Mychodea and the Mychodeaceae (Gigartinales, Rhodophyta) revisited: molecular analyses shed light on interspecies relationships in Australia’s largest endemic algal genus and familyGerald T. Kraft A C and Gary W. Saunders B
A School of Biosciences, University of Melbourne, Parkville, Vic. 3010, Australia.
B Department of Biology, University of New Brunswick, PO Box 4400, Fredericton, NB, E3B 5A3, Canada. Email: email@example.com
C Corresponding author. Present address: Tasmanian Herbarium, College Road, Sandy Bay Tas. 7005, Australia. Email: firstname.lastname@example.org
Australian Systematic Botany 30(3) 230-258 https://doi.org/10.1071/SB16058
Submitted: 12 December 2016 Accepted: 7 July 2017 Published: 20 October 2017
The red algal genus Mychodea Hook.f. & Harv. is not only Australia’s largest wholly endemic macroalgal genus, it and the family Mychodeaceae (of which it is the sole member) appear to be the largest completely endemic algal genus and family from any continental landmass in the world. Kraft’s 1978 morpho-taxonomic monograph credited Mychodea with 11 species varyingly distributed between Geraldton, Western Australia, south and eastward across the coasts of South Australia, Victoria and Tasmania, and northwards into southern New South Wales. Dismissed or discounted was every former extra-Australian attribution of the genus. In the over 40 years since completion of the research, further explorations of marine habitats in Australia have uncovered additional species, and the application of molecular-assisted taxonomic and phylogenetic methodologies has now allowed a substantial refinement of Mychodea systematics. We here document 19 Mychodea species, for 16 of which we have molecular data that support inferences of probable species relationships. To the 11 species treated by Kraft we now add 4 that are recently discovered, resurrect 2 that were synonymised with a third species in his 1978 work, and treat 2 species-level Western Australian entities that remain unnamed for lack of sufficient reproductive material. Mychodea is characterised by elaborate vegetative structures and some of the most complex fertilisation, diploidisation and embryogenesis processes of any red alga, which we detail and illustrate. Distinguishing features of the individual species are highlighted, some of which are particularly unusual.
Additional keywords: biogeography, macroalgae, molecular-assisted alpha taxonomy, morpho-taxonomy, phylogenetics.
Forty-three years ago, the first author had just completed a morpho-taxonomic monograph of the southern Australian red-algal genus Mychodea Hook.f. & Harv., which was then and still remains the continent’s most species-rich wholly endemic algal genus of any class or phylum. Although long enjoying this distinction, the genus and monotypic family to which it belongs were, before Kraft’s (1974) PhD thesis, among the most obscure and least-studied elements of Australia’s rich macroalgal flora.
When Kraft began his research in the early 1970s, the Mychodeaceae of Kylin (1932, p. 62) consisted of the genera Mychodea and Ectoclinium J.Agardh (synonymous with the previously published Neurophyllis Zanardini), together comprising some 18 species from not only Australia but also South Africa, the Caribbean, California and Mauritius. Kraft’s published monograph (Kraft 1978) subsumed Neurophyllis in Mychodea, removed all non-Australian species and records to other genera or discounted their being true members of Mychodea, and synonymised several on the basis of both nomenclatural and taxonomic grounds. In the end, he credited 11 wholly endemic Australian species to the genus. These results remained unchanged in a later shortened taxonomic treatment of the Mychodeaceae by Kraft and Womersley (1994) in the first volume of Womersley’s monograph of the marine red algae of southern Australia.
As had Kylin (1956, p. 238) before him, Kraft (1978, pp. 574–578) also speculated about possible phylogenetic relationships of the family within the order Gigartinales on the basis of features of vegetative architecture, gonimoblast initiation, carposporophyte development and tetrasporangial origin, and, in addition, proposed some possible paths of evolution taken by its many species.
In all, 10 of the 11 species that Kraft recognised and all those that he regarded as heterotypic synonyms had been named and described in varying degrees of detail by the year 1897, after which no further additions from Australia were made until his own (Kraft 1978, p. 563) proposal of Mychodea acanthymenia Kraft. Without exception, these species were obscure (although some were at times common in drift) and had never been cited as ecologically significant elements of the continent’s marine floras. References to them had mainly appeared as inclusions in unillustrated lists compiled during the late 19th and first three-quarters of the 20th centuries. However, when studied in detail they proved to have many distinctive features (as highlighted below) and to be among the most complexly organised red algae in terms both of vegetative structure and, especially, sexual reproductive processes. However, the number of actual species, as well as how the members might be related within the genus, were matters all but beyond guesswork to classical α-taxonomists before the advent of molecular-phylogenetic methodologies, which the second author has now applied to the bulk of the families of the order Gigartinales and most members of Mychodea. This has resulted in trees implying family and species relationships, the resurrection of two species subsumed by Kraft (1978), the naming and description of four new species, and the identification of at least two more that are not formally described for lack of more extensive and reproductive collections.
The genus Mychodea Hook.f. & Harv., London J. Bot. 6: 407 (1847)
The history of the genus and of the many additions to Hooker & Harvey’s initial two Tasmanian species, M. carnosa Hook.f. & Harv. and M. membranacea Hook.f. & Harv., were outlined by Kraft (1978, p. 518). The following is the suite of features that together form the anatomical criteria for recognising a member of the genus and family:
major axes are basically one of the following three types: cylindrical or nearly so (9 spp.); compressed, so that axes are approximately twice as broad as thick (4 spp.); and conspicuously flattened (6 spp.).
All are uniaxial, the apical cell either distinct (Figs 2A, C, 3A, 12D, 16E) or, in the case of the flattened forms like M. marginifera (Aresch.) Kraft, soon indistinguishable from surrounding outer cortical cells at the tips (Kraft 1978, figs 14B, 16G, 38C). Central-axial filaments are distinct (Figs 2B, 3A), at least distally, in most species (Kraft 1978, fig. 16G), where the successive cells link up around their primary pit-connections by short adventitious cells or filaments (Fig. 2B, C, H; Kraft 1978, figs 18G, 20B) that arise from the younger and connect to the older of the two cells (what Kraft called ‘conjunctor’ cells or filaments). Further elaborations of these conjunctors surround the central-axial filaments proximally and lead to the latter becoming largely indistinguishable in lower cross-sections of most species (Fig. 2D–F, I), the one major exception to this being Mychodea australis (Zanardini) Kraft (Fig. 2G, H), in which cells of the central-axial filaments are conspicuously inflated in cross- and long-sections throughout.
Surrounding the central core of filaments is a pseudoparenchymatous outer medulla, the interior cells of which in several species enlarge substantially (Figs 2D–F, 14D, 15D, 16D), imparting the feature for which Hooker and Harvey named the genus, ‘Mychodea’ being derived from a Greek word meaning ‘an internal cavity or secret chamber’ (Harvey 1860, pl. 142). Running from the central filamentous core and growing between the inflated medullary cells to hook up with the progressively smaller cells of the inner cortex in most species are varying numbers of ramified filaments of thin elongate cells (Figs 2D–F, 16F, G; Kylin 1956, fig. 237B). Medullas of compressed (Fig. 2I) and flattened (Fig. 2J) species do not generally display an interior layer of greatly inflated cells and have few, if any, intercalated filaments between the pseudoparenchymatous cells.
Holdfasts are thin (Fig. 3A) to fleshy (Fig. 3B) discs in most species, exceptions being Mychodea ramulosa J.Agardh (Fig. 3C) and M. minutissima G.W.Saunders & Kraft (Fig. 12B, C), in which erect axes arise from creeping stolons anchored by short haptera (Fig. 3C) or from secondarily attached lower laterals (Fig. 12B, C). In M. aciculare (J.Agardh.) Kraft and M. pseudoaciculare G.W.Saunders & Kraft, initially basal holdfasts and lower main axes disintegrate and are ultimately completely replaced by a thick weft of anchoring stolons (Figs 3D, E, 15A; Kraft 1978, fig. 18C) of an apparently obligate hydrozoan that course upwards through the interiors of all axes and branch orders. Two of the species (M. hamata Harv., Fig. 3F, and M. aciculare, Fig. 3G) have, in addition to basal holdfasts, the ancillary means of attaching and vegetatively propagating in the form of hooks, which are almost always present in the first, and less frequently so, in the second.
Thalli grow on a variety of substrata, including rocks, wooden pilings, seagrass blades (e.g. Posidonia ostenfeldii Hartog, Fig. 10H) or stems (mainly Amphibolis antarctica (Labill.) Sond. & Asch. ex Asch., Fig. 11F), larger algae such as the fucoids Acrocarpia paniculata (Turner) Aresch. and Cystophora spp., stalked tunicates and sponges. Three of the species (M. pusilla (J.Agardh.) Harv., M. spinulifera J.Agardh and M. echinocarpa Kraft & G.W. Saunders) appear to be obligate epiphytes of the woody stalks of Amphibolis antarctica, and other species, such as M. membranacea and M. terminalis Harv., are most frequently found on that seagrass.
Gametangial thalli are monoecious. Spermatangia occur in sunken ampullar clusters (Figs 4A, 12G, 13F, 16I) borne on subcortical mother cells scattered across the frond surfaces, usually proximal to the cystocarps.
Diploidisation is procarpic, the presumably fertilised carpogonium apparently first establishing a narrow connecting isthmus (Fig. 4E), then fusing fully onto the associated supporting cell (Fig. 4F), which then serves as the generative auxiliary cell from which the embryo (carposporophyte) develops. Supporting cells are intercalary within inner-cortical filaments (Fig. 5A).
Initial embryogenesis involves several inwardly and laterally directed narrow arms that radiate from the now-diploidised generative auxiliary cell (Fig. 5B). Multiple gonimoblast initials are cut off from the arms (Fig. 5C; Kylin 1956, fig. 236D), the filaments then entering the surrounding gametophytic tissue (Figs 5D–E, 6A; Kraft 1978, figs 9A, 13B, 15B, 17C, 19G, 22B) where they intermingle and form numerous fusions and secondary pit-connections (Fig. 6B, C), apparently diploidising large numbers of initially haploid gametophytic vegetative cells.
As the carposporophyte matures, diploid nuclei implanted by gonimoblasts into vegetative cells result in the budding off of carposporangial initials, which form irregular chains of carposporangia linked by primary and secondary pit-connections (Figs 6D, 7A; Kraft 1978, figs 4A–C, 6C, 9B, C, 11B, C, 13C, 15C, 19H, I).
Carposporophytes lack a Faserhülle (a differentiated jacket of surrounding sterile gametophytic filaments) in all (Figs 7B–E, 12I, 14E, 15F, 16J) but Mychodea herringtoniana, in which this feature is usually and uniquely prominent at intermediate stages of cystocarp development (Fig. 13I, J).
Cystocarps are non-ostiolate, the shedding of carpospores effected by the wearing away (Fig. 4D, E) or progressive splitting apart (Figs 7B, 13K, 14E) of cortical tissue external to the auxiliary cell.
Cystocarps are protuberant and occupy one of the following three basic positions on bearing axes: (1) subapical (Fig. 8D–F) or at the bases of spines or short spinous laterals (Fig. 8H, I), often appearing terminal (Fig. 8A, B) or sessile with the shunting aside or wearing away of branch apices (Fig. 8A); (2) submarginal at spatulate frond apices (Fig. 8C) or (in the sole case of M. herringtoniana, Figs 8G, 13A) lining lateral blade margins; and (3) lateral (Fig. 14B; Kraft 1978, fig. 5A) and either weakly (Fig. 8K) or sharply (Fig. 8L) bending the axes at their sites. In the sole case of M. minutissima (Figs 8M, 12C), cystocarps are single in stalked terminal subspherical or lanceolate pods.
Tetrasporangia are zonate and scattered in the outer cortex, with the exceptions of M. marginifera, M. australis (Fig. 9C) and M. sp_2WA, in which they occur in sori or slightly raised nemathecia. In the type species, M. carnosa, as well as M. terminalis and M. membranacea, the sporangia are intercalary (Figs 8M, 9A, 14G, H), each sporangium subtending one or two layers of outer cortical cells. In the other species they are terminal (Figs 9B, 13E), which led Kraft (1974) to initially place them all in Zanardini’s (1874, p. 499) genus Neurophyllis, a decision (as molecular data now indicate) fortunately not followed in his published monograph (Kraft 1978).
The present study aims to clarify the systematics of Mychodea through a combination of morphological and molecular-taxonomic approaches and to illustrate the utility of these disciplines in helping unravel the phylogenetic relationships between the many species of this distinctive Australian-endemic genus.
Materials and methods
In total, 97 specimens, including 77 specimens of Mychodea spp., were considered in the present study (Table S1, available as Supplementary material for this paper). Samples were processed in the field as outlined in Saunders and McDevit (2012), with pressed vouchers made when feasible. Extraction of DNA and COI-5P amplification followed Saunders and McDevit (2012), whereas LSU, rbcL and rbcL-3P amplification followed Saunders and Moore (2013). The primer pair used for amplification of COI-5P for each specimen was recorded with that accession on the Barcode of Life Database (BOLD) website (http://www.boldsystems.org/) and GenBank (Table S1). Amplified products were sent to Genome Quebec for sequencing, with data edited and aligned in Geneious, ver. 10.0.5 (Biomatters, Auckland, New Zealand). Additional data for all markers were obtained from GenBank (Table S1).
The COI-5P alignment (52 specimens, 664 sites) for Mychodea spp. was subjected to barcode gap analysis on BOLD to determine within-species variation and the distance to the nearest neighbour. The rbcL-3P data (included truncated full rbcL sequences; 37 specimens, 800 sites) for Mychodea spp. were screened for total differences in Geneious (there were no indels, all changes being substitutions).
For phylogenetic analyses, single-gene alignments for LSU (27 taxa, 2815 sites), rbcL (28 taxa, 1358 sites) and COI-5P (25 taxa, 664 sites; Table S1) were subjected to maximum-likelihood (ML) analyses using a GTR+I+G model, with partitioning by codon for the two protein-coding genes, in the program RAxML (Stamatakis et al. 2008), as implemented in Geneious. Robustness was determined with 500 replicates of bootstrap resampling. Because no discrepancies were noted, a multigene alignment (28 taxa, 4837 sites) of the three genes was generated and subjected to ML analyses as noted previously. The tree was rooted on other families of the Gigartinales known to be allied to the Mychodeaceae (Kraft and Saunders 2014; Table S1).
The range of habits exhibited by Mychodea species is illustrated below from dried herbarium specimens showing typical forms, or, in the cases of all but two of the taxa described or designated below as new, from either wet or dried voucher material used in the DNA analyses themselves. Several illustrations from the 1978 monograph in the Australian Journal of Botany (Vol. 26, pp. 515–610) have been utilised in the presentation below, with the publisher’s permission.
Micrographs made since 1978 have been made on a Zeiss Axioskop 2 (Carl Zeiss Microscopy LLC, Thornwood, NY, USA) fitted with an AxioCam MRc5 digital camera (Carl Zeiss Promanade 10, Jena, Germany), the images processed on Adobe Photoshop CS6, ver. 13.0 x64 (Softonic International, Barcelona, Spain).
The COI-5P region was successfully amplified for 52 specimens of Mychodea spp., which resolved as 14 genetic groups where only 9 were expected (Table 1). Within-genetic-group variation was typically low, 0–0.61%, whereas among groups the closest pair was M. membranacea and M. terminalis, which were nonetheless 2.98% divergent (Table 1). Amplification success of COI-5P was unusually low (68%; 52 of 77 specimens attempted; Table S1), and, therefore, rbcL-3P was generated for 21 specimens which, when combined with full rbcL data (n = 16; Table S1), generated an alignment representing 37 specimens for 800 bp. This facilitated the assignment of 22 additional specimens to genetic species groups and resolved two additional genetic groups, Mychodea perplexa G.W.Saunders & Kraft and M. pseudoaciculare (Table S1). Where replicate data were available within species, variation was low, typically 0–1 substitutions (0.125%), with M. aciculare having 0–2 substitutions (0.25%). The most closely related species pairs for rbcL-3P were Mychodea aciculare and M. pseudoaciculare, M. carnosa and M. membranacea, and M. australis and M. sp._2WA, which were all 13–14 bp divergent (>1.6%). As such, both COI-5P and rbcL-3P showed clear within- v. between-species gaps in support of a species-rich genus Mychodea endemic to Australia.
Phylogenetic analyses of our multigene alignment resolved essentially four lineages of species (Fig. 1). Mychodea minutissima was a reasonably deep sister to all of the other species of the genus. Mychodea disticha Harv., M. perplexa, M. pusilla, M. ramulosa and the closely related M. aciculare and M. pseudoaciculare formed a second strongly supported lineage (Fig. 1). Mychodea membranacea and M. terminalis joined the type, M. carnosa, in a strongly supported sister group to M. acanthymenia, M. australis, M. hamata, M. herringtoniana, M. marginifera and the unnamed Mychodea sp. 2WA (Fig. 1). The relationships resolved in our phylogenetic analyses matched well with the various anatomical features observed for Mychodea spp. (below).
Alpha-taxonomic results and description of new species
As a result of our combined morphological–anatomical and molecular datasets, the following species of Mychodea are recognised and the distinctive features of each enumerated. First to be treated will be the previously named species that we support, followed by both the distinctive and cryptic new species that our studies have discovered.
Specimens cited as ‘Hb Kraft’ are part of the first author’s personal herbarium that are not, as yet, incorporated into a registered herbarium. Where ‘Hb Kraft’ numbers are combined with those of registered-herbarium designations (e.g. AD, HO), the Kraft numbers are those assigned in field notebooks to these same specimens.
(A) The terete or subterete species
(1) Mychodea carnosa Hook.f. & Harv., London J. Bot. 6: 408 (1847)
Type citation: ‘Georgetown, Mr. Gunn.’ Type: s. loc., s. dat., leg. ign. s.n. (lecto, fide Kraft 1978: 520: BM 001039609).
Type citation: ‘Port Philipp: Harvey.’
Type: Iconotype, here designated: F.T.Kützing, Tab. phycol. 16(2): pl. 77c, d (1866).
Representative DNA barcode: HM918237 (COI-5P; GWS016530 [UNB]).
Thalli of this, the generitype species (Schmitz 1889, p. 441), are among the largest in the genus, reaching lengths of 40 cm and occurring from 3- to 40-m depths. Distinguishing features are the coarse, fleshy textures and dark colouring of fronds, which can dry to nearly black; thick lobate holdfasts (Figs 3B, 10A); growth on a variety of substrata such as rocks, wooden jetty pilings, large coarse algae, tunicates and sponges; irregularly radial branching; large warty subterminal cystocarps (Fig. 8A); and intercalary tetrasporangia (Fig. 9A). Records range from Geraldton, Western Australia, to eastern Victoria and around Tasmania, the type locality being Georgetown in northern Tasmania. This species is credited to Tierra del Fuego by Hariot (1892, p. 1434) and Ramírez and Santelices (1991, p. 275), both repeating the undocumented sole claim of Ardissone (1888, p. 214) for Mychodea carnosa, as well as for M. compressa (= M. disticha, below). Recognition of Mychodea in Ardissone’s day and as late as the 1940s (Børgesen 1943, p. 77; see Kraft 1978, p. 571) was fraught with uncertainties and frequently mistaken. The South American claims are discounted.
Kraft (1978, p. 520) regarded M. membranacea and M. terminalis as heterotypic synonyms of M. carnosa, both of which are now reinstated following consideration of molecular and morphological data.
No type material of M. mallardiae has been seen, as no specimens appear to be extant in L, MEL or TCD. In the absence of type material, the protologue illustration in Kützing (1866, fig. 77c, d) is the only available type element. The species was regarded by both J. Agardh (1876, p. 571) and Womersley (unpubl. notes of 20 February 1952) as synonymous with Mychodea terminalis, but is here treated as a synonym of M. carnosa.
(2) Mychodea membranacea Hook.f. & Harv., London J. Bot. 6: 408 (1847)
Type citation: ‘George Town, Mr. Gunn.’ Type: Tasmania. George Town, s. dat., R.C.Gunn 1318 (lecto, fide Kraft 1978: 520: TCD 0001833; isolecto: TCD 0001831).
Type citation: ‘Ad litus peninsulae Lefebre, Decbr.’ Type: F.Mueller s.n. (lecto, fide Kraft 1978: 529: MEL1007309); Mychodea gracilaria (Sond.) Sond. Fragm. 11: 106 (1881); Mychodea gracilaria (Sond.) Kraft, Austral. J. Bot. 26: 528 (1978), isonym.
Representative DNA barcode: HM915945 (COI-5P; UNB GWS001899).
Thalli reach 25 cm in length, have generally finer and more membranous axes, and are lighter in dried colouring than is M. carnosa, which is proximally much fleshier, although similar distally in radial branching. Thalli are frequently epiphytes of Amphibolis antarctica at 3–11-m depths, rarely also occurring on Posidonia australis or stalked tunicates as well as occasionally on rocks; cystocarps are lateral (the minute apices of the procarpic laterals are soon worn away), sessile, have prominent pericarps (Fig. 8K), and bend bearing branches only slightly (Figs 8K, 10C); tetrasporangia are intercalary (Fig. 9B). The species ranges from Rottnest Island, Western Australia, to South Australia.
Although Kraft (1974, p. 528) regarded M. membranacea as a synonym of M. carnosa, type material of both taxa having been collected from the same locality, the molecular evidence from recent Tasmanian collections, as well as Hooker and Harvey’s (1847, p. 408) explicit statement that cystocarps are sessile and lateral in M. membranacea rather than terminal, show that M. membranacea should be reinstated.
As remarked by Harvey, fragmentary and non-cystocarpic material of his Tasmanian collections of M. carnosa and M. membranacea is very difficult to attribute to one or the other of the two species, as well as to M. terminalis, although when specimens have intact holdfasts, the fleshy bases of M. carnosa contrast with the thin crusts of both M. membranacea and M. terminalis.
The type and only known material of M. gracilaria, from the Adelaide region of South Australia, consists of nine sheets in MEL, on which fragmentary axes lacking holdfasts are mounted. The lectotype (Fig. 14F) and several of the isolectotypes are tetrasporophytes, the tetrasporangia being intercalary (Fig. 14G, H, see below under M. perplexa) as they are in M. carnosa, M. membranacea and M. terminalis, but the thinness and texture of the axes seem most characteristic of M. membranacea.
(3) Mychodea terminalis Harv., in J.D.Hooker, Fl. Tasman. 3(2): 323 (1859)
Type citation: ‘above Georgetown, in the Tamar, Gunn…’ Type: Georgetown, s. dat., leg. ign. s.n. (lecto, fide Kraft 1978: 593: TCD 0011980; isolecto: LD3934).
Type citation: ‘‘Port Philipp’: Dr Ferd. Müller.’ Type: Port Phillip, s. dat., F.Mueller s.n. (syn: MEL0044784A; L.4063835).
Representative DNA barcode: KU986354 (COI-5; UNB DV023).
Thalli are characterised by generally firm texture, slender axes and reddish to tan colouring; thin crustose holdfasts primarily anchored to Amphibolis or Heterozostera seagrass stems (Fig. 10E), but also occasionally to other algae; irregularly radial branching; cystocarps that swell branch tips evenly and smoothly (Figs 8B, 10E; Kützing 1859, pl. 84, fig. 2, as Gigartina longipes); and intercalary tetrasporangia. Records range from the Perth area of Western Australia to the Port Phillip region of Victoria and around Tasmania, the type locality being the Adelaide region of South Australia. As he did with M. membranacea, Kraft (1978, p. 520) synonymised M. terminalis with M. carnosa, in this case on the grounds that cystocarps in both species form so close to the apices of axes as to appear terminal at maturity. However, larger numbers of collections, and especially the molecular evidence presented above, show clearly that M. terminalis should be removed from M. carnosa and reinstated.
(4) Mychodea pusilla (Harv.) J.Agardh, Acta Univ. Lund., new series 8: 34 (1872)
Type citation: ‘…near Emu Point, King George’s Sound…’ Type: King George’s Sound, Jan. 1854, W.H. Harvey s.n. (lecto, fide Kraft 1978, p. 535, fig. 33A: TCD0015300; isolecto: BM, Alg. Aust. Exsicc. 313-B (BM 000044096-000044100 and 001039638-38).
Type citation: ‘Port Philip Heads, W. H. H. Western Port, Dr Mueller.’ Type: Port Phillip Heads, W.H. Harvey s.n. (lecto, fide Kraft 1978, fig. 33D: TCD 0011974 (Trav. Set No. 80), Alg. Aust. Exsicc. 343-E).
Representative DNA barcode: KF026472 (COI-5P; UNB GWS024723).
Thalli are small (1–4 cm in length) and occur singly or more usually in caespitose clumps from low-intertidal to 12-m depths on the woody stems of its obligate host, Amphibolis antarctica. Colours range from reddish-brown to very dark, nearly black; branching is primarily subdichotomous (Figs 8F, 10F), but is frequently also irregularly proliferous (Fig. 10G) in forms that were given the separate name Mychodea fastigiata (Harvey) J.Agardh (Agardh 1872, p. 34; originally as Hypnea fastigiata by Harvey 1863, synop number 457), a species that was synonymised by Kraft (1978, p. 533) and confirmed by the second author’s molecular sequences. Cystocarps are intercalary, often embedded in axes at considerable distances from apices (Fig. 8F). Tetrasporangia are terminal. The range is from Geraldton, Western Australia, to Westernport, Victoria, and around Tasmania. It is often one of the most common and abundant Amphibolis epiphytes and a prominent drift component.
(5) Mychodea aciculare (J.Agardh) Kraft, Austral. J. Bot. 26: 555 (1978)
Type citation: ‘Hab. ad occidentales Nova Hollandiae oras’.
Type: (lecto, fide Kraft 1978, p. 557, LD 33993).
Type citation: ‘Ad litus peninsulae Lefebre, Decemb.’ Type: Ad litus mari inundatum penins. Levevre, Dec. 16 1847, F. Mueller s.n. (lecto, here designated: MEL0044483; isolecto: MEL5163780–82).
Type citation: ‘Ad oram Novae Hollandiae meridionalem, in sinu Port Adelaide.’ Type: ‘Port Adelaide, N.H. Dec. 1847. Gavin’ (lecto, here designated, S-A1664; isolecto: S-A1663).
Representative DNA barcode: HM917867 (COI-5P; UNB GWS015590 [UNB]).
Owing to Harvey’s early confusion and his illustration (Harvey 1860, pl. 142, fig. 1 only) of this species as Mychodea carnosa, M. aciculare was regarded as a heterotypic synonym of M. carnosa until resurrected by Kraft (1978, p. 555) on the basis of Areschoug’s type material in S. Thalli are wiry in texture (Fig. 3E), 10–40 cm in length, usually dark reddish-brown to nearly black, have needle-pointed apices for which the species was named (Fig. 8I), and irregularly radially forked and branched, the laterals occasionally ending in hooks (Fig. 3G). Hosts include rocks, jetty pilings, Amphibolis antarctica and various brown and red algae from 2- to 24-m depth. Thalli have a seemingly obligate relationship with the hydrozoan Plumularia flexuosa Bâle (Watson 1973, p. 188), as plantlets less than 10 mm in length are already fully colonised; hydrozoan stolons run from base to near the apices within the inner cortex of the alga and issue numerous polyp-bearing laterals that erupt through the algal surfaces (Kraft 1978, fig. 18A); as the thalli with their hydrozoan loads lengthen, the basal tissue degenerates, leaving the plants entirely anchored by a dense weft of hydrozoan stolons (Fig. 3D, E). Cystocarps are basal in lateral arcuate spines (Fig. 8I), and tetrasporangia are terminal. The range is from Cape Riche, Western Australia, to Westernport, Victoria, and around Tasmania.
(6) Mychodea spinulifera J.Agardh, Acta Univ. Lund., new series 33: 51 (1897)
Type citation: ‘Hab. Ad oras australes Novae Hollandiae; sub. no 156 ex Fowlers-bay a Haloran mihi missa.’ Type: holo, fide Kraft 1978: 560: LD 34058, illustrated by Kylin 1932, pl. 27, fig. 67).
This, along with M. herringtoniana and M. minutissima, is the most geographically restricted of all the Mychodea species, recorded as an epiphyte of drift and upper-sublittoral Amphibolis antarctica only from Fowlers Bay, South Australia, and Eucla, further west at the Head of the Great Australian Bight in Western Australia. Thalli are subterete to slightly compressed, reach 8 cm in length, are primarily dichotomously or subdichotomously branched, and are beset on all orders with myriad short simple or forked horizontally oriented spines (Fig. 10I). Axial structure is lacunate (Kraft 1978, fig. 20C), cystocarps occur basally and epibasally in the spines (Kraft 1978, fig. 20A), and tetrasporangia are terminal. Owing to the remoteness of its collection localities, this is one of the least frequently encountered species of the genus. Recent drift collections (1998) by the first author from Fowlers Bay failed to successfully sequence.
(7) Mychodea ramulosa J.Agardh, Acta Univ. Lund., new series 33: 50 (1897)
Type citation: ‘Specimina hujus tum ex Nova Hollandia occidental, tum a meridionali pauciora habui.’ Type: n.v. (lecto, fide Kraft 1978, p. 538, LD 33982).
Representative DNA barcode: KY250785 (COI-5P; GWS029573 [UNB]).
Thalli reach 20 cm in length, are coarse and fleshy in texture when living, and attached to rocks, jetty pilings, tunicates, Amphibolis antarctica, and various brown and red macroalgae from tidepools down to 12-m depths. Main axes are compressed and have a pleated, rather than smooth, surface (Fig. 8L), with the medullary cells surrounding central filaments being inflated as is typical of terete species. Branching is mostly marginal, although short laterals also form occasionally on the broad surfaces (Kraft 1978, fig. 10B). Most distinctive are the stoloniferous tangles of basal axes that form frequent hapteroid attachments to substrata (Fig. 3C). Cystocarps are sessile and cause the bearing axes to usually bend sharply at the sites (Figs 8L, 11A). Tetrasporangia are terminal. The range is from King George Sound, Western Australia, to Portland, Victoria.
(8) Mychodea disticha Harv., in J.D.Hooker, Fl. Tasman. 3(2): 323, pl. 129A (1859)
Type citation: ‘East coast, Gunn.’ Type: East Coast VDL, R Gunn s.n. (lecto, fide Kraft 1978, p. 542, fig. 36A; TCD 0011963).
Type citation: ‘Phillip Island, Western Port, W. H. H.’ Type: Phillip Id, Western Port, Victoria, W.H.Harvey s.n. (Alg. Aust. Exsicc. 414) (lecto, here designated: TCD 0011966; isolecto: MEL44486).
Type citation: ‘‘Philipp Island.’ Nova Hollandia: Dr Ferd. Mueller. 1865.’ Type: ‘Phillip Island, Ferd. Müller 1865’ (holo: L.4063986).
Type citation: ‘‘S. Australia’: Dr Curdie.’ Type (lecto, here designated: TCD0018169; isolecto: MEL44484; isolecto: LD34069 (Kylin 1932, pl. 27).
Representative DNA barcode: KY250791 (COI-5P; UNB GWS029583).
Thalli are robust, compressed, firm-textured and smooth-surfaced (Fig. 8D), reach 20–45 cm in length, and attach to rocks, calcified worm tubes and wooden jetty pilings, Amphibolis antarctica and Acrocarpia paniculata from the upper sublittoral down to 11-m depths. Apical cells are inconspicuous, and cross-sections are non-lacunate, consisting of 2–4 layers of uniformly sized outer medullary cells that abruptly transition to a shallow cortex (Fig. 2I). Laterals are simple or variously forked (Fig. 8D) and of varying lengths (Fig. 11B). Cystocarps are intercalary and subapical in the mostly distichous laterals (Fig. 8D), and lack differentiated pericarps (Fig. 7A). Tetrasporangia are terminal (Fig. 9C). The range is from Geraldton, Western Australia, to Westernport, Victoria, and around Tasmania.
This is the second Mychodea species credited to Pacific South America (Ardissone 1888; Ramírez and Santelices 1991, as M. compressa). As with M. carnosa, this record was based solely on the very doubtful claim of Ardissone (1888), which remains unconfirmed.
Kylin (1932, pl. 27, fig. 66) illustrated what he termed the ‘Originalexemplar’, but this is not the TCD lectotype as selected by Kraft in March of 1973. It is not known if Kylin’s illustration was taken from the original material available to J. Agardh.
(9) Mychodea hamata Harv., in J.D.Hooker, Fl. Tasman. 3(2): 323 (1859)
Type citation: ‘Georgetown and Port Arthur.’ Type: Port Arthur, VDL, W.H.Harvey s.n. (Alg. Aust. Exsicc. 415-K) (lecto, fide Kraft 1978, p. 547, TCD 0011969). Acanthococcus ewingii Harv. Phycol. Austral. 3: pl. 141 (1860), nom. illeg., nom. superfl.
Representative DNA barcode: KY250798 (COI-5P; UNB GWS029570).
Thalli reach 20 cm in length and grow on wooden pilings, Amphibolis antarctica, and a variety of macroalgae such as Gelidium asperum (C.Agardh) Grev. and Melanema dumosa (Harv.) Min-Thein & Womersley, from lower-eulittoral down to 17-m depths. Primary axes are smooth-surfaced (Fig. 8H), cartilaginous and subdichotomously branched, with numerous short distichous laterals (Fig. 11C), these often ending in tendril hooks (Figs 3F, 8H) that can effect secondary anchorage and vegetative propagation. Cystocarps form at the bases or midway along the lengths of short acute laterals (Fig. 8H), and tetrasporangia are terminal. The range is from Kangaroo Island and Encounter Bay, South Australia, to Westernport, Victoria, and around Tasmania. Along with M. pusilla, this is often the commonest Mychodea encountered in south-eastern Australia, especially in drift.
(B) The flattened species
(10) Mychodea marginifera (Aresch.) Kraft, Austral. J. Bot. 26: 551 (1978)
Type citation: ‘Ad oram Novae Hollandiae meridionalem in sinu Port Phillip.’ Type: ‘Port Phillip F. Mueller’ (lecto, fide Kraft 1978: 55: S-A28591; isolecto: S-A28593).
Type citation: ‘Port Phillip Heads, and Western Port, abundant, W. H. H.’ Type: Port Phillip Heads, W.H. Harvey 396-E (lecto, here designated: TCD 0011961).
Representative DNA barcode: KY250782 (COI-5P; UNB GWS017308).
Thalli range from 5 to 25 cm in length and grow from discoid holdfasts on rocks, jetty pilings, the tunicate Pyura pachydermata Herdman, the seagrass Amphibolis antarctica, and several macroalgae such as Acrocarpia paniculata and Gelidium asperum from lower-eulittoral down to 11-m depths. This is one of the commonest, most variable in habit, and occupies the widest variety of substrata of all the Mychodea species. Blades are typically flabellate with numerous marginal proliferations and broadly rounded apices in which series of submarginal cystocarps can occur (Figs 8C, 11E), but are also at times narrow, little proliferous, and have tips that accommodate only single cystocarps (Fig. 11F). Blade sections (Kraft 1978, fig. 16B, C) lack inflated medullary cells (i.e. are not lacunate), and uniaxial construction is only visible in cultured germlings (Kraft 1978, fig. 16G, H). Cystocarps lack differentiated pericarps and tetrasporangia are terminal. The range is from Geraldton, Western Australia, to Westernport, Victoria, and around Tasmania.
(11) Mychodea acanthymenia Kraft, Austral. J. Bot. 26: 563 (1978)
Type citation: ‘Cape Lannes, Robe, S.A. (Kraft, 10.ii.1973).’ Type: Cape Lannes, near Robe, S. Aust., 10 Mar. 1973, G. Kraft 4403 (holo: AD A44728). iso: BISH1000164).
Representative DNA barcode: HM915892 (COI-5P; UNB GWS000960).
Thalli reach 10 cm in length and occur mostly on rock (once on Coronaphycus elatus (C. Agardh) Metti) from just above extreme low-tide levels down to 18-m depths. Distinguishing features are flat spinous fronds, three orders of distichous spinous laterals on all surfaces (Fig. 11G), and cross-sections in which the central filamentous core is narrow within its surrounding layers of subisodiametric medullary cells in which traversing filaments are infrequent or absent (Fig. 2J). Cystocarps are positioned mostly midway along laterals and usually bear spines on mature pericarps (Fig. 8J); tetrasporangia are terminal. The range is from Kangaroo Island to Jervis Bay, New South Wales, and around Tasmania. This is the only Mychodea species reliably recorded from the eastern coast north of the state of Victoria (Millar and Kraft 1993, p. 22).
(12) Mychodea australis (Zanardini) Kraft, Austral. J. Bot. 26: 566 (1978)
Type citation: ‘Hab. Port Philip – (Ferd. von Mueller.).’ Type: n.v. (holo: Herb. G. Zanardini, Museo Civico di Storia Naturale, Venice, Italy (MCVE 786)).
Type citation: ‘Ad Tasmaniam R. Gunn.’ Type: n.v. (holotype: LD 34115 (apparently the basis for J. Agardh’s (1879) drawings (pl. 30, figs 1–8), which are mounted on sheets numbered LD34104 and LD34105 – Patrik Frödén, pers. comm.)).
Type citation: Hab. ad oras australes Novae Hollandiae (F. de Mueller!).’ Type: Port Phillip, s. dat., leg. ign. s.n. (lecto, here designated: MEL1007262).
Representative DNA barcode: HM918300 (COI-5P; UNB GWS016664).
Thalli are flattened, subdichotomously branched to 1–3 orders of variable widths that are beset with numerous short distichous laterals with rounded apices (Fig. 11H; Kraft 1978, fig. 42A–D). Anchorage is by a crustose holdfast to rock, jetty piles, tunicates, Amphibolis antarctica, and macroalgae such as Cystophora spp. and stolons of Caulerpa simpliciuscula (R.Br. ex Turner) C.Agardh from extreme low-tide levels down to 40–60-m depths. Frond surfaces are without proliferations, and cross- and long-sections are unique in the genus in having enlarged central-axial cells that usually have dark and dense cytoplasmic contents (Fig. 2G) that show through the surface as distinct lines in thalli that are bleached (Fig. 8E; Kraft 1978, fig. 23A, C). Cross-sections, like those of M. acanthymenia, have a narrow central-medullary core in which the connector cells and filaments are mostly confined to the poles of the central-axial cells (Fig. 2H) and few, if any, narrow filaments run from the central cable to connect to cells of the inner cortex (Fig. 2G; Kraft 1978, fig. 23F). Cystocarps are subterminal (Figs 8E, 11H), and tetrasporangia are produced in slightly raised sori separated from blade margins by a narrow sterile band. Sterile outer-cortical cells of the aggregates elongate and separate the tetrasporangia (Fig. 9D).
J. Agardh (1876) described Ectoclonium dentatum, from a mainland ‘fragmentum’ and a whole Tasmanian specimen, the latter illustrated (Agardh 1879) as fig. 1 of his plate 30. Patrik Frödén, Assistant Curator, Botanical Museum (LD), has suggested that Agardh depicts a habit somewhat different from the holotype specimen (LD34105) because he may have drawn a composite of loose correspondences to several branches of that specimen. Both the holotype and the basis for Agardh’s habit and anatomical (pl. 30, figs 2–8) illustrations are clear examples of robust Mychodea australis.
Species to be added to Mychodea as a result of post-1978 collections and molecular sequencing
(13) Mychodea minutissima G.W.Saunders & Kraft, sp. nov.
Type: Western Australia, Muttonbird Island (35°02′51″S, 117°40′52″E), west of Albany, –3 m on a rock wall, 6 Nov. 2010, G.W. Saunders & K. Dixon GWS024289. (holo: UNB).
DNA barcode of holotype: KY250796 (COI-5P; GWS024289 [UNB]).
Thalli of the holotype and paratype collections came from 3–9-m depths on rock or coral, forming dense clumps (Fig. 12A) or scattered individuals (Fig. 12B, C) of terete (Fig. 12A, C) or slightly compressed (Fig. 12B), simple or sparingly branched (Fig. 12C) axes less than 15 mm in length, anchored by numerous haptera where procumbent or down-growing lower axes (Fig. 12B, C) contact substrata, and also where erect axes come in contact with one another and adhere by basally constricted attachment pads (Fig. 12E). Axes range from 180 to 500 μm in diameter, their apices tapering to the single apical cells abruptly from thicker axes, especially those destined to become procarpic (Fig. 12D). Axes are distinctive in having central-axial filaments not jacketed by connector-cell derivatives and in lacking an abrupt size disjunction between outer medullary and inner cortical cells (Fig. 12F). Cystocarps occur singly in subspherical–elliptical to bilanceolate terminal pods (Figs 8M, 12C); slightly sunken spermatangial clusters are scattered in the outer cortex (Fig. 12G), and carpogonial branches are three-celled (Fig. 12H), with several borne on supporting cells throughout the medulla; mature cystocarps (Fig. 12I) lack any development of a Faserhülle and are surrounded by a pericarp in which large numbers of trichogynes persist (Fig. 12I). Tetrasporangia are unknown. Known only from the type locality and Port Denison, Western Australia.
This is by far the most unlikely species of Mychodea, for no other is so diminutive, inconspicuous, and likely to be overlooked because of its resemblance to any number of similar-appearing subtidal turf species that seldom attract attention or collection. The fact that the two collections are known, one gametangial and the other found and sequenced many years later by the second author, has been very fortuitous. Given its reasonably deep sister relationship to the other members of Mychodea and its anomalous habits, medullary structure and cystocarp position, attribution of this species to a separate genus was considered. However, given the spermatangial and carposporophyte similarities to typical Mychodeas, it was decided to include it in that genus, at least for the present.
Named for the small size and inconspicuous wiry habits of the thalli.
Specimens examined (paratype)
WESTERN AUSTRALIA: Port Denison (29°16′05.5″S, 114°54′48.5″E), 6–9 m, on coral, G. Kraft & R. Ricker Hb Kraft-16171, 11 Dec. 1989.
(14) Mychodea herringtoniana Kraft & G.W.Saunders, sp. nov.
Type: Victoria, Warrnambool (38°24′2.3″S, 142°28′33.8″E), drift on Acrocarpia paniculata behind the town breakwater, 13 Jan. 1995, R. Kraft [née Herrington] Hb Kraft-10448 (holo: HO 586764).
Representative DNA barcode: HM915998 (COI-5P; UNB GWS002136).
Thalli reach 10–40 cm in length and are epiphytes of drift Amphibolis antarctica, Acrocarpia paniculata (most frequently) and (once only) Perithalia caudata (Labill.) Womersley at Port MacDonnell, South Australia, and Warrnambool, Victoria, where they are occasionally found year-round. Blades are flat throughout, distally 3–13 mm wide and 750–1500 μm thick, linear, the fronds attached singly (Fig. 13A) or in clusters by small crustose holdfasts and are simple (Fig. 8G) or sparingly forked (Fig. 13A), or bearing basally constricted laterals from the margins (Figs 11D, 13B) or at sites of wounding (Fig. 13A). Primary distinguishing features are the aligned and closely spaced submarginal cystocarps (Figs 8G, 11D, 13A), often lengthy and lanceolate laterals on tetrasporophytes (Fig. 13B), and a narrow medulla of centrally aggregated filaments surrounded by one or two layers of inflated medullary cells between the filamentous centre and a deep inner cortex (Fig. 13C). Central-axial filaments are inconspicuous, the abutting cells linked by loose baskets of connector filaments (Fig. 13D). Terminal tetrasporangia 35–40 μm in length are scattered in the cortex and jacketed by two layers of somewhat elongated cortical cells (Fig. 13E), and spermatangia form in deep ampullar clusters within the outer cortex (Fig. 13F). Carpogonial branches are three-celled but often only one (Fig. 13G) or two persist on supporting cells. Multiple gonimoblast filaments radiate laterally and inwardly from the diploidised auxiliary cell (Fig. 13H), and at early stages a Faserhülle is prominent (Fig. 13I, J), unlike in other Mychodea species, although it does not always remain a distinct feature in mature cystocarps (Fig. 13K).
Early collections of this species were either sterile or tetrasporic, and although unusual in habit were filed with Mychodea marginifera because of that species’ notorious variability in blade morphology. With the discovery by Rebecca Herrington of the strikingly configured cystocarps and from the results of molecular analyses by the second author, the species status of this long-overlooked entity has been firmly established.
This species is named in honour of Rebecca Kraft (née Herrington), who made the first collections of cystocarpic thalli, recognised their difference from any Mychodea seen before, and who has been an essential partner in the endeavours of both authors in this and many other projects over the years.
Specimens examined (paratypes)
(15) Mychodea perplexa G.W.Saunders & Kraft, sp. nov.
Type: South Australia, Sceale Bay boat ramp (33°01′01.2″S, 134°11′24″E), 25 km south of Streaky Bay, drift on Amphibolis antarctica, 13 Jan. 2005, G. & R. Kraft & K.R. Dixon, Hb Kraft-11493 (holo: HO 586766; iso: UNB (GWS DV-012).
Representative DNA barcode: KY250755 (rbcL; UNB GWS029872).
Description: thalli are 12–30 cm in length, the fronds arising from a somewhat swollen basal pad (Fig. 14A) and consisting of fleshy percurrent axes 3–7 mm in diameter that radially bear three orders of progressively finer laterals (Fig. 14A, B). Primary branches are 3–16 mm in length. Cystocarps are sessile on both primary and higher-order laterals (Fig. 14A, C) and cause little bending of the bearing axes (Fig. 14C). Cross-sections are lacunate (Fig. 14D), with a single layer of greatly inflated cells surrounding the central core in which the central-axial filament becomes indistinguishable from cross-connector cells and their derivatives (Fig. 14D). Cystocarps lack Faserhüllen, the carposporangia within released by the spreading (Fig. 14E) and ultimate breakdown of the pericarp. Spermatangia occur in shallow clusters within the outer cortex in tissue adjacent to cystocarps.
This species, as pointed out above, combines the fleshy, terete and radially branched axes characteristic of M. carnosa with the sessile cystocarps of M. membranacea. However, its phylogenetic position, as indicated by the molecular data, is clearly among species with terminal, rather than intercalary, tetrasporangia and with non-sessile cystocarps (Fig. 1), leading to the prediction that its tetrasporangia, when finally found, can be expected to be terminal. For this reason, it was important to look at Mychodea gracilaria, the type locality being the Lefevre Peninsula of the Adelaide region, and which Kraft (1978, pp. 528–533) recognised as the species of choice for plants with sessile cystocarps and intercalary tetrasporangia when he erroneously relegated M. membranacea to synonymy with M. carnosa. It was thus necessary to re-examine the type collection of Sonder’s Acanthococcus gracilaria in MEL, which consists nine sheets of fragments, some tetrasporic (Fig. 14F) but none cystocarpic, to see whether the sporangia were terminal and could, therefore, be a name that the present species could at least provisionally bear. Spindly material, such as the isolectotypes, that has been dried for almost 170 years can be difficult to slice without the sections falling apart on rehydration, as was true in this case, but despite the breakup of slide-mounted tissue and a general absence of persistent primary and secondary pit-connections between the cortical cells (Fig. 14G), the positions of the tetrasporangial primordia in the subsurface layers, as well as the presence of closely associated, probably pit-connected cells distal to the tetrasporocytes, are strong indications that the sporangia are intercalary. The species is therefore provisionally placed in synonymy with Mychodea membranacea (above) and the anomalous Sceale Bay and Cape Jaffa collections designated as a new species. Known for certain only from the two South Australian localities where the two thalli of similar appearance and with the same defining DNA profiles were collected.
Named for the puzzlement this entity visited on both authors because of its lacunate cross-sections and sessile cystocarps, similar to those seen in M. membranacea, and its coarse fleshy axes, such as those of M. carnosa. However, by the second author’s DNA analyses, it falls into a clade in which all the other members have terminal, rather than the intercalary tetrasporangia of M. membranacea and M. carnosa. The puzzlement persists so long as tetrasporangia in molecularly confirmed specimens conforming to M. perplexa remain lacking.
Specimens examined (paratypes)
(16) Mychodea pseudoaciculare G.W.Saunders & Kraft, sp. nov.
Type: Tasmania, Windmill Point, Georgetown (41°06′34.8″S, 146°49′0.7″E), 23 m on invertebrate, 19 Jan. 2010, G.W. Saunders & K. Dixon GWS015052 (holo: UNB).
DNA barcode: KY250760 (rbcL: GWS015052 [UNB]).
This taxon is known only from the cystocarpic holotype specimen, which in all respects save sequence data is indistinguishable from M. aciculare. As in M. aciculare, the thallus has a wiry consistency (Fig. 15A), needle-like apices on dichotomous and lateral axes (Fig. 15B), and is infested by the largely endophytic hydrozoan Plumularia flexuos whose external polyps (Fig. 15C, D) arise from stolons that course upwardly through the outer medulla of the axes (Fig. 15D). Inner medullary cells are inflated (lacunate), and central-axial cells are surrounded by conjunctor cells and derivatives (Fig. 15D). Cystocarps form at the bases of short acicular laterals (Fig. 15E), and carposporophytes are similarly constructed of islands of carposporangia intermingled with a network of mixed carposporophytic and gametophytic cells (Fig. 15F). Spermatangial features are like those of M. aciculare (Kraft 1978, fig. 19J), and tetrasporangia are unknown.
This is the one truly ‘cryptic’ species discovered to date in Mychodea that has entirely sequence-based defining characters.
Named for the seemingly identical morphological features of the single collection to the widespread Mychodea aciculare.
(17) Mychodea sp._2WA
Representative DNA barcode: KY250797 (COI-5P; GWS024481 [UNB]).
The sole thallus of this entity is leafy, its subdichotomous blades 1.5–3.5 mm in width and bearing bluntly rounded extended proliferations and acute-tipped marginal spines (Fig. 15G). It resolves on molecular analyses as a distinct species most closely related to Mychodea australis, although differing greatly in habit and lacking the prominent central-axial filaments of that species. As expected, the tetrasporangia, which occur in slightly raised sori, are terminal. In view of the fact that a single non-gametangial thallus is all that is known, however, we withhold giving it a formal name in view of our ignorance of its range of thallus morphologies and critical carposporophyte features.
Members of Mychodea regarded as distinct species but for which molecular data are unavailable
(18) Mychodea sp._3WA
The two collections are of three individual specimens that grew as subterete, subdichotomous axes mostly lacking adventitious laterals and anchored by a swollen crustose holdfast (Fig. 15H). Cross-sections are lacunate, and tetrasporangia are intercalary. Prediction would group this entity with the carnosa–membranacea–terminalis complex of species characterised by intercalary tetrasporangia, but the habit is very unlike any of those species. As no further records are known and no molecular data have been acquired, we signal its distinctiveness without proposing a formal name.
(19) Mychodea echinocarpa Kraft & G.W.Saunders, sp. nov.
Type: Western Australia: Geraldton (28°48′02.9″S, 114°36′59.5″E), drift at end of Jarrah Street, 8 Oct. 1990, G. & R. Kraft, Hb Kraft-9998, (cystocarpic; holo: HO 586765).
The two thalli of this distinctive species are 6 cm (the gametophyte, Fig. 16B) and 8 cm (the tetrasporophyte, Fig. 16A) in length, both growing from small discoid holdfasts on Amphibolis antarctica and consisting of subdichotomous or pseudomonopodial compressed main axes covered by myriad simple and multifid spines on both the flattened and the marginal surfaces (Fig. 16C). The cystocarps form subapically within short laterals and bear numerous simple or forked spines on the pericarp (Fig. 16D). Spines and axes grow from rounded, rather than needle-like, tips in which the apical cells are prominent (Fig. 16E). Cross-sections are lacunate, with abrupt transitions between inner and outer medullary layers and the inner cortex, with the central-axial filament being either somewhat larger than surrounding filaments (Fig. 16F) or indistinguishable (Fig. 16G) from them. Tetrasporangia are terminal (Fig. 16H), and spermatangial ampullae are broad-based and scattered in the outer cortex (Fig. 16I). Carpogonial branches are three-celled, often with only one (Fig. 16J) or two being evident on supporting cells, and cystocarps have particularly thick pericarps that spread from the inside to effect carpospore release (Fig. 16K).
This distinctive species is represented by just two drift thalli collected 7 years and at sites a startling 3000+ km apart. Although material was unfortunately not prepared for DNA extraction, the thalli are so unambiguously mychodeoid in structure and represented by both a monoecious gametophyte and a tetrasporophyte that we feel justified in formally proposing them as a new species. Closest in habit is the geographically restricted Mychodea spinulifera (Fig. 10I; Kraft 1978, fig. 20A, 41A), in which the axes are narrower, terete or subterete, and the spines simpler, although both seem to be obligate epiphytes of the same seagrass host.
Named for the spinous cystocarps.
Specimen examined (paratype)
Kraft (1974) speculated some 43 years ago about how the various species of Mychodea might be related to one another. When his conclusions were compared with the implications of the molecular data reported in the present paper (Fig. 1), the exercise was instructive in showing how the virtual guesswork of the pre-molecular past can both conform to and wildly differ from the far more accurate pictures painted by today’s phylogenetic methodologies. In sum, the hypotheses framed (Kraft 1978, pp. 573, 574) on the basis of morphology and anatomy were mainly the following:
Fleshy, terete and radially branched species such as M. carnosa and M. membranacea (as M. gracilaria) were probably ancestral to progressively compressed, distichously branched forms such as M. ramulosa, M. disticha and M. hamata, ultimately leading to truly flattened species such as M. marginifera, M. acanthymenia and M. australis. What the molecular analyses (Fig. 1) showed (if we ignore the wiry, terete M. minutissima, which was not known until the present study) is a picture mostly very different from that which Kraft hypothesised, as terete and compressed, distichous, dichotomous and radial species occur in our Lineage 2 (Fig. 1: M. aciculare, M. disticha, M. pusilla, M. perplexa, M. pseudoaciculare and M. ramulosa). Lineage 3 (Fig. 1: M. carnosa, M. membranacea and M. terminalis) is composed wholly of terete or subterete species (this group was considered to consist of just M. carnosa by Kraft 1978), but it is in no way sister to all of the other clusters. Lineage 4 comprises one compressed (M. hamata) and five flattened species, so the latter do group together (Fig. 1).
Mychodea carnosa, being terete, radially branched, possessing a central-axial filament surrounded by many cross-connector filaments, and scattered intercalary tetrasporangia, and M. australis, being flattened, strictly distichous, displaying a prominent central-axial filament throughout with few cross-connectors limited to the poles of central-axial cells, and nemathecial terminal tetrasporangia, were hypothesised to be the most primitive and the most recently evolved species respectively. Evidence for M. carnosa as sister to all of the other included species is lacking, although M. australis is indicated as a recently evolved species in the third cluster (Fig. 1).
Species with subterminal cystocarps should group together, although in M. marginifera that might not hold because its cystocarps tend to form in series lining spatulate leading margins as opposed to developing singly near narrow branch apices. The lineage-3 carnosa–membranacea–terminalis group does hang together, as do (not so closely) marginifera and the newly added herringtoniana within Lineage 4, but expectations that species with cystocarps at varying distances from the base along laterals would cluster together are not fulfilled (Fig. 1).
Species with lacunate cross-sections should group together, which they do in the case of the carnosa–membranacea–terminalis clade, but others with such a medullary structure such as aciculare or pseudoaciculare, ramulosa, pusilla and the new perplexa are all on Lineage 2, rather than Lineage 3 (Fig. 1).
Species with intercalary, rather than terminal, tetrasporangia should group together. This is true of the species in Lineage 2, but without knowing how sporangia are positioned in M. perplexa we cannot be certain that this feature holds within the genus as a whole. Further, not knowing the derivation of tetrasporangia in M. minutissima makes it impossible to speculate on whether or not intercalary sporangia are a later development or a more recent reversion to an ancestral state in the genus. However, two firmly establishable facts are that M. australis belongs unquestionably in Mychodea rather than the segregate genus Neurophyllis (Fig. 1), and that intercalary v. terminal tetrasporangia are not grounds for the recognition of two genera, as Kraft (1974) initially opted for but fortunately abandoned in his published monograph (Kraft 1978).
Because both M. disticha and M. marginifera have apical cells and central-axial filaments distinguishable only in germling stages, before the axes come to appear multiaxial, it was implied that they might be closely related. Molecular data strongly refuted this hypothesis (Fig. 1).
The present study has re-emphasised one further point, made previously by Saunders et. al. (2004), viz. that the bizarre Western Australian-endemic genus and species Mychodeophyllum papillitectum Kraft and its monotypic family Mychodeophyllaceae are definitely taxa independent of the Mychodeaceae (Fig. 1) despite the similarities in vegetative structure (Kraft 1978, fig. 25C), polycarpogonial procarps (Kraft 1978, fig. 26A, B), the diploidisation process (Kraft 1978, figs 26D, 45B) and early gonimoblast development (Kraft 1978, figs 27A, 45D).
Conflicts of interest
The authors declare that they have no conflicts of interest.
All collectors listed in Table S1 are thanked for their contributions to this project. Tanya Moore generated the molecular data for this study, with assistance from past and present Saunders’ Laboratory members. The research was supported through funding to GWS from the Natural Sciences and Engineering Research Council of Canada, the Canada Foundation for Innovation and the New Brunswick Innovation Foundation. G. T. Kraft thanks the Australian Biological Resources Study, the School of Biological Sciences of Melbourne University, and the Tasmanian Herbarium’s (HO) Director Dr Gintaaras Kantvilas and Chief Administrative Officer Ms Kim Hill for material and research support. We particularly appreciate the helpful comments of two anonymous referees and the considerable assistance of Dr Peter G. Wilson, Associate Editor of Australian Systematic Botany, Brendan Lepschi, Curator of the Australian National Herbarium, Canberra, and Mr Andrew Bullen, Production Editor at CSIRO Publishing, Melbourne. We are grateful for the assistance of Dr Patrik Frȍdén (Assistant Curator, Botanical Museum LD), Dr Johannes Lundberg (Curator, Swedish Museum of National History, S), Dr Jo Wilbraham (Senior Curator, Algae, Natural History Museum. BM), Dr Pina Milne (Collections Manager, National Herbarium of Victoria, MEL) and Prof. John Parnell (Department of Botany, School of Natural Sciences, Trinity College Dublin, TCD) for providing accession numbers and details of Mychodea type materials in their respective herbaria. As always, we acknowledge the indispensable work of Prof. Michael and Gwendoline Guiry (National University of Ireland, Galway), whose creation and maintenance of AlgaeBase (www.algaebase.org/) makes projects such as ours so much easier, more accurately informed, and pleasurable to pursue.
ReferencesAgardh JG (1872) Bidrag till florideernes systematik. Lunds Universitets Årsskrift, Afdeling for Mathematik och Naturvetenskap 8, 60
Agardh JG (1876) ‘Species, Genera et Ordines Algarum. Vol. 3, Pt. 1. Epicrisis Systematis Floridearum.’ (Weigel: Leipzig, Germany)
Agardh JG (1879) Florideernes morphologie. Konglia Sveanska Veatenskaps-Akademiens Handligar – Ser. 4 15, 1–199.
Ardissone F (1888) Le alghe della Terra del Fuego racolte dal prof. Spegazzini. Rendiconti del Reale Istituto Lombardo di Scienze e Lettere – Ser. 2 21, 208–215.
Børgesen F (1943) Some marine algae from Mauritius. III. Rhodophyceae, pt. 2: Gelidiales, Cryptonemioales, Gigartinales. Det Kongliga Danske Videnskabernes Selskab – Biologiske Meddelelser 19, 1–85.
Hariot P (1892) Complément a la flore algologique de la Terre de Feu. Notarisia 7, 1427–1435.
Harvey WH (1860) ‘Phycologia Australica. Vol. 3.’ (Reeves: London, UK)
Harvey WH (1862) ‘Phycologia Australica. Vol. 4.’ (Reeves: London, UK)
Harvey WH (1863) ‘Phycologia Australica. Vol. 5.’ (Reeves: London, UK)
Hooker JD, Harvey WH (1847) Algae tasmanicae: being a catalogue of the species of algae collected on the shores of Tasmania by Ronald Gunn, Esq., Dr Heannerett, Mrs Smith, Dr Lyall and Dr J.D. Hooker; with characters of the new species. London Journal of Botany 6, 397–417.
Kraft GT (1974) Monograph of the Mychodeaceae, Dicranemaceae and Acrotylaceae (Gigartinales, Rhodophyta). PhD Thesis, University of Adelaide, SA, Australia.
Kraft GT (1978) Studies of marine algae in the lesser-known families of the Gigartinales (Rhodophyta). III. The Mychodeaceae and Mychodeophyllaceae. Australian Journal of Botany 26, 515–610.
| Studies of marine algae in the lesser-known families of the Gigartinales (Rhodophyta). III. The Mychodeaceae and Mychodeophyllaceae.CrossRef |
Kraft GT, Saunders GW (2014) Crebradomus and Dissimularia, new genera in the family Chondrymeniaceae (Gigartinales, Rhodophyta) from the central, southern and western Pacific. Phycologia 53, 146–166.
| Crebradomus and Dissimularia, new genera in the family Chondrymeniaceae (Gigartinales, Rhodophyta) from the central, southern and western Pacific.CrossRef |
Kraft GT, Womersley HBS (1994) Mychodeaceae. In ‘The Marine Benthic Flora of Southern Australia, Part IIIA’. (Ed. HBS Womersley) pp. 450–470. (Australian Biological Resources Study: Canberra, ACT, Australia)
Kützing FT (1859) ‘Tabulae phycologicae. Vol. 9.’ (F. Eberhardt: Nordhausen)
Kützing FT (1866) ‘Tabulae phycologicae. Vol. 16.’ (F. Eberhardt: Nordhausen)
Kylin H (1932) Die Florideenordnung Gigartinales. Acta Universitatis Lundensis 28, 1–88.
Kylin H (1956) ‘Die Gattungen der Rhodophyceen.’ (Gleerups: Lund, Sweden)
Millar AJK, Kraft GT (1993) Catalogue of marine and freshwater red algae (Rhodophyta) of New South Wales, including Lord Howe Island, South-western Pacific. Australian Systematic Botany 6, 1–90.
| Catalogue of marine and freshwater red algae (Rhodophyta) of New South Wales, including Lord Howe Island, South-western Pacific.CrossRef |
Ramírez ME, Santelices B (1991) Catálogo de las algas marinas bentónicas de la costa temperada del Pacífico de Sudamérica. Monografías Biológicas 5, 1–437.
Saunders GW, McDevit DC (2012) Methods for DNA barcoding photosynthetic protists emphasizing the macroalgae and diatoms. Methods in Molecular Biology 858, 207–222.
| Methods for DNA barcoding photosynthetic protists emphasizing the macroalgae and diatoms.CrossRef | 1:CAS:528:DC%2BC38XhvVWqu7bE&md5=aff64a9889122196d14a3f34225aeacfCAS |
Saunders GW, Moore TE (2013) Refinements for the amplification and sequencing of red algal DNA barcode and RedToL phylogenetic markers: a summary of current primers, profiles and strategies. Algae – Korean Phycological Society 28, 31–43.
| Refinements for the amplification and sequencing of red algal DNA barcode and RedToL phylogenetic markers: a summary of current primers, profiles and strategies.CrossRef | 1:CAS:528:DC%2BC3sXntVagsbs%3D&md5=8a1316b91b6e73ffe4ccb21aae89e345CAS |
Saunders GW, Chiovitti A, Kraft GT (2004) Small-subunit rDNA sequences from representatives of selected families of the Gigartinales and Rhodymeniales (Rhodophyta). 3. Delineating the Gigartinales sensu stricto. Canadian Journal of Botany 82, 43–74.
| Small-subunit rDNA sequences from representatives of selected families of the Gigartinales and Rhodymeniales (Rhodophyta). 3. Delineating the Gigartinales sensu stricto.CrossRef | 1:CAS:528:DC%2BD2cXjtFCqsro%3D&md5=e9ede18017c14f858567270a30878414CAS |
Schmitz F (1889) Systematische Ubersicht der bisher bekannten Gattungen der Florideen. Flora oder Allgemeine botanische Zeitung 72, 435–456.
Stamatakis A, Hoover P, Rougemont J (2008) A rapid bootstrap algorithm for the RAxML web-servers. Systematic Biology 57, 758–771.
| A rapid bootstrap algorithm for the RAxML web-servers.CrossRef |
Watson JE (1973) Pearson Island Expedition 1969. 9. Hydroids. Transactions of the Royal Society of South Australia 97, 153–200.
Zanardini G (1874) Phycae Australicae novae et minus cognitae. Flora 57, 486–490, 497–505.