–95. [Google Scholar]
  17. Erdtman G. The acetolysis method, a revised description. Svensk Botanisk Tidskrift. 1960;54:561–564. [Google Scholar]
  18. Felsenstein J. Confidence limits on phylogenies: an approach using the bootstrap. Evolution. 1985;39:783–791. doi: 10.1111/j.1558-5646.1985.tb00420.x. [DOI] [PubMed] [Google Scholar]
  19. Geyer CJ. Markov chain Monte Carlo maximum likelihood. In: Keramidas EM, editor. Computing Science and Statistics: Proceedings of the 23rd Symposium on the Interface. Fairfax Station: Interface Foundation; 1991. pp. 156–163. [Google Scholar]
  20. Givnish TJ, McDiarmid RW, Buck WR. Fire adaptation in Neblinaria celiae (Theaceae), a high-elevation rosette shrub endemic to a wet equatorial tepui. Oecologia. 1986;70:481–485. doi: 10.1007/BF00379892. [DOI] [PubMed] [Google Scholar]
  21. Givnish TJ, Renner SS. Tropical intercontinental disjunctions: Gondwana breakup, immigration from the boreotropics, and transoceanic dispersal. International Journal of Plant Sciences. 2004;165(4 Suppl):S1–S6. [Google Scholar]
  22. Goebel K. Morphologische und biologische Bemerkungen. 18. Brutknospenbildung bei Drosera pygmaea und einigen Monokotylen. Flora. 1908;98:324–335. [Google Scholar]
  23. Goloboff PA, Farris JS, Nixon KC. TNT, a free program for phylogenetic analysis. Cladistics. 2008;24:774–786. [Google Scholar]
  24. Hall TA. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symposium Series. 1999;41:95–98. [Google Scholar]
  25. Hasebe M, Omori T, Nakazawa M, Sano T, Kato M, Iwatsuki K. rbcL gene sequences provide evidence for the evolutionary lineages of leptosporangiate ferns. Proceedings of the National Academy of Sciences, USA. 1994;91:5730–5734. doi: 10.1073/pnas.91.12.5730. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Huber O. Guayana highlands versus Guayana lowlands, a reappraisal. Taxon. 1988;37:595–614. [Google Scholar]
  27. Huber O. Vegetation. In: Steyermark JA, Berry PE, Holst BK, editors. Flora of the Venezuelan Guayana 1. St Louis, MO: Missouri Botanical Garden; 1995. pp. 97–160. [Google Scholar]
  28. Huelsenbeck JP, Ronquist F. MRBAYES: Bayesian inference of phylogenetic trees. Bioinformatics. 2001;17:754–755. doi: 10.1093/bioinformatics/17.8.754. [DOI] [PubMed] [Google Scholar]
  29. Judziewicz EJ. A revision of Myriocladus (Poaceae: Bambusoideae: Bambuseae) Brittonia. 1998;50:430–446. [Google Scholar]
  30. Karlsson PS, Pate JS. Resource allocation to asexual gemma production and sexual reproduction in south-western Australian pygmy and micro stilt-form species of sundew (Drosera spp., Droseraceae) Australian Journal of Botany. 1992;40:353–364. [Google Scholar]
  31. Kondo K, Lavarack PS. A cytotaxonomic study of some Australian species of Drosera L. (Droseraceae) Botanical Journal of the Linnean Society. 1984;88:317–333. [Google Scholar]
  32. Länger R, Pein I, Kopp B. Glandular hairs in the genus Drosera (Droseraceae) Plant Systematics and Evolution. 1995;194:163–172. [Google Scholar]
  33. Lomolino MV, Riddle BR, Brown JH. Biogeography. 3rd edn. Sunderland, MA: Sinauer Associates; 2006. [Google Scholar]
  34. Lowrie A. Carnivorous plants of Australia 2. Perth: University of Western Australia Press; 1989. [Google Scholar]
  35. Lowrie A. Carnivorous plants of Australia 3. Perth: University of Western Australia Press; 1998. [Google Scholar]
  36. Lowrie A, Carlquist S. Eight new taxa of Drosera from Australia. Phytologia. 1992;73:98–116. [Google Scholar]
  37. Lowrie A, Conran JG. Drosera × sidjamesii (Droseraceae): systematics and ecology of a natural hybrid from Western Australia. Australian Systematic Botany. 2007;20:44–53. [Google Scholar]
  38. Macherey-Nagel . NucleoSpin® Plant: genomic DNA from plant user manual. KG, Düren, Germany: Macherey-Nagel GmbH & Co; 2007. February 2005/Rev. 3. [Google Scholar]
  39. Maguire B, Wurdack JJ. The botany of the Guayana Highland, Part II. Memoirs of the New York Botanical Garden. 1957;9:235–392. [Google Scholar]
  40. Mann P. Drosera gibsonii (Droseraceae), a new Pygmy Drosera from south-west Western Australia. Nuytsia. 2007;16:321–323. [Google Scholar]
  41. Miranda VFO, Martins VG, Furlan A, Bacci M., Jr Plant or fungal sequences? An alternative optimized PCR protocol to avoid ITS (nrDNA) misamplification. Brazilian Archives of Biology and Technology. 2010;53:141–152. [Google Scholar]
  42. McPherson S. Pitcher plants of the Americas. Blacksburg: McDonald & Woodward; 2006. [Google Scholar]
  43. McPherson S. Carnivorous plants and their habitats. Poole, UK: Redfern Natural History Productions; 2010. [Google Scholar]
  44. Muñoz J, Felicísimo AM, Cabezas F, Burgaz AR, Martínez I. Wind as a long-distance dispersal vehicle in the Southern Hemisphere. Science. 2004;304:1144–1147. doi: 10.1126/science.1095210. [DOI] [PubMed] [Google Scholar]
  45. Oxelman B, Lidén M, Berglund D. Chloroplast rps16 intron phylogeny of the tribe Sileneae (Caryophyllaceae) Plant Systematics and Evolution. 1997;206:393–410. [Google Scholar]
  46. Page RDM. TREEVIEW: an application to display phylogenetic trees on personal computers. Computer Applications in the Biosciences. 1996;12:357–358. doi: 10.1093/bioinformatics/12.4.357. [DOI] [PubMed] [Google Scholar]
  47. Posada D, Crandall KA. MODELTEST: testing the model of DNA substitution. Bioinformatics. 1998;14:817–818. doi: 10.1093/bioinformatics/14.9.817. [DOI] [PubMed] [Google Scholar]
  48. Rivadavia F, Kondo K, Kato M, Hasebe M. Phylogeny of the sundews, Drosera (Droseraceae), based on chloroplast rbcL and nuclear 18S ribosomal DNA sequences. American Journal of Botany. 2003;90:123–130. doi: 10.3732/ajb.90.1.123. [DOI] [PubMed] [Google Scholar]
  49. Rivadavia F. New chromosome numbers for Drosera L. (Droseraceae) Carnivorous Plant Newsletter. 2005;34:85–91. [Google Scholar]
  50. Ronquist F, Huelsenbeck JP. MRBAYES 3: Bayesian phylogenetic inference under mixed models. Bioinformatics. 2003;19:1572–1574. doi: 10.1093/bioinformatics/btg180. [DOI] [PubMed] [Google Scholar]
  51. Rull V. Biogeography of the ‘Lost World’: a palaeoecological perspective. Earth-Science Reviews. 2004;67:125–137. [Google Scholar]
  52. Schlauer J. A dichotomous key to the genus Drosera L. (Droseraceae) Carnivorous Plant Newsletter. 1996;25:67–88. [Google Scholar]
  53. Schlauer J. Carnivorous Plant Database. 2007 www.omnisterra.com/bot/cp_home.cgi . (last accessed 4 April 2011) [Google Scholar]
  54. Seine R, Barthlott W. On the morphology of trichomes and tentacles of Droseraceae Salisb. Beiträge zur Biologie der Pflanzen. 1993;67:354–366. [Google Scholar]
  55. Seine R, Barthlott W. Some proposals on the infrageneric classification of Drosera L. Taxon. 1994;43:583–589. [Google Scholar]
  56. Sheikh SA, Kondo K. Differential staining with orcein, Giemsa, CMA, and DAPI for comparative chromosome study of 12 species of Australian Drosera (Droseraceae) American Journal of Botany. 1995;82:1278–1286. [Google Scholar]
  57. Steyermark JA. Flora of the Guayana Highland: endemicity of the generic flora of the summits of the venezuela tepuis. Taxon. 1979;28:45–54. [Google Scholar]
  58. Swofford DL. PAUP*: phylogenetic analysis using parsimony (*and other methods), version 4·0. Sunderland, MA: Sinauer Associates; 2002. [Google Scholar]
  59. Sytsma KJ, Litt A, Zjhra ML, et al. Clades, clocks, and continents: historical and biogeographical analysis of Myrtaceae, Vochysiaceae, and relatives in the Southern Hemisphere. International Journal of Plant Sciences. 2004;165:S85–S105. [Google Scholar]
  60. Takahashi H, Sohma K. Pollen morphology of the Droseraceae and its related taxa. Science Reports of the Research Institutes Tohoku University (Biology) 1982;38:81–156. [Google Scholar]
  61. Thompson JD, Higgins DJ, Gibson TJ. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Research. 1994;22:4673–4680. doi: 10.1093/nar/22.22.4673. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Thorne RF. Major disjunctions in the geographic ranges of seed plants. Quarterly Review of Biology. 1972;47:365–411. [Google Scholar]
  63. Vidal-Russell R, Nickrent DL. Evolutionary relationships in the showy mistletoe family (Loranthaceae) American Journal of Botany. 2008;95:1015–1029. doi: 10.3732/ajb.0800085. [DOI] [PubMed] [Google Scholar]
  64. Yesson C, Culham A. Phyloclimatic modeling: combining phylogenetics and bioclimatic modeling. Systematic Biology. 2006;55:785–802. doi: 10.1080/1063515060081570. [DOI] [PubMed] [Google Scholar]
  65. Walker JB, Sytsma KJ. Staminal evolution in the genus Salvia: molecular genetic evidence. Annals of Botany. 2007;100:375–391. doi: 10.1093/aob/mcl176. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. White TJ, Bruns T, Lee S, Taylor J. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis MA, Gelfand DH, Sninsky JJ, White TJ, editors. PCR protocols – a guide to methods and applications. UK edn. San Diego: Academic Press; 1990. pp. 315–322. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data
supp_110_1_11__index.html (1.4KB, html)
supp_mcs096_mcs096supp.pdf (960.8KB, pdf)
supp_mcs096_mcs096supp1.txt (12.9KB, txt)
supp_mcs096_mcs096supp2.txt (99.4KB, txt)
supp_mcs096_mcs096supp3.txt (14KB, txt)

Articles from Annals of Botany are provided here courtesy of Oxford University Press

RESOURCES