Elachanthus, Isoetopsis and Kippistia are nested in the genus Minuria (Asteraceae: Astereae)

Sampling

In addition to six species of Minuria and one species each of Elachanthus, Isoetopsis and Kippistia, we selected one or two species to represent each of the subtribes and major genera of Australian Astereae. We did not have a sample of E. glaber Paul G.Wilson available. Voucher information is available in Appendix A1.

We included sequences of the following taxa published by Genomics for Australian Plants (GAP): Tetramolopium sp. Mt Bowen (GAP sample 80121, https://data.bioplatforms.com/gap-illumina-shortread/bpa-gap-illumina-shortread-82312-83654), Elachanthus pusillus F.Muell. (GAP sample 80054, https://data.bioplatforms.com/gap-illumina-shortread/bpa-gap-illumina-shortread-82245-83654), Isoetopsis graminifolia Turcz. (GAP sample 80076, https://data.bioplatforms.com/gap-illumina-shortread/bpa-gap-illumina-shortread-82267-83654), Kippistia suaedifolia F.Muell. (GAP sample 80079, https://data.bioplatforms.com/gap-illumina-shortread/bpa-gap-illumina-shortread-82270-83654), and Chondropyxis halophila D.A.Cooke (GAP sample 80040, https://data.bioplatforms.com/gap-illumina-shortread/bpa-gap-illumina-shortread-82231-83654). We also included sequences of a sample published by the Plant And Fungal Tree Of Life project (PAFTOL), Minuria leptophylla DC. (Sequence Read Archive Accession ERR7619546, https://www.ncbi.nlm.nih.gov/sra/ERR7619546).

Laboratory procedures

Genomic DNA was extracted from 5–15 mg of silica-dried leaf tissue or herbarium material by using a DNeasy Plant 96 kit (QIAGEN, Venlo, Netherlands), following the manufacturer’s instructions. DNA concentration was quantified using a Fluorskan fluorescent microplate reader (Thermo Fisher, Waltham, MA, USA) and the Quant-iT 1X dsDNA HS kit (Thermo Fisher), following the manufacturer’s instructions. DNA was either diluted to 10 ng μL^–1 in a total volume of 100 μL and transferred into 0.65-mL Bioruptor tubes, or remained undiluted and 55 μL was aliquoted into an 8-microTUBE-50 AFA Fiber Strip V2 (Covaris, Woburn, MA, USA; catalogue number PN520174). Following the manufacturer’s instructions, DNA was sonicated to a mean size of ~250 bp in the Bioruptor Pico (Diagenode, Denville, NJ, USA) for four cycles of 15 s on and 90 s off, and after a brief pulse centrifugation, for a further three cycles of 15 s on and 90 s off. Following the manufacturer’s instructions, some samples were sonicated to a mean insert size of ~250 bp in the Covaris LL220 ultrasonicator (4°C; 0.5-mm X- and Y-dither at 10 mm s^–1; peak incident power 450 W, duty factor 15%, cycles per burst 1000, treatment time 120 s per 8 strips). Samples were either diluted to 15 ng μL^–1 by using ultrapure water, or were concentrated to a volume of 15 μL in a SpeedVac (Eppendorf, Hamburg, Germany) vacuum centrifuge set to 60°C.

Libraries were built from <1 to 5 ng of DNA by using the QIAGEN Ultralow Input Library kit (Qiagen, Melbourne, Vic., Australia). Sequence capture was conducted on pools of 16 libraries byusing the Angiosperms353 (Johnson et al. 2019) MYbaits kit (Daicel Arbor Biosciences, Ann Arbor, MI, USA), following the manufacturer’s instructions. Enriched libraries were sequenced on Illumina NovaSeq. 6000 SP with v1.5 paired-end 2 × 150-cycle chemistry with 8-bp dual indexing.

Bioinformatics

Demultiplexed reads were quality filtered and paired with TRIMMOMATIC (ver. 0.39, see http://www.usadellab.org/cms/index.php?page=trimmomatic; Bolger et al. 2014) with illuminaclip:adapters, fa:4:20:10, minimum length of 30, and average quality of 25, and then further filtered with bbduk (ver. 38.90, see https://github.com/BioInfoTools/BBMap/blob/master/sh/bbduk.sh) with entropy of 0.8, entropy window of 20, and entropy mask t. Reads were assembled against target sequences by using hybpiper-nf (see https://github.com/chrisjackson-pellicle/hybpiper-nf; Jackson et al. 2023), a Nextflow pipeline adapted from HybPiper (ver. 1, see https://github.com/mossmatters/HybPiper; Johnson et al. 2016) against a target file designed for broad representation of Asteraceae by mining transcriptome data for angiosperms353 targets (McLay et al. 2021).

The results of HybPiper’s paralog finder were analysed with the monophyletic outgroup (MO) algorithm as implemented in paragone-nf (see https://github.com/chrisjackson-pellicle/ParaGone; Jackson et al. 2023), a Nextflow pipeline for the four gene tree-based paralogy-resolution algorithms collated by Yang and Smith (2014). We chose this algorithm because it returns at most one ortholog group for each locus, producing a more complete sample × gene matrix than do alternative algorithms that return more ortholog groups with, on average, fewer sequences.

For both paralogy resolution and phylogenetic analysis, we selected one representative each from the following four tribes closely related to Astereae as outgroups: Dimorphotheca pluvialis (L.) Moench (Calenduleae, Sequence Read Archive, SRA, Accession ERR7618447, https://www.ncbi.nlm.nih.gov/sra/ERR7618447), Cotula coronopifolia L. (Anthemideae, Genomics for Australian Plants, GAP sample 80045, https://data.bioplatforms.com/gap-illumina-shortread/bpa-gap-illumina-shortread-82236-83654), Ewartia nubigena (F.Muell.) Beauverd (Gnaphalieae, GAP sample 79643, https://data.bioplatforms.com/gap-illumina-shortread/bpa-gap-illumina-shortread-81656-83650), and Abrotanella nivigena (F.Muell.) F.Muell. (Senecioneae, GAP sample 80011, https://data.bioplatforms.com/gap-illumina-shortread/bpa-gap-illumina-shortread-82202-83654).

Paralogy resolution using the MO algorithm returned 312 of the originally 352 loci, excluding the remainder for absence of any outgroup sequence or non-monophyly of the ingroup, potentially indicating paralogy in the outgroup. We excluded another three ortholog groups that had data for fewer than 10 samples, leaving 309. Retained ortholog groups had data for 10–32 samples (median 31), and samples had data from 229 to 309 ortholog groups (median 296.5).

Custom-written Python scripts (see https://bitbucket.csiro.au/projects/NRCA/repos/bioinformatics-and-phylogenetics/browse) were used to ensure that gene alignments were in frame and to concatenate them into a supermatrix. The concatenated dataset comprised 32 samples and 197,877 characters, of which 32,227 were parsimony informative, 40,839 variable but uninformative and 124,811 constant.

Phylogenetic analysis

A phylogeny of the concatenated supermatrix was inferred with IQ-TREE (ver. 2.2.0.5, see http://www.iqtree.org/; Minh et al. 2020), partitioning the alignment by codon positions and under automatic partition and model testing (Chernomor et al. 2016; Kalyaanamoorthy et al. 2017). Testing resulted in the three codon-position partitions being maintained, with the first two under the GTR + F + I + G4 model, and the third under GTR + F + G4. One thousand ultrafast bootstrap (UFB) replicates were used to estimate branch support (Hoang et al. 2018).

We also inferred a phylogeny with the short-cut species-tree inference approach implemented in ASTRAL-PRO (ver. 1.1, see https://github.com/chaoszhang/A-pro; Zhang et al. 2020) that accepts multiple sequences per terminal under the assumption that they represent paralogs. We aligned the 351 output fasta files from HybPiper’s paralog finder by using MAFFT (ver. 7.490, see https://mafft.cbrc.jp/alignment/software/; Katoh and Standley 2013) under automatic algorithm choice and inferred gene trees using IQ-TREE (ver. 2.2.0.5) under automatic model testing. We concatenated all resulting gene trees into a single text file and removed paralog identifiers with a custom R script. ASTRAL-PRO was run with default parameters, inferring branch local posterior probability (LPP) as the clade support value.

Taxonomic history and morphology

Minuria was described by de Candolle (1836), and last fully revised by Lander and Barry (1980a), although several species have been described or transferred to the genus since (Lander 1987; Short 1991; Short and Hosking 2000). As currently recognised by the Australian Plant Census (see https://biodiversity.org.au/nsl/services/search/taxonomy, accessed 25 September 2023), it comprises 12 species and 1 putative phrase-named taxon. Its morphology varies from small shrubs less than 100 cm tall to perennial or annual herbs. The involucre consists of three or four rows of uniform (undifferentiated), herbaceous phyllaries with a scarious margin. Capitula have several rows of radiating ray florets of varying colours exceeding the length of the involucre and yellow disc florets. Pappi are markedly dimorphic between ray and disc florets. The fertile ray cypselae are glabrous or variously hairy and bear pappi of barbellate or smooth bristles. The sterile disc cypselae are glabrous or pubescent and bear pappi of free barbellate bristles or scales ending in bristles.

Several of the species currently placed in Minuria were originally described in the genera Elachothamnos DC., Minuriella Tate and Therogeron DC., but these genera have long been synonymised (Bentham 1867; Lander and Barry 1980a).

Kippistia and its only species K. suaedifolia were described by Mueller (1859). It is an aromatic shrub up to 60 cm tall, with somewhat succulent leaves. The involucre consists of three rows of uniform (undifferentiated), herbaceous phyllaries with a scarious margin. Capitula have several rows of yellow ray florets exceeding the involucre and yellow disc florets. Pappi are markedly dimorphic between ray and disc florets. The sterile or fertile ray cypselae are basally pubescent, and their pappi are either free barbellate bristles or basally united into a scale-like cup. The fertile disc cypselae are glabrous, and their pappi are basally united into a scale-like cup but apically free, barbellate bristles.

Bentham (1867) synonymised Kippistia under Minuria. The genus was reinstated by Lander and Barry (1980b) on the basis of its leaves being strongly aromatic (v. no or weak scent in Minuria), ray florets being yellow like the disc florets (v. ray florets being a different colour from the disc), style branches of ray florets without stigmatic lines (v. having papillose stigmatic lines), and ray florets being often sterile and disc florets fertile (v. ray florets being fertile and disc florets sterile). Lander and Barry (1980b) also discussed the morphology of ray pappi, but variation in Kippistia suaedifolia is too great for a clear contrast (bristles free or connate v. always free).

Isoetopsis and its only species I. graminifolia were described by Turczaninow (1851). It is a small, nearly stemless annual with capitula sessile between linear leaves. The involucre consists of two rows of uniform, herbaceous phyllaries with a scarious margin. Capitula have several rows of pistillate outer florets with strongly reduced, non-radiating, greenish corollas, and yellowish-green disc florets. Cypselae produced by the outer florets are densely covered with long sericeous trichomes and bear pappi of erect scales about as long as the cypselae. The disc florets are epappose and do not produce cypselae, because their styles remain mostly closed and lack stigmatic lines.

Since its description, Isoetopsis has undergone neither synonymisation nor addition of new species, but because of its reduced habit, its tribal placement was long controversial. Originally considered to be part of Anthemideae, it was transferred to Astereae only in 1973 on the basis phyllary morphology, its pappus of scales, and pollen morphology (Robinson and Brettell 1973). However, in his comprehensive cladistic study of Gnaphalieae, Anderberg (1991) argued for its affiliation with that tribe on the basis of the presence of ectomycorrhiza, a divided phyllary stereome, and narrow apical anther appendages. Molecular phylogenies have since confirmed the placement of Isoetopsis in Astereae as being correct (Bayer and Cross 2002).

Elachanthus was described by Mueller (1853) with one species, E. pusillus. The second species currently accepted by the Australian Plant Census, E. glaber, was added by Wilson (1965). A third name, E. occidentalis S.Moore, is considered to be of uncertain application. Both accepted species are annual herbs up to 6 cm tall. The involucre consists of two or three rows of uniform, mostly herbaceous phyllaries with a thin scarious margin. The receptacle is naked. Capitula have several rows of pistillate outer florets with strongly reduced, non-radiating, greenish corollas and yellowish-green discflorets. Cypselae produced by the outer florets are densely covered with long sericeous hairs and bear pappi of slightly patent scales about as long as the cypselae. The disc florets do not produce cypselae. Elachanthus pusillus and E. glaber are very similar to each other but differ in the stems and leaves of the former species being puberulous (v. glabrous), its cypsela surfaces uniformly hairy (v. hairs restricted to the areas between longitudinal ribs), and pappi of disc florets of free bristles (v. linear scales that are basally fused) (Wilson 1965). Elachanthus has never been synonymised, and neither of its species have ever been assigned to other genera.

The morphological differences between the genera as currently circumscribed are summarised in Table 1.

Morphological affinities of the four genera

Of the characters used to justify segregation of Kippistia at the genus level, ray floret colour is highly variable in Minuria, including within several species (Lander and Barry 1980b). The fertility of disc cypselae, the sterility of ray cypselae, and the absence of stigmatic lines on the style branches of ray florets appear to be of greater importance. However, it is plausible to assume that all three reproductive traits are directly and causally connected, as one or the other type of floret would have to be fertile for the species to persist, and the stigmatic lines missing in the sterile florets are another term for the stigmatic surface required to allow fertilisation in Asteraceae (Roque et al. 2009).

In other words, the main arguments for the separation of Kippistia from Minuria would have to be the fertility of disc florets and stronger leaf aroma. Even without phylogenetic data, this difference would suggest apomorphic segregation, i.e. it would have to be considered likely that fertile ray cypselae and sterile disc cypselae are the ancestral state for Minuria and Kippistia, and that K. suaedifolia is potentially a member of Minuria that evolved to switch these states, as our phylogeny now indicates to be the case.

As discussed above, Kippistia had been synonymised under Minuria from 1867 until 1980, and our results suggest that Bentham’s (1867) synonymisation reflects evolutionary relationships.

By contrast, Elachanthus and Isoetopsis have never been suggested to be part of Minuria, and in fact Isoetopsis has historically been considered difficult to place even in a tribe (Robinson and Brettell 1973; Anderberg 1991; Bayer and Cross 2002). However, with the benefit of our phylogeny prompting us to undertake to a direct comparison, it may now seem surprising that this was the case.

Elachanthus and Isoetopsis are strikingly similar with annual habit and linear leaves (Fig. 2c, d), nearly indistinguishable involucre (Fig. 2g, h), and equally long-sericeous cypselae and pappi of scales about as long as the cypselae (Fig. 2k, l). The main differences between the two genera are the absence of elongated stems in Isoetopsis and its straight, instead of slightly patent, orientation of the pappus scales. Isoetopsis graminifolia may have evolved from an Elachanthus-like ancestor through the shortening of stem internodes.

Fig. 2.

(a–d) Habit, (e–h) phyllaries and (i–l) cypselae of representative specimens of Kippistia, Minuria sens. str., Elachanthus and Isoetopsis. (a, e, i) Kippistia suaedifolia (Hj.Eichler 21259). The cypsela (i) is immature but shows the scale-like bases of the pappus elements. (b, f, j) Minuria annua (W.E.Mulham s.n., CANB 616618). (c, g, k) Elachanthus pusillus (M.D.Crisp 638), (d, h, l) Isoetopsis graminifolia (A.N.Schmidt-Lebuhn 1500 & A.H.Thornhill). All vouchers specimens for these photographs are part of the collections at CANB.

It is more understandable that Elachanthus and Isoetopsis have traditionally not been associated with Minuria, given that in contrast to them, Minuria is predominantly perennial and has multiple rows of showy ray florets. However, the morphological variation represented by this larger genus encompasses all elements required to make the evolution of Elachanthus and Isoetopsis from within it plausible.

In terms of habit, Minuria includes annuals even in its current circumscription, and M. annua is, despite its elongated stems, reminiscent of Isoetopsis in how subtending leaves overtop the capitula (Fig. 2b). The phyllaries of Minuria vary in shape from linear to lanceolate, and they are arranged in three or four, as opposed to two or three, rows, but their uniform morphology, green herbaceous stereome and scarious margin match those of the smaller genera (Fig. 2e–h). Elachanthus and Isoetopsis share with Minuria, as currently circumscribed (i.e. excluding Kippistia), that only the outer florets form fertile cypselae, because the inner are male. Although the indumentum of the cypselae is variable in Minuria, some species, including M. leptophylla and M. annua, are densely covered in sericeous hairs strikingly similar to those in Elachanthus and Isoetopsis (Fig. 2j–l). Finally, several species of Minuria exhibit a tendency for the base of pappus bristles to be broadened into scales (e.g. M. multiseta P.S.Short), which makes it plausible that entirely scale-like pappi could have evolved from this ancestral state.

Minuria DC., Prod. 5: 298 (1836)

Type: Minuria leptophylla DC., fide N.S.Lander & R.Barry, Nuytsia 3(2): 222 (1980).

Eurybiopsis DC., Prodr. 5: 260 (1836). Type: Eurybiopsis macrorhiza DC.

Therogeron DC., Prodr. 5: 283 (1836). Type: Therogeron denticulatum DC., fide N.S.Lander & R.Barry, Nuytsia 3(2): 222 (1980).

Elachothamnos DC., Prodr. 5: 398 (1836). Type: Elachothamnos cunninghamii DC.

Isoetopsis Turcz., Bull. Soc. Imp. Naturalistes Moscou 24(1): 174–175 (1851). Type: Isoetopsis graminifolia Turcz.

Elachanthus F.Muell., Linnaea 25: 410 (1853). Type: Elachanthus pusillus F.Muell.

Kippistia F.Muell., Rep. pl. Babbage’s exped. 3(1): 12 (1859). Type: Kippistia suaedifolia F.Muell.

Minuriella Tate, Trans. Proc. & Rep. Roy. Soc. South Australia 23: 288 (1899); Minuria sect. Minuriella (Tate) Lemée, Dict. gen. pl. phan. 4: 491 (1932). Type: Minuriella annua Tate.

Erect or prostrate, small shrubs, or annual or perennial herbs. Stems herbaceous or woody, glabrous or variously pubescent. Leaves alternate, sometimes clustered, sessile, linear, lanceolate, ovate, obovate, spathulate, or falcate, glabrous or variously pubescent, sometimes with leaves over-topping the capitula; margin entire, undulating, finely serrate or conspicuously dentate; apex obtuse, acute or acuminate. Capitula pedunculate or sessile between subtending leaves, solitary or rarely clustered, terminal. Involucral bracts in 2–4 rows, linear to lanceolate or narrow-elliptic, undifferentiated, grading in size or dimorphic, glabrous or variously pubescent, with 0–2 prominent ribs; herbaceous with scarious margin, entire or denticulate; apex acute to acuminate, entire or fimbriate, green or tinged pink. Receptacle naked, flat to noticeably convex. Ray florets many, in 2 or more rows, estaminate except in M. suaedifolia; ligules white, violet, mauve, blue, lilac, lavender to pink, or yellow, often variable in the same species, conspicuous and exceeding the involucre, or greenish and strongly reduced in some annuals; floral tube glabrous; style branches subulate to lanceolate, with conspicuous papillose stigmatic lines except in M. suaedifolia, cypsela fertile except in M. suaedifolia, brown, reddish-brown, red, orange or yellow, ±prominently ribbed, glabrous or variously pubescent, slightly flattened; pappus of several to many barbellate or smooth bristles or of several scales. Disc florets staminate except in M. suaedifolia, yellow or yellowish-green, pentamerous, rarely tetramerous; floral tube glabrous or variously pubescent with multicellular, biseriate hairs; anther bases obtuse; style branches subulate or narrowly lanceolate, pubescent on dorsal surfaces to below point of bifurcation; cypsela sterile except in M. suaedifolia, glabrous or pubescent with notched twin-hairs, translucent or opaque, white, straw-coloured or reddish-brown, flattened; pappus very variable, of undifferentiated or dimorphic capillary or barbellate bristles or scales or branching towards apices, or a cup of fused scales surmounted by 1–8 bristles, or absent.

Minuria graminifolia (Turcz.) Schmidt-Leb. & Steph.H.Chen, comb. nov.

Isoetopsis graminifolia Turcz., Bull. Soc. Imp. Naturalistes Moscou 24(1): 174–175, Tab. III (1851). Type citation: ‘Nova Hollandia. Drum. IV. n. 207.’ Type: Nova Hollandia, s. dat., J.Drummond IV. n. 207 (holo: KW001001452 [image seen]; iso: G00302776 [image seen], G00302777 [image seen], K000890262 [image seen], MEL2279209 [image seen]).

We here follow the advice of Mosyakin et al. (2019) (especially pp. 382–383) to treat original material of Turczinanow names held at KW as holotypes. However, in this instance, mounted on the same sheet as J.Drummond 207 (KW001001452) are two other gatherings of the same species, namely, J.Drummond 390 (‘Nova Hollandia, Drummond coll. V n. 390’, KW001001451) and F.Mueller s.n. (‘Murray’, KW001001453). There is no indication on the sheet which of the 14 plants mounted together belong to which of the three gatherings, with the exception of a single plant that touches a blue label reading ‘390’ and can therefore be inferred to belong to J.Drummond 390. Although the sheet contains the holotype of Isoetopsis graminifolia, which of the plants on the sheet comprise the holotype is an open question. The material at G, K, and MEL is clearly labelled as J.Drummond 207 and these specimens are considered to represent isotypes.

Minuria glabra (Paul G.Wilson) Schmidt-Leb. & Steph.H.Chen, comb. nov.

Elachanthus glaber Paul G.Wilson in Hj. Eichler (Ed.), Suppl. J.M. Black’s Fl. S. Australia (2nd edn, 1943–1957): 304–305 (1965). Type: east of Flinders Range, Koonamore (~60 km north of Yunta), Koonamoore Homestead, 16 Aug. 1956, Hj.Eichler 12515 (holo: AD 95731053 [image seen]).

Minuria pusilla (F.Muell.) Schmidt-Leb. & Steph.H.Chen, comb. nov.

Elachanthus pusillus F.Muell., Linnaea 25: 411 (1853). Type citation: ‘In collibus siccis ab Arkaba ad Cudnaka’. Type: Akaba & Cudnaka, an dürren Hügeln, N. Holl. Austr., Oct 1851, F.Mueller s.n. (lecto, here designated: MEL2160665 [image seen]; Cudnaka, (isolecto: GH00006518 [image seen]); Arkaba, N. Holl. austr. interior (isolecto: MEL604807 [image seen], MEL604808 [image seen]).

?Elachanthus occidentalis S.Moore, J. Linn. Soc., Bot. 34: 196. 1899. Type citation: ‘Juxta Coolgardie floret et fructificat mens. Aug.’ Type: West Australian Goldfields, 1895, S.Moore s.n. (NY 00168310 [image seen]).

Of the three syntypes of Elachanthus pusillus at MEL of which we have seen images (JSTOR Global Plants, see https://plants.jstor.org, accessed 16 September 2023), MEL604808 comprises only a single plant, and MEL604807 comprises only two plants and some fragments, whereas MEL2160665 comprises five plants and some fragments. More importantly, label data of MEL2160665 are the closest match of all syntypes to the type citation, mentioning both Arkaba and Kanyaka (as ‘Akaba’ and ‘Cudnaka’ respectively) and the habitat in a traditional German handwriting script, which can reasonably be translated to Latin as ‘in collibus siccis’. We therefore designate this specimen as the lectotype. A fourth specimen at GH differs from the MEL specimens in not mentioning Arkaba and being labelled only ‘Cudnaka’, but refers to the publication of the name (‘Linnaea 1852, p. 411’), suggesting that it may be part of the same gathering.

The status of Elachanthus occidentalis is beyond the scope of this study, but we here tentatively consider it to be a potential synonym of Minuria pusilla on the basis of indumentum and geographic occurrence.

Minuria suaedifolia (F.Muell.) F.Muell. ex Benth., Fl. Austral. 3: 499 (1867)

Kippistia suaedifolia F.Muell., Rep. Pl. Babbage’s exped. 13 (1859); Therogeron suaedifolius (F.Muell.) Kuntze, Revis. Gen. Pl. 1: 369 (1891), as suaedifolia. Type citation: ‘Stuart’s Creek’. Type: Between Streaky Bay & Venus Bay, s. dat., Babbage s.n. (neo: MEL 70481 [image seen], fide N.S.Lander & R.Barry, Nuytsia 3(2): 21 (1980)).

Minuria kippistiana F.Muell., Victoria Parliamentary Papers Votes Proc. Legisl. Assembly 3(19): 17 (1861), nom. inval., nom. nud.