Australian Systematic Botany Australian Systematic Botany Society
Taxonomy, biogeography and evolution of plants
RESEARCH ARTICLE

Molecular phylogeny of subgenus Polypompholyx (Utricularia; Lentibulariaceae) based on three plastid markers: diversification and proposal for a new section

Richard W. Jobson A C , Paulo C. Baleeiro A and Markus S. Reut B
+ Author Affiliations
- Author Affiliations

A National Herbarium of New South Wales, Royal Botanic Gardens and Domain Trust, Mrs Macquaries Road, Sydney, NSW 2000, Australia.

B Jagiellonian University in Kraków, Department of Plant Cytology and Embryology, 9 Gronostajowa Street, PO-31-044 Kraków, Poland.

C Corresponding author. Email: richard.jobson@rbgsyd.nsw.gov.au

Australian Systematic Botany 30(3) 259-278 https://doi.org/10.1071/SB17003
Submitted: 5 January 2017  Accepted: 12 July 2017   Published: 20 October 2017

Abstract

Phylogenetic relationships among all of the 47 recognised species and 10 putative new taxa of Utricularia subgenus Polypompholyx, were assessed using maximum parsimony and Bayesian inference analyses of DNA sequences representing the plastid rps16 intron, trnL–F intron and spacer regions and the trnD–T intron. We found strong jackknife and posterior-probability support for a monophyletic subgenus Polypompholyx and a sister relationship between the sections Polypompholyx + Tridentaria and Pleiochasia. Within the section Pleiochasia, are two well-supported major clades, each containing three supported clades. Our S-DIVA biogeographic analysis, using five major Australian drainage basins and New Zealand as geographic areas, estimated two early vicariance events between south-western and north-western mainland regions, corresponding with known periods of increased aridity at 15 and 6 million years ago. Subsequent dispersal events were estimated between northern and south-eastern Australia, with recent dispersal of species from south-western regions to the south-east and New Zealand occurring between 4 million and 1 million years ago. There were 28 speciation events inferred within the north-western region, followed by 9 for the south-western and south-eastern regions, indicating that the north-western monsoonal savanna habitats are a biodiversity hotspot for the lineage. We also show the evolutionary shifts in growth habit, and show that lifecycle corresponds strongly with shifts in seasonality between temperate and monsoonal regions. On the basis of our molecular phylogenetic results and morphology, we here designate a new sectional ranking for subgenus Polypompholyx.

Additional keywords: Australia, biogeography, character evolution, Lamiales, phylogenetics.

Introduction

The Lentibulariaceae is a diverse family of carnivorous plants consisting of the three monophyletic genera Pinguicula L., Genlisea A.St.-Hil. and Utricularia L. (Juniper et al. 1989; Jobson et al. 2003). In his monograph of Utricularia, Taylor (1989) delimited the following two subgenera: Polypompholyx (Lehm.) P.Taylor (three species), including two sections (Tridentaria P.Taylor and Polypompholyx); and Utricularia P.Taylor (211 species), including 35 sectional groupings. Currently, the genus is divided into 3 subgenera, including Polypompholyx, Utricularia and Bivalvaria sensu Müller and Borsch (2005). One of these subgenera, Polypompholyx, now includes section Pleiochasia Kamiénski (Fig. 1; Müller and Borsch 2005; Reut and Jobson 2010), which used to be placed in subgenus Utricularia sensu Taylor (1989).


Fig. 1.  Inflorescences of representative species for each of the clades. Section Pleiocahsia: A. Urticularia menziesii (Clade A); B. U. hamiltonii (Clade B); C. U. ameliae (Clade C); D. U. lasiocaulis (Clade D); E. U. holtzei (Clade E); F. U. dunstaniae (Clade E); G. U. dunlopii (Clade F); H. U. kimberleyensis (Clade F). Section Polypompholyx: I. U. multifida. For floral size refer to Taylor (1989) and Jobson (2013). Images used with permission: William Archer (A) and Wayne Cherry (B–I).
Click to zoom

In a recent molecular phylogeny of the subgenus Polypompholyx across 26 of the 37 then recognised species, using a single chloroplast marker rps16, Reut and Jobson (2010) supported a sister relationship between subgenus Polypompholyx and Utricularia + Bivalvaria with moderate support (Jackknife support (JK) = 89%), but did not resolve the relationship between sections Polypompholyx and Pleiochasia. A polytomy contained a moderately supported section Polypompholyx (JK = 83%) and two clades of section Pleiochasia, with one containing mostly tropical species (JK = 100%), whereas the other clade was weakly supported (JK = 70%) and contained widespread species. Since the publication of Reut and Jobson (2010), there have been nine additional species described for section Pleiochasia (Wakabayashi 2010; Jobson 2012, 2013; Jobson and Baleeiro 2015).

With inclusion of all 47 currently recognised species for subgenus Polypompholyx, we here expand the sampling of Reut and Jobson (2010), with the addition of two chloroplast markers that provide additional branch support among the clades.

In the present study, we aimed to (1) determine the relationship between sections Polypompholyx and Pleiochasia, (2) determine the relationships between previously recognised species and 9 recently described species, (3) examine the relationships of 10 undescribed taxa and discuss morphological characters that distinguish them from related taxa, (4) examine biogeographic patterns between and among major clades and (5) propose a new section for subgenus Polypompholyx.


Materials and methods

Taxon sampling

Extractions of total genomic DNA were made from silica-dried, herbarium sheet, and fresh plant material. Extractions were performed using the DNeasy Plant DNA Mini kit (QIAGEN, Hilden, Germany), following the manufacturer’s protocol. GenBank accession numbers and voucher information for published and unpublished sequences are listed in Table A1 in Appendix 1. Published rps16 sequences obtained from GenBank represented 16 ingroup taxa. The current study includes a total of 205 accessions, including seven outgroup and 198 ingroup representatives. Of these, 193 are newly sampled accessions, representing 47 recognised species and 10 taxa from subgenus Polypompholyx (Table A1).

DNA isolation, amplification and sequencing

Sequences were compiled for the plastid markers of the rps16 intron, trnLUAAtrnFGAA (trnL–F) intron and spacer region, and the trnDGUCtrnTGGU (trnD–T) intron. Data for all markers were included in a matrix of previously published rps16 sequences obtained from GenBank (Reut and Jobson 2010). Primers used for polymerase chain reaction (PCR) amplification and sequencing are described in Shaw et al. (2005). Difficulty obtaining trnD amplification for some samples required use of the internal trnY primer of Shaw et al. (2005). All PCR and sequencing reactions were carried out as described in Reut and Jobson (2010).

Sequence alignment

The forward and reverse sequences were assembled in Geneious, ver. 5.3.6 (http://www.geneious.com, accessed 8 March 2011; Kearse et al. 2012), and consensus sequences were generated and aligned in Bioedit (Hall 1999), and further adjusted manually. Gaps were treated as missing data. The sequence matrix consisted of 205 accessions representing 198 ingroup and 7 outgroup representatives (Table A1).

Maximum-parsimony analysis

All characters of the 205 accession matrix were analysed using equal weights, and gaps were treated as missing data. The three datasets were subjected to parsimony analysis both separately and combined. The three separate analyses produced congruent topologies, which allowed the use of the concatenated dataset. Reconstructions and jackknife support analyses for internal branches were performed as for Reut and Jobson (2010) using NONA (Goloboff 1998). A strict-consensus tree was calculated from all most-parsimonious trees, collapsing all branches of potential zero length.

Bayesian phylogenetic analysis

To investigate the nucleotide substitution model that best fit each of the three datasets we used the Akaike information criterion (AIC) implemented in the program jMODELTEST, ver. 2.1.7 (Guindon and Gascuel 2003; Darriba et al. 2012). On the basis of the results of the above analyses, we selected the GTR + G model for each of the three datasets. For phylogenetic analysis of the concatenated three-cpDNA dataset, containing 205 accessions, we estimated Bayesian posterior probabilities (PP) using three independent runs of one million generations, using four chains with sampling of trees every 1000 generations, and with a burn-in involving the first 25% of the samples. All parameters were set to be unlinked and with all other priors for the analysis set flat for the multinomial distribution (i.e. as Dirichlet priors). Stationarity was investigated by examining plots of the –lnL across generations in Tracer, ver. 1.6 (A. Rambaut, M. A. Suchard, D. Xie and A. J. Drummond, see http://beast.bio.ed.ac.uk/Tracer). Parameters of the effective sample size (ESS) were mostly greater than 1000. Remaining trees were used to construct a 50% majority-rule consensus tree (Fig. 2).


Fig. 2.  The 50% majority-rule consensus tree of the Bayesian inference analysis from the combined three cpDNA dataset across the subgenus Polypompholyx and seven outgroup taxa. Recognised sections are shown with thick bars, Clades A–F described in text are shown with thin bars (subclades shown on branches). Outgroup (subgenera Utricularia and Bivalvaria) is shown as collapsed branch. Branch lengths represent substitution rate; scale bar: 0.02. Posterior probability is shown as heat bar; black (1.0–0.95: strong support), blue (0.94–0.84: weak support); red (unsupported).
Click to zoom

Bayesian molecular-clock analysis

We investigated divergence times across the subgenus Polypompholyx lineage using the concatenated rps16–trnL–F–trnD–T dataset, with multiple samples of the same monophyletic species randomly reduced to a single representative per taxon, resulting in a matrix of 57 ingroup and 6 outgroup taxa. Outgroup taxa consisted of two representatives from subgenus Utricularia and four for subgenus Bivalvaria (Table A1). To obtain estimates of divergence time, we used BEAST, ver. 1.8.3 (Drummond et al. 2012), with an uncorrelated lognormal relaxed molecular clock model for divergence-date estimates, using 20 million generations. Four independent runs were performed to determine the number of generations required to allow all node-time estimates to reach stationarity. For each run, the starting trees were randomly selected, with enforced monophyly on all taxa representing both the outgroup and ingroup (subgenus Polypompholyx). Two nodes were constrained on the basis of two divergence-time priors estimated from across the whole genus in a previous study (Ibarra-Laclette et al. 2013). Ibarra-Laclette et al. (2013) used a log-normal prior to constrain the common ancestor of Lentibulariaceae to 42 million years ago (31–54 million years ago; Bell et al. 2010). In the current study, we implemented the uncorrelated log-normal relaxed clock by setting a lognormal prior on the estimated divergence time of the common ancestor of genus Utricularia to 31 million years ago (21.3–42.2 million years ago), with the common ancestor of subgenus Polypompholyx constrained to 15.5 million years ago (9.5–22.7 million years ago). A Yule Process prior was used for the tree, with default distributions used for the remaining priors. Multiple independent runs were performed for 10 million generations, with the first 10% of the generations discarded as burn-in. Stationarity was investigated by examining plots of the –lnL across generations in Tracer, ver. 1.6 (A. Rambaut et al., see http://beast.bio.ed.ac.uk/Tracer). Parameters of the ESS were mostly greater than 1000. A maximum clade-credibility tree was generated using the program TreeAnnotator, ver. 1.7.4 (Fig. 3; Drummond et al. 2012).


Fig. 3.  Maximum-credibility diversification time tree from BEAST analysis. Node-time estimates are based on substitution rates of the combined three-plastid sequence matrix. Dataset was randomly reduced to include single accession per species or taxon. Divergence dates (Ma) are shown at nodes. Posterior probability (PP) and jackknife (JK) support is shown above branches (PP/JK; PP < 0.85 not shown). Letters denoting clades (A–F) are discussed in the text. Geological epoch and time scale is shown at the base of the figure.
Click to zoom

Biogeographic analysis

We used the maximum clade-credibility tree with time estimates in an analysis using statistical dispersal–vicariance analysis (S-DIVA; Yu et al. 2010; Sun et al. 2016), implemented in RASP. We also ran the data in both S-DEC and BayArea programs implemented in RASP (Yu et al. 2015) and obtained similar results

Biogeographic areas were based on 11 major surface-water drainage divisions including Tasmania (Stein et al. 2014), with addition of a 12th area representing New Zealand (Fig. 4). For our analysis, we combined drainage divisions and areas to reduce the overall number to six major regions of the distribution: south-west (SW), north-west (NW), north-east (NE), south-east (SE), Lake Eyre (LE), and New Zealand (NZ) (Fig. 4). For NE, we combined Carpentaria coast and north-eastern coast, and for SE, we grouped all of the south-eastern temperate drainages including South Australian Gulf, south-eastern coast, Tasmania, and Murray–Darling. Using a parsimony ancestral-state reconstruction, we also defined two broad biogeographic regions representing seasonality (monsoonal tropics v. temperate) based on distributions either north or south of the Tropic of Capricorn (Fig. 5).


Fig. 4.  Ancestral-state reconstruction with dispersal–vicariance predictions for six biogrographic regions using S-DIVA. Input trees include diversification times from BEAST analysis, shown with geological epoch at base of the tree. Region(s) optimised to nodes are shown as most likely states (MLS): A, south-west (SW); B, north-west (NW); C, north-east (NE); D, south-east (SE); E, Lake Eyre (LE); and F, New Zealand (NZ). Nodes estimated as vicariance events are shown with outer green ring, and nodes estimated as dispersal events are shown with outer blue ring. All event-node MLS frequencies are shown. All regions (except NZ) are shown on map coloured in accordance with legend.
Click to zoom


Fig. 5.  Mirrored parsimony ancestral-state reconstructions: left-hand side, climate seasonality (monsoonal tropics v. temperate); right-hand side, lifestyle (annual v. perennial). Pagel’s correlation analysis provided a significant (P = 0.008) correlation (refer to Materials and methods).
Click to zoom

The complexity of past geological and climatic changes across currently arid barriers between NW and SW (Pepper et al. 2013), and SW and SE regions (Byrne et al. 2011; Crisp and Cook 2013), and a lack of data on dispersion for any of the species (Taylor 1989), provided justification for a conservative approach to the coding of range constraints in RASP, with all areas being coded equally. With no reliable fossil records for genus Utricularia, constraints on fossils was set to zero. The maximum ancestral-area size was constrained to three because there was a single species in our study distributed in three of the six designated areas (U. fenshamii) (Table 1, Fig. 4).


Table 1.  S-DIVA dispersal-vicariance results, showing event nodes across the Bayesian maximum-credibility diversification time tree
Ancestral areas: A, south-west (SW); B, north-west (NW); C, north-east (NE); D, south-east (SE); E, Lake Eyre (LE); F, New Zealand (NZ). The most likely state (MLS) is that with highest frequency. The RASP route is shown as →, direction; ^, alternate route; |, isolation of distribution
Click to zoom

Ancestral-state reconstructions

A correlation analysis of the characters seasonality (monsoonal tropics v. temperate) and life cycle (annual v. perennial) was performed in the Mesquite software package, ver. 3.03 (W. P. Maddison and D. R. Maddison, see http://mesquiteproject.org), using Pagel’s correlation method (Pagel 1994). We used a four-parameter model with 5000 simulations to estimate maximum-likelihood values and statistical significance (P < 0.05) for null hypothesis, namely, that Character state X evolves independently of Character state Y (Fig. 5).


Results

Sequences and alignment

The rps16 matrix contained 1133 characters, of which 378 were parsimony informative, the trnL–F matrix contained 1323 characters, of which 290 were parsimony informative, and the trnD–T matrix contained 1258 characters, of which 289 were parsimony informative.

Maximum-parsimony analysis

Phylogenetic analysis of the concatenated rps16, trnL–F and trnD–T matrix resulted in 9345 equally parsimonious trees of a length 2351, with a consistency index (CI) of 0.58 and retention index (RI) of 0.89. The strict-consensus tree (figure not shown) was well resolved, with well-supported branches across most of the tree (support values shown on Fig. 3).

Bayesian inference and divergence-time estimates

The topology of the 50% majority-rule tree from the Bayesian analysis of the 205 accession matrix (Fig. 2) was congruent with the maximum-parsimony topology. The ultrametric maximum-credibility tree derived from the Bayesian inference analysis from the reduced taxon dataset (containing a single representative for each species or taxon) is shown (Fig. 3). The analysis incorporated two calibration points and provided diversification-time estimates for all nodes in the tree (Fig. 3). The divergence-date node age 95% highest posterior density (HPD) intervals varied from the mean by up to 37%, indicating high levels of uncertainty (Fig. 3).

Figs 2 and 3 show that subgenus Polypompholyx forms a well-supported monophyletic lineage (PP = 1.00, JK = 100%) comprising two well-supported clades containing the monophyletic sections Tridentaria and Polypompholyx (PP = 1.00, JK = 100%) that together form a sister relationship with a clade containing a well-supported monophyletic section Pleiochasia (PP = 1.00, JK = 100%). Within section Pleiochasia are two major clades; the first is well supported (PP = 1.00, JK = 100%), consisting of Clade A (PP = 1.00, JK = 100%), which is sister to a weakly supported clade (PP = 89, JK = 60), containing Clade B (PP = 1.00, JK = 98%) and Clade C (PP = 1.00, JK = 96%), which, together, are sister to the second well-supported major clade (PP = 1.00, JK = 99%) consisting of Clade D (PP = 1.00, JK = 94%), which is sister to a supported clade (PP = 0.97, JK = 52%) containing Clade E (PP = 1.00, JK = 98%) and Clade F (PP = 0.98, JK = 60%) (Figs 2, 3).

Biogeographic pattern

Global results from the S-DIVA analysis estimated 17 dispersal and 9 vicariance events across the lineage (Fig. 4). S-DIVA also estimated number and directionality of dispersal events across the lineage between the following areas: NW > SW, SE > NW, SE > NE, SE > LE and SE > NZ each occurred once; SW > SE and NW > SE occurred twice; and NW > NE occurred six times. Estimates of speciation within areas across the lineage include the following: SW = 9, NW = 28, NE = 1 and SE = 9. Divergence of subgenus Polypompholyx from sister subgenus Utricularia + Bivalvaria was calibrated to 31 million years ago, indicating that the lineage may have occurred in Australia for c. 15 million years, before divergence of the two major clades (Figs 24). Our S-DIVA results suggest a centre of origin SW–NW (Fig. 4), followed by a vicariance event c. 15.5 million years ago between the two regions, with subsequent key vicariance and dispersal events shown in Fig. 4 and Table 1.


Discussion

Inter- and infrageneric relationships

In the current study, sections Tridentaria and Polypompholyx formed a well-supported sister relationship, and, together, are sister to a monophyletic clade containing two major clades consisting of all recognised species of section Pleiochasia (Taylor 1989). The phylogeny of Reut and Jobson (2010) did not resolve this relationship, instead finding a polytomy containing section Polypompholyx and two major clades of section Pleiochasia. Utricularia westonii P.Taylor, not included in Reut and Jobson (2010), and representing the monotypic section Tridentaria endemic to Cape Le Grand, Western Australia (WA), is sister to a section Polypompholyx, with inclusion of recognised members of both U. multifida and U. tenella, each of which were found to be monophyletic (Figs 2, 3).

Our two samples of U. multifida represent the northern and southern ends of the distribution in SW WA, whereas the two samples representing U. tenella represent SW WA and southern South Australia (SA), although the distribution of this species extends across into Tasmania (Tas; Taylor 1989; Table A1). The major morphological characters that distinguish sections Tridentaria and Polypompholyx from section Pleiochasia are presence of four v. two calyx lobes and a bladder-trap with a divided v. simple dorsal appendage (Taylor 1989). Characters distinguishing sections Polypompholyx and Tridendaria include a trap stalk that is distally inflated v. stalks not inflated, possessing a short bifid trap dorsal appendage v. dorsal appendage deeply trifid, and a trap door with numerous short tripping glands v. tripping glands long and bristle-like (Taylor 1989).

Clades A–C

The Clade 1 of Reut and Jobson (2010) is supported by the current study with expansion of sampling to contain all recognised species, containing Clades A–C (Figs 2, 3), characterised morphologically by all species having a glabrous peduncle (Taylor 1989).

Clade A contains species geographically restricted to SW WA in all cases except U. violacea R.Br., which also extends eastwards into SA, Victoria (Vic.) and Tas. (Fig. 4). The phylogeny of Reut and Jobson (2010) included U. paulineae Lowrie and U. violacea that together grouped within a weakly supported polytomy containing the clades of Group B (U. fistulosa, U. hamiltonii, U. singeriana, U. triflora) and the ‘U. dichotoma group’ (U. dichotoma and allies; Fig. 2).

Here, we show relationships across all recognised species of the Clade A, with the recently described stoloniferous U. paulineae (Lowrie 1998), sister to a clade of non-stoloniferous, rosulate species (Fig. 2). This clade is defined by the presence of basisolute bracts and bracteoles, with the terrestrial red-flowered U. menziesii (Fig. 1A) sister to a clade of the affixed-aquatic U. volubilis R.Br. and U. helix P.Taylor (Fig. 6), both of which possess a twining peduncle; the former has a larger violet entire corolla lower lip, whereas U. helix has a smaller mauve three-lobed corolla lower lip (Taylor 1989). These species are sister to a polytomy containing recently described species U. petertaylorii Lowrie (Lowrie 2002), U. violacea and sister species U. inaequalis A.DC. and U. benthamii P.Taylor (Fig. 2).


Fig. 6.  Parsimony ancestral-state reconstruction of habit mapped onto the BEAST maximum-credibility diversification-time tree. Shading on branches represent character states shown in legend. Scale represents million years (Ma).
Click to zoom

Clade A is sister to Clades B and C, with the former containing a natural grouping of mostly terrestrial species characterised by a possession of a non-inflated hollow peduncle and an often polymorphic corolla lower lip that is either entire or three-lobed (Fig. 2; Taylor 1989).

Within Clade B, we find the two Subclades B1 and B2 (Fig. 2). Subclade B1 includes two representatives of the Northern Territory (NT) endemic Utricularia linearis Wakabayashi in a polytomy with Queensland (Qld) endemic U. blackmanii R.W.Jobson, and U. fistulosa that has a distribution across the Kimberley, WA (Fig. 2).

Subclade B2 forms a polytomy containing U. triflora P.Taylor, represented by accessions from across its NT distribution (Fig. 2). Other members of the polytomy include northern Qld endemic sister species U. albiflora R.Br. and U. terrae-reginae P.Taylor (Fig. 1B) that occupy a similar distribution across the Qld latitude of 17°S (Taylor 1989). It also includes U. singeriana, which is sister to a grouping of U. hamiltonii F.E.Lloyd, and two potentially new taxa U. sp. Kununurra (C.Glover 81), endemic to the Kununurra region of WA, and U. sp. Theda (M.D.Barrett 2056) that is distributed across Kimberley region of WA (Fig. 2). Taylor (1989) considered U. singeriana and U. sp. Kununurra (C.Glover 81) conspecific, although on the basis of the current phylogenetic position of the latter, and the observations of Cowie (2010), it is here shown to be closely allied with U. hamiltonii (Fig. 2). Cowie (2010) found that this entity is not closely related to accessions that fit the type specimen of U. singeriana.

Both U. hamiltonii and U. sp. Theda (M.D.Barrett 2056) are characterised with a single-flowered inflorescence and peduncles that become recurvate during maturation of the capsule (R. W. Jobson, pers. obs.). This character does not occur in other members of subgenus Polypompholyx (Taylor 1989), except in the multi-flowered inflorescence of U. tubulata F.Muel. (Clade C), although in that species, it involves only the deflexion of the pedicel (Taylor 1989).

Clade C contains the two species U. byrneana R.W.Jobson and U. tubulata, together forming a sister clade to members of the U. dichotoma Labill. complex (Reut and Jobson 2010; R. W. Jobson, M. S. Reut and B. J. Plachno, in preparation). Both U. byrneana and U. tubulata are distributed in the Kimberley region of WA, with the distribution of the latter extending across to NT, coastal regions of the Gulf of Carpentaria and the type location in north-eastern Qld (Taylor 1989). These two species are morphologically and ecologically divergent, with the terrestrial U. byrneana occupying edges of swamps, possessing a non-inflated hollow peduncle, corolla spur as long as the lower lip, and seed structure typical of most other members of Pleiochasia (Jobson and Baleeiro 2015). In contrast, U. tubulata is highly modified for the fully suspended aquatic habit (Fig. 6) with inflated hollow peduncle, along with a corolla spur twice the length of the lower lip, and seeds coated in specialised appendages (Taylor 1989; Reut and Jobson 2010).

The well-supported sister clade contains the U. dichotoma complex that groups several newly described well-supported species (Jobson 2013), previously recognised under U. dichotoma (Taylor 1989). The first branching clade contains U. barkeri R.W.Jobson and U. grampiana R.W.Jobson (Fig. 2). These two species are characterised by possession of basisolute bracts and bracteoles and a light mauve or cream corolla upper lip speckled with purple flecks (Jobson 2013). Utricularia barkeri occupies swampland across south-eastern SA, southern Vic. and Tas., whereas U. grampiana is endemic to high-elevation seepages (altitude 450–1160 m) of central western Vic. (Jobson 2013).

Sister to U. barkeri and U. grampiana is a clade containing three recognised species and several unrecognised taxa that all have basifixed bracts and bracteoles (Taylor 1989; Jobson 2013). The first branching clade contains four population representatives of the subaquatic U. sp. Foster (R.W.Jobson 1137) restricted to wallum swamps of central and lower northern-coast regions of New South Wales (NSW; Figs 2, 6). These specimens may fit the type description of U. speciosa R.Br. (Brown 1810), subsequently synonymised under U. dichotoma by Taylor (1989).

The next branching clades contain two new geographically isolated species, namely U. fenshamii R.W.Jobson and U. ameliae R.W.Jobson (Fig. 2). These two species occupy discharge mound springs of the Great Artesian Basin (GAB). Utricularia ameliae is represented by two specimens, each from its two known populations. This species has the most geographically isolated distribution within the subgenus, located in the Diamantina region of far-western Qld (Fig. 4). Utricularia ameliae is characterised by its single-flowered inflorescence, white corolla upper lip, and transversely oblong shape of the lower lip (Jobson 2013). Utricularia fenshamii is represented by samples from across its distribution in mound spring habitats from the Lake Eyre region of SA, western NSW, and southern and northern-central Qld (Fig. 2).

Utricularia ameliae and U. fenshamii are within a polytomy that also contains a clade of U. beaugleholei Gassin and an allied taxon U. beaugleholei Gassin ssp. A (R.W.Jobson 1203) distributed across SW NSW and Vic. Utricularia beaugleholei is a recently described species characterised by presence of 4–11 yellow ridges at the base of the corolla lower lip, and a larger corolla than other members of the complex (Gassin 1993). The polytomy also contains the widely distributed U. sp. Nowra (R.W.Jobson 1243), occurring from south-western WA to SA, Vic. and coastal NSW. This taxon may fit the morphological description and type material of U. oppositiflora R.Br. (Brown 1810), a previously recognised species synonymised under U. dichotoma by Taylor (1989).

The fifth clade within the polytomy is also well supported, and contains two subclades. The first includes specimens from Tas. that fit the morphological description of U. dichotoma s.s. (Taylor 1989), along with U. monanthos and U. novae-zelandiae (Fig. 2), a grouping that has been previously suggested on the basis of morphology (Reut and Fineran 1999, 2000). The second clade is well supported, and includes a grouping of U. dichotoma specimens representing populations distributed across central Qld and the north-western slopes of NSW, occupying swamps associated with the recharge spring system of the GAB in Qld, and similar habitat not associated with GAB in NSW. The second grouping within the clade is well supported and includes U. dichotoma specimens from higher-elevation tablelands of the Great Dividing Range from central Qld, south to the Furneaux islands in Bass Straight, Tas. (Figs 2, 4). A study that is focused on the relationships and taxonomy within the U. dichotoma complex is currently underway, using fast evolving nuclear markers and morphology (Jobson et al., in preparation).

Clades D–F

The Clade 2 of Reut and Jobson (2010) is also supported by the current study (PP = 1.00, JK = 99%), with expansion of sampling to contain all recognised species that form the three Clades D–F (Fig. 2), with all groupings containing many species characterised by the presence of hairs or glands on the peduncle (Taylor 1989). In the current study, Clade D contains population samples representing the respective distributions of U. kamienskii F.Muell., U. leptorhyncha O.Schwarz. and U. lasiocaulis F.Muell. (Taylor 1989). The three accessions of U. kamienskii, that represent populations from across its NT distribution, are together sister to a clade containing accessions representing many of the varied forms of U. lasiocaulis and U. leptorhyncha from across their distributions. In the current study, we found that the relationships among the varied forms of U. leptorhyncha, distributed from the Kimberly, WA, to NT, and U. lasiocaulis with a similar distribution that also extends across to Qld, are generally unresolved (Fig. 2). We are currently sampling more broadly across the U. lasiocaulis and U. leptorhyncha complex, using fast-evolving nuclear markers and morphology to better determine species boundaries (R. W. Jobson and P. C. Baleeiro, in preparation).

Clade E was weakly supported as monophyletic in the study of Reut and Jobson (2010). Minute or small stature and presence of a five-lobed corolla lower lip are characters found in all members except in U. arnhemica P.Taylor, a mostly medium-sized species with a three-lobed corolla lower lip, and the largest bladder-traps in the subgenus (~1.5 cm long; Taylor 1989).

In the current study, we resolved most of the species relationships within Clade E, with Arnhem Land, NT, endemic U. rhododactylos P.Taylor found to be sister to all other members of the clade (Fig. 2). The sister group contains U. arnhemica populations from across its NT distribution. Utricularia quinquedenata F.Muell. ex P.Taylor, the smallest flowered species in the subgenus, is distributed across tropical Australia with several disjunct populations near Brisbane, Qld (Taylor 1989). This species is sister to a clade containing U. capilliflora F.Muell, a species with capillary appendages that arise from the corolla upper lip lobes, and is itself sister to a polytomy (Clade E1) that contains three recognised species with capillary appendages that arise from the two outer-most lower lip lobes, namely U. dunstaniae F.E.Lloyd, U. lowriei R.W.Jobson and U. anntenifera P.Taylor (Fig. 2). These taxa form a well-supported clade in a polytomy that includes NT endemics U. cheiranthos P.Taylor and U. holtzei F.Muell., with all taxa sharing a corolla lower lip limb with a ±semi-circular marginal rim of the palate (Taylor 1989). Utricularia lowriei is a newly discovered endemic species from Cape York, Qld (Jobson 2013), whereas U. anntenifera is endemic to the Kimberley region of WA (Taylor 1989). Utricularia dunstaniae is distributed from Jabiru to Darwin, NT, with a disjunct occurrence in the Kimberley, WA (Taylor 1989). However, we found that the morphologically divergent Kimberley specimens of U. dunstaniae, here referred to as U. sp. Bearded (R.W.Jobson 2308), are sister to U. anntenifera, whereas the U. dunstaniae accessions from the NT fit the type specimen, and are supported as sister to U. lowriei (Figs 2, 3).

Clade F is a moderately supported clade comprising three well-supported subclades of taxa distributed across the Kimberley region of WA (Taylor 1989). Clade F1 contains two unresolved U. tridactyla P.Taylor accessions from near Kununurra, WA, that are sister to two accessions of U. kenneallyi P.Taylor. These two species are characterised by possessing a three-lobed corolla lower lip and basifixed bracts and bracteoles (Taylor 1989), and are weakly sister to a third entity here designated U. sp. Basalt (M.D.Barrett 3576), that is found in the north-western Kimberley region (Fig. 2).

The two populations representing the northern Kimberley endemic U. kenneallyi differ from U. tridactyla with the presence of trap-door dorsal appendage divided into three setae v. a single broad fimbriate appendage, lateral appendages reduced to single setae on each side of the trap-door rim v. fimbriate lateral appendage set away from the trap-door rim, being single-flowered v. two-flowered, and with the base of peduncle being densely hispid v. sparsely papillose (Taylor 1989). The above clade is sister to a well-supported, undescribed taxon possessing a densely papillose peduncle base that is here designated U. sp. Papilose (R.W.Jobson 2657) (Fig. 2), and was previously misidentified as U. leptorhyncha (Taylor 1989).

Sister to the above taxa is a well-supported grouping containing three Kimberley accessions of an unnamed entity previously misidentified as U. arnhemica (R.W. Jobson, pers. obs.). We found that the type of U. arnhemica fits best the populations from NT (grouped within Clade E, Fig. 2). We, therefore, designate the three Kimberley populations as an undescribed taxon U. sp. Sandstone (M.D.Barrett 1335) (R. W. Jobson, P. C. Baleeiro and M. D. Barrett, in preparation).

Sister to Clade F1, are the F2 and F3 clades, with F2 containing two populations of the Kimberley endemic U. georgei P.Taylor, which is, in turn, sister to U. kimberleyensis C.A.Gardner (Taylor 1989). Although both species have two short ridges at the base of the corolla lower lip, the lip is deeply three-lobed in U. georgei, and entire in U. kimberleyensis (Taylor 1989).

Clade F3 is well supported and contains four recognised species, including U. uniflora sampled from across its distribution (Taylor 1989). On the basis of overall appearance, U. uniflora was previously thought to be allied with U. dichotoma (Taylor 1989). Here, we have confirmed the results of Reut and Jobson (2010) and have found a sister relationship with a well-supported grouping containing two populations of U. dunlopii P.Taylor from Jabiru and Darwin, NT, the recently described Kimberley endemic lithophyte U. wannanii R.W.Jobson (Jobson and Baleeiro 2015), and U. sp. Fanged (R.W.Jobson 2682; Fig. 2). Both U. dunlopii and U. wannanii have elongated upper (2×) and lower (3×) corolla lip lobes; in U. dunlopii, these are filiform and pinkish-bronze (Taylor 1989), whereas in U. wannanii the lobes are subulate and white (Jobson and Baleeiro 2015). These two species share similar connate bracts and bracteoles; however, bladder trap structure, overall floral form and size, and habitat are divergent (Fig. 6; Taylor 1989; Jobson and Baleeiro 2015). Utricularia dunlopii shares a similar floral form to that of U. capilliflora of Clade E (Fig. 2), with ascending filiform corolla upper lip lobes; however, the bracts and bracteoles are different and the lower lip lobes are three- and five-numerous respectively (Taylor 1989). Whereas U. sp. Fanged (R.W.Jobson 2682) bears little resemblance to it sister species U. dunlopii and U. wannanii, it has previously been confused with U. kimberleyensis s.s. of Subclade F2 (Fig. 2; Taylor 1989). This undescribed entity is distributed from the Edith River, NT, to the Kimberley region and is characterised by its two white or cream-coloured claw-like ridges at the base of the corolla lower lip (R. W. Jobson, P. C. Baleeiro and M. D. Barrett, in preparation).

Diversification times and biogeographic shifts

The ancestral-state reconstruction of biogeographic characters of Jobson et al. (2003) optimised the common ancestor of Utricularia to Tropical America, with divergence of subgenus Polypompholyx being optimised to Australia and Utricularia + Bivalvaria being optimised to South America. The biogeographic results of Jobson et al. (2003) were interpreted as fitting best the boreotropics hypothesis (Lavin and Luckow 1993), that suggests an initial dispersal across the boreal region and subsequent vicariance events.

Recent molecular divergence dating estimated that the split between subgenus Polypompholyx and subgenus Utricularia + Bivalvaria may have occurred c. 31 million years ago (21.3–42.2 million years ago, 95% HPD; Ibarra-Laclette et al., 2013) in the early Oligocene when the Australian landscape began drying out with increased seasonality (Toon et al. 2015). From their analysis, Ibarra-Laclette et al. (2013) estimated that the split between sections Polypompholyx (one taxon) and Pleiochasia (two taxa) occurred c. 15.45 million years ago (9.5–22.7 million years ago, 95% HPD), suggesting that the lineage may have been present in Australia for c. 15 million years before the mid-Miocene (Crisp and Cook 2007, 2013). In the current study, we have repeated the divergence dating of Ibarra-Laclette et al. (2013) with an expanded sampling in subgenus Polypompholyx, including all recognised species (Fig. 3). We found that during the evolution of the subgenus Polypompholyx lineage, there were four major biogeographic shifts; the first occurred c. 15 million years ago during the mid-Miocene, at the establishment of the two major lineages (subgenus Polypompholyx node), whereas three other events may have occurred between 7 and 1 million years ago across multiple taxa independently in different clades (Fig. 4).

At the basal node (Node 113, Fig. 4), the S-DIVA vicariance–dispersal results showed that the most likely states (MLS) are the SW and NW drainage basins and predicted that these basins may have been involved in the first major vicariance event occurring c. 15 million years ago (Fig. 4). This was possibly preceded by a period of linkage between these two basins via the Pilbara–Gascoyne (GP) and North Western Plateau (NWP) drainage basins (shown in yellow; Fig. 4), during the mid-Miocene, when Australia began a process of rapid aridification (Bowman et al. 2010; Byrne et al. 2011; Toon et al. 2015), largely restricting aquatic plants to moist coastal regions in the SW and NW. About this time, a coastal incursion of the far-western edge of GP+NWP occurred, with subsequent drying of the surrounding region (Byrne et al. 2011). The incursion was a likely dispersal barrier until it receded to the current coastline at c. 10 million years ago, then possibly providing suitable oligotrophic swampy linkage habitats until GP+NWP drainages dried up c. 6 million years ago (Byrne et al. 2011). This process of loss, gain, and loss of linkage habitats between SW and NW corresponds with the first vicariance event at the establishment of the two major clades (Clades A–C and D–F), with a subsequent dispersal event NW > SW c. 12 million years ago (Node 110), and the second major vicariance event at Node 88 (Clades A and B–C) between SW and NW at c. 7 million years ago, which may have led to the establishment of a second SW clade, Clade A (Node 64; Fig. 4).

Each of the two SW clades contain a species with a disjunct distribution in SE (U. violacea and U. tenella), predicted to have dispersed SW > SE c. 2 million years ago (Nodes 59 and 111), across the currently arid Nullarbor Plain (NP; Fig. 4). This pattern approximately corresponds with the geological history of the NP; a region inundated c. 14 million years ago, receding to the current coastline at c. 9 million years ago, with reversion to wet climates across the NP during the mid-Pliocene c. 3.5 million years ago (Sniderman et al. 2007, 2016). This process may have provided suitable oligotrophic swampy habitats up until a period of extreme cooling and drying that began during the Pleistocene from 2 to 1 million years ago (Byrne et al. 2011).

Divergence of Clades A–C and D–F c. 12 million years ago (Node 110) shows the MLS as NW and was estimated to have been a dispersal event (Fig. 4). Species diversification within NW was far higher than in any other area, with an estimated 28 speciation events. These occurred from c. 9 million years ago, and mainly involved Clades C and D–F, containing species that currently occupy seasonally wet savanna habitats (Figs 2, 4B, 5). Although we have found no cases of dispersal from NE to NW, a strong pattern is observed in the S-DIVA analysis for dispersal from NW to NE, involving six dispersal events, with a single speciation event in NE (Fig. 4B; Table 1). In three cases, dispersal to NE is followed by vicariance events that involve endemics of far-northern Qld, namely, U. blackmanii (Node 85) and U. albiflora-U. terrae-reginae (Node 82) in Clade C, and U. lowriei (Node 90) in Clade F. All three events are estimated to have occurred between c. 3 and 1.5 million years ago (Fig. 4). The former species is restricted to the basalt plateaus of the Einasleigh Uplands region, whereas the latter three are restricted to Cape York (11–15°S latitude) (Taylor 1989). Although savanna habitats expanded across much of the Gulf plains during the late Pleistocene and early Holocene, a low-lying semi-arid region of the Gulf known as the ‘Carpentaria gap’ (Fujita et al. 2010; Catullo et al. 2014) is thought to have formed during the Pleistocene (Catullo et al. 2014), and was a likely dispersal barrier between NW and NE. Species with distributions spanning the NW and NE areas include U. tubulata (Clade B), U. lasiocaulis (Clade D), and U. quinquedentata (Clade E) (Fig. 2), with all three species dispersing from NW to NE between c. 4 and 2 million years ago (Nodes 65, 107, 95 respectively; Fig. 4).

Clade B contains the sister species U. byrneana (NW endemic) and U. tubulata (NW–NE) that, together, are sister to the U. dichotoma complex clade containing a diverse assemblage of species widely distributed in the NE–SE–LE–SW (Fig. 4). These clades diverged c. 3.6 million years ago and was estimated to have differentiated as a result of a NW and SE vicariance event (Node 72) with first branching species, U. grampiana and U. barkeri, occuring near the south-western limits of their SE distribution (Fig. 4; Jobson 2013). Four dispersal events were estimated to have occured within the clade, with the first occurring during the divergence of clades U. dichotomanovae-zelandiaemonanthos and U. fenshamiiameliaebeaugleholei c. 1.4 million years ago (Node 73) dispersing from SE to NE–LE–NZ (Fig. 4B). In a subclade of Clade C in Fig. 4, a vicariance event occurred 0.44 million years ago between SE and NZ (Node 67) involving U. novae-zelandiae, with dispersal from SE to NE (Node 66) for U. dichotoma. A vicariance event between SE and LE was estimated c. 1.3 million years ago at Node 72, involving U. ameliae (LE) and U. fenshamiibeaugleholei. Two subsequent events among the latter species involve dispersal, from SE to NE–LE at c. 1.3 million years ago (Node 71) in U. fenshamii, and from SE to SW c. 1 million years ago (Node 70) in U. sp. Nowra (R.W.Jobson 1243; Fig. 4). Utricularia fenshamii is mostly restricted to mound spring habitats distributed across the GAB (mostly LE), whereas U. ameliae is known only from two mound springs habitats within the arid LE region; the remote nature of the sites may have been an isolating barrier during speciation.

Shifts in lifecycle with seasonality

Lifecycle shift (i.e. duration of growth) from an annual ancestor to perennial was estimated to have occurred twice independently within the mostly monsoonal tropical clades, namely Clades D–F (section Lasiocaules, the present study; Fig. 5). For Clades A–C, the ancestral state was estimated to be perennial with five independent reversals to the annual lifecycle (Fig. 5). All three members of sections Polypompholyx and Tridentaria, distributed across the SW, are annual. The annual lifecycle tends to correspond with biogeography (Fig. 5) and mostly involves the terrestrial habit (Fig. 6), where plants undergo seasonal drying of habitat across savanna biomes of northern Australia, but also those of the Mediterranean climate of the SW (Taylor 1989). The perennial lifecycle tends to correspond with species occurring in temperate regions of SE, or those occupying permanently wet habitats of the other regions (Figs 4, 5; Taylor 1989). We examined whether or not shifts in lifecycle (x) are dependent on seasonality (y) (log-likelihood x = –40.08367; y = –36.82635; difference = 3.2573) and we showed a significant (P = 0.008) correlation between shifts in lifecycle (annual v. perennial) and seasonality (temperate v. monsoonal tropics).

Evolution of growth habit

We found that the terrestrial habit, in which stolons, traps and rhizoids are embedded in a wet substrate, with leaves mostly aerial, is optimised as the ancestral habit (Fig. 6). The ancestral habit across the sister lineages of subgenus Utricularia–Bivalvaria was also found to be the terrestrial state (Jobson et al. 2003). A shift to the affixed aquatic habit, in which stolons, traps and rhizoids are mostly embedded in substrate with all vegetative organs submerged, has occurred four times independently across subgenus Polypompholyx (Fig. 6). A single shift from terrestrial to the suspended aquatic habit (all vegetative organs floating in water column) has arisen once in U. tubulata; a shift accompanied by verticillate growth form, inflated peduncle and modified seed-coat appendages (Taylor 1989; Reut and Jobson 2010) (Fig. 6). Reut and Jobson (2010) found strong correspondence between shifts to the aquatic habit and evolution of filiform appendages around the bladder trap mouth. The lithophytic habit, with growth in a thin layer of substrate on wet rock surfaces, has arisen once in U. wannanii, from a terrestrial ancestor (Fig. 6; Jobson and Baleeiro 2015).


Proposal for new sectional division

Taylor (1989) distinguished section Pleiochasia from all other sections of the genus on the basis of the presence of traps with a single dorsal appendage, along with lateral or ventral trap appendages (or both), in combination with a lack of scales on the peduncle. We, here, delimit section Pleiochasia as always possessing a glabrous peduncle, whereas all three clades (D–F, Fig. 2) of section Lasiocaules contain many members having ornamentation (hairs or glands) on the peduncle.

Utricularia section Lasiocaules R.W.Jobson & Baleeiro, sect. nov.

Type: Utricularia lasiocaulis F.Muell.

Minute to medium size, perennial or annual herbs. Terrestrial, affixed subaquatic or lithophyte. Leaves from the peduncle base or rosulate, or on stolons, very narrowly linear to narrowly linear-obovate, elliptic, or very narrowly obovate to obovate. Traps lateral, basal, terminal or subterminal from the peduncle base, stolon nodes or internodes, filiform dorsal and lateral appendages, ventral wing lobes when present fimbriate or ciliate, less frequently reduced and entire, subulate, or narrowly deltoid. Peduncle base shortly hispid, minutely glandular or glabrous; scales absent; bracts and bracteoles connate, basifixed, or basisolute to semi-basisolute. Corolla violet, pink, white, cream or apricot (brown); upper lip entire with apex emarginated or divided into two long filiform lobes; lower lip slightly 3-lobed to deeply 3–5 oblong- or filiform-lobed; spur ovoid from a narrowly cylindrical base, inflated scrotiform, subulate, or narrowly conical–cylindrical, apex rounded or acute. Seeds ovoid, narrowly obovoid, oblong-obovoid, obovoid or ellipsoid (Fig. 1D–H).

Distribution

With the exception of U. uniflora, which is distributed from south eastern Queensland to Tasmania, all other species are restricted to northern Australia.


Utricularia sect. Pleiochasia Kamiénski

Very small to large size, perennial or annual herbs. Terrestrial, affixed aquatic or suspended aquatic. Small stem present and stolons absent or opposite (stem absent and stolons present). Leaves rosulate from the stem, or from stolon nodes, narrowly linear to obovate. Traps lateral, from the stem or peduncle base, less frequent from stolons, long-stalked, one simple subulate dorsal appendage, a pair of linear or narrowly oblong, apically fimbriate, lateral appendages and a pair of marginally fimbriate, ventral wings present or absent. Peduncle base glabrous; scales absent; bracts and bracteoles basisolute or basifixed. Corolla purple–mauve, white, pink or red; upper lip generally constricted in the middle, apex entire to emarginate, or deeply lobed; lower lip limb almost semi-circular, reniform, flabellate or transversally oblong or elliptic, apex entire or slightly lobed; spur broadly conical to narrowly subulate or cylindrical, apex rounded, obtuse or shortly bilobed. Seeds narrowly obovoid, ovoid, broadly ovoid, oblong-ovoid, oblong-obovoid, ellipsoid or cylindrical (Fig. 1A–C).

Distribution

All states, New Zealand, New Caledonia.



Conflicts of interest

The authors declare that they have no conflicts of interest.



Acknowledgements

We thank the curators and staff of AD, AK, BRI, CANB, CHR, CNS, DNA, HO, MEL, NE and PERTH, who provided loans, images, and access to herbarium material. We also thank Matt Barrett (PERTH), Russell Barrett (NSW) and Wayne Cherry (NSW) for providing material and assisting with fieldwork. This work was funded by a grant provided to R. W. Jobson through the Australian Biological Resources Study (ABRS) National Taxonomy Research Grant Programme (RFL212-45). Scientific collection permits were obtained from relevant State and Federal Government agencies.


References

Bell CD, Soltis DE, Soltis PS (2010) The age and diversification of the angiosperms re-revisited. American Journal of Botany 97, 1296–1303.
The age and diversification of the angiosperms re-revisited.CrossRef |

Bowman DMJS, Brown GK, Braby MF, Brown JR, Cook LG, Crisp MD, Ford F, Haberle S, Hughes J, Isagi Y, Joseph L, McBride J, Nelson G, Ladiges PY (2010) Biogeography of the Australian monsoon tropics. Journal of Biogeography 37, 201–216.
Biogeography of the Australian monsoon tropics.CrossRef |

Brown R (1810) Prodromus floræ Novæ Hollandiæ et Insulæ Van-Diemen: exhibens characteres plantarum quas annis 1802–1805. (R. Taylor et socii: London, UK)

Byrne M, Steane DA, Joseph L, Yeates DK, Jordan GJ, Crayn D, Aplin K, Cantrill DJ, Cook LG, Crisp MD, Keogh JS, Melville J, Moritz C, Porch N, Sniderman JMK, Sunnucks P, Weston PH (2011) Decline of a biome: evolution, contraction, fragmentation, extinction and invasion of the Australian mesic zone biota. Journal of Biogeography 38, 1635–1656.
Decline of a biome: evolution, contraction, fragmentation, extinction and invasion of the Australian mesic zone biota.CrossRef |

Catullo RA, Lanfear R, Doughty P, Keogh JS (2014) The biogeographical boundaries of northern Australia: evidence from ecological niche models and a multi-locus phylogeny of Uperoleia toadlets (Anura: Myobatrachidae). Journal of Biogeography 41, 659–672.
The biogeographical boundaries of northern Australia: evidence from ecological niche models and a multi-locus phylogeny of Uperoleia toadlets (Anura: Myobatrachidae).CrossRef |

Cowie ID (2010) Notes on the identity, distribution and conservation status of the threatened plant species ‘Utricularia singeriana’ F.Muell. (Lentibulariaceae). The Beagle: Records of the Museums and Art Galleries of the Northern Territory 26, 119–121.

Crisp MD, Cook LG (2007) A congruent molecular signature of vicariance across multiple plant lineages. Molecular Phylogenetics and Evolution 43, 1106–1117.
A congruent molecular signature of vicariance across multiple plant lineages.CrossRef | 1:CAS:528:DC%2BD2sXlvVOrtbw%3D&md5=6133292eafeac6b6796b3da613933d6bCAS |

Crisp MD, Cook LG (2013) How was the Australian flora assembled over the last 65 million years? A molecular phylogenetic perspective. Annual Review of Ecology Evolution and Systematics 44, 303–324.
How was the Australian flora assembled over the last 65 million years? A molecular phylogenetic perspective.CrossRef |

Darriba D, Taboada GL, Doallo R, Posada D (2012) jModelTest 2: more models, new heuristics and parallel computing. Nature Methods 9, 772
jModelTest 2: more models, new heuristics and parallel computing.CrossRef | 1:CAS:528:DC%2BC38XhtFWmsbfP&md5=73923fa81b12f7189024e36149b9c579CAS |

Drummond AJ, Suchard MA, Xie D, Rambaut A (2012) Bayesian phylogenetics with BEAUti and the BEAST 1.7. Molecular Biology and Evolution 29, 1969–1973.
Bayesian phylogenetics with BEAUti and the BEAST 1.7.CrossRef | 1:CAS:528:DC%2BC38XhtFagu7fO&md5=f63ae629377ad80ee4088ee8ef3c1021CAS |

Fujita MK, McGuire JA, Donnellan SC, Moritz C (2010) Diversification and persistence at the arid-monsoonal interface: Australia-wide biogeography of the Bynoe’s gecko (Heteronotia binoei; Gekkonidae). Evolution 64, 2293–2314.
Diversification and persistence at the arid-monsoonal interface: Australia-wide biogeography of the Bynoe’s gecko (Heteronotia binoei; Gekkonidae).CrossRef | 1:CAS:528:DC%2BC3cXhtFaltL3M&md5=23fcab1fa3e6fdf563269368c098195fCAS |

Gassin R (1993) Utricularia beaugleholei (Lentibulariaceae: subgenus Utricularia: section Pleiochasia), a new species from south-eastern Australia. Muelleria 8, 37–42.

Goloboff PA (1998) ‘NONA.’ (Published by the author: Túcuman, Argentina)

Guindon S, Gascuel O (2003) A simple, fast and accurate method to estimate large phylogenies by maximum-likelihood. Systematic Biology 52, 696–704.
A simple, fast and accurate method to estimate large phylogenies by maximum-likelihood.CrossRef |

Hall TA (1999) BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symposium Series 41, 95–98.

Ibarra-Laclette E, Lyons E, Hernández-Guzmán G, Pérez-Torres CA, Carretero-Paulet L, Chang T-H, Lan T, Welch AJ, Abraham Juárez MJ, Simpson J, Fernández-Cortés A, Arteaga-Vázquez M, Góngora-Castillo E, Acevedo-Hernández G, Schuster SC, Himmelbauer H, Minoche AE, Xu S, Lynch M, Oropeza-Aburto A, Cervantes-Pérez SA, de Jesús Ortega-Estrada M, Cervantes-Luevano JI, Michael TP, Mockler T, Bryant D, Herrera-Estrella A, Albert VA, Herrera-Estrella L (2013) Architecture and evolution of a minute plant genome. Nature 498, 94–98.
Architecture and evolution of a minute plant genome.CrossRef | 1:CAS:528:DC%2BC3sXotFSms7c%3D&md5=3186ae31e181a689c52166ed822ac7a6CAS |

Jobson RW (2012) A new species of Utricularia (Lentibulariaceae) from northern Queensland, Australia. Telopea 14, 49–57.
A new species of Utricularia (Lentibulariaceae) from northern Queensland, Australia.CrossRef |

Jobson RW (2013) Five new species of Utricularia (Lentibulariaceae) from Australia. Telopea 15, 127–142.
Five new species of Utricularia (Lentibulariaceae) from Australia.CrossRef |

Jobson RW, Baleeiro PC (2015) Two new species of Utricularia (Lentibulariaceae) from the North West region of Western Australia. Telopea 18, 201–208.
Two new species of Utricularia (Lentibulariaceae) from the North West region of Western Australia.CrossRef |

Jobson RW, Playford J, Cameron KM, Albert VA (2003) Molecular phylogenetics of Lentibulariaceae inferred from plastid rps16 intron and trnL–F DNA sequences: implications for character evolution and biogeography. Systematic Botany 28, 157–171.
Molecular phylogenetics of Lentibulariaceae inferred from plastid rps16 intron and trnL–F DNA sequences: implications for character evolution and biogeography.CrossRef |

Juniper BE, Robbins RJ, Joel DM (1989) ‘The Carnivorous Plants.’ (Academic: London, UK)

Kearse M, Moir R, Wilson A, Stones-Havas S, Cheung M, Sturrock S, Buxton S, Cooper A, Markowitz S, Duran C, Thierer T, Ashton B, Meintjes P, Drummond A (2012) Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics 28, 1647–1649.
Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data.CrossRef |

Lavin M, Luckow M (1993) Origins and relationships of tropical North America in the context of the boreotropics hypothesis. American Journal of Botany 80, 1–14.
Origins and relationships of tropical North America in the context of the boreotropics hypothesis.CrossRef |

Lowrie A (1998) A new species of Utricularia (Lentibulariaceae) from the south-west of Western Australia. Nuytsia 12, 37–41.

Lowrie A (2002) Utricularia petertaylorii (Lentibulariaceae), a new species from the south west of Western Australia. Nuytsia 14, 405–410.

Müller K, Borsch T (2005) Phylogenetics of Utricularia (Lentibulariaceae) and molecular evolution of the trnK intron in a lineage with high substitutional rates. Plant Systematics and Evolution 250, 39–67.
Phylogenetics of Utricularia (Lentibulariaceae) and molecular evolution of the trnK intron in a lineage with high substitutional rates.CrossRef |

Pagel M (1994) Detecting correlated evolution on phylogenies: a general method for the comparative analysis of discrete characters. Proceedings of the Royal Society of London 255, 37–45.
Detecting correlated evolution on phylogenies: a general method for the comparative analysis of discrete characters.CrossRef |

Pepper M, Doughty P, Keogh JS (2013) Geodiversity and endemism in the iconic Australian Pilbara region: a review of landscape evolution and biotic response in an ancient refugium. Journal of Biogeography 40, 1225–1239.
Geodiversity and endemism in the iconic Australian Pilbara region: a review of landscape evolution and biotic response in an ancient refugium.CrossRef |

Reut MS, Fineran BA (1999) An evaluation of the taxonomy of Utricularia dichotoma Labill., U. monanthos Hook.f., and U. novae-zelandiae Hook.f. (Lentibulariaceae). New Zealand Journal of Botany 37, 243–255.
An evaluation of the taxonomy of Utricularia dichotoma Labill., U. monanthos Hook.f., and U. novae-zelandiae Hook.f. (Lentibulariaceae).CrossRef |

Reut MS, Fineran BA (2000) Ecology and vegetative morphology of the carnivorous plant Utricularia dichotoma (Lentibulariaceae) in New Zealand. New Zealand Journal of Botany 38, 433–450.
Ecology and vegetative morphology of the carnivorous plant Utricularia dichotoma (Lentibulariaceae) in New Zealand.CrossRef |

Reut MS, Jobson RW (2010) A phylogenetic study of subgenus Polypompholyx: a parallel radiation of Utricularia (Lentibulariaceae) throughout Australasia. Australian Systematic Botany 23, 152–161.
A phylogenetic study of subgenus Polypompholyx: a parallel radiation of Utricularia (Lentibulariaceae) throughout Australasia.CrossRef |

Shaw J, Shafer HL, Leonard OR, Kovach MJ, Schorr M, Morris AB (2005) The tortoise and the hare ii: relative utility of 21 noncoding chloroplast DNA sequences for phylogenetic analysis. American Journal of Botany 92, 142–166.
The tortoise and the hare ii: relative utility of 21 noncoding chloroplast DNA sequences for phylogenetic analysis.CrossRef | 1:CAS:528:DC%2BD2MXht1Klsbc%3D&md5=92e13057054b00f2f11658f12b17c693CAS |

Sniderman JMK, Pilmans B, O’Sullivan PB, Kersaw AP (2007) Climate and vegetation in southeastern Australia respond to southern hemisphere insolation forcing in the late Pliocene–early Pleistocene. Geology 35, 41–44.
Climate and vegetation in southeastern Australia respond to southern hemisphere insolation forcing in the late Pliocene–early Pleistocene.CrossRef |

Sniderman JMK, Woodhead JD, Hellstrom J, Jordan GJ, Drysdale RN, Tyler JJ, Porch N (2016) Pliocene reversal of late Neogene aridification. Proceedings of the National Academy of Sciences of the United States of America 113, 1999–2004.
Pliocene reversal of late Neogene aridification.CrossRef |

Stein JL, Hutchinson MF, Stein JA (2014) A new stream and nested catchment framework for Australia. Hydrology and Earth System Sciences 18, 1917–1933.
A new stream and nested catchment framework for Australia.CrossRef |

Sun Y, Li Y, Vargas-Mendoza CF, Wang F, Xing F (2016) Colonization and diversification of the Euphorbia species (sect. Aphyllis subsect. Macaronesicae) on the Canary Islands. Scientific Reports 6, 34454
Colonization and diversification of the Euphorbia species (sect. Aphyllis subsect. Macaronesicae) on the Canary Islands.CrossRef | 1:CAS:528:DC%2BC28Xhs1Smu7jI&md5=0a0ff665b0f58010ccf54320c1e2caf5CAS |

Taylor P (1989) The genus Utricularia. In ‘Kew Bulletin, Additional Series XIV’. (HMSO: London, UK)

Toon A, Crisp MD, Gamage H, Mant J, Morris DC, Schmidt S, Cook LG (2015) Key innovation or adaptive change? A test of leaf traits using Triodiinae in Australia. Scientific Reports 5, 12398
Key innovation or adaptive change? A test of leaf traits using Triodiinae in Australia.CrossRef | 1:CAS:528:DC%2BC2MXhtlCjtrrP&md5=f8866410c33cb565ad925566adbbe540CAS |

Wakabayashi H (2010) Utricularia linearis (Lentibulariaceae), a new species from the Howard Springs, Northern Territory, Australia. The Journal of Insectivorous Plant Society 61, 88–92.

Yu Y, Harris AJ, He XJ (2010) S-DIVA (statistical dispersal-vicariance analysis): a tool for inferring biogeographic histories. Molecular Phylogenetics and Evolution 56, 848–850.
S-DIVA (statistical dispersal-vicariance analysis): a tool for inferring biogeographic histories.CrossRef |

Yu Y, Harris AJ, Blair C, He XJ (2015) RASP (reconstruct dispersal-vicariance analysis): a tool for historical biogeography. Molecular Phylogenetics and Evolution 87, 46–49.
RASP (reconstruct dispersal-vicariance analysis): a tool for historical biogeography.CrossRef |






Abstract PDF (1.1 MB) Export Citation