Plant Systematics and Evolution

, Volume 304, Issue 5, pp 699–721 | Cite as

New insights into variation, evolution and taxonomy of fairy lanterns (Thismia, Thismiaceae) with four new species from Borneo

Original Article
  • 71 Downloads

Abstract

Fairy lanterns (Thismia Griff.) is a genus of poorly known mycoheterotrophic plants with unclear infrageneric classification. Commonly used approaches that utilize just a single or few traits in higher-level taxonomy lead to an apparently artificial system. In this contribution, four new species from Sarawak, northern Borneo, are described and illustrated. Thismia acuminata, T. laevis and T. nigra belong undoubtedly to section Sarcosiphon. Thismia viridistriata exhibits a high morphological variability with some individuals fitting section Scaphiophora based on the presence of a column on the top of the mitre, but otherwise perfectly matching the morphological characteristic of section Sarcosiphon. Five-locus (SSU, ITS, LSU, matR, atpA) phylogeny inference revealed paraphyly or polyphyly in the studied infrageneric taxa and showed that the importance of some traits traditionally used in Thismia taxonomy was overestimated. Most of the markers provide comparable phylogenetic signal; LSU performs best and is highly recommended for further phylogenetic studies. On the other hand, ITS is hypervariable and thus informative only within (sub)sections, as well as on intraspecific level, as proven in T. viridistriata with two distinct ITS (and also LSU and matR) alleles and two morphotypes within a small geographic area, which leads to an assumption of strong reproductive isolation even among nearby populations. For delimitation of species, the key trait appears to be the structure of the connective and any of the molecular markers used here.

Keywords

Fairy lanterns Kelabit Highlands Mycoheterotrophy Pulong Tau Sarcosiphon Thismia 

Introduction

Fairy lanterns (Thismia Griff.; Thismiaceae, or Burmanniaceae s.l.) are a genus of small, mycoheterotrophic plants distributed mainly in Southeast Asia, tropical America, Japan, Australia and New Zealand. Around 72 species are currently recognized in the genus, of which more than half is restricted to Southeast Asia, particularly the Malesian floristic region (Hroneš 2016; Cooper 2017; Kumar et al. 2017; Sochor et al. 2017; Sujanapal et al. 2017; Suetsugu et al. 2017; Tsukaya et al. 2017; Chantanaorrapint and Suddee 2018; Suetsugu et al. 2018). As obligate mycoheterotrophs, fairy lanterns are dependent on their fungal symbionts that are further associated with their autotrophic plant symbionts and other organisms (Bidartondo 2005). Fully functional biotic networks may therefore be the crucial factor for Thismia occurrence and survival. That is possibly the main reason why these plants are found predominantly in primary tropical rainforests (Merckx et al. 2013b) which are renowned for richness of biotic interactions. Furthermore, the mycorrhizal association is often highly specific and phylogenetically conserved in Thismiaceae and distribution of the fungal symbionts may thus limit occurrence and spread of the plants (Merckx et al. 2012, 2017). On the other hand, the associated fungi are probably widespread and enable relatively rapid expansions of fairy lanterns into new areas, as was observed in T. rodwayi F.Muell. and T. hillii (Cheeseman) N.Pfeiff. (Merckx et al. 2017). Most Thismia species are considered extremely rare, though. The underlying reasons may vary from true rarity of the species (resulting from their narrow niche), to our poor ability to detect them. Their ephemeral aboveground growth and inconspicuous appearance in combination with their habitats (mostly difficultly accessible tropical rainforests) and phenology (flowering predominantly during rainy seasons) make them quite a problematic study object.

Besides the difficult detection, several other factors complicate taxonomical studies on fairy lanterns. Obligate mycoheterotrophy resulted in many reductions in vegetative morphology, and only very few characters apart from those on flowers can be used in taxonomy. The systematics of Thismia is mostly based on organization of perianth lobes. According to the latest monographic treatment of the genus by Jonker (1938), the Old World species with free perianth lobes are treated in section Thismia (Euthismia Schlr.) with two subsections—Odoardoa Schlr. (with all perianth lobes equal in size and shape) and Brunonithismia Jonker (with reduced outer lobes). Sections Rodwaya Schltr. and Glaziocharis (Taub. ex Warm.) Hatus. include species whose inner tepal lobes are bent inwards but are usually not or only imperfectly connate. Species with inner tepals connate and forming a mitre above the flower opening were included in section Sarcosiphon (Blume) Jonker, but only if they lack any filiform appendages on the top of the mitre. If appendages are present, Jonker (1938) treated such species in the separate genera Scaphiophora Schltr. and Geomitra Becc. (with one or three appendages, respectively), which were recently reduced to sectional status (Kumar et al. 2017). However, such systems based on absolute importance of a single trait may not reflect the true phylogenetic relationships and may thus be artificial.

Species of Thismia are traditionally delimited by various traits, from overall habitus to inner flower anatomy (Jonker 1948; Hroneš 2016). Nevertheless, many species are only known from type collections, sometimes very poor (single flowering or even fruiting individual), covering often just a small fragment of true natural variation of the species. Common problem, typical for old plant names but underlined by the limited use of standard (i.e. pressed) herbarium specimens in Thismia, is insufficient description with incomplete documentation. Moreover, many species are desperately inaccessible not only for morphological evaluation, but also for DNA analyses due to their rarity and very limited usability of herbarium specimens. Little is known also about population (genetic) structure, reproductive biology, dispersal abilities or population dynamics, among others, despite some recent advances (Mar and Saunders 2015; Merckx et al. 2017). Consequently, making any natural taxonomic system, as well as defining a proper species concept, is problematic. This fact is becoming apparent with increasing number of Thismia collections, new taxa descriptions and involvement of molecular methods during the last decade (Merckx et al. 2017).

During our recent expedition to the Kelabit Highlands, Borneo, eight undescribed species of Thismia were discovered, some of which exhibited extraordinarily common occurrence as well as extreme variability in outer morphology. In this paper, we aim 1. to describe formally four of these species as new to science; 2. provide data on their morphological and molecular (ribosomal and mitochondrial DNA) variation; and 3. discuss usefulness of different morphological traits and DNA loci for taxonomy and phylogeny inference in fairy lanterns.

Materials and methods

Plant material for this study was collected during a 10-day expedition in the vicinity of Pa’Umor, Pa’Lungan and Bario villages, the Kelabit Highlands, Sarawak, Malaysia. Thismia plants were searched systematically in several localities that were mostly selected in advance based on our previous experiences. Each population of each found morphological type (separated by ca > 100 m from the neighbouring population) was studied morphologically in the field camps using hand lenses (20 × − 60 × magnification) and macrophotography. Specimens of whole plants were taken from each population in form of pressed herbarium and/or alcohol (70% ethanol) specimen. Piece of tissue (e.g. pedicel, capsule, part of stem) from one individual per population was silica gel-dried for subsequent DNA extraction. Several fruiting individuals of unknown identity were also added to the molecular analyses, as well as samples of T. hexagona Dančák, Hroneš, Kobrlová & Sochor, T. inconspicua Sochor & Dančák, T. brunneomitra Hroneš, Kobrlová & Dančák, T. neptunis Becc. and T. pallida Hroneš, Dančák & Rejžek from the type localities in Brunei, western Sarawak and Sabah, respectively.

DNA was extracted by the CTAB method (Doyle and Doyle 1987), and sequence data were generated for three nuclear and two mitochondrial loci. The small subunit of ribosomal DNA (SSU rDNA) and internal transcribed spacers (ITS) were amplified and sequenced with primers NS1, NS6, NS5 and ITS4 (White et al.1990) or 268rev (Kuzoff et al. 1998) and large subunit of ribosomal DNA (LSU rDNA) by primers N-nc26S6 and 2134rev (Kuzoff et al. 1998). Amplification of the LSU region adjacent to ITS2 was attempted by primers N-nc26S1 and 1229rev or 1499rev, but most reactions did not provide specific product and were therefore not analysed further. The mitochondrial atpA gene was amplified and sequenced with primers developed by Eyre-Walker and Gaut (1997) and matR gene with primers 26F and 1002R (Meng et al. 2002). All PCRs were performed with Kapa polymerase (Kapa Biosystems) following a standard protocol with 37 to 40 cycles and annealing temperature of 56 °C (rDNA, atpA) or 47 °C (matR). The PCR products were purified by precipitation with polyethylene glycol (10% PEG 6000 and 1.25 M NaCl in the precipitation mixture) and sequenced in both directions by Sanger method at Macrogene Europe.

Sequences were edited and aligned in Geneious (ver. 7.1.7., Biomatters) and deposited in NCBI GenBank under accession numbers MG008338–MG008393 and MG765543–MG765546. Alleles of each locus were distinguished based on single-nucleotide polymorphisms (sequences differing at least in one mutation were considered different alleles). For characterization of molecular variation at each locus, twelve (morpho)species were selected whose data from all five loci were available (one individual per species); seven from sect. Sarcosiphon and five from sect. Thismia. Maximum parsimony trees were computed in MEGA (ver. 6.06; Tamura et al. 2011) by subtree-pruning-regrafting method and MP search level of 5 and tested by bootstrapping with 1000 replications. Bayesian phylogeny inference was computed in MrBayes (ver. 3.2.4; Ronquist et al. 2012) with 107 generations, sampling every 3000th generation, in two independent runs, each with 4 chains; first 2.5 × 106 generations (25%) were excluded as burn-in. Substitution model for each locus was determined in jModelTest (ver. 2.1.4; Darriba et al. 2012) using Akaike information criterion (AIC). The models used are shown in Table 1. The individual gene trees were subsequently visually compared among loci and analytical methods. Phylogeny inference from concatenated data from the five loci was performed by the same methods as above for the sample set including additional eleven species whose sequences (at least four loci for the ingroup taxa) could be downloaded from NCBI Genbank (Online Resource 1). The substitution models for this analysis were determined in PartitionFinder2 (ver. 2.1.1; Lanfear et al. 2017) using the corrected AIC (AICc) and a greedy search, and partitions were set subsequently according to the loci in MrBayes (ITS1, ITS2 and 5.8S rDNA were treated as separate partitions; see Table 1).
Table 1

Comparison of different loci in terms of total and parsimony-informative variability and usefulness for maximum parsimony (MP) and Bayesian (BI) phylogeny inference; ITS locus covers only the ITS1–5.8S rDNA–ITS2 region

Locus

SSU

ITS

LSU

atpA

matR

Alignment length [bp]

1713

691

957

1167

1003

Variable SNPs (% of total)

189 (12%)

492 (71%)

239 (25%)

50 (4%)

155 (15%)

Informative SNPs (% of variable SNPs)

125 (66%)

354 (72%)

179 (75%)

20 (40%)

70 (45%)

# most parsimonious trees

7

1

2

6

6

MP tree length

262

851

395

61

140

# splits with BS ≤ 50%

2

1

2

4

3

# splits with PP < 95%

4

5

3

4

5

Substitution model (based on AIC in jModelTest/AICc in PartitionFinder2)

GTR + I + G/GTR + I + G

GTR + I + G/GTR + I + G, SYM + Ga

TIM3 + G/GTR + G

TVM + I/GTR + I + G

TIM1 + G/GTR + G

SNP single-nucleotide polymorphism, BS bootstrap value, PP Bayesian posterior probability, AIC Akaike informative criterion, AICc corrected AIC

aFor 5.8S rDNA region of the sequence

Results

Nine different morphological types (morphotypes) of Thismia were discovered in the field. Two of them were later found to belong to the same species; eight species were therefore detected. Six of them were mitriform species with coralliform roots. One of these was morphologically unique species not similar to the rest of the species, and it will be described separately as Thismia kelabitiana ined. (Dančák et al. submitted). In another morphospecies, Thismia aff. nigra, only immature individuals were found, and it cannot be formally described until additional material is collected. The four remaining species of this group are formally described here as T. acuminata, T. laevis, T. nigra and T. viridistriata. Based on morphology, three of them belong undoubtedly to section Sarcosiphon and the remaining one, T. viridistriata, either to this section, or, due to the presence of a column on top of the mitre in some individuals, to the genus Scaphiophora sensu Jonker (1938) which is now reduced to Thismia (Merckx et al. 2013a; Kumar et al. 2017). The remaining two species were T. cornuta Hroneš, Sochor & Dančák (Hroneš et al. 2018) and an undescribed species, which is not treated here and will be described in a separate paper.

Detailed morphological evaluation could be performed in all discovered morphotypes except for the one that was only found with one developing (but not open) flower and possibly represented different species. This was treated as T. aff. nigra in further analyses (i.e. similar but not identical to T. nigra). Intraspecific morphological variability was generally low, except for T. viridistriata with two distinct morphotypes and several transitional individuals (Table 2). The best character that distinguishes all of the species was revealed to be connective morphology which was very stable within and among populations of the same species. Morphological variation is further discussed in Taxonomic treatment.
Table 2

Variation within and among the four new Thismia species in selected phenotypic characters

 

T. acuminata

T. laevis

T. nigra

T. viridistriata

Height (cm)

3–5

5.5–10.5

2.5–7

2–4.5(–9)

Number of leaves

3–6

4

3–6

2–6

Fruiting pedicel length (cm)

ca 1.5

up to 11

up to 9

up to 6

Number of flowers

1–2(–4)

4

2–5

1–5

Flower size

15–19 mm long, 5–6 mm wide

18 mm long, 7 mm wide

10–12 mm long, 4.5–6 mm wide

13–21 mm long, 4.5–6 mm wide

Perianth tube colour

Greyish or light beige to sometimes brownish orange, with 12 dark brown (to almost blackish) or reddish brown longitudinal stripes

Greyish, with dark brown to blackish apex and ribs and brown stripes between the ribs (together making 12 longitudinal dark stripes)

Dark greenish-grey with 1 transversal and 12 vertical blackish stripes

Greyish to rich brown or reddish brown with darker transversal stripe in the middle and 12 dark brown (to almost blackish) or reddish brown longitudinal stripes

Mitre processes

Absent

3 short obtuse processes

3 short obtuse processes or rarely absent

3 short processes or a single column

Connective apex

Densely hairy with ca 4–5 ± irregular short processes

Glabrous with no processes

2 small lobes with ca 5 long simple hairs

1–2 lobes with ca 5 ca 0.15–0.2 mm long simple hairs

Stigmatic lobes

Deeply divided into 2 triangular lobes or rather shallowly trilobed with the middle lobe smaller

Divided into 2 large and 1 small lobe

Deeply divided into 2 long triangular lobes

Shallowly bilobed

All of the five studied genomic loci could be amplified and sequenced easily in most samples. ITS region could not be amplified in T. hexagona using primer ITS4 due to two point mutations in the primer complementary sequence in this species. The region was therefore amplified and sequenced with primer 268rev, which anneals further in the LSU region.

In the dataset of twelve studied species from two sections, each locus was clearly able to distinguish between different morphologically defined species, with an exception of atpA in T. nigra and T. aff. nigra. The highest variability was observed in the ITS region and most of this variation was parsimony informative (Table 1). It also exhibited the highest intraspecific polymorphism (ten SNPs and one indel in T. viridistriata and one SNP in T. cornuta; Online Resource 2). Other intraspecific polymorphisms were detected at LSU and matR (two and one SNPs in T. viridistriata, respectively, resulting in two alleles at each locus that corresponded to the ITS alleles; Online Resources 1 and 2), and at SSU (three SNPs in T. cornuta forming three alleles). Only very few short conserved domains were detected at the ITS locus among species (all in 5.8S rDNA region), and the alignment was therefore very ambiguous, particularly among sections. Each of the remaining loci provided both considerable polymorphism (see Table 1) and multiple conserved motives that allowed for unambiguous alignment.

Ribosomal and mitochondrial DNA markers performed comparably in resolving phylogenetic relationships (Online Resource 3). The best resolving markers according to the number of maximally parsimonious trees and the number of highly supported splits were ITS and LSU (Table 1, Online Resource 3). These were also the only markers that provided a phylogenetic signal (although slightly different from each other) within the “core Sarcosiphon” clade (Online Resource 3). Considering serious ambiguities in alignment of ITS and therefore its limited usability among distant lineages, LSU remains the most suitable marker for whole-genus phylogeny reconstruction, although it has never been used in fairy lanterns until now. Nevertheless, slight incongruencies among loci call for necessity of multilocus phylogenetic inferences.

After inclusion of other species for which at least four of the studied loci were publicly available (Online Resource 1), a well resolved Bayesian multilocus phylogenetic tree with most branches well supported was obtained (Fig. 1). Bootstrap consensus maximum parsimony tree exhibited similar topology, except for a few splits (Online Resource 3). Four lineages were clearly distinguished that correspond to several infrageneric taxa—sect. Thismia, sect. Sarcosiphon (including sections Scaphiophora and Geomitra), and sections Rodwaya and Glaziocharis (sensu Kumar et al. 2017) sharing two evolutionary lineages. Thismia clavigera Becc., assigned to sect. Geomitra, groups with species of section Sarcosiphon. Three of the four species described here form a well-separated cluster (T. nigra, T. laevis and T. acuminata), whereas T. viridistriata forms different, basal lineage within the Sarcosiphon clade (Fig. 1; Online Resource 3).
Fig. 1

Bayesian tree inferred from five-locus concatenated alignment; posterior probabilities shown above branches. Infrageneric placement follows the treatment of Kumar et al. 2017, but note that, e.g. Thismia clavarioides was originally placed in sect. Rodwaya

Discussion

Connective morphology is the most appropriate character for species distinction

Despite the recent wave of interest that resulted in description of many new species and some new insights into biology and evolution of fairy lanterns, the genus Thismia remains a poorly understood plant genus. For instance, no or only poor and/or indirect data are available on population structure, reproduction systems, clonality, life cycle, dispersal ability, specificity of biotic interactions, etc. Intraspecific variation is also virtually unknown as most species are known only from the type collections. These facts greatly limit the development of taxonomical concepts in the group, including species concept (i.e. definition of species boundaries) and delimitation of higher taxa, particularly sections and subsections. Different approaches are therefore appearing. Whereas some authors treat morphologically atypical populations as infraspecific taxa (e.g. T. hexagona var. grandiflora; Tsukaya et al. 2014) or conspecific without further taxonomical distinction, e.g. T. clandestina Blume (Chantanaorrapint et al. 2015, Suetsugu et al. 2017), T. crocea Becc. and T. versteegii J.J.Sm. (Jonker 1938), or formerly conspecific T. rodwayi and T. hillii (Hunt et al. 2014), the common approach is describing new species. The main diagnostic characters for delimitation of species range from external traits (coloration, appendages on perianth lobes, etc.; e.g. Hunt et al. 2014; Chantanaorrapint et al. 2016; Tsukaya et al. 2017) to inner flower structure (e.g. Hroneš et al. 2015), or usually combination of several traits.

During our fieldwork, several populations of two distinct morphotypes were detected that shared inner flower morphology, but differed in outer characters, particularly in colour of flower, presence and length of terminal appendage on top of mitre, shape of perianth tube, flower size, number of stems from one root, number of flowers per stem and stem length (see Taxonomic treatment of T. viridistriata). Both morphotypes exhibited only low variation in molecular markers (Online Resources 2 and 3); two related alleles were detected at each of the ITS, LSU and matR loci; one allele was detected in both of the morphotypes and the second allele (rarer) in one of them (Online Resource 1). Moreover, some populations appeared transitional between both morphotypes in their morphology. Therefore, the observed variation could be considered as intraspecific. For delimitation of this species, the main diagnostic traits were inside the flower—mainly structure of connectives. Shape of stigma was also constant in this case, although it can be variable in Thismia and may reflect the anthesis phase (Sochor et al. 2017; see also Taxonomic treatment). Connective morphology is crucial also in distinguishing T. nigra and T. brunneomitra, which are strikingly similar to each other in external appearance, but well differentiated in molecular markers (Fig. 1, Online Resources 1 and 2). Importance of connective appendages morphology in taxonomy on the species level was noticed earlier (e.g. Tsukaya et al. 2014; Hroneš et al. 2015; Chantanaorrapint et al. 2016) and may be critical for comparison of geographically distant populations among which some differentiation can be expected, at least in sect. Sarcosiphon. Though, relatively little attention appears to be paid to this feature in some of the new species descriptions (e.g. Jarvie 1996; Kiew 1999; Chiang and Hsieh 2011; Tsukaya and Okada 2012; Dančák et al. 2013; Mar and Saunders 2015; Tsukaya et al. 2017). Another species-diagnostic marker could be any of the conservative genomic regions analyzed here. Unfortunately, herbarium specimens (particularly old and pressed ones) usually do not permit analysis of either of these traits (those on connectives or molecular ones) and the use of old names that are based on insufficient protologues may therefore be very difficult. An illustrative example is Thismia clandestina originally described from Java (see, e.g. Smith 1911) and recently reported from Thailand, although slightly differing from the Javan population (Chantanaorrapint et al. 2015; Suetsugu et al. 2017).

Perianth morphology has been overestimated in infrageneric taxonomy

Delimitation of higher infrageneric taxa (mainly sections and subsections) is also very unclear and awaits wider phylogenetic analyses and subsequent thorough revision (Hroneš 2016). The currently accepted concepts (e.g. Govaerts et al. 2007; Kumar et al. 2017) consider mainly modifications of petal lobes, as introduced by Schlechter (1921) and Jonker (1938). Although our preliminary screening covers only six infrageneric taxa (T. sect. Thismia subsect. Odoardoa and Brunonithismia, T. sect. Rodwaya, T. sect. Glaziocharis, T. sect. Sarcosiphon and T. sect. Geomitra), four of them (i.e. all with more than one sampled species) appear poly- or paraphyletic. Sect. Rodwaya includes at least two unrelated, yet morphologically very similar lineages (Fig. 1; Merckx et al. 2017; Sochor, Dančák, Hroneš, unpublished data). Subsect. Odoardoa is only monophyletic if subsect. Brunonithismia is included, and sect. Sarcosiphon should include sect. Geomitra. Interestingly, although very similar in morphology to the “core Sarcosiphon” clade (T. nigra, T. laevis, T. acuminata, etc.), T. viridistriata clearly forms a basal lineage of the whole SarcosiphonGeomitra clade (i.e. including T. clavigera and morphologically very distinct T. kelabitiana; Fig. 1). This position is well supported by every method and marker used (Online Resource 3). Also section Glaziocharis, if distinguished by the presence of appendages on outer perianth lobes (Kumar et al. 2017), is clearly polyphyletic (Fig. 1). These contradictions between phylogeny and classical taxonomy may indicate either convergent evolution or limited anagenesis in morphology of perianth in some distantly related lineages. Traits such as the presence of mitre and terminal appendages on perianth lobes therefore appear to have been overestimated in infrageneric taxonomy.

In contrast, root morphology should receive more attention in taxonomy. Root systems of Old World Thismia are of two architectural types—vermiform and coralliform (see, e.g. Jonker 1948, Imhof et al. 2013). Already Jonker (1938) pointed out the importance of root system in infrageneric taxonomy of Thismia as he used it in characteristics of his sections. He was also aware of coralliform roots of Scaphiophora but not Geomitra. However, due to scarcity of species with coralloid roots this was generally not emphasized in later research (see, e.g. Merckx et al. 2006). Nevertheless, it seems that root morphology could indeed be a good character of systematic importance as the coralliform roots are exclusively found in one of the four evolutionary lineages studied here (see Fig. 1) that corresponds to section Sarcosiphon, including Geomitra and Scaphiophora. Roots of the other groups are vermiform (or tuberous in neotropical outgroup taxa). Interestingly, the occurrence of coralliform roots is associated with other traits, e.g. species with coralliform roots always have perfect mitre and outer perianth lobes absent, considerably reduced or at least modified. To conclude, proper delimitation of infrageneric taxa and their morphological characterization needs to be critically re-evaluated in Old World fairy lanterns.

Thismia populations may be reproductively isolated at small geographic scale

Thanks to the two morphotypes and two ITS/LSU/matR alleles, as well as to the relatively common occurrence in the studied region, T. viridistriata may represent a unique model for further studies on evolution in fairy lanterns. Although the morphotypes were not always clearly separated and transitional forms were observed in some populations, no mixed signal was detected in the ITS or LSU sequences despite the fact that these are multi-copy nuclear markers that are usually able to detect recent, yet not always ancient hybridization (Kovarik et al. 2005; Sochor et al. 2015). Thismia viridistriata therefore appears to be a single species with multiple isolated populations (or other evolutionary units) that do not interbreed with each other (or only in an extent that precludes complete homogenization of both molecular and phenotypic traits in the metapopulation). Existence of cryptic species (although in an early stage of speciation) within T. viridistriata cannot be ruled out either, but the present data do not allow any finer taxonomical treatment. Interestingly, the studied area was relatively small (ca 10 × 18 km) and both of the morphotypes and genetic lineages were discovered even within one river valley just a few hundred meters apart, mostly not high enough above the water level, so that dispersal of seeds and whole plants by water is, hypothetically, well possible. Isolation by distance, physical barriers, environmental conditions or absence of the fungal symbiont is therefore hardly imaginable. Such a pattern, nevertheless, may be shaped in other ways of reproductive isolation. Based on a characteristic complex inner morphology of the flower and probable (yet tentative) protandry, outcrossing via cross-pollination by flies is generally presumed to be the dominant reproduction mode (see Mar and Saunders 2015 for details). These physical mechanisms preventing selfing, nevertheless, would be insufficient in case of uniclonal population structure or asexual seed development, for instance. Another mechanism leading to the among-population differentiation in both molecular and phenotypic traits could be strong preference of pollinators for particular flower shape and/or colour, which could also explain the differences in outer flower morphology (but see Kay and Sargent 2009). Nevertheless, all of these examples are only speculative at this state of knowledge and need to be systematically studied, before any conclusion can be made.

Taxonomic treatment

Thismia acuminata Hroneš, Dančák & Sochor, sp. nov.—HOLOTYPE: Malaysia, Sarawak, Kelabit Highlands, Pa’Lungan village, Arur Bedalawid, 3.0 km N of the village, 1152 m a. s. l., 3°50′14″N, 115°31′08″E, 15 Jan 2017, Sochor et al. BOR6/17 (holotype: SAR [accession number Sochor/BOR-6/17]; isotype: OL [accession number 35271]) (Figs. 2, 3).
Fig. 2

Thismia acuminata: a habit; b outer view of connective tube; c ovaries with stigmas; d longitudinal section of connective tube (from Sochor et al. BOR6/17). Drawn by R. Melichárková

Fig. 3

Thismia acuminata: a, b habit; c outer view of connective tube; d longitudinal section of connective tube; e detail of appendages on connectives; f, g ovaries with two types of stigma; h mature capsule with seeds (from Sochor et al. BOR6/17)

Etymology: From the Latin acuminatus, reflecting acuminate shape of mitre.

Diagnosis: Similar to T. laevis Sochor, Dančák & Hroneš but differing in tall acuminate mitre and by having several small lobes and tomentum of long hairs on connective apex and nose-shaped central lobe of lateral appendage.

Description: Achlorophyllous ground herb, ca 3–5 cm tall. Roots short, clustered, coralliform, light beige to brown. Stem 14–50(–70) mm long, creeping, ascending to erect, sparsely branched or unbranched, pale to rich brown, pinkish, greyish to almost orange; branching sympodial, with one bud covered in bracts of the flower in anthesis and growing as the main stem after anthesis, pedicels indiscernible during anthesis, elongating during fruit maturing to ca 1.5 cm. Leaves 3–6 below the first flower, spirally arranged, scale-like, triangular, acute (to acuminate), entire, sometimes with prominent central rib, 2.2–3.0(–6) × 1.2–1.5(–2.0) mm, of the same colour as stem. Floral bracts usually 3, of the same shape and colour as leaves, but slightly larger (4–6 × 1.5–1.7 mm). Flowers 1–2 (–4) per stem, actinomorphic, 15–19 mm long, 5–6 mm wide at widest part; perianth tube narrowly urceolate, widest at its upper quarter, often with slight narrowing above the middle; outer surface greyish or light beige to sometimes brownish orange, with 12 dark brown (to almost blackish) or reddish brown longitudinal stripes, with 6 longitudinal ribs; outer perianth lobes absent; inner perianth lobes bent upwards, connate at top and forming mitre with 3 holes; holes elliptic to rounded, 3–4 mm wide, 2–3 mm high; mitre tall and narrow, acuminate, ca 6 mm high, perfectly connate at the tip (without processes), with 3 longitudinal ribs, dark reddish brown to almost blackish with paler base (columns); annulus formed by 12 low crenate lobes, hairy. Stamens 6, pendent from the apical margin of the perianth tube; filaments free, short, curved downwards, hairy, light brown-violet to brown-pinkish; connectives broad and flattened, laterally connate to form a tube, densely hairy to tomentose in the apical half, sparsely hairy to glabrous in the basal half, ca 4 mm long, pale violet or pinkish to whitish; interstaminal glands not prominent, oblong, ca 0.5 mm long and 0.25 mm wide, placed between bases of lateral appendages; apical end of individual connective far exceeding lateral appendage, bearing several (ca 4–5) ± irregular short processes (sometimes almost hidden in the tomentum in young flowers); lateral appendage almost glabrous above, dark brown-violet, composed of 3 lobes—central lobe nose-shaped with irregularly dentate and straight to slightly convex margin, lateral lobes straight, with tuft of hairs on the upper margin. Style rather short but distinct (ca 0.3 mm), dark brown; stigma 3-lobed, light beige to whitish, papillose and sometimes with sparse long hairs; each lobe usually deeply divided, forming two triangular lobes or sometimes trilobed (the middle lobe being smaller), laterally adnate to the adjacent stigmatic lobes. Capsule cup-shaped, longitudinally ribbed, dark (violet-)brown to almost blackish, ca 5 mm in diameter, on elongated pedicel. Seeds light yellow-brown, ellipsoid, ca 0.3–0.4 × 0.16 mm.

Habitats: Thismia acuminata was observed in closed canopy primary lower montane tropical rainforest in humus-rich soil in a river ravine.

Distribution area: The species is known only from the type locality in the Kelabit Highlands, Sarawak, Borneo.

Proposed conservation status: The population of T. acuminata is protected against anthropogenic disturbance thanks to its location in the Pulong Tau National Park. Although the inconspicuous appearance and ephemeral growth may contribute to overlooking of the species in the field, the total population size can be estimated to < 50 mature individuals at the present state of knowledge. Therefore, we suggest evaluating the species as critically endangered (CR) based on criterion D of IUCN Red List Categories and Criteria (IUCN 2012).

Notes: Despite the fact that the observed population consisted of 8 individuals with ca three well preserved and two old flowers, quite large variation was observed in flower colour (reddish to dark brown or almost blackish), colour of stem (light brown or greyish to dark reddish brown) and shape of stigmatic lobes (bilobed with deep incision between the lobes to trilobed with middle lobe being smallest). Tomentum on connective apex varied probably due to the age of the flower.

Having coralliform roots, mitre-forming inner perianth lobes and reduced outer perianth lobes, T. acuminata clearly belongs to section Sarcosiphon. Based on outer morphology, the most similar species is probably T. laevis which differs in acute and rather low and wide mitre. From this and other related species (e.g. T. brunneomitra, T. clandestina, T. episcopalis Becc., T. nigra), T. acuminata also differs in having several small lobes and tomentum of long hairs on connective apex and nose-shaped central lobe of lateral appendage. Lobed and hairy (although described as ciliolate and minutely verrucose) connective apex is present also in the New Guinean T. versteegii, which nevertheless has different shape of lateral appendage (described as obtrapeziform with inflexed lateral margins), smaller flowers (7.25 mm long above the ovary and 4.7 mm in diameter), light yellow-orange perianth with red-brown stripes and mitre, small but apparent outer perianth lobes, rather low mitre with obtuse apex and slit-like mitre openings (which, nevertheless, may only reflect an early anthetic phase).

Thismia laevis Sochor, Dančák & Hroneš, sp. nov.—HOLOTYPE: Malaysia, Sarawak, Kelabit Highlands, Pa’Lungan village, Arur Bedalawid, 3.0 km N of the village, 1108 m a. s. l., 3°50′13″N, 115°31′11″E, 16 Jan 2017, M. Sochor et al. BOR9/17 (holotype: SAR [accession number Sochor/BOR-9/17]) (Figs. 4, 5).
Fig. 4

Thismia laevis: a habit; b detail of flower; c section of flower with outer view of stamens; d longitudinal section of connective tube; e ovary with stigma (from Sochor et al. BOR9/17). Drawn by R. Melichárková

Fig. 5

Thismia laevis: a habit; b detail of flower; c section of flower and outer view of stamens; d outer view of stamens; e section of connective tube; f upper view of mitra; g stigma; h capsule with seeds (from Sochor et al. BOR9/17)

Etymology: From the Latin laevis (smooth) referring to the smooth apex of connectives.

Diagnosis: Similar to T. episcopalis Becc., but differing in glabrous, shallowly bilobed apex of connectives, greyish perianth tube with 12 (dark) brown stripes and dark (red-)brown to blackish mitre.

Description: Achlorophyllous ground herb, ca 5.5–10.5 cm tall. Roots short, clustered, coralliform, light brown. Stem ca 13 cm long (but mostly under ground or in leaf litter), creeping, ascending to erect, sparsely branched, pale brown to pinkish; branching sympodial, with one bud covered in bracts of the flower in anthesis and growing as the main stem afterwards; pedicel very short or almost missing, elongating markedly (to ca 11 cm) after anthesis. Leaves around 4 below the first flower, spirally arranged, scale-like, triangular, acute, entire, 4–4.5 mm long and 1.5–2 mm wide at base, of the same colour as stem. Floral bracts of the same size and shape as leaves, pinkish, enveloping base of flower and a young bud. Flowers around 4 per stem, actinomorphic, 18 mm long, 7 mm wide below top of perianth tube; perianth tube narrowly urceolate with slight constriction above the middle, of 6 fused tepals; outer surface with 6 longitudinal ribs, greyish, with dark brown to blackish apex and ribs and brown stripes between the ribs (together making 12 longitudinal dark stripes); outer perianth lobes absent; inner perianth lobes bent upwards, connate at top and forming mitre with 3 holes; holes elliptic, 5.5 mm wide, 3.5 mm high; mitre with 3 distinct sutures, 3 ribs and short obtuse processes on the top, 6 mm tall, dark (red-)brown to blackish with paler apex and margins of holes; annulus formed by 12 low lobes, each consisting of several small lobes. Stamens 6, pendent from the apical margin of the perianth tube, dull orange; filaments free, short, curved downwards; connectives broad and flattened, except for lateral appendage and area around anthers glabrous, laterally connate to form a tube, ca 4 mm long; interstaminal glands not prominent, flat, oblong, ca 0.5 mm long, placed between bases of lateral appendages; apical end of individual connective shallowly bilobed, without any processes, far exceeding lateral appendage; lateral appendage large, (rather sparsely) hairy on margin, box-shaped, composed of 3 lobes—central lobe convexly down-curved and with slightly concave margin, making acute angles with lateral lobes from outer view. Style short, dark brown, stigma 3-lobed, ± smooth to smoothly papillose, each lobe with 2 large and 1 small lobe on apical margin, curved upwards and together forming triangular cup-like structure. Capsule cup-shaped, pinkish-brown, ca 7 mm in diameter, on elongated pedicel (to 11 cm long). Seeds beige, ellipsoid, ca 0.4 × 0.16 mm.

Habitats: Thismia laevis was observed in primary lower montane tropical rainforest under closed canopy in humus-rich soil in a river ravine.

Distribution area: The species is known only from two nearby localities in the Kelabit Highlands, Sarawak, Borneo.

Proposed conservation status: Population of T. laevis is protected against anthropogenic disturbance thanks to its location in the Pulong Tau National Park. Although wider occurrence of the species can be expected, the total population size can be estimated to < 50 mature individuals due to low population densities. Therefore, we suggest evaluating the species as critically endangered (CR) based on criterion D of IUCN Red List Categories and Criteria (IUCN 2012).

Additional specimens examined: Malaysia. Sarawak Kelabit Highlands, Pa’Lungan village, Arur Bedalawid, 3.4 km N of the village, 1224 m a. s. l., 3°50′26″N, 115°30′54″E, 4 Feb 2016, M. Sochor and Z. Egertová BOR-PL/16 (OL [accession number 34700]).

Notes: Based on corraliform roots, inner perianth lobes forming mitre and reduced outer perianth lobes, T. laevis can be placed in the sect. Sarcosiphon. Morphologically most similar species is T. clandestina which was described from West Java in 1850 based on a fruiting individual (Blume 1850). Our current understanding of the species’ floral morphology therefore comes from later descriptions of plants from the type area (Smith 1911) and Southern Thailand (Chantanaorrapint et al. 2015, but see Suetsugu et al. 2017). Both of these populations share several characteristic traits with T. laevis, such as overall habitus, connective apex far exceeding lateral appendage and 3-lobed stigmatic lobes with the middle lobe smaller (not clear for the Javan plants). Nonetheless, they differ in having 3-lobed, hairy connective apex, straight margin of lateral appendage and acuminate (rather than acute) and rather narrow mitre. While the plants from Java do not differ from T. laevis in perianth colour (greyish with dark brown to blackish stripes), Thai plants are orange-brown. Another similar species is T. episcopalis. Both species share similar shape o mitre (although more obtuse in T. episcopalis), but T. episcopalis differs clearly in shape of connective apex (hairy with 3 teeth), flower colour (orange-yellow) and shape of perianth tube (rather short urceolate without constriction above the middle). T. acuminata differs mainly in having several small tomentose lobes on connective apex, nose-shaped brown-violet central lobe of lateral appendage and narrow, tall, acuminate mitre. Superficially similar is also recently described Australian T. lanternatus W.E.Cooper, particularly in colour of perianth tube, mitre and connectives, and in position and shape of lateral appendage. Nevertheless, it differs, e.g. in underground stem, flattened mitre, connectives with 3- or 4-toothed apices and, importantly, in vermiform roots (Cooper 2017).

Thismia nigra Dančák, Hroneš & Sochor, sp. nov.—HOLOTYPE: Malaysia, Sarawak, Kelabit Highlands, Pa’Lungan village, Arur Bedalawid, 3.4 km N of the village, 1224 m a. s. l., 3°50′26″N, 115°30′54″E, 15 Jan 2017, M. Sochor et al. BOR8/17 (holotype: SAR [accession number Sochor/BOR-8/17]; isotype: OL [accession number 35274]) (Figs. 6, 7).
Fig. 6

Thismia nigra: a habit; b detail of flower with annulus; c outer view of stamens; d section of connective tube (a from Sochor et al. BOR14/17 & BOR15/17; b from Sochor et al. BOR19/17; c, d from Sochor et al. BOR8/17). Drawn by R. Melichárková

Fig. 7

Thismia nigra: a, b, c habit; d stigma; e outer view of stamens; f section of connective tube (a from Sochor et al. BOR19/17; b from Sochor et al. BOR15/17; c–f from Sochor et al. BOR8/17)

Etymology: From the Latin nigrum (black) referring to the colour of the perianth.

Diagnosis: Similar to T. brunneomitra Hroneš, Kobrlová & Dančák, but with smaller flowers (10–12 mm long vs. 16–17 mm), apex of connective with two small lobes and one tuft of long simple hairs, and lateral appendage skirt-like with concave curved lower margin.

Description: Achlorophyllous ground herb, ca 2.5–7 cm tall. Roots short, clustered, coralliform, light brown to orange-brown. Stem 0.9–8 cm tall, usually erect, sometimes basal part creeping in leaf litter, sparsely sympodially branched or unbranched, with 1–4 flowers, glabrous, pale brown to dark reddish brown or greyish; pedicel of the same colour as stem, indiscernible or very short during anthesis, elongating markedly (up to ca 9 cm) after anthesis. Leaves 3–6 per branch, spirally arranged, scale-like, triangular, acute, entire, 3–4 mm long, ca 2 mm wide at base, pale brown or greyish-brown to reddish brown. Floral bracts 3, similar to leaves but slightly longer (3.5–5 mm), sometimes keeled and sparsely dentate, enveloping base of flower. Flowers solitary or usually 2–5 per stem, bisexual, actinomorphic, 10–12 mm long, 4.5–6 mm wide below the top of perianth tube; perianth tube urceolate, of 6 fused tepals; outer surface with 6 (or sometimes more) longitudinal ribs, dark greenish-grey with 1 transversal and 12 vertical blackish stripes; outer tepal lobes absent; inner tepal lobes well developed, bent upwards, connate at top and forming mitre with 3 holes, 3–3.7 mm tall; holes elliptic to almost rounded, 1–2 mm wide, ca 2.5 mm high; mitre dark brown to blackish with greyish base, usually topped by 3 short obtuse processes or rarely rounded on top. Annulus not prominent, blackish, ca 0.9 mm thick and with opening ca 1.3–1.8 mm in diameter, flat with raised inner margin formed by six low rectangular lobes. Stamens 6, hanging from top of perianth tube (outer margin of the annulus), blue; filaments flat and rather thin (ca 0.6 mm wide, 40–60% of the width of connective), hairy in lower part; connectives flattened, connate to form tube; interstaminal glands flat, circular, ca 0.32 mm in diameter, placed between lateral appendages; individual connectives shortly hairy in the apical quarter, with 2 small lobes (sometimes not prominent) and several (usually around 5) long simple transparent hairs in the middle of the apex; lateral appendage exceeding the connective apex, skirt-like, with concavely curved lower margin and 2 rather small lateral lobes, with long hairs on margins. Ovary inferior, cup-shaped, blackish in upper part and pale otherwise, longitudinally ribbed, often covered in bracts; stigma 3-lobed, each lobe deeply divided into 2 long triangular lobes and laterally adnate to the neighbouring lobes, papillose. Capsule cup-shaped to funnel-shaped, ca 6 mm in diameter, pinkish-brown to dark brown.

Habitats: All populations were found in primary lower montane tropical rainforest in elevation above 1200 m a.s.l. near streams or rivers, usually around the line of maximum water level, sometimes also in fresh fluvial sediments, in sandy, clayey and humus-rich soils. All populations were accompanied by other mycoheterotrophic species, such as Epirixanthes spp., Sciaphila spp., Thismia viridistriata and Thismia laevis.

Distribution area: The species is known from five nearby localities in the Pulong Tau National Park, Kelabit Highlands, Sarawak.

Additional specimens examined: Malaysia. Sarawak Kelabit Highlands, Pa’Lungan village, Arur Dutu, 5.6 km N of the village, 1208 m a. s. l., 3°51′40″N, 115°31′22″E, 19 Jan 2017, M. Sochor et al. BOR14/17 (SAR [accession number Sochor/BOR-14/17]); Kelabit Highlands, Pa’Lungan village, Arur Dutu, 5.7 km N of the village, 1210 m a. s. l., 3°51′43″N, 115°31′24″E, 19 Jan 2017, M. Sochor et al. BOR15/17 (SAR [accession number Sochor/BOR-15/17]); Kelabit Highlands, Pa’Lungan village, Arur Dutu, 5.8 km N of the village, 1210 m a. s. l., N3°51′46″, E115°31′26″, 20 Jan 2017, M. Sochor et al. BOR19/17 (SAR [accession number Sochor/BOR-19/17]); Kelabit Highlands, Pa’Lungan village, Arur Dutu, 5.9 km N of the village, 1210 m a. s. l., 3°51′48″N, 115°31′28″E, 20 Jan 2017, M. Sochor et al. BOR46/17.

Proposed conservation status: In the studied area, the species seems to occur locally more commonly than is usual in most other Thismia species and is under protection of the Pulong Tau National Park. The fact that it is so far known only from this small region, nevertheless, could be attributed to its possible stenoendemism. Considering its low population densities (usually only one or a few individuals), the total population size can be estimated to < 250 mature individuals at the present state of knowledge. Therefore, we suggest evaluating T. nigra as endangered (EN) based on criterion D of IUCN Red List Categories and Criteria (IUCN 2012).

Notes: Similarly to the previous species described here, T. nigra belongs to section Sarcosiphon. Based on the morphology, the closest relative is undoubtedly T. brunneomitra which is known from Brunei, less than 90 km from the localities of T. nigra. Both species exhibit similar overall appearance (habitus, flower shape and colour), and even colour of connectives. Nevertheless, they differ in shape of connectives, flower size and number and also ecology, among other traits that may vary. Whereas T. brunneomitra has three large lobes on connective apex, which are ended by glandular hairs, T. nigra has only two small lobes in the centre of connective apex with only one tuft of a few long, simple, straight or hooked hairs. Shape of lateral appendage also differs in the two species (compare Figs. 6c, 7e and Hroneš et al. 2015). Flowers of the former species are in number of 1–2, ca 1.7 × 1 cm long, but those of T. nigra are usually 2–5 and smaller (ca 1.1 × 0.5 cm). Also, stigma in T. brunneomitra is shallowly notched whereas in T. nigra is deeply divided, although this trait may be generally variable within Thismia species. And although both species were found always on river banks in primary rain forest, the locality of T. brunneomitra is in lowland Dipterocarp forest ca 105 m a.s.l., whereas T. nigra grows in lower montane forest > 1200 m a.s.l. Differentiation of both species was further detected in molecular markers, particularly LSU (Fig. 1; Online Resource 2).

Another similar species, T. brunneomitroides Suetsugu & Tsukaya, described recently from Thailand, has ivory-coloured perianth, two teeth with glandular hairs on apical margin of the connective and convex margin of lateral appendage. T. episcopalis, known from western Sarawak, differs in having more flowers per stem (up to 7), orange-yellow flowers, taller mitre (ca 5 mm), no projections on mitre, 3 teeth on connective apex and lateral appendage not reaching connective apex. Three other similar species, T. laevis, T. acuminata and T. clandestina differ in the connective apex that is of different shape and far exceeding lateral appendage, in trilobed stigmatic lobes with middle lobe being smallest, in the absence of projections on mitre (only in T. laevis very small projections are present) and usually in colour and shape of flowers.

Thismia viridistriata Sochor, Hroneš & Dančák, sp. nov.—HOLOTYPE: Malaysia, Sarawak, Kelabit Highlands, Pa’Lungan village, Arur Bedalawid, 3.2 km N of the village, 1182 m a. s. l., 3°50′20″N, 115°31′01″E, 16 Jan 2017, M. Sochor et al. BOR11/17 (holotype: SAR [accession number Sochor/BOR-11/17]) (Figs. 8, 9).
Fig. 8

Thismia viridistriata: a habit; b, c flowers of two different morphotypes; d outer view of stamens; e section of floral tube (a from Sochor et al. BOR11/17; b from Sochor et al. BOR7/17; c from Sochor et al. BOR3/17; d from Sochor et al. BOR13/17; e from Sochor et al. BOR23/17). Drawn by R. Melichárková

Fig. 9

Thismia viridistriata: a, b two different morphotypes; c plant with typical green-striped ovary; d detail view of floral inner structure—stigma, vertical ribs on perianth tube and longitudinal section of connective tube; e outer-bottom view of stamens; f section of connective tube; g top of mitre; h example of the species’ habitat diversity—a plant growing on a bare vertical rock (a, c from Sochor et al. BOR3/17; b, f, g from Sochor et al. BOR7/17; d, h from Sochor et al. BOR23/17; e from Sochor et al. BOR11/17)

Etymology: From the Latin viridis (green) and striae (stripes) referring to the green longitudinal stripes on ovary, sometimes very prominent.

Diagnosis: Thismia viridistriata differs from congeneric species in having the following combination of morphological traits: coralliform roots, inner perianth lobes forming mitre, mitre topped by a single cylindrical column or three short obtuse processes, prickle-like protrusions on ovary and perianth, greenish stripes on ovary, vertical ribs partly detached from inner surface of perianth, and apex of connectives with many short hairs and one tuft of several long hairs.

Description: Achlorophyllous ground herb, ca 2–4.5(–9) cm tall. Roots short, densely clustered, coralliform, light beige, forming a dense, spheroidal, tuber-like system, 1–2.5 cm in diameter. Stem 0.5–3(–8) cm long, creeping, ascending or erect, simple or sparsely branched, brown to whitish; branching sympodial, branches short (usually < 1 cm); pedicels indiscernible or very short during anthesis, elongating markedly during fruit maturing (up to ca 6 cm). Leaves usually 2–6, spirally arranged, scale-like, triangular, acute, entire, 3–6 mm long and 1.5–2 mm wide at base, of the same colour as stem. Floral bracts 3, of the same size and appearance as leaves. Flowers 1–5 per stem, actinomorphic, 13–21 mm long, 4.5–6 mm wide in the widest point; perianth tube urceolate (widest in the middle or upper third); outer surface with 6 verrucose longitudinal ribs, greyish to rich brown or reddish brown with darker transversal stripe in the middle and 12 dark brown (to almost blackish) or reddish brown longitudinal stripes; perianth inside with 12 transparent vertical ribs partly detached from its surface, most prominent in lower third to half (below connectives); outer perianth lobes usually absent or very short, obtusely triangular; inner perianth lobes bent upwards, connate at top and forming mitre with 3 holes; holes discorectangular or ellipsoid, 1.2–2 × 2–3.5 mm; mitre ± hemispherical, almost smooth, verrucose or covered by short prickle-like protrusions, especially on the 3 central ribs, top of mitre with a single cylindrical, up to ca 6 mm long column (usually dark brown) or with 3 short protrusions (usually dark pink to reddish), basal part of mitre pinkish or light brown to greyish, upper part darker (reddish or dark brown to blackish); annulus thin, often markedly raised, divided into 6 rectangular lobes. Stamens 6, pendent from the apical margin of the perianth tube; filaments free, curved downwards, glabrous adaxially, covered by long straight hairs on margins abaxially, dark violet to dark reddish brown; connectives broad and flattened, laterally connate to form a tube, ca 2–2.5 mm long, adaxially, apically and abaxially below lateral appendage densely covered by short transparent hairs, greyish-green (especially in section) or sometimes pinkish; interstaminal glands not prominent, ± flat, slightly oblong, ca 0.45 mm long, placed between lateral appendages and sometimes hidden under their lateral lobes; apical end of individual connective with 1 wide lobe (sometimes divided by notch into 2 short lobes) bearing several (usually 5) transparent, ca 0.15–0.2 mm long hairs; lateral appendage not reaching the connective apex, distinctly 3-lobed (central lobe being smallest), densely hairy on margins. Ovary inferior, cup-shaped, light brown with 12 grey-green to dark green stripes, usually covered by long prickle-like protrusions; stigma 3-lobed, individual lobes shallowly bilobed, curved upwards and together forming triangular cup-like structure. Capsule cup-shaped, 5–9 mm in diameter, pinkish, light brown to greyish-brown, usually with darker longitudinal stripes. Seeds beige, ellipsoid, ca 0.5–0.75 × 0.2–0.3 mm.

Habitats: Eleven of the twelve known localities are placed in primary lower montane tropical rainforest in elevation of ca 1100 to 1200 m a.s.l. One rich population was found on a borderline between primary forest and old forest clearing (possibly former pastures or fields, now overgrown by Pteridium aquilinum). All individuals were discovered near streams or rivers, usually just above the line of maximum water level, but sometimes also in fresh fluvial sediments or on drier slopes up to ca 20 m above the river; in sand, humus-rich soils or even on almost barren rocks. At most localities, it grew together with other mycoheterotrophic taxa, mainly Epirixanthes kinabaluensis, Sciaphila spp., Exacum tenue, Burmannia lutescens agg. and other Thismia species.

Distribution area: The species is known from 11 localities within an area of ca 10 × 18 km in the Kelabit Highlands, Borneo, Malaysia.

Proposed conservation status: Although T. viridistriata seems to be one of the most common Thismia species in the Kelabit Highlands, it is known from just a small geographic area (minimal extent of occurrence 121 km2 and minimal area of occurrence ca 20 km2). Also, the fact that it had not been detected earlier may imply its limited distribution area or its rare occurrence elsewhere. Four of the detected populations are furthermore unprotected from human activities, and one of them has already probably vanished due to road construction between Bario and Ba’kelalan. The region experienced an extensive logging and development in the last decade, which very likely influenced the total population size and further loss of populations can be expected in the future outside the Pulong Tau National Park. Therefore, we suggest evaluating the species as endangered (EN) based on criterion C2ai of IUCN Red List Categories and Criteria (IUCN 2012).

Additional specimens examined: Malaysia. Sarawak Kelabit Highlands, Pa’Umor village, Anak Kadi Ridge, 4.4 km SSE of the village, 1195 m a. s. l., 3°42′1″N, 115°31′28″E, 13 Jan 2017, M. Sochor et al. BOR3/17 (SAR [accession number Sochor/BOR-3/17]); Kelabit Highlands, Pa’Lungan village, Arur Bedalawid, 3.1 km N of the village, 1175 m a. s. l., 3°50′18″N, 115°31′3″E, 15 Jan 2017, M. Sochor et al. BOR7/17 (SAR [accession number Sochor/BOR-7/17]); Kelabit Highlands, Pa’Lungan village, Arur Bedalawid, 3.0 km N of the village, 1164 m a. s. l., 3°50′15″N, 115°31′16″E, 16 Jan 2017, M. et al. BOR12/17 (SAR [accession number Sochor/BOR-12/17]); Kelabit Highlands, Pa’Lungan village, Arur Dutu, 5.5 km N of the village, 1206 m a. s. l., 3°51′38″N, 115°31′21″E, 19 Jan, M. Sochor et al. BOR12/17 (SAR [accession number Sochor/BOR-12/17], OL [accession number 35266]); Kelabit Highlands, Long Rebpun (Rapung), Pa Tabanul, 6.1 km NE of Pa’Lungan village, 1093 m a. s. l., 3°51′6″N, 115°33′26″E, 18 Jan 2017, M. Sochor et al. BOR16/17 (SAR [accession number Sochor/BOR-16/17]); Kelabit Highlands, Pa’Lungan village, Upper Pa’Lungan river, 5.7 km N of the village, 1200 m a. s. l., 3°51′43″N, 115°31′20″E, 20 Jan 2017, M. Sochor et al. BOR17/17 (SAR [accession number Sochor/BOR-17/17]); Kelabit Highlands, Pa’Lungan village, a ridge between Upper Pa’Lungan river and Arur Dutu, 5.8 km N of the village, 1230 m a. s. l., 3°51′45″N, 115°31′23″E, 20 Jan 2017, M. Sochor et al. BOR18/17 (SAR [accession number Sochor/BOR-18/17]); Kelabit Highlands, Bario, Arur Dalan village, 0.9 km WSW of the longhouse, 1120 m a. s. l., 3°45′8″N, 115°26′16″E, 22 Jan 2017, M. Sochor and Z. Egertová BOR23/17 (SAR [accession number Sochor/BOR-23/17]); Kelabit Highlands, Pa’Umor village, Pa’Pidap river catchment, 3.1 km SSE of the village, 1135 m a. s. l., 3°42′41″N, 115°30′57″E, 7 Feb 2016, M. Dančák et al. MD2016/431 (OL [accession number 355512]).

Notes: Compared to other known Thismia species, T. viridistriata exhibits an extremely high variability in outer macromorphological characters. The greatest variation was observed in colour of flower, particularly mitre (from pink through brown to greyish-brown) and in the presence of cylindrical projection on top of mitre and its length (from absent to ca 6 mm). Outer perianth lobes are sometimes well developed but are frequently lacking with no relation to the presence or absence of the mitre column. Some variation was observed also in shape of perianth tube (widest in the middle or in upper third), flower size (13–21 mm, incl. mitre), number of stems from one root (1–6), number of flowers per stem (1–5) and stem length (0.5–8 cm). Other traits, such as presence and shape of prickle-like processes on flowers, colour of connectives (greyish-green to rather pinkish) or bilobed connective apex, can be partly attributed to flower age and preservation.

As a mitriform species with distinct apical column on the top of the mitre in some individuals, T. viridistriata could be placed into the genus Scaphiophora, together with T. appendiculata Schltr. and T. gigantea (Jonker) Hroneš. However, this genus is now generally reduced to Thismia (e.g. Maas et al. 1986; Merckx et al. 2013a; Kumar et al. 2017). The example of Thismia viridistriata clearly shows that this approach is correct because the main diagnostic trait, i.e. the presence of the apical column, may vary even within the species. This also brings into question the existence of Scaphiohora even as a section of Thismia. With coralliform roots, inner perianth lobes forming mitre and outer perianth lobes reduced or lacking, T. viridistriata (as well as the other species of Scaphiophora) can well be placed into section Sarcosiphon. T. viridistriata differs from members of this section by, e.g. the presence of cylindrical column on top of mitre, prickle-like protrusions on flowers (especially on ovary), vertical ribs partly detached from the inner surface of perianth tube, and shape of connectives. Individuals that lack the mitre column may be superficially similar to T. brunneomitra, T. nigra and T. acuminata which differ (beside the above mentioned) in having dark brown to blackish flowers, narrowly urceolate perianth tube, narrower mitre and ± smooth outer surface of flower. In our phylogenetic analysis, T. viridistriata forms a distinct, basal lineage within the Sarcosiphon clade (Fig. 1) and the morphological similarity with the above-mentioned species may therefore reflect either convergent evolution, or limited morphological anagenesis in both lineages.

Identification key of Thismia sect. Sarcosiphon (incl. Scaphiophora and Geomitra, excl. Indochinese species with vermiform roots)

1a. Mitre at its top with 1–3 appendages, at least 2 mm long … 2

1b. Mitre at its top without any appendages or with very short, up to 2 mm long processes … 8

2a. Mitre at its top with 3 appendages … 3

2b. Mitre at its top with 1 appendage (column), sometimes trilobed at its top … 4

3a. Perianth yellow-orange to pink-red; connective apex triangular, acute … T. clavigera

3b. Perianth blue-green; connective apex rounded to slightly emarginate … T. betung-kerihunensis

4a. Appendage at top of mitre 2–6 mm long … 5

4b. Appendage at top of mitre at least 6 mm long … 6

5a. Perianth tube verrucose, greyish to rich brown or reddish brown; outer perianth lobes absent or very short, obtusely triangular; connective apex with 1 wide lobe (sometimes divided by a notch into 2 short lobes) bearing several short hairs … T. viridistriata

5b. Perianth tube smooth, yellowish white; outer perianth lobes small, ear shaped; connective apex with 1 filiform median appendage…T. appendiculata

6a. Mitre appendage dark brown to blackish, apically distinctly trilobed, the lobes acute, mitre black … T. hawkesii

6b. Mitre appendage orange to red, apically clavate to very shallowly trilobed, the lobes obtuse; mitre yellow, orange or pink … 7

7a. Mitre fleshy, its margin reflexed and extending downwards over the upper perianth tube, covering their openings; mitre appendage clavate; connective apex without appendages … T. tectipora

7b. Mitre slender, with large distinct openings; mitre appendage shallowly trilobed at its apex; connective apex with 5 appendages … T. gigantea

8a. Outer perianth lobes present, distinct … 9

8b. Outer perianth lobes absent or indistinct … 11

9a. Mitre thick, flashy; connective apex bilobed; mitre purplish brown … T. yorkensis

9b. Mitre thin, slender; connective apex with one central lobe and two smaller lateral lobes; mitre yellowish or bluish … 10

10a. Outer perianth lobes deeply dentate, bright yellow; mitre elevated by three filiform pillars … T. kelabitiana ined.

10b. Outer perianth lobes smooth with a single central short tooth, blue; mitre not elevated by filiform pillars … T. goodii

11a. Annulus and stamens sunken inside perianth tube … T. crocea

11b. Annulus and stamens positioned at the perianth tube opening not sunken inside perianth tube … 12

12a. Ovary and perianth tube covered by prickle-like protrusions; vertical ribs elevated and partly detached from inner surface of perianth tube … T. viridistriata

12b. Ovary and perianth tube verrucose to smooth; vertical ribs not distinctly elevated on inner surface of perianth tube … 13

13a. Lateral appendage of connective clearly exceeding connective apex … 14

13b. Lateral appendage of connective not exceeding connective apex … 15

14a. Flowers 16–17 mm long; connective apex with 2 wide, obtusely triangular lateral lobes and narrowly triangular middle lobe, each lobe with tuft of glandular hairs … T. brunneomitra

14b. Flowers 10–12 mm long; connective apex with 2 small lobes and several long simple transparent hairs in the middle of the apex … T. nigra

15a. Connective apex smooth, without any hairs, teeth or processes … T. laevis

15b. Connective apex at least hairy or dentate … 16

16a. Mitre apex with 3 short acute processes; ovary densely verrucose; perianth tube ivory … T. brunneomitroides

16b. Mitre apex without any processes; ovary glabrous or sparsely verrucose; perianth tube yellow-orange to brown … 17

17a. Connective densely hairy to tomentose, bearing ca 4–5 irregular short processes at apex … T. acuminata

17b. Connective glabrous to sparsely hairy, bearing at most 3 processes at apex … 18

18a. Connective apex shallowly trilobed or obtuse … T. clandestina

18b. Connective apex with 3 sharp triangular teeth … T. episcopalis

Notes

Acknowledgements

We thank John Rian Pasan and David Attu for guiding services, Nur Afiza Binti Umar (Sarawak Forestry Department) for providing us with all the permits, Zuzana Egertová for field assistance and staff of SAR herbarium, particularly Nur Safinas Binti Jelani, for kind cooperation. We are also grateful to Renata Melichárková for line drawings. MH was supported by the project no. IGA PrF-2018-001 from the Internal Grant Agency of the Palacký University, and MS was supported by Grant No. LO1204 (Sustainable development of research in the Centre of the Region Haná) from the National Program of Sustainability I, MEYS. The research was conducted under the permit No. NCCD.907.4.4(JLD.13)-337 issued by Sarawak Forestry Department.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.

Supplementary material

606_2018_1504_MOESM1_ESM.pdf (465 kb)
Supplementary material 1 (PDF 464 kb)
606_2018_1504_MOESM2_ESM.pdf (19 kb)
Supplementary material 2 (PDF 19 kb)
606_2018_1504_MOESM3_ESM.pdf (726 kb)
Supplementary material 3 (PDF 725 kb)
606_2018_1504_MOESM4_ESM.txt (406 kb)
Supplementary material 4 (TXT 405 kb)

References

  1. Bidartondo MI (2005) The evolutionary ecology of myco-heterotrophy. New Phytol 167:335–352.  https://doi.org/10.1111/j.1469-8137.2005.01429.x CrossRefPubMedGoogle Scholar
  2. Blume CL (1850) Burmanniaceae. Mus Bot Lugduno-Batavum 1(5):65–80Google Scholar
  3. Chantanaorrapint S, Suddee S (2018) Thismia thaithongiana (Dioscoreaceae: Thismieae), a new species of mycoheterotroph from an unusual habitat. Phytotaxa 333:287–292.  https://doi.org/10.11646/phytotaxa.333.2.14 CrossRefGoogle Scholar
  4. Chantanaorrapint S, Tetsana N, Sridith K (2015) Notes on Thismia clandestina (Thismiaceae), a little-known mycoheterotrophic species. Polish Bot J 60:71–74.  https://doi.org/10.1515/pbj-2015-0013 Google Scholar
  5. Chantanaorrapint S, Wai JS, Poopath M, Tetsana N (2016) Thismia claviformis (Thismiaceae), a new species from the Thai-Malay Peninsula. Phytotaxa 280:292–296.  https://doi.org/10.11646/phytotaxa.280.3.8 CrossRefGoogle Scholar
  6. Chiang P-Y, Hsieh T-H (2011) Thismia huangii (Thismiaceae), a new species from Taiwan. Taiwania 56:138–142Google Scholar
  7. Cooper WE (2017) Thismia hawkesii W.E.Cooper and T. lanternatus W.E.Cooper (Thismiaceae), two new fairy lantern species from the Wet Tropics Bioregion, Queensland, Australia. Austrobaileya 10:130–138Google Scholar
  8. Dančák M, Hroneš M, Sochor M, Kobrlová L, Hédl R, Hrázský Z, Vildomcová A, Sukri RS, Metali F (2013) A new species of Thismia (Thismiaceae) from Brunei Darussalam, Borneo. Phytotaxa 125:33–39.  https://doi.org/10.11646/phytotaxa.125.1.5 CrossRefGoogle Scholar
  9. Darriba D, Taboada GL, Doallo RR, Posada D (2012) jModelTest 2: more models, new heuristics and parallel computing. Nat Meth 9:772.  https://doi.org/10.1038/nmeth.2109 CrossRefGoogle Scholar
  10. Doyle JJ, Doyle JL (1987) A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochem Bull 19:11–15Google Scholar
  11. Eyre-Walker A, Gaut BS (1997) Correlated rates of synonymous site evolution across plant genomes. Molec Biol Evol 14:455–460.  https://doi.org/10.1093/oxfordjournals.molbev.a025781 CrossRefPubMedGoogle Scholar
  12. Govaerts R, Wilkin P, Saunders RM (2007) World checklist of Dioscoreales: yams and their allies. Royal Botanic Gardens, KewGoogle Scholar
  13. Hroneš M (2016) Genus Thismia Griff. (Thismiaceae) in Asia and Australasia with focus on the Malesian floristic region. Palacky University, OlomoucGoogle Scholar
  14. Hroneš M, Kobrlová L, Taraška V, Popelka O, Hédl R, Sukri RS, Metali F, Dančák M (2015) Thismia brunneomitra, another new species of Thismia (Thismiaceae) from Ulu Temburong, Brunei Darussalam. Phytotaxa 234:172–178.  https://doi.org/10.11646/phytotaxa.234.2.7 CrossRefGoogle Scholar
  15. Hroneš M, Rejžek M, Sochor M, Svátek M, Kvasnica J, Egertová Z, Pereira JT, Nilus R, Dančák M (2018) Two new species of Thismia subsect. Odoardoa (Thismiaceae) from Borneo. Pl Ecol Evol (in press).  https://doi.org/10.5091/plecevo.2018.1387
  16. Hunt CA, Steenbeeke G, Merckx VSFT (2014) Thismia megalongensis (Thismiaceae), a new species of Thismia from New South Wales. Telopea 16:165–174.  https://doi.org/10.7751/telopea20147809 Google Scholar
  17. Imhof S, Massicotte HB, Melville LH, Peterson RL (2013) Subterranean morphology and mycorrhizal structures. In: Merckx VSFT (ed) Mycoheterotrophy. Springer, New York, pp 157–214CrossRefGoogle Scholar
  18. IUCN (2012) and Criteria: version 3.1, 2nd edn. IUCN, Gland. Available at: http://www.iucnredlist.org/technical-documents/categories-and-criteria
  19. Jarvie JK (1996) Thismia lauriana (Burmanniaceae), a new species from Central Kalimantan. Blumea 41:57–259Google Scholar
  20. Jonker FP (1938) A monograph of the Burmanniaceae. Meded Bot Mus Herb Rijks Univ Utrecht 51:1–279Google Scholar
  21. Jonker FP (1948) Burmanniaceae. In: van Steenis CGGJ (ed) Flora Malesiana I, 4. Noordhoff-Kolf, Leiden, pp 13–26Google Scholar
  22. Kay KM, Sargent RD (2009) The role of animal pollination in plant speciation: integrating ecology, geography, and genetics. Annual Rev Ecol Evol Syst 40:637–656.  https://doi.org/10.1146/annurev.ecolsys.110308.120310 CrossRefGoogle Scholar
  23. Kiew R (1999) Thismia goodii (Burmanniaceae), the Blue-capped Thismia, a new species from Borneo. Gard Bull Singapore 51:179–182Google Scholar
  24. Kovarik A, Pires JC, Leitch AR, Lim KY, Sherwood AM, Matyasek R, Rocca J, Soltis DE, Soltis PS (2005) Rapid concerted evolution of nuclear ribosomal DNA in two Tragopogon allopolyploids of recent and recurrent origin. Genetics 169:931–944.  https://doi.org/10.1534/genetics.104.032839 CrossRefPubMedPubMedCentralGoogle Scholar
  25. Kumar P, Gale SW, Li J, Bouamanivong S, Fischer GA (2017) Thismia nigricoronata, a new species of Burmanniaceae (Thismieae, Dioscoreales) from Vang Vieng, Vientiane Province, Laos, and a key to subgeneric classification. Phytotaxa 319:225–240.  https://doi.org/10.11646/phytotaxa.319.3.2 CrossRefGoogle Scholar
  26. Kuzoff RK, Sweere JA, Soltis DE, Soltis PS, Zimmer EA (1998) The phylogenetic potential of entire 26S rDNA sequences in plants. Molec Biol Evol 15:251–263.  https://doi.org/10.1093/oxfordjournals.molbev.a025922 CrossRefPubMedGoogle Scholar
  27. Lanfear R, Frandsen PB, Wright AM et al (2017) PartitionFinder 2: new methods for selecting partitioned models of evolution for molecular and morphological phylogenetic analyses. Molec Biol Evol 34:772–773.  https://doi.org/10.1093/molbev/msw260 PubMedGoogle Scholar
  28. Maas PJM, Maas-van de Kamer H, van Benthem J, Snelders HCM, Rübsamen T (1986) Burmanniaceae. Fl Neotrop Monogr 42:1–189Google Scholar
  29. Mar SS, Saunders R (2015) Thismia hongkongensis (Thismiaceae): a new mycoheterotrophic species from Hong Kong, China, with observations on floral visitors and seed dispersal. PhytoKeys 46:21–33.  https://doi.org/10.3897/phytokeys.46.8963 CrossRefGoogle Scholar
  30. Meng S-W, Chen Z-D, Li D-Z, Liang H-X (2002) Phylogeny of Saururaceae based on mitochondrial matR gene sequence data. J Pl Res 115:71–76.  https://doi.org/10.1007/s102650200011 CrossRefGoogle Scholar
  31. Merckx VSFT, Schols P, Maas-van de Kamer H, Maas P, Huysmans S, Smets E (2006) Phylogeny and evolution of Burmanniaceae (Dioscoreales) based on nuclear and mitochondrial data. Amer J Bot 93:1684–1698CrossRefGoogle Scholar
  32. Merckx VSFT, Janssens SB, Hynson NA et al (2012) Mycoheterotrophic interactions are not limited to a narrow phylogenetic range of arbuscular mycorrhizal fungi. Molec Ecol 21:1524–1532.  https://doi.org/10.1111/j.1365-294X.2012.05472.x CrossRefGoogle Scholar
  33. Merckx VSFT, Freudenstein JV, Kissling J, Christenhusz MJM, Stotler RE, Crandall-Stotler B, Wickett N, Rudall PJ, Maas-van de Kamer H, Maas PJM (2013a) Taxonomy and classification. In: Merckx VSFT (ed) Mycoheterotrophy. Springer, New York, pp 19–101CrossRefGoogle Scholar
  34. Merckx VSFT, Smets EF, Specht CD (2013b) Biogeography and conservation. In: Merckx VSFT (ed) Mycoheterotrophy. Springer, New York, pp 103–156CrossRefGoogle Scholar
  35. Merckx VSFT, Gomes SIF, Wapstra M, Hunt C, Steenbeeke G, Mennes CB, Walsh N, Smissen R, Hsieh TH, Smets Bidartondo MI (2017) The biogeographical history of the interaction between mycoheterotrophic Thismia (Thismiaceae) plants and mycorrhizal Rhizophagus (Glomeraceae) fungi. J Biogeogr 44:1869–1879.  https://doi.org/10.1111/jbi.12994 CrossRefGoogle Scholar
  36. Ronquist F, Teslenko M, van der Mark P, Ayres DL, Darling A, Hohna S, Larget B, Liu L, Suchard MA, Huelsenbeck JP (2012) MrBayes 3.2: efficient Bayesian phylogenetic inference and model choice across a large model space. Syst Biol 61:539–542.  https://doi.org/10.1093/sysbio/sys029 CrossRefPubMedPubMedCentralGoogle Scholar
  37. Schlechter FRR (1921) Die Thismieae. Notizbl Königl Bot Gart Berlin 8:31–45Google Scholar
  38. Smith JJ (1911) Beiträge zur Kenntnis der Saprophyten Javas. IV. Zur Systematik von Thismia clandestina Miq. und Thismia Versteegii J. J. Smith. Ann Jard Bot Buitenzorg 24:55–59Google Scholar
  39. Sochor M, Vašut RJ, Sharbel TF, Trávníček B (2015) How just a few makes a lot: speciation via reticulation and apomixis on example of European brambles (Rubus subgen. Rubus, Rosaceae). Molec Phylogen Evol 89:13–27.  https://doi.org/10.1016/j.ympev.2015.04.007 CrossRefGoogle Scholar
  40. Sochor M, Sukri RS, Metali F, Dančák M (2017) Thismia inconspicua (Thismiaceae), a new mycoheterotrophic species from Borneo. Phytotaxa 295:263–270.  https://doi.org/10.11646/phytotaxa.295.3.7 CrossRefGoogle Scholar
  41. Suetsugu K, Tsukaya H, Tagane S, Suddee S, Rueangruea S, Yahara T (2017) Thismia brunneomitroides (Thismiaceae), a new mycoheterotrophic species from southern Thailand. Phytotaxa 314:103–109.  https://doi.org/10.11646/phytotaxa.314.1.9 CrossRefGoogle Scholar
  42. Suetsugu K, Tsukaya H, Tagane S et al (2018) Flora of Bokor National Park VII: Thismia bokorensis (Burmanniaceae), a new species representing a new generic record. Phytotaxa 334:65–69.  https://doi.org/10.11646/phytotaxa.334.1.10 CrossRefGoogle Scholar
  43. Sujanapal P, Robi AJ, Dantas KJ, Sumod M, Merckx VSFT (2017) Thismia (Thismiaceae): the first record of the mycoheterotrophic genus to the Flora of India with a new species revealing the phytogeographical significance of Western Ghats. Blumea 62:97–102.  https://doi.org/10.3767/blumea.2017.62.2.04 CrossRefGoogle Scholar
  44. 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. Molec Biol Evol 28:2731–2739.  https://doi.org/10.1093/molbev/msr121 CrossRefPubMedPubMedCentralGoogle Scholar
  45. Tsukaya H, Okada H (2012) A new species of Thismia (Thismiaceae) from West Kalimantan, Borneo. Syst Bot 37:53–57.  https://doi.org/10.1600/036364412X616639 CrossRefGoogle Scholar
  46. Tsukaya H, Suleiman M, Okada H (2014) A new variety of Thismia hexagona Dančák, Hroneš, Koblová et Sochor (Thismiaceae) from Sabah, Borneo, Malaysia. Acta Phytotaxon Geobot 65:141–145Google Scholar
  47. Tsukaya H, Suetsugu K, Suleiman M (2017) Thismia bryndonii (Thismiaceae), a new species from Maliau Basin, Sabah, Borneo. Phytotaxa 312:135–138.  https://doi.org/10.11646/phytotaxa.312.1.13 CrossRefGoogle Scholar
  48. White TJ, Bruns T, Lee S, Taylor JW (1990) Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis MA, Gelfand DH, Sninsky JJ, White TJ (eds) PCR protocols: a guide to methods and applications. Academic Press, New York, pp 315–322Google Scholar

Copyright information

© Springer-Verlag GmbH Austria, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Centre of the Region Haná for Biotechnological and Agricultural Research, Department of Genetic Resources for Vegetables, Medicinal and Special PlantsCrop Research InstituteOlomoucCzech Republic
  2. 2.Department of BotanyPalacký UniversityOlomoucCzech Republic
  3. 3.Department of Ecology and Environmental SciencesPalacký UniversityOlomoucCzech Republic

Personalised recommendations