J OQURNAL of the ADELAIDE

BOTANIC GARDENS

AN OPEN ACCESS JOURNAL FOR AUSTRALIAN SYSTEMATIC BOTANY

flora.sa.gov.au/jabg

Published by the STATE HERBARIUM OF SOUTH AUSTRALIA on behalf of the BOARD OF THE BOTANIC GARDENS AND STATE HERBARIUM

State Herbarium of South Australia PO Box 2732 Kent Town SA 5071 Australia

Board of the Botanic Gardens and Reit State Herbarium

of South Australia Department of Environment, Water and Natural Resources

Journal of the Adelaide Botanic Gardens 27 (2014) 1-5

Placynthium australiense sp. nov. (lichenised Ascomycota, Placynthiaceae) from South Australia

Patrick M. McCarthy’ & Gintaras Kantvilas°

“ Australian Biological Resources Study, G.P.O. Box 787, Canberra, Australian Capital Territory 2601 E-mail: Patrick.McCarthy@environment.gov.au

’Tasmanian Herbarium, P.O. Box 5058, UTAS LPO, Sandy Bay, Tasmania 7005 E-mail: Gintaras.Kantvilas@tmag.tas.gov.au

Abstract

Placynthium australiense P.M.McCarthy & Kantvilas (Peltigerales, Placynthiaceae) 1s described from hard mesic limestone in south-eastern South Australia. It 1s characterised by having discontinuous colonies of mainly squamulose thalli that lack a hypothallus and prothallus, as well as comparatively broad, |-septate ascospores. Three other lichens, Candelariella aurella (Hoftm.) Zahlbr., Endocarpon pallidum (Nyl.) Nyl. and Verrucaria calciseda DC., are reported

for the first time from South Australia.

Key words: biodiversity, lichen, new species, taxonomy, Placynthium, South Australia.

Introduction

The lichen genus Placynthium (Ach.) Gray (Pel- tigerales, Placynthiaceae), with about 25 species, grows mainly on dry to moist or inundated calcareous and siliceous rocks in temperate regions of the Northern Hemisphere. The usually dark to blackish thallus contains a cyanobacterial photobiont, often has a well- developed prothallus, and ranges in habit from crustose- areolate and densely coralloid-isidiate to squamulose with entire or dissected margins, or rosette-like with short to elongate lobes. Apothecia are black, lecideine, with an amyloid hymenium, mainly 8-spored Peltigera- type asci, and hyaline, transversely septate ascospores.

The earliest records of Placynthium from Australia were of P nigrum (Huds.) Gray from Tasmania (Bratt & Cashin 1975) and, later, that species and P. subradiatum (Nyl.) Arnold from southern New South Wales (Weber 1977). Henssen (1984) reported the latter from the Australian Capital Territory, while Allen et al. (2001) cited collections of P nigrum from South Australia and the A.C.T. Records of both species from Victoria (McCarthy 2013) require confirmation. In this contribution, a new species, P. australiense, 1s described from hard mesic limestone in south-eastern South Australia, while three associated lichens, Candelariella aurella (Hoftm.) Zahlbr., Endocarpon pallidum (Nyl.) Nyl. and Verrucaria calciseda DC., are new records for the state.

Methods

Observations and measurements of photobiont cells, thallus and apothecium anatomy, asci, ascospores and conidia were made on hand-cut sections mounted in

Published online: 18 Feb. 2014 ¢ flora.sa.gov.au/jabg

l

water and dilute KOH (K). Asci were also observed in Lugol’s Iodine (1), with and without pretreatment in K.

Taxonomy

Placynthium australiense P.M.McCarthy & Kantvilas, Sp. nov. Thallus ater, epruinosus, non lobatus, areolatus vel ex squamulis constans, hypothallo prothalloque destitutus; algae Scytonema pertinentes; squamulae rotundatae vel irregulares, 0.3—3 mm latae, 0.1—0.6 mm crassae, dispersae vel in tumulis aggregatae, superficie laevi, nodulosa vel coralloideo-isidiata, margine vulgo effigurata. Apothecia abunda, atra, lecideina, 0.25— 0.66 mm diametro, ascosporis uniseptatis, comparate latis, 9-15 um longis, 5.5—8 um latis. MycoBank No.: MB 807101 Typus: SouTtH AUSTRALIA. Murray River Region: beside Marne River, 10 km NE of Springton, 34°40°12”S, 139°09°56”E, alt. 280 m, on gently sloping limestone slabs in pasture with Eucalyptus camaldulensis, 12 Apr. 2013, PM. McCarthy 4010 (holo.: AD; iso.: CANB, HO).

Thallus epilithic, initially crustose, richly rimose or areolate and 30—60 um thick, the areoles becoming larger and thicker and often subsquamulose or squamulose, these structures scattered and often resembling minute rosettes, or aggregated and forming colonies (2—) 5—10 (-15) mm wide, greenish-black to black, dull, not swollen and not noticeably gelatinous when wetted, epruinose at all stages of development; often a range of immature, mature and post-mature thall1 occupying much of an area of limestone up to 10—20 cm wide, but generally discontinuous and interspersed with small colonies of crustose lichens (mainly Lecanoraceae, Physciaceae, Teloschistaceae and _ Verrucariaceae).

ISSN 0313-4083 (Print) « ISSN 2201-9855 (Online)

P.M. McCarthy & G. Kantvilas

; een”

ry o* bt yt re shat 1 uy r

- aN

a + \ a

te we tea ie) %

Cd ee ast

a Fgh 8 ES ee oe a7 +) ee ieee 3

ay

weRl i won, De aa

wits = & =. 3 eee

Oe Pe at’

a = oy . rea fe

2 2

* ,a

=

ey Ft +

¥

ee

J. Adelaide Bot. Gard. 27 (2014)

Fz? bs . ¥ sa eietay A a4 ad " tag yt ra =z A SY

1 7 v * . iS sf" re A Ki rt Sa ap

ia et DTT: se

ocatet

ara.

i] a* bal tay + ey + Fas gt ay ot Ta ie a 4 * ee Se * zis ee ™ ry 4 " * a

a Fr *

A r "relate x . 7

Pr er: +s Se a.

= .* *

a a” a

at

cs ps |

hem ge ha ee ae

haere a? Ce aS a

Heelan my si , te i % : _ eS - a - * = 4 oe 1 af ™ be -- 4 iax 12 a et a , a PPTL Bos 26 Sora oa aa

LJ tal * a . eel g ee = ia % . i , he 5 tag ate Oe ee er ee eee el + aes at oo * 1, F at Penang” _* trie te r . © a = # a

B * a + ee eee fase pee, ae" ase, ate op ot a or ke a ee See a a ad ae pat dated 2.) * _ = Fpaat* a

a pa es r 2 ase *

Fig. 1. Placynthium australiense. Habit of thalli and apothecia. Scale bar: 1 mm. — Holotype.

Sguamules and areoles rounded to angular or irregular, closely aggregated in a crust or more scattered and often laminally short-lobulate (with lobules 40-80 um wide) or marginally effigurate, (0.3—) 0.6—2.2 (-3) mm wide and 0.1—0.35 (—0.6) mm thick, thickest when the thallus is densely isidiate; squamules and areoles with a blackish underside, often markedly constricted at their attachment to the substratum, lacking dedicated attachment organs, such as rhizines. Thallus margin usually indistinct, rarely with contiguous to somewhat discrete simple lobes that are tightly appressed to the substratum, to | mm long and 0.15—0.25 mm wide. Thallus surface rather smooth, or granulose, nodulose or isidiate; isidia 20-50 (—80) um diam. and up to 0.25 mm long, simple to irregularly branched, finger-like, furcate or + coralloid, usually more or less erect, others tilted to horizontal and smooth to contorted. Thalline anatomy paraplectenchymatous, the cells 5—10 (-15) um diam., largest towards the thallus interior, indistinctly corticate; cortical zone yellowish-brown to olive-brown,

amorphous or with rounded and comparatively thick- walled cells 6-12 um diam. Photobiont cyanobacterial, Scytonema-like, consisting of scattered or irregularly clustered cells and short to moderately long filaments, occupying almost the entire thallus; cells in filaments yellowish-brown, 8-12 wm wide and 4-10 um long. Hypothallus absent. Prothallus usually absent around thalli as well as isolated squamules and areoles; a few squamules with an indistinct and discontinuous blue- black, fimbriate prothallus extending up to 0.3 mm beyond the margin. Apothecia sparse to very numerous, laminal, usually solitary, adnate to superficial, not constricted at the base, lecideine, (0.25—) 0.42 (—0.66) mm diam. [7 = 60], jet-black, usually matt, occasionally slightly glossy (mainly immature apothecia), colour unchanged when wetted; disc usually plane, occasionally slightly to moderately concave or convex at maturity, the surface smooth to minutely and irregularly uneven; proper exciple concolorous with the disc or a little paler, smooth, usually entire, sometimes faintly undulate

J. Adelaide Bot. Gard. 27 (2014)

C

New Placynthium species from South Australia

a= - * a ' ~~ e = pe A + Looe. a = hs

if 4 na Lam aie} 2; fie * e 7a

P pe 4 Y 6

—_ Te

i, +

Fig. 2. Placynthium australiense. A, B sectioned apothecia (semi-schematic); C distal cells of paraphyses; D ascospores. Scale bars: A 0.5 mm;

B 0.1 mm; C, D 10 um. — A-D holotype.

or flexuose, 30-80 um thick, often slightly raised, persistent or becoming almost excluded (especially around the most convex apothecia); in section uniformly dark, consisting of radiating, tightly packed hyphae, continuous below the hypothecium, (30—) 40—80 (—100) um thick at the sides, 60—120 (—150) um thick at the base; cells at the margin ellipsoid to elongate-ellipsoid, 8-17 x 5-10 um, with thick greenish-black walls. Hypothecium pale to medium greenish-brown or rather dark golden or orange-brown, 50—80 (—100) um thick, not inspersed with oil droplets or granules, I+ deep blue (fading) to almost black (without pretreatment in K) or I+ deep blue (fading) to reddish-brown (with pre- treatment). Hymenium 70-100 (—110) um thick, not inspersed with oil droplets or granules, I+ persistently deep blue to bluish-black (with and without pretreatment in K), subtending a greenish-black or violet-blackish epihymenium 10-15 (—20) um thick. Paraphyses unbranched to sparingly branched and anastomosed distally (scattered branches or anastomoses at all levels of the hymenium), short- to rather long-celled, 3-4 (—6) um wide, thin-walled, remaining coherent in water and K; shape of apical cells very variable even within a single apothecium, ranging from strongly capitate, with the apical cell rounded or somewhat pointed and 4—5 (—6) um wide and with a hyaline or partly dark green to violet-blackish wall, or the distal 3—5 cells a little shorter and broader than more proximal cells, or the apical cells little altered 1n size and shape other than being included

within the pigmented epihymenium. Asci narrowly to broadly clavate or clavate-cylindrical, 8-spored, 58—75 x 12-17 um [n =20], with an external amyloid cap and a thin internal amyloid sheet adjacent to the apex of the ascoplasma. Ascospores colourless, 1-septate, overlapping-uniseriate to irregularly biseriate in the asci, ellipsoid, usually slightly constricted at the septum, uniformly thin-walled, lacking a distinct perispore, (9—) 12 (—15) x (5.5—) 7 (8) um [n = 90]; cells more or less identical in size and shape; apices rounded to subacute; contents usually granular and guttulate. Pycnidia spherical, semi-immersed to almost fully immersed in the thallus, 70-100 um diam., with a greenish-black apex and a hyaline conidiogenous layer; conidiophores short-celled, 10-20 um long. Conidia bacilliform, 3—6 (—7) x c. 0.7 um. Fig, 1-3.

Remarks. Placynthium australiense 1s characterised by its blackish, epruinose, non-lobate thallus with squamules and areoles that are neither subtended by a hypothallus nor delimited by a prothallus; it contains a Scytonema-like photobiont. The squamules are rounded or irregular in shape, (0.3—) 0.6—2.2 (—3) mm wide and 0.1—0.35 (—0.6) mm thick, scattered or contiguous in small groups, with a smooth, nodulose or coralloid- isidiate surface, commonly with an effigurate margin, abundantly fertile with comparatively small, black, lecideine apothecia (0.25—) 0.42 (—0.66) mm diam. and with comparatively broad, l-septate ascospores measuring (9—) 12 (—15) x (5.5—) 7 (—8) um.

P.M. McCarthy & G. Kantvilas

Placynthium can be subdivided more or less evenly into species with distinct, radial, elongate, marginal lobes and those with thalli that lack well-defined lobes, although individual squamules can have minutely and shallowly effigurate margins. Placynthium australiense lies comfortably among the latter, its distinctive and persistently 1-septate spores setting it apart from all but three known taxa.

Placynthium tremniacum (A.Massal.) Jatta, from Iceland, the British Isles, continental Europe, Macaro- nesia, North Africa and Central Asia, has a glossy brown thallus, dark brown apothecia, an indistinct prothallus and narrower ascospores (9-16 x4—6 um: Clauzade & Roux 1985; Jergensen 2007; Gilbert & James 2009; Burgaz 2010). However, its relationship with P nigrum has yet to be fully resolved, because while most recent authors have regarded the persistently 1-septate spores of P tremniacum as being diagnostic for a distinct species, Czeika & Czeika (2007) examined syntype material, observed a minority of 3-septate propagules, and reduced P tremniacum to synonymy under P. nigrum.

Placynthium anemoideum (Servit) Gyeln., from the British Isles, France, Croatia and Turkey, has isolated areoles with raised crenulate margins (Czeika & Czeika 2007: Fig. 6b) which are not unlike those of P. australiense, but the ascospores are smaller (9-10 x 5—6 um) and the apothecia are only c. 0.3 mm diam. (Clauzade & Roux 1985; Czeika & Czeika 2007; Gilbert & James 2009).

Finally, ascospores very similar to those of P australiense occur in P. tantaleum (Hepp) Hue, which is known from Greenland, Iceland, Svalbard, the British Isles, continental Europe (Clauzade & Roux 1985; Czeika & Czeika 2007; Jorgensen 2007; Gilbert & James 2009; Burgaz 2010), south-western (Schultz 2002, as P nigrum (Huds.) Gray) and north-eastern U.S.A., Canada (Henssen 1963, as P nigrum var. tantaleum (Hepp) Arnold) and Central Asia. However, P. tantaleum has a glossy, mottled grey-brown thallus to 3 cm wide, resting on “a voluminous blue-green prothallus” (Jorgensen 2007) or on “a distinctive blue-black hypothallus” (Gilbert & James 2009).

Placynthium nigrum, the most common and widely distributed species, has comparatively elongate, 1-—3- septate spores [8—25 <x 4-6 um (Clauzade & Roux 1985); 7-22 <x 3.5-6 um (Schultz 2002); 10-15 (—20) x 4-6 um (Jorgensen 2007); 8-18 < 4-6 um (Czeika & Czeika 2007); 7-22 x 3.5-6 um (Gilbert & James 2009); 7.5—20 x 2.5—7.5 um (Gilbert & James 2009)]. However, when Henssen reported P nigrum from Argentina, Chile and New Zealand (Henssen 1984), she circumscribed the species “in the broad sense including var. tantaleum (Hepp) Arnold (see Henssen 1963)” and noted “only two-celled spores have been observed in the specimens from the Southern Hemisphere” (Henssen 1984). Recently, Jorgensen (2007) observed that while P tantaleum usually occurs 1n and beside rivers, specimens

J. Adelaide Bot. Gard. 27 (2014)

Fig. 3. Placynthium australiense. Immature (A), submature (B) and mature (C) asci stained with Lugol's lodine. Scale bar: 10 um. — HO isotype.

from drier habitats with thin-walled, l-septate spores “appear only to be part of the variation of P nigrum’. Among the South Australian and Tasmanian specimens examined, there seems to a clear distinction, based on ascospore septation and dimensions, between P. nigrum as it 1s commonly circumscribed (see above) and P nigrum sensu Henssen (1984) with comparatively short and broad 1-septate ascospores. Whether or not both can be accommodated within the variation of P. nigrum will require further study.

There have been inconsistencies 1n the description of the asci of Placynthium by previous authors. Whereas Jorgensen (2007) referred to an amyloid cap and internal sheets, Schultz (2002) mentioned an amyloid tube. Our observations are illustrated in Fig. 3 and interpreted thus: when young, the asci have an intensely amyloid external cap and a thin, internal, amyloid sheet at the base of the tholus adjacent to the apex of the ascoplasm (A). As the asci mature, a distinct, beak-like ocular chamber develops which pushes up through the inner amyloid sheet; the two sides of this sheet become orientated more or less vertically, somewhat approximating the appearance of an amyloid tube-like structure (B). With further development, this inner amyloid structure becomes more squashed and less prominent, although its vestiges can still be evident at the edges of the ocular chamber; the external cap remains prominent throughout

(C).

J. Adelaide Bot. Gard. 27 (2014)

Distribution & habitat. Placynthium australiense 1s known only from gently sloping, slab-like outcrops of hard limestone at its type locality in the Murray River region of south-eastern South Australia. Associated lichens include Aspicilia contorta (Hoftm.) Kremp., Buellia albula (Nyl.) Miull.Arg., Caloplaca mereschkowskiana S.Y.Kondr. & Kérnefelt, Caloplaca spp., Candelariella aurella (Hoffm.) Zahlbr.*, En- docarpon pallidum (Nyl.) Nyl.*, Lecania_ turicensis (Hepp) Miull.Arg., Lecanora dispersa (Pers.) Sommertf., L. sphaerospora Miill.Arg., Placidium sp., Rinodina bischoffii (Hepp) A.Massal., Yoninia aff. aromatica (Sm.) A.Massal., Verrucaria calciseda DC.*, V. muralis Ach. and . nigrescens Pers. The three species marked with an asterisk are new records for South Australia.

Key to the Australian Species of Placynthium

1. Central areoles commonly eroded, leaving arcs or rings of contiguous, radiating marginal lobes 1—1.5 mm long and Q.1—0.25 mm wide; prothallus and hypothallus absent .. .

Se ot Be 4 F0, Rete obo ha eriiag ciple Placynthium subradiatum

1: Central areoles not eroded; thallus without marginal lobes;

prothallus and hypothallus present or absent

2. Thallus a wide-spreading crust of more or less uniform,

eranulose- and coralloid-isidiate areoles; prothallus

usually extending well beyond the thallus margin; hypothallus often thick and visible between the areoles

eT Lees etenlsysePeta ren & ted eases Placynthium nigrum s.l\at.

2: Thallus of rounded to irregular squamules that are

scattered or aggregated in small groups; prothallus absent or very faint; hypothallus absent ............

sok S Reb t Hadadloe ea coHee eho Placynthium australiense

Additional specimens examined

Placynthium nigrum s.str.

SOUTH AUSTRALIA. Murray River Region: 15 km S of Angaston, on marble in pasture, 21 Oct. 1981, J.A.Elix 9243 (CANB 9602209, det. A. Henssen, Sep. 1982).

TASMANIA. Vale of Belvoir, alt. 840 m, on limestone outcrops in buttongrass moorland and heath, 16 May 1987, G.Kantvilas 61/87 (HO 569326); Tiger Road, 1 km NW of Florentine River bridge, 42°35’S, 146°26’E, alt. 370 m, on limestone boulder at edge of wet forest, 17 Dec. 2003, G.Kantvilas 742/03 (HO 524542).

Placynthium nigrum s.l\at. (incl. sensu Henssen 1984)

SOUTH AUSTRALIA. Yorke Peninsula: Innes National Park, West Cape, 35°14’°49”S, 136°497°31”E, alt. 45 m, on clifftop limestone outcrops and boulders, 14 Apr. 2013, P.M.McCarthy 4029 (AD). Kangaroo Island: Cape Borda, 35°45’S, 136°35’E, alt. 100 m, on semi-exposed limestone rocks in dense heathy vegetation, 27 Sep. 1994, A. Streimann 54956 (AD, B, CANB); track to Cape Gantheaume, 36°04’S, 137°27°E, on coastal limestone outcrops in heathland, 29 Sep. 2008, G.Kantvilas 321/08 (AD, HO); Ravine des Casoars, 35°48’S, 136°35’E, alt. 5 m, on limestone outcrops in coastal heathland, 24 Sep. 2012, G.Kantvilas 453/12 (AD, HO); Point Ellen, 36°00’S, 137°11°E, alt. 5 m, on limestone outcrops in coastal heathland, 26 Sep. 2013, G.Kantvilas 214/13 (AD, HO).

TASMANIA. c. 6 km SW of Hardwood Hill, alt. 75 m, on limestone outcrops in sedgeland heath, 25 Apr. 1985, G.Kantvilas 178/85 (HO 308227, MB); Mole Creek, alt. 350 m, on limestone outcrop in paddock, 19 Feb. 1984, PJames &

New Placynthium species from South Australia

G.Kantvilas 364/84 (BM, HO 308225, MB [det. A.Henssen]); W of Mole Creek, near Liena Road and Mersey Forest Road junction, | km NW of Florentine River bridge, 41°34’S, 146°15’E, alt. 350 m, on limestone outcrop in pasture, 19 Feb. 1984, G.Kantvilas 365/84 & P.James (BM, HO 564173); Eleven Road, Florentine Valley, 42°37’S, 146°26’E, alt. 430 m, on limestone boulder at edge of wet forest, 17 Dec. 2003, G.Kantvilas 744/03 (HO 524548); Flinders Island, Killiecrankie Bay, W of township, 39°50’S, 147°50’E, alt. 5 m, on outcrop of soft calcarenite along seashore, 2 Apr. 2007, G.Kantvilas 150/07 (HO 544228).

Acknowledgements

The first author is grateful to Jack Elix for his company and assistance in the field, and for the loan of specimens. Jean Jarman assisted with preparing Figure 3 for publication.

References

Allen, D., Lumbsch, H.T., Madden, S. & Sipman, H. (2001). New Australian and Australian state lichen and lichenicolous lichen reports. Journal of the Hattori Botanical Laboratory 90: 269-291.

Bratt, G.C. & Cashin, J.A. (1975). Additions to the lichen flora of Tasmania I. Papers and Proceedings of the Royal Society of Tasmania 109: 17-20.

Burgaz, A.R. (2010). Nuevos datos sobre el género Placynthium en la Peninsula Ibérica. Botanica Complutensis 34: 5—19.

Clauzade,G. & Roux, C. (1985). Likenoj de Okcidenta Etrropo. Ilustrita Determinlibro. Bulletin de la Société Botanique du Centre-Ouest, nouvelle série, numéro spécial 7: 1-893.

Czeika, H. & Czeika, G. (2007). Placynthium in den Alpen und Karpaten sowie in benachbarten Gebieten. Herzogia 20: 29-51.

Gilbert, O.L. & James, P.W. (2009). Placynthium (Ach.) Gray (1821). In: Smith, C.W., Aptroot, A., Coppins, B.J., Fletcher, A., Gilbert, O.L., James, P.W. & Wolseley, P.A. (eds), The Lichens of Great Britain and Ireland, pp. 714— 718. (British Lichen Society: London).

Henssen, A. (1963). The North American species of Pla- cynthium. Canadian Journal of Botany 41: 1687-1724. Henssen, A. (1984). Placynthium arachnoideum, a new lichen from Patagonia, and notes on other species of the genus in

the Southern Hemisphere. Lichenologist 16: 265-271.

Jorgensen, P.M. (2007). Placynthiaceae. In: Ahti, T., Jorgensen, P.M., Kristinsson, H., Moberg, R., Sochting, U. & Thor, G. (eds), Nordic lichen flora 3: Cyanolichens, pp. 714—718. (Nordic Lichen Society & Museum of Evolution, Uppsala University: Uppsala).

McCarthy, P.M. (2013). Checklist of the lichens of Australia and its island territories. (Australian Biological Resources Study: Canberra). http://www.anbg.gov.au/abrs/lichenlist/ introduction.htm! [Version 17 September 2013].

Schultz, M. (2002). Placynthium. In: Nash III, T.H., Ryan, B.D., Gries, C. & Bungartz, F. (eds), Lichen flora of the Greater Sonoran Desert region |: 397-399. (Lichens Unlimited: Tempe, Arizona).

Weber, W.A. (1977). Placynthium (Ach.) S.F.Gray, a genus of lichens previously unreported from Australia. Muelleria 3: 250;

J OQURNAL of the ADELAIDE

BOTANIC GARDENS

AN OPEN ACCESS JOURNAL FOR AUSTRALIAN SYSTEMATIC BOTANY

flora.sa.gov.au/jabg

Published by the STATE HERBARIUM OF SOUTH AUSTRALIA on behalf of the BOARD OF THE BOTANIC GARDENS AND STATE HERBARIUM

State Herbarium of South Australia PO Box 2732 Kent Town SA 5071 Australia

Board of the Botanic Gardens and Reit State Herbarium

of South Australia Department of Environment, Water and Natural Resources

Journal of the Adelaide Botanic Gardens 27 (2014) 7-21

Drosera murfetu (Droseraceae): a new species from Tasmania, Australia

Allen Lowrie? & John G. Conran?

“6 Glenn Place, Duncraig, Western Australia 6023

° Australian Centre for Evolutionary Biology and Biodiversity and Sprigg Geobiology Centre, School of Earth & Environmental Sciences, Benham Building, DX 650 312, The University of Adelaide, South Australia 5005 Email: john.conran@adelaide.edu.au

Abstract

Drosera murfetii Lowrie & Conran is a new species from south-west Tasmania, Australia, bearing mainly non-carnivorous leaves and usually several large flowers per inflorescence. It is compared to its morphologically closest taxon, Drosera arcturi Hook., with which it often crows. The two species are described and illustrated, a key to species and a comparison table of characters is provided. Drosera arcturi 1s lectotypified.

Keywords: Drosera arcturi, D. murfetii, Droseraceae, taxonomy, new species, Tasmania.

Introduction

Drosera arcturi Hook. from alpine southern Australia and New Zealand is very variable across its range and in Tasmania sometimes produces large, robust plants (Morris 2009), including a form bearing several conspicuous, basal glabrous leaves, a usually single, large trapping leaf, with obvious glandular hairs, and often several flowers per scape, each with 4—5 larger, ereenish-white or cerise stigmas (Lowrie 1998, Clayton 2003, Gibson 2010). This form is widespread in south- west Tasmania and contrasts strongly with the type form (with which it often co-occurs at more elevated altitudes). The type form of D. arcturi is generally a smaller plant and always bears numerous trapping leaves, highly reduced (or no) scale-like glabrous leaves, almost always single flowers per scape, and usually three stigmas (rarely four). The so-called ‘giant form’ of D. arcturi was considered by Gibson (1998, 1999, 2010) to be at least an ecotype, while Clayton (2003) thought that the two forms should deserve at least subspecific status; both authors suggested that further study was needed. However, the fact that both morphotypes occur together over such a wide geographic range and in the same habitats, but without apparent intermediate forms, suggests instead that they may be distinct albeit closely related species, as subspecies are generally regarded as being geographically (Schlauer 1996) and/ or ecologically allopatric (Stace 1989).

Although D. arcturi also occurs in New Zealand, Where it grows mainly in montane to subalpine bogs southwards from 39°S (Allan Herbarium 2000, Salmon 2001), it extends down to sea level in the far south of its range in both countries (Allan Herbarium 2000, CHAH 2013). Several New Zealand morphotypes have been

Published online: 1 Apr. 2014 ¢ flora.sa.gov.au/jabg

given taxonomic status in the past (Colenso 1890, 1896, 1899), but these were later deemed to be relatively trivial size or foliage colour variants and reduced to synonymy (Cheeseman 1906). Specimens with 4—5 large, greenish white or cerise stigmas are absent from both New Zealand and mainland Australian populations and any large plants occurring there also lack the 2—5 large, non-carnivorous basal leaves seen in the Tasmanian ‘siant form’ plants. Although isolated plants in New Zealand very rarely bear two flowers per peduncle, the plants and flowers are otherwise identical to the ‘type’ form both in size and morphology (Conran, pers. obs.)

The two Australian morphs appear to differ con- sistently from each other for a range of characteristics relating in particular to the glabrous leaves, flower and stigma number, and there is a lack of apparent intermediate forms for these features when the morphs co-occur. Accordingly, this study compares the morphology of the two forms and examines the taxonomic implications of these differences.

Methods

Living and dried specimens of D. arcturi across its geographic range and covering all known Australian and New Zealand morphotypes were examined under dissecting microscope and/or by SEM. Descriptions and, where possible, type specimens for the formerly segregated names from New Zealand were investigated, as well as D. arcturi collections at AD, HO, MEL, OTA, PERTH, Gunn’s and Brown’s collections at K and BM, and personal collections by A. Lowrie and D.E. Murfet.

To assess character variability in the complex, up to five complete specimens per herbarium sheet, re- presenting a total of 354 specimens from 107 accessions

ISSN 0313-4083 (Print) * ISSN 2201-9855 (Online)

A. Lowrie & J.G. Conran

J. Adelaide Bot. Gard. 27 (2014)

Table 1. T-test comparisons between PATN dendrogram groups for the characters used in the cluster analysis assuming unequal variances, as well as Kruskal Wallace and r° values from the cluster and ordination analyses of 354 specimens from 107 accessions.

D. arcturi D. murfetiz

Character code mean + SD ders SD E# 0.56 + 0.83 2.91 + 0.88 EL (mm) 4.31 +6.02 33.77 + 16.67 EW (mm) 0:87 + 1.11 5.48 + 1.47 G# 4.92 + 1.44 1.34 + 0.57 GP (mm) 17.21 + 7.75 24.46 + 14.63 GL (mm) 23.20 + 9.36 42.71 + 22.21 SC (mm) 44.08 + 17.31 92.46 + 33.13 F# 1.01 + 0.10 1.20 + 0.50 SL (mm) 6.33 + 1.46 8.29 + 1.74 SW (mm) 1.83 + 0.47 2.43 + 1.05 PL (mm) 6.57 + 1.48 8.99 + 1.85

t df P KW y? Do dS 1, 334.38 <Q.0001 224.30 0.887 21.43 1, 290.83 <Q.0001 130.51 0.737 33.12 1, 194.18 <Q.0001 145.05 0.727 32.29 1, 282.69 <Q.0001 251.54 0.768 5.58 be 27 <Q.0001 18.67 0.202 10.24 1, 193.96 <Q.0001 106.03 0.125 16.61 1, 224.95 <Q.0001 185.10 0.769 4.74 1, 167.23 <Q.0001 6.26 0.203 a Bea 1, 305.54 <Q.0001 88.50 0.684 6.65 1, 208.42 <Q.0001 57.56 0.367 Lao. 1, 296.40 <Q.0001 115.85 0.667

Table 2. Summary comparison of distinguishing morphological character and character states differences between Drosera arcturi and D. murfetii

derived from measurements of 354 specimens from 107 accessions.

Character Drosera arcturi Basal, glabrous leaf size (mm) 3-27 x 0.7-4 Glandular leaf number 2—12

Glandular leaf length (cm) 1.3-12

Scape length (cm) 0.8—9

Flower number L 2)

Petals (mm) 3-10 x 2.5-7 Style number 3 (4)

Stigma colour Seed size (mm)

collected from Australia and New Zealand, were coded for 11 morphological characters:

. Glabrous leaf number (E#)

. Longest glabrous leaf length (mm) (EL)

. Longest glabrous leaf width (mm) (E'W)

. Glandular leaf number (G#)

. Longest glandular petiole length (mm) (GP) . Longest glandular lamina length (mm) (GL) . Scape length (mm) (SC)

. Flowers per scape (F#)

. Sepal length (mm) (SL)

10. Sepal width (mm) (SW)

11. Petal length (mm) (PL).

These data were then transformed by log,,+1 for continuous data and square root + 0.5 for counts (Zar 1996) and then subjected to cluster analysis using Gower Association, flexible UPGMA with a beta value of 0.0. Ordination used non-metric semi-strong multidimensional scaling (SSH) with 100 random starts and a cutoff value of 0.9 in the program PATN v. 3.1.2 (Belbin & Collins 2008), with character/specimen cluster relationships in the ordination space explored using Principal Component Correlation (PCC) analysis (Faith 1991). In addition, character mean comparisons were undertaken on the raw data using t-tests in JMP 4.0.3 (SAS Institute 2000) for the two groups defined by the PATN cluster analyses, with the assumption of unequal variances.

Seed micromorphology is considered important in Drosera at subgeneric and sectional level (Dwyer 1983, Boesewinkel 1989), as well as between related species

\O COND NA B WN

greenish-white to pale yellow

0.5—1 (1.3) x 0.40.8

Drosera murfetii

(l=) 3=5(=12) * 5=15

(O—) 1-2 (-4) 3.5—21.7 3—20

1-4

5-14 x 4-5 4—5

ereenish-white or reddish-cerise

(0.8—) 1-2 x (0.4—) 0.6—1

(Susandarini et al. 2002, Lowrie 2005, Lowrie & Conran 2007, 2008). Accordingly, ten seeds per sample from 14 samples, representing both morphs and including material from New Zealand, were measured and the two morphotypes compared using the Student t-test on log, +1 transformed data, allowing for unequal variances. SEM micrographs were also produced at The University of Adelaide Microscopy Centre (CEMMSA), following the methods of Susandarini et al. (2002), with additional information on the seeds of New Zealand material of D. arcturi obtained from Webb & Simpson (2001).

Results

The cluster analysis (Fig. 1) and ordination (Fig. 2) both show that there are two clearly separated morphotypes within the D. arcturi complex, differing consistently for nearly all of the characters measured. One cluster represents all of the (O—) 1—2 (-4) trapping- leaved, multi-flowered specimens with multiple, large, broad, glabrous leaves from SW Tasmania, including all samples reported to have cerise stigmas; the other, consists of the multiple glandular-leaved, single- flowered and widespread type form with no or few, small, linear, glabrous leaves from New South Wales, Victoria, Tasmania and New Zealand. The vegetative morphological differences between the two forms, as well as inflorescence and floral characteristics clearly support the justification to treat these two taxa as distinctive species (Tables 1, 2), particularly as they co-

J. Adelaide Bot. Gard. 27 (2014)

0.28 D. arcturi

0.21

Gower dissimilarity oO BR

0.07}

|

0.0

f

A new Drosera from Tasmania (Droseraceae)

D. murfetii

ii

Fig. 1. Gower Metric flexible UPGMA clustering dendrogram of 354 specimens from 107 accessions in the Drosera arcturi complex from SE Australia,

Tasmania and New Zealand.

occur across a wide part of their respective ranges with no obvious intermediacy seen in our analysis (Stace 1989). Accordingly, a new species, Drosera murfetii, 18 proposed and described here and D. arcturi is redefined. Illustrations, descriptions and a comparison table for both species are provided.

Character comparisons between the two clusters (Table |) show that there are significant mean differences for all features used in the analyses, with D. murfetii

D. murfetii L

stress: 0.09 -2.015

2.293

EF Fyy

EL \ a

on average larger for all features, as well as generally bearing more flowers per scape (1.e. >1) and having larger, broader and more numerous glaborous leaves. In contrast, the type form of D. arcturi (which also included all type specimens and related accessions from Australian and New Zealand taxa in the same cluster) has few, small and linear (if any) glabrous leaves, more numerous glandular, trapping ones, but was otherwise significally smaller for the features measured here.

1.046

a A D. arcturi

GH -1.232 1.324

Fig. 2. Non-metric semi-strong hybrid 3-D multidimensional scaling ordination of the Drosera arcturi complex specimens and clusters from the dendrogram shown in Fig. 1; white = D. arctur; red = D. murfetii; Principal Canonical Correlation vector arrows indicate direction of increasing score

in the ordination space for the characters listed in Table 1.

A. Lowrie & J.G. Conran

(

=:

J. Adelaide Bot. Gard. 27 (2014)

Fig. 3. Seeds of Australasian alpine Drosera arcturi complex species under SEM. A-C Drosera arcturi: A seed; B close up of funicle; C detail of testa. D-F D. murfetii: D seed; E close up of funicle; F detail of testa. Scale bars: A-D 200 um; E-F 100 um. — A-C A.Lowrie 3178; D-F A.Lowne 3187

& D.E.Murtet.

The ordination (Fig. 2) shows that the two clusters had virtually no overlap. The stress value of 0.09 is well within the limits deemed acceptable for dimensional distortion (Belbin & Collins 2008) and the PCC plot shows the same pattern as the t-test and Kruskal- Wallace (KW) comparisons between the clusters from the dendrogram, albeit with much lower r* (<0.4) for glandular leaf size features, flower number and sepal width, indicating that these were not as important in the ordination space as the other features (all >0.6).

10

The features with the highest t, KW and 7’ scores were glabrous leaf number, length, width, glandular leaf number, scape length, sepal and petal length.

The overall character ranges for features observed to differ between the two taxa are listed in Table 2 (including some excluded from the cluster analysis as there were too many specimens missing those characteristics and/or the binary nature of the feature may have skewed the result). The seeds of the two forms, although similar in shape and testal morphology to each

J. Adelaide Bot. Gard. 27 (2014)

other (Fig. 3), as might be expected for closely related taxa, differ in overall size; the seeds of Drosera murfetii being significantly longer (¢t = 14.762, df 1, 81.87, P <Q.0001) and wider than those of D. arcturi (t = 8.505, df 1, 90.62, P <0.0001), though there is some overlap.

Amended key to the species of Tasmanian Droseraceae

The key to Tasmanian Droseraceae of Morris (2009) should be amended to incorporate the new species by inserting a new couplet after number 5, as follows:

5. Leaves orbicular, < 2 cm long; flowers 2-20 ........... sg nehiRe GET recht Ee Ure eioe enat acraserneee er cee D. glanduligera

5: Leaves linear-oblong, 3—20 cm long; flowers 1—4 Sa. Basal leaves broad, spreading, glabrous; inner leaves (O—) 1—2 (-4), glandular, erect D. murfetii 5a: Basal leaves small or absent; glandular leaves numerous, often semi-erect D. arcturi

Taxonomy

1. Drosera arcturi Hook.

J. Bot. (Hooker) 1: 247 (1834), emend.; Icon. Pl. 1: tab. 56 (1837). — Type citation: “summit of Mt Arthur Mr Gunn, (n. 139 [sic].)”’, corrected by Diels, Pflanzenr. 4(112) [Heft 26]: 64 (1906): “Mount Arthur (Gunn n. 129 [...])”. Type sheet: “Summit of Mt Arthur, Van Dn’s [Diemen’s| Land, 129. W. Gunn 1832”, bottom collection on sheet. Lectotype (here designated, Fig. 6): K 000215043!*, pro parte [top middle specimen], ex Herb. Hook. Remaining Syntypes: K 00215043!* pro parte [remaining five specimens], MEL 96027A!

Drosera polyneura Colenso, Trans. & Proc. New Zealand Inst. 22: 460 (1890) [°1889”"]. — Type citation: “Swampy spots, base of Mount Tongariro, County of East Taupo; 1889.” Syntype: Swampy spots, base of Mount Tongariro, County of East Taupo, [1889,| A. Hill s.n. KK 000659184).

Drosera ruahinensis Colenso, Trans. & Proc. New Zealand Inst. 28: 593 (1896). — Type citation: “Ruahine Mountain-range: Mr. H.Hill, 1895; Mr. A. Olsen, 1895” Syntype: Ruahine Mountain-range, east side, wet spots near summits, [1895,] A.Olsen s.n. (X 000659178).

Drosera atra Colenso, Trans. & Proc. New Zealand Inst. 31: 269 (1899). — Type citation: “Ruahine Mountain-range, east side, wet spots near summits.” Syntype: Ruahine Mountain-range, east side, wet spots near summits, [1898,] A.Olsen s.n. (WELT 23581).

Drosera ligulata Colenso, Trans. & Proc. New Zealand Inst. 31: 269 (1899). — Type citation: “Euahine [sic] Mountain-range, east side, wet spots near summits; 1898.” Syntype: Ruahine Mountain-range, east side, in low-lying wet spots, [1898,] A.Olsen s.n.; n.v., fide Colenso (1899),

A cold tolerant, fibrous-rooted alpine perennial herb; rhizome 1—5 (>10) cm long, covered with persistent dead leaf bases. Hibernaculum bud-like, derived from glabrous basal leaf sheaths covering rhizome apex. Glabrous leaves at emergence (Q—) 1-5, 3-27 x 0.7-4 mm, lanceolate, apically acute, glabrous, often short- lived and absent by anthesis; g/andular leaves 2-12, narrowly lanceolate-spathulate, loosely distichous,

1]

A new Drosera from Tasmania (Droseraceae)

Fig. 4. Drosera arcturi. A plant; B lamina; C sepal; D petal; E gynoecium; F style and stigma. Scale bars: A 10 mm; B-D 5 mm; E-F 1 mm. — Drawn by A. Lowrie, 1988, from live material from the Snowy Mountains, New South Wales.

semi-erect to spreading, 1.3-12 cm long, green to ereenish-bronze, red or almost black, persisting after senescence; petioles stem-sheathing, 0.6—0.8 mm wide at sheath apex, 1.5—1.8 mm wide at lamina, glabrous; lamina 9-70 x 2—6 mm, narrowly oblong, apex rounded, adaxial surface bearing large, sub-marginal insect- catching glands and smaller glands within, abaxial surface glabrous with a prominent midrib. /nflorescence terminal, scapose, | (—2)-flowered, 0.8—9 cm long, 0.4—0.5 mm diam., glabrous, enclosed basally by leaf sheaths, persistent, erect after senescence. Flower shortly pedicellate, erect in fruit; pedicel 3—12 mm; floral bract subulate, 2—2.5 x 0.40.6 mm. Sepals greyish-green, 4—10 x 1-3 mm, narrowly elliptic, adaxially concave, horizontal to recurved at anthesis, margins entire, apex obtuse, irregularly crenate, both surfaces glabrous. Petals white, 3-10 x 2.5—7 mm, obovate, entire, apex broadly rounded. Stamens 5, 4—6 mm long, filaments white, anthers and pollen yellow. Ovary 44.5 x 2.5- 2.7 mm, ellipsoid, green, ovules numerous. Styles 3 (—4), greenish-white to pale yellow, 0.5—0.6 mm long (excluding stigmas), semi-erect, fusiform, flat in cross- section; stigmas + reniform-peltate, 0.6-0.7 =< 0.2—0.3

J. Adelaide Bot. Gard. 27 (2014)

A. Lowrie & J.G. Conran

tr * 7. 1 he, wee

¥ aS.

Oo. — > © OS a ©O 7 3 “NO > ? OG CD) - © Q.. | @ C=: ny Oo ®O > . = cy 20 Sat =H —- © Oo - =e c= O- OF — ae oO. Oa a, Pou TO —— = SE Bg

D close up of flower with 4 styles. Scale bars

J

-C habi

Fig. 5. Drosera arcturi in Australia and New Zealand. A-B Hartz Range, Tasman

Otago, New Zealand

12

J. Adelaide Bot. Gard. 27 (2014)

mm, stigma adaxially papillate. Fruit ellipsoid, 10—11 x 44.5 mm. Seeds c. 50-60, brown, 0.5—1 (—1.3) x 0.4—0.8 mm, irregularly lachrymoid to obovoid, + longitudinally flattened; funicle sulcate; testa minutely reticulate-foveolate, anticlinal cells walls sinuous. Alpine sundew. Figs. 3a—c, 4—7.

Etymology. The epithet arcturi comes from the Latinised form of the Ancient Greek ApKktovpoc (Arcturus), meaning “Guardian of the Bear’, and from which the name Arthur is sometimes thought to be derived; it 1s referring to the type locality of Mt Arthur, Tasmania.

Distribution and ecology. Victoria, New South Wales, Tasmania & New Zealand: Drosera arcturi 1s a relatively common herb of the alpine and sub-alpine regions of mainland Australia and Tasmania, as well as southern New Zealand (Fig. 10), extending to middle altitudes or even down to sea level in western and south-western Tasmania and southern New Zealand. It srows in sphagnum bogs on wet heathlands, as well as in black peat soils along the margins of small mountain streamlets and rivulets, or the boggy margins of lakes. It also frequently grows in cushion plants such as Donatia novae-zealandiae Hook.f. (Stylidiaceae) and Dracophyllum minimum F.Muell. (Ericaceae) (Cameron 1981) and sometimes grows intermixed with D. murfetii, with which it is sympatric over much of south- west Tasmania, but with no clear evidence of hybrids or intermediate plants.

Conservation status. Not currently threatened.

Flowering period. October—January (during the snow- free season).

Lectotypification of Drosera arcturi. William Hooker (1834) gives the type citation for D. arcturi as: “Summit of Mount Arthur Mr Gunn, (n. 139. [sic])”, but this appears to be an error, as J.D. Hooker (1860) cited the collection as “Gunn 129” in his F7. Zasmaniae. The type sheet at Kew holds two collections (Fig. 6), but as noted by Diels (1906), the lower sheet gathering with hand notation “summit of Mt Arthur, Van Dn’s [Diemen’s| Land, 129. W. Gunn 1832” (K 000215043) represents the type material of Drosera arcturi. Nevertheless, as no lectotype has been nominated previously, we accordingly here nominate the top middle specimen of that collection as the lectotype, with the other 5 duplicate specimens from that collection and the specimens on the duplicate at MEL regarded as residual syntypes. In contrast, the upper sheet gathering with hand notation “129 [over] 1842 Mt Wellington 1/3/39 [1 Mar. 1839] & 31/1/40 [31 Jan. 1840]” (K 000215085), represents material of D. arcturi from two separate, later gatherings from Mount Wellington and is not considered to be type material.

The gazetteer (Geoscience Australia 2013) lists three Mt Arthur locations in Tasmania:

|. Wellington Park, NE of Mt Wellington (42°53’6"S,

147°13’5”E), which overlooks the city of Hobart 2. Mt Arthur State Reserve, SW of Port Arthur near the

13

A new Drosera from Tasmania (Droseraceae)

original English convict penal settlement (43°9’14’S, 147°48°51”E)

3. Lilydale, NW of Launceston (41°16°44”S, 147°17’

17”E).

However, of these, Mt Arthur at Lilydale, north-west of Launceston, appears to be where Gunn’s original specimens of Drosera arcturi were collected in 1832. Buchanan (1988) noted that Gunn was a resident of Launceston, when he sent his first consignment of herbarium specimens to Hooker in 1832, and that these were gathered from east of Launceston, principally from Ben Nevis and Mt Arthur, including the type material for Drosera arcturi.

Robert Brown collected material of what is now D. arcturi (Bennett no. 4850: BM 001050170) from near the summit of ‘Table Mountain’ (now part of Mt Wellington) in 1804 and labelled it with the unpublished manuscript name “Drosera lingulata” (Chapman et al. 2001). Un- fortunately, this excellently preserved flowering material (Fig. 7) was not included in his Prodromus (Brown 1810), nor was it apparently seen by Hooker. Gibson (2010, p. 11) referred to this collection, saying “Robert Brown collected samples of the ‘Giant’ form of D. arcturi from Port Davey in March 1804, which were later described as D. ‘lingulata’.”’ However, the Brown collection at K is labelled “In paludosis summitatis Montis Tabularis prope fluvio Derwent Feb: 18 Mar 1804” [in a swamp at the summit of Table Mountain near the Derwent River] and Brown 1s not known to have visited Port Davey in Feb—Mar 1804, but was instead collecting in the region around Hobart (Vallance et al. 2001). This collection is also clearly conspecific with the type form of D. arcturi and was placed into that cluster in the analyses. As the epithet “/ingulata’ was never published validly it has no taxonomic standing, nor does the Brown collection represent type material.

Colenso (1890, 1896, 1899) described several taxa from New Zealand which are clearly part of the D. arcturi complex and examination of photographs of specimens of D. atra (WELT 23581), D. polyneura (K 000659184), and D. ruahinensis (K 000659178) shows that they all fall well within the variability seen for D. arcturi sens. str. to where they were reduced in synonymy by Cheeseman (1906). These specimens also fell into the D. arcturi clusters in the numerical analyses.

Specimens examined (specimens marked * were used in the numerical analysis)

NEw SouTtH WALEs: Found below Snowy Crossing and not far below Charlotte Pass Road, Kosciusko Nat. Park, 6 Feb. 1975, A.M. Ashby 5128 (AD 97529011); Near site of old Soil Conservation Hut near Blue Lake, 3 Feb. 1988, A.M Buchanan 10806 (*HO 129367); Near Bett’s Creek S of The Paralyser (c. 9 km ENE of Mt Kosciusko), 24 Jan. 1957, Hj.EKichler 13461 (*AD 95736072); Near Bett’s Creek S of The Paralyser (c. 9 km ENE of Mt Kosciusko), 24 Jan. 1957, Hj.Eichler 13472 (*AD 95736029).

TASMANIA: | km N of Reservoir Lakes, I1 Jan. 1984, D.G.Adams 7 (“HO 76687); Lake Salome, Walls of Jerusalem, 25 Jan. 1982, MJ Brown 117 (“HO 65857); In paludosis summitatis Montis Tabularis prope fluvio Derwent, 18 Feb.

A. Lowrie & J.G. Conran J. Adelaide Bot. Gard. 27 (2014)

Dresera arctuci Hoelk . | .

ROYAL BOTANIC GARDENS KEW eo Newker's Soucwal id Senay 12244 (1934) WLU Determinavit : K00027 5085

16/1) 198% | |

7

* ; ¢ ; } ‘ +

» tk + TTP Peet ee

_¢ = 4 i — — -

re . 6 * + . a >?

Om aas® a aoeeer” ** 2s 3® «*

ere) Ta tt 0 te

= ‘ i 4

yo 2

?~ —

r ~

gS 2's

S cw

= és —_

*

A Oo Oo > a oO o & ©) ° = ms o " a o =) © Ww

Determinavit: eum eneeme eee ekeBee SBE eeeeneansant*

Droséra arctur, Hook.

Hooker's Sounal af. Retan

4 *,

SOC ARUEE Cede rR aS, ee +

ROYAL BOTANIC GARDENS KEW @

. tae

WUWEMNNM “Koon 15043

.

Fig. 6. Type sheet of Drosera arcturi with two collections by Gunn. The lectotype (K 000215043, collected in 1832) is indicated by a thick black line;

residual syntypes by a dashed line. The upper collection (K 000215085, mixed collection from 1839 and 1840) is not type material. Image copyright of the Board of Trustees of the Royal Botanic Gardens, Kew; used with permission.

14

J. Adelaide Bot. Gard. 27 (2014) A new Drosera from Tasmania (Droseraceae)

TENA MSTORY DEPART en se ee CONDON Wh) ll] Hl.

M001050 169

—s - " - - - ; ; . * . , : y ' =f ; od , My , : P 4 Po = f _ _ «* 7 — 7 ’ / Ja ad a= A, - Y. oF Sa ae Ca faite i = : - ° . Y oe Farida Sree. 4 .

Tv?

Bi hb Sa . / | SS See — ——_ ! NATURAL 5 4 = 2. Oo J 1Q Hl st 0 RY copyright reserved MUSEUM

TURAL HISTORY MUSEVI LONDON HEA MEPARTMENT OF BOTANY

| i

M0010501 70

-

> :

Ly [ fre ter ; Kooi’

‘ as ;

Lycra ees pit (idee 4 - 2 4

A heticttee dunpmnthe a

i- ts belt C2PCE oe five » earrc, bit Mar liaty

2 ' , f }

Fig. 7. Robert Brown's 1804 collection of Drosera arcturi (Bennett no. 4850: BM 001050170) from near the summit of “Table Mountain” (now part of Mt Wellington) and labelled with the unpublished manuscript name “Drosera lingulata’. The upper collection on the sheet (BM 001050169) is possibly a duplicate specimen. Image copyright of the The Picture Library, The Natural History Museum, London; used with permission.

[5

A. Lowrie & J.G. Conran

1804, R.Brown 4850 (*BM 001050170, Fig. 7); Near Giblin River, 7 km from mouth, 23 Jan 1986, 4.\é/ Buchanan 8091 (HO 142845); Lake Ewart, 4 Feb. 1987, A.M. Buchanan 9895 (*HO 409074, mixed collection); SW Nat. Park, Middle Giblin River Basin, 8 km NW of Mt Gaffney, 13 km N of Mulcahy Bay, 17 Feb. 1989, J. R. Croft 10158 (*“HO 410760); Eagle Tarn, Mt Field Nat. Park, 2 Jan. 1948, WM. Curtis s.n. (“HO 3309); Mt Wellington, near Hobart, | Feb. 1947, WM. Curtis s.n. (“HO 42667); Cradle Mountain Nat. Park, Lake Dove, near boat shed, 7 Jan. 1960, Aj.Eichler 16465 (AD 96107068); Near Cradle Mt, Along track to Artist Pool, 9 Jan. 1960, G.Ford s.n. (“AD 98142217); Gordon River, 27 Nov. 1937, H.D.Gordon s.n. (HO 3322); Mt Wellington, 1 Mar. 1839 & 31 Jan. 1840, R.C.Gunn 129/1842 (*K 000215085, Fig. 6); Hartz Mt, 18 Dec. 1966, /.H. Hemsley 6072 (*HO 3317); Mt Wellington plateau S, 27 Jan. 1968, .H. Hemsley 6503 (*HO 3318); Upper Maxwell Valley, 15 Dec. 1978, S./.Jarman s.n. (*HO 30380); Surrounds of Lake Esperance, Hartz Mts, allt. 1000 m, 23 Jan. 2005, A.Lowrie 3105 & D.E.Murfet (AD, HO, MEL, NSW, PERTH); Near Ben Lomond alpine village, alt. 1425 m, 26 Jan. 2005, A.Lowrie 3135 & D.E.Murfet (*AD, HO, MEL, NSW, PERTH); On Belton & Belcher Lakes walk trail c. 500 m from road to Lake Dobson, Mt Field, alt. 1090 m, 22 Jan. 2006, A.Lowrie 3173 & D.E.Murfet (“AD, HO, MEL, NSW, PERTH); Saddle c. halfway between Mt Arthur & Mt Wellington summits, overlooking Hobart, alt. 1180 m, 23 Jan. 2006, A.Lowrie 3178 & D.E.Murfet (*AD, HO, MEL, NSW, PERTH); Start of the Hobartian Walk Trail on summit of Mt Arthur near road to Mt Wellington summit, alt. 1120 m, 23 Jan. 2006, A.Lowrie 3180 & D.E.Murfet (*AD, HO, MEL, NSW, PERTH); Near Lady Tarn, Hartz Mts, alt. 1000 m, 24 Jan. 2006, A.Lowrie 3185 & D.E.Murfet (*AD, HO, MEL, NSW, PERTH); Second Bar Lake, 11 Mar. 1984, A.Moscal 6885 (HO 401488); Julian Lakes, 29 Dec. 1984, A.Moscal 9113 (“HO 404030); Abbotts Lookout, 24 Mar. 1985, A.Moscal 10353 ("HO 143814); Hartz Mts Nat. Park, Lake Esperance area, 23 Jan. 2005, D.E. Murfet 4893 (“AD 178970, HO); Hartz Mts Nat. Park, Lake Esperance area, 23 Jan. 2005, D.E.Murfet 4594 (“AD 178971, HO); Mt Ben Lomond, near ski lodge area, 26 Jan. 2005, D.E. Murfet 4912 (SAD 178963, HO); Lake Dobson in the Mt Field Nat. Park, 22 Jan. 2006, D.E.Murfet 5184 (“AD 190697, HO); Lake Belcher/Belton walking trail, Mt Field Nat. Park, 22 Jan. 2006, D.E.Murfet 5185 (*AD 190655, HO); Mt Wellington on saddle on way up from Mt Arthur, alt. 1180 m, 23 Jan. 2006, D.E.Murfet 5190 (*AD 190792, HO); Mt Arthur, off Mt Wellington road near sharp bend, 23 Jan. 2006, D.E.Murfet 5191 (*AD 190791, HO); Ladies Tarn in the Hartz Mts Nat. Park, alt. 1000 m, 24 Jan. 2006, D.E.Murfet 5196 (“AD 190654, HO); Hartz Mts Nat. Park, Ladies Tarn, 27 Dec. 2007, D.E.Murfet 5846 (*AD 216591, HO); Mt Field Nat. Park, Lake Dobson, 28 Dec. 2007, D.E.Murfet 5851 (*AD 216748, HO); Pine Lake on the Highland Lakes Highway, 31 Dec. 2007, D.E.Murfet 5902 (“AD 216520, HO); Ben Lomond Nat. Park, Land of Little Sticks, 26 Mar. 1979, MG. Noble 28509 (*“HO 73420); Edge of Tarn, Wombat Moor, Mt Field Nat. Park, 26 Mar. 1932, O.Rodway 104 (*HO 3311); Upper reaches of North West Bay River, Mt Wellington, 6 Feb. 2000, A.C. Rozefelds 1626 (“HO 502869); Near Lake Dobson huts, Mt Field, 6 Jan. 1978, J. M.B.Smith 246 (“HO 36364); N end of Tarn Shelf, Mt Field Nat. Park, 21 Mar. 2006, M Visoiu 176 (*HO 539027, mixed collection); Lake Augusta, 13 Feb. 2009, M Visoiu 537, J.Balmer & M.Van Slageren (*HO 551187; Millennium Seedbank Voucher).

Victoria: Mt Baw Baw, NE of Ski Village, 22 Jan. 1966, A.C. Beauglehole 15272 (MEL 538741A); Bogong High Plains, Watchbed Creek, 27 Jan. 1966, A.C. Beauglehole 15629

16

J. Adelaide Bot. Gard. 27 (2014)

(MEL 538743A); Mt Cope, 21 Feb. 1982, E.A.Chesterfield 1609 (MEL 2130844A); Bogong High Plains, surroundings of Mt Nelse, alt. 1900 m, 13 Feb. 1958, M7. Eichler 14531 (“AD 96105210); Bogong High Plains, Rocky Knobs, alt. 1800 m, 3 Feb. 1958, AHj.Eichler s.n. (“AD 96105056); Alpine Nat. Park, Bogong High Plains, Head of Middle Creek, about 100 m N of start of track to Cope Hut, alt. 1700 m, 18 Mar. 2009, J.A.Jeanes 2148 (MEL 2325711A); Mt Mueller, Gippsland, alt. 1540 m, s.dat., J.G.Luehmann s.n. & C.H.French Jr (MEL 95997A); Bogong High Plains, Rocky Valley, alt. 1560 m, 20 Jan. 1959, 7°B.Muir 704 (*AD 96435006); Bridge crossing on walk to Derrick Hut, S of Mt Lock, 2 Jan. 2005, D.E.Murfet 4866 (*AD 178932, HO); Near Mt Cope on Bogong High Plains, 13 Jan. 2005, D.E.Murfet 4871 (“AD 178939, HO); Bogong High Plains, 20 Jan. 1936, R.7’Patton s.n. (MEL 96023A); Mt Erica, s.dat., G. Weindorfer s.n. (MEL 580029A); Bogong High Plains. (c. 130 km SE of Wangaratta), s.dat., J.H. Willis s.n. (MEL 96024A); Echo Flat, Lake Mt, near Marysville, alt. 1480 m, 6 Dec. 1943, JH. Willis s.n. (MEL 2217024A); Echo Flat, Lake Mt, c. 10 miles [16 km] NE of Marysville, alt. 1450 m, 25 Jan. 1948, JA. Willis s.n. (MEL 95999A); Baw Baw Mts, 2 Jan. 1905 A.B.Williamson s.n. (MEL 620164A).

New ZEALAND: Mt Sebastopol, Canterbury, 3 Feb. 1946, G.TS. Bayliss s.n. (OTA); Devils Elbow Stream, Mt Cook, Canterbury, 11 Feb. 1968, A.M Buchanan s.n. (HO 80813); Upper West Matakitaki River valley, 15 Jan. 1992, A.M.Buchanan 12263 (HO 538074); Near summit of Mt Whitecoomb, Umbrella Mts, 14 Dec. 1985, K...M. Dickinson & B.D.Race s.n. (*OTA); Little Pomohaka Headwater, Umbrella Ecology District, 6 Jan. 1986, K./.M Dickinson & B.D.Race s.n. (“OTA); Maungatua, Central Otago, 23 Mar. 1985, J. Eason s.n. (*“OTA); Lammermoors, 20 Dec. 1977, D. Holdsworth s.n. (“OTA); Maungatua, Central Otago, s.dat., J.E.Holloway s.n. (OTA); Lewis Pass, Canterbury, 4 Jan. 1968, PN.Johnson s.n. (“OTA); Red Mt, NW Otago, | Feb. 1975, W.G.Lee & A.F-Mark s.n. (“OTA); Wiarau Valley, Red Hills, 30 Nov. 1972, W.G.Lee s.n. (*OTA); Maungatua, Central Otago, s.dat., A.B.Lloyd s.n. (*“OTA); Maungatua, Central Otago, s.dat., A.F’ Mark s.n. (OTA); Blue Mts, 31 Dec. 1964, A.F Mark s.n. (©OTA); Red Creek Barrier U, Pyke River , 9 Dec. 1969, A.F! Mark s.n. ““OTA); Maungatua, Central Otago, Mar.1951, A.F!} Mark s.n. (OTA); Olivine Ledge, W. Otago, 17 Feb. 1968, A.F! Mark s.n. (*OTA); Mueller Valley, Turnbull River, Mt Aspiring Nat. Park, 21 Jan. 1969, A.FiMark s.n. (*OTA); Mueller Valley, Turnbull River, Mt Aspiring Nat. Park, 21 Jan. 1969, A.F’} Mark s.n. (*“OTA); Upper Makarora Valley, Young Range, Mt Aspiring Nat Park, Fiordland, 9 Feb. 1969, A.F! Mark s.n. (*OTA); Upper Aparima Valley, Takitimu Mts, N Southland, | Feb. 1971, A.FiMark s.n. (*OTA); Waitutu Forest, W Southland, 13.May 1985, A..F’ Mark s.n. (*“OTA); Lake Sylvester, NW Nelson, 17 Dec. 1967, A.F’ Mark & N.M.Adams s.n. (*OTA); Arthurs Pass, Arthurs Pass Nat. Park, 30 Dec. 1967, A.F’- Mark & N.M.Adams s.n. (*OTA); Borland Saddle, Hunter Mts, Fiordland, 7 Jan. 1968, 4.F’ Mark & N.M.Adams s.n. (“OTA); Key Summit, Fiordland, 10 Jan. 1968, A.F’Mark & N.M.Adams s.n. (*“OTA); Mt Richmond, Two Thumb Range, Canterbury, | Jan. 1969, A.F’Mark & N.M.Adams s.n. (*OTA); Williamson Flat, Joe — Arawhata River, Mt Aspiring Nat. Park, Fiordland, 16 Jan. 1968, A.F. Mark & M.L.Burke s.n. (*OTA); Lake Tehafer, 27 Dec. 1958, D. Scott s.n. “*OTA); Ajax Swamp, Catlins, SE Otago, 29 July 1975, A. Mark, PH.Johnson & G.T:S.Bayliss s.n. (*OTA); Secretary Island, Feb. 1959, J Murray 4463 (*OTA)); Gardiners Rd area, 16 Feb. 1980, J F'West s.n. (*OTA); Gardiners Rd area, Glendhu, 16 Feb. 1980, ,FiWest s.n. (“OTA);

J. Adelaide Bot. Gard. 27 (2014)

2. Drosera murfetii Lowrie & Conran, sp. nov.

A Drosera arcturi foliis viridis et rubis, foliis basalis

2—5, glabris, (1—) 3—5 (—12) longis, foliis distalis (0-)

l—2 (—4), glandulosis, 3.5—21.7 cm longis, styliis 4-5;

viridi-albis ad purpureo-ceriseis stigmata, semina

(0.8—) 1-2 x (0.4) 0.6—I1 mm, differt.

Typus: shores of Lake Esperance, Hartz Mountains,

Tasmania, 43°13’44’S 146°467°14”E, alt. 1000 m, A.

Lowrie 3181 & D.E. Murfet, 24 Jan 2006. — Holotypus:

HO. Isotypi: PERTH; MEL, AD, NSW, distribuendi.

Drosera arcturi auct. non Hook.: Marchant, FI. Austral. 8:

54 (1982), pro parte; Lowrie, Carniv. Pl. Austral. 136 (1998), pro parte; Morris, Fl. Tas. Online, Droseraceae 3 (2009), pro parte.

A cold tolerant, green, turning red with age or ex- posure, fibrous-rooted alpine perennial herb; rhizome 1-5 (©10) cm long, covered with persistent dead leaf bases. Hibernaculum bud-like, derived from non- elandular basal leaf sheaths covering rhizome apex. Leaves at emergence 2—5, (1—) 3-5 (—12) x 0.5—1.5 cm, lanceolate, apically acute, glabrous; glandular leaves (OQ—) 1—2 (-4), linear to narrowly lanceolate-spathulate, loosely distichous, erect, 3.5—21.7 cm long, persisting after senescence; petioles stem-sheathing, 1.5—2.5 cm long, 0.8—1 mm wide at sheath apex, 1.5—2 mm wide at lamina, glabrous; lamina 22—55 (>75) x 2-6 mm, more or less linear to lanceolate or narrowly oblong, apex blunt to rounded, adaxial surface bearing large, sub-marginal insect-catching glands and smaller glands within, abaxial surface glabrous with a prominent midrib (occasional plants bear no trapping leaves at all). Inflorescence terminal, scapose, 1—4-flowered, 3-20 cm long, 0.5—0.6 mm diam, glabrous, enclosed basally by leaf sheaths, persistent, prostrate after senescence. Flowers shortly pedicellate, erect in fruit; pedicels 1.5- 12 mm long; floral bract subulate, 1-2 x 0.4—0.5 mm. Sepals greyish-green, 10-16 <x 2.3—2.5 mm, narrowly elliptic, adaxially concave, horizontal at anthesis, margins entire, apex obtuse, irregularly crenate, both surfaces glabrous. Petals white, 5-14 x 4-5 mm, obovate, entire, apex broadly rounded. Stamens 5, 6—7 mm long, filaments white, anthers and pollen yellow. Ovary 7-8.5 x 4.2—5.1 mm, ellipsoid, green, ovules numerous. Styles 4—5, white, basally black, 0.8-2 x 0.2—-0.3 mm (excluding stigmas), semi-erect, fusiform, flat in cross-section; stigmas greenish-white throughout most of its range or reddish-cerise (Hartz Mountains), + reniform-peltate, 1.5-—2.5 x 0Q.5—0.6 mm, stigma adaxially papillate. Fruit ellipsoid, 7-11 x 4.5—7.7 mm. Seeds c. 80—90, dark brown, (0.8—) 1-2 x (0.4—) 0.6—1 mm, irregularly lachrymoid to obovoid, + longitudinally flattened; funicle sulcate; testa minutely reticulate- foveolate, anticlinal cells walls sinuous. Giant alpine sundew. Figs. 3d-f, 8, 9.

Etymology. The species is named in honour of Denzel Edwin Murfet (1957—), communications technician, amateur botanist, carnivorous plant, Stylidium and orchid enthusiast, who collected the species in the Hartz Range, Tasmania.

17

A new Drosera from Tasmania (Droseraceae)

Bhd balts ize LIL Ee

Fig. 8. Drosera murfetii. A plant; B lamina; C sepal; D petal; Egynoecium; F style and stigma, posterior view; G style and stigma, anterior view. scale bars: A 10 mm; B-D 5 mm; E-G 1 mm. — Drawn by A. Lowrie, 2007, from type material (A.Lowrie 3187 & D.E.Murfet).

Distribution and ecology. Widespread from sea-level to alpine areas of western and south-western Tasmania (Fig. 11) in wet heathlands, near the shores of tarns, and at higher altitudes, commonly grows in cushion plants (Donatia novae-zealandiae and Dracophyllum minimum), often alongside Drosera arcturi.

Conservation status. Drosera murfetii throughout its range 1s common and not threatened.

Flowering period. November—January, with a peak flowering period for D. murfetii observed by AL and DEM at the end of Dec. 2007 (during the snow-free season).

Distinguishing characters. The morphologically most similar species to D. murfetii 1s D. arcturi, whose dis- tinguishing characters are given in parenthesis. Drosera murfetii 1s distinguished by having: green and red foliage (green to greenish-bronze, red or almost black) conspicuous, glabrous basal leaves, usually (1—) 3-5 (—12) cm long above the soil surface (glabrous basal leaves absent or small, linear, mostly 3—12 mm long, but occasionally up to 27 mm); glandular leaves commonly 1, sometimes 2, rarely up to 4, usually 3.5—8 cm long, but sometimes up to 21.7 cm long or longer (glandular leaves 3-12, 3-5 cm long, but sometimes up to 12 cm

A. Lowrie & J.G. Conran

> ,

- » + . A ; ~ ? : ' : = ; ‘ . ; ; .

' “ i ; Fs . i, \* ° 2

. : k

| ~ ~ BD a Pa ' P .

" a a f ba . “

a .

_

* ” = a . i, y : L s ‘ =. a ra r rae , ; . . af ; - ' _ ~ : - > * e ' ° , —_—— .

Fig. 9. Drosera murfetii in the Hartz Mountains, Tasmania. A plants showing broad glabrous basal leaves and few, erect trapping leaves; B close up of flower with white stigmas; C plant showing multiple flowers per scape; D close up of cerise stigma form. Scale bars: A, C 10 mm; B, D5 mm. — Photos: A. Lowrie.

=

long); styles 4—5 (styles 3-4); seeds (0.8—) 1—2 x (0.4—) 0.6—1 mm (seeds 0.5—1 (—1.3) x 0.4—0.8 mm).

Notes. Throughout its range, D. murfetii is more common than D. arcturi at lower elevations and particularly at or near sea level in south west Tasmania. Drosera

J. Adelaide Bot. Gard. 27 (2014)

> . f \ A , ’ _ i - ra - 4, iA > : 4 a _ —— -/ , / ’ | ’ 5) FS « *f ‘

18

.*

murfetii coexists with D. arcturi at higher elevations, where it commonly grows side by side with D. arcturi, sharing the same cushion plants. Despite extensive field explorations carried out by AL and DEM in Jan. 2005, Jan. 2006 and Dec. 2007 at several sites 1n south-west Tasmania, no intermediate forms were found in terms of

J. Adelaide Bot. Gard. 27 (2014)

the possession of large glabrous leaves and few, erect (versus numerous, semi-erect) trapping leaves.

Specimens examined (specimens marked * were used in the numerical analysis)

TASMANIA: Moores Garden, 14 Jan. 1984, D.G.Adams 41 (*HO 76650); 1 km NE of the W-most causeway on the S side of Lake Augusta, 5 Jan. 1971, W.R.Barker 1030 (FAD 97114123); Ooze Lake, 31 Jan. 1981, A.Brown 221 (*HO 327549); Head of D’Entrecasteaux River, 22 Mar. 1984, A.M. Buchanan 3039 (*HO 76816); Jubilee Range, 13 Jan. 1985, A.M. Buchanan 5241 (*HO 120685); 2 km W of Granite Tor, 22 Jan. 1985, A.M. Buchanan 5469 (*HO 122115); Giblin River c. 7 km from mouth, 11 Jan. 1986, A.M. Buchanan 7804 (*HO 405459); S ridge of Mt Gaffney, 14 Jan. 1986, A.M. Buchanan 7862 (“HO 405516); N peak of Mt Gaffney, 14 Jan. 1986, A.M. Buchanan 7880 (*“HO 405532); SE ridge of the Lawson Range, 25 Jan. 1986, A.M. Buchanan 8107 (“HO 404985); Low hills behind Coffin Bay, Port Davey, 5 Jan. 1987, A.M. Buchanan 9244 ("HO 405370); Ridge E of Deadmans Creek, 21 Jan. 1987, A.M. Buchanan 9761 ("HO 402904); Lake Ewart, 04 Feb. 1987, A.M Buchanan 9895 (HO 409074, mixed collection); Between Deadmans Bay and Lousy Bay, 19 Jan. 1987, A.M. Buchanan 9638 (*HO 123668); Schnells Ridge, 27 Jan. 1998, A.M. Buchanan 15037 (HO 324160); Mt Read S side of summit, 21 Feb. 2008, 4. Buchanan 16916 (HO 548612); Lake Dobson, Mt Field Nat. Park, 22 Jan. 1949, N.T-Burbidge 3259 (HO 3310); Darwin Plateau, 25 Dec. 1984, P-Collier 220 (*HO 116495); Mt Field Nat. Park (Lake Belcher track, beyond saddle), 2 Feb. 1969, E.M Canning s.n. (“AD 97026096); Mt Solitary summit, 25 Mar. 1990, P.Collier 4641 ("HO 142923); Lake Dobson, Mt Field Nat. Park, 23 Jan. 1944, WMCurtis s.n. (‘HO 42666); Near Cradle Mt, along track to Artist Pool, 9 Jan. 1960, G.Ford s.n. (*AD 98204400); Denison Range, Reeds Peak area, 30 Nov. 1978, C.Harwood s.n. (“HO 29560); Hamilton Range, 29 Jan. 1977, S.J. Jarman s.n. (HO 411337); Propsting Range, 16 Feb. 1977, S.J Jarman s.n. “(HO 410974); Mt Humboldt, 17 Jan. 1978, S.J. Jarman s.n. (“HO 411973); Mt Rugby, 16 Feb. 1978, S.J Jarman s.n. (HO 411018); Mt Lee, 3 Mar. 1978, S.J. Jarman s.n. (“HO 411087); Upper tributaries of the Giblin River, E of the Lawson Range, 31 Dec. 1978, S./.Jarman s.n. (*“HO 30472); Moth Creek, Bathurst Harbour, 31 Jan. 1962, H.J.King s.n. ("HO 3315); Melaleuca Creek, 31 Jan. 1929, F°'H.Long s.n. (HO 3323); Shores and surrounds of Lady Tarn, Hartz Mts, 27.Dec. 2007, A.Lowrie 3740 & D.E.Murfet (*AD, MEL, NSW, PERTH); Mt Lyell camp, Julia Creek, 8 Dec. 1980, M.K.Macphail s.n. (“HO 37056); Hamilton Moraine, Basin Lake, 9 Dec. 1980, M.K.Macphail s.n. (“HO 37115); Jubilee Range, 8 Jan. 1985, A.Moscal 9187 (HO 97042); Jubilee Range, 13 Jan. 1985, A.Moscal 9255 (*HO 400964); Jubilee Range, 14 Jan. 1985, A.Moscal 9333 ("HO 95532); Jubilee Range, 16 Jan. 1985, A.Moscal 9372 (“HO 95146); Mulcahy River plain, 13 Jan. 1986, A.Moscal 11631 (*HO 144694); Lawson Range, 25 Jan. 1986, A.Moscal 11944 (*HO 402114); Wilson Bight, 13 Jan. 1987, A.Moscal 13935 (*HO 409828); Hartz Mts Nat. Park, Lake Esperance area, 23 Jan. 2005, D.E.Murfet 4894 & A.Lowrie (*AD 178971, HO); Lake Dobson in the Mt Field Nat. Park, 22 Jan. 2006, D.E.Murfet 5183 & A.Lowrie (“AD 190696, HO); 200 m from Ladies Tarn in the Hartz Mts Nat. Park, 24 Jan. 2006, D.E.Murfet 5192 & A.Lowrie (*AD 190794, HO); Ladies Tarn in the Hartz Mts Nat. Park, 24 Jan. 2006, D.E.Murfet 5195 & A.Lowrie (*AD 190793, HO); Anthony Rd 8.3 km N of Zeehan Highway, 27 Jan. 2006, D.E.Murfet 5208 & A.Lowrie (*AD 190708, HO); Hartz Mts Nat. Park, Lake Esperance, 27 Dec. 2007, D.E.Murfet 5843 & A.Lowrie (*AD 216534, HO);

19

A new Drosera from Tasmania (Droseraceae)

Hartz Mts Nat. Park, Ladies Tarn, 27 Dec. 2007, D.E.Murfet 5844 & A.Lowrie (*“AD 216536, HO); Mt Field Nat. Park, Lake Dobson, 28 Dec. 2007, D.E.Murfet 5850 & A.Lowrie (*AD 216513, HO); Queenstown, along Anthony Rd toward Tullah, 29 Dec. 2007, D.E.Murfet 5867 & A.Lowrie (*AD 216515, HO); Hartz Mts, 29 Jan. 1938, A.M Olsen s.n. (*HO 411530); Hartz Mts, 22 Jan. 1939, A.M. Olsen s.n. (“HO 312393); Mt Wellington, 31 Dec. 1892, L.Rodway 190 (*HO 3316); Twelvetrees Range, 12 Jan. 1980, D. Jan Morris 8038 (HO 32596); Dove Lake, 31 Dec. 1908, L.Rodway 231 (*HO 3312); Near Arve River, below hut, Hartz Mts, no date, J. Somerville s.n. (*HO 3313); Moth Creek, Bathurst Harbour, 31 Jan. 1962, S.Stanier s.n. (“HO 3314); N end of Tarn Shelf, Mt Field Nat. Park, 21 Mar. 2006, M Visoiu 176 ("HO 539027, mixed collection); Sharlands Peak, Frenchmans Cap Nat. Park, 2 Jan. 1981, TJ. Wardlaw s.n. ("HO 541098); Charles Range, summit of Southern Peak, SW aspect, 17 Feb. 1987, J.M. Wells (?*HO 315576).

Discussion

Drosera arcturi occurs in the alpine regions of Victoria and New South Wales on the mainland of Australia, as well as Tasmania and New Zealand. The species is placed in subgenus Arcturia (Planch.) Schlauer [= Drosera section Psychophila auct. non. Planch. sensu Diels (1906) and Marchant et al. (1982)], along with D. murfetii. The New Zealand endemic Drosera stenopetala J.D.Hook and D. uniflora Willd. from southern South America are also traditionally included here (Diels 1906), but Culham (1993) found that pollen of D. arcturi was distinct from D. uniflora. This led Schlauer (1996) to transfer the latter to subgen. Ptycnostigma (Planch.) Drude, albeit with reservations on its placement. Although molecular studies also widely separated these two species, D. stenopetala was found to be sister to D. uniflora in an isolated lineage inside Drosera L. subgen. Drosera, rather than close to D. arcturi (Rivadavia et al. 2003, Yesson & Culham 2006). The latter 1s instead sister to most of the genus, with only the isolated South African D. regia Stephens sitting below it in the phylogenetic tree of Cameron et al. (2002).

Species concepts and taxon variability in Droseraceae vary considerably, with many taxa that are currently recognised at specific level within Australia (CHAH 2008—) reduced 1n rank or placed into synonymy by some overseas researchers (Schlauer 1996, 2006—), largely due to differences in opinion on species definitions. Nevertheless, recent research on Drosera has found that there are generally more species than were thought previously within so-called species complexes, with the trend being to recognise more rather than fewer entities, usually as a result of detailed multi-character comparisons (e.g. Fleischmann et al. 2008, 2011; Riva- davia et al. 2009; Gonella et al. 2014).

Cluster analysis and ordination is a standard method for the determination of morphometric variability within and between plant taxa (e.g. Stace 1989; Quinn & Keogh 2002) and the technique was applied recently to the polymorphic Drosera peltata Thunb. complex (Gibson

A. Lowrie & J.G. Conran

SE Australia

Melbourne 6 §

on 9 Iasmania

Drosera arcturi

Fig. 10. Distribution map of Drosera arcturi. Map derived from AVH and specimens examined at AD, HO, MEL, OTA, and PERTH.

et al. 2012). As with the present study, the use of multi- character comparisons including morphometrics, found that there were clear differences between the different entities in the complex, allowing the determination of exactly how many taxa there are for the group.

When applied to D. arcturi, this approach shows that there are two discrete entities which differ in a range of features. Although D. murfetii is statistically significantly larger for nearly all measured features, there is still wide variability and overlap with D. arcturi for most of the size-related characteristics. Similarly, although the majority of D. arcturi specimens examined had three stigmas, while those of D. murfetii mostly had five, there were still plants in both taxa which possessed four.

The most obvious difference between the two species relates to the possession of broad, glabrous non- trapping leaves. These were completely absent in D. arcturi, where, if there were still glabrous leaves present at anthesis, they were tiny linear structures, the leaves instead all being typical, mostly semi-erect to spreading Drosera glandular trapping leaves. In contrast, all the D. murfetii plants examined bore broad, obviously photosynthetic outer leaves with only 1—2 larger, erect, inner gland-bearing leaves, or in the case of ten plants examined, apparently no trapping leaves at all. These features, combined with the overlapping distributions of the plants in south-western Tasmania and absence in the field of any obvious intermediate plants for these features, even when both taxa are growing in mixed populations, support our conclusion that they represent discrete taxa at species level.

Acknowledgements

The Directors of herbaria AD, HO, MEL, OTA and PERTH are thanked for access to specimens, as are various Australian Botanical Liaison Officers at K. The Board of Trustees of the Royal Botanic Gardens, Kew, and The Picture Library, The Natural History Museum, London, are thanked for providing images of Gunn’s and Brown’s collections. The School of Earth & Environmental Sciences is thanked for provision of facilities for some of this research.

20

J. Adelaide Bot. Gard. 27 (2014)

41

-42

-43

Drosera murfetii

-44

143.9 144.5 145.9

146.5

147.5 148.5

Fig. 11. Distribution map of D. murfetii. Map derived from AVH and specimens examined at AD and HO.

References

Allan Herbarium (2000). Ned Tipu o Aotearoa—New Zealand plant names database. (Landcare Research, New Zealand: Lincoln). http://nzflora.landcareresearch.co.nz/ [accessed: 11 Dec. 2011].

Belbin, L. & Collins, A. (2008). PATN version 3.1.2. (Blatant Publications Pty Ltd: Brisbane).

Boesewinkel, F.D. (1989). Ovule and seed development in Droseraceae. Acta Botanica Neerlandica 38: 295-311. Brown, R. (1810). Prodromus flora Nova-Hollandiae et insulae Van Diemen. (Richard Taylor & Son: London). Buchanan, A.M. (1988). The Tasmanian collecting localities of Ronald Gunn and Joseph Milligan. (Tasmanian Herbarium: Hobart). [Zasmanian Herbarium Occasional

Publication No. 1]

Cameron, K.M., Wurdack, K.J. & Jobson, R.W. (2002). Molecular evidence for the common origin of snap-traps among carnivorous plants. American Journal of Botany 89: 1503-1509.

Cameron, M. (1981). A guide to flowers & plants of Tasmania. (Reed: Chatswood, N.S.W.).

CHAH (2008-). Australian Plant Census—APC, IBIS data- base. (Council of Heads of Australasian Herbaria Inc.: Canberra). http://chah.gov.au/chah/apc/ [accessed: 28 Jan. 2014].

CHAH (2013). Australias Virtual Herbarium—AVH. (Council of Heads of Australasian Herbaria Inc.: Canberra). http:// avh.ala.org.au/ [accessed: 28 Jan. 2013].

Chapman, A.R., Moore, D.T., Rees, R.G. & Groves, E.W. (2001). Robert Brown’ Australian botanical specimens, 1801-1805 at the BM: Introductory notes to the botanical collection made by Robert Brown (1773-1858) in Australia, 1801-1805. (Western Australian Herbarium: Perth; The Natural History Museum: London). http:// florabase.calm.wa.gov.au/brown/#fields [accessed: 12 Jul. 2011].

Cheeseman, T.F. (1906). Manual of the New Zealand flora. (John Mackay, Government Printer: Wellington).

Clayton, C. (2003). Carnivorous plants in Tasmania, Australia. A field guide and cultural notes to the indigenous species. (Published by the author: Somerville, Vic.).

Colenso, W. (1890). A description of a some newly-discovered phaenogamic plants, being a further contribution to-

J. Adelaide Bot. Gard. 27 (2014)

wards the making-known the botany of New Zealand. Transactions of the New Zealand Institute 22: 459-493.

Colenso, W. (1896). Phaenogams: a description of a few more newly discovered indigenous plants; being a further contribution towards the making known the botany of New Zealand. Transactions of the New Zealand Institute 28: 591-613.

Colenso, W. (1899). Phaenogams: a description of a few more newly discovered indigenous plants; being a further contribution towards the making known the botany of New Zealand. 7ransactions of the New Zealand Institute 31: 266-281.

Culham, A. (1993). A taxonomic revision of sectional and generic delimitations in the Droseraceae: evidence from pollen morphology, seed morphology and chloroplast DNA intergenic spacers. American Journal of Botany 80 (suppl.): 142.

Diels, L. (1906). Droseraceae. In: Engler, A. (ed.), Das Pflan- zenreich 4(112) [Heft 26]: 1-136. (Engelmann: Leipzig).

Dwyer, T.P. (1983). Seed structure in carnivorous plants. Carnivorous Plant Newsletter 12: 8-22, 33-38, 45—50.

Faith, D.P. (1991). Effective pattern analysis methods for nature conservation. In: Margules, C.R. & Austin, M.P. (eds), Nature conservation: cost effective biological surveys and data analysis, pp. 47—53. (CSIRO Publishing: Melbourne).

Fleischmann, A., Gibson, R. & Rivadavia, F. (2008). Drosera ericgreenii (Droseraceae), a new species from the fynbos of South Africa. Bothalia 38: 141-144.

Fleischmann, A., Robinson, A.S., McPherson, S., Heinrich, V., Gironella, E. & Madulid, D.A. (2011). Drosera ultramafica (Droseraceae), a new sundew species of the ultramafic flora of the Malesian highlands. Blumea 56: 10—15.

Geoscience Australia (2013). Gazetteer of Australia, 2012 release. (Geoscience Australia: Canberra). http://www.ga. gov.au/place-name/ [accessed: 18 Mar. 2014].

Gibson, R.P. (1998). Observations on a selection of Tasmanian carnivorous plants. Carnivorous Plant Newsletter 27: 90- 92.

Gibson, R.P. (1999). Drosera arcturi in Tasmania and a com- parison with Drosera regia. Carnivorous Plant Newsletter 28: 76-80.

Gibson, R.P. (2010). Highlights from a late summer trip to Tasmania in 2010. Carniflora Australis 7(4): 8-21.

Gibson, R.P., Conn, B.J. & Bruhl, J.J. (2012). Morphological evaluation of the Drosera peltata complex (Droseraceae). Australian Systematic Botany 25: 49-80.

Gonella, P.M., Rivadavia, F., Sano, P.T. & Fleischmann, A. (2014). Exhuming Saint-Hiulaire: revision of the Drosera villosa complex (Droseraceae) supports 200 year-old neglected species concepts. Phytotaxa 156: 1-40.

Hooker, J.D. (1860). The botany of the Antarctic voyage of HM. discovery ships ‘Erebus’ and ‘Terror’ in the years 1839-1843 under the command of Captain Sir James Clark Ross. Vol. 3. Flora Tasmaniae. Part 1. Dicotyledones. (Lo- vell Reeve and Co.: London).

Hooker, W.J. (1834). Contributions towards a flora of van Dieman’s Land; from collections sent by R. W. Lawrence, and Ronald Gunn, Esqrs., and by Dr. Scott. Journal of Botany (Hooker) 1: 241-258.

ZA

A new Drosera from Tasmania (Droseraceae)

Lowrie, A. (1998). Carnivorous plants of Australia, Vol. 3. (University of Western Australia Press: Perth).

Lowrie, A. (2005). A taxonomic revision of Drosera section Stolonifera. Nuytsia 15: 355-393.

Lowrie, A. & Conran, J.G. (2007). A revision of the Drosera omissa / D. nitidula complex (Droseraceae) from south- west Western Australia. Taxon 56: 533-544.

Lowrie, A. & Conran, J.G. (2008). A review of Drosera whittakeri s. lat. (Droseraceae) and description of a new species from Kangaroo Island, South Australia. Telopea 12: 147-165.

Marchant, N.G., Aston, H.I. & George, A.S. (1982). Dro- seraceae. In: George, A.S. (ed.), Flora of Australia 8: 9-66. (Australian Government Publishing Service: Canberra).

Morris, D.I. (2009). 96 Droseraceae. In: Duretto, M.F. (ed.), Flora of Tasmania online [ISBN 978-1-921599-31-6 (PDF) /. Version 2009.1: 1-6. (Tasmanian Museum & Art Gallery: Hobart).

Quinn, G.P. & Keough, M.J. (2002). Experimental design and data analysis for biologists. (Cambridge University Press: Cambridge).

Rivadavia, F., Kondo, K., Kato, M. & Hasebe, M. (2003). Phylogeny of the sundews, Drosera (Droseraceae), based on chloroplast rbcL and nuclear 18S ribosomal DNA sequences. American Journal of Botany 90: 123-130.

Rivadavia, F., Vicentin1, A. & Fleischmann, A. (2009). A new species of sundew (Drosera, Droseraceae), with water- dispersed seed, from the floodplains of the Northern Amazon Basin, Brazil. Ecotropica 15: 13-21.

Salmon, B. (2001). Carnivorous plants of New Zealand. (Ecosphere Publications: Auckland, NZ).

SAS Institute (2000). JMP IN Version 4.0.3 (Academic). (SAS Institute Inc.: Cary, NC).

Schlauer, J. (1996). A dichotomous key to the genus Drosera L. (Droseraceae). Carnivorous Plant Newsletter 25: 67-

Schlauer, J. (2006—). World carnivorous plant list: nomenclatu- ral synopsis of carnivorous phanerogamous plants. (Last update 27 Jan. 2012). http://www.omnisterra.com/botany/ cp/data/jans.db [accessed: Feb. 2014].

Stace, C.A. (1989). Plant taxonomy and biogeography, ed. 2. (Edward Arnold: London).

Susandarini, R., Collins, G.G., Lowrie, A. & Conran, J.G. (2002). Morphological variation within the Drosera indica (Droseraceae) complex in northern Australia. Australian Journal of Botany 50: 207-214.

Vallance, T.G., Moore, D.T., Groves, E.W. & Maslin, B.R. (2001). Natures investigator: The diary of Robert Brown in Australia, 1801-1805. (Australian Biological Resources Study: Canberra).

Webb, C.J. & Simpson, M.J.A. (2001). Seeds of New Zealand gymnosperms and dicotyledons. (Manuka Press: Christ- church).

Yesson, C. & Culham, A. (2006). Phyloclimatic modeling: combining phylogenetics and bioclimatic modeling. Systematic Biology 55: 785-802.

Zar, J.H. (1996). Biostatistical analysis, ed. 3. (Prentice Hall: Englewood Cliffs, New Jersey).

J OQURNAL of the ADELAIDE

BOTANIC GARDENS

AN OPEN ACCESS JOURNAL FOR AUSTRALIAN SYSTEMATIC BOTANY

flora.sa.gov.au/jabg

Published by the STATE HERBARIUM OF SOUTH AUSTRALIA on behalf of the BOARD OF THE BOTANIC GARDENS AND STATE HERBARIUM

State Herbarium of South Australia PO Box 2732 Kent Town SA 5071 Australia

Board of the Botanic Gardens and Reit State Herbarium

of South Australia Department of Environment, Water and Natural Resources

Journal of the Adelaide Botanic Gardens 27 (2014) 23-24

New combinations for the Phillip Island wheat grass, Anthosachne kingiana subsp. kingiana (Poaceae)

R. Govaerts

Royal Botanic Gardens, Kew, Herbarium Building, Richmond TW9 3AE, United Kingdom E-mail: R.Govaerts@kew.org

Abstract

Phillip Island wheat grass has been known under many different scientific names. The taxon is now mostly placed in the genus Anthosachne Steud., following recent molecular phylogenetic data. Unfortunately no correct name 1s currently available under that genus which 1s rectified here. The new combinations Anthosachne kingiana (Endl.) Govaerts and Anthosachne kingiana subsp. multiflora (Banks & Sol. ex Hook.f.) Govaerts are made.

Key words: nomenclature, new name, grass, Poaceae, Anothosachne, Phillip Island,

Norfolk Island, Lord Howe Island, Australia.

Introduction

Since the early 1990’s I have maintained a database on plants recorded as being extinct. The entire database is not currently online, though the data feed into the World checklist of selected plant families (WCSP 2014). Species recorded as being extinct are indicated with a dagger (1) after the relevant geographical code. When transferring the Phillip Island wheat grass from the genus Elymus L. to Anthosachne Steud., I noticed a nomenclatural problem, which I resolve in this paper.

The Phillip Island wheat grass is a rare species listed as critically endangered in the Australian legislation protecting rare and endangered plants and animals under the name Elymus multiflorus subsp. kingianus (Endl.) de Lange & R.O.Gardner (Department of the Environment 2014). Until recently, 1t was considered to be endemic to Phillip Island, off Norfolk Island, and was thought to be extinct together with the two other endemics, Hibiscus insularis Endl. and Streblorrhiza speciosa Endl., as nearly all the vegetation on the islet had been eaten by goats, pigs and rabbits. When these feral animals were removed, Phillip Island wheat grass was able to recolonize the area, and was rediscovered on Phillip Island in 1987 (Sykes & Atkinson 1988). The species has now also been found on Norfolk Island and Lord Howe Island (Green 1994: 469) and therefore can no longer be considered as being endemic to Phillip Island.

After the removal of feral animals, the near-endemic Abutilon julianae Endl. also reappeared (Green 1994). This species was first described from Norfolk Island, but seems to have become extinct there a long time ago and can now only be found on Phillip Island. Another, previously unknown species, Achyranthes margaretarum de Lange (2001), was discovered recently, adding a third endemic species to the island.

Published online: 29 May 2014 « flora.sa.gov.au/jabg

At the species level Elymus multiflorus (Banks & Sol. ex Hook.f.) A.Léve & Connor is found in coastal eastern Australia (Connor 1990; de Lange et al. 2005). So it seems that the only true endemics to Phillip Island are Achyranthes margaretarum, Hibiscus insularis and Streblorrhiza speciosa, of which the last two are still considered to be extinct (Green 1994).

Discussion

The Phillip Island wheat grass was previously classi- fied within the genus E/ymus. Recent morphological and molecular studies (Barkworth & Jacobs 2011), however, have shown this taxon to be part of an Australasian eroup, distinct from Elymus, and the name Anthosachne has been resurrected to accommodate this group (Bark- worth & Jacobs 2011).

When updating my personal database on plants recorded as being extinct, I noticed a problem with the newly published combination, Anthosachne multiflora (Banks & Sol. ex Hook.f.) C.Yen & J.L.Yang subsp. kingiana (Endl.) Barkworth & S.W.L.Jacobs. It seems that the basionym of the species name, dating from 1853, is later than the basionym of the infraspecific name, which has priority from 1833, thus making the subspecies superfluous and illegitimate (ICN Art. 52.1; McNeill et al. 2012). This error seems to have started when Connor published the name Elymus multiflorus (Banks & Sol. ex Hook.f.) A.Léve & Connor var. kingianus (Endl.) Connor (Connor 1990), rather than making the combination under Elymus kingianus, and was perpetuated thereafter. This is hereby corrected.

Nomenclature

Anthosachne kingiana (Endl.) Govaerts, comb. nov.

Triticum kingianum Endl., Prodr. Fl. Norfolk. 21 (1833). — Festuca kingiana (Endl.) Steud., Syn. Pl. Glumac. 1:

ISSN 0313-4083 (Print) * ISSN 2201-9855 (Online)

R. Govaerts

316 (1854). — Agropyron kingianum (Endl.) Petrie ex Laing, Trans. & Proc. New Zealand Inst. 47: 18 (1915). — Elymus kingianus (Endl.) A.L6ve, Feddes Repert. 95: 469 (1984). — Elymus multiflorus var. kingianus (Endl.) Connor, Kew Bull. 45: 680 (1990), nom. superfl. — Elymus multiflorus subsp. kingianus (Endl.) de Lange & R.O.Gardner, New Zealand J. Bot. 43: 571 (2005), nom. superfl. — Anthosachne multiflora subsp. kingiana (Endl.) Barkworth & S.W.L.Jacobs, Telopea 13: 50 (2011), nom. superfl.

Anthosachne kingiana (Endl.) Govaerts subsp. Kingiana

This is the correct name for the subspecies from Norfolk, Phillip and Lord Howe Islands, commonly known as the Phillip Island wheat grass.

Anthosachne kingiana subsp. multiflora (Banks & Sol. ex Hook.f.) Govaerts, comb. nov.

Triticum multiflorum Banks & Sol. ex Hook.f., Bot. Antarct. Voy. II (Fl. Nov.-Zel.) 1: 311 (1853). — Agro- pyron multiflorum (Banks & Sol. ex Hook.f.) Kirk ex Hack. in Cheeseman, Man. New Zealand FI.: 921 (1906). — Agropyron kirkii Zotov, Trans. & Proc. Roy. Soc. New Zealand 73: 233 (1943), nom. superfl. — Elymus multiflorus (Banks & Sol. ex Hook.f.) A.L6ve & Connor, New Zealand J. Bot. 20: 183 (1982). — Anthosachne multiflora (Banks & Sol. ex Hook.f.) C.Yen & J.L.Yang, Xiao mai zu sheng wu xi tong xue 3: 232 (2006).

This is the correct name for the subspecies from south-eastern Australia and New Zealand, commonly known as the short-awned wheat grass.

24

J. Adelaide Bot. Gard. 27 (2014)

References

Barkworth, M.E. & Jacobs, S.W.L. (2011). The Triticeae (Gramineae) in Australasia. Telopea 13(1—2): 37-56.

Connor, H.E. (1990). Elymus (Gramineae) on Norfolk Island. Kew Bulletin 45: 680.

de Lange, P.J. & Murray, B.G. (2001). A new Achyranthes (Amaranthaceae) from Phillip Island, Norfolk Island eroup, South Pacific Ocean. New Zealand Journal of Botany 39: 1|-8.

Department of the Environment (2014). Elymus multiflorus subsp. kingianus. In: Species profile and threats database. (Department of the Environment: Canberra). http://www. environment. gov.au/cgi-bin/sprat/public/publicspecies. pl?taxon 1d=82413 [Accessed: 8 Mar. 2014].

Green, P.S. (1994). Flora of Australia, vol. 49: Oceanic islands I, (Australian Government Publishing Service: Canberra).

McNeill, J., Barrie, F.R., Buck, W.R., Demoulin, V., Greuter, W., Hawksworth, D.L., Herendeen, P.S., Knapp, S., Mar- hold, K., Prado, J., Prud’homme van Reine, W.F., Smith, G.F., Wiersema, J.H. & Turland, N. (2012). International code of nomenclature for algae, fungi, and plants (Mel- bourne Code). (Koeltz Scientific Books: K6nigstein). [Regnum Vegetabile 154].

Sykes, W.R. & Atkinson, I.A.E. (1988). Rare and endangered plants of Norfolk Island. (Botany Division, Department of Scientific and Industrial Research, New Zealand: Christchurch).

WCSP (2014). World checklist of selected plant families. (Royal Botanic Gardens: Kew, Richmond). http://apps. kew.org/wesp/ [Accessed: 25 Apr. 2014].

J OQURNAL of the ADELAIDE

BOTANIC GARDENS

AN OPEN ACCESS JOURNAL FOR AUSTRALIAN SYSTEMATIC BOTANY

flora.sa.gov.au/jabg

Published by the STATE HERBARIUM OF SOUTH AUSTRALIA on behalf of the BOARD OF THE BOTANIC GARDENS AND STATE HERBARIUM

State Herbarium of South Australia PO Box 2732 Kent Town SA 5071 Australia

Board of the Botanic Gardens and Reit State Herbarium

of South Australia Department of Environment, Water and Natural Resources

Journal of the Adelaide Botanic Gardens 27 (2014) 25-34

Three new species of, and realignments in, Rhododendron sect. Schistanthe (Ericaceae)

L.A. Craven ft

Australian National Herbarium, CSIRO Plant Industry, G.P.O. Box 1600, Canberra, Australian Capital Territory 2601

Abstract

Three new species are described, two from Sulawesi and one from New Guinea: Rhododendron dissilistellatum Craven, R. gumineense Craven and R. torajaense Craven. The circumscriptions of R. seranicum J.J.Sm. sensu Argent, R. radians J.J.Sm. sensu Argent & Sleumer, and R. javanicum (Blume) Benn. sensu Argent are considered to be excessively broad and the taxonomy of these three species 1s amended. One new name and six new combinations resulting from realignments in taxonomic status are proposed: R. biappendiculatum Craven, R. cladotrichum (Sleumer) Craven, R. cockburnii (Argent, A.Lamb & Phillipps) Craven, R. extraneum (Sleumer) Craven, R. kinabaluense (Argent, A.Lamb & Phillipps) Craven, R. minahasae (Sleumer) Craven, R. pala-

wanense (Argent) Craven.

Keywords:

Rhododendron, Rhododendron sect.

Schistanthe, vireya rhododendrons,

Ericaceae, new species, realignments, taxonomy, Malesia

Introduction

Fieldwork in support of the doctoral research of Gillian Brown (School of Botany, The University of Melbourne, and CSIRO Plant Industry, Canberra) into the morphological systematics of Rhododendron sect. Schistanthe Schltr. (at that time called sect. Vireya (C.B.Clarke) H.F.Copel.) was undertaken in Sulawesi, Indonesia, in 2002 (see also Brown 2004, 2006a, 2006b). During the fieldwork, several collections of non-fertile plants were made and living material (seed or cuttings) was obtained of these in order to grow them on to flowering, at which point the plants could then be identified. In several other cases, herbarium collections were identified to a preliminary point only and detailed investigations into their identity were not then made.

The identification of the Sulawesi collections has recently been revisited and I have concluded that the taxonomy of R. radians J.J.sm. sensu Argent (2006) and Sleumer (1966), and of R. javanicum (Blume) Benn. sensu Argent (2006) is unsatisfactory. The diffe- rences between the several taxa included in these two species are as great as the differences between other taxa accepted as “good” species by these two authors. Accordingly, realignments in the taxonomy of these two species and of R. seranicum J.J.Sm. sensu Argent (2013) are proposed below. Notwithstanding any possible interpretation to the contrary, all the new species in the present contribution are accepted as species insofar as the provisions of the /nternational

‘The author passed away on 11 July 2014, three weeks after submitting the final version of this paper. — Editor.

Published online: 21 Aug. 2014 ¢ flora.sa.gov.au/jabg

Code of Nomenclature (McNeill et al. 2012) relating to acceptance of taxa by their authors are concerned.

Based upon studies of several DNA _ sequences conducted in the Benjamin D. Hall laboratory, Seattle, USA, Goetsch et al. (2011) published an inferred phylogeny of sect. Schistanthe. This research was used as the basis for proposing a new classification of Schistanthe (Craven et al. 2011). An important finding of the phylogenetic research was that there was no support for the formal recognition of subsections which are now known to be artificial assemblages of species with certain common morphological traits. In the newly proposed classification, the subsections are treated as informal groups. The advantage of this approach 1s that the informal groups, which can be keyed, are useful for identificatory purposes. In the text below, the names of the informal groups are given in Roman font.

It is worth noting in the context of taxon recognition in the present discussion that Sleumer (1966) apparently used the rank ‘variety’ when he was uncertain of the status of the entity involved (Argent 2006: 6). In other words, when Sleumer was uncertain as to whether or not an entity should be recognised at species rank or perhaps even not recognised at all, he ranked it as a variety. This led to his including taxa of uncertain acceptance within a “good” species, the expanded circumscription thus making that species itself of uncertain status. Argent (2006) has used the ranks of variety and subspecies in infraspecific taxonomies, but has not given a rationale for their application. In some cases he apparently has followed Sleumer in using the rank variety, e.g., for R. radians (see below).

ISSN 0313-4083 (Print) * ISSN 2201-9855 (Online)

L.A. Craven

Taxonomy

1. On the Euvireya group in western Malesia and Sulawesi

1.1. Rhododendron javanicum sensu Argent (2006)

Argent (2006) adopted a very broad circumscription for R. javanicum, one that 1s difficult to justify and that was criticised by Craven (2007). One significant issue 1s that there is no character state or combination of character states that uniquely defines the species R. jJavanicum sensu Argent. In fact, if one accepts that the summed character states of R. javanicum sensu Argent circumscribe a species of sect. Schistanthe then at least several other species accepted by Argent as being distinct species of the section would have to be included within his concept of R. javanicum.

It seems the taxa mentioned with respect to the discussion of R. javanicum sensu Argent (2006) are probably closely related; in Goetsch et al. (2011), clade 4 consists of western Malesian species that include several of the treated or mentioned taxa. It is pertinent also to note that within clade 4, R. brookeanum H.Low ex Lindl. and R. javanicum are in different, strongly supported clades, thus supporting Sleumer (1966) who maintained these two as distinct species.

Very little is known about the extent of sympatry within this group of species, which is unfortunate, as biotic sympatry can be a powerful indicator of genetic distinctness. Sleumer (1966) indicates that R. teysmannii Mig. and R. javanicum rarely occur in biotic sympatry, which 1s evidence they are reproductively isolated.

Several characters that are diagnostic for the western Malesian sect. Schistanthe species included by Argent (2006) in R. javanicum are listed in Table 1, with the relevant states indicated for each taxon as far as the states can be determined. The character states predominantly have been taken from Argent (2006) and Sleumer (1966), supplemented with data from Callard (2014), protologues and personal observations. Where a state was not given by Argent (2006) or Sleumer (1966) for one of the infraspecific taxa they treated, but those authors indicated that the infraspecific taxon was morphologically the same as the autonymic, infra- specific taxon except for particular, designated features, the applicable state of a character has been determined by inference. Typically branchlets of taxa within the complex are scaly and lack hairs, but when Sleumer (1966) and/or Argent (2006) state that an organ 1s glabrous it is interpreted as there being neither scales nor hairs. Characters that I consider to be valuable for characterising species include leaf arrangement, leaf lamina anatomy, outer perulae shape, corolla form, ovary surface and the ovary-style interface.

The leaves of Rhododendron species are spirally arranged on each seasonal growth flush, either pseudo- whorled or dispersed, and the possession of one or other of these two latter states 1s diagnostic. Occasionally, a

26

J. Adelaide Bot. Gard. 27 (2014)

species, in which the pseudowhorled condition is the usual state, may have dispersed leaves on a particularly vigorous growth or “water shoot” but this axis reverts to the pseudowhorled condition once the abnormally vigorous growth has slowed. Taxa which always possess the dispersed state are R. kinabaluense (Argent, A.Lamb & Phillipps) Craven, R. palawanense (Argent) Craven and R. schadenbergii Warb., the remainder have pseudowhorled leaves, with the exception of R. moultonii Ridl., in which the leaves are in an intermediate pseudowhorled-dispersed state.

Leaf lamina anatomy is an especially interesting feature. In R. javanicum, the leaves of fresh material are relatively thin and the primary (the major lateral) veins are distinct; the leaves appear “dryish” when snapped. In some other taxa, such as R. brookeanum and R. kinabaluense, the leaves are thickened and the primary veins are obscured, apparently by being more or less immerged within the lamina; the leaves of fresh material appear “subfleshy” when snapped. The “subfleshy” leaves apparently are, to one extent or another, sunken between the primary veins and can perhaps be best described as “puckered”. The “dryish” leaves are flat between the primary veins. The subfleshy, puckered leaf type is well shown in photographs in Argent (2006: 246, R. brookeanum, & 247, R. cockburnii). The dryish leaf type is depicted well in Callard (2014: Species Gallery, as Rhododendron javanicum subsp. Javanicum).

Anatomical studies underway at the Erik Nilsen laboratory at Virginia Tech, Blacksburg, will demonstrate the structural differences between the two leaf types and provide answers as to whether or not distribution of these leaf anatomies within Euvireya are congruent with the phylogenetic clades recovered by Goetsch et al. (2011) (E. Nilsen, personal communication). As it 1s of particular interest to those interested in physiological adaptations and as many sect. Schistanthe species are epiphytic, one hopes the Nilsen lab will also consider the potential significance of the subfleshy leaf anatomy to such aspects as water management. My impression 1s that these leaf features reflect major genetic differences, possibly indicating separate evolutionary paths, and therefore should be acknowledged in classification by distinct, specific-level, taxonomic recognition as opposed to uncritical “lumping”’.

The shape of the outer perulae (inflorescence bud scales) and the nature of any indumentum also are valuable in species delimitation. Unfortunately, because of their essential ephemeral nature these usually are not available on herbarium specimens and data are only directly available for R. brookeanum and R. javanicum — collectors usually only collect rhododendrons when they are in flower, by which time the perulae have largely fallen, and non-insightful collectors usually do not collect buds. The corolla of all taxa is funnel-shaped, with the exception of R. cockburnii, in which the corolla is tubular-funnel-shaped with a constriction in the tube. Surface features of the ovary, whether scaly, hairy or

Rhododendron sect. Schistanthe (Ericaceae)

J. Adelaide Bot. Gard. 27 (2014)

(79s ayqs ‘AITeY 0] Sutsadyy AAvao Ajasuap (jas (jasands ags Maa LO 01 Buisadyy tavao SNOLGUIO Ayeos pure Asrey ATQeiqrea JO snosqeys ay As 01 Surode Ajidniqe ATeAo snoigery]s Arey AyeOs Ajasreds ay As AJOA JO 0} SurtJade} AIeAO snoiqeys ays 04 Butsadvs Arrey (dnagv ¥ Kavao pue (qos aggs 04 Butsadvy (Ate (jadnagn + davao pup yas snoiqeys ays 04 Butsadvy (Ate (adnagnv = ksvao pup yas ap Aas 0} Sutsode} Arey Ajidniqe + ATeAo pur Ayeos JIVJIONUI soRyns a] A4qs-ATeAGC) AJVAGC)

(F)

CE-E (-S 7) = padvys-jauunf

(Z-) 9-S

Ww

9-6

(¥-) G°¢-€ (-G’°Z)

ys

(J-) S-G'h (-¥)

¢

(/-) S-G'F (-4)

(Z-) S-S'y (-) (ur) ysuaT

padvgs-jauunf{

podeys-jouuny

podeys-jauuny

podeys-jouung

podeys-jauuny

podeys-jouuny

UOTIOINSUOD ® UM ogni ‘podeys

-jpuUny-TeyN qn

padvys-jauunf

podeys-jouuny

‘EUTLUNIEQNS 31€AO

SNOLGUIS SPUOAINU

-ayvingns (jq4oys ‘4000 (7pvo0Lg

Arey SABMIE Ayreou

SNOLGUIS SVUOAINU

-aypngns Mjq4oys 1000 (7pvo0Lg

(Ate jou Wvas xv]

AyeOs AJXE] JO snoiqeys

ATBOS ATXP] snoiqry]s ‘MeUOIONUI -ajeyngns Apsoys AITeY

‘78AO0 Aypeoiq 10U ‘ATBOS AJXE]

SNOLGUID 10U QIDUIUNIVGNS awao ‘10 Arey “ATeOS

SNOLQUIS

‘aqpuiunspgns aywao Arey OU “ATeOS

SNOLQUIS

‘QIDUIUNIVNS agvao SNOIQP]S -/+

snosqerys Arey

AJares ‘ATBOS

(yas (79s Ayeos AyeOs Ayeos ATXE] Ayeos Ayeos Ayeos AXE] (ze

jou Wvos ou Jo Arrey ‘Ajeos

(A20Y Asrey you ‘ATeos qou Wvos —- A|xey] syoTyOuRIq [10U IO

Arrey3] “qos Arey “qpas Arrey pure (ys

Arey

10U ‘ATeOS ATBOS ATXE]

yy WIOFep]OIOZ odeys sepnsod JaING Sd¥FINS [OOIpsg sd"ZINS JeaT sovJANs JopWoursg

‘SOJE]S/SISJOLIEYO JO UOISSNISIP JOJ 1X9} BAS ‘dnolb eASJIANG OU} Ul Sa1I09ds UOJPUEPOPOYY UEISEeI\| WSISEM USAQIO JO Saje}s Ja}OeJeYD “| B|qe]

sutaa Areurtid sulINosqo pure pousyory) JOU

(UOTVUTIUOD spoou) pomnosqo Ay[Ny 10U A]qissod sutaa Areuttid pue

pouayory) sdeysod

(UOTeUTIUOD spoou) pornosqo Ay]Ny 10U A]qissod Suto Areurtid pur

pousyoryi sdeysod

poainosqo sutaa Areurtid

pur pousyory) poainosqo sutaa Areutid

pure pousyoryy

sutaa Areurtid SuUTINOsqO pur pouaxpry1 10

sutoa Areurtid SuTINOsqoO pur

pousyoruy 10u Apuaredde

poinosqo sutaa Areurlid

pure pousyory)

pomnosqo Ay]ny 10u Ayqissod

suraa Areurtid

yng pousxprya Ajiuoredde

pajloymopnasd

pasiadstp

passadstp

pasiadstp pajioyMopnasd

IIETPOUTIOIUT

possadstp

poytoyMopnasd

Os AJ2SOO] oi Apysn ‘payoymopnasd

pajtoymopnasd

payzoymopnasd

pajtoymopnasd

pajzoymopnoasd

BUIWIE] JeaT WoW SuvsIe JeaT

MuUUvUSha, “Y

11DAIQUIPVYIS “M

asuaupnyud “yy

11U0[JNOUL “MY

as UIN|VqQVULA OL

wngiupavl “y

ANLIVAD “MY

WINIUDAIXA “MY

71UA AGG IOI Oo

WNYILAJOPV]I “MY

WNUVIYOOL “M

UOXETL

(996) Jawina|s Jo/pue (9002) juaby WO) sdUaJaJUI Ag PaAUEp ale Sa/ej/, Ul Saye}S Ja}OeIeUD “SUONEAJASGO |eUOSIEd ae JO (pL. OZ) puelieD ‘(996],) sawNals ‘(990Z) JUebuy WO) Uaye} UB9g SAY JUO! UEWOY, Ul Sa}ejs sa}OeIeUD

2

L.A. Craven

elabrous, are useful and easily scored. Generally, the particular states are consistent within a taxon, but in R. gracile and R. teysmannii hairs may be present or absent and in R. palawanense the ovary may be glabrous or scaly. Another useful feature involving the style is the ovary/style interface. The ovary may taper gradually to the style or may taper abruptly and apparently this is a difference between R. javanicum, in which the ovary tapers to the style, and R. brookeanum, in which it more or less abruptly tapers to the style.

There are other characters, the states of which have diagnostic value, such as flower orientation (erect, sem1- erect, spreading, semipendulous, etc.), inflorescence shape (domed or flat), anther colour (yellowish or dark reddish/purple) and pollen colour (at least grey or cream), that could be used for circumscribing the species, but data are not available for sufficient taxa to include them in Table 1. Assessment of inflorescence shape is difficult when a plant is not flowering to the maximum of its potential; reduced nourishment, reduced vigour, etc., can result in the production of flat inflorescences instead of the usual domed inflorescences that that genotype may produce when in a state of optimal vigour.

The name R. brookeanum var. extraneum Sleumer is not mentioned by Argent (2006), neither in the index nor in the text dealing with the R. javanicum complex. This is not just an isolated instance of a name either not appearing in Argent’s book or not being satisfactorily accounted for; it happens quite frequently. Surprisingly, Argent has not even included some of his own names, such as R. brookeanum subsp. brookeanum var. kina- baluense (Argent, A.Lamb & Phillipps) Argent, R. brookeanum subsp. gracile (Lindl.) Argent, and several others relevant to the present discussion as far as their taxonomic and/or nomenclatural status 1s concerned.

Having considered the variation recorded for the taxa listed in Table 1 together with variation in other organs such as leaf lamina shape and size, etc., I believe the taxonomy of this complex is best resolved with all the taxa being accepted at species rank, as listed below. The species can be identified using the keys in Argent (2006: 234-237), with the exception of R. extraneum (see below).

Additional synonyms are given under the relevant taxa in Argent (2006) and Sleumer (1966).

R. brookeanum H.Low ex Lindl.

J. Hort. Soc. London 3: 823 (1848). — R. javanicum subsp. brookeanum (H.Low ex Lindl.) Argent & Phillipps, Bot. J. Linn. Soc. 85: 15 (1982).

Notes. Rhododendron brookeanum may be distinguished from R. javanicum sensu stricto by the puckered leaf lamina; the ovate subacuminate, glabrous outer perulae; and the ovary more or less abruptly tapering to the style. In R. javanicum the leaf lamina is flat; the outer perulae are broadly ovate, shortly subulate-mucronate, glabrous; and the ovary tapers to the style.

28

J. Adelaide Bot. Gard. 27 (2014)

R. cladotrichum (Sleumer) Craven, comb. et stat. nov.

Basionym: Rk. brookeanum var. cladotrichum Sleumer, Reinwardtia 5: 224 (1960). — R. javanicum subsp. cladotrichum (Sleumer) Argent, Rhododendrons of subgenus Vireya 247 (2006).

Notes. Sleumer (1966) separated R. cladotrichum (as var. cladotrichum) from R. brookeanum (as var. brookeanum) by the branchlets being hairy and the leaf lamina midrib being densely short-hairy on both surfaces in R. cladotrichum and the branchlets and leaf lamina being glabrous in Rk. brookeanum.

R. cockburni (Argent, A.Lamb & Phillipps) Craven, comb. et stat. nov. Basionym: R. javanicum subsp. cockburnii Argent, A.Lamb & Phillipps, Notes Roy. Bot. Gard. Edinburgh, 42: 113 (1984). — R. brookeanum subsp. cockburnii (Argent, A.Lamb & Phillipps) Argent, Edinburgh J. Bot. 52: 364 (1995).

Notes. The pseudowhorled leaves, the subfleshy and puckered leaf lamina and the corolla with its constricted tube are diagnostic for the species. The corolla is depicted in Argent (2006: 247, as R. javanicum subsp. cockburnii).

R. extraneum (Sleumer) Craven, comb. et stat. nov.

Basionym: R. brookeanum var. extraneum Sleumer, Reinwardtia 5: 225 (1960).

Notes. Rhododendron extraneum was treated by Sleumer (1966) as a variety within R. brookeanum. Rhododendron extraneum occurs on Sumatra whereas the other three of the four infraspecific taxa of R. brookeanum sensu Sleumer (1966) are from Borneo. It may well be that the present taxon fell into Sleumer’s “uncertain status” category, discussed above, and he placed it as a variety within R. brookeanum. The two specimens of R. extraneum in L, Meyer 4582 (the type) and Meyer 5390, cannot be keyed to a taxon in Argent’s key to sect. Euvireya sensu Argent (F. Adema, personal communication) and this supports its status as a distinct species.

Sleumer (1966) separated this taxon from R. gracile (as var. gracile) on the basis of the leaf lamina being narrowly ovate and the corolla being 5 (—6) cm long in R. gracile and the leaf lamina being narrowly ovate- oblong and the corolla being 4 (—5) cm long in R. extraneum.

R. gracile H.Low ex Lindl.

J. Hort. Soc. London 3: 84, fig. (1848). — R. brookeanum var. gracile (H.Low ex Lindl.) Henslow, J. Roy. Hort. Soc. 13: 261, f. 42 (fol.), 43a (i1.) (1891). — R. javanicum subsp. gracile (H.Low ex Lindl.) Argent, A.Lamb & Phillipps, Notes Roy. Bot. Gard. Edinburgh, 42: 114 (1984). — R. brookeanum subsp. gracile (H.Low ex Lindl.) Argent, Edinburgh J. Bot. 52: 364 (1995).

Notes. Fide Argent et al. (1984) the leaf lamina is smooth, not puckered, and up to 3.5 cm wide. The species 1S well depicted in Argent (2006: 247, as R.

J. Adelaide Bot. Gard. 27 (2014)

Javanicum subsp. gracile). Argent (1995) considered the taxon as “clearly requiring at least subspecific status”’.

R. Javanicum (Blume) Benn.

Pl. Jav. Rar. 85 (1838), excl. t. 19, which 1s R. teysmannii fide Sleumer (1966). — Vireya javanica Blume, Buydr. FI. Ned. Ind. 15: 854 (1826).

Notes. See under R. brookeanum for the differences between these two species.

R. kinabaluense (Argent, A.Lamb & Phillipps)

Craven, comb. et stat. nov.

Basionym: R. javanicum var. kinabaluense Argent, A.Lamb & Philltpps, Notes Roy. Bot. Gard. Edinburgh, 42: 113 (1984). — R. brookeanum var. kinabaluense (Argent, A.Lamb & Phillipps) Argent, Edinburgh J. Bot. 52: 363 (1995). — R. javanicum subsp. kinabaluense (Argent, A.Lamb & Phillipps) Argent, Rhododendrons of subgenus Vireya 248 (2006).

Notes. Rhododendron kinabaluense 1s characterised by its dispersed leaves, puckered leaf lamina, and appressed and often emarginate outer perulae.

R. moultoniu Ridl.

J. Straits Br. Roy. As. Soc. 63: 61 (1912). — R. javanicum subsp. moultonii (Ridl.) Argent, Bot. J. Linn. Soc. 85: 16 (1982). — R. brookeanum var. moultonii (Ridl.) Argent, Edinburgh J. Bot. 52: 364 (1995).

Notes. This species 1s characterised by the intermediate, dispersed-pseudowhorled arrangement of its leaves, the puckered leaf lamina, and the ovary abruptly tapering to the style.

R. palawanense (Argent) Craven, comb. et stat. nov.

Basionym: R. javanicum subsp. palawanense Argent, Gard. Bull. Singapore 56: 90 (2004).

Notes. Rhododendron palawanense 1s _ characterised by the dispersed leaves, flat leaf lamina, acute (never emarginate) outer perulae, and grey pollen.

R. schadenbergii Warb.

in Perkins, Fragm. Fl. Philippines 172 (1905). — R. java- nicum var. schadenbergii (Warb.) Sleumer, Reinwardtia 5: 195 (1960). — R. javanicum subsp. schadenbergii (Warb.) Argent, Edinburgh J. Bot. 52: 364 (1995).

Notes. Rhododendron schadenbergii 1s characterised by the dispersed leaves, apparently flat leaf lamina, and cream pollen. Argent (1995) comments that this plant “warrants at least subspecific status’.

R. teysmanni Miq.

FI. Ned. Ind., Eerste Biyv. 3: 585 (1861). — R. javanicum var. teysmannii (Migq.) King & Gamble, J. As. Soc. Bengal 74: 75 (1905). — R. javanicum subsp. teysmannii (Miq.) Argent, Rhododendrons of subgenus Vireya 249 (2006).

Notes. This species is characterised by the flat leaf lamina, the very commonly hairy pedicels, and the more or less densely pubescent ovary. As noted above, R. teysmannii and R. javanicum may occur in biotic sympatry which is evidence in support of their being distinct species.

29

Rhododendron sect. Schistanthe (Ericaceae)

1.2. Anew name for Rhododendron seranicum subsp. sparsthirtum

Rhododendron biappendiculatum Craven, nom. et Stat. nov. Replaced synonym: R. seranicum J.J.Sm. subsp. sparsi-

hirtum Argent, Rhododendrons, camellias and magnolias 2013, 127 (2013) (as sparsihirtus).

Notes. Rhododendron seranicum subsp. sparsihirtum was distinguished from subsp. seranicum on the basis of it having hairs at the base of the filaments, on the disk and inside the corolla towards the base (Argent 2013). Argent also noted that the Binney plant, which is the type accession and the only specimen cited by Argent, “superficially looks identical to the other accessions of R. seranicum from Sulawesi, only on dissection of the flowers can the distinguishing hairs be seen” (Argent 2013).

The following material also originated from the Gunung Sojol complex in Sulawesi: A.Rouse s.n., Australia, Victoria, cultivated in Hawthorn East, 20 July 2012 (CANB) [provenance: Indonesia, Sulawesi Tengah, the western lower slopes of the Gunung Sojol complex, on Tinombo-Sipatoh path, open steep grassy slope, leg. L.A. Craven, 23 July 2002, living material only collected]. This material must have been collected very close to, if not from the same, population from which living material was collected by David Binney, as, at the time of our visit in 2002, there was only the one path from the last village, Sipatoh; this was used by rattan harvesters, etc., to access the Gunung Sojol massif. The Binney-collected material subsequently afforded the type specimen of subsp. sparsihirtum. The Rouse material differs in several particulars from R. seranicum, as that species was circumscribed by Sleumer (1966). Sleumer noted that a slightly differing specimen, 1.e., to R. seranicum in the Moluccas, had been collected from Sulawesi, but, as he gave the distribution of R. seranicum as being the Moluccas only, it can confidently be taken that his 1966 description and concept of the species is based upon Moluccan specimens only. The Rouse collection differs from R. seranicum sensu Sleumer in at least the following: the indumentum features noted by Argent (2013); the biappendiculate anthers (blunt or minutely apiculate in Rk. seranicum); the narrowly oblong-obovoid ovary, brownish when dry (oblong-conical, blackish when dry in R. seranicum); and the pedicels not, or scarcely, widened at the apex (thickened at apex in R. seranicum).

Argent (2013) also stated that the Gunung Sojol plant

agrees quite closely to R. javanicum ssp. schadenbergii (Warb.) Argent a Philippine sub-species that was recorded from Manado in the north of Sulawesi (Sleumer, 1960). However the plants from Sulawesi have much shorter anthers than 1s recorded for R. javanicum ssp. schadenbergii

from the Philippines and it seems safer for the present to keep that subspecies restricted to the Philippines.

L.A. Craven

It seems therefore that an association of the Gunung Sojol plant with the plant from the Moluccas may not be so strong that its inclusion within the circumscription of R. seranicum is the best possible taxonomic disposition for the Gunung Sojol plant. Given the several differences between the two plants noted above, and given that additional differences may be discovered in future (no material from the Moluccas has been available to me and I have relied on the data in Argent (2013) and Sleumer (1966), and that obtainable from the Rouse collection in CANB), my conclusion is that the Gunung Sojol plant should be recognised as a distinct species.

The epithet sparsihirtum 1s not especially appropriate at species rank in sect. Schistanthe given that many of its species, when hairy, are sparsely so and for that reason a new name is given. Although Sleumer (1966) indicates that a tendency towards the appendiculate condition may be present in R. seranicum, the biappendiculate anthers in the Gunung Sojol plant are a striking feature.

1.3. A new species from Sulawesi

Rhododendron torajaense Craven, sp. nov.

From R. rhodopus Sleumer it differs in the coriaceous, acuminate leaf lamina; the weakly sweet-fragrant flowers; the corolla being 45-50 mm long including the lobes; and the anthers being 3 mm long. In R. rhodopus the leaf lamina is obtusely acuminate to narrowly acute and cartilaginous, the flowers are strongly feijoa fruit-scented, the corolla 1s 65—79 mm long including the lobes, and the anthers are 6 mm long.

Holotypus: Indonesia, Sulawesi Selatan, Tana Toraja, Batutumonga, on way to Gunung Sesean, Lat. 02° 54’ 32” N, Long. 119° 53’ 02” E, alt. 1359 m, 12 July 2002, G.Brown, L.Craven & L.Juswara 78 (BO). Isotypus: CANB.

Shrub to | m tall. Branchlets glabrescent (initially moderately densely scaly), c. 2.5-4 mm in diameter. Leaves 4—6-whorled (on vigorous growth the pseudo- whorls loose with the proximal leaves discretely inserted). Leaf lamina narrowly elliptic, elliptic, narrow- ly ovate or ovate, 85—170 x 35—75 mm, base cuneate to rounded, apex acuminate, margin subrevolute; abaxially and adaxially with moderately dense sessile scales with the rim irregular (but not incised) and with a raised centre; midrib prominent abaxially, adaxially prominent proximally and becoming slightly impressed towards the apex, lateral veins distinct, 8-14 per side. Petiole 12—277 mm long. /nflorescence a 7—8-flowered umbel, flowers erect to spreading. Pedicel 14-22 mm long. Calyx obsolete. Corolla short-salverform, 45-50 mm long including the lobes, white, weakly sweet-fragrant, tube 28-30 mm long, lobes 17—20 mm long and spreading at right angles to the tube, sparsely scaly outside and with or without pubescent hairs on the very proximal region of the tube. Stamens 10, unequal in length, slightly exserted; filaments 30-35 mm long, densely pubescent-hairy for the proximal c. 2/3 and glabrous above; anthers oblong, c. 3 mm long. Disk densely hairy. Ovary narrowly fusiform, 11—12 mm long, scaly

30

J. Adelaide Bot. Gard. 27 (2014)

Fig. 1. Rhododendron torajaense. Cultivated at the Rhododendron species Botanic Garden, Washington, USA. Photo: Hank Helm.

and very densely hairy; style c. 18 mm long, exserted to c. 8 mm, densely hairy excepting the glabrous distal c. 4 mm; stigma subdiscoid, c. 3.5 mm in diameter. Fruit not available. Fig. 1.

Distribution and ecology. Indonesia, Sulawesi Selatan, Tana Toraja, on way to Gunung Sesean, Batutumonga, Lat. 02° 54’ 32” N, Long. 119° 53’ 02” E, alt. 1359 m, on a disturbed roadside bank. Only known from one wild location.

Conservation status. This species is best given the conservation status Vulnerable according to the criteria of the IUCN Red List (UCN 2012), although, as the necessary surveys have not been undertaken, Data Deficient would also be applicable. The type collection, the only collection seen, was made in the village of Ba- tutumonga. The Batutumonga region is much disturbed with rainforest in the area above the village considerably cleared and converted to anthropogenic grassland.

Etymology. The specific epithet is derived from the region Tana Toraja.

Notes. The corolla form has been given as “short- salverform” in the description above. “Tubular- campanulate” might also be applicable. Argent (2006: 364) gives several examples of corolla form in sect. Schistanthe species and his illustration for “cylindrical” is the closest to the corolla form in R. torajaense. However, the term cylindrical is better applied to the corolla of species such as R. perakense King & Gamble in which the corolla lobes are erect, 1.e. in line with the tube. In R. torajaense, the lobes are spreading at right angles to the tube and cylindrical is not the best term to use. Short-salverform is more appropriate.

The flowers of R. torajaense have a weak, sweet fra- erance. It is not rich as in R. konori Becc., nor “spicy” as are the flowers of some other fragrant species, but

J. Adelaide Bot. Gard. 27 (2014)

~ Lo —

7

Rhododendron sect. Schistanthe (Ericaceae)

Fig. 2. Rhododendron gumineense. Cultivated at East Hawthorn, Victoria, Australia. Photo: Andrew Rouse.

is a little reminiscent of gardenias. The first plant of Brown et al. 78 to flower at Melba did so in February 2011 and the second plant flowered in June 2011. The fragrance of the latter was very much weaker than that of the summer-flowering plant, perhaps due to the cold weather, but the scent was so slight that it is difficult to describe in qualitative terms. The fragrance of the flowers of R. rhodopus is very different and in this species the flowers of the field-collected material (Brown, Craven, Juswara & Ramadhanil 129, BO, CANB) had a fragrance very much like that of the ripe fruit of fe1joa, Acca sellowiana (Myrtaceae).

2. Anew species of the Phaeovireya group from New Guinea

Rhododendron gumineense Craven, sp. nov.

From R. beyerinckianum Koord. it differs in having the leaf lamina narrowly elliptic to elliptic, 22-35 <x 7—16 mm, base cuneate and often very narrowly so, apex obtuse or shortly acuminate to very narrowly acute to very narrowly acuminate; anthers 1.2—1.6 mm long; and the ovary subcylindrical and tapering to the style. In Rk. beyerinckianum the leaf lamina is narrowly ovate, to broadly elliptic, obovate or subcircular, 30-60 x 10-35 mm, base broadly tapering or rounded, apex obtuse, broadly acute, sometimes apiculate; anthers 2—2.5 mm long; and the ovary elongate conical or subovoid, usually abruptly tapering distally.

Holotypus: Australia, Victoria, Hawthorn East, culti- vated in greenhouse, A.Rouse s.n., 29 December 2012 (CANB).

31

Shrub to 40 x 40 cm. Branchlets terete, densely scaly, c. 0.8-1 mm in diameter. Leaves 3—5-pseudo- whorled. Leaf lamina narrowly elliptic to elliptic, 22— 35 x 7-16 mm, base cuneate and often very narrowly SO, apex obtuse or shortly acuminate, very narrowly acute, or very narrowly acuminate, margin revolute; abaxially with dense dendroid scales each on a tubercle and with the limb deeply dissected; adaxially with the scales quickly deciduous and occasional hairs present along the midrib; midrib prominent abaxially, impressed to flat adaxially, lateral veins obscure abaxially and prominent adaxially, 5—8 per side. Petiole 7-12 mm long. Inflorescence 1—2-flowered, the flowers spreading to subpendulous. Outer perulae ovate, 8—8.5 mm long, acuminate with the acumen broad. Pedicel 6—9 mm long, scaly. Calyx obsolete. Corolla pink, tubular-curved with the tube flaring distally, c. 23-29 mm long including the lobes, tube curved, c. 17—20 mm long and 6-9 mm in diameter, lobes 6—9 mm long; outside the tube and lobes moderately scaly, inside glabrous. Stamens 10, slightly exserted; filaments c. 22—26 mm long, glabrous, anthers light brown, oblong, 1.2—1.6 mm long, base subapiculate. Disk scaly. Ovary subcylindrical and tapering to the style, 6 x 2 mm, densely scaly; style 15—20 mm long, slightly exserted and approximating the stamens in length, scaly to within 4-5 mm of the apex; stigma subcapitate, c. 1.7 mm in diameter. Fruit immature. Fig. 2.

L.A. Craven

J. Adelaide Bot. Gard. 27 (2014)

Table 2. Character states of three Sulawesi Rhododendron species in the Solenovireya group. States have been taken from Argent (2006), Sleumer

(1966) or are personal observations. Leaf scales

Taxon Leaf lamina

R. minahasae obovate to elliptic, base cuneate to truncate,

apex acute to obtuse

R. pubitubum elliptic or obovate elliptic, sometimes obovate, broadly attenuate towards the base, the very base rounded, not rarely subcordate, apex very broadly attenuate, obtuse to nearly rounded, occasionally very shortly retuse

R. radians ovate to ovate-oblong, base slightly to distinctly cordate, apex shortly obtusely

attenuate, sometimes nearly rounded

incised

Distribution and ecology. Papua New Guinea, Chimbu Province, Gumine. Nothing is known about the ecology of R. sumineense. However, the elevation of Gumine 1s c. 1700 m and consequently the forest type in the region would probably be lower- to mid-montane rainforest. Whether the species is epiphytic in rainforest or terrestrial on open ground is unknown. It is only known from one natural location.

Conservation status. Rhododendron gumineense 1s best given the conservation status Data Deficient according to the criteria of the IUCN Red List (UCN 2012). Nothing is known about its population size or ecology.

Etymology. The specific epithet is derived from the place name Gumine.

Notes. Rhododendron gumineense keys to R. beyer- inckianum Koord. in Argent (2006), from which it differs in the features given in the diagnosis above. Additionally, the species differs from R. beyerinckianum in habit, being a very small shrub in contrast to R. beyerinckianum which is a shrub or tree from 1-5 m tall. Initially I considered that the Gumine plant might represent a primary hybrid, but I have grown it to flowering from seed obtained from my own plant and there were no morphological differences between the parent plant and those plants grown from seed.

Even though described from cultivated material, R. gumineense has a known provenance. It was collected at Gumine by D.Stanton, Wollongong, NSW, Australia, in 1971, and sent to the Australian Rhododendron Society, Olinda, Victoria, Australia by L.Searle (Kundiawa, Chimbu Province, PNG) in 1974.

In cultivation in Australia, apparently from its first introduction, the species has been misidentified as R. xschoddei Sleumer, which is a plant possibly of hybrid

funnel-shaped and inserted on a pedestal or foot of lamina origin

marginal zone irregularly dentate or bitten, with a + persistent, blackish and sunken centre

almost regularly rounded ovate-acuminate, acute, in outline to distinctly

oe

Outer perulae Style

+ obtuse, subsericeous at hairy and scaly almost to the apex or practically the top glabrous

orbiculate to obovate, subdensely hairy and densely short laxly scaly below (the

scales well visible

there), laxly hairy in the middle, glabrous for the distal 10 mm

subsericeous on both surfaces

laxly patent-pubescent to the lower 5/6 and bearing some scales in the lower part

+ glabrous dorsally

origin and is best treated as a nothospecies until its biological status can be verified.

3. On the Solenovireya group in Sulawesi

Sleumer (1966) recognised two varieties within the Sulawesi species R. radians: var. radians and vat. minahasae Sleumer. Argent (2006) added a third taxon, var. pubitubum (Sleumer) Argent, differentiating it from the other two varieties by “the corolla tube being densely hairy throughout its length and with the lobes hairy outside along the middle line’. Differences bet- ween these three taxa are given in Table 2.

Scanning electron micrographs of R. dissilistellatum Craven, R. minahasae and R. radians showing abaxial scales and stomata are given in Fig. 3. The distribution of stomata relative to scales is worthy of note in R. minahasae and R. radians. In R. minahasae the sto- mata and scales are quite evenly distributed (Fig. 3B) whereas in R. radians the scales are inserted in a shallow depression in a zone which lacks stomata (Fig. 3D). A reason for this distribution pattern is not immediately evident. Material of R. pubitubum has not been available for examination.

Based on the data given in Table 2 plus the scale distribution data, it is my conclusion that these three taxa should be treated at species rank. A new combination for R. minahasae 1s provided below.

Living material of another species collected in 2002, of which fertile material was not available in the field, subsequently flowered in cultivation. It keyed to R. radians in both Argent (2006) and Sleumer (1966), which is not surprising as long-tubular, white-flowered plants would key to this species alone. The newly flowered plant proved to be distinct from all other Sulawesi species of Solenovireya and is described be- low as R. dissilistellatum Craven.

J. Adelaide Bot. Gard. 27 (2014)

The Sulawesi species of the Solenovireya group (sensu Craven et al. 2011) can be distinguished by the following key:

1. Corolla 40—45 mm long including the lobes . . R. amabile 1: Corolla more than 50 mm long including the lobes 2. Leaf lamina narrowly elliptic to narrowly obovate .... R. dissilistellatum 2: Leaf lamina ovate to ovate-oblong, obovate to elliptic, or obovate-elliptic

3. Leaf lamina ovate to ovate-oblong R. radians

3: Leaf lamina obovate to elliptic, or obovate-elliptic

4. Perulae + obtuse, subsericeous at the apex or practically glabrous; style hairy and scaly almost to TOEVAPSK ihe ey ete Sees R. minahasae

4: Perulae orbiculate to obovate, densely short seri- ceous on both faces; style hairy and scaly excepting the glabrous distal 10 mm R. pubitubum

Rhododendron dissilistellatum Craven, sp. nov.

From R. radians J.J.Sm. it differs in the leaf lamina being narrowly elliptic to narrowly obovate with the base cuneate to obtuse (in R. radians the lamina is ovate to ovate-oblong with the base cordate), the outer perulae being elliptic to broadly elliptic (ovate- acuminate in R. radians), and the style being hairy to within c. 2 mm of the apex (hairy 1n the proximal 5/6 of the style in R. radians).

Holotypus: Australia, Australian Capital Territory, Melba, cultivated in greenhouse (from living material of G.Brown, L.Craven, L.Juswara & Ramadhanil 114), L.A.Craven 15044, September 2010 (BO). Isotypi: A, CANB, CEB, E, L, P. wa

Lax shrubs to c. 60 x 70 cm. Branchlets moderately densely scaly and with moderately dense short hairs, c. 1.5—2 mm in diameter. Leaves 5—10-pseudowhorled. Leaf lamina narrowly elliptic to narrowly obovate, 25-64 x 7-19 mm, base cuneate to obtuse, apex retuse, margin subrevolute; abaxially with the scales sessile, mostly flat and rarely subfunnel-shaped, and with scattered hairs throughout; adaxially with sessile, funnel-shaped to flat, subentire to dentate scales with a broadish flattened centre, and with short scattered hairs that are more dense proximally along the midrib; midrib prominent abaxially, impressed adaxially, lateral veins more or less obscure, c. 5—6 per side, reticulate venation obscure. Petiole c. 1-3 mm long. /nflorescence a 7/—15-flowered umbel, flowers in a more or less spreading band. Outer maximally-developed perulae elliptic to broadly elliptic, c. 8-11 mm long, obtuse to rounded and often secondarily retuse through rupturing occurring at anthesis. Pedicel 7-13 mm long, scaly and hairy. Calyx obsolete. Corolla white, salverform, c. 59- 79 mm long including the lobes, tube straight, c. 50-67 mm long and 2—3 mm in diameter, lobes c. 9-12 mm long, outside the tube laxly scaly and proximally with a few hairs, inside hairy to about % to the apex, lobes glabrous. Stamens 10, scarcely exserted; filaments c. 50-65 mm long, hairy for about ’2 way to their apex, anthers light brown, oblong, c. 1.4—1.8 mm long, base

a3

Rhododendron sect. Schistanthe (Ericaceae)

Fig. 3. Scanning electron micrographs of the abaxial leaf surfaces of Rhododendron species. A R. dissilistellatum. B—-C R. minahasae: B (a) indicates scales; (b) indicates contaminant on the leaf surface; C arrow indicates the foot or pedestal that is of lamina origin. D R. radians. — A Craven 15044; B-C Alston 15800; D Brown et al. 128.

obtuse to rounded. Disk hairy. Ovary 6.5—7 x 1.2 mm, subcylindrical and tapering to the style, densely scaly and hairy; style c. 60-65 m long, exserted to c. 12 mm, scaly proximally, hairy to within c. 2 mm of the apex, stigma c. 2 mm in diameter. Fruit not available. Fig. 3A, 4.

Distribution and ecology. Indonesia, Sulawesi Tengah, the western lower-mid slopes of the Gunung Sojol complex, Balukang, on Tinombo-Siboang path (between camp 3 and camp 1), Lat. 00° 28’ 19” N, Long. 120° 08’ 27” E, alt. 1153-1344 m, lower montane rainforest. It is only known from one natural location.

Conservation status. This species 1s best given the conservation status Data Deficient according to the criteria of the IUCN Red List (UCN 2012). Although known from only the one location in the wild, this location is unlikely to be disturbed by forest conversion operations or slash and burn gardening in the foreseeable future. Additionally, it is possible the species is widely distributed around the lower-mid slopes of the Gunung Sojol complex.

Etymology. The specific epithet 1s derived from the Latin dissilio, tly apart, burst and stella, star. A plant in full flower, especially on a dull day, brings to mind a certain type of modern firework, 1n which the primary rocket explodes into numerous smaller rockets that themselves explode, giving rise to numerous radiating shafts of brilliant light (often white); the inflorescences are analogous to these final displays.

L.A. Craven

Other specimens examined.

INDONESIA, SULAWESI TENGAH: the western lower slopes of the Gunung Sojol complex, Balukang, on Tinombo- Siboang path (between camp 3 and camp 1), Lat. 00° 28’ 19” N, Long. 120° 08’ 27” E, alt. 1153-1344 m, lower montane rainforest, G.Brown, L.Craven, L.Juswara & Ramadhanil 114, 23 July 2002 (BO, CANB).

R. amabile Sleumer Reiwardtia 5: 127 (1960).

R. minahasae (Sleumer) Craven, comb. et stat. nov.

Basionym: R. radians var. minahasae Sleumer, Rein- wardtia 5: 130 (1960).

R. pubitubum Sleumer Reinwardtia 5: 126 (1960). — R. radians var. pubitubum (Sleumer) Argent, Rhododendrons of subgenus Vireya 170 (2006).

R. radians J.J.Sm. Bull. Bot. Buitenzorg II, 1: 403, t. 51 (1920.

Acknowledgments

The fieldwork in Sulawesi took place in the company of Gill Brown (The University of Melbourne), Lina Juswara (LIPI, Bogor) and Ramadanil Pitopang (Tadulako University, Palu); their fellowship made this productive trip all the more enjoyable. The Research Foundation of the American Rhododendron Society is thanked for its support of the fieldwork. Financial support was also provided by CSIRO Plant Industry and CSIRO International Scientific Liaison. The staff of Kebun Raya Indonesia Bogor are thanked for their assistance in obtaining export documents.

I thank Dr Frits Adema and Gerard Thijsse, Leiden, for their assistance in the provision of information on specimens held at L; the responsibility for acting upon those observations remains with me but the unambiguity of their several reports was extremely helpful.

Mark ‘Talbot, Microscopy Centre, CSIRO Plant Industry, produced the scanning electron micrographs and Siobhan Duffy, Visual Resources Unit, CSIRO Plant Industry, produced the final plate. Photographs of R. gumineense and R. torajaense were kindly made available by Andrew Rouse and Hank Helm, respectively.

Elizabeth Minchin, Classics, ANU, as ever is thanked for her comments on my suggested epithets. The informed comments of two reviewers assisted ereatly in improving the manuscript.

References

Argent, G.C.G. (1995). New combinations and synonymy in some rhododendrons of section Vireya. Edinburgh Jounal of Botany 52: 363-365.

Argent, G.C.G. (2004). New species and comments on Rho- dodendron (Ericaceae) from the island of Palawan, Philippines. Gardens’ Bulletin Singapore 56: 79-94.

Argent G.C.G. (2006). Rhododendrons of subgenus Vireya. (Royal Horticultural Society: London).

Argent, G.C.G. (2013). Current taxonomy: Rhododendron seranicum J.J.Sm. ssp. sparsihirtus Argent. Rhododen- drons, camellias and magnolias 2013: 126—127.

34

J. Adelaide Bot. Gard. 27 (2014)

Fig. 4. Rhododendron dissilistellatum. Cultivated at Melba, A.C.T., Australia. Photo: Lyn Craven.

Argent, G.C.G., Lamb, A. & Phillipps, A. (1984). New taxa and combinations in vireya rhododendrons from Sabah (Borneo). Notes of the Royal Botanic Gardens Edinburgh 42: 113-120.

Brown, G.K. (2004). Phylogeny and biogeography of Rho- dodendron sect. Vireya. Unpublished PhD thesis. (School of Botany, The University of Melbourne: Melbourne).

Brown, G.K., Craven, L.A., Udovicic, F. & Ladiges, PY. (2006a). Phylogeny of Rhododendron section Vireya (Ericaceae) based on two non-coding regions of cpDNA. Plant Systematics & Evolution 257: 57-93.

Brown, G.K., Craven, L.A., Udovicic, F. & Ladiges, P.Y. (2006b). Phylogenetic relationships of Rhododendron section Vireya (Ericaceae) inferred from the ITS nrDNA region. Australian Systematic Botany 19: 329-342.

Callard, C. (2014). Vireya rhododendrons. http://www.vireya. net/ [Accessed: 12 Jan. 2014].

Craven, L.A. (2007). Rhododendrons of subgenus Vireya, Argent, G. 2006. Royal Horticultural Society: London. Journal American Rhododendron Society 61: 90-92. [Book review].

Craven, L.A., Danet, F., Veldkamp, J.-F., Goetsch, L.A. & Hall, B.D. (2011). Vireya rhododendrons: their mono- phyly and classification (Ericaceae, Rhododendron sect. Schistanthe). Blumea 56: 153-158.

Goetsch, L.A., Craven, L.A. & Hall, B.D. (2011). Major speciation accompanied the dispersal of Vireya rhodo- dendrons (Ericaceae, Rhododendron sect. Schistanthe) through the Malayan archipelago: Evidence from nuclear gene sequences. Jaxon 60: 1015-1028.

IUCN (2012). IUCN Red List Categories and Criteria, Version 3.1, ed. 2. (UCN Species Survival Commission: Gland, Switzerland & Cambridge, U.K.).

McNeill, J., Barrie, F.R., Buck, W.R., Demoulin, V., Greuter, W., Hawksworth, D.L., Herendeen, P.S., Knapp, S.., Marhold, K., Prado, J., Prud’>homme van Reine, W.F., Smith, G.F., Wiersema, J.H. & Turland, N.J. (2012). International Code of Nomenclature for algae, fungi, and plants (Melbourne Code). (Koeltz Scientific Books: Koenigstein). [Regnum Vegetabile 154].

Sleumer, H. (1966). Rhododendron. In: van Steenis, C.G.G.J. (ed.), Flora Malesiana I(6): 474—668. (Wolters-Noordhoff: Groningen). [Reprinted by Wolters-Noordhoff as An account of Rhododendron in Malesia with an introduction and index].

J OQURNAL of the ADELAIDE

BOTANIC GARDENS

AN OPEN ACCESS JOURNAL FOR AUSTRALIAN SYSTEMATIC BOTANY

flora.sa.gov.au/jabg

Published by the STATE HERBARIUM OF SOUTH AUSTRALIA on behalf of the BOARD OF THE BOTANIC GARDENS AND STATE HERBARIUM

State Herbarium of South Australia PO Box 2732 Kent Town SA 5071 Australia

Board of the Botanic Gardens and Reit State Herbarium

of South Australia Department of Environment, Water and Natural Resources

Journal of the Adelaide Botanic Gardens 27 (2014) 35-39

Notes on Hibbertia (Dilleniaceae) 10. Hibbertia hirsuta (subgen. Hemistemma: H. vestita group) does occur in South Australia and Tasmania

H.R. Toelken

State Herbarium of South Australia, PO. Box 2732, Kent Town, South Australia 5071 E-mail: hellmut.toelken@sa.gov.au

Abstract

A review of Hibbertia hirsuta (Hook.) Benth., which is prompted by its recent rediscovery in South Australia, includes a full description and a discussion of the importance of an understanding of the position and number of stamens in a flower to place the species with allied ones. Its affinities to H. dispar Toelken and H. basaltica A.M.Buchanan & Schah. are assessed and its

possible cleistogamy discussed.

Keywords: Dilleniaceae, Hibbertia, taxonomy, cleistogamy, South Australia, Tasmania.

Introduction

When Joe Quarmby consulted Bob Bates about a plant from the Meadows Creek tributary that he could not identify, the latter immediately recognised it as a new species for the South Australian flora. He soon confirmed its natural occurrence in the area by finding a second population of the species a few kilometres away along the creek. This new evidence was needed particularly as the species, Hibbertia hirsuta (Hook.) Benth., as it was identified, has lately been described as a Tasmanian endemic (Curtis 1956, Stone & Curtis 1971, Curtis & Morris 1975, Gray 2010). However, a closer examination of the literature shows that Mueller (1882, 1889) and Tate (1883, 1889, 1890) had already recorded its existence in South Australia, although all recent literature including three editions of the Flora of South Australia (Black 1926, 1952, Jessop 1986), have ignored it in spite of Tate (1883) citing details of a specimen(s) by J.G.O. Tepper and R. Tate. The only herbarium material found to date is a specimen collected by J.G.O. Tepper in 1881 (MEL119774), which has been overlooked for more than one hundred years, although it was correctly identified at the time. It 1s significant that Tepper also collected his specimen along Meadows Creek “on temporarily inundated flat’, while Quarmby (J/.Quarmby 189) describes it as “seasonally wet depression”. AHibbertia hirsuta is _ extremely localised in South Australia, while it is more widespread in Tasmania, but it has not been recorded from Victoria, which is geographically closer to and more similar in climate to Tasmania.

One reason for overlooking the species for so long may be its very small flowers, as reported by Joe Quarmby (pers. comm.). Hibbertia pallidiflora Toelken has similar small flowers, which also gradually open (Toelken 1995, 2000), but as plants of that species are

Published online: 29 Oct. 2014 « flora.sa.gov.au/jabg

larger and more common they are better represented in herbaria.

Floral biology

Field observations in South Australia did not yield the envisaged breakthrough in the understanding of the breeding system of the species, except that it showed a very high percentage of seed set. An examination of the morphology of the flower together with different possible scenarios of the pollination biology are ex- plored below. It is hoped that these will stimulate further interest, new observations and research into the problem.

Particularly important in HA. hirsuta is the reduction of the number of petals and stamens to usually one, rarely few. Similar to specimens from Tasmania, South Australian plants examined displayed short styles bearing slightly incurved stigmas, which are at the upper level of the anther(s). This flower arrangement differs from that of H. pallidiflora and the vegetatively similar H. sessiliflora Voelken.

The small flowers, “which often remain unopened” as Curtis (in Stones & Curtis 1971) described it, cause difficulties in interpreting the breeding system of the species. The evidence that flowers never open, Bob Bates (pers. comm.) reported them to be cleistogamous, is at present incomplete. They might only open for a short time or at hours not coinciding with visits by collectors, particularly as they grow in not easily accessible localities in South Australia.

The specimens show, though not clearly, the petal(s) protruding above the apex of the clasping calyx lobes, as also illustrated by Stones & Curtis (1971, pl. 92). These authors also reported a reduction of petals, but never the absence thereof. The photo in Wapstra et al. (2010, p. 107) shows one flower with the calyx partially opened.

ISSN 0313-4083 (Print) * ISSN 2201-9855 (Online)

H.R. Toelken

J. Adelaide Bot. Gard. 27 (2014)

Table 1. Morphological comparison of Hibbertia hirsuta, H. basaltica and H. dispar to show similarities and differences.

Characters

Hairs on branches

Leaf lamina

Hairs on leaves, upper surface

Hairs on leaves, lower surface

Outer calyx lobes

Outer calyx outside hairs

Inner calyx lobes

Stamen number and position around ovary

Anthers Seeds Habit

H. hirsuta

scattered larger and smaller fascicled hairs, rarely few simple hairs

(2.2—) 3.5—5 (—6.4) x (0.7—) 0.9-1.3 (—1.6) mm

scattered longer and shorter simple hairs

scarcely raised central vein and undersurface with + dense fascicled hairs

4-4.4 x 1.2-1.4 mm

strigose, Coarse simple over fascicled hairs

3.8—4 x 1.6—-1.8 mm

1 (2) on one side

0.8—1 mm long 1.8—2 x1.5—1.6 mm

prostrate to scrambling

H. basaltica

dense larger and smaller fascicled hairs and few longer simple hairs

(2.4—) 3.5—-6 (-7.4)

x 1-1.4 mm

scattered simple, rarely forked hairs

recessed central vein and undersurface with few simple over short fascicled hairs

4.5-4:7 5.1) « 2.3-25-mm

+ strigose, few simple over mainly fascicled hairs

4.3—4.7 (—4.8) x 2.5—2.8 mm (3-) 4 or 5 (-7) on one

and 1 on other side

1-1.2 mm long 1622 x 1.2—15-mim

prostrate to procumbent

H. dispar

dense larger simple over smaller simple and/or fascicled hairs

(1.8—) 2.5-6.5 (—10.4) x (0.8—) 1-1.3 mm

scattered simple hairs

flush/recessed central vein with mainly long over shorter simple hairs

(5.2—) 5.4—5.8 (—6.1) x 1.8-2.6 mm

+ strigose, few simple over mainly fascicled hairs

(5—) 5.2—5.6 (—6) x 2.6—3.2 mm (2—) 4-6 (7) on one

and (0) 1 or 2 on other side (1—) 1.1-1.4 mm long 1.7-2 x 1.4-1.6 mm

prostrate to decumbent

A flower of the South Australian plants examined displayed only one petal clasping a dehisced anther with some pollen in it and a cluster of pollen grains on each of the adjoining stigmas.

Plants of H. hirsuta show a very high degree of seed development, which could indicate selfing and, in fact, cleistogamous flowers, if the flowers can be demonstrated to remain closed. More field work is needed, as it 1s, for instance, not clear whether a polli- nating agent is still needed to burst the anthers by vibration in order to spread the pollen, as usual in Hibbertia species (Bernhardt 1986). The stigmas of flowers of the South Australian H. hirsuta examined were covered with pollen although not, or perhaps no longer, in contact with the anther. One could speculate on the continued presence of petal(s) inviting pollinating vectors to visit the flowers. Even though they are closed or partially closed (see Wapstra et al. 2010), visiting insects might provide the thoracic vibration required to dehisce the anther(s) to cause selfing/cleistogamy by violently distributing the pollen onto the adjacent stigmas. It is unlikely that such a visitor will forcefully open the flower, as pollen, the only reward in hibbertias, is sparse and was still found scattered in some flowers with developing seed.

In a detailed study of the floral ontogeny in 12 species of Hibbertia from different groups in the genus, Tucker & Bernhardt (2003) discussed the possible importance of the reduction of the number of stamens in specialisation and speciation in Hibbertia, with parti- cular reference to a form of H. fasciculata R.Br. ex DC., which has its anthers reduced to three (Tucker

36

& Bernhardt 2000). While this form of H. fasciculata differs only in the reduced number of stamens and flowers open normally, H. hirsuta displays also a reduction in the number of petals and, apparently, the flowers are not opening fully. This changes not only the floral biology, but has also obscured its affinity to the two related species (below) with anthers on both sides of the ovary. The specialisation of the breeding system of H. hirsuta needs further examination, as discussed above. The possibility that this is an inbreeding species, makes the disjunct, but morphologically indistinguishable populations of H. hirsuta in Tasmania and South Australia even more remarkable.

Relationships

A number of smaller species of Hibbertia, e.g. H. humifusa F.Muell. and HA. sessiliflora, with a superficially similar habit and morphology, occur throughout the genus. Raheem (2013, pp. 3-9, 3-10) placed AH. hirsuta next to H. dispar TYoelken in his molecular cladogram (ITS data) of the eastern Australian species. Hibbertia basaltica A.M.Buchanan & Schah, a very similar species to H. dispar (cf. Toelken 2013), was not included in his study. This close affinity of the three species is not only reflected in their very similar morphology (Table 1), but also in a progressive loss of stamens, especially of the anterior ones closest to the bract. Furthermore, the number of petals in H. hirsuta is also reduced to one, rarely two. Toelken (2013) had for that reason not included AH. hirsuta in his treatment of the H. vestita group, because it did not conform to the criterion of the group, 1.e. stamens on both sides of

J. Adelaide Bot. Gard. 27 (2014)

the ovary. He, however, discussed the wide range of variation, particularly in the number of stamens found in H. dispar.

Hibbertia basaltica and H. dispar are found in eastern Victoria and Tasmania, H. hirsuta in South Australian and Tasmania. Herbarium specimen data indicate that South Australian plants of H. hirsuta have been recorded only from temporarily wet areas in forest, while Tasmanian plants occur in drier habitats, such as, “amongst rocks in open sclerophyll forest” (RD. Hoogland 11748), or, as Gray (2010) describes it, “in dry grassy places, open heath and woodlands”. The very similar H. basaltica, as well as H. dispar, 1s also mainly recorded from rocky or gravelly outcrops; Toelken (2013) cited a number of different habitats for different populations of H. dispar, among them there are forms from Flinders and Clark Island in Bass Strait, “recorded from wet areas near a lagoon” and “in sedgeland on marshy flats”.

Taxonomy

Hibbertia hirsuta (Hook.) Benth.

Fl. Austral. 1: 26 (1863); Spicer, Handb. Pl. Tasmania 100 (1878); F.Muell., Census Pl. Tasmania 3 (1880); F.Muell., Syst. Census 1: 1, 141 (1882), includes S.A.; Tate, Trans. Proc. Rep. Roy. Soc. S. Austral. 6: 111 (1883), incl. S.A.; F.Muell., Second Syst. Census 1: 1 (1889), incl. S.A.; Tate, Fl. Extratrop. S. Austral. 14, 205 (1890), incl. S.A.; Tate, Trans. Proc. Rep. Roy. Soc. S. Austral. 12: 70 (1889), incl. S.A.; Rodway, Tasmanian Fl. 4 (1903); W.M.Curtis, Student Fl. Tasmania 1: 21 (1956); M.Stones & W.M.Curtis, Endemic Fl. Tasmania 3: 172, t. 92 (1971); W.M.Curtis & D.I.Morris, Student Fl. Tasmania edn 2, 1: 22 (1975); M.Wapstra et al., Tasmanian Pl. Names Unravelled 107, fig. (2010); A.M.Gray in M.Duretto (ed.), Fl. Tasmania Online: Dilleniaceae (2010). — Pleurandra hirsuta Hook., Companion Bot. Mag 1: 273 (1836); Hook.f., J. Bot. (Hook.) 2: 42 (1840); Walpers, Repert. Bot. Syst. 1: 64 (1842); Hook.f., Fl. Tasmania 1: 17 (1855). — Type citation: “Mr. Gunn (n. 445) — Gathered on dry hills, near Hobart Town, by Mr. Backhouse”. Lectotype (designated here): Tasmania, Hobart Town, dry hills, J.Backhouse sub Gunn 445 (K000687477!). Remaining Syntype: Vian] D[iemen’s] Land, [Tamar River, Whirl- pool Reach, Sep. 1835,] R.C.Gunn 445 (K000687476!).

Suffrutex up to 0.15 m tall, prostrate to scrambling, moderately branched; branches thin-wiry, with leaf bases scarcely raised and scarcely decurrent, pilose but soon glabrescent. Vestiture on many parts persistent, sparse, consisting of fascicled and/or longer simple hairs; on branches sparse but denser below the inflorescence, a range of larger and smaller erect-spreading fascicled hairs (S—10 often unequally long arms) overtopped by scattered and on some branches very few porrect simple hairs, usually wearing off soon; on leaves above sparse, with scattered long and short erect hairs and rarely with a few forked hairs on the flanks of the revolute margins; on leaves below sparse with few fascicled hairs under scattered erect long and short simple hairs on the revolute margins and with dense spreading fascicled

SY

Notes on Hibbertia 10 (Dilleniaceae)

hairs (8—12 often unequal arms on the central vein but + similar on the undersurface) and usually without simple hairs; on bracts moderately dense, above and below with few small fascicled hairs overtopped by longer antrorse simple hairs; on outer calyx lobes moderately dense, outside with short fascicled hairs (2-4 usually unequal arms) overtopped by coarse antrorse simple hairs, inside moderately dense with finer forked hairs overtopped by coarser antrorse simple ones; on inner calyx lobes outside sparse, with scattered coarser simple hairs mainly along the central ridge over more widespread finer forked hairs, but glabrous membranous margins, inside glabrous. Leaves with pronounced intrapetiolar tuft of hairs up to 1 mm long and often + decurrent on both sides of the petiole; petiole O—-0.4 mm long; lamina narrowly elliptic to rarely linear, (2.2—) 3.5-5 (-6.4) x (0.7—) 0.9-1.3 (—1.6) mm, gradually tapering into petiole, acute, rarely obtuse, above grooved to incurved and pilose, below exposing a tomentose undersurface between the slightly revolute margins and a narrow central vein, which is often visible to the leaf apex. Flowers sessile, single, terminal on terminal and axillary short shoots on distal branches; buds ellipsoidal; bracts lanceolate to almost triangular, 2—2.3 x 0.5—0.6 mm, bluntly acute, pubescent. Calyx unequally long and accrescent; outer calyx lobes 3, lanceolate, 44.4 x 1.2— 1.4 mm, slightly longer than inner ones, acute, without ridge or recurved distal margins, outside strigose, inside strigose on upper half; inner calyx lobes broadly elliptic to obovate, 3.8—4 x 1.6—1.8 mm, obtuse, with + broad membranous margins, outside pubescent, inside glabrous. Petal apparently only one, linear, with incurved apex, 3.2-3.7 mm long, shielding stamen. Stamens 1, rarely 2; filaments stiffly erect; anthers narrowly obloid and scarcely broader than and + continuous with filament, 0.8—1 mm long and positioned with apex between the stigmas. Pistils 2; ovaries obovoid, with 2 ovules per ovary and attached above one another, pubescent; erect styles attached to the apex and curved inwards (sideway) to place stigmas next to anther. Seeds obloid to obovoid, 1.8—2 x 1.5—1.6 mm, brown; aril scarcely fleshy, with slightly lobed margins of the membranous cup covering the lower half of the seed.

Distribution and ecology. Known in South Australia (Southern Lofty region) from “clay soils on flats near creek line with Eucalyptus viminalis subsp. cygnetensis, E. obliqua, E. camaldulensis over Acacia melanoxylon, A. verticillata, Exocarpus cupressiformis, Leptospermum continentale, Microlaena stipoides, Goodenia ovata, Hibbertia crinita, Haloragis heterophylla’ (J.Quarmby 189) and from “heavy grey clay with seasonal wet depressions with Eucalyptus ovata over Leptospermum continentale, Banksia marginata, Gahnia_ sieberiana, Pteridium esculentum, Chorizandra enodis, Haloragis heterophylla and Gonocarpus tetragynus” (J.Quarmby 188), while Gray (2010) recorded that south-eastern Tasmanian (TSE region) plants were growing “in dry grassy places, open heath and woodlands”.

H.R. Toelken J. Adelaide Bot. Gard. 27 (2014)

P

ie? on

‘ah ROYAL BOTANIC ¢ GARDENS KEW

S MOE ?

\ K000687478 ] iN

™ com

é # i Rh

Ric atat i 9 {SJ he * ty c } . : -| | sg) N a ROYAL BOTANIC GARDENS KEW PF om . SKE: XQ ROYAL = BOTANIC GARDENS KEW

Tn OS ee WOU K000687477 tee nn [ ? . 4 K000687476 mar =<] os ag woe Peete pt |

\ A Deas bot, —aee,|

tS? ZZ \ sa Bin, ey, ' eal J : ee ee mee oa 7)

- » -—f P Ss mad ” = F- 4 f é,* of - ” — al r Ss ee “of 2. £ <f@¢e? seu wm, PROG Lit a

a2 * 7 a - - ; oe, él Se = Ae < nf yd en eee at. 2 ae Pie Ha wo, es ee

; ° ‘7 > we - a <> s / ras y i ‘. _ - ~~ oF . ‘ ~- ——_ . —_ : “ ye te # Ci e “A rg J LA ‘ Na” ‘ a a oa S 2 a # a = ’ se AS « ‘ * a ?. i _ : a ~s - : | «* ' - . 7 « f ~ ’ : - ra : < . - are cw<re er « .'-% wo co’ tier an - = ras J -_ ie ot ne ~ “ =o ~tous ae - 7 : * 4 ; = we - il & wy 7 - pf wt f . <" * - Ae a Ts ” ri # ad - Bg fz o a - ssf “4 eo t- . y a -_ 5 « a of — <> s we

- - ¢ a * 7 ~~ oa. - iv YX. ~* a ei» 4 , ¥ . a - = Fee i‘ = j i

" .

. - A P ¥ a - - . Ae 7 « yw Sas "A ” ri a wt 4 Oe et nl ee ee Se bags San i me. weet me -ees - : - - - r x 7 : yf ‘ ’ ss a St a PA . a J . - r r , . ~ - a a - & "9 Sf 4 - - Fy of go tre F - | — “ a 7? re we - - sa oe « er creer fee nT - ; © a we a, z os “a , - q - r fl —

“a : ed ‘ — ' a.

- —_~ PLORA AUS ETRALLENS IS,

| : , muned by Ar KENT AM = bie ra ih lx hk

Fig. 1. Type sheet of Hibbertia hirsuta: Lectotype indicated in red (000687477: . J.Backhouse sub Gunn 445), The remaining syntype, indicated in yellow (K000687476: R.C. Gunn 445), is accompanied with a note by Gunn: “Sent in 1835 as received from Mr. Backhouse / which he collected at Hobart Town. — Those now sent / are not in flower, but were gathered by myself / at Whirlpool-reach on the Tamar early in Sept." / 1835 — growing on a dry hill-side, — the habit / evidently dwarf and procumbent, — or prostrate. —-". — Reproduced with kind permission of the Board of Trustees of

the Royal Botanic Gardens, Kew.

38

J. Adelaide Bot. Gard. 27 (2014)

Conservation status. Very rare in South Australia; “scattered but not common” in Tasmania (Gray 2010).

Diagnostic features. Although the vegetative organs are similar to several small species of Hibbertia, the flowers are easily distinguished by the reduction of the number of petals and stamens to one or two.

Variation. Except for some range of variation in the denseness of the vestiture in local forms, as for instance in north-eastern Tasmania, where it 1s much denser, H. hirsuta shows remarkably little variation for a suspected inbreeding species.

Typification. A number of early specimens of this species have been accumulated on two sheets at K. On sheet A (Fig. 1), the specimen collected by J. Backhouse (left centre; inscribed “445. Dry Hills. Hobart Town”; K000687477), which is cited in the protologue of Pleu- randra hirsuta Hook., 1s here selected as lectotype, as it is flowering. The specimen gathered by R.C. Gunn (right centre; KO00687476) is a syntype. It was collected at Whirlpool Reach near Launceston in Sep. 1835 and was sent to W.J. Hooker some time after the Backhouse specimen (see note by Gunn; Fig. 1), but both specimens were presumably available to W.J. Hooker when describing the species in Apr. 1836. Collections by J.D. Hooker on the type sheet, and three collections of Gunn collected in the 1840s, mounted on sheet B, are irrelevant to the typification.

Specimens examined

SOUTH AUSTRALIA: R. Bates 84343, Kuitpo Forest, 5.x.2011 (AD, HO, MEL); J. Quarmby 188, Meadows Creek tributary, 27.x.2011 (AD); J,Quarmby 189, Kuitpo, 27.x.2011 (AD); J.G.O.Tepper MEL1I1I97774, Meadows Creek, 10.x11.1881 (MEL).

TASMANIA: M.Allen HO758&47, Stringy Bark Gully, Warrane, 12.x.1975 (HO); M/Allen HO76346, Proctors Road, 25.1x.1975 (HO); A.M. Buchanan 560, Finger Post Hill, 7.x.1981 (HO); J. Bufton MEL695427, Port Arthur, 1894 (MEL); PCollier 134, S Blackmans Bay, 26.x1.1984 (HO); P.Collier 1770, near Molesworth, 23.x.1986 (HO); P Collier 4308, 6 km SW Tooms Lake, 14.x.1989 (HO); WM Curtis HO3114, Conara, Tunbridge, 20.x1.1952 (HO); A.V.Giblin HO3120, Mt Nelson Road, 15.x.1929 (HO); Gunn 445/184], Grass Tree Hill, Hobart, 12.x.1840 (K); Gunn 445/1842, Hobart Town, 1.1839 (K); Gunn 445/23.x.1S544, George Town (K, NSW); M Hart HO29331, Sandy Bay, 17.x1.1947 (HO); R.D.Hoogland 11748, summit of Mt Nelson, 14.1.1970 (CANB, K); 1. Murfet HO3115, c. 300m SE Mt Nelson Signal Station, 30.x.1957 (HO); L.Rodway 13, Queens Domain, Hobart, 15.x1.1897 (HO); L.Rodway HO311I9, Mt Nelson Range, x.1917 (HO); L.Rodway NSW243046, Bellevive, 5.1x. 1892 (NSW).

Acknowledgements

I am grateful to Joe Quarmby and Bob Bates for providing additional information and photographs to complete this paper.

Thanks are also extended to the following herbaria for the loan of specimens and/or assistance received while working there: BM, BRI, CANB, HO, K, MEL & NSW.

ay

Notes on Hibbertia 10 (Dilleniaceae)

References

Bentham, G. (1863). Dilleniaceae. In: Flora Australiensis, a description of plants of the Australian territory |: 26. (Reeves: London).

Bernhardt, P. (1986). Bee-pollination in Hibbertia fasciculata (Dilleniaceae). Plant Systematics and Evolution 152: 231- 241.

Black, J.M. (1926). Dilleniaceae. In: Flora of South Australia 3: 386-389. (Government Printer: Adelaide).

Black, J.M. (1952). Dilleniaceae. In: Flora of South Australia edn 2, 3: 574-577. (Government Printer: Adelaide).

Curtis, W.M. (1956). The students flora of Tasmania |: 21. (Tasmanian Government Printer: Hobart).

Curtis, W.M. & Morris, D.I. (1975). The student's flora of Tasmania edn 2, 1: 22. (Tasmanian Gov. Printer: Hobart).

Gray, A.M. (2010). Dilleniaceae. In: Duretto. M.F. (ed.). Flora Tasmania Online. (Tasmanian Herbarium, Tasmanian Museum & Art Gallery: Hobart). www.tmag.tas.gov.au/ floratasmania [accessed 28 July 2014].

Jessop, J.P. (1986). Dilleniaceae. In: Jessop, J.P. & Toelken, H.R. (eds), Flora of South Australia edn 4, 1: 354-358. (Government Printer: Adelaide).

Mueller, F. (1882). Systematic Census of Australian Plants |: 1, 141. (Victorian Government: Melbourne).

Mueller, F. (1889). Second systematic Census of Australian Plants |: 1. (Victorian Government: Melbourne).

Raheem, |.A.A.A. (2012). Phylogeny of the SE Australian clade of Hibbertia subg. Hemistemma (Dilleniaceae). Unpublished Ph.D. Thesis. (The University of Adelaide: Adelaide).

Stones, M. & Curtis, W.M. (1971). The endemic flora of Tasmania 3: 172, pl. 92. (The Ariel Press: London).

Tate, R. (1883). Additions to the flora of South Australia. Transactions & Proceedings of the Royal Society of South Australia 6: 111.

Tate, R. (1889). A census of the indigenous flowering plants and vascular cryptogams of Extra Tropical South Australia. Transaction & Proceedings of the Royal Society of South Australia \2: 70.

Tate, R. (1890). Flora of Extratropical South Australia, pp. 14, 205. (Education Department: Adelaide).

Toelken, H.R. (1995). Notes on Hibbertia (Dilleniaceae). 1. New taxa from south-eastern Australia. Journal of the Adelaide Botanic Gardens 16: 59-72.

Toelken, H.R. (1998). Notes on Hibbertia (Dilleniaceae). 2. The Hibbertia aspera-empetrifolia complex. Journal of the Adelaide Botanic Gardens 18: 107-160.

Toelken, H.R. (2000). Notes on Hibbertia (Dilleniaceae). 3. H. sericea and associated species. Journal of the Adelaide Botanic Gardens 19: \1—54.

Toelken, H.R. (2013). Notes on HAibbertia subgen. Hemi- stemma (Dilleniaceae) 9. The eastern Australian H. vestita eroup, including H. pedunculata and H. serpyllifolia. Journal of the Adelaide Botanic Gardens 26: 31-69.

Tucker, S.C. & Bernhardt, P. (2000). Floral ontogeny, pattern formation, and evolution in Hibbertia and Adrastaea (Dilleniaceae). American Journal Botany 87 (12): 1915-— 1936.

Tucker, S.C. & Bernhardt, P. (2003). Developmental evidence for stamen number reduction in populations of Hibbertia fascicularis (Dilleniaceae). Plant Systematics and Evolution 240: 107-114.

Wapstra, M., Wapstra, A. & Wapstra, H. (2010). Zasmanian plant names unravelled, p. 107. (Fullers Bookshop: Launceston).

J OQURNAL of the ADELAIDE

BOTANIC GARDENS

AN OPEN ACCESS JOURNAL FOR AUSTRALIAN SYSTEMATIC BOTANY

flora.sa.gov.au/jabg

Published by the STATE HERBARIUM OF SOUTH AUSTRALIA on behalf of the BOARD OF THE BOTANIC GARDENS AND STATE HERBARIUM

State Herbarium of South Australia PO Box 2732 Kent Town SA 5071 Australia

Board of the Botanic Gardens and Reit State Herbarium

of South Australia Department of Environment, Water and Natural Resources

Journal of the Adelaide Botanic Gardens 27 (2014) 41-45

Additions to the genus Lecidella (lichenised Ascomycetes: Lecanoraceae)

Gintaras Kantvilas? & John A. Elix?

“ Tasmanian Herbarium, P.O. Box 5058, UTAS LPO, Sandy Bay, Tasmania 7005 Email: Gintaras.Kantvilas@tmag.tas.gov.au

’ Research School of Chemistry, Building 137, Australian National University, Canberra, Australian Capital Territory 0200 Email: John.Elix@anu.edu.au

Abstract

Two new taxa in Lecidella K6rb. are described: L. leucomarginata Kantvilas & Elix, from Kangaroo Island (South Australia) and south-western Western Australia and L. granulosula var. lecanorina Kantvilas & Elix, from Kangaroo Island (South Australia) and the coast of southern New South Wales. Both display the unusual feature of an unpigmented proper excipulum, densely inspersed with crystals. A key to Lecidella in temperate Australia is presented.

Keywords: biodiversity, Kangaroo Island, Lecidea, lichens, temperate Australia.

Introduction

Lecidella K6rb. 1s a genus of lichens characterised by a crustose thallus containing a trebouxioid photobiont, biatorine, typically dark coloured apothecia with a persistent proper excipulum composed of radiating thick-walled hyphae, clavate, amyloid, eight-spored asci referred to as Lecidella-type (Hafellner 1984; Fig. 2), an hymenium that typically separates readily in KOH, simple, hyaline, non-halonate ascospores, and curved, filiform conidia; thallus chemistry is dominated by xanthones in the majority of species (Fletcher et al. 2009; Kantvilas & Elix 2013). With 12 species (McCarthy 2014), Lecidella is well-represented in the Australian lichen flora, occurring mostly in temperate latitudes on rock, bark or wood. The eight Tasmanian species were recently revised by Kantvilas & Elix (2013), who also studied comparative material from elsewhere in southern Australia. Morphological and anatomical relationships with superficially similar genera, such as Japewiella Printzen, Carbonea (Hertel) Hertel, Yasmidella Kantvilas, Hafellner & Elix and others, were discussed therein and are not repeated here. In this paper, we describe two further taxa from temperate latitudes.

Methods

The study is based chiefly on collections of the authors, housed in the Tasmanian Herbartum (HO) and the Australian National Herbarium (CANB). Descriptions are based on hand-cut sections of the thallus and ascomata examined with high-power light microscopy. Mounting media included water, 15% KOH (K), Lugols Iodine after pretreatment with K

Published online: 29 Oct. 2014 « flora.sa.gov.au/jabg

(IKI), ammoniacal erythrosin and 50% HNO, (N). Dimensions of asci and ascospores are based on 30 and 70 observations respectively. The latter are presented in the format: 5" percentile-average—95" percentile; outlying extreme values are given in parentheses. Chemical constituents were identified by thin layer chromatography (Elix & Enrnst-Russell 1993) and comparison with authentic samples. Nomenclature of pigments follows Meyer & Printzen (2000).

Taxonomy

Lecidella leucomarginata Kantvilas & Elix, sp. nov.

Lecidellae flavovirenti Kantvilas & Elix fortasse simi- lissima, sed sorediis destitutis, excipulo proprio hyalino vel pallide griseo, pigmentum deficienti, sed crystallis insperso, et ascosporis aliquantum parvioribus, 10-15 um longis, 6-8 um latis, differt.

Mycobank No.: MB810540.

Typus: SouTH AUSTRALIA. Kangaroo Island, Western Cove, 35°44’S 137°35’E, 0.5 m alt., on bleached dead wood of Melaleuca in salt marsh, 27 Sep. 2013, G. Kantvilas 234/13 (holo: HO; iso: AD, BM).

Thallus crustose, effuse, bright lemon-yellow, rimose-areolate, esorediate, lacking a_ prothallus, forming irregular, undelimited patches to 12 cm wide or more, 0.25—1 mm thick; individual areoles with upper surface unevenly verruculose and with edges mostly lifting from the substrate; photobiont cells globose, 6-18 um diam. Apothecia biatorine, abundant, 0.5— 1.3 mm wide, sessile, basally constricted to adnate, sometimes sunken in the thallus surface, scattered and roundish, or crowded, misshapen and fused together; disc plane at first, soon becoming convex, grey-black

ISSN 0313-4083 (Print) * ISSN 2201-9855 (Online)

G. Kantvilas & J.A. Elix

J. Adelaide Bot. Gard. 27 (2014)

Fig. 1. Habit of new Lecidella taxa. A L. leucomarginata; B L. granulosula var. lecanorina. Scale bar: 1 mm. — A G. Kantvilas 234/13 (holotype),

B G. Kantvilas 375/13 (holotype).

to bluish black to jet black, sometimes a little mottled, matt, epruinose. Proper excipulum colourless to pale erey or yellowish grey, rarely dark grey, sometimes piebald, glossy, sometimes a little flexuose, persistent but increasingly less obvious in older, more convex apothecia, sometimes with adhering thallus fragments, in section 25—60 um thick, lacking blue-green pigments, composed of radiating, branched hyphae 3—5 um thick, densely inspersed, especially towards the outer edge, with minute, yellowish brown to reddish brown crystals that fluoresce yellow in polarised light and dissolve fleetingly yellowish in K. Hypothecium 80-150 um thick, colourless to pale yellow-brown, occasionally yellowish orange, more intensely coloured towards the base, intensifying yellow-orange in K and N, sometimes inspersed with scattered oil droplets. Hymenium 60—70 um thick, separating readily in K, in the upper part usually dilutely greenish blue, K + grey-green, N+ crimson and overlain by yellowish brown crystals that dissolve in K, towards the base mostly colourless. Asci 45-55 x 12-20 um, of the Lecidella-type. Paraphyses 1.5—2 um thick, sparsely branched; apices not markedly capitate, sometimes expanded to 2.54 um, unpigmented or with a thin internal cap and a faint external blush of blue-green, N+ crimson pigment. Ascospores hyaline, (9—) 10—/2.4—-15 (—16) x (5.5—) 6—-6.5—8 um, broadly ellipsoid to oblong. Conidiomata pycnidia, black, emergent, resembling apothecial initials, in section with wall blue-green, N+ crimson; conidia thread-like, curved, 15—20 x 1 um. Fig. 1A.

42

Chemical composition. Vhiophanic acid (major), 3-O- methylthiophanic acid (+ minor), atranorin (+ minor), chloroatranorin (+ minor), arthothelin (4 minor or trace); thallus K—, C+ orange, KC-—, P—, UV+ orange.

Etymology. The specific epithet refers to the distinctive colourless proper excipulum of the apothecia of this species.

Remarks. This 1s a distinctive and remarkable species on account of its colourless proper excipulum, a feature also seen in L. granulosula var. lecanorina (described below). One of the defining generic characters of Lecidella is the black apothecial margin, comprised of radiating, thick hyphae coated in dark pigments and becoming swollen and almost parenchymatous towards the outer edge (Kantvilas & Elix 2013). In that regard, L. leucomarginata might appear to be misplaced in Lecidella. Yet, all the other salient characters, notably the anatomy of the thallus, its chemical composition consisting of xanthones, the distinctive ascus type, the hymenium that separates readily in KOH, the hyaline, broadly ellipsoid ascopores and the thread-like, curved condia, are consistent with Lecidella. Indeed, on close inspection, the anatomy of the excipulum can also be interpreted as being of the Lecidella-type, except that instead of being swollen with pigment, the hyphae are densely inspersed with crystals. However, there is no hint of any development of a parenchyma at the outer edge. Traces of blue-green (cinereorufa-green) pigment, characterisitic of most Lecidella species, can be detected in the hymenium by the application of HNO,

J. Adelaide Bot. Gard. 27 (2014)

which yields a crimson reaction. Perhaps the most likely alternative generic placement for this species that was considered was Japewiella. However, this genus has a very distinctive excipulum composed of a loose reticulum of radiating, branched and anastomosing hyphae c. | um thick in a gelatinous matrix (Printzen 1999; Kantvilas 2011).

In the Australian flora, this new species appears to be most closely related to L. flavovirens Kantvilas & Elix, a corticolous species that likewise has a bright lemon-yellow thallus containing thiophanic acid as the major constituent, but which differs chiefly in becoming entirely sorediate. These species differ further in that L. flavovirens has unequivocally black apothecia, and although the proper excipulum lacks blue-green pigment, it is constructed in the typical Lecidella way, with the hyphae instead swollen with a dark brown pigment. In addition, the ascospores of L. flavovirens are also slightly longer and broader: (12—) 13-17 (18) x (6—) 7-10 um (Kantvilas & Elix 2013).

Ecology and distribution. On Kangaroo Island, this new species 1s known only from the type locality, where it was collected from remnant, fire-killed, bleached Melaleuca stags (several metres tall) 1n a salt marsh. The site is extremely degraded by draining, burning, some clearing and stock grazing. The habitat of this Lecidella is very species poor with respect to lichens and the only other species with which it was associated was Ramboldia crassithallina Kalb, a common lignicolous lichen on Kangaroo Island. In the near vicinity of the type locality where living individuals of Melaleuca remain, the epiphytic flora is richer and includes species of Ramalina, Usnea, Lecanora and Parmotrema, but no Lecidella was encountered. Lecidella leucomarginata also occurs in south-western Western Australia where it erows on the bark of Acacia and Melaleuca in open, dry sclerophyll woodland. Here associated species included Austroparmelina pruinata (Mull. Arg.) A.Crespo, Divakar & Elix, Baculifera xylophila (Malme) Marbach, Buellia reagenella_ Elix, Flavoparmelia rutidota (Hook.f. & Taylor) Hale, Haematomma eremaeum R.W.Rogers, Lecanora caesiorubella Ach., Pertusaria subarida A.W.Archer & Elix, Punctelia subalbicans (Stirt.) D.J.Galloway & Elix and Ramalina inflata subsp. australis G.N.Stevens.

Additional specimens examined

WESTERN AUSTRALIA. Wotto Nature Reserve, First North Road, 21 km NE of Eneabba, 29°42’29”S 115°24’37”E, 275 m alt., 5 May 2004, JA. Elix 28868 (CANB, HO, PERTH); same locality, 5 May 2004, JA. Elix 28875 (CANB); Gwambygine Nature Reserve, 11 km S of York, 31°58’24”S 116°48’38”E, 245 m alt., 4 Apr. 2006, JA. Elix 31756, 37406 (CANB); Walebing, Quarrell Range, Moora-New Norcia Road, 22 km S of Moora, 30°41°38’S 116°12’20”E, 275 m alt., 2 Apr. 2006, JA. Elix 37549 (CANB); Fourteen Mile Brook Nature Reserve, 13 km NW of Narrogin along Wandering Road, 32°50’°08"S 117°06’07”E, 335 m alt., 6 Apr. 2006, JA. Elix 43229 (CANB).

43

Two new taxa in Lecidella (Lecanoraceae)

Lecidella granulosula var. lecanorina Kantvilas & Elix, var. nov. Respectu praesentiam chodatini, hypothecium incolo- ratum et ascosporas 9.5—14.5 um longas, 5—7.5 um latas, varietati granulosulae similis, sed excipulo proprio hyalino vel pallide griseo, pigmentum destituto, crystallis flavo-brunneis dense insperso differt. Mycobank no.: MB810541. Typus: SourH AUSTRALIA. Kangaroo Island, Creek Bay Farm, 35°50’S_ 138°06’E, 85 m alt., on rock in mallee woodland, 12 Sep. 2013, G. Kantvilas 375/13 (holo: HO).

Thallus crustose, pale yellowish grey to greyish ereen, minutely granular, esorediate, lacking a prothallus, forming small, irregular, undelimited “islands” 2—3 cm wide amongst other lichens; individual granules mostly 0.05—0.1 mm wide, dispersed or contiguous; photobiont cells globose, 6-18 um diam. Apothecia biatorine, 0.4—0.7 mm wide, sessile, basally constricted, scattered, roundish; disc plane at first, soon becoming undulate or convex, grey to brownish grey to grey-black, sometimes a little mottled, matt, epruinose. Proper excipulum colourless to pale grey or yellowish grey, sometimes partly or wholly dark grey, glossy, persistent, in section 40-60 um thick, lacking blue-green pigments, composed of radiating, branched hyphae 4—5 um thick, densely inspersed, especially towards the outer edge, with minute, yellowish brown to reddish brown crystals that fluoresce yellow in polarised light and dissolve to form a pale yellow solution in K. Hypothecium 50-100 um thick, colourless to pale yellow-orange especially at the base, intensifying yellow-orange in K and N, not inspersed. Hymenium 50—75 um thick, mostly separating readily in K, in the upper part overlain by a thick layer of yellowish brown granules that dissolve in K, usually also with patches of greenish blue, K + grey-green, N+ crimson pigment, especially adjacent to the excipulum, towards the base mostly colourless. Asci 50-60 x 12—20 um, of the Lecidella-type. Paraphyses 1—2 wm thick, mainly simple, occasionally sparsely branched towards the apices; apices variable, not capitate or expanded to 3.5—4.5 um, unpigmented or with the apical cell with an internal, blue-green cap, and the subterminal cell faintly blue-green, N+ crimson. Ascospores hyaline, (8—) 9.5— 11.8—14.5 (—16) x 5—6.4-7.5 (—8) um, broadly ellipsoid to oblong. Conidiomata not seen. Fig. 1B, 2.

Chemical composition. Chodatin (major), 1soarthothelin (minor), thiophanic acid (minor), 2,5,7-trichloro- lichexanthone (minor); thallus K—, C+ orange, KC-, P-, UV+ orange.

Etymology. The infraspecific epithet alludes to the fact that the apothecia, with their pale coloured proper exciple, somewhat resemble those of a Lecanora.

Remarks. The granular to areolate thallus containing chodatin as a major compound (an uncommon sub- stance, not least in Lecidella), the predominantly colour- less hypothecium and the relatively small ascospores indicate unequivocally that this new lichen is closely

G. Kantvilas & J.A. Elix J. Adelaide Bot. Gard. 27 (2014)

Key to Lecidella in temperate Australia!

1. Growing on rock or on bryophytes overgrowing rock 2. Hypothecium yellow-brown to dark brown, with the colour intensifying in K and N a. Vicario present: diploicin and thurine1one-absent. «.-.. 60s 6 ak viene aea edad notaes te a ee eee L. sublapicida 3: Vicanicm absent: cdiplorcin anid: Thurinetone presente. yn gate oi pk ee te ee es ee eR gin Shon’ eae a de L. carpathica 2: Hypothecium colourless or at most pale yellowish brown at the base 4. Thallus rimose-areolate, usually containing atranorin (K+ yellowish) together with other compounds 5. Thallus areoles rather lumpy and verruculose, lacking a prothallus, containing atranorin and zeorin

(sometimes only in trace amounts); apothecia 0.4—1.5 mm wide; mostly on calcareous or nutrient- CHTICHEC'SUBSHPALA cg o oes 12 Re ES ee 2 Pa IS ae MAHESH EDR ESR PARA pA Ta OS RP HEE RE EE SRS L. stigmatea

5: Thallus areoles + plane, growing on a blackish, effuse prothallus; apothecia to 0.8 mm diam.; on siliceous substrata 6. Diploicin and/or caloploicin, + thiophanic acid present; arthothelin and thuringione absent ...... L. buelliastrum 6: Diploicin, caloploicin and thiophanic acid absent; arthothelin and thuringione present ......... L. enteroleucella 4: Thallus granular, lacking atranorin but containing chodatin 7. Proper excipulum black, in section densely infused with blue-green pigment and lacking crystals . . birds him rlast a ites ¥clee a abderb arty netted detip dats teins iabee ube a anlar at beh dutauteate a eat batechans L. granulosula var. granulosula 7: Proper excipulum colourless to pale yellowish grey or mottled dark grey, in section lacking blue- ereen pigment but densely inspersed with crystals that dissolve pale yellow in KOH . L. granulosula var. lecanorina 1: Growing on bark or wood 8. Thallus not sorediate 9, Thallus containing atranorin only (C—); hymenium inspersed with oil droplets and crystals that do not CUSSOING ISIE sired eaters i'd, & Mes lat «Boe SRAM severely esa REE yer. bing ANE CD Wyong afta Mar aad we Ge MRA PER Ane EONS te dN L. destituta 9: Thallus containing xanthones (C+ orange); hymenium not inspersed, or at most with an epihymenial layer of granules that dissolve in K 10. Hypothecium colourless to pale brownish; apices of paraphyses with an internal cap of pigment ...... L. xylogena

10: Hypothecium pale yellow-brown to yellow-orange, colour intensifying in K; apices of paraphyses unpigmented or with an external cap of blue-green, N+ crimson pigment

11. Ascospores 15—24 x 8—13 um; excipulum in section mainly opaque brown, with blue-green, N+

CTIMSOM PISIMGIE Abie SA OSS. ce sinew see cocks eps wide Wd bowls Hd nd mais od Oh Som eng ea db Ay oe sae Bas L. montana 11: Ascospores 9-15 x 5—8 um; excipulum in section mainly colourless to dilute brown, with blue- Sree. NP CFIISON: PISNIE ALTNS- CAPSS 26.4 cae sce Se bie SG pd thie Rew oe Daw EE We L. elaeochroma

8: Thallus sorediate

12. Proper excipulum colourless to pale yellowish grey or mottled dark grey, in section lacking blue-green pigment but densely inspersed with crystals that dissolve pale yellow in KOH .................. L. leucomarginata

12: Proper excipulum black, in section lacking crystals, opaque dark brown or with blue-green pigment.

13. Ascospores 12—18 x 6—10 um; apices of paraphyses unpigmented or with an external cap of blue- ereen, N+ crimson pigment; excipulum in section opaque dark brown, lacking blue-green pigment; SOREMIA ALESIS Hh CISCRELS: SORTA 6 cn ee beta. eth ARN wld Ape RN 9) 4: Wa SECA a, OPER WH AO NE SD ha, WO SPE L. flavovirens

13: Ascospores 7-14 x 4.5—9 um; apices of paraphyses with an internal cap of pigment; excipulum in section grey-green to pale brownish, usually with blue-green pigment towards the outer edge; soredia not in discrete soralia, arising from a dissolution of the thallus ................ 00.000 L. xylogena

‘Australian records of L. asema (Nyl.) Knoph & Hertel (Elix 2008, 2010) are misdeterminations of L. flavovirens.

related to L. granulosula (Nyl.) Knoph & Leuckert as being correlated with an absence of pigment or that he delimited and described by Kantvilas & Elix (2013), | saw any specimens with an unpigmented excipulum. No and by Knoph (1990) and Rambold (1989) (under its — crystals were observed in any Australian specimens of synonym L. chodati). Yet the pale, unpigmented proper — var. granulosula studied; we were unable to study any excipulum of the specimens studied is so distinctive and non-Australian material. In several species of Lecidella, visually striking, and unusual for the genus as a whole — specimens from exposed habitats tend to have more (see also discussion under L. /eucomarginata above) intensely pigmented apothecia. In the case of var. that we feel that they deserve taxonomic recognition. /lecanorina, specimens from exposed habitats certainly Specimens of var. granulosula invariably contain have darker apothecial discs, but this does not translate cinereorufa-green pigment in their excipulum, which — into a higher concentration of blue-green pigment; the is constructed of radiating, swollen hyphae typical amount of pigmentation tends to be + constant across of Lecidella. The new variety instead has a totally all specimens, with most of the darker coloration of the unpigmented excipulum which is densely inspersed disc deriving from crystals, and the blue-green pigment with crystals, identical to what is displayed by L. being very dilute and located mainly in that part of the leucomarginata. In his discussion of L. granulosula (as — epihymenium adajcent to the excipulum.

L. chodati), Knoph (1990) mentions that crystals are In Tasmania, there is a _ further, undescribed seen “occasionally”, but he makes no mention of crystals saxicolous taxon with an unpigmented excipulum.

44

J. Adelaide Bot. Gard. 27 (2014)

This species has a well-developed, yellowish, areo- late thallus containing atranorin, 2,5,7-trichloro-3-O- methylnorlichexanthone (both major) plus zeorin and isoarthothelin (minor). It grows in sheltered underhangs in dry sclerophyll forest. We have excluded it from this study pending further collections.

Ecology and distribution. This new variety 1s known from South Australia (Kangaroo Island) and the coast of southern New South Wales, and this widely disjunct distribution suggests that it is probably more widespread, albeit inconspicuous and overlooked. It has been recorded from sheltered aspects on coastal rocks (chiefly sandstone and laterite) where it grew in rather nutrient- enriched conditions, together with Buellia aeruginosa A.Nordin, Owe-Larsson & Elix, Candelariella vitellina (Hoftm.) Miill.Arge., Halecania subsquamosa (Mill. Arg.) van den Boom & Mayrhofer, Lecanora dispersa (Pers.) Sommerf., Verrucaria fusconigrescens Nyl., Xanthoria ligulata (K6rb.) P.James and species of Caloplaca and Amandinea. The type specimen is from a more inland locality in dry lowland woodland, where it grew beneath a canopy of mallee eucalypts, associated with Xanthoparmelia subprolixa (Nyl. ex Kremp.) O. Blanco et al. and species of Caloplaca.

Additonal specimens examined

SOUTH AUSTRALIA. Kangaroo Island, northern end of Ante- chamber Bay, 35°46’S 138°04’E, 5 m alt., 22 Sep 2013, G. Kantvilas 264/13 (HO).

JERVIS BAY TERRITORY. Bristol Point, 35°08’S 159°44’E, 1 malt., 17 Nov 2012, G. Kantvilas 606/12 (HO).

Acknowledgements

We thank Jean Jarman for producing the photographs for this paper and preparing the line drawing for publication.

References

Elix, J.A. (2008). Additional lichen records from Australia 67. Australasian Lichenology 63: 2-9.

Elix, J.A. (2010). Additional lichen records from Australia 72. Australasian Lichenology 66: 60-69.

Elix, J.A. & Enrnst-Russell, K.D. (1993). A catalogue of standardized thin-layer chromatographic data and _ bio- synthetic relationships for lichen substances, 2nd edn. (Australian National University: Canberra).

Fletcher, A., Purvis, O.W. & James, PW. (2009). Lecidella Korb. (1855). In: Smith, C.W., Aptroot, A., Coppins, B.J., Fletcher, A., Gilbert, O.L., James, P-W. & Wolseley, P.A. (eds), The lichens of Great Britain and Ireland, pp. 519- 525. (British Lichen Society: London).

Hafellner, J. (1984). Studien in Richtung einer natiirlicheren Gliederung der Sammelfamilien Lecanoraceae und Leci- deaceae. Beiheft zur Nova Hedwigia 79: 241-371.

Kantvilas, G. (2011). The lichen genera Japewia and Jape- wiella in Australia. Muelleria 29: 99-103.

Kantvilas, G. & Elix J.A. (2013). The lichen genus Lecidella (Lecanoraceae), with special reference to the Tasmanian species. Muelleria 31: 31-47.

45

Two new taxa in Lecidella (Lecanoraceae)

Fig. 2. The Lecidella-type ascus (with amyloid portions stippled), ascospores and paraphyses (with pigmented areas stippled), as exemplified by Lecidella granulosula var. lecanorina. Scale bar: 10 um. — G. Kantvilas 375/13 (holotype).

McCarthy, P.M. (2014) Checklist of the lichens of Australia and its island territories. (Australian Biological Resources Study: Canberra). www.anbg.gov.au/abrs/lichenlist/intro duction.html accessed: 27 June 2014].

Meyer, B. & Printzen, C. (2000). Proposal for a standardized nomenclature and characterization of insoluble lichen pigments. Lichenologist 32: 571-583.

Knoph, J.-G. (1990). Untersuchungen an gesteinsbewohnended xanthonhaltigen Sippen der Flechtengattung Lecidella (Lecanorales, Lecanoraceae) unter besonderer Bertick- sichtigung von aussereuropdischen Proben exklusive Amerika. Bibliotheca Lichenologica 36: 1-183.

Printzen, C. (1999). Japewiella gen. nov., a new lichen genus and a new species from Mexico. Bryologist 102: 714-719.

Rambold, G. (1989). A monograph of the saxicolous lecideoid lichens of Australia (excl. Tasmania). Bibliotheca Licheno- logica 34: 1-345.

J OQURNAL of the ADELAIDE

BOTANIC GARDENS

AN OPEN ACCESS JOURNAL FOR AUSTRALIAN SYSTEMATIC BOTANY

flora.sa.gov.au/jabg

Published by the STATE HERBARIUM OF SOUTH AUSTRALIA on behalf of the BOARD OF THE BOTANIC GARDENS AND STATE HERBARIUM

State Herbarium of South Australia PO Box 2732 Kent Town SA 5071 Australia

Board of the Botanic Gardens and Reit State Herbarium

of South Australia Department of Environment, Water and Natural Resources

14-03 Lang Ol.cs6.indd 47

Journal of the Adelaide Botanic Gardens 27 (2014) 47—56

Goodenia valdentata (Goodeniaceae), a new rare species endemic to Davenport Range, South Australia

PJ. Lang

State Herbarium of South Australia, P.O. Box 2732, Kent Town, South Australia 5071 E-mail: peter.lang@sa.gov.au

Abstract

A new arid zone species endemic to Davenport Range in South Australia 1s formally described as Goodenia valdentata P.J.Lang. It belongs in subsection Goodenia and fits within a group of species allied to G. grandiflora Sims. Previously, G. valdentata has been confused with Goodenia saccata Carolin and G. chambersii F.Muell., and characters distinguishing the three species are tabulated. Goodenia valdentata appears most closely related to G. chambersii, with G. saccata a more distant relative strongly linked to G. albiflora Schldl. by the synapomorphies of crenulate- laciniate corolla wing margins and flattened, short, white hairs inside the corolla. Based on its restricted range G. valdentata warrants listing as rare in South Australia.

Keywords: biodiversity, new species, taxonomy, Australia, Goodenia subsect. Goodenia.

This new species of Goodenia Sm. was first col- lected by B.G. (Bruce) Andrews! in October 1968 on Nilpinna Station. It 1s represented by nine collections in the State Herbarium of South Australia (AD), all from the vicinity of Davenport Range, a small isolated range on the boundary between Nilpinna and Peake Stations, c. 750 km NNW of Adelaide. Six of the existing AD collections have hitherto been identified as Goodenia saccata Carolin and the other three as G. chambersii F.Muell. (State Herbarium of South Australia 2013). Although related to these species, comparison of her- barium material shows that the new taxon is consistently and clearly distinct from both.

Goodenia valdentata P.J.Lang, sp. nov.

Typus: F./.Badman per W.R.Barker 5267, 28 Aug. 1986, 0.5 km SE of Levi Springs Bore, 28°22’35’S 136°09°32”E, Lake Eyre Region, South Australia, (holo.: AD 99524087; iso.: CANB, K). Goodenia saccata auct. non Carolin: Jessop, List Vasc. PI. S. Austral. ed 4, 87 (1993), as to LE occurrence.

Perennial, erect-stemmed subshrub to 80 x 80 cm. Stems branched, with a fine, velvety indumentum of dense, short, simple hairs and sparse, short, glandular hairs; prominently ridged, becoming woody. Leaves usually with 1—6 widely spaced, patent, narrow petiolar lobes or pinnae 2—5 mm long, arising along the petiole 7-20 (—25) mm long; /amina narrowly ovate to ovate, 16-36 (-46) x 9-25 mm, deeply and coarsely dentate with 14—25 (—29) teeth (including 1—9 (—13) secondary teeth), primary teeth 1.5—5 mm long with an elon- gated terminal tooth, 2.5—7.5 (-9) mm long, forming

' B.G. Andrews was briefly working in the area for North Broken Hill Ltd and developed an interest in the local flora; he submitted a total of 79 collections to the State Herbarium, all made in 1968.

Published online: 24 Dec. 2014 «¢ flora.sa.gov.au/jabg

an acute to acuminate apex; midrib prominent on abaxial surface, strongly and sharply raised at base, eradually shallower towards apex and disappearing near base of the terminal tooth (but continuing almost to the apex in upper inflorescence leaf-bracts); mid vein only weakly evident in lower two thirds of adaxial surface; both surfaces finely velvety with patent, stiff, short, fine, simple hairs, c. 0.1 mm long, and obscure, sparse, short glandular hairs. /nflorescence: flowers solitary, or rarely 2-flowered, in axils of upper stems, forming weakly differentiated leafy racemes or rarely thyrses; bracts leat-like, becoming more elongate and deeply-toothed up the stem; peduncles 1—1.5 mm long; bracteoles very narrowly lanceolate, 6—10.5 x 0.8—1.2 mm, with mid-rib on abaxial side; pedicels angular, 5—l11 mm long, persistent, articulate 1.2-2 mm below base of ovary. Hypanthium with 5 prominent, velvety ribs continuous with calyx lobes and covered in dense, simple hairs, intervening areas with mid-dense, short, glandular hairs, 5—7.8 mm long, grading into a basal stalk, 1.2—2 mm long. Calyx lobes green, leafy, narrow- lanceolate, 5.7-8 <x 1.1-1.5 mm, with a midrib in the lower 1/2—2/3, often with 1—3 (-4) weak teeth. Corolla a bright, rich yellow, often with purple veining inside on the lower part, 25-30 mm long, articulate; outside pubescent with dense, patent, short, fine, simple hairs; inside glabrous apart from fine, soft hairs at base of the tube, rows of prominent enations present in basal parts; pouch on upper 1/3—1/2 of the hypanthium, 2.5—4.5 mm long, shallow; /obes distinct from wings, lanceolate to narrow-triangular, with apices acute to acuminate, with a narrow blunt tip extending a little beyond the wings; wings entire, with sparse, marginal setae 0.4—0.7 mm long, not projected beyond lobe apex but gradually, or abruptly, narrowed apically to form a

ISSN 0313-4083 (Print) * ISSN 2201-9855 (Online)

19/12/2014 12:48:42 PM

PJ. Lang J. Adelaide Bot. Gard. 27 (2014)

—*

Fig. 1-3. Goodenia valdentata. 1 habit and habitat: flowering shrub on sandstone outcrop; 2 flowers and foliage; 3 detail of leaves. — D.E.Murfet 7034 et al. Photos: D.J. Duval, S.A. Seed Conservation Centre.

48

14-03 Lang Ol.cs6.indd 48 19/12/2014 12:48:44 PM

14-03 Lang Ol.cs6.indd 49

J. Adelaide Bot. Gard. 27 (2014)

Fig. 4. G. valdentata. Flowers, showing tapered corolla wings, and style with fine spreading hairs visible on the upper part. — D.E.Murfet /034 et al. Photo: D.J. Duval, S.A. Seed Conservation Centre.

combined acute to obtuse apex. Adaxial corolla lobes diverging from fused corolla at 7-10 mm from base, 12-18 mm x 1.9—2.5 mm; wings + equal, 7-12 x 1.4—2 mm; lacking distinct auricles. Abaxial corolla lobes separating at a further 3.5-—6.5 mm past divergence of adaxial lobes, 10.5-16 =< 2.5-3.5 mm; wings 7-9 x 1.5—2.8 mm. Stamen filaments linear, 7—7.5 mm long; anthers narrowly oblong to linear, 2.2—3.7 x 0.5-0.7 mm, apiculate, with small, ovate extension 0.3—0.6 mm long. Style 16—21 mm long, pale cream with some short, glandular hairs and soft, fine, spreading hairs, 0.5—1 mm long, for almost its entire length, but becoming purple and with longer hairs to 1.5 mm near junction with indusium. /ndusium inclined forwards at almost right- angles to style, pale purplish-grey, obovoid-trapezoid, 2—2.6 mm long, dorsiventrally compressed, with apical margin lightly convex from above and 3.84.1 mm wide, base 2.1—2.5 mm wide, glabrous except for some long hairs in a darker, purplish zone at its base, orifice beset with dense, white bristles 0.4—0.6 mm long. Fruit ellipsoid, 1O—14 < 5—7 mm, tapered at base into a short stalk, 0.5—2.5 mm long, 2-valved; valves extending 2-3 mm beyond attachment of calyx lobes, apices rostrate and usually bifid, sometimes splitting towards base and giving the appearance of a 4-valved fruit; septum 2/3-— 3/4 the length of loculus. Seeds light brown, broadly elliptic, 3.6-4.1 <x 2.3—2.6 mm, papillulose, surrounded by a raised thickened margin with a squared-off edge. Fig. 1-4, 7, 10.

Diagnostic features. The new species has bracteolate pe- dicels and falls in subsection Goodenia (Carolin 1992), comprising over 53 species. Within this subsection it belongs with a loose group of species allied to G.

49

New Goodenia from Davenport Range (Goodeniaceae)

me - . / -

4 : . > ; " +—_ ; by ad ‘ _ j 7 . i 4 . ¥ = - “ aff ~~ a ; * ” :

¢

> .

Fig. 9. G. chambersii. Flowers, showing the abruptly terminated corolla wings and absence of hairs around the style apex. — D.J. Duval 1686 & M.J. Thorpe. Photo: D.J. Duval, S.A. Seed Conservation Centre.

grandiflora Sims, a widespread and variable species extending from coastal N.S.W. and Qld to the ranges of central Australia in N.T., and W.A. The group also includes G. brunnea Carolin, G. kingiana Carolin, and G. macmillanii F.Muell., as well as G. chambersii, G. saccata and possibly G. albiflora Schltdl., all species that have been previously treated as G. grandiflora or varieties thereof. Species of this group are shrubby perennials (excepting G. albiflora which is a renascent rhizomatous perennial), and they are usually long-lived, and found on rocky substrates. Goodenia valdentata 1s clearly distinct from all these species, but is most like G. grandiflora, G. chambersii and G. saccata in its general appearance. It differs from G. grandiflora by its narrower leaves that are never cordate at the base, the more deeply serrate leaf margin, and consistently ciliate corolla wings (with long sparse setae) that taper to the lobe apex. Goodenia valdentata most closely resembles the two South Australian endemic species, G. chambersii and G. saccata, with which it has been previously confused. Diagnostic characters separating these three taxa are summarised in Table 1.

Goodenia valdentata has distinctive corolla wings that are gradually to steeply tapered apically (Fig. 4) rather than abruptly terminated. Combined with the main body of the lobe this produces an obtuse to acute apex. In contrast, the combined apex in G. chambersii is truncate (Fig. 5) or often emarginate where the wings are angled forwards beyond the apex; in G. saccata it 1s truncate with the lobe apex protruding as a short acute extension.

The three species can also be separated on leaf cha- racters alone. The leaves of Goodenia valdentata are distinctive by their longer lateral teeth (to > 3 mm)

19/12/2014 12:48:46 PM

P.J. Lang

Fig. 6. Isotype of G. saccata (detail), showing short white flattened hairs on inside of corolla and crenulate-laciniate wing margins. — Hy.Eichler 12676 (MEL 23070). Photo: JSTOR Global Plants.

and the attenuate terminal tooth that is especially pronounced on the upper leaf-bracts (Fig. 3 & 10). They are consistently more elongate than the ovate-orbicular leaves of G. chambersi and approach those of G. saccata in their general shape, but differ by their deeper, more regular teeth. By contrast G. saccata has more secondary teeth, producing a somewhat irregular almost bidentate margin. The leaves of both G. valdentata and G. saccata are strictly speaking pinnatisect and bear 1-6 small petiolar pinnae. The pinnae in G. valdentata are narrower and less well developed than in G. saccata, while in G. chambersii they are always absent.

An unusual feature of G. valdentata is the develop- ment of 1—3 weak teeth on many of the calyx lobes. This was observed in four of the six collections of G. valdentata with flowers but in only seen once in G. saccata (on the holotype), and not at all in G. chambersii.

Affinities. Goodenia valdentata is geographically situ- ated between G. chambersii and G. saccata species (Fig. 13) and shares some morphological features of each. Goodenia chambersii 1s found on breakaways in eypseous landscapes around Coober Pedy, Arckaringa and Evelyn Downs, and also extends southwards in an arc on residuals of the Stuart Range, well to the west of the range of G. valdentata. Goodenia saccata 1s endemic to the northern Flinders Ranges and occurs as three discrete major populations on basement rocks, in the Arkaroola-Yankaninna, Blinman-Parachilna-Moolooloo and Leigh Creek-Beltana areas.

Goodenia valdentata matches G. chambersii in its larger, bright yellow flowers with a mostly glabrous indusium, prominent enations on the inside of the corolla, and entire wings that are distinctly ciliate with widely-spaced setae, particularly in their lower part. The latter character was among three highlighted by Carolin (1992) in justifying the recognition of G. chambersii as distinct from G. grandiflora. This status is supported by the molecular phylogenetic analysis of Goodeniaceae by Jabaily et al. (2012) where G. chambersii and G.

14-03 Lang O1.cs6.indd 50

SO

J. Adelaide Bot. Gard. 27 (2014)

Fig. 7. G. valdentata seed. — LHS: BS69-25204; RHS: B.Andrews, 10 Oct. 1968.

grandiflora emerged as well separated sister species. The distinctive seed features of G. chambersii (more prominent sculpturing and a smoother, more tapered rim) are not shared by G. valdentata or G. saccata, which both have seeds with a squared rim typical of other species in the G. grandiflora group (Fig. 7-9). Goodenia valdentata is close to G. saccata in general leaf shape, presence of petiolar pinnae, its seed characters and in having spreading hairs distributed over the entire length of the style. However, G. saccata differs by its lighter and slightly glaucous foliage, predominantly white flowers (see Notes, p. 54), larger corolla tube pouch, and sweet scent (noted as especially strong in the late afternoon). A similar set of features is found in G. calcarata F.Muell.’ and G. albiflora and is indicative of a pollination syndrome, possibly for night-flying insects, such as moths. More importantly, Goodenia saccata has two distinctive apomorphic fea- tures that are absent in G. valdentata, but shared with G. albiflora: stiff, short, white, flattened-clavate (to narrow-oblanceolate) hairs on the inside of the corolla, and irregularly crenulate-laciniate corolla wing margins (Fig. 6). Its closest relationship appears to be with G. albiflora rather than G. valdentata.

Overall, the new species appears more closely related to G. chambersii than to G. saccata, despite the obviously divergent leaf morphology of the former.

Distribution. Goodenia valdentata is endemic to Daven- port Range and associated drainage lines, in the Lake Eyre Region of South Australia (Fig. 13 & 14). It is not to be confused with the Davenport Range of Northern Territory, but is part of a system generally referred to as the Peake and Denison Ranges (Twidale 2007, Ambrose et al. 1981), comprising small isolated remnants of the Adelaide Geosyncline fold belt, from which the Mt

2 Goodenia calcarata (F.Muell) F.Muell. exhibits a similar polli- nation syndrome, but has the pouch extended into a prominent nectary spur. However, it appears less closely related and also differs by its ephemeral habit, entire corolla wing margins, and stiff purple hairs on the inside of the corolla that are slightly thickened but not flattened as in G. saccata and G. albiflora.

19/12/2014 12:48:48 PM

J. Adelaide Bot. Gard. 27 (2014) New Goodenia from Davenport Range (Goodeniaceae)

Fig. 8. G. chambersi seed. — LHS: B.Lay 638; RHS: E.H./sing, 22 Oct. Fig. 9. G. saccata seed. — LHS: K.H.Brewer 640; RHS: L.D.Williams 1955. 11722.

Lofty and Flinders Ranges were formed. The range — from Biological Survey of S.A. quadrats, data on sur- straddles the boundary of Nilpinna Station and The rounding vegetation indicate a very sparse tree or shrub Peake Station, and populations of G. valdentata occur — stratum dominated by Acacia species (A. ligulata, A. on both properties. salicina and A. tetragonophylla) or red gum (Eucalyptus

Habitat. The major rock type in the Davenport Range is camatdutensis) in the watercourses.

sandstone of the Burra Group from the Late Proterozoic = Conservation status. The known range of G. valdentata (S.A. Dept Mines and Energy 1982). Collectors’ field spans 21 <x 13 km, based on the seven more precisely notes describe G. valdentata as occurring on rocky located collections (Fig. 14). With such a restricted outcrops or hill slopes (4 collections) or watercourses — extent, it clearly warrants listing as a Rare species. There (5 collections). In all but one case, mention is made of | are no obvious immediate threats to the conservation the substrate being rocky or having a moderate to dense __ of this species, but it could also qualify as Vulnerable cobble strew. It is recorded as growing in: crevices of _ if its distribution within this area is found to be only deep red-brown rocks, skeletal brown soil, clayey sand, = sparse and patchy. Using the IUCN 2001 categories as and silty clay loam. The vegetation of Davenport Range adopted by the National Parks and Wildlife Council is mostly Open Woodland dominated by mulga and (2003), a plant may be assigned Vulnerable (VU) status chenopods (Gee et al. 1996). For the two collections — under criterion D2 for a restricted area of occupancy (as

Amended key to South Australian + shrubby, cauline-leaved species of subsection Goodenia (‘Group B’ of Carolin & Cooke 1986) 1. Leaves with petiolar pinnae (these often reduced) 2. Leaf lamina lanceolate, viscid, with simple hairs almost absent........ 0.0... 0. cc ee eee nes G. brunnea 2: Leaf lamina narrowly to broadly ovate (-triangular), finely velvety, with abundant short, simple hairs 3. Leaf margin irregularly dentate/bidentate, longest lateral teeth 1-3 mm long; corolla mostly off-white

WV CCTEN ATCT CA TNA TIO TAR OTITIS! aa gy cre se BENE ie ts ete Sav wl age AS a BENE, dtm Go Ev we Be aU IEER 8078 DEIN S'la Gh Few G. saccata 3: Leaf margin strongly and evenly dentate, longest lateral teeth 3-5 mm long; corolla yellow with + STINT TR ANONITS: , sch os ace eh eile Lene dant ened Wee amet Na hiees Ate Sv aitc Less ect eas aide h Beh ine eee eet Seng Ge den Beh Py kas a G. valdentata

1: Leaves simple, with petioles lacking pinnae, or sessile 4. Leaves and stems with plentiful glandular and/or non-glandular hairs 5. Leaves petiolate, with abundant simple hairs

6. Leaves broadly ovate to orbicular, green, with fine short erect hairs ......... 0... 0. ccc eee ee eee G. chambersii 6: Leaves linear to narrowly elliptic, silvery-grey, with appressed T-shaped hairs ...................5 G. mueckeana 5: Leaves sessile and stem-clasping, with abundant glandular hairs 7. Leaves strongly stem clasping, mostly > 3 cm long; indusium entire .......... 0... 0. cece eee eee G. amplexans 7: Leaves weakly stem clasping, mostly < 3 cm long; indusium notched ................... 000005 G. benthamiana 4: Leaves and stems glabrous or glabrescent, without conspicuous hairs $; [Leaves Sironely SAUCOUS) TOWEPS WHITES: osc LE fla ner eaten ew ot ne bol ate a nee tea be ate nok Habeas kone sess G. albiflora 8: Leaves green, flowers yellow 9, Leaves narrowly elliptic to oblanceolate; corolla with minute simple hairs outside .................. G. vernicosa 9: Leaves ovate to elliptic, rarely narrower; corolla lacking simple hairs 10. Leaves thick; sepals 2-4 mm long; bracteoles to 1.5mm long .......... 0... 0. ccc eee ees G. varia 10: Leaves thin; sepals 3—11 mm long; bracteoles to6mm long ........ 0... 0... eee ees G. ovata 5]

14-03 Lang Ol.cs6.indd 51 c, 19/12/2014 12:48:50 PM

P.J. Lang

14-03 Lang Ol.cs6.indd 52

peAiesel juBuAdOS

eeuIsnNy yNos JO wnueqiay aes

Fig. 10. Goodenia valdentata. Holotype, Fu.Badman per WR. Barker 5267.

52

J. Adelaide Bot. Gard. 27 (2014)

KU9952408 7

State Herbarium of South Australia

AVIA AD99524087

Goodenia Va dentate:

KE 22 Ney. 2013

DET. P.J. Lang

STATE HERBARIUM OF SOUTH AUSTRALIA (AD) Goodeniaceae / AD99524087

Goodenia saccata Carolin det. D.E. Symon, 15 Nov 1995 South Australia. Region 2; Lake Eyre, c. 0.5 km SE of Levi Springs Bore. (Autoderived cvord, 28 22°35"S, 136 09°32"E; precn 2)

Common. Rock outcrop on W end of rocky hill; deep red-brown rocks; S-facing slope. In rock crevices. Mainly seedlings. Corolla mid yellow. Style and indusium light grey.

F.J. Badman per W.R. Barker 5267 28 Aug 1986

f -

Duplicates tu: CA! %

Wh Databased in: ADHERB (crobins; geoc.OK) ANNI

19/12/2014 12:48:57 PM

New Goodenia from Davenport Range (Goodeniaceae)

J. Adelaide Bot. Gard. 27 (2014)

‘adAjojoy 9/92) 48/YoIF IH ‘eyea9es eluapooy ‘Z| ‘bi4 “Buls} aeuajey ‘9 Jo edAjeled e ‘yozg¢ Bus)’ ‘Isuequieyo ejuepooy *|y ‘bi

woo12F FO S808 an TU ag peoyyes = 2279 GepeT FO Ice P £5*IG) Balyy Wears * SdEy

) | GxOparlad TAewiaey “RPTeaseny*g

:

er We ya, |

| ; | rt ; - " : = t a | 77 eee relels . ; : > | ' . HewA tad TH $0 Geoe—ople = Wa Clyp -ee = Bua) tH Esan ’ : Tes ye Pa -- : . wt. a - 7, ie] ' th ; an te —T> SLA") —-—*. Velw> Fx, ‘Se Tee ce ey Twos ; A ; 4 = * f ” “7 “e 4 u i ry ' 7 r , 7 ' ELT elalt wh Ta Lams a on = rir ; | ere i> 7 "HOR oh ra = a ; mY cor epu leu ST UeEDOO | — jaryv7soywv , = < wih DRS soiviagy en veF jeg ap) woah Hl ~ 20°°Tae at WITWHLSNY HINOS 40 WHinvuusH 3 VNWHIsnY HLNOsS 40 WhiiveEwsM Silvis f Wee ~~ @ weal ; j cy a ee Try / yn » ' = ’ d — ° Le ee ao ‘ - = —_r- “ ceune we nal i > , wlhs "sy Wre YT) eure i worm ho F bm ATG) eapeey ® ye e709 ela ate ® “syAy - ' ay | her 4 > den —— ¥ 5k fk “ — _- Wis oun) « ‘ — . ' eal 4 e ‘3 Fy hu Py Yrs oy wa) AE sti 3 30) oye 3 si ru. ive 4 Nat 2 Wy ragby PHS TY Oy

opyr

ber ey ty Pha + SAIN

X

at | , ; ‘a LY?) SMa vreirruaese

| SdALOTOH

234ole 20 41/Se24/09 -NV-0g Jo Lh arian

i6av aa Tin

eyeamatiy HINGE FO MUETILEG-WH &/FES.

BLO01L2S60V

| | - av pasjesey pybuAdos seam CEOSTSLE ly Byeaysny YMVOS — — <a — -

JO WNUEQMUEH EIS

Pce

wars

19/12/2014 12:49:01 PM

53

14-03 Lang Ol.cs6.indd 53

P.J. Lang

” Goodenia valdentata

¢ Goodenia chambersii

Goodenia saccata

Fig. 13. Distribution of Goodenia valdentata, G. chambersii and G. saccata, based on AD collections.

opposed to extent) of “typically less than 20 km”. Currently there is insufficient information to estimate the area of occupancy. Collectors’ field notes merely describe it as [locally] moderately common or common (5 collections). To ascertain whether it qualifies as Vul- nerable, details on population sizes and their coverage within the range area are required, so at this stage only Rare status 1s recommended.

Etymology. The epithet valdentata 1s formed from a contraction of the Latin words valde, strongly, and dentatus, toothed, and refers to the leaf margin with prominent teeth, which are more pronounced than in related species.

Notes. One factor contributing to the confusion of G. valdentata with G. saccata is the description of white or yellow flower colour for the latter. Goodenia saccata appears to be a consistently white-flowered species, but its flowers were first described as “cream to yellow” (Carolin 1986), and later as “yellow or off-white” (Carolin 1992). It is unlikely that Carolin would have seen fresh material and these descriptions are probably based on the uniform yellowish tinge to the corolla seen on some herbarium specimens, which is probably an artefact of drying and aging.

Flowers of the holotype sheet (Fig. 12) appear particularly yellowish, but on other plants from the

14-03 Lang 01.cs6.indd 54

54

J. Adelaide Bot. Gard. 27 (2014)

’ A - ~y —_ ~ — - i *- 5 p— / -.~ r * ra vr = le « J -_ p f 4 , 4 ° s _ » £ ft - _ - = -" + - ”. - ’ ~ i 4 : ~f , Ld + Ms . f , ; i. w t : ' - : 7 fleece —. < ue OM - 4 . : + >... ai A r \4 ge. . ~~ "ae ‘ A ‘ : : f ~ - -_ | + @. of. a aT, eng , ii oe | 7 . : ad —\ —sT)| os Are oes J ' es Eo VI WO Jj ed ee r - fo | iets 5 ' ee ee nd - 4 : : : poge, Meee a Z . ¢ sea } ose v' 7 f Ff, — , se | : ; a ¢ in 4. 7 . ie Ve i+ ; 9 * % : 'F ° ' ”

: 4 oo +

=

- f ~ ob an a

‘ “ 7 -™ Lav — s ' ® P 7 s fr, =~) + maT « ct ,- 4 ~~ a 4 y - UP Le eee 2s on p= * is | gat ie or iYais je - > = % a | Re, Le Ce eel ce = — . s—= COIN ob a eee om § ~S aa Py peeitar ati 8 < J a7 * ST he Le ry 7 7 ae ee A -

>= 4 ,# ~~, { * tie aT! ” a + ‘ a 5 i = (ff - ris

ey - 4 - r] | A + 1 Vem | ae ~

4 . ™ eae’ rr , rs a | ye 4 - ;

re 7

- + is ad a a | OWN

- ]

34 7

che

>

s ie

‘_ r

r 7

' - Thad

ia -

ees) s, ( An ox

tae Ps. ‘ears, & if Viargaret . ie. SP LiviagiyaiCtl . . ee : ~ = eee eal es

end - ‘ _*.

- 4 ? i get im. wave 7

Fig. 14. Detail of Goodenia valdentata distribution (+) based on the seven collections with a resolution of 1 km or better.

same collection (e.g. the isotype, MEL 23070) they are clearly white. The possibility of yellow-flowered variants within the population cannot be eliminated completely, but it seems unlikely given the absence of field-based records of yellow-flowered forms, and the following evidence. More common species known in the field to be consistently white-flowered such as G. albiflora and G. calcarata do sometimes produce dried specimens with a yellowish tone on the flowers (e.g. R.Taylor 507 and D.J.Duval 1544, respectively). Also, one herbarium specimen of G. saccata with yellowish flowers (K.H. Brewer 546) has field notes describing the flowers as white. In the four other AD collections of G. saccata with field notes that refer to flower colour, it is also given as white. Further support is provided by recent field observations of the two eastern populations, one by the author and several by K. Brewer, who re- ported that the flowers were always predominantly white or “off-white”, although sometimes suffused with yellow towards the base.

Goodenia valdentata collections examined other than type, all from AD

SOUTH AUSTRALIA. Lake Eyre: B.Andrews, 10 Oct. 1968, Nilpinna Station; -!/.Badman 51438, 18 Apr. 1992, Peake Station, Edith Spring (Dups: BRI, NY); F'/.Badman 7051, 12 Aug. 1993, 13 km NW of Mt Margaret, Peake Station; FE-J.Badman 7070, 12 Aug. 1993, 12 km NW of Mt Margaret, Peake Station; R.Bates 59193, s.dat. [c. 15 Aug. 2001 based on adj. nos.], Davenport Range (SW gorge), Nilpinna (Dups: MA, SIU); JL.Chivell & A.Shalekoff BS69-25204, 1 Mar. 1996, 15.7 ENE Nilpinna; D.E.Murfet 7034, D.J.Duval & 7’S.Te, 9 Oct. 2010, Davenport Range, c. 10 km ESE of Nil- pinna Station; A.C.Robinson & S.A.Bemmer BS69-28610, | Mar. 1996, 23 km ENE Nilpinna Homestead.

19/12/2014 12:49:05 PM

J. Adelaide Bot. Gard. 27 (2014)

New Goodenia from Davenport Range (Goodeniaceae)

Table 1. Diagnostic characters distinguishing Goodenia chambersii, G. valdentata and G. saccata.

Indumentum of stems & leaves

simple hairs

glandular hairs

Petiolar pinnae number

shape

Lamina

shape

length/width ratio

Leaf margins

number of teeth secondary teeth

lateral tooth length (max.) terminal tooth length

Flowers per axil

Corolla colour

pouch

enations vestiture inside

wing margins wing marginal setae wing termination

lobe & wing apex Style

spreading hairs

indusium surface

Seed

mature colour margin

rim

sculpture

G. chambersi

mid-dense sparse to mid-dense, conspicuous

none

broadly ovate to orbicular

0.9-1.3

crenate-dentate

9-20

rare

Q.8—3 mm to 4.5 mm

I-3

yellow shallow, on upper 1/2 of hypanthium

yes pilose in lower 1/3—1/2

entire

yes

not tapered apically, often projecting beyond lobe

truncate or emarginate

only in lower 1/2—2/3, and near base of indusium mostly glabrous

mid to dark (reddish-) brown flush or slightly raised tapered or rounded papillose-verrucose

G. valdentata

very dense sparse, inconspicuous

1-6

narrowly oblong

narrowly ovate to ovate

1.4—2.0

prominently & acutely dentate elongate terminal tooth

14-25

few

3—5 mm

to 7.5 mm

1 (-2)

bright yellow

shallow, on upper 1/3-1/2 of hypanthium

yes

pilose only near base

entire

yes

gradually to steeply tapered apically, not exceeding lobe

acute or obtuse

entire length

mostly glabrous

pale brown

strongly raised,

flat with squared-off edges papillulose

G. saccata

mid-dense mid-dense, prominent

1-6 narrowly triangular to triangular

(narrowly-) ovate to broadly ovate-triangular 1.0—2.2

irregularly compound-dentate

18-45 many 1—3 mm to 4mm

I-3

mostly off-white

prominent, on upper 2/3—3/4 of hypanthium

absent, or a few weak ones

hispid, flattened straight white hairs in lower 2/3—3/4

crenulate-laciniate

no

usually abruptly terminated short of lobe apex

truncate with protruding tip

entire length

hirsute with soft hairs

pale brown

strongly raised,

flat with squared-off edges papillulose

14-03 Lang Ol.cs6.indd 55

Goodenia chambers, selected specimens (of 27 seen), all from AD (except otherwise indicated)

SouTH AUSTRALIA. Lake Eyre: D./. Duval 1888, 1 Sep. 2010, 4 km E Arckaringa Station; E.AH Jsing 3626A, 15 Oct. 1950, Evelyn Downs, 90 miles SW of Oodnadatta (Dups: MEL, K, BM, P, TI, US, CANB); £.AlIsing, 22 Oct. 1955, Evelyn Downs, Oodnadatta; PJ.Lang BSOP-438, 12 May 2000, Just inside western boundary of Arckaringa Station; B.Lay 638, 11 Oct. 1971, Residuals of Stuart Range, c. 80 km NW of Millers Creek Station, Balta Baltana; J. McDouall Stuart (expedition), 1859, N.W. interior of South Australia (lectotype: MEL 23048; isotype: K; images seen); 7’ Webb 9, 21 Sep. 2010, c. 10 km E of Coober Pedy; H.P Vonow 3535 et al., 14 Dec. 2010, Arckaringa Station, S of Wurley Hole; J.H. Willis, 23 Jul. 1979, Allandale Homestead, SE of Oodna- datta (Dup. of MEL 2118360).

55

Goodenia saccata, selected specimens (of 25 seen), all from AD

SouTH AUSTRALIA. Flinders Ranges: H.W. Andrew, Aug. 1920, Mt Bayley, near Beltana; K.H.Brewer 546 & T'S.Te, 26 Nov. 2010, Glass Gorge road, c. 5 km from Parachilna Gorge; K.H.Brewer 640 & K.Waugh, 9 Dec. 2011, between Yankaninna Homestead and the outstation at Owieandana; P.E.Conrick AD28, Aug. 1979, Arkaroola; Hj.Eichler 12676, 17 Sep. 1956, Gammon Ranges, near mouth of gorge of Arcoona Creek, south of Arcoona Bluff Range (holotype: AD 95710018; isotypes: MEL 23070, B, L, images seen; BRI, NSW, BM, K, NE, P, TI, UC, US, n.v.); Hj. Eichler 12997, 26 Sep. 1956, Aroona Dam, 6.5 km SW of Copley (Dup.: CAL); E.H Ising 511, 7 Oct. 1918, Witchen’s Well, Moo- looloo; D.N.Kraehenbuehl 23, 5 Oct. 1958, Near junction of Moolooloo and Parachilna roads (Dup.: RSA); D.E.Symon

19/12/2014 12:49:06 PM

P.J. Lang

6762, 5 Sep. 1969, W end of Parachilna Gorge (Dups.: NSW, K, B, L); L.D.Williams 11722, 4 Nov. 1980, 12.2 km [N of] Yankaninna HS.

Acknowledgments

I am grateful to Dan Duval of the S.A. Seed Con- servation Centre, Botanic Gardens of South Australia, for providing images of the plant taken in the field. | also thank Kieran Brewer for sharing details of his recent field observations of Goodenia saccata, Ainsley Calladine for producing the images of herbarium sheets, and Graham Bell for information on B.G. Andrews from the State Herbarium archive.

14-03 Lang Ol.cs6.indd 56

56

J. Adelaide Bot. Gard. 27 (2014)

References

Ambrose, G.J., Flint, R.B. & Webb, A.W. (1981). Precambrian and Palaeozoic geology of the Peake and Denison Ranges. South Australian Geological Survey Bulletin 50: 5—71.

Carolin, R.C. (1986). Goodenia saccata; Goodenia brunnea. In: Jessop, J.P. & Toelken, H.R. (eds), Flora of South Australia 3: 1403; 1393. (Government Printer: Adelaide).

Carolin, R.C. (1992). Goodenia. In: George, A.S. (ed.), Flora of Australia 35: 147-281. (ABRS, Australian Biological Resources Study: Canberra).

Carolin, R.C. & Cooke, D.A. (1986). Goodenia. In: Jessop, J.P. & Toelken, H.R. (eds), Flora of South Australia 3: 1388-1406. (Government Printer: Adelaide).

Gee, P., Gee, I. & Read, J. (1996). An annotated bird list from the Davenport Range, South Australia. South Australian Ornithology 32: 76-81.

Jabaily, R.S., Shepherd, K.A., Gustafsson, M.H.G., Sage, L.W., Krauss, S.L., Howarth, D.G. & Motley, T.J. (2012). Systematics of the Austral-Pacific family Goodentaceae: Establishing a taxonomic and evolutionary framework. Taxon 61: 419-436.

National Parks and Wildlife Council (2003). Review of the status of threatened species in South Australia: Proposed schedules under the South Australian National Parks and Wildlife Act 1972. Discussion Paper. (Department of Environment & Heritage, South Australia: Adelaide).

State Herbarium of South Australia (2013). Census of South Australian Plants, Algae and Fungi. (Department of Environment, Water and Natural Resources: Adelaide). flora.sa.gov.au/census.shtml [accessed: Aug. 2013].

South Australian Department of Mines and Energy (1982). Geological Map South Australia, 1:2 000 000 scale. (Ade- laide).

Twidale, C.R. (2007). Ancient Australian Landscapes. (Ro- senberg Publishing: Dural, N.S.W.).

19/12/2014 12:49:06 PM