Royal Botanic Gardens ROYAL BOTANIC gardens Melbourne Melbourne national herbarium of victoria Muelleria publishes research papers on Southern Hemisphere plant, algal and fungal systematics, particularly relating to Australia and the Australian states of Victoria and Tasmania, and to the collections of the National Herbarium of Victoria. Acceptable submissions include: taxonomic revisions; phylogenetic and biogeographical studies; short papers describing new taxa, documenting nationally significant new records, or resolving nomenclatural matters; historical analyses relevant to systematics; any other research contributing to our knowledge of plant, algal or fungal diversity. Muelleria is published annually or semiannually by the National Herbarium of Victoria, Royal Botanic Gardens, Melbourne. Manuscripts should be sent in triplicate to: The Editor, Muelleria Royal Botanic Gardens, Melbourne Bird wood Avenue South Yarra Vic. 3141 Australia Muelleria may be found on-line at http://www.rbgmelb.org.au/muell . A cumulative index to scientific names and authors, as well as Intructions to Contributors may also be found on this address. Format should be copied from the most recent edition of Muelleria , or 'Instructions to Contributors’ may be obtained from the Editor. Twenty-five reprints of each accepted paper are provided free of charge. Subscription details can be obtained from the address above. Editor James Grimes Assistant Editor Teresa Lebel Editorial Advisory Committee Marco Duretto Tom May Jim Ross Neville Walsh Student Editors Rodney Jones Bryan Mole © 2000 ISSN 0077-1813 MUELLERIA CONTENTS Volume 14, 2000 Page Contributed Papers The Taxonomy of Boronia anemonifolia and B. rigens ( Boronia sect. Cyanothamnus, Rutaceae) — Peter G. Neish and Marco F. Duretto 3 The lichens of Nothofagus cunninghammii-dom i n ated and Acacia melanoxylon- dominated forests in the Otways, Victoria. — Sharon Ford, Maria Gibson and Geoff Duke 17 A note on Lepidium strictum (S. Watson) Rattan (Brassicaceae) in Victoria — Neville Scarlett 3 1 The stomata of bluegums ( Eucalytpus spp.) — Denis J. Carr 33 Identified types in the Wilhelm Hillebrand collections in the herbarium of the Royal Botanic Gardens, Melbourne — James W. Grimes 41 Additions to the Hygrocybeae of Victoria. 1. — A. M. Young 51 A revision of Agrostis billardierei R.Br. (Poaceae) — A. J. Brown and N.G. Walsh 65 Two new species of Thelymitra (Orchidaceae) from southeastern Australia — Jeffrey A. Jeanes 9 1 Spirogyra cardinia (Zygnemataceae, Zygnematophyceae, Chlorophyta): A new species of freshwater alga from Victoria, Australia — Simon H. Lewis and Timothy J. Entwisle 99 Book Review Flora of Australia, Vol. 12 — James W. Grimes 103 Muelleria volume 13 was distributed on 9 June 2000 fefl I " 1 ’ ' ' Tf *23*5gf- Muelleria 14:3 ( 2000 ) The taxonomy of Boronia anemonifolia and B. rigens ( Boronia sect. Cyanothamnus, Rutaceae) Peter G. Neish 1 and Marco F. Duretto Royal Botanic Gardens Melbourne, Birdwood Ave, South Yarra 3141 Australia 'Author for correspondence; e.mail: pneish@rbgmelb.org.au Abstract The taxonomy of Boronia anemonifolia A. Cunn. and B. rigens Cheel are discussed. Two new sub- species, B. anemonifolia subsp. aurifodina P.G.Neish and B. anemonifolia subsp. wadbilligensis P.G.Neish are described, and B. anemonifolia var. variabilis (Hook.f.) Benth. is raised to subspe- cific rank. All subspecies of B. anemonifolia are illustrated. The original description of B. rigens Cheel was based on discordant elements, and so is also revised. Boronia anemonifolia , B. vari- abilis, B. polygalifolia Sm. var. robusta Benth., B.dentigera F. Muell., and B. dentigeroides Cheel are lectotypified. Introduction While revising the east coast species of Boronia Sm. sect. Cyanothamnus (Lindl.) F. Muell. for the Flora of Australia treatment of Rutaceae it became apparent that the cir- cumscriptions of a number of taxa needed revision. Here, we revise B. anemonifolia A. Cunn. and B. rigens Cheel. The remaining east coast species of Boronia sect. Cyanothamnus are discussed in a forthcoming paper by Duretto, except for B. coerulescens F. Muell. which was revised by Wilson (1998). Material and Methods Material: Herbarium specimens were made available from AD, BRI, CANB, HO, LUND, MEL, NE, NSW, PERTH and TCD; cibachromes and 35 mm transparencies were received from K, and 35 mm transparencies were received from BM. Herbarium abbre- viations follow Holmgren et al. (1990). These specimens were augmented with material collected in the field from the entire range of each taxon. Five plants per population were collected where possible. Leaf Anatomy: The central portion of leaves of all taxa were fixed in 70% ethanol. Fixed material was dehydrated through a graded ethanol series up to 100% ethanol, infil- trated with 100% LR-White (London Resin) through a resin/ethanol series, and poly- merised at 60°C. Sections 2 pm in thickness were cut on a Reichert Ultracut ultra-micro- tome and stained with 0.05% toluidine blue solution (pH 4.4). Taxonomic Descriptions: Descriptive terminology follows Briggs and Johnson (1979) and Weston (1990) for inflorescence structure, and Murley (1951) for seed surfaces. Conservation codes follow the format of Briggs and Leigh (1996). Taxonomy 1. Boronia anemonifolia A. Cunn. in B. Field, Geographical Memoirs of New South Wales 330 (1825). Type citation: "Verge of the Regent’s Glen, Blue Mountains.” Type: Regents Glen. Blue Mountains, N.S. Wales, A. Cunningham No.43, Oct. 1822 (lectotype, here designated, K, ex. hb. Cunningham., n.v., (cibachrome MEL 2047064, photgraph AD); isolectotypes MEL 256802, NSW); rocky declivities and precipitous descents, ? A. Cunnigham (probable isolectotype K n.v. (cibachrome MEL 2047065 ; photograph AD 99543144)). Erect, much branched shrub to 2.5 m tall, the branches terete to slightly quadrangular 4 P.G. Neish & M.F. Duretto with moderate cork development on older branches, the stems and leaves glandular punc- tate; the whole plant glabrous to pubescent, with hairs often restricted to area between the decurrent leaf bases on stems, larger stems becoming glabrous with age, all hairs simple, erect, straight or curved, to 0.5 mm. Leaves simple or 3-5-foliolate, or bipinnate with the lower leaflets ternate, the leaf in outline 4 — 1 8 ( — 3 5 ) mm long, 2-25(— 30) mm wide; peti- ole 2— 9(— 1 6) mm long, canaliculate above; rachis segments 3-7 mm long, canaliculate above, sometimes winged, widest at distal end; leaflets or simple leaves narrow-cuneate to cuneate or narrow-elliptic to elliptic or ovate or obovate, strongly conduplicate or plane, concolourous, dorsiventral, 2— 9(— 13) mm long, 1— 4(— 6) mm wide, tip often divid- ed into three acute to obtuse lobes (Fig. 1), midrib slightly raised abaxially, spongy mes- ophyll continuous under midvein. Inflorescence axillary, l-6(-9+)-flowered; peduncle 1-8 mm long; prophylls unifoliolate or tridentate, persistent with mature fruit, 0.5— 5(— 8) mm long, to 1.5(-3) mm wide; metaxyphylls to 0.5 mm long, sometimes absent; anthopodium 0.5— 4(-6)mm long. Sepals broad-ovate, imbricate in bud, persistent with mature fruit, 1.5— 2.5(— 3.0) mm long, 0.8-1. 5 mm wide, 1/5 to 2/3 of the length of the petals, tip acuminate; adaxial surface glabrous or margins ciliate or sparsely to densely puberulous; abaxial surface glabrous or sparsely puberulous. Petals white to pink, imbri- cate in bud, persistent or caducous with mature fruit, (3.5— )4 — 5.5(— 6.5 ) mm long, 2-3 mm wide, midvein not raised on abaxial surface; adaxial surface with few scattered simple appressed hairs; abaxial surface glabrous or glabrescent or sparsely puberulous; tip with a small but distinct incurved hook. Stamens 8, erect, all fertile; filaments gradually taper- ing to anther connective, pilose on margins below slightly glandular tip, antesepalous fil- aments 1.3-2 mm long, antepetalous filaments 1.0-1. 5 mm long; anthers attached sub- apically on filament, anther connective maroon; anther-apiculum prominent, glabrous or with a few simple hairs. Disc entire, glabrous, entirely within staminal whorl. Ovary glabrous, or with a few hairs between carpels; style pilose at base; stigma rounded, as wide or slightly wider than style. Cocci glabrous or with a few simple hairs along suture, 3-5 mm long, 1.8-2. 2 mm wide. Seeds elliptical in outline, 2.5 mm long, 1.2-1. 5 mm wide, black to black-brown, dull, slightly tuberculate. without wax crystals between tuberculae. Typification: A single collection was cited in the protologue of B. anemonifolia: “Verge of the Regent’s Glen, Blue Mountains.” Cunningham collections ( Cunnigham 43) matching this information are lodged at K, MEL and NSW. The specimen lodged at K is in excellent condition, was part of Allan Cunnigham’s Australian herbarium and is des- ignated the lectotype. Another specimen lodged at K, with only ‘rocky declivities and precipitous descents’ written on a label, was annotated by Melville in 1955 who stated it was part of the type. The specimen is probably part of the type collection and is here con- sidered to be a probable isolectotype. Taxonomic History. Boronia anemonifolia has a complex taxonomic history. Cunningham (1825) characterised B. anemonifolia as having trifoliolate leaves with each leaflet divided into two or three at the apex. Later, Hooker (1834) described a similar taxon from Tasmania as B. variabilis Hook, and Mueller ( 1 855) described B. dentigera F. Muell. from Victoria. Mueller (1860-62) placed these latter two species, along with B. anemonifolia, B. tetrathecoides DC. (= B. polygalifolia Sm.), B. hyssopifolia Sieber ex J.D. Hook. (- B. nana Hook. var. hyssopifolia Melville), B. nana, and B. bipinnata Lindl. in synonomy under B. polygalifolia, concluding that B. polygalifolia was “one of the most variable species of the vegetable kingdom”. He retained this synonymy in his 'Census of Australian Plants’ (Mueller 1882, 1889) and in his 'Key to. ..Victorian Plants’ ( Mueller 1 888). Bentham ( 1 863) did not follow Mueller, and instead divided B. anemoni- folia into, what he called, three 'tolerably distinct races’: viz. var. dentigera (F. Muell.) Benth. (which included Cyanothamnus tridactylitis Bartl., = Boronia anemonifolia subsp. anemonifolia) from New South Wales, Victoria. Tasmania and Western Australia; var. variabilis (Hook.) Benth. (see subsp. variabilis (Hook.) P.G. Neish) from Tasmania; and Boronia anemonifolia 5 Hgure 1. Habit and flowers of B. anemonifolia. A-B B. anemonifolia subsp. anemonifolia'. A flower x3 (Neish et al. 42, MEL 2032477)', B habit xl ( Coveny 11044, CANB 333125)-, C-D B. anemonifolia subsp. aurifodina: C flower x3 ( Neish et al. 20. MEL 2032453); D habit xl (hoiotype, Corrick 7892, MEL 603327 ); E-F B. anemonifolia subsp. wadbilligensis: E flower x3 (Neish et al. 44. MEL 2032480 ); F Habit xl (hoio- type, Telford 3661, CANB); G-H B. anemonifolia subsp. variabilis (Tasmania): G flower x3 ( Neish et al. 68, MEL 2032648 ); H habit xl (. Moscal 8395, HO 403360); I B. anemonifolia subsp. variabilis (New South Wales): habit xl (Coveny et al 16565 CANB 468216) 6 P.G. Neish & M.F. Duretto var. anethifolia (Lindl.) Benth. (= B. anethifolia Lindl.) from Queensland and New South Wales. The Western Australian distributional data was based on his and Bartling’s (1848) erroneous assumption that the type of C. tridactylitis was from Western Australia. Baker (1899) extended the range of var. dentigera to include Mittagong and Sugar Loaf Mountain near Braidwood. Ewart (1930) adopted Bentham’s (1863) varieties in his ‘Flora of Victoria’. Rodway (1903) lists two varieties of B. anemonifolia in Tasmania, viz. var. dentigera and var. variabilis. Curiously, Rodway (1903) includes B. variabilis Hook, in synonomy under var. variabilis but does not include B. dentigera F. Muell. in synonomy under var. dentigera. Both of these varietal combinations had been made by Bentham (1863), but it cannot be assumed that Rodway was aware of these combinations. Reference to the work of Bentham is absent from the entire Rutaceae section of his work. He is inconsistent throughout his treatment in including synonomies and authors, and in author abbrevia- tions. Rodway (1903) did make a new combination, B- pinnata Sm. var. citriodora (Hook.) Rodway, but he did not state it was new. As it cannot be assumed that Rodway (1903) was or was not intentionally describing new combinations, var. dentigera (F. Muell.) Rodway (see subsp. anemonifolia) and var. variabilis (Hook.) Rodway (see subsp. variabilis), are considered to be illegitimate names. Of these two varietal names only the latter is listed by Chapman (1991). Cheel (1928) reinstated B. dentigera and described B. dentigeroides Cheel which was “similar in general appearance to B. dentigera but the leaves are more compound, being twice ternate, and the leaflets more or less flattened and dentate at the apex’’. Boronia dentigeroides and B. dentigera were placed in synonomy under B. anemonifolia by Willis (1957). Willis (1973), following Bentham (1863), retained B. anemonifolia var. vari- abilis, noting that it graded into the dentigeroides form of B. anemonifolia through the Furneaux group in Bass Strait. Curtis (1975) adopts a broad veiw of the species and recognises no varieties for Tasmania, while Weston and Porteners (1991 ) indicate that B. anemonifolia var. anemonifolia is the only variety found in New South Wales. Four subspecies of B. anemonifolia are recognised in this treatment, including two newly described. The rank of subspecies is considered appropriate due to the presence of gradation between subspecies. For example subsp. anemonifolia grades into subsp. vari- abilis throughout east Gippsland in Victoria and in the south-east New South Wales, while subsp. aurifodina approaches subsp. anemonifolia in central Victoria. Morphology is consistent within populations and it seems that intermediates are not the result of hybrids, but simply a product of the variability of the species. Distribution and ecology. Boronia anemonifolia is found in south-eastern Australia from near Warwick (SE Qld), through eastern New South Wales and eastern and central Victoria, to the islands of Bass Straight and northern and eastern Tasmania (Fig. 2). It is mainly found growing on and around sandstone and granite outcrops in eucalypt wood- land or forest, where it can be the dominant component of the shrubby understorey, or heath or shrubland in exposed areas. It flowers for most of the year, but mainly from August to February. Fruiting specimens have been collected between October and January. Etymology: The specific epithet alludes to the resemblance of the leaves to those of species of Anemone L. (Ranunculaceae). Key to subspecies 1 . Leaves simple or if trifoliate then petiole very much longer than leaflets; petals 5-6 mm long (central Vic.) lb. subsp. aurifodina 1 : Leaves trifoliolate, pinnate or bipinnate, petiole c. same length as leaflets; petals (3.5-)4— 5.5(— 6.5) mm long (NSW. Vic., Tas.) 2 2. Petals not persistent with mature fruit, inflorescence (l-)3-6(-9+)-flowered; Boronia anemonifolia 7 prophyUs simple, to 1.5 mm long; sepals 1-2 mm long, 1/5 to l/3(-l/2) of the length of the petals; leaflets glabrous (rarely glabrescent) (Qld, NSW, Vic., Tas.) „ “ : Id. subsp. variabilis 2: Petals Persistent with mature fruit, inflorescence l-3(-6)-flowered; prophylls simple or tridentate, usually 1-8 mm long; sepals 1.5-3 mm long, (l/4-)l/2-3/4 length of petals; leaflets glabrescent to pubescent (NSW, Vic.) & 3 3. Leaflets strongly conduplicate, < 2 mm wide, lateral leaflets usually tridentate (NSW, ,, la. subsp. anemonifolia 3. Leaflets flat, 2-5 mm wide, fleshy, lateral leaflets usually entire (Wadbilliga Plateau, subsp. wadbilligensis la. Boronia anemonifolia A. Cunn. subsp. anemonifolia. Cyanothamnus tridactylites Bartl. in Lehmann, PI. Preiss 2, 227. (1848). Type citation: “In regionibus interioribus Austrahae meridionah-occidentalis m. Octobri a. 1 840. hb. Preiss No. 2628”. Type: hb. preiss. 2628 (holotype LD 95036-0984 transparencies MEL 2044581 NSW- isotvne MEL 256803). ’ ' yp Boronia dentigera F. Muell. , Trans. & Proc. Victorian Inst. Advancem. Sci. 32 ( 1855) Type citation: “On sandhills near the La Trobe River, and in McCrae’s Island. Also, near the Pendland [Pentlandl Hills according to Mr. Dallachi.” Type: Latrobe River, // v Mueller [26.iv.1853] (lectotype, here designated, MEL 257409)-, Latrobe River F. i> Mueller, May 1853 (syntype K n.v. (photograph AD 99548108); McCrae’s Island, F. Mueller, v.1853 (syntype MEL 257413); Gippsland, Nov. Holl. aust., Dr F. Mueller 8 P.G. Neish & M.F. Duretto [1853?] (probable syntypes MEL 257411 , MEL 257412 (both from Sonder’s herbari- um)); Between Bacchus Marsh and Pentland Hills [Myrniong], F.v. Mueller [ 1853] (pos- sible syntype MEL 256801)', Wimmera River. N.W. Viet.. Dallachy [1853?] (possible syntype MEL 257410 (from Sonder's herbarium)). B. anemonifolia var. dentigera (F. Muell.) Benth., FI. aust. 1: 321 (1863). B. anemonifolia var. dentigera (F. Muell.) Rodway, Tasmanian FI. 22 (1903) nom illeg. non B. anemonifolia var. dentigera (F. Muell.) Benth. ( 1863). Boronia anemonifolia ssp. A (Typical Form) sensu Ross (1996, p. 129). Illustrations : G.R. Cochrane, B.A. Fuhrer, E.R. Rotterham & J.H. Willis, Flowers and Plants of Victoria t. 309 (1968); A. Fairley and P. Moore, Native Plants of the Sydney District, 235, t. 817 ( 1989); P.H. Weston and M.F. Porteners, FI. New South Wales 2: 230 (1991). Shrub to 1 .5 m tall. Branches glabrous or glabrescent or pubescent between decurrent leaf bases, leaves glabrescent to pubescent. Leaves trifolioldte or rarely bipinnate with the lower leaflets ternate, the leaf in outline 4— 1 5(— 20) mm long, 2— 1 4(— 1 9) mm wide; peti- ole 2—7 (— 1 0) mm long; leaflets strongly conduplicate or rarely flattened, narrow-cuneate, often canaliculate, tips tridentate; terminal leaflet 2-7(-9) mm long, 1 — 2(— 2.5) mm wide, about as long as lateral leaflets. Inflorescence 1-3-flowered; peduncle 1-8 mm; prophylls often tridentate, ( 1 — )2— 5(— 8) mm long, glabrescent to pubescent; anthopodium 1-5 mm long. Sepals glabrous or margins ciliate, or sparsely to moderately densely pubescent, 1 .5— 2(— 3) mm long, 0.8-1. 4 mm wide, (l/4-)l/2 the length of the petals . Petals glabrous or with few hairs along midrib or at tip, (3.5-)4-5.5(-6.5) mm long, 2— 2.5( — 3) mm wide, persistent with mature fruit. Selected specimens examined (of c. 250 collections): New South Wales; Rylstone Dam on Cudgegong River, near wall, J.H. Willis s.n., 6.x. 1969 (MEL); Bundanoon Creek, below Echo Pt., Constable s.n., 263.1956 (NSW 36706): Blue Mountains, at base of Fort Rock, near Blackheath, 33°38'08”S 150°16'23”E. P.G. Neish 33 M.F. Duretto & I Thompson, 28.X.1995 (MEL); 200m W of Pass, c. 3km E of Endrick R. on Braidwood-Nerriga Rd, 35°05'S 150°08’E, P.G. Neish 40 M.F. Duretto & 1 Thompson. 30.x. 1995 (MEL, NSW, CANB); Tianjara Falls, along track heading N from falls along cliff line, 35°08’S 150°20'E, R.J. Rudd 241. 17.xi. 1992 (CANB); Dr Georges Mountain summit, 36°40'S 149°54'E, N.G. Walsh 1675, 2.xii. 1 986 (BRI. HO, MEL); mountain peak 2.5km direct NE of Mount Poole. Yambulla State Forest, 37°ir30’’S I49°32'15"E, J.D. Briggs 1996 & D. Albrecht. 21.vii.1986 (CANB); Victoria: Jawbones, R. Webb 4 & E. Richards, 26. ix. 1965 (CANB); Brisbane Ranges NP, 4 km along Reids Road, going SW along the road from Rowsley, 37°44’S 144°19’E. V Stajsic 496, 27.ix.1991 (AD. BRI. CANB. MEL, NSW); near sum- mit of northern end of Cathedral Range. T.B. Muir 2554, 4,viii. 1 962 (MEL); Ballantyne Hills, Suggan Buggan, A.E. Orchard 2491 , 25.viii.1970 (AD); Nunniong Plateau, along saddle running north c. 1 km east of end of Brumby Point track, 37°03'20"S 148°05’()0"E, M.G. Corrick 10164, 23.1987 (MEL); Mt Kaye, East Gippsland. N.A. Wakefield 3532. 20.X.1946 (MEL); 5 miles [c. 8 km] ESE of Rosedale Post Office, 1 mile N of Limestone Quarry Road, A.C. Beauglehole 38190, 16.ii.1972 (MEL); Holey Plains Park. Pipeline Trk near intersection with Rosedale-Stradbroke Rd. 38 1 3 ’ S 146 52' F. P.G. Neish 81 & A. Vadala, ll.iv.1996 (MEL); Holey Plains State Park, Chessum Road, 0.2 km west of its junction with Kelly track, T18, 38°13‘20"S 146°55’00”E, M.G. Corrick 9991 & B.J. Conn, 14.x. 1986 (CANB. MEL); slopes above south bank of Yambulla Creek, c. 200 m within Victoria, c. 2 km SW of Yambulla Peak. 37°16’50”S 149°23'30"E, N.G. Walsh 2136. D. Albrecht & J. Westoway, 93x4988 (CANB. HO, MEL). Notes: Boronia anemonifolia subsp. anemonifolia is variable in leaf size and in the number and shape of the leaflets. The various forms grade into each other. A typical spec- imen is illustrated in Figure 2(A, B). Boronia dentigera was based on one of these forms. Plants matching the type of B. dentigera are found at Holey Plains State Park near Sale in eastern Victoria. These plants are notable in having very pink petals, even when dry, pubescent foliage and a prostrate habit and may warrant further investigation. Bentham (1863) and Rodway (1903) both state that B. anemonifolia var. dentigera (- Boronia anemonifolia 9 subsp. anemonifolia) is found in Tasmania, however all Tasmanian specimens examined are referable to the subsp. variabilis (see notes under subsp. variabilis). Distribution and ecology. This subspecies is found on the tablelands and in coastal areas south from Rylstone (NSW), through eastern Victoria and west to the Brisbane Ranges. It is found on sandy or rocky soils in dry woodlands or heath often on exposed sandstone or granitic outcrops. At MEL there is a collection of B. anemonifolia subsp. anemonifolia that was sup- posedly collected from the Grampians (W Vic.): viz. Grampians, Miss Cowle s.n., x.1904 (MEL). Other collections made at the “Jawbones” (eg. R. Webb 4 & E. Richards [CANB]) have also been assumed to have come from the Grampians. The “Balconies” or “Asses ears” in the Grampians sometimes go by the name of the “Jawbones”, but here the “Jawbones” are more likely to be from the Cathedral Ranges (south-central Victoria), where a number of collections of subsp. anemonifolia have been made. A search of the “Jawbones” in the Grampians did not locate any trace of this species. The only specimen seen of subsp. anemonifolia from the Northern Tablelands of New South Wales is from Mt Spirabo ( E.F. Constable s.n. NE 29213). However, this specimen is a duplicate of a specimen lodged at NSW (NSW 56122) which is referable to subsp. variabilis. These two specimens are clearly not from the same plant and the absence of any other speci- mens of subsp. anemonifolia in the Northern Tablelands suggests that this record is spu- rious. A search of the area around Mt Spirabo located only subsp. variabilis. Conservation status : Boronia anemonifolia subsp. anemonifolia is a fairly common subspecies that is adequately represented in reserves across its entire range. lb. Boronia anemonifolia subsp. aurifodina RG. Neish subsp. nov. A vaiietate typicafoliis unifoliolatis vel petioli quam foliolis multo longioribus, petal- is 5-6 mm longis differt. Type: Victoria: Rushworth Forest, west of Bailieston, 36°45’S, 144°56’E, M.G. Corrick 7892 , 25.x. 1 981 (holotype MEL 603327 ; isotypes HO 52023, AD 98230303 ). Boronia anemonifolia ssp. C (Rushworth) sensu Ross (1996, p. 129). Shrub to 1.2 m tall. Branches glabrous or pubescent between decurrent leaf bases. Leaves simple with trifid tips, though first few leaves on branches can be trifoliolate with leaflets very much smaller than petiole, entire leaf 5-15 mm long, simple leaves 0.5-2.5 mm wide, trifoliolate leaves to 6 mm wide; petiole 3-12 mm long; leaflets strongly condupli- cate or flattened, very narrow-cuneate to 3 mm long, 0.5-2 mm wide if leaf compound. Inflorescence ( l-)3(-6)-flowered; peduncle 1-4 mm; prophylls simple, 1.5-5(-7) mm long, sometimes tridentate, glabrous or pubescent; anthopodia 1-3 mm long. Sepals glabrous or margins ciliate or sparsely to densely pubescent, 1.5-2. 5 mm long, 1. 0-1.5 mm wide, c. 1/3 ot the length of the petals. Petals 5—6 mm long, 3—4 mm wide, glabrous or glabrescent, persistent with mature fruit. Other specimens examined: Victoria: Inglewood Flora Reserve, N of Rifle Range alon® Kingower Rd. c. 1.5 km SW of Town Centre, H36, N.G. Walsh 2451, 25.viii.1987 (MEL)fnistoric area, Castlemaine-Chesterton area, A.C. Beauglehole 69011, E.R. Perkin & F.H. Blake, 2.ix 1981 (MEL); Puzzle Flat, Bealiba, J.H. Willis s.n., 4,ix.l936 (MEL); Murderer’s Hill, Dunolly, M.E. Phillips 424-425, 20.ix. 1 96 1 (CANB); Mandurang, Koolamurt Park, Spring Gully Resevoir, M38 36°49’S 144°18'E, J. Samson s.n., 2.x. 1997 (MEL 2042600)- McEldrew’s Paddock, 1 mile ’[c. 1 6 km] NE of Eaglehawk near Bendigo, R. Melville 1248, P. Morris, M. Cohn, H. Henkel & W. Perry, 30. ix. 1952 (K n.v.. MEL); c. 18.5km SSW of Bendigo P.O., A.C. Beauglehole 50092 2 viii 1975 (MEL); Lightning Hill Road, Bendigo. F. Robbins ACB 7259, 1 1 ,ix. 1 947 (MEL); Diamond Hill, 17 km SSW of Bendigo, A.C. Beauglehole 50045, 29.vii.1995 (MEL); Moormbool, 36°48’S 144°54'E, N.A. Wakefield s.n., ix. 1939 (QRS 14601); Kamarooka S.F., c. 400 m N of Campbell’s Rd on road to Kamarooka, 36°34’S 144°23’E, PG. Neish 17 & K.R. Henshall 6 viii 1995 (MEL NSW); Whipstick, F. Robbins ACB7266, 25.ix.1949 (MEL); Central Whipstick. along road 5. W. Perry s.n., 10.ix.1961 (MEL 530742 ); Rushworth State Forest, between main Graytown-Whroo 10 P.G. Neish & M.F. Duretto Road and Darrochs (Derricks) Dam. J.H. Willis s.n ., 4.ix.l974 (MEL): Graytown State Forest. S of intersection of Old Kilmore Road with Graytown-Rushworth Road. 36 L 44 S 144°57 E, R.J. Fletcher 153 , 23.viii.1993 (MEL); Rtishworth State Forest, on Whroo-Graytown Rd. corner of Darrochs Rd and extending to Johnson's Reef Road, 36°45S 144°47 E. R.J. Fletcher 4, 4.x. 1991 (MEL, PERTH); Rushworth forest reference area, M42, A.C. Beauglehole 69729, 4.xi. 198 1 (CANB. MEL); 4.5 miles [c. 7.2 km] SE of Maldon, M46, F. Filson 5111, 17.viii. 1963 (MEL). Notes: Boronia anemonifolia subsp. aurifodina is segregated on the basis of its sim- ple, strongly conduplicate leaves with tips divided into three very small points (Fig. 2D), and its larger petals (5-6 mm long). (Fig. 2C). Specimens from the Rushworth Forest have consistently simple foliage while those from the other populations often have trifo- liolate leaves on the first few nodes of a branch. Plants from the Castlemaine area approach subsp. anemonifolia in leaf shape, but are distinguished by having a petiole much longer than the leaflets. Distribution and ecology: The subspecies is confined to central Victoria in the gold- fields area and is common in the Rushworth State Forest and widespread through the "Whipstick Forest” around Bendigo (Fig. 2). It is found in low open eucalypt woodland with a diverse understorey. The populations of this subspecies west ot Bendigo towards Wedderburn represent the western most occurrence of B. anemonifolia. Conservation status: Boronia anemonifolia subsp. aurifodina is restricted in range and is found mainly in state forests. Due to uncertainty regarding the long-term protec- tion provided by State Forests the subspecies is regarded as not adequately preserved. Monitoring of all known populations is necessary to determine the viability of this sub- species and a management plan might be appropriate. A conservation code ot 2R is con- sidered appropriate at this stage. Etymology: The subspecific epithet is derived from the Latin for gold-mine aurifodi- na ( aurum (gold) and fodire (to dig)), and refers to the subspecies being confined to the goldfield region of central Victoria. lc. Boronia anemonifolia subsp. wadbilligensis RG. Neish subsp. nov. A varietate typica foliolus applanatis latioribus (2-5 mm latis) differt. Type: SE of Wadbilliga Trig. 46km ESE of Cooma, South Coast, New South Wales, 36°23’S 149°35'E, /. R. Telford 3661 19.xi. 1973 (holotype CANB, transparency MEL 2059439 : isotypes PERTH n.v.. K n.v., L n.v., A n.v .) Shrub to 1 (—2) m tall. Branches pubescent, hair density greater between decurrent leaf bases, becoming glabrous with age. Leaves glabrescent to pubescent, 3-5-foliolate, the leaf in outline 8-18 mm long, 4-10(-18) mm wide; petiole 3-8 mm long; leaflets flat- tened. entire or tridentate at tip, obovate, oblanceolate or cuneate; terminal leaflet 4-9 mm long, 2 — 4( — 6) mm wide, shorter or the same length as the lateral leaflets. Inflorescence ( l-)3-flowered; peduncle 3-5 mm long; prophylls simple or tridentate, glabrescent to pubescent, 3-6 mm long, 1—1 .5( — 3) mm wide; anthopodium 2-3.5 mm long. Sepals 2-3 mm long. 1.0-1. 5 mm wide, 1/2 to 3/4 the length of the petals; abaxial surface moderately densely pubescent. Petals (4 — )4.5 — 5( — 6) mm long, 2-3 mm wide, per- sistence unknown; abaxial surface sparsely pubescent with hairs concentrated along the midrib. Fruit and seed not seen. Other Specimens Examined: New South Wales, Southern Tablelands: A prominent bluff 2.4 km direct north of Wadbilliga trig, 36°19'S 149°36'E. J.D. Briggs IRIS and P.H. Weston, 27.iii.1985 (CANB. HO. MEL, NSW); Spur SE of Wadbilliga trig., 36°20’S 149°36’E, D.F. Bla.xell 491. 25.iv.1971 (CANB. MEL, NSW); c. 0.5 km north of Wadbilliga trig, P. Gilmour W057, 22.iv.1983 (CANB): Wadbilliga NR Brogo Wilderness area, c. 15 km direct ENE of Kybean beside Razor Back Fire Trail, 36°20’S 149°36’E, 5. Donaldson 662. D. Mallinson. AM. Lyne & I.R. Telord, 15.xii. 1995 (CANB n.v.. MEL. NSW n.v, PERTH n.v); Wadbilliga trig, fire trail, 14 km E of Tuross River crossing. 36°21’S 149°37’E, M.D. Crisp 1238 and I.R. Telford, 30.ix.1975 Boronia anemonifolia 11 (CANB); On rocky outcrop on Spur 1.5 km SE of Wadbilliga trig., 36°21’S 149°37'E, P.G. Neish 44, M. Duretto & 1. Thompson , l.xi.1995 (MEL); Head of Tuross River, R.H. Cambage s.n., xi.1908 (NSW); Northern end of Wadbilliga Mtn Plateau, I. Olsen 2373 , 13.x. 1974 (NSW). Notes: Boronia anemonifolia subsp. wadbilligensis differs from subsp. anemonifolia in having pinnate leaves with wider and thicker leaflets (2—5 mm wide) and usually entire lateral leaflets, and from subsp. variabilis by the pubescent leaves, petals, and the longer prophylls and sepals (Figs. 2E, F). Distribution and ecology: This subspecies is restricted to the Wadbilliga Plateau in south-eastern New South Wales (Fig. 2). It is found in eucalypt woodland or low Allocasuarina nana (Sieber ex Sprengel) L. Johnson heath on rocky outcrops and ridge tops between 1200 and 1300 m. Flowering material has been collected in October and December. Conservation status: All known collections of B. anemonifolia subsp. wadbilligensis have been made within five kilometres of each other in Wadbilliga National Park. A con- servation code of 2RC+ is appropriate. Further field work is required to ascertain the full range of this subspecies. Etymology: The subspecific epithet is derived from the name of the major landmark, the Wadbilliga trig, point, within the distributional range of this subspecies. Id. Boronia anemonifolia subsp. variabilis (Hook.) P.G. Neish, comb. nov. Boronia variabilis Hook., J. Bot. (Hooker) 1: 255 (1834). Type citation: “Mr. Lawrence, (1831), Mr. Gunn, (n. 8.) - 6 Mr. Gunn, (n. 214) - y Mr. Gunn, (n. 303.), who observes that it is called Lemon-plant. Type: Van D. Land [Tasmania], Mr Gunn n. 214 (lectotype, here designated and by implication by Hooker, J. Bot. (Hooker) 2, 418 ( 1 840), K [ex lib. hook., 4 sprigs on lower half of sheet) n.v., cibachrome MEL 2041296. B. anemonifolia var variabilis (Hook.) Benth., FI. aust. 1; 321 (1863). B. anemonifolia var. variabilis (Hook.) Rodway, Tasmanian FI. 22 (1903) nom illeg. non B. anemonifolia var. variabilis (Hook.) Benth. (1863). Boronia dentigeroides Cheel, J. Roy. Soc. N.S. Wales 62: 301 (1929). Type citation: “Braidwood, W. Baeuerlen [sic]; Clyde Mountain, near Nelligen. J.L. Boorman; Belmore Falls, W. Forsythe; Menangle, Mr. Harper; Timburra (Stuart) ex Herb. Melbourne, labelled B. polygalifolia var. anemonifolia-, Flinder’s Island (Gulliver), labelled B. anemonifolia .” Type: Timburra [Timbarra, E of Tenterfield, 29°0I’S 152°13’E], C. Stuart s.n. (lectotype, here designated, NSW 377539 ; isolectotypes MEL 270372, MEL 270373)- Marengenburg, Braidwood, W Bauerlen s.n., ix. 1 890 (syntype NSW 385530): Braidwood District, W. Bauerlen s.n., ix. 1 884 (possible isosyntype MEL 251075); Braidwood District, W. Bciuerlen s.n., xii.1884 (possible isosyntype MEL 270174); Braidwood, N.S.W., W. Bciuerlen s.n. (isosyntype K n.v., photograph AD 99548104)- Clyde Mountain, near Nelligen [35°33’S I49°57’E], J.L. Boorman s.n., iii.1909 (syntype NSW 385289); Clyde Mountain or Sugarloaf Mountain, J.L. Boorman s.n., ix 1915 (pos- sible syntype NSW 385321); Belmore Falls, W. Forsythe (syntype ? NSW n.v.); Menangle, Mr Harper s.n., viii.1894 (syntype NSW 385576); Flinder’s Island. Gulliver (syntype ? NSW n.v.). Boronia anemonifolia ssp. B (Wilsons Promontory) sensu Ross (1996, p. 129). Illustrations: N.C.W. Beadle and L.D. Beadle, Students Flora of North Eastern New South Wales Part IV, 554, Fig. 243B (1980); W.M. Curtis, The Student's Flora of Tasmania, 105 (1975); M. Cameron, Guide to Flowers and Plants of Tasmania, 110 ( 1 98 1 ). Shrub to 1 .5(— 2.5) m tall. Branches glabrous or pubescent between decurrent leaf bases oi rarely around entile stem (Sensation Gorge, Tas.), leaves prominently glandular, glabrous or glabrescent. Leaves 3-5-foliolate or bipinnate, the leaf in outline 1 3— 25(— 35) mm long 9-25(-30) mm wide; petiole 5— 9(- 1 6) mm long; leaflets simple, flattened or P.G. Neish & M.F. Duretto 12 conduplicate, entire or tridentate, broad-cuneate to oblanceolate, 3— 8(— 13) mm long, 1^4 mm wide, terminal leaflets usually shorter than lateral leaflets. Inflorescences ( 1 -)3-6(-9+)-flowered; peduncle 2-10 mm long; prophylls glabrous, simple, linear, 0.5—1 ,5(— 2.5) mm long; anthopodium 1.5-4(-6) mm long. Sepals glabrous. 1-2 mm long, 0.8-1. 2 mm wide. 1/5— 1/3(— 1/2) of the length of the petals. Petals 3-5 mm long. 2.5-3 mm wide, caducous, sometimes lately; abaxial surface glabrous or glabrescent. Selected specimens examined (of c. 200 collections): Queensland: About lkm ENE of Gambubal Forest Station, E of Warwick. 28°14’S 152°23’E, A.R. Bean 10980 6.X.1996 (BRI n.v„ NSW n.v., MEL); Paddy’s Knob, Condamine Gorge, Lot 13, Parish of Emu vale, 28" 17 S 152 22 E. K. Sparshott KS45 & P. Sparshott , 9.iii. 1993 (BRI); Mt Mitchell [possibly 28°04’S 152°23’E], anon (MEL); New South Wales: On slope above Boonoo Boonoo River, c. 100 m upstream from Boonoo Boonoo falls, 28°48’S 152°10’E, R.G. Coveny 16565 & A.J. Whalen , 14.X.1993 (BRI, CANB. NSW, NBG n.v., P n.v., NY n.v., CHR n.v.); Hill c. 1.8km NNE of Mt Spirabo, 29°20 00 S 15' ,o 05’05”E P.G. Neish 31 M.F. Duretto & I Thompson. 26.x. 1995 (MEL); Pheasant Mtn, 2 km E of Backwater on Paddy’s Gully Rd, 30°03’S 15 1°55'E, J.B. Williams s.n.. 30.x. 1965 (NE 49121 , NSW)- Mt Currockbillly, Budawang Range, Southern Tablelands, 35°25’S 150°02’E, A. Sikkas and 1 r Telford BR405 , 7,xii. 1973 (CANB. NSW); 4.4 km west of bridge at Numeralla, then 100 m south of road, 36°11’S 149°18’E, J.D. Briggs 1885 , 15.vi. 1985 (BRI, CANB, MEL); 6.2 km SW of Bronte trig, and 6.6 km E of Bemboka Peak, 36°36‘10”S 149°42’30”E. D.E. Albrecht 2953. 4.xi.l986 (CANB. MEL); The Pinnacles, Ben Boyd NP, 36°50’30"S 149°55’20’’E, D.E. Albrecht 2282. 1 0.xii. 1 985 (MEL. NSW): Victoria: Corner Inlet, Sunday Island, C. I.Skewes s.n.. 2- 6.ix.l952 (AD, CANB, MEL); Snake Island, Corner Inlet on NW edge of “The Gulf’, 3.25 miles [c. 5.2 km] WSW of Sunday Island, C.N. Rossiter s.n., 14.vi.1965 (MEL 516718): Port Arlington, Dickinson s.n.. 1870 (MEL): Tasmania: track to Seal Rocks, King Island, 40°06’S 143°58’E, MIH. Brooker 5855, 3,iv. 1978 (CANB, HO); Furgusons Gully, Flinders Island, 40°11’S 148°04’E, P.J. Cullen s.n., 21 ,ix. 1990 (HO 126142): Burgess Cove, Rocky Cape, 40°52’S 145°30 ’e] P Collier 1 126, 1 1 ,i. 1986 (HO); Mt Roland, 41°28’S 146°16'E, D.A. & A. V. Ratkowsky 1308, 27.x. 1974 (HO); Sensation Gorge, 6 km W of Mole Creek, 41°33’S 146°20 E, P. Collier 5123, 12. i. 1991 (HO); Sensation Gorge, c. 5 km W of Mole Creek, 41°33’S 146' 20 E, P. G. Neish 57-62, M.F. Duretto & K.R. Henshall, 6. i. 1996 (MEL); Cape Tourville, below the lighthouse, 4T°07’S 148°22’E J. Armstrong 879., 18,v. 1976 (HO); Bluff River Gorge, 42°32’S 147°40’E, P Collier 1553, 31.viii.1986 (HO); Midway Point. 42°48’S I47°31’E, A Moscal 8395, 3.vm.l984 (AD. HO): base of St Pauls Dome, C. Stuart s.n., xi. 1848 (MEL); slope near St Pauls Dome. W.H. Archer s.n. (NSW 3855335). Taxonomy: When describing B. variabilis, Hooker (1834) cited tour specimens which he classified into three forms (see Type citation). Of these four collections, Hooker ( 1 840) excludes all but the B form (Gunn 214) from B. variabilis indicating that he thought his a and y forms of B. variabilis were actually forms of B. tetrandra Labill. These speci- mens can be confidently placed in B. pilosa Labill. s. lat. which Hooker ( 1840) also con- sidered to be a form of B. tetrandra. Following Hooker ( 1 840) the K specimen of Gunn 2 14 is chosen as the lectotype of B. variabilis. When Cheel (1928) described B. dentigeroides he included all mainland populations of subsp. variabilis in it. However, his description was preceded by a week by an account by Penfold (1928) who described the essential oils of B. dentigeroides. In doing so, Penfold legitimately described B. dentigeroides. It was obviously not Penfold's intention to formally describe this species as he gives Cheel as the authority and states that the botany of this new species, Boronia dentigeroides , is fully described by its author. Mi. . Cheel, in the current issue of the Society's Journal . Later authors have quite light y ignored Penfold’s description. Chapman ( 1 99 1 ) in the Australian Plant Name Index lists B. anethifolia A. Cunn. var. variabilis (Hook.) Benth. Bentham (1863) did not make this combination but did relegate B. anethifolia to varietal status under B. anemonifolia. Notes: W.J. Hooker (1834. 1840) cited only collections from Tasmania when dis- cussing B. variabilis, while J.D. Hooker’s (1855) concept of B. variabilis included plants from New South Wales and south-eastern Australia. Later, B. variabilis was considered a Boronia anemonifolia 13 Tasmanian endemic (c.f. Bentham 1863; Curtis 1956) and Cheel (1928) included all mainland plants in his new species, B. dentigeroides, a name that never gained wide acceptance. Following Bentham (1863), Willis (1957) considered B. variabilis a variety of B. anemonifolia and noted that the two varieties were “grading imperceptibly on the islands of Bass Strait into the dentigeroides form of B. anemonifolia.” Curtis (1975) included B. variabilis in synonymy under B. anemonifolia and stated that “the species is polymorphic.” Mainland populations of subsp. variabilis have 3-5-foliolate or bipinnate leaves and leaflets usually with rounded tridentations at the apex (Fig. II). Plants from Tasmania have 3— 7-foliolate leaves and often quite narrow leaflets that are entire and sometimes recurved (Fig. 1H). These two forms grade into each other, but are united by the presence of veiy short sepals, short prophylls, usually glabrous leaves and pubescence on stems being confined to between the decurrent leaf bases. Specimens of subsp. variabilis from St Paul’s Dome in the north-west of the island (eg. Stuart s.n., xi. 1 848, MEL 275677; Archer s.n., NSW 385335 ) resemble subsp. anemonifolia in some respects. The long prophylls, long sepals, and persistent petals are charactei istic of subsp. anemonifolia , but the glabrous, wider leaves and numerous flow- ers resemble that of subsp. variabilis. These specimens are here treated to be subsp. vari- abilis. The small population of plants at Sensation Gorge (Tasmania; Collier 5123, Neish 57- 62) are notable in being pubescent over the entire stem, rather than just between the decunent leaf bases, and in having glabrescent leaves. Subspecies variabilis is known from south-eastern Queensland from collections near Warwick (A.R. Bean 10980, Sparshott 45). The Sparshott 45 collection from Paddys Knob is notable for its slightly pubescent foliage. No other specimens have been recorded from Queensland except for a collection labelled only as Mount Mitchell (MEL 275678) which may be near Toowoomba. Boronia anemonifolia subsp. variabilis differs from subsp. anemonifolia by having entire, flattened and usually glabrous leaflets, caducous petals, minute prophylls and smaller, but more numerous flowers, and from subsp. wadbilligensis by the smaller pro- phylls and glabrous to glabrescent leaves (Figs 1G-I). Distribution and ecology. Boronia anemonifolia subsp. variabils is found in south- eastern Queensland, the Northern, Central and Southern Tablelands and Central and South Coast of New South Wales, on the Bass Straight Islands, across northern Tasmania and on the hillsides around Hobart (Fig. 2). In Victoria, the subspecies is known only from Snake and Sunday Islands north of Wilsons Promontory and an 1870 collection from Portarlington on the Bellarine Peninsula. This area has been heavily degraded since settlement and tecent seaiches in the immediate vicinity of Portarlington have failed to locate any plants of B. anemonifolia. Boronia anemonifolia subsp. variabilis is found in heath, open woodland or open forest on sandy and rocky soils sometimes on or near sand- stone or granite outcrops. Conservation status'. A common subspecies that is adequately represented in reserves over its full range. Etymology. The subspecific epithet refers to the variable nature of the foliage which can be trifoliolate, pinnate or bipinnate. 2. Boronia rigens Cheel, ./. c£ Proc. Roy. Soc. New S.Wales 62: 297 (1929), a nom. et stat. nov. for Boronia poly galifolia Sm. var. robusta Benth., FI. aust. 1; 321 (1863). Type citation. Port Jackson, Sieber, n. 283; Blue Mountains, A. Cunningham; Moreton Island, F. Mueller ." Type : New Holland, Sieber FI. Novae holl. 283 (lectotype! here des- ignated, K n.v., cibachrome MEL 2041262; isolectotypes MEL 257414, MEL 62147); Moreton Island | label locality information probably incorrect], F. Mueller s.n., viii.1855 (syntypes MEL 257415, MEL 25741 6); Blue Mountains, A. Cunningham (syntype ? K n.v.). 14 P.G. Neish & M.F. Duretto Illustrations: A. Fairley and P. Moore, Native plants of the Sydney District, 235, t. 8 1 6 (1989); PH. Weston and M.F. Porteners, FI. New South Wales 2: 230 ( 1991 ). Prostrate to erect shrub to 30 cm tall and wide, the branches terete with moderate cork development on older branches; the stems and leaves not obviously glandular to slightly glandular; the whole plant glabrescent to pubescent, with hair density greatest between decurrent leaf bases on stems, leaves glabrescent or sparsely to moderately pilose; all hairs simple, erect, straight, to 1 mm long. Leaves smooth, 3(-5-7)-foliolate, rarely bip- innate. the leaf in outline 4— 1 0(— 1 2) mm long, 4-12 mm wide; petiole 1.5-5 mm long; rachis segments 1-3 mm long; leaflets linear to narrow elliptic, plane to carinate, con- colourous, dorsiventral, (2— )4— 8(— 1 2) mm long, (0.5—) 1—2 mm wide, lamina dotted with small sunken glands, hairs often concentrated along abaxial midrib, tip acute, midrib raised abaxially and distinctly red or brown, spongy mesophyll continuous under mid- vein. cell layer immediately above abaxial epidermis and below the midvein with sec- ondary thickening. Inflorescence axillary, 1-3-flowered,, glabrous or sparsely to moder- ately dense pilose; peduncle 0.5-2 mm long; prophylls unifoliate, 0.5-1. 5 mm long; metaxyphylls to 1.5 mm long, sometimes absent; anthopodium 0.5-1. 5 mm long. Sepals broad ovate to deltate, imbricate in bud, persistent with mature fruit, 1. 2-2.0 mm long, 0.5-1. 2 mm wide, c. 1/2 length of petals, tip acute; adaxial surface and margin covered in short woolly hairs that are c. 0.2 mm long, becoming glabrous towards base; abaxial surface glabrous or sparsely to moderately pilose, hairs concentrated along the midrib. Petals white or pink, imbricate in bud, 2.5-3.5 mm long, 1. 0-2.5 mm wide, glabrous or margins sparsely ciliate, persistent with mature fruit, glands often concentrated along midrib, midvein not raised on the abaxial surface, tip with a small but distinct incurved hook. Stamens erect. 8, all fertile; filaments gradually tapering to anther connective, pilose on margins below the slightly glandular, obtuse tip; antesepalous hlaments c. 1 .5 mm long; antepetalous filaments 1.0-1. 2 mm long; anthers attached sub-apically on the filament, anther-connective maroon; anther-apiculum prominent, glabrous. Disc entire, entirely within staminal whorl, glabrous. Ovary glabrous, or with a few hairs between carpels; style pilose at base, or whole style pilose; stigma rounded, as wide or slightly wider than style. Cocci glabrous or sparsely pubescent, 2. 5-3. 5 mm long, 1. 5-2.0 mm wide. Seeds elliptical in outline, 2—4 mm long, 1—2 mm wide, black or black-brown, dull, slightly tuberculate, microscopic wax crystals on surface between tuberculae. Selected specimens examined (of c. 100 collections): New South Wales: Head ot Coricudgy Ck, 3 miles [c. 4.8 km] N of Mt Coricudgy, L.A.S. Johnson s.n., 30.viii. 1 95 1 (NSW 17273): 1 mile east of Currant Mountain Gap, Rylstone District, B.G. Briggs s.n., 6.viii. 1 96 1 (NSW 385503): Bunboori Ck, via Waratah Ridge, Blue Mountains NP, 33°25’S 150°17'E, P. Hind 6340, 1 l.iii.1991 (NSW); Mount Banks, on North slope, 33°36’S 150°21 'E, D. Benson 2319 & D. Keith, 15. ii. 1985 (NSW); La Perouse, 33°59'S 151°14’E. R. Coveny 1 1 199, 28.vii.1982 (CANB); 3.8 km along Tevotts Road, c. 6 km SW of Bundanoon, 34°41’S 150°18'E, P. Ollerenshaw 1740, 28.vii.1985 (CANB); Northern Budawang Range, Wog Wog Creek - Corang Trig track, 35°27’S 150°05'E. I P. Telford 9549, 15.viii. 1976 (CANB); Endeavour Reserve, around base of communi- cation tower, c. 1 km South of Bombala, 36°55'S 149°15’E, P.G. Neish 74 & A.J. Vadala, lO.iv. 1996 (MEL). Notes: Cheel s concept of B. rigens was based on discordant elements that included specimens of B. nana from South Australia (Mt. Lofty Range) and Victoria. Although superficially similar to B. nana var. nana. B. rigens can be distinguished by leaflets with distinctly raised midribs on the abaxial surface that are usually red and darker than the rest of the leaf. This dark colouration is due to large pigment-filled cells in the midrib. Distribution and ecology: Boronia rigens is found between Bombala and Rylstone on the Southern and Central Tablelands and in coastal areas around Sydney, New South Wales (Fig. 2). It is usually found in woodland or heath on sandy or stony soils associat- ed with sandstone. One of the type collections of B. polygalifolia var. robusta is a specimen collected by Boronia anemonifolia 15 Ferdinand Mueller which supposedly came from Moreton Island near Brisbane, south- eastern Queensland. No other collections of B. rigens have been made north of Rylstone and it is assumed that Mueller’s specimens were mislabelled. Conservation status'. Though found over a limited area, B. rigens is well represented in National Parks and is not considered to be rare or threatened. Etymology. The specific epithet is derived from the Latin, rigens, for stiff or rigid and is possibly referring to the stiff or rigid habit of the plant as compared to the lax B. poly- galifolia and B. nana. Acknowledgements We thank the Directors and Curators of BRI, NE, NSW, CANB, HO, AD. PERTH. LUND and TCD who provided necessary loan material; Don Foreman (while working as ABLO at K) for locating type material at K and BM; Ian Thompson, Katherine Henshall and Anthony Vadala for assistance with field work; The Bendigo Field Naturalists Club who provided assistance with fieldwork and localities of the subsp. aurifodina. This proj- ect was funded by the Australian Biological Resources Study, Grant Ref N°. 94/536. References Baker, R.T. (1899). Contributions to a knowledge of the flora of Australia. Proceedings of the Linnean Society of New South Wales 24, 437-447. Lehman, J.G.C. ( 1 848). Diosmeae. "Plantae Preissianni.’ 2, 226-8. Sumptibus Meissneri: Hamburg. Bentham, G. (1863). ‘Flora australiensis’, Vol. 1. (Lovell, Reed and Co.: London.) Briggs, B.G., and Johnson, L.A.S. (1979). Evolution in the Myrtaceae - evidence from inflores- cence structure, Proceedings of the Linnean Society of New South Wales 102 , 157-256. Briggs, J.D., and Leigh, J.H. (1996). ‘Rare or Threatened Australian Plants.’ Revised edn. (CSIRO Australia: Collingwood). Cheel, E. (1928). Further notes on the genus Boronia. Journal and Proceedings of the Royal Society of New South Wales 62,290-302. Chapman, A.D. (1991). ‘Australian Plant Name Index, A-C.’ Australian Flora and Fauna serial No 12. (Australian Government publishing Services: Canberra.) Cunningham, A. (1825). ‘A specimen of the indigenous botany of the mountainous country between the colony round Port Jackson and the settlement of Bathurst; being the results of observations made in the months of October, November and December, 1 822.’ In B Field (Ed) ‘Geographical Memoirs of New South Wales’, pp. 323-65. (John Murray: London.) Curtis, W.M. (1956). ‘The Students Flora of Tasmania’, Part 1 . (University of Tasmania: Hobart ) Curtis, W.M. (1975). ‘The Students Flora of Tasmania’, Part 1. 2nd Edn. (Government Printer Tasmania.) Ewart, A.J. (1930). ‘Flora of Victoria.’ (University of Melbourne Press: Melbourne.) Holmgren, P.K., Holmgren, N.H., and Barnett, L. (1990). ‘Index herbariorum. Part 1. The Herbaria of the World.’ 8th Edn (New York Botanical Garden: New York). Hooker, J.D. (1855). ‘The botany of the antartic voyage of H.M. discovery ships Erebus and Terror in the years 1 839-1843, Part III Flora Tasmaniae, Vol. I Dictyledons.’ (Lovell Reeve- London ) Hooker. W.J. (1834). Contributions towards a flora of Van Dieman's Land. Hooker’s Journal of Botany 1 , 241-258. Hooker, W.J. (1840). Contributions towards a flora of Van Dieman’s Land. Hooker’s Journal of Botany 2 , 399-42 1 . J Mueller, F.J.H. (1855). Description of fifty new Australian plants, chiefly from the colony of Victoria. Transactions and Proceedings of the Victorian Institute for the Advancement of Science 1854 - 1855 , 28-48. Mueller, F.J.H. (1860-1862). ‘The plants indigenous to the colony of Victoria, Vol. 1 Thalamiflore’ (Victorian Government: Melbourne.) Mueller. F.J.H. (1882). ‘Systematic census of Australian plants. Part 1 - Vasculares ’ (Victorian Government: Melbourne.) Mue J' ei i’ FJ H - (l888) - ‘ Ke y t0 the system of Victorian plants.’ (Victorian Government Printer Melbourne.) 16 P.G. Neish & M.F. Duretto Mueller, F.J.H. (1889). ‘Systematic census of Australian plants. Part 1 - Vasculares.’ 2nd edn. (Victorian Government: Melbourne.) Murley, M.R. (1951). Seeds of Cruciferae of northeastern North America. The American Midland Naturalist 46 , 1 - 81 . Penfold, A.R. (1928). The essential oil of a new species of anemone leaf Boronia rich in ocimene. Journal and Proceedings of the Royal Society of New South Wales 62 , 263-271. Rodway, L. (1903). ‘The Tasmanian flora.’ (John Vail, Government Printer, Hobart.) Ross, J.H. (1996). ‘A Census of Vascular Plants of Victoria.’ (Royal Botanic Gardens Melbourne: Melbourne.) Weston, P.H. (1990). Notes on Boronia (Rutaceae) in New South Wales, including descriptions of three new species. Telopea 4 , 121-128. Weston, P.H., and Porteners, M. ( 1991). Boronia. In ‘Flora of New South Wales, Volume 2'. (Ed G. Harden.) pp. 227-36. (New South Wales University Press: Sydney.) Willis, J.H. (1957). Flora of Victoria and South Australia. Victorian Naturalist 73. 193-194. Willis, J.H. (1973). ‘A handbook to plants in Victoria, Vol. II: Dicotyledons’ (Melbourne University Press: Melbourne.) Wilson, P.G. (1998). New names and new taxa in the genus Boronia (Rutaceae) from Western Australia, with notes on seed characters. Nuytsia 12 , 119-154. Muelleria 14:17 (2000) The lichens of Nothofagus cunninghamii- dominated rainforests and Acacia melanoxylon -dominated forests in the Otways, Victoria. Sharon Ford, Maria Gibson and Geoff Duke Biological and Chemical Sciences Deakin University, Rusden Campus 662 Blackburn Road, Clayton, VIC, 3168. Abstract Lichens occurring in rainforests dominated by Nothofagus cunninghamii (Hook.) Oerst. and in forests dominated by Acacia melanoxylon R. Br. were examined in the Otway Ranges, southwest Victoria. A total of 1 10 species were recorded, 93 occurred in N. cunninghamii rainforests and 67 in A. melanoxylon forests. Fifty of these species were common to both forest types. In total, 17 lichen species are newly reported for Victoria. Introduction The Otway Ranges are approximately 200 km southwest of Melbourne, Victoria (D.N.R.E. 1996). Calder (1990) described the region as an island, with Bass Strait to the south and the vast basalt plains to the north effectively isolating the area. There are no comparable regions closer than 400 km away (Calder 1990). The Otway Ranges extend approximately 80 km, and contain a wide variety of forest types including examples of heathlands, coastal vegetation, wet mountain forests and cool temperate rainforest (D.N.R.E. 1996). The Otways have a temperate climate. Most rain falls between May and September, with the mean annual rainfall varying between 1750 to 2000 mm (Brinkman & Farrell 1990). Mean maximum temperatures range from 20°C on the coast to 27°C inland in the warmest months, January and February (Brinkman & Farrell 1990). The coldest months see average minimums of 3 to 4°C (Brinkman & Farrell 1990). There is little published information on the lichen flora of the Otways region, although historically the region has attracted some attention from past collectors, for example, R. Filson in 1963-4, J.H. Willis in 1955 and 1963 and A.C. Beauglehole in 1953. Indeed, there are comparatively few studies on lichens in Victorian rainforests. Louwhoff (1995) examined the lichen floristics of Mt. Donna Buang Scenic Reserve, which included some small pock- ets of rainforest, and Wedin (1995), in his review of the lichen family Sphaerophoraceae in the southern hemisphere, examined the genera Bunodophoron A. Massal. and Lefidium Wedin from Victorian rainforest. Recently, a revision of the lichen genus Usnea Dill, ex Adans. included species found in rainforests in Victoria (Stevens 1999). Cool temperate rainforests are cited as one of the most important vegetation commu- nities in the Otways (D.N.R.E. 1996) and are dominated by Nothofagus cunninghamii (Hook.) Oerst. Many Victorian rainforests are co-dominated by N. cunninghamii and Atherosperma moschatum Labill., however, the latter species appears to be absent from the Otways (Busby & Brown 1994; Peel 1999), and is apparently being replaced by Hedycarya angustifolia A. Cunn. for the most part, but also in part by Pittosporwn bicol- or Hook., Olearia argophylla (Labill.) Benth. and Acacia melanoxylon R.Br. Dicksonia antarctica Labill. is always present as an understorey along with ground ferns such as Blechnum wattsii Tindale and/or Polystichum proliferum (R. Br.) Presl. The Otways N. cunninghamii- dominated rainforests are classified as “callidendrous rainforest” (after Jarman et al. 1991 ) or “Otways cool temperate rainforest” (after Peel 1999). Acacia melanoxylon is a fast growing species (Floyd 1989) that forms almost pure stands in the Otways. The species is dominant in areas of past disturbance and frequent 18 S. Ford, M. Gibson and G. Duke fire, which tends to favor the establishment of Acacia species (Busby & Brown 1994). Physiognomical ly, these almost pure stands of A. melanoxylon might be classified as either callidendrous or thamnic rainforests (after Jarman et al. 1 99 1 ) as they have a struc- tural resemblance to true rainforest (closed canopies, multi-strata vegetation, ferny under- storey and high dominance of epiphytes). However, they are floristically not considered rainforests unless there is significant co-dominance by N. cunninghamii or other recog- nised rainforest trees (D.C.F.L. 1987; Peel 1999). These forests are considered ‘tall open forests’ by Ashton and Attiwill (1994). Other associated species include Nematolepis squamea (Labill.) Paul. G. Wilson, Coprosma qitadrifida (Labill.) Robinson, Olearia argophylla and Pomaderris aspera Sieb. ex DC. Dicksonia antarctica and B. wattsii are often present as understorey species. This paper examines some of the similarities and differences in lichen composition between rainforests dominated by Nothofagus cunninghamii (Hook.) Oerst. and the struc- turally similar tall open forests dominated by Acacia, melanoxylon R. Br. in the Otways, Victoria. The study forms part of a much larger, ongoing investigation examining the lichen communities of cool temperate rainforest throughout Victoria. Methods A total of 20 randomly selected quadrats were sampled, 10 in rainforests dominated by Nothofagus cunninghamii and 10 in forests dominated by Acacia melanoxylon. Due to the very limited extent of rainforest in Victoria, the more traditional map-grid method of randomly selecting sites was not appropriate. Victorian rainforests occur as small pockets in moist gullies and along creeks and rivers (Busby & Brown 1994; D.C.F.L. 1987). Those of the Otways are no exception. Methods using larger scale grids would certainly miss most of these pockets. Instead, potential sites of both N. cunning- hamii- dominated rainforest and, for consistency, A. melanoxylon-dominated forests were located through study of maps and by field surveys. Appropriate sites were listed and a computer-generated random number table was used to select actual study sites. Quadrats measuring 20m x 20m (after Louwhoff 1992, 1995) were placed in the approximate centre of the selected forest pockets, where edge effect was considered mini- mal. Five trees of each species present in four size classes (Table 1) were sampled where they were present in the quadrat. Different size classes of trees represent different ages and, hence, have varying bark characteristics. This factor is known to influence lichen floristics (Adams & Risser 1971; Griffin & Conran 1994; Kantvilas 1990), so each size class must be examined to obtain an accurate representation of the lichen flora present in the forest pocket under investigation. An analysis of lichen variation between size class- es was not undertaken as this would require a much larger sample size than that obtained. Lichen species and percent cover-abundance were recorded up to a height of 2 metres. Table 1 . Size classes used for the major tree species in Nothofagus cunninghamii - dominated rainforests and Acacia melanoxylon - dominated forests in Victoria. Size class Radius at chest height (cm) Small <5 Medium 5-10 Large 10-20 Extra Large 20+ Lichens of Nothofagus and Acacia 19 this being the practical limit of accessibility. Recently fallen branches and sticks were collected from the forest floor in each quadrat as a representation of what was present in the canopy (Jarman & Kantvilas 1995a). Nomenclature follows Filson (1996), Flora of Australia Volumes 54 and 55 (1992 and 1994 respectively). Wedin (1995) for the genera Bunodophoron and Lefidium and Stevens (1999) for the genus Usnea. Distinct lichen species not able to be matched against avail- able literature were assigned “sp. 1” etc. A 2-dimensional Non-metric Multidimensional Scaling (NMDS) ordination tor the average cover of each lichen species in the 20 quadrats was performed using CLUSTER. MDS and CONPLOT in the statistical package PRIMER (Plymouth Routines in Multivariate Ecological Research). NMDS constructs a configuration of the samples, which attempts to satisfy all the conditions imposed by the rank similarity matrix (Clarke & Warwick 1994). Bray-Curtis similarity matrix was used, which is appropriate for delin- eating groups of sites with distinct community structure and patterns of abundance (Clarke & Warwick 1994). The data were given 4th root transformation in order to even- ly weight both the dominant species and the rare species (Clarke & Warwick 1994). A dendrogram results from the CLUSTER option of PRIMER, and is presented as further explanation of the relationships between the similarities of quadrats. Results In total, 495 trees, representing 10 different vascular species, were examined (Table 2). A total of 1 10 lichen species were recorded, 93 from rainforests dominated by N. cunning- hamii and 67 from forests dominated by A. melanoxylon. Fifty species were common to both forest types (Appendix I). Of the 93 lichen species occurring in N. cunninghamii rainforests, 66 were recorded in the canopy and 38 in the lower trunk region. Thirteen species were found in both the canopy and lower trunk regions (Fig. 1 ). Similarly, the canopy of A. melanoxylon forests had a high- er species richness than the lower trunk region with 43 and 35 species respectively and 12 species common to both. The lower trunk region of both forest types appeared to be com- parable in terms of lichen species richness over the 20 quadrats examined. Table 2. Number of trees of each species sampled. Tree species Total sampled Nothofagus rainforest Acacia forest Nothofagus cunninghamii 113 102 11 Hedycarya angustifolia 98 45 53 Acacia melanoxylon 91 7 84 Dicksonia antarctica 68 43 25 Coprosma quadrifida 53 19 34 Pittosporum bicolor 27 1 1 16 Olearia argophylla 19 9 10 Nematolepis squamea 13 0 13 Pomaderris aspera 11 0 11 Cyathea australis 2 0 2 Total: 495 236 259 20 S. Ford, M. Gibson and G. Duke 100 90 80 70 d40 Z 30 20 10 0 Not ho fag us Acacia Forest Type Figure 1. Lichen species-richness of the lower trunk and canopy, showing over- lap of species common to both. In total, 30 families were represented in the forests studied, 29 occurred in N. cun- ninghamii rainforests and 24 occurred in A. melanoxylon forests. All families recorded in A. melanoxylon forests were found in N. cunninghamii rainforests, with the exception of the Ramalinaceae. Six families were exclusive to N. cunninghamii rainforest: Arthoniaceae, Chrysothricaceae, Coccotremataceae, Lecanactidaceae, Nephromataceae and Teloschistaceae. In all cases, representatives from these families were not recorded frequently in the study areas, therefore, there may be insufficient data from which to draw conclusions. Ten families were well represented in N. cunninghamii rainforests, with four or more species being recorded for each (Table 3). In A. melanoxylon forests, six families were recorded with more than four species each. In many cases, these families were repre- sented by only one genus, for example, the Thelotremataceae by the genus Thelotrema Ach., the Pertusariaceae by Pertusaria DC. and the Usneaceae by Usnea Dill, ex Adans. The families Cladoniaceae, Lobariaceae. Pertusariaceae, Sphaerophoraceae and Thelotremataceae appear to be important groups for N. cunninghamii rainforests as they have large numbers of representative species, and were rarely recorded in A. melanoxy- lon forests (Table 3). Similarly the Collemataceae appears to be an important family in the A. melanoxylon forests. □ Canopy □ Overlap ■ Trunk Table 3. Commonly represented families in Nothofagus cunninghamii- dominated rainforests and Acacia melanoxylon- dominated forests in the Otways. Family Nothofagus -dominated Acacia-dominated No. Genera No. Species No. Genera No. Species Cladoniaceae 2 4 Collemataceae 2 4 Hypogymniaceae 2 8 2 5 Lecideaceae 4 5 3 5 Lobariaceae 1 7 Pannariaceae 1 5 2 4 Parmeliaceae 6 12 5 8 Pertusariaceae i 6 Sphaerophoraceae 2 6 Thelotremataceae i 4 Usneaceae i 5 1 7 Lichens of Nothofagus and Acacia 21 Canopy species were included in the above figures, however, due to the haphazard nature of collecting lichens from the canopy, these data were not included in the follow- ing analyses. It is possible to indicate which species are present in the canopy, but the absence of species could just be a reflection of the sampling procedure. The number of species per quadrat ranged from as low as two species in A. melanoxy- lon forest, and up to 20 species in N. cunninghamii rainforest (Fig. 2). On average, N. cunninghamii rainforests supported a higher species richness (mean of 12.8 species) than A. melanoxylon forests (mean of 8.7 species), however, an independent samples t test revealed that this difference was not significant (p = 0.085). Figure 2. Species richness per quadrat in the two forest types. A t test on the difference between the means was not significant with p = 0.085. On consideration of the common species from the two forest types (that is species occurring in six or more of the ten quadrats), there appears to be a distinct flora for each forest type, with two species, Lepraria sp. and Thelotrema lepadinum (Ach.) Ach. being common to both (Table 4). Three species, Bunodophoron australe (Laurer) A. Massal, Metus conglomeratus (F. Wilson) D.J. Galloway & P. James and Pyrenula nitida (Weig.) Table 4. Comparison of common species (occurring in six or more of the ten quadrats) in the lower trunk region of Nothofagus cunninghamii- dominated rainforests and Acacia melanoxylon - dominated forests. Nothofagus cunninghamii rainforest Acacia melanoxylon forest Bunodophoron australe Bunodophoron murrayi Cladia aggregatci Metus conglomeratus Pseudocyphellaria dissimilis Pseudocyphellaria glabra Pyrenula nitida Thelotrema subclenticulatum Lepraria sp. Lepraria sp. Thelotrema lepadinum Thelotrema lepadinum 10 species Bacidia buchananii Leptogium victorianum 4 species 22 S. Ford, M. Gibson and G. Duke Table 5. Comparison of common species (occurring in six or more of the ten quadrats) from the canopy of Nothofagus cunninghamii- dominated rainforests and Acacia melanoxylon- dominated forests. Nothofagus cunninghamii rainforest Acacia melanoxylon forest Coccotrema cucurbitula Menegazzia myriotrema Menegazzia sp. 1 Pertusaria novaezelandiae Sarrameana tasmanica Thelotrema sp. B Maronea constans Maronea constans Megalaria grossa Megalaria grossa Parmelia tenuirima Parmelia tenuirima Thelotrema lepadinum Thelotrema lepadinum 10 species Hypogymnia mundata Normandina pulchella Parmelina quercina Pertusaria gibherosa Pyrrhospora laeta Tephromela atra 11 species Ach., were not recorded in A. melanoxylon forests. In the canopy (Table 5), differences in species composition also can be seen, with four species, Maronea constans (Nyl.) Hepp., Megalaria grossa (Pers. ex. Nyl.) Hafellner, Parmelia tenuirima Hook. f. & Taylor and T. lepadinum being common to both forest types. A NMDS ordination configuration of average cover abundance of lichen species in each quadrat revealed two groups of quadrats with 2 outlier quadrats (Fig. 3). The dotted lines enclose groups of sites with a similarity of 32 percent, as shown in the dendrogram (Fig. 4). There appeared to be a lichen flora that was characteristic of N. cunninghamii rainforest and a lichen flora characteristic of A. melanoxylon forest as both forests clus- tered separately. The exceptions were quadrat numbers 07A and 02A which showed a high similarity to the quadrats dominated by N. cunninghamii. The dendrogram (Fig. 4) clustered these two quadrats with the N. cunninghamii forests at 55 and 50 percent simi- larity, respectively. Quadrat 07A was dominated by A. melanoxylon , but had a subdomi- nant canopy of N. cunninghamii. Quadrat 02A appeared to have a fairly common collec- tion of lichen species, with no rare or unusual species, which may contribute to its simi- larity to N. cunninghamii forests. Indeed, this quadrat may be clustered arbitrarily with the other A. melanoxylon forests on the NMDS (Fig. 3). The outliers 09A and ION recorded the lowest number of species for their respective forest types. A NMDS ordina- tion was also done for presence/absence data and lichen cover per area of tree sampled. Results were essentially the same with only minor variations in distance configuration. New Records All lichen species have previously been found in Australia, but a total of 17 new records for Victoria were recorded during this study (Table 6). Of these, Bunodophoron murrayi (Ohlsson) Wedin, Coccotrema cucurbitula (Mont.) Mull. Arg., Graphis insidiosa (C. Lichens of Nothofagus and Acacia 23 Figure 3. Mean cover of lichen species. NMDS of 20 sites based on 4 lh root transformed abundances and Bray-Curtis similarities (Stress = 0.14). N = quadrats dominated by Nothofagus cunninghamii\ A = quadrats dominated by Acacia melanoxylon. 06A 03A I 05A / I \ 04A Figure 4. Mean cover of lichen species. Dendrogram of 20 quadrats using aver- age clustering from Bray-Curtis similarities on 4 th root transformed cover abundance data. : ION 09A 03N 08N Q6N L 09N — Q2A r ^ 01 N 02N 04 N 05N 07N 07A i 03A Q6A Q1A 05A 04A 08A i k_2 , | , I | • | | 10A 20 30 40 50 60 70 80 90 100 24 S. Ford, M. Gibson and G. Duke Table 6. Lichen species newly recorded for Victoria. Species Herbarium No.* Spp. description ref. Bacidia laurocerasi MEL; SF 120 Galloway 1985 Bunodophoron murrayi MEL 2085466: SF1 10 Galloway 1985; Wedin 1995 (as Sphaerophorus murrayi ) Cliostomum griffithii MEL 2085462: SF 108/ 109 Galloway 1985 Coccotrema cucurbitula MEL 2085468 Galloway 1985; Kantvilas 1990a Graphis insidiosa MEL 2085463: SF119 Galloway 1985; Kantvilas & James 1991 Lecidea immarginata MEL, SF123 Kantvilas & James 1991 Lepraria lobificans MEL 2085461: SF 142/1-57 Kantvilas & James 1991 Menegazzia eperforata MEL 2085469: SF158 Flora of Australia 1992 Micarea prasina MEL 2085467 Coppins 1983 Parmelia protosulcata MEL 2085464 Flora of Australia 1994 Parmotrema robustum MEL 2060624: SF1 16 (from W. Prom) Flora of Australia 1994 Phaeographis exaltata MEL 2085465: SF1 17/118 Galloway 1985; Kantvilas & James 1991 Phlyctis subuncinata MEL 2085470 Galloway 1985; Kantvilas & James 1991 Sarrameana albidoplumbea MEL 2085471 Kantvilas & Vezda 1995 Usnea oncodes MEL 2085460: SF144/150 Stevens 1999 Usnea pycnoclada SF149 Stevens 1999 Usnea xanthopoga MEL 2085472: SFI51 Stevens 1999 * MEL = Melbourne Herbarium, Royal Botanic Gardens, SouthYarra SF = Personal herbarium of Sharon Ford Knight & Mitt.) Hook, f., Lecidea immarginata R. Br. ex Cromb., Menegazzia eperfora- ta P. James & D.J. Galloway, Micarea prasina Fr., Parmelia protosulcata Hale, Usnea pycnoclada Vainio and U. xanthopoga Nyl. are not surprising additions to Victoria’s lichen flora as they are currently known from Tasmania and New South Wales. These records simply fill in the previous disjunct distribution. Cliostomum griffithii (Sm.) Coppins, Lepraria lobificans Nyl., Phaeographis exaltata (Mont. & v.d. Bosch) Mull. Arg., Phlyctis subuncinata Stirton, Sarrameana albidoplumbea (Hook. f. & Taylor) Farkas and Usnea oncodes Stirton are new for the mainland and Bacidia laurocerasi (Delise ex Duby) Zahlbr. and Parmotrema robustum (Degel.) Hale have their southern most record in the Otways. Menegazzia e perforata and P. robustum were recently report- ed from the rainforests at Wilsons Promontory by one of the authors (Ford & May, 1 998), however, no formal publication of these records was made at the time. Discussion Currently, the number of macrolichen species in Tasmanian rainforests stands at 355 species, 200 macrolichens and 155 crustose lichens (Jarman & Kantvilas 1995a). The total number of lichen species currently known for Victorian rainforests is 149 species, Lichens of Nothofagus and Acacia 25 with 94 macrolichens and 54 microlichens (compiled from Ford unpubl.; Ford & May 1998; Louwhoff 1995). Victorian rainforest lichen figures are projected to increase con- siderably as more work is completed. The lichens of Tasmanian rainforests and wet scle- rophyll forests have received much attention in the past decade (Brown et al. 1994; Jarman & Kantvilas 1994; Jarman & Kantvilas 1995a; Jarman & Kantvilas 1995b; Kantvilas 1990; Kantvilas 1995; Kantvilas & James 1991; Kantvilas & Jarman 1991; Kantvilas & Jarman 1993), allowing for comparison of the lichen flora of the Otways with published lists for Tasmanian forests. A large proportion of the lichens found during this study may be considered “ubiqui- tous wet forest” species (Kantvilas & Jarman 1993). Of the 31 lichen species listed as ubiquitous by Kantvilas & Jarman (1993, p. 219), 19 were recorded from the Otways. Of these, all 19 species were found in N. cunninghamii rainforest, and 14 species were found in A. melanoxylon forest. Common examples include: Bacidia buchananii (Stilt.) Hellb., Cladia aggregate, i (Sw.) Nyl., Parmelia tenuirima Hook f. & Taylor, Psoroma microphyl- lizans (Nyl.) D.J. Galloway and Pseudocyphellaria glabra (Hook f. & Taylor) C.W. Dodge. Of the 26 “typical rainforest species” listed by Jarman and Kantvilas (1993), seven were found in this study. All seven species were recorded in N. cunninghamii rain- forest, and four species (Sarrameana albidoplumbea (Hook f. & Taylor) Farkas (syn. Bacidia albidoplumbea), Micarea prasina Fr., Thelotrema lepadinum (Ach.) Ach. and Usnea oncodes Stirton) were found in the tall open forests dominated by A. melanoxylon. Seven out of 34 “non-rainforest species” (Kantvilas & Jarman 1993), or species that gen- erally characterise sclerophyll forest, were also found, five in N. cunninghamii rainforest and 6 in A. melanoxylon forest. Many of the lichens recorded in this study are considered “widespread” lichens for the three Tasmanian rainforest suballiances: callindendrous, thamnic and implicate (Jarman & Kantvilas 1995a; refer to Jarman et al. 1991 for rainforest descriptions). Of the 44 widespread rainforest lichens listed (Jarman & Kantvilas 1995a), 28 were found in the Otways. Twenty-five of these species were recorded from N. cunninghamii rain- forest and 15 species from A. melanoxylon forest. Of the 18 species listed as being restricted to callidendrous rainforest, five species were recorded in A. melanoxylon forests. These were S. albidoplumbea, Bunodophoron murrayi, Menegazzia myriotrema, P. tenuirima and Pseudocyphellaria dissimilis (Nyl.) D.J. Galloway & P. James. These five species and an additional three species (Arthothelium interveniens (Nyl.) Zahlbr., Bunodophoron ramuliferum (I.M. Lamb) Wedin and Lecanactis abietina (Ach.) Korb.) were found in N. cunninghamii rainforest. It was interesting to note that some typical rainforest lichen species are recorded in A. melanoxylon- dominated tall open forest. The observed structural similarities between N. cunninghamii rainforests and A. melanoxylon forests may lead to microclimatic correla- tions, explaining some of the similarity in lichen species composition. The presence of N. cunninghamii as a component of the canopy in some A. melanoxylon dominated forests also may be a contributing factor. Just as different forms of rainforest have a ubiquitous group of lichen species and a number that are characteristic of each forest type (Jarman & Kantvilas 1995a), the results suggest that N. cunninghamii rainforests and A. melanoxylon forests also have both a distinct flora and a shared flora. Seventy-nine per- cent of lichen families were represented in both forest types, while 45 percent of species were common to both. The remainder are divided. 39 percent of lichen species are con- fined to N. cunninghamii rainforest and 15 percent of species confined to A. melanoxylon forests. In studies comparing rainforest to sclerophyll forest, rainforests consistently record- ed higher species richness (Jarman & Kantvilas 1994; Kantvilas & Jarman 1993; Louwhoff 1995). However, the differences may be minimal, for example Kantvilas & Jarman (1993) recorded 79 species in disturbed rainforest in Tasmania and 72 species in surrounding sclerophyll forest, a difference of seven species. Jarman and Kantvilas (1994) 26 S. Ford, M. Gibson and G. Duke recorded 72 and 78 lichen species in two types of Eucalyptus-dom'maled sclerophyll for- est, compared to 83 species in rainforest, with a minimum difference of 5 species. Louwhoff (1995) recorded 56 species in N. cunninghamii- dominated rainforest and 42 species in Eucalyptus- dominated sclerophyll forests at Mt. Donna Buang Scenic Reserve, Victoria, a difference of 14 species. A considerably greater difference in species number was found between the two for- est types considered here, when compared with similar studies comparing forest types in south-eastern Australia. Overall (canopy inclusive), a difference of 26 species was found, with higher species richness recorded in N. cunninghamii rainforests (93 spp. compared to 67 spp. in A. melanoxylon forests). Lichenologically, cool temperate rainforests are recognised as one of the most floristi- cally diverse forest types (Jarman & Kantvilas 1995a; 1994; Kantvilas 1990). The general- ly low species richness in A. melanoxylon tall open forests may be an indication that these forests are quite distinct from N. cunninghamii- dominated rainforest despite their structur- al similarity. However, they share a high proportion of lichen species with the N. cunning- hamii rainforests, some of which are considered true rainforest lichen species. Acknowledgments The authors would like to thank Deakin University (Rusden Campus), Barwon Water and the Department of Natural Resources and Environment (Apollo Bay Office) for their con- tinuing support during this research. Thank you also to Gintaras Kantvilas, Jack Elix, Nell Stevens, Patrick McCarthy and Alan Archer for their kind aid in the identification and confirmation of certain species. This paper was presented, in part, at the Australasian Lichenologists Conference held in Dunedin, New Zealand, 15-19 November 1998. References Adams, D. B. and Risser, P. G. (1971). Some factors influencing the frequency of bark lichens in North Central Oklahoma. American Journal of Botany 58: 752-757. Ashton, D.H. and Atti will, P.M. (1994). Tall open-forests. In ‘Australian Vegetation’. (Ed R.H. Groves) pp. 157-196. (Cambridge University Press: Melbourne). Brinkmann, R. and Farrell. S. ( 1 990). Statement of resources, uses and values for the Otways Forest Management Area. Department of Conservation and Environment, Victoria. Brown, M. J., Jarman, S. J. and Kantvilas, G. (1994). Conservation and reservation of non-vascular plants in Tasmania, with special reference to lichens. Biodiversity and Conservation 3: 263-278. Busby, J. R. and Brown, M. J. (1994). Southern Rainforests. In ‘Australian Vegetation’. (Ed R.H. Groves) pp. 131-156. (Cambridge University Press: Melbourne). Calder, J. (1990). Parks: Victoria’s National and State Parks. Victorian National Parks Association, Canterbury Press Pty. Ltd., Melbourne. Clarke, K. R. and Warwick, R. M. (1994). Change in Marine Communities: an approach to statistical analysis and interpretation. Natural Environment Research Council, Plymouth. United Kingdom. Department of Conservation. Forests and Lands. (1987). Victoria’s Rainforests: an overview. Victoria, Australia. Department of Natural Resources and Environment. (1996). Otway National Park management plan. Department of Natural Resources and Environment (National Parks Service), Victoria. Filson, R. B. (1996). Checklist of Australian Lichens and Allied Fungi. Australian Biological Resources Study, Canberra. Flora of Australia. (1992). Flora of Australia Volume 54. Lichens - Introduction, Lecanorales /. Australian Government Publishing Service, Canberra. Flora of Australia. (1994). Flora of Australia Volume 55, Lichens - Lecanorales 2, Parmeliaceae. Australian Biological Resources Study, Canberra. Floyd, A. G. (1989). Rainforest trees of mainland south-eastern Australia. Inkata Press, Melbourne. Lichens of Nothofagus and Acacia 27 Ford, S. E. and May, T. (1998). A cryptogamic extravaganza. Report of the FNCV expedition. 22-25 May, 1998, to commemorate the centenary of Wilsons Promontory National Park. Field Naturalists Club of Victoria. Griffin, M. and Conran, J. G. (1994). Ecology of the corticolous lichens on Pinus radiata at five sites of increasing age near Linton, Victoria, Australia. Australian Journal of Ecology 19: 328-335. Jarman, S. J. and Kantvilas, G. (1994). Lichens and bryophytes of the Tasmanian world heritage area II - Three forest sites at Pelion Plains. Tasf orests 6: 102-120. Jarman, S. J. and Kantvilas, G. ( 1 995a). A Floristic Study of Rainforest Bryophytes and Lichens in Tasmania’s Myrtle-Beech Alliance. Tasmanian NRCP Report No. 14. Forestry Tasmania and Department of the Environment, Sport and Territories, Canberra, Australia. Jarman, S. J. and Kantvilas, G. (1995b). Epiphytes on an old Huon pine tree (Lagarostrobos franklinii) in Tasmanian rainforest. New Zealand Journal of Botany 33: 65-78. Jarman, S.J., Kantvilas, G. and Brown, M.J. (1991). Floristic and ecologiccd studies in Tasmanian rainforest. Tasmanian NRCP Report No. 3. Forestry Commission, Tasmania, and Department of the Arts, Sport, the Environment, Tourism and Territories, Canberra, Australia. Kantvilas, G. (1990). Succession in rainforest lichens. Tasforests 2: 167-171. Kantvilas, G. (1995). A revised key and checklist for the macrolichens in Tasmanian cool temperate rainforest. Tasforests 7: 93-127. Kantvilas, G. and James, P. W. (1991). Records of crustose lichens from Tasmanian rainforest. Mycotaxon 41: 271-286. Kantvilas, G. and Jarman, S. J. (1991). Lichens and bryophytes of the Tasmanian world heritage area 1 - Mt. Sprent. In Banks, M. R., Smith, S. J., Orchard, A. E. and Kantvilas, G„ eds. Aspects of Tasmanian Botany - A Tribute to Winifred Curtis. Royal Society of Tasmania, Hobart, pp. 149-162. Kantvilas, G. and Jarman, S. J. (1993). The cryptogamic flora of an isolated rainforest fragment in Tasmania. Botanical Journal of the Linnean Society 111: 21 1-228. Louwhoff, S. (1992). Lichens of Mt. Donna Buang ( Preliminary Report). Unpublished Master of Applied Science. Biological and Chemical Sciences, Deakin University. Melbourne, Australia. Louwhoff, S. (1995). The Lichen Flora of Mt. Donna Buang Scenic Reserve, Victoria. Unpublished Master of Applied Science. Biological and Chemical Sciences, Deakin University. Melbourne, Australia. Peel, B. (1999). Rainforests and cool temperate mixed forests of Victoria. Department of Natural Resources and Environment, Melbourne. Stevens, G. N. (1999). A revision of the lichen family Usneaceae in Australia. Bibliotheca Lichenologica 72: 1-128. Wedin, M. (1995). The lichen family Sphaerophoraceae (Caliciales, Ascomycotina) in temperate areas of the southern hemisphere. Symbolae Botanicae Upsaliensis 31: 1-102. Appendix 1. Lichen species list for the Otway Ranges, Victoria. Presence in two forest types - Nothofagus- dominated rainforests and Acac/a-dominated forests. Species List Nothofagus rainforest Acacia forest Arthothelium interveniens (Nyl.) Zahlbr. + Bacidia buchananii (Stirt.) Hellb. + + Bacidia laurocerasi (Delise ex Duby) Zahlbr. N + Buellia sp. 1 + Buellia sp. 2 + + Bunodophoron australe (Laurer) A. Massal. + Bunodophoron insigne (Laurer) Wedin + + Bunodophoron murrayi (Ohlsson) Wedin N + + Bunodophoron patagonicum (C.W. Dodge) Wedin + Bunodophoron ramuliferum (I.M. Lamb) Wedin + Bunodophoron sp. + Ccdoplaca sp. + Chrysothrix candelaris (L.) J.R. Laundon + Cladia aggregata (Sw.) Nyl. + + 28 S. Ford, M. Gibson and G. Duke Cladonia ochrochlora Florke + Cladonia ramulosa (With.) JR. Laundon + Cladonia rigida (Flook f. & Taylor) Hampe + Cladonia subradiata (Vain.) Sandst. + Cliostomum griffithii (Sm.) Coppins N + + Coccotrema cucurbitula (Mont.) Mull. Arg. N + Coenogonium implexum Nyl. + + Collema fasciculare (?var. microcarpum) (Mull. Arg.) Degel. + Collema cf. laeve Hook f. & Taylor + Collema subconveniens Nyl. + + Degelia gayana (Mont.) Arv. & D.J. Galloway + Dimerella lutea (Dicks.) Trevis + + Graphis insidiosa (C. Knight & Mitt.) Hook. f. N + Grapliis tenella Ach. + + Heterodermia hypocaesia (Yasuda) Awasthi + Hypogymnia enteromorphoides Elix + Hypogymnia mundata (Nyl.) Oxner ex Rass. + + Hypogymnia subphysodes (Kremp.) Filson + Hypogymnia (cf. tasmanical) + Hypotrachyna sinuosa (Sm.) Hale + Lecanactis abietina (Ach.) Korb. + Lecidea immarginata R. Br. ex Cromb. N + Lefidium tenerum (Laurer) Wedin + Lepraria lobificans Nyl. N + + Lepraria sp. + + Leptogium victorianum F. Wilson + + Maronea constans (Nyl.) Hepp. + + Megalaria grossa (Pers. ex . Nyl.) Hafellner + + Menegazzia confusa P. James + + Menegazzia eperforata P.James & D.J. Galloway N + Menegazzia myriotrema (Mull. Arg.) R. Sant. + + Menegazzia norstictica P. James + Menegazzia platytrema (Mull. Arg.) R. Sant. + + Menegazzia sp. 1 + Menegazzia subpertusa P. James & D.J. Galloway + Metus conglomeratus (F. Wilson) D.J. Galloway & P. James + Micarea prasina Fr. N + + Miearea spp. agg. + Nephroma australe A. Rich + Normandina pulchella (Borrer) Nyl. + + Ochrolechia sp. + Parmelia cunninghamii Cromb. + Parmelia protosulcata Hale N + + Parmelia tenuirima Hook. f. & Taylor + + Parmeliella nigrocincta (Mont.) Mull. Arg. + Parmelina endoleuca (Taylor) Hale + Parmelina labrosa (Zahlbr.) Elix & J.Johnst. + + Parmelina quercina (Willd.) Hale + + Parmelinopsis afrorevoluta (Krog & Swinscow) Elix & Hale + + Parmelinopsis subfatiscens (Kurok.) Elix & Hale + + Parmotrema chinense (Osbeck) Hale & Ahti + + Parmotrema robustum (Degel.) Hale N + Peltigera dolichorrhiza (Nyl.) Nyl. + + Lichens of Nothofagus and Acacia 29 Pertusaria gibberosa Mull. Arg. + + Pertusaria novaezelandiae Szatala + + Pertusaria truncata Kremp. + Pertusaria sp. 1 + + Pertusaria sp. 2 + Phaeogr aphis exaltata (Mont. & v.d. Bosch) Mull. Arg. 1 + Phlyctis subuncinata Stirton N + + Physcia adscendens (Fr.) H. Olivier + Pseudocyphellaria billardierei (Delise) Rasanen + Pseudocyphellaria colensoi (C. Bab. ex Hook, f.) Vain. + Pseudocyphellaria dissimilis (Nyl.) D.J. Galloway & P. James + + Pseudocyphellaria glabra (Hook. f. & Taylor) C.W. Dodge + + Pseudocyphellaria multijida (Nyl.) D.J. Galloway & P. James Pseudocyphellaria rubella (Hook. f. & Taylor) + + D.J. Galloway & P. James + Pseudocyphellaria sp. A + Psoroma asperellum Nyl. + Psoroma durietzii P. James & Henssen + Psoroma leprolomum (Nyl.) Rasanen + + Psoroma microphyllizans (Nyl.) D.J. Galloway + + Psoroma sp. 1 + Punctelia borreri (Sm.) Krog + Pyrenula nitida (Weig.) Ach. + Pyrenula sp. 1 + + Pyrrhospora laeta (Stirt.) Hafellner + + Ramalina inflata Hook. f. & Taylor + Ramboldia brunneocarpa Kantvilas & Elix + Rimelia reticulata (Taylor) Hale & A.Fletcher + Sarrameana albidoplumbea (Hook.f. & Taylor) Farkas N + + Sarrameana tasmanica Vezda & Kantvilas + + Tephromela citra (Huds.) Hafellner + + Thelotrema decorticans Mull. Arg. + Thelotrema lepadinum (Ach.) Ach. + + Thelotrema subdenticulatum (Zahlbr.) G. Salisb. + + Thelotrema sp. 1 + Thelotrema sp. 2 + + Usnea oncodeoides G.N. Stevens + + Usnea oncodes Stirton N + + Usnea ca. punctulata G.N. Stevens + Usnea pycnoclada Vainio N + Usnea rubicunda (Stirton) var. spilota (Stirton) G.N. Stevens + + Usnea scabrida subsp. tayloriana G.N. Stevens + Usnea subeciliata (Motyka) Swinscow & Krog + + Usnea xanthopoga Nyl. N + Total Species: 110 ( N 17) 93 67 N = Newly Recorded for Victoria 50 species common to both forest types ' Muelleria 14:31 (2000) A note on Lepidium strictum (S. Watson) Rattan (Brassicaceae) in Victoria, Australia. Neville Scarlett Department of Botany, La Trobe University, Bundoora, 3083. Victoria, Australia Abstract Lepidium strictum (S. Watson) Rattan is shown to be the correct name for the species to which the name L. pubescens Desv. has been misapplied by Australian authors, following Thellung (1906). Introduction In his paper “The South American species of Lepidium" (1945), Hitchcock showed that L. strictum (S. Watson) Rattan differs diagnostically from L. pubescens Desv. in having relatively large nectary glands c. 0.5 mm long and prominently reticulate-veined silicu- lae 2.5-3 mm long, and thus Thellung (1906) erred in reducing the former species to syn- onymy with L. pubescens. Australian records of L. pubescens are referable to L. strictum. All relevant collections in MEL match Hitchcock’s description and figure of the latter species (Hitchcock 1945), as do the descriptions and figures of L. pubescens in Hewson ( 1982) and Entwisle ( 1996). Taxonomy Lepidium strictum (S. Watson) Rattan, Anal. Key West Coast bot. 25 2 nd edn (1888). Type : near Placerville [California, United States], Rattan (holotype GH n.v., fide Hitchcock, Madrono 3: 272 (1936)). Lepidium oxycarpum Torr. & Gray var. (?) strictum S. Watson, Bot. California 1: 46 ( 1876). Lepidium pubescens auct. non Desv., J. Bot. (Desvaux) 3: 165, 180 (1814): Thellung, Mitt. Bot. Mus. Univ. Zurich 28: 247 (1906); Hitchcock, Madrono 3; 272 (1936); Willis, Hand. PL Victoria 2: 175 (1973); Hewson, Brunonia 4: 276 (1982); Entwisle, FI. Victoria 3: 420 (1996). Lepidium reticulatum Howell. FI. N.W. Amer. i: 64 (1897), non Thellung, Mitt. Bot. Mus. Univ. Zurich 28: 253 ( 1906) = Lepidium oblongum Small. Thellung (1906) lists a number of other misapplied or synonymous names but only one of these is relevant in an Australian context; the illustration cited in Willis (1973) for L. pubescens: Bettfreund, Flor. Argent. 2: t. 78 (1900), is referred by Thellung to L. bonar- iense L. Discussion The only collection of L. pubescens seen by Thellung was Desvaux’s type specimen from Peru. Regarding L. pubescens , Hitchcock (1945) states “It seems remarkable that the identity of this species has remained uncertain so long. The large, pubescent-margined silicles and sharply toothed leaves are to be matched in no other American species.” He adds: "The material which Thellung and I called L. pubescens (= L. strictum ) differs among other ways in having very prominently reticulate and much smaller fruits, longer glands, persistent sepals and different leaves.” In Australia, L. strictum is a rare weed of urban areas, confined to Victoria (Entwisle 1996). Although the type is from California, Hitchcock (1945) states “The species 32 N. Scarlett appears to be a rather recent introduction to North America as it is found chiefly near habitations.’' Rollins (1993) describes it as a sporadic species, mostly in towns and cities, occurring in Utah, California and Oregon. It is thus probable that L. strictum is native only to Chile, where it occurs between latitudes 27° and 34°S, mainly around Valparaiso and Santiago, with one northern collection from the coastal town of Caldera in Atacama province (Hitchcock 1945). It is interesting to speculate as to whether the species was introduced directly into Australia from Chile or secondarily via California. The transito- ry appearance at Tenterfield, N.S.W. of L. oblongum Small (California to Arkansas and south to Mexico and Guatemala) appears to support the latter possibility (see Hewson, 1982). The three other American species recorded for Australia shed no light on the prob- lem as L. bonciriense L. (SE South America), L. densiflorum Schrad. (North America) and L. virginicum L. (North and South America) are widespread and successful weeds, now established outside their original ranges in the Americas and adventive in Europe, Australia and New Zealand (Garnock- Jones 1988; Hernandez Bermejo & Clemente 1993; Hewson 1982; Hitchcock 1936; Rollins 1993;'Ryves 1977). Acknowledgements Thanks are due to Dr R.F. Parsons, Mr N.G. Walsh and the anonymous referees who read and criticised the manuscript. References Entwisle, T.J. (1996). Brassicaceae — Lepidium. In ‘Flora of Victoria, Volume 3 . (Eds N.G. Walsh and T.J. Entwisle.) pp. 415-23. (Inkata Press: Melbourne.) Garnock-Jones, P.J. (1988). Brassicaceae (Cruciferae - Lepidium L.). In ‘Flora of New Zealand. Volume 4’. (Eds C.J. Webb, W.R. Sykes and P.J. Garnock-Jones.) pp. 425-436. (Botany Division. D.S.I.R.: Christchurch.) Hernandez Bermejo, J.E., and Clemente. M. (1993). Lepidium (L.) R. Br. In ‘Flora Iberica. Volume 4’. (Eds S. Castroviejo, C. Aedo, C. Gomez Campo, M. Lainz, P. Montserrat, R. Morales, F. Munoz Garmendia, G. Nieto Feliner, E. Rico. S. Talavera, and L. Villar.) pp. 311—327. (Real Jardin Botanico: Madrid.) Hewson, H.J. (1982). The genus Lepidium L. (Brassicaceae) in Australia. Brunonia 4, 217-308. Hitchcock, C.L. ( 1936). The genus Lepidium in the United States. Madrono 3, 265-320. Hitchcock, C.L. ( 1945). The South American species of Lepidium. Lilloa 1 1 . 75-134. Rollins, R.C. (1993). ‘The Cruciferae of Continental North America.’ (Stanford University Press: Stanford.) Ryves, T.B. (1977). Notes on wool-alien species of Lepidium in the British Isles. Watsonia 11 . 367-372. Thellung, A. (1906). Die Gattung Lepidium (L.) R. Br. Eine monographische Studie. Mitteilungen aus dem botanischen Museum der Universitdt Zurich 28, 1-340. Willis, J.H. (1973). ‘A Handbook to Plants in Victoria, Volume 2’. (Melbourne University Press: Carlton.) Muelleria 14:33 (2000) The Stomata of Bluegums ( Eucalyptus spp.) Denis J. Carr Visiting Fellow, RSBS, ANU, Canberra 2001, Australia Abstract Measurements of the sizes of the stomata of juvenile and adult leaves of Tasmanian blue-gum (• Eucalyptus globulus Labill.) made over a long period and in diverse countries are collated. It is shown that in E. globulus , the sizes are independent of locality and whether the trees are natural or cultivated, that is almost certainly they are genetically determined. The stomata of adult leaves of different species of blue gums are larger, sometimes as much as twice as large, as those of the juve- nile leaves. The upper surface of the juvenile leaves of three of the blue gum species bears few or no stomata and the internal structure of the (bifacial) leaf reflects this. The (unifacial) adult leaves always have stomata on both surfaces. The large adult/juvenile stomatal size ratio appears to be unique to bluegums since it does not occur in a range of other species investigated. Introduction Alhough Vesque (1882) is usually credited with drawing attention to the usefulness of studies of leaf anatomy to systematics Mueller (1883) was already aware of and using features of leaf anatomy in his research. He wrote: “By the aid of the microscope we may yet hope to be able to obtain characteristics of diagnostic value from the anatomy of leaves sufficiently positive to recognise ordinal and even perhaps generic groups...” “I was enabled, for instance, to demonstrate the existence of Epacridae in New Guinea from the microscopic comparison of the leaf epidermis of a species, brought from thence with- out flowers or fruits, with the very peculiar cuticle of many Epacridae easily recognized microscopically. Mueller was the first to investigate the microscopic features of the leaf epidermis of eucalypts. In his Eucalyptographia (1879-1884) he illustrated the stomata of 39 species. He not only discussed differences in sizes of the stomata but also the dif- ferences betweeen species in stomatal frequency (number per unit area of leaf surface). His younger contemporary, J.H. Maiden (1909-1928) referred (Vol. 1, p. 9) to Mueller’s tentative classification based on the differences in distribution of stomata on the two leaf surfaces, but dismissed it, without citing supporting evidence, as “not reliable.” Beyond this Maiden makes no further reference at all to stomata. Nevertheless, he does remark {ibid) “The anatomical characters of the leaves of Eucalyptus offer, however, much room for research.” In carrying out the statistically elaborate work for his 1973 doctoral thesis on Geographical variation in Eucalyptus globulus Labill. Kirkpatrick collected materials of bluegums, as the group of species to which E. globulus belongs is known, throughout south-eastern Australia. His two papers on the topic (Kirkpatrick 1974, 1975) divide the group into four sub- species of which E. globulus proper is almost entirely Tasmanian with some outliers in southern Victoria, and E. maidenii F.Muell. is almost entirely restricted to the seaward slopes of south-western New South Wales. I differ from him in what follows in treating the entities he regards as subspecies as species in their own right (cf. also Chippendale 1988). The literature cited by Kirkpatrick is almost exhaustive of that on the bluegums: but there still remain some aspects of the bluegums, including some older literature, nei- ther mentioned nor discussed. One of these aspects goes back to the work of Mueller. It concerns the stomata of E. globulus. To Mueller, that name covered all the species (or subspecies) of bluegum now recognised (except E. maidenii) so it is difficult to be sure that his work on stomata really concerned only the Tasmanian species. Nevertheless, 34 D.J. Carr the illustration to his description of the species (Mueller 1879-84) is, as far as one can tell, true to the concept of E. globulus put forward by Labillardiere in 1800 and 1806. Nevetheless, whereas E. globulus (in Labillardiere’s sense) has inflorescences each con- sisting of a single flower, an unnumbered picture of a three-flowered inflorescence is included in the Plate (Decade 6) without any comment in the legend. 1 think that Mueller would have been cautious enough to realise the possility of his specimen not being E globulus sensu stricto. In constructing Table 1 I was unable to trace two references to E. globulus in early pharmacological journals, referred to by Maiden (1909-1920, Vol 1, p 7) which might contain measurements of stomata. Lacking funds to collect for my work, materials from the "wild,” I requested, but was unable to obtain, access to Kirkpatrick’s specimens but they had been returned to the collector. However there are plantings of most species of Table 1. Stomatal size* and frequency in E. globulus Author Juvenile Adult Size pm Frequency mm 2 Size pm Frequency mm 2 Mueller upper 70 upper 60 lower 132 lower 51 87 Briosi upper upper lower 40-60 162 lower 80-100 59 Johnson upper upper 53 86 lower 29 83.8 lower 54 86 Carr and Carr ( 1 ) lower 27 142 lower 53 Carr and upper upper (na) Carr (2) lower 26.5 191 lower (na) Carr and Carr (3) lower 27.6 127 lower 53.3 Carr and Can - (4) lower 35.4 lower 58.3 Maiden & Mansfield (1968) lower 31 Ridge. upper upper 56 129 R.W. (1980) lower lower 57 63.6 *The size measurements in all Tables are of stomatal length (micrometres) (means of 30 determinations) . 35 Stomates of bluegums bluegums in the Canberra area (Pryor 1951) and 1 turned to these for materials. We did have some specimens of E. globulus obtained from A.K. Cameron. Subsequent to Mueller’s research on E. globulus a monograph appeared in Italy on the leaf of Eucalyptus globulus which included a number of measurements and drawings of stomata (Briosi 1892). Since then a number of other measurements have been made, pub- lished or not, on the stomata of bluegums including E. globulus, which (in its narrow sense) ( contra Kirkpatrick) is not widely grown on the mainland of Australia but is or was widely grown abroad. However, in the last few years, extensive plantings of E globulus have been made for timber and woodchips in southern and south-western Western Australia and in western Victoria. Table 1 shows three remarkable things. One is the extrordinary consistency of the measurements on a single species, especially of the stomatal lengths over more than one hundred years and in places separated by the diameter of the earth, and irrespective of seed provenance. This attests to inherent that in E. globulus stomatal length is under strict genetical control. The only aberrant data are those of Briosi, which are twice those of all the other observers. The second remarkable thing is that in all but one case (Mueller) there were no stomata recorded on the upper side of the juvenile leaves. The third, is that in all cases the stomata on the lower side of the juvenile leaves were about or just under one half the length of those of the adult. Omitting Briosi’s measurement the mean of the juvenile stomatal lengths is 29.4 pm with a standard deviation (SD) of 3.06; that of the adult stomata 54.6 SD 1.95. Thus the ratio of the two means is 1.86. This is also the case in the measurements of Briosi, which leads me to believe that his measurements suffer from a systematic error, perhaps in the calibration of his instruments. If his magnifica- tions (given as x300) were actually x450 (as were Mueller’s) the adult stomata illustrat- ed in Briosi’s Plate III Fig 2 would be 55 pm, not 80 to 100 as stated. The Carr and Carr measurements were from specimens collected by A.K.Cameron and A. M. Gray in Tasmania. They had no stomata (Carr and Carr 2 & 4) or very few (1 and 3) (less than 1 mm 2 ) on the upper surface of the juvenile leaves; another specimen from Flinders Island, Bass Strait had none. The juvenile leaves of Mrs Carr's undated and unnumbered specimen from Squeaky Bay, Victoria, about as close a locality as one can get to Tasmania on the mainland of Australia, also lacked upper surface stomata. On the other hand, a specimen collected (as L E globulus ') by Cameron in Toorak, Melbourne, Victoria, had stomata on both surfaces of the juvenile leaves, with a frequency of 31 (upper) and 42 (lower) mm 2 . Johnston (1926) recorded “a few” stomata along the midrib on the upper surface of the juvenile leaves, evidently too few to record their frequency (c.f Carr and Carr 1 & 3). The material she used was a plant grown in a greenhouse at Manchester. The material of Meidner and Mansfield (1968) was also a seedling grown in an English greenhouse. Mueller, whose concept of the species E. globulus was extreme- ly wide (as already mentioned), found stomata on the upper surface of the juvenile leaves, reported by none of the other authors (but c.f. the Toorak specimen of Cameron). One cannot therefore be certain that Mueller’s data refer to E. globulus in the narrower, mod- ern sense. The size reported in Table 1 is from a drawing, together with its magnification, of a single stoma. The stomata of the juvenile leaves of E. globulus (in the narrow sense) are therefore essentially restricted to the lower surface. This has implications for the internal structure of the leaf, which differs in this from the adult leaf which, having stomata on both sides, has also a layer or layers of palisade parenchyma below each leaf surface (Figs. 1-4). This, of course was already known to Briosi (1892: 106-107) but has escaped the notice of subsequent writers (e.g. Maiden. 1909-1920 (Vol 1: 7; 6: 287); Penfold & Willis 1961), who cite his publication. Nevertheless Maiden (1:7) described Briosi’s mono- graph as “a masterly paper.” Knowledge thus won early has subsequently been forgotten. We must ask if these findings extend to the other bluegunt species. 36 D.J. Carr Figures 1-4. All illustrations of leaves of Eucalyptus globulus from the publication by Briosi.(1892). 1 Stomata of a juvenile leaf. 2 Stomata of an adult leaf. 3 Section of a juvenile leaf with stomata only on the lower surface. 4 Section of a mature adult leaf. Materials and Methods Ausing cuticle preparations were obtained by chemical maceration in a heated mixture of glacial acetic acid and 100 vol. hydrogen peroxide (1:5), stained and mounted in glyc- erin-jelly. The measurements were made using a calibrated eyepiece micrometer and fre- quency measurements with an eyepiece graticule. Observations The observations (Tables 2-6) show that the juvenile leaves of E st.johnii (R.T.Baker) (a species subsumed by Kirkpatrick into his E. pseudoglobulus) and E bicostata Maiden also lack stomata on the upper surface and therefore have an internal structure like that of the juvenile leaves of E. globulus. But the juvenile leaves of the other species always have stom- ata on both surfaces, like the adult leaves of all. Thus, it matters not that such juvenile leaves are (as Briosi called them) ‘"horizontal," the two surfaces illuminated unequally, and there- fore (according to him) theoretically predisposed to be bifacial and have stomata confined to the lower surface whereas the vertically hanging adult leaves, as expected, are unifacial. Stomates of bluegums 37 These facts stand, of course, in contrast to one of Kirkpatrick’s statements (1974). “No dis- continuities are evident in any of the 30 adult and juvenile characters studied over the total iange ol blue gum. The results reported below (Table 2) are quite preliminary and they must be taken as tentative and requiring confirmation and expansion. To go further with the comparisons would require a considerable research undertaking. Table 2. E. bicostata , size of stomata (pm). Juvenile Adult Anzac Parade 33.5 55.8 Yarralumla 30.4 56.8 Ratio of adult/juvenile 1.78. Of bluegum trees growing in Canberra E. bicostata is by far the commonest. In addition to cultivated trees of E.st.johnii 1 was also fortunate enough to have access to type material of that species collected by the eponymous E. St John from the Lerderder^ Gorge in Victoria. Table 3. E. st johnii. (= E. pseudoglobulus of Kirkpatrick) (lower surfaces) Juvenile Adult Lerderderg Gorge size 32.6pm 57.4 pm (E. St.John) freq. 266 224.5 ANU campus site 36.58 46.17 freq. 187 — near Mosque size 39.68 54.7 The ratio of the (size) means) adult/juvenile is 1.47 As juvenile leaves of E. maidenii were not accessible I was forced to examine a number of herbarium specimens. One of these, Beauglehole 33943 , had originally been identified as E.st.johnii, then was re-identified by Brooker as E. pseudoglobulus and subsequently used under that name as the voucher for a drawing of its three-flowered inflorescence in the Flora of Australia (Chippendale 1988). Its identification seems to me to be wrong and I believe that it is an unusual specimen of E. maidenii. For instance the juvenile leaves have stomata on both surfaces like E. maidenii but unlike E pseudoglobulus. The following are measurements of stomatal length in micrometres from the lower surfaces of the leaves. Table 4. E. maidenii , size of stomata (lower sufaces) Juvenile Adult de Beuzeville 588 Eden NSW 29.05 58.7 de Beuzeville 587 Eden NSW 32.7 60.45 Beaglehole 33682 Genoa, Victoria 32.83 51.4 Beaglehole 33943 Tara Range, Vic. 28.163 62.46 The overall ratio of means of adult/juvenile sizes is 1 .9. In some specimens (e.g. Beauglehole 33943) it is greater than 2. 38 D.J. Carr Thus the ratios of the means of adult to juvenile sizes are as follows: maidenii 1 .9, glolndus 1.86, bicostata 1.67, st.johnii (= peudoglobulus) 1.47. Clearly the adult stoma- ta of E. maidenii are by far the largest of any of the blue gums. We are left with a puzzle: what are the so-called E. globulus specimens in and around Melbourne which superficially look like that species but have juvenile leaves with large numbers of stomata on the upper surface? They cannot be E. pseudoglobulus which lacks such stomata. The late J.H Willis also had difficulty with such specimens. “ Efucalyptus ) st.johnii sometimes overlaps with occurrences of typical E. globulus (e.g. on Wilson’s Promontory, Phillip Island and the Otways), where puzzling intermediate populations occur” (1973, 2: 419). Such so-called "intermediates” would include the specimen exam- ined by Mueller (Table 1 ) and the Cameron specimen from Toorak. Evidently some trees identical with the real, Tasmanian, E. globulus exist in southern Victoria (like the one from Squeaky Bay, Wilsons Promontory), as one might expect since Tasmania has been isolated from the mainland only since the end of the last ice-age. But it is not possible immediately to classify the other specimens from the vicinity of Melbourne. During one of my visits to California 1 collected a specimen of a locally grown vari- ety (E. globulus var .compacta Maiden) near Santa Barbara. It is listed in Chippendale (1988) as a hybrid. Table 5. E. globulus var. compacta. (size of stomata pm) Juvenile Adult upper 39.76 upper 38.29 lower 33.18 lower 41.25 For comparison: Cameron Toorak (as “E. globulus") upper 26. 1 8 upper 39.0 lower 33,32 lower 40 Ratios adult/juvenile (lower surfaces) compacta 1.2; Cameron Toorak 1.2) The var. compacta ratio (Table 5) might suggest some relationship with E pseudoglobulus , some specimens of which (e.g. those on the ANU campus) have a ratio of 1 .2, rather than with E globulus. However the fact that the juvenile leaves have stom- ata on both surfaces, unlike E. pseudoglobulus. does not support that possibility. The data added for the Toorak specimen of Cameron labelled E. globulus present a striking sim- ilarity except for the lengths of the stomata on the upper surfaces of the juveniles. Further work is suggested on the blue gums of Santa Barbara and those around Melbourne. Unfortunately the Santa Barbara tree appeared to be sterile, at least I saw on it no fruits or flowers. Discussion The ability to illuminate relationships within the bluegums by examining the leaf anato- my suggests that it. rather than characters such as the number of buds in the inflorescence, might be the better determinant. For instance, it is quite clear that stomatal sizes and dis- tribution easily distinguish between E. pseudoglobulus and E. maidenii (as in the case of Beauglehole 33943). A corollary of the measurements reported above, which include some measurements on juvenile foliage not from seedlings but from what are called “reversion shoots" on adult trees, is that the characters of the juvenile leaves are repeated when, in the production of these reversion shoots, the adult tree produces another set of juvenile leaves. These shoots usually begin with a few' leaves, identical in shape and glaucousness to those of the sapling Stomates of bluegums 39 and to these characters we may now add those of the stomata. The sizes and locations of the stomata are identical with those of the seedling and sapling. Then as the reversion shoot develops, leaves identical with the adult leaves in all characters, including those of the stom- ata, are produced. This is quite a remarkable performance since it implies that a whole suite of juvenile characters, presumably all genetically determined and including cell size (since stomatal size appears so), is switched on and then in the course of the early development of the succeeding leaf primordia they are switched off and all replaced simultaneously by a different suite of adult characters, also genetically determined. Of course this performance merely repeats what occurs during normal “phase change” (juvenile to adult) during the normal growth of the tree, but in its recurrence and spontaneity it is very striking. During the slow development of the bark-enclosed residual meristem (which gives rise to the rever- sion shoot), in turn derived from an accessory bud on the growing shoot, there must be some reprogramming of the genetic material of the cells of the meristem to an embryonic state like that of the seedling; this state then changes rapidly during reversion shoot growth back to the adult condition (see Carr 1984). In a previous publication (Carr & Carr 1991 ), we stated (without citing data) that “the size of the guard cells (of juvenile leaves of eucalypts) is smaller than those of the adult leaves. This I re-investigated in 7 species growing in my own garden or nearby in Canberra. The results are given below in Table 6. Table 6. Comparisons of measurements of sizes of juvenile and adult stomata in some species of Eucalyptus. Species ratio adult/juvenile significance quadrangulata Deane & Maiden 1.220 p< 0.05 cinerea F Muell. ex Benth, 1.019 ns bridgesiana R.T.Baker 1.136 ns melliodora Cunn. ex Schauer. 1.205 ns leucoxylon F.Muell. 1.1072 ns sideroxylon Cunn. ex Woods 1.1065 ns perriniana F.Muell. ex Rodway 1.1535 p< 0.001 It will be clear that the results are equivocal. In all cases the measurements showed that the adult stomata were a little larger but the data when subjected to statistical test the differences were significant only in two cases (ns = not significant). The great differences in the bluegums appear therefore to be unique to that group of species. The number of stomata measured in each case was 30 for each surface and the tests were by t test. Perhaps if a larger sample of the stomata were measured in each case the differences observed might achieve significance. However, the differences are too small to warrant a fuller investigation. Acknowledgements This work was completed during the tenure of a Visiting Fellowship in RSBS. I am grate- ful to the outgoing Director of the School, Prof. Barry Osmond for that appointment and for help in completion of the manuscript. References Briosi, G. (1892). Intorno alia anatomia delle foglie dc\V Eucalyptus globulus Labill. Atti 1st. Bot. Univ. Pavia B, series 2, 7-151. Carr, D. .1. (1984). Positional information and plant morphology. Chapter 12. In ‘Positional Controls in Plant Development'. (P.W. Barlow & D.J. Carr, eds.) pp. 350-374. (Cambridge University Press. Cambridge) I 40 D.J. Carr Carr. D.J. and Carr, S.G.M. (1991). Development of the stomatal complexes during ontogeny in Eucalyptus and Angophora (Myrtaceae) Australian Journal of Botany 39 . 43-58. Chippendale, G.M. (1988). Flora of Australia. Vol 19 Myrtaceae - Eucalyptus , Angophora. (Australian Government Publishing Service. Canberra ) Johnson. E.D. (1926). A comparison of the juvenile and adult leaves of Eucalyptus globulus. New Phytologist 25 , 202-263 . Kirkpatrick, J.B. (1974). The numerical intraspecific taxonomy of Eucalyptus globulus Labill. (Myrtaceae). Botanical Journal of the Linnean Society 69 . 89-104 Kirkpatrick, J.B. (1975). Geographical variation in Eucalyptus globulus. Forestry and Timber Bureau. Buletin. No 4 , (Australian Government Publishing Service, Canberra). Labillardiere, J.-J. (1800). Relation du voyage a la recherche de La Perouse (H.J. Jansen: Paris). Labillardiere, J.-J. (1806). Novae Hollandiae Plantarum Specimen. 2 vols. (D. Huzard: Paris) Maiden. J.H. (1909-1928). A critical revision of the genus Eucalyptus. Vols l-8.(Government Printer: Sydney) Meidner, H and Mansfield. T. A. (1968). Physiology of stomata. (McGraw-Hill: London ) Mueller. F.von. (1879-1884) Eucalyptographia, Decades I-X (John Forbes, Government Printer: London). Mueller, F von (1883). Observations on new vegetable fossils of the auriferous drifts. 2nd Decade. (Government Printer: Melbourne & London) Penfold. A.R. and Willis, J.L. (1961). The Eucalypts.(Leonard Hill (Books) Ltd: London) Pryor, L.D. (1951). Trees in Canberra. (Dept, of the Interior: Canberra). Ridge, R. W. (1980). A comparison of the anatomy and ecophysiology of some species of Eucalyptus. Honours. B Sc. Thesis, University of Western Australia. Vesque, M.J. (1882). L’espece vegetale consideree au point de vue de l’anatomie comparee. Annales des sciences naturelles; botanique. 13 , 5-50. Willis, J.H. (1973). A Handbook to Plants in Victoria. 2 Vols. (Melbourne University Press: Melbourne). Muelleria 14:41 (2000) Identified types in the Wilhelm Hillebrand collections in the herbarium of the Royal Botanic Gardens, Melbourne. James W. Grimes Plant Sciences and Biodiversity, Royal Botanic Gardens, Melbourne, Birdwood Ave, South Yarra, VIC 3141 Australia email: jgrimes@rbgmelb.org.au A list of clearly identified and probable types from the Hawaiian collections of Wilhelm Hillebrand is presented. Of the well over 200 taxa published by Hillebrand, there are 102 verified types, and 4 possible types deposited at MEL. A short discussion of how these collections came to be deposit- ed at MEL is provided. Introduction Wilhelm Hillebrand, Prussian physician, lived for 20 years in Hawai’i and amassed a per- sonal herbarium of 12,000-15,000 sheets (Lammers 1994). This herbarium was the basis of his Flora of the Hawaiian Islands (1888). published by his son two years after his death. When Hillebrand died his herbarium was bequeathed to the Berlin-Dahlem Museum, and his collections were among those destroyed during World War II. Duplicates of his collections are fairly widely distributed (Stafleu & Cowan 1979), and it is reported in Taxonomic Literature // that a set of duplicates is present at MEL. Nonetheless, no information about these collections has yet been published; consequent- ly, the collections at MEL are often not included in discussions of Hillebrand's duplicates (■ cf. Sayre 1975). Ferdinand Mueller, first Government Botanist of Victoria, managed to acquire an impressively large set of duplicate collections from around the world. Unfortunately it Abstract Ex herb. Vi. Hillebrand. Figure 1 . An example of the kind of label found on all of Hillebrand's types lodged at MEL. 42 J.W. Grimes Flora Hawaiensis. Coll. Dr. W. Hillebrand, IK Hab. Figure 2. An example of a second kind of label found on Hillebrand’s types of ferns. cannot be known for certain when and how most of these specimens came to be at MEL as F. Mueller’s correspondence was destroyed during a move of the MEL herbarium in 1934-35 (Pescott 1982). In the case of the Hillebrand collections there is contradictory information. For instance, all collections at MEL have labels stating ‘ Ex Museo botanico Berolinensi .’ (Fig. 1 ), leading one to believe the plants were distributed from B. and, given the consistency in the labels, that they came in one shipment. However, the pteridophytes have not only this kind of label, but also a second (Fig. 2). It is possible that the ferns came in a separate shipment. Mueller lists Hillebrand in his Annual Report of the Government Botanist to Parliament, 1864-65. (MEL library archives) as having donated dried plant specimens: but Hillebrand’s collections were bequeathed to B after his death, and. in 1865 he was still living in Hawai’i. These collections therefore are undoubtedly not part of the Reliquae Hillebrandianae. It is reported (Corn 1983) that Hillebrand and Mueller became friends when Hillebrand went plant collecting with Mueller during a visit to Australia, and Corn states (1983) that the two are known to have exchanged plant material. Though the two men collected together, in only one instance have I found that a Hillebrand collection from Australia is cited as a syntype for a species described by Mueller (i.e.. Erysimum blennodes)', however, no syntype is deposited at MEL. Furthermore, though other species are named by Mueller in honour ot Hillebrand (e.g, Eriostemon hillebrandii, Veronica hillebrandii), no duplicates ot any collections made by Hillebrand in Australia have been found at MEL. How large the set of his collections from Australia, and if he kept them in his personal herbarium, is not known. It is likely they were destroyed at B. Corn (1983) first brought attention to the duplicate set of Hillebrand's collections deposited at the Royal Botanic Gardens, Melbourne. She noted that she photographed 569 sheets of flowering plants and ferns, and distributed copies of these photos (Corn, 1983. unpublished ms.). Apparently no collections of lower plants or fungi are represent- ed. Most of the foreign material that Mueller acquired for MEL remains unmounted. And though Corn did search through the Polynesian set, she did miss some specimens. It is therefore likely that there is somewhat more than the 569 sheets reported by Corn deposited in MEL. As Hillebrand’s own herbarium bequeathed to Berlin was destroyed, and duplicates of his collections are widely distributed, typification of his names remains a problem (Lammers 1994). It is unlikely that any herbarium has a complete set of his duplicates. As was typical for 19 th century collectors, types were not clearly designated by Hillebrand types 43 Hillebrand. It is in the interest of informing researchers who may have to designate lecto- or neotypes, that a list of clearly identified or probable duplicates of the syntypes present in MEL is herein presented. If lectotypification or neotypification has already been done, that is noted. It should not be assumed that this list is complete, as the unmounted mate- rial in the foreign herbarium at MEL still surpasses 250,000 sheets. Following the basionyms, synonyms are provided and are taken from Wagner et al. (1990), unless otherwise stated. Following in single quotes is any information on the label (other than the plant name) presented verbatim. When a type has been annotated by another researcher that it also noted. Ferns are presented first, then flowering plants in alphabetical order by family. As Hillebrand did not assign collection numbers, the MEL herbarium number of each sheet is given. Of the 560-some collections by Hillebrand reported at MEL, verified types are listed below. Though an insignificant number of collections in comparison to the estimated size of Hillebrand's herbarium, the number of duplicate types is substantial. List of verified types Pteridaceae Aspidium glabrum Mett. var. soripes W.F. Hillebr., FI. Haw. Is., 577. "Molokai, Sandw. Islds. Juli 1870. Leg. W.H.’ (MEL 587939). Asplenium insititum Brack, var grandipinna W.F. Hillebr., FI. Haw. Is., 599. ‘Kaala; Oahu. 3000 ft. No. 43 e . coll. D. Baldwin.’ (MEL 1520643). Asplenium mannii W.F. Hillebr. var dareoides W.F. Hillebr., FI. Haw. Is., 595. ‘Makaleha, Sandw. Is. coll. W.H.’ (MEL 587940). Asplenium marginale W.F. Hillebr., FI. Haw. Is., 613. "Molokai.’ (MEL 1520441). Asplenium pseudofalcatum W.F. Hillebr., FI. Haw. Is., 599. ‘Oahu, Nuuanu, Molokai (Halawa).’ (MEL 587944) and ‘Oahu, Sandw. Is.’ (MEL 587945). Asplenium sphenotomum W.F. Hillebr., FI. Haw. Is., 599. "Kauai, coll. Knudsen.’ (MEL 587947). Polypodium tamariscinum Kaulf. var montanum W.F. Hillebr., FI. Haw. Is., 556. ‘Kauai, Sandw. Islds.’ (MEL 1520613). Pteris irregularis Kaulf. var. linearis W.F. Hillebr. FI. Haw. Is., 628. ‘as Pteris irregularis Kaulf. b. Sandw. Islds.’ (MEL 1520621). Apocynaceae Vallesia macrocarpa W.F. Hillebr., FI. Haw. Is., 297 = Pteralyxia macrocarpa (W.F. Hillebr.) K. Schum. ‘Oahu. leg. Hbd.’ (MEL 587762). Araliaceae Triplasandra meiandra W.F. Hillebr., FI. Haw. Is., 152 = Tetraplasanda oahuensis (A. Gray) Harms "Wailupe.' (MEL 587937). Asteraceae Argyroxiphium virescens W. F. Hillebr., FI. Haw. Is., 219. ‘OH. Maui. Sandw Is le<* Lydgate.’ (MEL 1520480). Campylotheca macrocarpa W.F. Hillebr. b var. ovatifolia W.F. Hillebr.. FI. Haw. Is., 214 = Bidens macrocarpa (A. Gray) Sherff. As 'Campvlotlieca macrocarpa var. b Oahu leg. Hnd.' (MEL 1520559). Campylotheca molokaiensis W.F. Hillebr., FI. Haw. Is., 212 = Bidens molokaiensis (W.F. Hillber.) Sherff. "Molokai. Sandwich Inseln. leg. Hbd.’ (MEL 1520478). Dubautia laxa Hook & Arn. var. hirsuta W.F. Hillebr., FI. Haw. Is., 223 = Dubautia laxa subsp. hirsuta (W.F. Hillebr.) G. Carr "W. Maui. Sandw. Inns. leg. Hbd’ (MEL 1520387). Hesperomannia arbuscula W.F. Hillebr., FI. Haw. Is., 232. "W. Maui Sandw Ins Le<* Bishon.’ (MF.L /“OOWF 44 J.W. Grimes Lipochaeta hcistata W.F. Hiilebr., FI. Flaw. Is., 208 = Lipochaeta heterophylla A. Gray ‘Lanai, Sandw. Ins. leg. Hbd.’ (MEL 78098). Raillardia molokaiensis W.F. Hiilebr., FI. Haw. Is., 226. ‘Molokai. Sandwich Islands, leg. Hbd.' (MEL 78108). According to Carr (1990, Man. FI. PI. Hawaii, p. 292) this is a hybrid between Dubautia linearis subsp. opposita and D. scabra susbsp. leiophylla. Remya mauiensis W.F.Hillebr., FI. Haw. Is., 194. ‘Maui. Sandw. Ins. leg. Hbd.' (MEL 78112). Tetramolopium filiforme Sherff. var .filiforme, Bot. Gaz. 95: 498 ( 1934). ISOTYPE, fide T. Lowrey, 1986. As ‘ Tetramolopium tenerrimum N. Oahu. leg. Hbd.’ (MEL 78119). Tetramolopium lepidotum (Lessing) Sherff var. luxurious W.F. Hiilebr., FI. Haw Is., 199. LECTOTYPE, desig. by T. Lowrey, 1986 = Tetramolopium lepidotum subsp. lepidotum. As ‘ Tetramolopium chamissonis b luxurious. Oahu. leg. Hbd.’ (MEL 78115). Wilkesia grayaya W.F. Hiilebr., FI. Haw. Is., 220 = Argyroxiphium grayanum (W.F. Hiilebr.) Degener. ‘W. Maui, Sandw. Inf. leg. Hbd.’ (MEL 78120). Brassicaceae Lepidium arbuscula W.F. Hiilebr., FI. Haw. Is., 10. ‘Oahu. Waianae. leg. Hillebrand.’ (MEL 1520481). Campanulaceae Clermontia coerulea W.F. Hiilebr., FI. Haw. Is., p. 243. ISOLECTOTYPE design. Rock, 1919 .fide T. Lammers, 1991. = Clermontia clermontioides (Gaudich.) A. Heller subsp. clermontioides ‘Hawaii, leg. Hillebrand.’ (MEL 587783). Clermontia macrocarpa Gaudich. var. rosea W.F. Willebr., FI. Haw. Is., p. 240. LECTOTYPE. design. T. Lammers 1991: 39 = Clermontia kakeana Meyen. 'Oahu, leg. Hillebrand.' (MEL 587786). Clermontia multiflora W.F. Hiilebr.. FI. Haw. Is., 242. ‘Oahu. leg. Hillebrand.' (MEL 587767). Ver. T. Lammers, 1 1 1/89. Though Lammers annotated this as an ISOLECTOTYPE. it was not cited in his monograph of Clermontia (1991 ). Clermontia multiflora W.F. Hiilebr. var. micrantha W.F. Hiilebr., FI. Haw. Is., 242. ISOLECTOTYPE, design, by T. Lammers 1991: 59. ‘Maui. leg. Hillebrand.’ (MEL 587837). Clermontia pallida W.F. Hiilebr., FI. Haw. Is., 241. ISOLECTOTYPE. design, by T. Lammers, 1991: 50. ‘Molokai. Sandw. Ins. leg. Hillebrand.' (MEL 587839). Cyanea comata W.F. Hiilebr., FI. Haw. Is., 256. SYNTYPE, ann. T. Lammers, III/89 ‘Haleakala, W. Maui. Sandw. Ins. leg. Hillebrand.’ (MEL 587857). Cyanea gibsonii W.F. Hiilebr., FI. Haw. Is.. 263. LECTOTYPE design. T. Lammers, 1988: 503 = C. macrostegia W.F. Hiilebr. subsp. gibsonii (W.F. Hiilebr.) Lammers. ‘Lanai. Sandw. Ins. leg. Hillebrand.’ (MEL 587859) Cyanea holophylla W. F. Hiilebr., FI. Haw. Is., 257. ISOLECTOTYPE, design, by T. Lammers. 1994: 559. = C. scabra W. F. Hiilebr. (MEL 587861). Cyanea macrostegia W.F. Hiilebr., FI. Haw. Is. 263. ISOLECTOTYPE, design, by T. Lammers, 1994: 562. = Cyanea macrostegia W.F. Hiilebr. subsp. macrostegia. ‘West. Maui. Sandw. Ins. leg. Hillebrand.’ (MEL 587856). Cyanea scabra W.F. Hiilebr.. FI. Haw. Is., 256. SYNTYPE, ver. T. Lammers, in annot. 1989. ‘Kaanapali, W. Maui. Sandwich I. leg. Hillebrand.’ (MEL 587930). Cyanea solanacea W.F. Hiilebr., FI. Haw. Is., 259. SYNTYPE, ver. T. Lammers, in annot. 1989. ‘Kalae auf Molokai. Sandw. Ins. leg. Hillebrand.’ (MEL 587931). Cyanea solenocalyx W.F. Hiilebr., FI. Haw. Is., 258. SYNTYPE. ver. T. Lammers, in annot. 1989. ‘Molokai, Sandw. ins. leg. Hillebrand.’ (MEL 587929). Cyanea superba A. Gray var. velutina Rock, Monogr. Stud. Haw. Lobeliod. 157. 1919. ISOTYPE. ver. T. Lammers, in annot. 1989. = Cyanea superba A. Gray subsp. superba. As ‘ Cyanea superba Gray var. b. Oahu. leg. Hillebrand.’ (MEL 587769). Hillebrand types 45 Delissea lacinata W.F. Hillebr., FI. Flaw. Is., 249. SYNTYPE, ver. T. Lammers, in annot. 1989. ‘Oahu. leg. Hillebrand.’ (MEL 587773). Lobelia hypoleuca W.F. Hillebr., FI. Haw. Is., 238. ‘West Maui. Sandw. Ins. leg. Hillebrand.’ (MEL 587836). Rollandia grandifolia Hillebr. FI. Haw. Is., 245. SYNTYPE, ver. T. Lammers, in annot. 1989 = Rolandia crispa Guadich. 'Oahu. leg. Hillebrand.’ (MEL 587834). Cyperaceae Carex montis-eeka W.F. Hillebr., FI. Haw. Is., 486. ‘Mount Eeka, Sandw. Ins. leg. Hillebrand.’ (MEL 587892). Cyperus hillebrandii Boeck., Flora 63: 436 (1880). ISOTYPE, ver. J. H. Kern, 1953 'Haleakala, leg. Hillebrand.’ (MEL 587896). Cyperus polystachys Rottb. var 6 pallidus W.F. Hillebr., FI. Haw. Is., 463 = Pycreus polystachyos (Rottb.) P. Beauv. ‘Sandwich insula, leg. Hillebrand.’ (MEL 587900). Rhynchospora spicaeformis W.F. Hillebr., FI. Haw. Is., 477. = Rhynchospora chinensis Nees & Meyen subsp. spiciformis (W.F. Hillebr.) T. Koyama. ‘W. Maui, Sandw. Ins.’ (MEL 587912). Euphorbiaceae Antidesma pulvinatum W.F. Hillebr., FI. Haw. Is., 403. ‘Hawaii, leg. Hillebrand.’ (MEL 587938). Euphorbia hookeri Steud. 6 var integrifolia W.F. Hillebr., FI. Haw. Is., 397 = Chamaesyce multiformis (Hook. & Arn.) Croizat & Degener. ‘Maui. Sandwich ins. leg. Hillebrand.’ (MEL 1520552). Geraniaceae Geranium tridens W.F. Hillebr., FI. Haw. Is., 55 = G. cuneatum Hook subsp. t ride ns (W.F. Hillebr.) Carlq. & Bissing. ‘Haleakala. Sandw. Ins. leg. Hillebrand.' (MEL 1520566). Geranium bundle W.F. Hillebr., FI. Haw. Is., 56. ‘On the summit of Mt Eika. leg. Hbd.' (MEL 1520565). Gesneriaceae Cyrtandra biserrata W.F.Hillebr., FI. Haw. Is., 329 ‘Molokai. Sandwich Inseln. Leg. Hbd.’ (MEL 587948). Cyrtandra filipes W.F. Hillebr., FI. Haw. Is., 336. ‘W. Maui. Sandwich Inseln. lea. Hbd.’ (MEL 587918). Cyrtandra gracilis W.F. Hillebr., FI. Haw. Is., 333. ‘Oahu. Leg. Hbd.’ (MEL 587919). Cyrtandra grayana W.F. Hillebr., FI. Haw. Is., 330. ‘Maui. Sandwich Inseln. Lea Hbd’ (MEL 587920). Cyrtandra latebrosa W.F. Hillebr. FI. Haw. Is., 337 = Cyrtandra hawaiensis C. B. Clarke. ‘Oahu. leg. Hbd.’ (MEL 587932). Cyrtandra lydgatei W.F. Hillebr., FI. Haw. Is., 335 ‘Maui. Sandwich Ins. leg. Hbd ’ (MEL 587922). Cyrtandra macrocalyx W.F. Hillebr., FI. Haw. Is., 329. ‘Molokai. Sandwich Inseln leg Hbd.’ (MEL 587924). Cyrtandra paritiifolia W. F. Hillebr., FI. Haw. Is., 328 = C. hashimotoi x C.platyphylla. ‘Maui. Sandwich Inseln. leg. hbd.’ (MEL 587926). Cyrtandra procera W.F. Hillebr., FI. Haw. Is., 329. ‘Molokai. Sandw. Inseln. Leg Hbd ’ (MEL 587926). Goodeniaceae Scaevola cylindrocarpa W.F. Hillebr., FI. Haw. Is., 268 = Scaevola chamissoniana Gaudich. ‘Lanai. Sandwich Inselnn.’ (MEL 587831). Scaevola procera W.F. Hillebr., FI. Haw. Is., 268. ‘Molokai. Sandwich Inseln’ (MEL 587832). 46 J.W. Grimes Lamiaceae Haplostachys rosmarinifolia W.F. Hillebr., FI. Flaw. Is., 347 = Haplostachys linearifolia (Drake) Sherff. 'Molokai, Sandwich Inseln.’ (MEL 1520492). Phyllostegia ambigua (A. Gray) W.F. Hillebr. var. longipes W.F. Hillebr., FI. Haw. Is., 350 = Phyllostegia warshaueri St. John. ‘Maui. Sandwich Inseln. leg. Lydgate.' (MEL 587949). Phyllostegia hispida W.F. Hillebr., FI. Haw. Is., 353. 'Molokai. Sandwich Insel. leg. Hbd.' (MEL 587950). Stenogyne bifidia W.F. Hillebr., FI. Haw. Is., 358. ’Molokai. Sandwich Inseln..’ (MEL 587951). Stenogyne viridis W.F. Hillebr., FI. Haw. Is., 361. 'Maui. Sandwich Inseln. leg. Hbd. (MEL 587954). Loganiaceae Labordia grayana W.F. Hillebr., FI. Haw. Is., p. 290 = Labordia hedyosmifolia Baill. ‘Maui. Sandwich Inseln. leg. Hbd.' (MEL 587790) Labordia lophocarpa W. F. Hillebr., FI. Haw. Is., p. 289 = Labordia waiolani Wawra ‘Molokai. Sandwich Ins.’ (MEL 587792). Labordia triflora W.F. Hillebr., FI. Haw. Is., 293. Molokai. Sandwich Inseln. leg. Hbd. (MEL 587795). Malvaceae Hibiscus kokio W. F. Hillebr. Flora 56: 174 (1873). 'Molokai. leg. Hillebrand. (MEL 587718). Menispermaceae Cocculus virgatus W. F. Hillebr., FI. Haw. Is., 8 = Cocculus trilobus (Thunb.) DC. 'Lanai. Sandw. Ins. leg. Hillebrand.’ (MEL 587721). Myrsinaceae Enibelia pacifica W.F. Hillebr., FI. Haw. Is., 282. ‘Maui. Sandwich Ins. leg. Hbd. (MEL 587796). Myrsine lanaiensis W.F. Hillebr., FI. Haw. Is., 281. 'Lanai. Sandwich In. leg. Hbd. (MEL 611633). Nyctaginaceae Pisonia sandwicensis W.F. Hillebr., FI. Haw. Is., 369. Lanai. Sandiwch Ins. leg. Hillebrand.’ (MEL 1520475). Orchidaceae Habenaria holochila W.F. Hillebr., FI. Haw. Is., 432 = Platanthera holochila (Hillebr.) Kraenzl. ‘Kanai. leg. Knudsen.' (MEL 1520529) Piperaceae Peperomia hypoleuca Miq. b var. mantis eeka W.F. Hillebr., FI Haw. Is., 422 —P. eekana C.DC. ‘West Maui. Sandwich Ins. leg. Hillebrand.’ (MEL 1520494). Peperomia macraeana C. DC. b var. pic la W. F. Hillebr., FI. Haw. Is., 421 - P. macraeana C. DC. ‘W. Maui. Sandw. Inseln. leg. Hillebrand. (MEL 587955). Peperomia mauiensis Wawra b var. parvifolia W.F. Hillebr., FI. Haw. Is., 425 = P. ligustrina W.F. Hillebr. 'West. Maui. Sandwich I. leg. Hillebrand.’ (MEL 587956). Pittosporaceae Pittosporum cauliflorum Mann var fulvum W.F. Hillebr., FI. Haw. Is.. 25 = P. confertiflorum A. Gray. ‘Ewa, Waiahua,. Oahu. leg. Hillebrand. (MEL 587722). Pittosporum glomertum var acutisepala W.F. Hillebr., FI. Haw. Is., 23 = Pittosporum glabrum Hook. & Arn. 'Niu, Oahu. leg. Hillebrand.’ (MEL 587724). Pittosporum kauaiense W.F. Hillebr.. FI. Haw. Is., 25. Kauai, leg. Knudsen. (MEL 587725). Haas (1977: 143) lectotypified this taxon on the description. Hillebrand types 47 POACEAE Agrostis sandwicensis W.F. Hillebr., FI. Flaw. Is., 515. L Oahu. leg. Hillebrand. (MEL 1520483). Eragrostis grandis W.F. Hillebr., FI. Haw. Is., 528. "Molokai. Sandw. Ins. leg. Hillebrand.’ (MEL 1520486). Eragrostis pheloides W. F. Hillebr., FI. Haw. Is., 530 = Eragrostis variabilis (Gaudich.) W.F. Hillebr. ‘Haleakala: Sandw. I.’ (MEL 1520487). Eragrostis thyrsoidea W. F. Hillebr., FI. Haw. Is., 529 = E. variabilis (Gaudich.) W. F. Hillebr. "Oahu. leg. Hillebrand.' (MEL 1520488). Eragrostis variabilis (Gaudich.) b var. ciliata W.F. Hillebr., FI. Haw. Is., 529 — Eragrostis variabilis (Gaudich.) Steud. ‘Sandwich Inseln.' (MEL 1520523). Panicum imbricatum W.F. Hillebr.. FI. Haw. Is., 501 = Dichantheliaai cynodon (Reichardt) C. A. Clark & Gould ‘Mt Eika: Sandwich. I. leg. Hillebrand.’ (MEL 1520489). Panicum nephelophilum Gaudich. var xerophilum W.F. Hillebr. FI. Haw. Is. 498 = Panicum xerophilum (W.F. Hillebr.) Hitchc. ‘Maui. Sandwich I. leg. Hillebrand. (MEL 1520491). Primulaceae Lysimachia rotundifolia W.F. Hillebr., FI. Haw. Is., 284, nom. illeg. = L. hillebrandii J.D.Hook. ex A.Gray. ‘Oahu. leg. Hbd.’ (MEL 587803). Rhamnaceae Aiphitonia ponderosa W.F. Hillebr., FI. Haw. Is., 81. ‘Molokai. Sandw. In. leg. Hbd. (MEL 1520399). Gouania hillebrandii Oliver in W.F. Hillebr., FI. Haw. Is., 83. ‘Gullies of Lahaina, Sandwich Ins. leg. Hbd.' (MEL 1520401). Rubiaceae Coprosma cymosa W.F. Hillebr., FI. Haw. Is., 186. 'Hawaii, leg. Hbd.’ (MEL 1520424). Coprosma stephanocarpa W.F. Hillebr., FI. Haw. Is., 187 = Coprosma foliosa A. Gray. ‘OH. Maui. Sandwich Ins. leg. Hbd. (MEL 1520413) and ‘Maui. Sandwich Ins. leg. Hbd.’ (MEL 1520414). Kadua foliosa W.F. Hillebr., FI. Haw. Is., 164 = Hedyotis foliosa (W.F. Hillebr.) Fosberg ‘Maui. Sandwich Inseln.’ (MEL 1520424). Kadua formosa W.F. Hillebr., FI. Haw. Is., 165 = Hedyotis formosa (W.F. Hillebr.) Fosberg 'Maui. Sandwich Ins. leg. Hbd.’ (MEL 1520425). Kadua littoralis W.F. Hillebr., FI. Haw. Is., 166 = Hedyotis littoralis (Hillebr.) Fosberg. ‘Molokai. Sandwich. Ins. leg. Hbd.’ (MEL 1520429). Kadua remyi W.F. Hillebr., FI. Haw. Is., 162 = Hedyotis schleclitendahliana Steud. ‘Lanai. Sandw. Ins.’ (MEL 1520432). Rutaceae Fagara hawaiiensis (Hillebr.) Sherff var. subacuta Sherff. ‘Lanai, Sandw. I.’ (MEL 520437). Published by Hillebrand as Zanthoxylum hawaiiensis b var. FI. Haw. Is., 76. Pelea molokaiensis W.F. Hillebr., FI. Haw. Is., 65. ‘Kalea. Molokai. Sandwich Ins. leg. Hillebrand.’ (MEL 587806). Pelea orbicularis W.F. Hillebr., FI. Haw. Is., 67. ‘Mount Eka, W. Maui. Sandwich Ins. leg. D. Lydgate.’ (MEL 587807) Pelea sandwicensis A. Gray var. lucens W.F. Hillebr., FI. Haw. Is., 66 = Pelea oahuensis H. Lev. ‘Oahu. leg. Hillebrand.’ (MEL 587811) Pelea volcanicci A. Gray b var. grandifolia W.F. Hillebr., FI. Haw. Is., 67 = Pelea volcanica A. Gray. ‘Hawaii, leg. Hillebrand.’ (MEL 587873 & 587874). Sapindaceae Sapindus oahuensis W. F. Hillebr., FI. Haw. Is., 85. ‘Kalihi. Sandw. Inseln. leg. Hbd.' (MEL 1520574). 48 J.W. Grimes Sapotaceae Chrysophyllum polynesicum W.F. Hillebr., FI. Haw. Is., 277 = Nesoluma polynesicum (W.F. Hillebr.) Baill. ‘Lanai. Sandw. Ins. leg. Hbd..' (MEL 587814). Sideroxylon spathulatum W.F. Hillebr., FI. Haw. Is., 277 = Pouteria sandwicensis (A. Gray) Baehni & Degener. ‘Lanai. Sandwich Ins. leg. Hbd.' (MEL 587813). SOLANACEAE Solatium incompletum Dunal b var. mauiensis W.F. Hillebr., FI. Haw. Is., 305 = Solatium incompletum Dunal. ‘Maui. Sandwich Inseln. leg. Hbd.’ (MEL 1520575). VlOLACEAE Viola helioscopia W.F. Hillebr.. FI. Haw. Is., 17 = Viola chamissoniana Ging. ‘Makaha, Oahu. leg. Hbd.’ (MEL 587733). Viola robusta W.F. Hillebr.. FI. Haw. Is., 16 = Viola chamissoniana Ging. subsp. robusta (W.F. Hillebr.) W. Wagner, Herbst & Sohmer. ‘Molokai'. Sandwich I.’ (MEL 587734). Possible types Pteridaceae Pteris irregularis Kaulf. b var. linearis. W.F. Hillebr.. FI. Haw. Is., 628. The label on the MEL sheet says only "Pteris irregularis Kaulf. b. Sandw. Islds.” (MEL 1520621) Lindsaya alexandri W.F. Hillebr., FI. Haw. Is., 622. In the protologue the syntypes were collected by Prof. Alexander and Mr. Lydgate, and by Knudsen. The MEL sheet has no collector noted. (MEL 1520442). Amaranthaceae Nototrichum liumile W.F. Hillebr., FI. Haw. Is., 373. The label reads Wainae. leg. Hillebrand.’ In the protologue, Hillebrand cites "Oahu! Cape Kaena, near the coast. Waina is near Cape Kaena on the eastern side of the island ot Oahu. (MEL 587937). Rutaceae Pelea orbicularis Hillebr.. FI. Haw. Is., 67. The label ot the MEL sheet (MEL 587807) reads ‘Mount Eika, W. Maui, Sandwich Is. leg. F. Lydgate’ The protologue gives this same locality for one of the syntypes, but does not cite Lydgate. In other cases (c.f. Pittosporum kauaiense) the collector, it not Hillebrand, was cited. Nonetheless, it is likely that this sheet is a type. Rubiaceae llobea timonioides Hook.. Ic. pi. t. 1070 (1870). The protologue reads ‘western end of the Kokala Range, Dr. Hillebrand. n. 186.’ The sheet at MEL (MEL 1520405) says only ‘ Bobea timonioides [J.D.Hook.] Hbd.. Hawaii, leg. Hbr. Types mistakenly reported to be at MEL The types of the following names are reported by Corn (1983. unpublished), who visited MEL and photographed types of Hillebrand’s that she could identify. These photographs have been shared by other workers. In this manuscript Corn repot ts the following types as being at MEL. These reports are mistaken. Peperomia ligustrina Hillebr. FI. Haw. Is., 424. Corn lists this name under Piese nt Name .’ and then has an entry for the ‘ Name in Hbd. 1888 Under Name in Hbc L 1 888 ’ Corn lists ‘ Peperomia mauiensis Wawra b var. parvifolia. Actually, both / . ligustrina and P. mauiensis b var. parvifolia are published on the same page. No duplicate of the specimen cited in the protologue of P. ligustrina, ( i.e . ‘W. Maui! Waihee, and forms with shorter and proportionally broader, somewhat obtuse, 1- nerved leaves from Kaanapali , about 2000 ft. above the sea. ) has been located at Hillebrand types 49 MEL. The type of Peperomia mauiensis Wawra b var. parvifoliafq.v.) is erroneously included in Corn’s manuscript of type-photos as the type of the former name. Cyrtandra malacophylla C.B. Clarke, Monogr. phan. 5: 227 (1883). The protologue reads “Kauai, Hillebrand, in hb. K.’ The sheet Corn shows in photo has a label reading ‘Maui...’, and is in actuality the type of Cyrtandra parritifolia ( q.v .). References Corn, C.A. (1983). The Melbourne Hillebrand collection of Hawaiian plants. Newsletter, Hawaiian Botanical Society. 22 , 20-2 1 . Corn, C.A. (1983, unpublished manuscript, distributed by author). The Hillebrand Hawaiian Collection at Melbourne. [A manuscript of photographs of sheets from the Hillebrand Herbarium deposited at MEL. Some photographs are annotated with type designations.] Deposited at the Library, Royal Botanic Gardens, Melbourne. Haas, J.E. (1977),. The Pacific species of Pittosporum Banks ex Gaertn. (Pittosporaceae). Allertonia 1, 73-167. Hillebrand, W. (1888). Flora of the Hawaiian Islands. (B. Westermann & Co.: New York). Lammers, T.G. (1988) New taxa, new names, and new combinations in the Hawaiian Lobelioideae (Campanulaceae). Systematic Botany 13 , 496-508. Lammers, T.G. (1991) Systematics of Clermontia (Campanulaceae-Lobelioideae). Systematic Botany Monographs 32, 1-97 Lammers T.G. (1994) Typification of the names of hawaiian Lobelioideae (Campanulaceae) published by Wilhelm Hillebrand or based upon his specimens. Taxon 43 , 545-572. Lowrey, T.K. (1986). A biosystematic revision of Hawaiian Tetramolopium (Compostiae: Astereae). Allertonia 4 , 203-265. Pescott, R.J.M. (1982), The Royal Botanic Gardens, Melbourne. A history from 1845 to 1970. (Oxford University Press: Melbourne.) Rock, J.F. (1919). A mongraphic study of the Hawaiian species of the tribe Lobelioideae family Campanulaceae. Memoirs of the Bernice Pauahi Bishop Museum 7 , 1-395. Sayre, G. (1975). Cryptogamae exsiccatae. An annotated bibliography of exsiccate of Algae, Lichenes, Hepaticae and Musci. V. Unpublished Exsiccatae. I. Collectors. Memoirs of the New York Botanical Garden 19 , 21-ATb. Stafleu, F. and Cowan, R. (1979). Taxonomic Literature - II. Vol. II, Regnum Veg. 98 , 1-991. Wagner, W.L., Herbst, D.R., and Sohmer, S.H. (1990). ‘Manual of the Flowering Plants of Hawai’i.' (University of Hawai’i Press, Bishop Museum Press: Honolulu). . Muelleria 14:51 (2000) Additions to the Hygrocybeae (Fungi, Hygrophoraceae) of Victoria. I. A. M. Young Hon. Associate, Queensland Herbarium, Brisbane Botanic Gardens Mt Coot-tha, Mt Coot-tha Road, Toowong, Qld, 4066, Australia. mail address: Bee Cottage, Langton Rd, Blackbutt, Qld 4306, Australia, email: tyoung@bigpond.com Abstract Herbarium material from the Jumping Creek Nature Walk of the Warrandyte State Park (Victoria) includes previously described taxa both known from and new to Australia, as well as five unde- scribed species. Taxa previously known from Australia include Hygrocybe austropratensis A. M. Young, H. cheelii A. M. Young, H. leucogloea A. M. Young, H. rodwayi (Massee) A. M. Young and H. virginea var. virginea (Wulfen : Fr.) P.D. Orton & Watling. Two taxa, Hygrocybe psittacina var. perplexa (A.H.Sm. & Hesler) Boertm. and H. virginea var. fuscescens (Bres.) Arnolds, are new records for Australia. Three new species are described: Hygrocybe arcohastata, H. fuhreri and H. saltorivula. The herbarium material also contains three collections which represent two species believed to be new taxa, but for which it is considered there is insufficient material for a valid description and deposition of a suitable type. Descriptions of these two taxa have been provided to assist with further studies of the Jumping Creek taxa. Introduction The Warrandyte State Park is located 25 kilometres north-east of Melbourne (Victoria, Australia) and is on the east bank of the Yarra River. Within the park is the Jumping Creek Nature Walk which passes through various vegetation communities, including the kunzea thickets that contained the fungi examined during this study. The park receives an annual rainfall of about 600 mm with a greater part of this rain- fall occurring during the late autumn and winter months of May to August. Rainfall is somewhat erratic and there can be extended dry periods. The dry, siltstone ridges of the park support open stands of Eucalyptus polyanthemos (red box) sometimes mixed with E. radiata (narrow-leaved peppermint). The slopes running towards the river are mostly alluvial soils and the trees supported include E. viminalis (manna gum), E. melliodora (yellow box) and E. goniocalyx (long-leaved box). These slopes also have an extensive shrub understorey which includes Pomaderris prunifolius, Correa glabra and Kunzea eri- coides (burgan) (Anon. 1997; B.A.Fuhrer, pers. comm.). Parts of the reserve bordering the nature walk are dominated by very dense thickets of K. ericoides about 2-3 metres in height. These thickets replace the usual dominant trees ( Eucalyptus spp.) which are sometimes slowly killed as an indirect result of the feeding activities of Manorina melanophrys (bell miner). The bell miner is a native bird species that feeds on, but also encourages the spread of, lerp insects which can, in large infestations, severely damage eucalypt foliage. The kunzea thickets produce a very extensive and dense canopy which prevents eucalypt germination and re-growth but which also provides a microclimate with ideal conditions for the formation of dense moss and lichen beds on the ground-surface. The moss species present include Thuidium furfurosum and Ptychomnion aciculare and they in turn provide an ideal habitat for fungal species that require such con- ditions (B.A.Fuhrer, pers. comm.). The moss beds, with their maintenance of suitable con- ditions of temperature, moisture and humidity, are the critical factor that allows the Hygrophoraceae to flourish within the park, because the rainfall is too erratic and the nor- mal woodland or forest litter within the park is too exposed to dessication to permit occur- rence of the Hygrophoraceae unless there are very unusual climatic conditions. 52 A.M. Young Materials and Methods Fourteen, air-dried collections from the Jumping Creek Nature Walk locality were exam- ined. Field notes accompany each collection, but these are often incomplete and the notes were supplemented using information obtained from very high quality photographs that were made of each collection under natural conditions. Standardised colour codes were not provided with the herbarium collections. All material has been deposited at the National Herbarium of Victoria (MEL). Descriptions and illustrations are provided for the new taxa and for those species which are either not illustrated in previous papers (Young & Wood 1997; Young 1999) or which require additional text or diagrams as a result of new information. The habit-sketch shows basidiome dimensions. Transverse sections (either drawn or photographed) were not pro- vided with the collection material. The microstructures of the pileus, hymenophoral trama and stipe are generally not depicted because they usually conform to standard forms (Young & Wood 1997). For each illustrated specimen, 20 spores and 10 basidia were selected at random, drawn and measured. Scale bars are provided for all drawings: habit sketches, 10 mm; all microstructures, 10 pm. The derived parameter ‘Q’ is defined as the quotient of the length divided by the width of the relevant spore or basidium; the mean ‘Q’ is the quotient of the mean length and width respectively. This paper lists several species of Hygrophoraceae originally collected and described from Europe for which no types are designated (Boertmann pers. comm.). This problem has already been addressed (Young 2000) and where types for European taxa do not exist, the species concepts of Boertmann (1995) are used. Species: Information and Descriptions 1. Hygrocybe austropratensis A.M. Young, Austrobaileya 5: 546 (1999). Type: New South Wales. Lane Cove Bushland Park, 33°49'S 15I°10"E, 7.vi. 1 998. R. & E. Kearney s.n. (holotype DAR 73916 ; isotype BR1). Illustration: Young (1999), p. 546. Habitat: Gregarious or caespitose on soil amongst moss. Material examined: Victoria. Warrandyte State Park. 23.V.1996, B.A.Fuhrer 2055 (MEL 2063194). Remarks: The macrocharacters of the Jumping Creek material agree with those of the type description. The swollen stipe base observed in the holotype material is also present in the Victorian collection, suggesting that the characteristic is not peculiar to the Lane Cove collections. A swollen stipe base does not appear in the closely related European Hygrocybe pratensis (Pers. : Fr.) Murrill; other characters which also separate H. pratensis are dis- cussed in Young (1999). The spores of the Victorian material (5.5-8 x 4.5-6 pm, mean 6.8 x 5.3 pm. Q: 1.1 -1.5, mean Q: 1.28) are slightly smaller than those of the type (6— 8.3(— 9) x 5-7.3 pm. mean 7.5 - 6.3 pm, Q: 1.1-1. 4. mean Q: 1.2) but the two ranges overlap so extensively that the difference is not considered significant. The basidia of the Jumping Creek collection also show the same smaller size (44—59 x 6-8 pm. mean 50.7 x 6.8 pm. Q: 6. 1-8.8. mean Q: 7.48) compared with the type (53-69 x 6-8 pm, mean 62.0 x 6.8 pm, Q: 6.6— 1 0.2(— 1 2.6). mean Q: 9.20) but again the overlap is considerable and the differences are not considered significant. Only the type collection was previously known. 2. Hygrocybe cheelii A.M. Young, Austrobaileya 5: 547 (1999). Type: New South Wales. Gladesville, 17. vi. 1916. J.B.Clelaml s.n. (holotype AD 3418). Cantharellus lilacinus Cleland & Cheel, Trans. & Proc. Roy. Soc. S. Australia 43: 271 (1919). Type: New South Wales. Gladesville, 17. vi. 1916. J.B.Cleland s.n. (holotype AD 3418). Camarophyllus Hygrocybeae of Victoria 53 lilacinus (Cleland & Cheel) E.Horak, New Zealand J. Bot. 28; 203 (1990) non Hygrocybe lilacina (C.Laest. ex P.Karst.) M. Moser, Die Rohrlinge unci Bldtterpilze (Agaricales) 3 ed„ 64 (1967). Illustrations-. Young (1999), p. 547; Willis (1963), plate 9, fig.l as Cantharellus lilac- inus ; Cleland & Cheel (1919), Plate 29, fig.l. Habitat : Gregarious amongst leaf mould and moss under Kunzea ericoicles. Material examined : Victoria. Warrandyte State Park, 4.vi.l995, BA FuhrerI937 (MEL 2063188). Remarks-. The macrocharacters of the basidiomes in MEL 2063188 conform very closely to those ot the holotype description (Cleland & Cheel 1919) and the material pre- viously described from the Lane Cove Bushland Park (Young 1999). The differences in spore dimensions are insignificant: the Jumping Creek Nature Walk material has spores that measure 7-10 x 4.5-5. 5(— 6.5) pm, mean 8.4 x 5.4 pm, Q: 1.3-1. 8, mean Q: 1.56; the holotype description gives the spores as 7-8.5 x 4.5-5. 5 pm, and re-examination of the type gave spores measuring 6.0-8.5 x 4.5-6.0 pm, mean 7.2 x 5.5 pm, Q; 1.2-1. 7 and mean Q: 1 .5. The holotype collection is in very poor condition due to insect attack and original preservation. Obtaining critical information from this material is quite difficult especial- ly since almost nothing remains of the lamellae. For these reasons, two collections, B.A Fuhrer 1937, (MEL 2063188), and hb young 2118 (BRI) are here nominated as exemplar material as each is considered identical to the holotype in all significant respects (Young 1999). The collection in the Queensland Herbarium (BRI) is particularly significant as it contains material that was collected within approximately 10 kilometres of the original 1916 collection of the holotype at Gladesville, New South Wales. 3. Hygrocybe leucogloea A. M. Young, in Young & Wood, Austral. Svst. Bot. 10: 976 (1997). Type: New South Wales. Mt. Wilson, 33°30’S 150°22’E, 29.iv.1989, A.E.Woocl s.n. (holotype UNSW 89/87). Illustration: Young & Wood (1997), p. 984. Habitat: Gregarious or caespitose in moss amongst litter. Material examined: Victoria. Warrandyte State Park, 23.V.1996, B A Fuhrer 2060 (MEL 2063192). Remarks: The first Victorian record of Hygrocybe leucogloea was from the Black Range State Forest (Young 2000). The Jumping Creek Nature Walk collection has spores that measure 6— 7(— 8) x 4-5(-5.5) pm, mean 6.7 x 4.5 pm, Q: 1.3-1. 7, mean Q: 1.50, a slightly smaller mean and reduced upper range when compared with spores of the holotype collec- tion [(6. 3-)6.5-7. 9(-8. 5) x 4.0-5. 6 pm, mean 7.2 x 4.8 pm, Q: 1.2-1. 7, mean Q: 1.50], Although medallion clamps are present on the hyphae of the ixotrichoderm in the holotype collection, none were found on the relevant hyphae of the Jumping Creek Nature Walk col- lection although clamps were abundant. This difference is thought to be part of a morpho- logical range of basidiome variations likely to be encountered in this taxon. 4. Hygrocybe psittacina var. perplexa (A.H.Sm. & Hesler) Boertm., Fungi of Northern Europe. 1: 82 (1995). Type: U.S.A. Michigan, Cheboygan County, T.E. Brooks & A. H. Smith 21491 (holotype MICH 10924, n.v.). Hygrophorus perplexus A.H.Sm. & Hesler, Sydowia 8: 328 (1954). Hygrocybe perplexa (A.H.Sm. & Hesler) Arnolds Persoonia 12: 477 (1985). Illustration: Boertmann (1995), p. 83. 54 A.M. Young Pileus 15-30 mm, orange-brown to reddish brown but darker at the centre, convex becoming plane or upturned but then retaining an umbo at the centre, very glutinous, strongly striate at the margins and for up to one-third of the pileal diameter, margins paler and a little crenulate. Lamellae adnate to narrowly adnate, pinkish buff, thick, distant, margins concolorous and even. Stipe 30—45 x 2-3 mm, orange-red but paler at the apex, smooth, cylindrical, very glutinous. Spores 6— 8(— 9) x 4-5.5 pm, mean 7.0 x 4.8 pm, Q: 1 .2— 1 .6(— 1 .8), mean Q: 1.46, broadly ellipsoid to obovoid or sub-lacrymoid or occasionally sub-globose, smooth, hya- line, thin-walled, inamyloid. Basidia (27-)31-47 x 7-9.5 pm, mean 38.2 x 8.0 pm, Q: 3. 9-6. 2, mean Q: 4.78, (2-) 4-spored, clavate, hyaline, thin-walled, clamped and some- times of medallion form. Cystidia absent. Hymenophoral trama regular in the upper two- thirds of the lamellae but sub-regular to irregular in the lower one-third nearest the mar- gins; regular trama composed of short, parallel, thin-walled, hyaline elements 33-35 x 5-22 pm, clamps present; sub-regular to irregular trama composed of ellipsoid to sub- globose, hyaline, thin-walled elements 15-5 x 7-3.5 pm, clamps not seen. Lactifers may be present in the regular trama section and appear as meandering, contorted, hyaline hyphae 3- 5 pm diameter, but do not exhibit the typical very high refractive index and may be overlooked. Pileipellis a very well defined ixotrichoderm 50-120 pm in thickness and composed of thin-walled, hyaline, septate, cylindrical hyphae 1.5-3. 5 pm diameter, usually with apices that are rounded or slightly swollen, clamped and sometimes of Figure 1. Hygrocybe psittacina var. perplexa. A habit sketch: B spores; C basidia. Hygrocybeae of Victoria 55 medallion form. Stipitipellis a well developed ixotrichoderm composed of hyaline, thin- walled, septate hyphae 1.5— 5mm diameter, medallion clamps usually present. (Fig. 1) Habitat : Gregarious in moss and litter under Kunzea ericoides. Material examined'. Victoria. Warrandyte State Park, 10.vi.1994 BAFuhrer 1994 (MEL 2063197). Remarks: The material in MEL 2063197 agrees almost exactly with the description of the taxon given in Boertmann (1995) which has spores measuring (6— )7— 8.5(— 9) x (4-)4.5-5.5(-6) pm, Q: 1 .4-1.9 mean Q: 1 .5-1.6, and basidia measuring 36-50 x 7-8 pm which have medallion clamps at their bases. Basidiomes ot Hygrocybe psittacina var. perplexa are leadily separated from brownish red basidiomes ot H. graminicolor , because the latter have umbilicate pilei and arcuate or decurrent lamellae with cheilocystidia embedded in a glutinous thread on their margins. This is the first confirmed record of this Northern Hemisphere taxon for Australia. The species was first described from North America but is also known from Europe (Boertmann 1995; Arnolds 1990) and Japan (Imazeki, Otani & Hongo 1988). 5. Hygrocybe rodwayi (Massee) A. M. Young, in A.M. Young & A.E.Wood, Austral. Syst. Bot. 10: 923 (1997). Hygrophorus rodwayi Massee, Bull. Misc. Inform. Kew 1899: 178 (1899). Type: Tas. Kingston Rd. (nr. Hobart), undated , L.Rodway 137 (holotype K). Camarophyllus rodwayi (Massee) Monks & A. K. Mills in Banks et al. (eds). Aspects of Tasmanian Botany - A Tribute to Winifred Curtis 13 (1991). Illustrations: Young & Wood (1997), p. 925; Fuhrer & Robinson (1992), p. 39. Habitat: Gregarious in moss. Material examined : Victoria. Warrandyte State Park, 23 May 1996 .BAFuhrer 2059 (MEL 2063193). Remarks. This collection has sub-globose to globose spores measuring (4.5-)5.5-6(-7) x (4-)4.5-5.5 pm, mean 5.7 x 4.8 pm, Q: 1. 0-1.4, mean Q: 1.18. This agrees very well with the holotype collection, which has similar spores measuring 5-7 x 4-6 pm, mean 5.8 x 5. 1 pm, Q: 1 .0—1 ,3(— 1 .4), mean Q: 1.15. The photographic material accompanying the Warrandyte collection depicts basidiomes that have a strong resem- blance to the other common white taxon, Hygrocybe virginea var. virginea (Wulfen : Fr.) P.D. Orton & Watling, but the two are always separable microscopically because H. vir- ginea var. virginea has much larger, ellipsoid spores measuring 8—1 1 (—12) x 5-8 pm. Hygrocybe rodwayi occurs in eastern Australian forests from the Sydney region to Tasmania (Young & Wood 1997; Young 2000). 6. Hygrocybe virginea (Wulfen : Fr.) P.D.Orton & Watling, Notes Roy. Bot. Card. Edinburgh 29: 132 (1969). Agaricus virgineus Wulfen, in Jacq., Misc. austr. 2: 104 (1781). Type: none designated. A. virgineus Wulfen : Fr., Syst. mycol. 1: 100 ( 1 821 )• Hygrophorus virgineus (Wulfen : Fr.) Fr., Epicr.: 327 (1838); Camarophyllus virgineus (Wulfen : Fr.) PKumm., Fiihr. Pilzk.: 1 17 (1871 ). Agaricus niveus Scop., FI. cam.. Ed. 2, 2: 430 (1772). Type: none designated. A vir- gineus var. niveus (Scop. ) Fr., Syst. mycol. 1 : 1 00 ( 1 82 1 ); Hygrophorus niveus (Scop.) Fr., Epicr.: 327 (1838); Camarophyllus niveus (Scop.) Wunsche, Pilze: 1 15 (1877). Key to varieties of Hygrocybe virginea 1 . Pileus pure white; spores ellipsoid, rarely constricted 6a. var. virginea 1. Pileus brown at the centre; spores ellipsoid, often cylindrical and constricted var. fuscescens Figure 2. / D Hygrocybe virginea. var. virginea. A spores, var . fuscescens. B spores; C habit sketch; D basidia. 6a. var. virginea (Fig. 2) Illustrations'. Young & Wood ( 1997), p. 929; Boertmann ( 1995), p. 49. Habitat'. Gregarious in moss under Kunzea ericoides. Material examined: Victoria. Warrandyte State Park. 10. vi. 1995. B.A.Fuhrer 1942 (MEL 2063190)'. Warrandyte State Park. lO.vi. 1995, B.A.Fuhrer 1943 (MEL 2063198). Remarks: These collections have macroscopic characters that agree very closely with those previously described for Australian material (Young & Wood 1997; Young 1999). There are minor microscopic variations. Collection MEL 2063190 contains basidiomes with 2-spored basidia and spores that display occasional small constrictions measuring 8— 1 1 (— 1 2 ) x 5-8 pm, mean 9.6 x 6.4 pm. Q: 1.2-1. 8. mean Q: 1.50; clamps are absent throughout the basidiome. Collection MEL 2063198 contains basidiomes that have 4-spored basidia and spores measuring 6 — 8(— 9) x 3. 5-6. 5 pm, mean 7.3 x 4.6 pm, Q: 1 .3-1 .8(— 1 .9), mean Q: 1 .6. Basidiomes with either or both 2-spored and 4-spored basidia are reported in Boertman (1995) and the Australian material agrees fully with that description. Previous papers (Young & Wood 1997; Young 1999) recorded this taxon as Hygrocybe virginea (Wulfen ; Fr.) P.D. Orton & Watling. The Jumping Creek collections conform very closely with European descriptions of var. virginea. 6b. var. fuscescens (Bres.) Arnolds, Persoonia 12: All (1985). Type: none designated. Hygrophorus niveus var. fuscescens Bres., Iconogr. rnycol. 1: pi. 330 (1928). Illustration: Boertmann (1995), p. 51. Hygrocybeae of Victoria 57 Spores 8-1 1 x 4-6 pm, mean 9.5 x 4.9 pm, Q: 1 .5-2.3, mean Q: 1 .93, long-ellipsoid to cylindrical and often strongly constricted. Basidia 2-spored. The remainder of the micro- scopic characters conform with var. virginea. (Fig. 2) Habitat: Gregarious amongst leaf litter and moss. Material examined : Victoria. Warrandyte State Park, 4 June 1995, B.A.Fuhrer 1936 (MEL 2063191). Remarks: Macroscopically, this taxon differs from var. virginea only in that the cen- tre of the pileus is distinctly brown-tinted. Microscopically, the spores differ as present- ed in the key. This is the first record of this taxon for Australia. Its characters agree very well with the those in the description by Boertmann (1995). 7. Hygrocybe arcohastata A. M. Young, sp. nov. Pileus 20-40 mm latus, atro-olivaceo-viridis deinde aurantiacus vel aurantiaco-ruber, conicus dein lato-conicus vel applanatus vel umbonatus, glaber, sub-viscidus, ad margin- em aequalis vel crenulatus, striatus. Lamellae adnatae vel arcuatae, virello-flavae dein subaurantiacae, ad marginem concolorae. Stipes 20-40 x 2-4 mm, super sub-viridis, sub- malvinus, cylindricus, glaber, siccus, ad basim sub-aurantiacus. Sporae 7. 5-9. 5 x 4.5-6 pm. Q: 1 .4—1 .8(— 2.0), ellipsoideae, hyalinae. Basidia 38-50 x 6— 9(— 1 0.5) pm, Q: (3.6-)4.2-6.5, 4-spora, ad basim fibulata medallionae. Cystidia nulla. Trama hymenophoralis regularis, haud fibulata. Epicutis pilei sub-ixocutem formans; hyphae cuticularis hastatae, pigmentae. Gregaria vel caespitosa in musco sylvestri. Type: Victoria. Warrandyte State Park, 24. v. 1996, B.A.Fuhrer 2064 (holotype MEL 2063201). Pileus 20-40 mm., at first deep olive-green but dark purple-tinted at the centre and yel- low-tinted at the margin, changing to orange or orange-red with the colour change com- pleted before the pileus is fully expanded; conical becoming broadly conical and finally more or less plane with a distinct umbo, smooth, slightly viscid, margin even to crenulate and striate. Lamellae adnate with a decurrent tooth or arcuate, greenish yellow becoming orange-tinted with age, margins even and concolorous. Stipe 20-40 x 2-4 mm; pale green near the lamellae, mauve-tinted in the middle section and orange-tinted towards the base; more or less cylindrical but a little tapered at the base; smooth, dry. Dried material becomes brownish pink to orange. Spores 7. 5-9. 5 x (4.5— )5— 6 pm, mean 8.5 x 5.3 pm, Q: 1.4-1 .8(-2.0), mean Q: 1.60, ellipsoid, smooth, hyaline, thin-walled, inamyloid. Basidia 38-50 x 6— 9(— 1 0.5) pm, mean 44.7 x 8.3 pm, Q: (3.6-)4.2-6.5, mean Q: 5.32, 4-spored, with medallion clamps. Cystidia absent. Hymenophoral trama regular and composed of cylindrical, thin-walled, hyaline, inflated and ellipsoid or moniliform elements 19-125 x 4.5-20.5 pm, clamps absent. Pileipellis a weak ixocutis of repent, cylindrical, hyaline, septate hyphae 2-4.5 pm diameter, with spear-like, tapered, acute, pigment-encrusted apices; clamps absent. Stipitipellis a cutis of repent, cylindrical, thin-walled, hyaline, septate hyphae 1 .5-5.5 pm diameter, clamps absent, pigment granules often encrusted on hyphal walls. (Fig. 3) Habitat: Gregarious or caespitose amongst moss in eucalypt woodland. Material examined: Known only from the type. Remarks: The absence of clamps throughout the basidiome (except at the bases of the basidia where they are of medallion form) indicates that this taxon belongs in sub-genus Humidicutis Singer. Hygrocybe arcohastata does not approach any European or North American taxa. but it is close to a group of three New Zealand species, Hygrocybe con- spicua E.Horak, H. luteovirens E.Horak. and H. multicolor (Berk. & Broome) E.Horak (Horak 1990). Hygrocybe conspicua can be separated from H. arcohastata because the former has a brilliant orange pileus and similarly coloured lamellae as well as smaller spores (6-7 x 4-4.5 pm). Hygrocybe multicolor also has smaller spores (5.5-7 x 4-5 pm) 58 A.M. Young Figure 3. Hygrocybe arcohastata. A habit sketch; B spores; C basidia; D cuticular hyphae. and both pileus and lamellae are olive-green becoming bluish green with age. Hygrocybe luteovirens has spores that have a lower upper limit to their range of length and are nar- rower (6-8 x 3.5 — 4.5 pm), has olive-green pilei that gradually change to yellow, yellow- brown, brown or reddish brown with age rather than the intense orange or orange-red of H. arcohastata , and medallion clamps that have the structure of a normal clamp with a central opening rather than the extremely large ‘dough-nut’ shape of the medallion clamps found in II. arcohastata. The only other taxon that is similar is the Japanese species Hygrophorus oli- vaceoviridis Hongo (Hongo 1967). This differs in that the pilei remain olive-green, the lamellae are yellowish and the basidia are much smaller (34-40 x 7.5-8 pm). Etymology: Latin, arcus - a rainbow; Latin, hastatus - armed with a spear; referring to the numerous colours exhibited by this taxon and the long, spear-like endings of the cuticular hyphae. 8. Hygrocybe fuhreri A.M. Young, sp. nov. Pileus 10-20 mm latus, aurantiaco-brunneus, convexus deinde umbilicatus, glaber, siccus, ad marginem striatus, crenulatus dein aequalis. Lamellae decurrentes, sub-auran- tiaco-bubalinae, distantes, ad marginem concolorae. Stipes 30-45 x 2-4.5 mm. sub- aurantiaco-flavus, glaber, siccus, cylindricus, cavus. Sporae 8-10.5 x 4-4.5 (-6) pm, Q: I. 7-2.1. ellipsoideae vel cylindricae, hyalinae, aliqout constrictae. Basidia (40— )49— 59 x 5.5-7 pm, Q: 7. 1-8.7, 4-spora, fibulata. Cystidia nulla. Trama hymenophoralis regularis, fibulata. Epicutis pilei cutis formans. Gregaria in humo vel musco sylvestri. Txpe: Victoria. Warrandyte State Park, 23. v. 1996, B.A.Fuhrer 2054 (holotype MEL 2063199). Pileus 10-20 mm. reddish orange to orange-brown and darker at the centre, convex and umbilicate, dry, smooth, margin striate, even and a little crenulate. Lamellae decurrent, pale orange-buff, distant, margins even and concolorous. Stipe 30-45 x 2^4.5 mm, orange-yellow, cylindrical but tapered at the base, dry, smooth, hollow, often sinuous. Hygrocybeae of Victoria 59 Spores 8-10.5 x 4— 5.4(— 6) pm, mean 9.0 x 5.1 pm, Q: 1 .7-2.1, mean Q: 1.78, ellip- soid to cylindrical, a majority showing some constriction, smooth, hyaline, inamyloid. Basidia (40— )49— 58 x 5.5-7 pm , mean 57.7 x 6.7 pm, Q: 7. 1-8.7, mean Q: 7.69, 4- spored, clamped. Cystidia absent. Hymenophoral trama regular and composed of short, cylindrical to inflated-ellipsoid, hyaline, thin-walled, elements 22-83 x 4.5-13 pm, clamps present. Pileipellis a cutis of repent, cylindrical, hyaline, thin-walled, septate hyphae 2-5.5 pm diameter, clamps present. Stipitipellis a cutis of repent, cylindrical, hya- line, thin-walled, septate hyphae 1.7—4. 2 pm diameter, clamps present. (Fig. 4) Habitat : Gregarious in moss and litter in eucalypt woodland. Material examined'. Known only from the type. Remarks: The regular hymenophoral trama composed of short elements places this species in subgenus Pseudohygrocybe M.Bon. No taxon from either Sri Lanka (Pegler 1986) or North America (Hesler & Smith 1963) approaches H. fuhreri closely. The only European species which shares some similar characteristics according to both Arnolds (1990) and Boertman (1995) is Hygrocybe constrictospora Arnolds; however, that species has bright red pilei which do not (or rarely) become depressed or umbilicate, bright yellow stipes, smaller basidia (35-45 x 5-7.5 pm) and spores that exhibit very strong constrictions rather than the minor constrictions that are present in H. fuhreri. Several orange and/or yellow species from New Zealand (Horak 1990) approach H. fuhreri but each is readily separated: Hygrocybe elegans E. Horak has a trichoderm on the pileus; Hygrocybe procera has much larger spores (1 1—1 7x 7-10 pm); Hygrocybe ceri- nolutea E.Horak does not have constricted spores and the pileus does not become umbil- icate; and Hygrocybe blanda E.Horak has much smaller spores (5.5-7 X 4-4.5 pm). Etymology: named after Mr B.A. Fuhrer of Ringwood, Melbourne Victoria; photog- rapher, naturalist and mycologist. Figure 4. Hygrocybe fuhreri. A habit sketch; B spores; C basidia. 60 A.M. Young 9. Hygrocybe saltorivula A.M. Young, sp. nov. Pileus 20-35 mm latus, aurantiacus vel aurantiaco-rubus, conicus deinde umbonato- applanatus, viscidus, ad marginem sub-striatus, sub-flavus. Lamellae adnatae, ventri- cosae, aurantiaco-rosae, ad marginem sub-flavae. Stipes 30-40 x 2-4 mm, aurantiaco- roseus vel aurantiacus, glaber, lubricus deinde siccus, cavus. Sporae 7.5— 9(— 9.5) x 4-5 pm, Q: 1 .6-2.2, ellipsoideae vel cylindricae, hyalinae, aliquot constrictae. Basidia 37-49 x 5.5-9 pm, Q: 5. 2-5. 7, 4-spora, fibulata. Cystidia nulla. Trama hymenophoralis regu- laris, fibulata. Epicutis pilei ixotrichoderm formans. Gregaria in musco sylvestri. Type'. Victoria. Warrandyte State Park, 23.V.1996. B.A.Fuhrer 2053 (holotype MEL 2063200). Pileus 20-35 mm, orange, orange-red to cherry-red, broadly conical becoming plane but with a central umbo, viscid, smooth, margins even, a little striate and slightly yellowish. Lamellae adnate, ventricose, orange-pink, margins yellowish or a little paler. Stipe 30-40 x 2-A mm, orange-pink or orange especially when older, smooth, slippery or soon dry, cylindrical, hollow. Spores 7.5— 9(— 9.5 ) x 4-5 pm, mean 8.3 x 4.4 pm, Q: 1 .62.2, mean Q: 1 .89, ellipsoid to cylindrical, hyaline, smooth, thin-walled, inamyloid, a majority strongly constricted. Basidia 37—49 x 5.5-9 pm, mean 42.7 x 7.2 pm, Q: 5. 2-7. 2, mean Q: 5.95, 4-spored, clavate, thin-walled, hyaline, clamped. Cystidia absent. Hymenophoral trama regular and composed of cylindrical, ellipsoid or monilitorm elements which are hyaline, thin-walled 37-1 10 X 7-30 pm, damps present; lactifers present as highly refractive, winding, thin- walled, hyaline hyphae 1 .5-4 pm diameter. Pileipellis an ixotrichoderm up to 50 pm deep, composed of hyaline, thin-walled, clamped hyphae 2.5-4 pm diameter, apices often swollen. Stipitipellis a weak ixocutis of repent, hyaline, thin-walled, clamped, paitially gelatinised hyphae 1-3.5 pm diameter. (Fig. 5) Habitat'. Gregarious amongst deep moss in eucalypt woodland. Material examined'. Known only from the type. Remarks'. The regular hymenophoral trama composed of short elements places this species in subgenus Pseudohygrocybe M.Bon. The basidiomes of Hygrocybe saltorivula strongly resemble mature fruiting bodies of Hygrocybe cerasinomutata A.M. Young which change from cherry-red to golden orange (Young & Wood 1997); however, the lat- ter is easily separated microscopically by its very regular hymenophoral trama of tubular, aseptate elements. Hygrocybe saltorivula also approaches Hygrocybe julietae (G.Stev.) E.Horak from New Zealand but that species has a convex pileus, arcuate lamellae and Figure 5. Hygrocybe saltorivula. A habit sketch; 15 spores; C basidia. Hygrocybeae of Victoria 61 much smaller spores (6-8 x 3-3.5(-4) pm). The South American species Hygrocybe araucana Singer differs by having a convex pileus and a dry cutis (Horak 1979). Etymology. Latin, saltus - to jump; Latin, rivulus - a small stream; from the name of the type locality. Taxa with Limited Collections During this study, three collections made from the Jumping Creek Nature Walk area were considered to be new species but were found to consist of only two or three basidiomes which is considered insufficient material for a valid description and deposition of a suit- able type. These collections have provided well defined characters for two taxa. The descriptions are provided here to facilitate further studies. Hygrocybe sp. JCR1 Pileus 10-20 mm, pale orange, luteous or slightly ochre tinted yellow, strongly viscid to near glutinous, convex with a central umbo or umbonate, smooth, margins somewhat crenulate. Lamellae deeply decurrent, white or pale cream-coloured, margins concolor- ous. Stipe 30-40 x 2-4 mm, whitish to pale yellow or pale orange but paler or near white near lamellae, smooth, dry, tapered downwards. Spores 8.5-1 1(-1 1.5) x 4-6 pm, mean 9.8 x 5.1 pm, Q: 1.7-2. 5, mean Q: 1.91, ellip- soid to lacrymoid or subcylindrical, hyaline, thin-walled, inamyloid; subcylindrical spores may occasionally be slightly constricted. Basidia (28— )37— 49(— 52) x 4.5-7 pm, mean 43.3 x 5.6 pm, Q: 5.5-10.3, mean Q: 7.68, 2-spored with scattered 1-spored basidia, clamps absent. Cystidia absent. Hymenophoral trama irregular and composed of hyaline, thin-walled, cylindrical, septate, often branching hyphae 24-78(-100) x 2— 5(— 8.5) pm, clamps absent. Pileipellis a well developed, loose ixocutis or reduced ixotrichoderm composed of repent to loosely interwoven, cylindrical, hyaline, thin- walled, septate hyphae 1-5 pm diameter, clamps absent. Stipitipellis a cutis composed of repent, cylindrical, hyaline, thin-walled, septate hyphae 1-3.5 pm diameter, clamps absent. (Fig. 6) Habitat : Gregarious in moss under Kunzea ericoides. Figure 6. Hygrocybe sp. JCR1. A habit sketch; B spores; C basidia. 62 A.M. Young Material examined: Victoria. Warrandyte State Park, lO.vi. 1 995, B.A.Fuhrer 1941 (MEL 2063189): Warrandyte State Park, 10. vi. 1995, B.A.Fuhrer 1945 (MEL 2063196). Remarks : This species should be very easy to recognise in any future study since high- ly glutinous pilei in sub-genus Cuphophyllus (defined by an irregular hymenophoral trama) appear to be uncommon. Hygrocybe sp. JCR2 PUeus 10-20 mm, bright scarlet, fading a little with age, obtusely conical to conical, dry (possibly developing a slightly lubricous surface during maturation) and may appear very finely squamulose in dry conditions; margins striate, yellowish and even to sub-crenulate. Lamellae broadly adnate and often with decurrent tooth, buff pink, margins concolorous and even. Stipe 30-40 x 3-4 mm, orange-yellow with or without pink tints, cylindrical, dry, hollow, smooth but a fine pruinose, yellowish layer more or less covering the stipe surface when immature but becoming more scattered and flocculose with maturity. Spores 8.5-1 1(— 1 1.5) x 5-6(-7.5) pm, mean 9.7 x 5.7 pm, Q: (1.3— )1 .4—1 .9(— 2.0), mean Q: 1.69, ellipsoid to obovoid or sub-cylindrical, smooth, hyaline, thin-walled, inamyloid; sub-cylindrical spores may occasionally exhibit small but distinct constric- tions. Basidia 42-50 X 8-9.5 pm, mean 47.0 x 8.8 pm, Q: 4.9-6. 3, mean Q: 5.38, 4- spoied, clamped. Cheilocystidia and pleurocystidia absent. Hymenophoral trama regular and composed of a mixture of cylindrical and inflated moniliform elements which are hyaline, thin-walled, septate 30-100 x 5-27 pm, clamps present; lactifers occasionally present as highly refractive, branching, often contorted hyphae 4-8.5 pm diameter. Pileipellis a cutis of repent, hyaline, thin-walled, septate, cylindrical and somewhat inflat- ed hyphae 3—9 pm diameter, overlying a subcuticle of moniliform elements similar to those of the hymenophoral trama; lactifers similar to those of the hymenophoral trama present in the subcuticular layer. Stipitipellis a cutis of hyaline, thin-walled, repent, cylin- drical, septate hyphae, 2.5-5 pm diameter, clamps present. The cuticular hyphae exhibit some weak gelatinisation. Caulocystidia present as hyphal fascicles up to 100 pm high and composed of tangled or contorted hyphae which are cylindrical, rounded at the apices, hyaline, thin-walled, clamped and 3-5 pm in diameter. (Fig. 7) Habitat ; Gregarious in moss. Material examined : Victoria. Warrandyte State Park, 21.V.1996, B.A.Fuhrer 2043 (MEL 2063195). Remarks : The dried material exhibits a ‘varnished appearance’ on both pileus and stipe surface which is indicative of an ixocutis. Microscopic examination did not show the presence of a gluten layer but did show some gelatinisation of surface hyphae on both pileus and stipe; however, there were very few adherent spores on the stipe which is con- trary to the case if an ixocutis were developed. If an ixocutis is present in juvenile basid- iomes, it will be very weak and the surfaces will be lubricous at most. This spectacular taxon should be easily identified as there is no other Australian species that shows the fine, yellow pruinose layer over the immature stipe with residues on the mature stipe. It appears to approach the European Hygrocybe coccinea (Schaeff. : Fr.) P.Kumm. but that species has a convex pileus and does not have caulocystidia on the stipe. Acknowledgements The author wishes to thank Mr Bruce Fuhrer of Ringwood. Melbourne for his consider- able generosity in not only allowing access to his Jumping Creek collections but also in providing superb field photographs of the material. His help and willing assistance dur- ing the examination of these collections, both with additional field notes and descriptions of the park flora, is also very gratefully acknowledged. This investigation was completed Hygrocybeae of Victoria 63 under a grant provided by the Australian Biological Resources Study for research into the species of the Hygrophoraceae of south-eastern Australia. References Anon. (1997). ‘Jumping Creek Nature Walk.’ Parks Victoria, Melbourne. Arnolds, E. (1990). Tricholmataceae pp. in Bas, C„ Kuyper, Th.W., Noordeloos, M.E. & Vellinga, E.C. (Eds.) Flora Agaricina Neerlandica 2, 1-137. Boertmann, D. (1995). The genus Hygrocybe. Fungi of Northern Europe 1, 1-184. Cleland, J.B. & Cheel, E.C. (1919). Australian Fungi: Notes and Descriptions. No. 3. Transactions & Proceedings of the Royal Society of South Australia 43 , 262-315. Fuhrer. B.A. & Robinson. R. (1992). ‘Rainforest Fungi of Tasmania and South-East Australia.’ (CSIRO: Melbourne.) Grgurinovic, C.A. (1997). ‘Larger Fungi of South Australia.’ (Botanic Gardens of Adelaide and " State Herbarium and the Flora and Fauna of South Australia Handbooks Committee: Adelaide.) Hesler, L..R. & Smith, A.H. (1963). ‘North American Species of Flygrophorus.’ (University of Tennessee Press: Knoxville.) Hongo, T. (1967). Notes on the Japanese larger fungi 19. Journal of Japanese Botany 42 . 151-159. Horak, E. (1979). Fungi, Basidiomycetes, Agaricales y Gasteromycetes secotioides. Flora Criptogamica de Tierra del Fuego 11 , 1-524. Horak, E. (1990). Monograph of the New Zealand Hygrophoraceae (Agaricales). New Zealand Journal of Botany 28 . 255-309. Imazeki, R„ Otani, Y. & Hongo, T. (1988). ‘Fungi of Japan.’ (Yama-kei Publishers Co., Ltd.: Tokyo.) 64 A.M. Young Pegler, D.N. (1986). Agaric Flora of Sri Lanka. Kew Bulletin Additional Series 12 . 1-519. Willis. J.H. ( 1963). 'Victorian Toadstools and Mushrooms.’ 3rd edn. (The Field Naturalists Club of Victoria: Blackburn.) Young, A.M. (1999). The Hygrocybeae (Fungi, Basidiomycota, Agaricales, Hygrophoraceae) of the Lane Cove Bushland Park. New South Wales. Austrobaileya 5, 535-564. Young, A.M. (2000). Additions to the Hygrophoraceae of South Eastern Australia. Muelleria 12 . 3-36. Young, A.M. & Wood, A.E. (1997). Studies on the Hygrophoraceae (Fungi, Homobasidiomycetes, Agaricales) of Australia. Australian Systematic Botany 10 . 91 1-1030. Muelleria 14:65 (2000) A Revision of Agrostis billardierei R. Hr. (Poaceae) A. ./. Brown * and N. G. Walsh** *State Chemistry Laboratory, Agriculture Victoria, Cnr. Sneydes and South Roads, Werribee, 3030 **Royal Botanic Gardens Melbourne, Birdwood Avenue, South Yarra, 3141 Abstract Observations of growth habit and habitats and statistical analysis of detailed morpholog- ical measurements on specimens of Agrostis billardierei R. Br. and some closely related taxa has resulted in a re-evaluation of the rank and status of some of them. New combi- nations are made for Agrostis billardierei var. robusta Vickery, A. billardierei var. colli- cola D. Morris, A. billardierei var. filifolia Vickery and A. aemula var. setifolia Vickery. The new names for these taxa are A. robusta, A. collicola, A. punicea war. filifolia and A. punicea var. punicea respectively. Agrostis billardierei var. tenuiseta D. Morris is con- firmed as a good taxon. Introduction As a result of recent extensive surveys and collections of Agrostis in Western Victoria by the senior author, the status of some published taxa was in obvious need of reassessment. In particular, the morphological and ecological relationships between A. billardierei var. fdifolia Vickery and A. aemula var. setifolia Vickery appeared to be a lot closer than implied by their taxonomic status. In contrast, a number of the recognised varieties of A. billardierei R. Br. appeared to be more distinct than their status would indicate. For exam- ple, James and Brown (2000) found highly significant differences in both morphological assessment and DNA analysis, when comparing single populations of A. billardierei var. robusta and A. billardierei var. fil folia. Vickery (1941) carried out the last major revision of the Australian species of Agrostis. Working with a limited number of specimens, she recognised three varieties of A. billardierei , viz. A. billardierei var. billardierei, from coastal New South Wales, Victoria, Tasmania and South Australia (20 specimens recorded), A. billardierei var. fdi- folia, from inland Victoria and Tasmania (2 specimens recorded) and A. billardierei var. robusta Vickery, also from inland Victoria and Tasmania (2 specimens recorded). Since 1941, many other specimens of A. billardierei have been collected in south-eastern Australia. Black (1943) added A. billardierei vox. fdifolia (2 specimens recorded) and Jessop and Toelken (1986) added A. billardierei var. robusta (2 specimens recorded) to the South Australia flora. Morris (1990) recognised additional varieties of A. billardierei from Tasmania, viz. A. billardierei var. collicola D. Morris (2 specimens recorded) and A. billardierei var. tenuiseta D. Morris (5 specimens recorded). Agrostis billardierei var. robusta was separated from var. billardierei on the basis of its taller, rather rigid habit, non-littoral habitat, narrow, inrolled, acuminate leaves, more conspicuously scaberulous lemmas and smooth to scaberulous-sided glumes with very scabrous keels (Vickery 1941). Agrostis billardierei var. filifolia was separated from the typical variety on the basis of its more slender and rather taller habit, non-littoral habitat, very narrow, almost filiform leaves, rather few-flowered and more capillary panicles, long anthers and smooth-sided glumes (Vickery 1941). Agrostis billardierei var. collico- la and var. tenuiseta were separated from var. billardierei on the basis of their slender, straight, sub-terminal awns (or awns sometimes absent in the case of the latter) and from each other on the basis of their differing habitat and stature (the former growing in non- littoral highlands and being smaller with few spikelets) (Morris 1990). 66 A.J. Brown and N.G. Walsh Vickery ( 1941 ) also described Agrostis oemula var. setifolia based on 3 specimens from Tasmania. Walsh and Entwisle (1994) recorded this taxon for Victoria. Agrostis aemula var. setifolia was separated from var. aemula on the basis of its very narrow, setaceous or sub- filiform leaves, more lax panicle, longer lemmas, awn insertion much below the middle, and long anthers (Vickery 1941). Vickery also noted: “It shows a very close superficial resem- blance to 4 . billardierei vax.fdifolia, but differs in the very pubescent lemma.” Taxonomic keys appearing in current Floras (e.g. Jessop & Toelken 1986, Simon 1993, Walsh & Entwisle 1994, Curtis & Morris 1994) largely follow Vickery’s key (Vickery 1941). Using morphological characterisation, this paper presents a statistical assessment of the relationships between these taxa and an enhanced description of each taxon. In addi- tion, geographic distribution and environmental niches are noted. Methods MORPHOLOGY Field collections of Agrostis specimens were made throughout southern Victoria over the growing seasons of 1993/94 to 1998/99 and measurement of a range of morphological characters were made (Appendix 1 ) for each specimen. The majority of the collection sites were located in south-west Victoria but some specimens were collected from Gippsland (eastern Victoria). Additional collections were made of A. billardierei var. tenuiseta from the east coast of Tasmania, where it is endemic. As A. billardierei var. col- licola is confined to a few Tasmanian highland localities (Morris 1990) and fresh field collections were not readily available, measurements were made on the few extant herbarium collections. Total height was measured from the base of the above ground plant parts to the high- est part of the plant (almost always, the inflorescence) while basal tussock height was a more subjective measure of the height of the major leafy part of the tussock. C ulm height was taken from the plant base to the base of the highest flag leaf. Basal leaf width meas- urements were of the widest non-senescent leaf and flag leaf and inflorescence measure - ments related to the largest inflorescence in each specimen. The width of conduplicate or involute leaves was measured without flattening these leaves out (in accord with pub- lished descriptions and keys). Leaf roll was assessed tor the majority of green leaves. Inrolling, or conversely, flattening sometimes occurred where leaves had senesced, but these were not used for leaf width measurements. Where an inflorescence was not exsert- ed from the leaf sheath, the length of the visible peduncle was recorded as zero. The height of the inflorescence was taken from the base of the flag leaf or from the lowest panicle whorl if exserted. Spikelet measurements were made under a binocular micro- scope (20x mag.) on one ‘typical’ spikelet selected from each specimen. Occasionally, where an anther was found to be missing, an anther from another spikelet of similar age and size on the same specimen was measured. Pedicel and spikelet colour was assessed as the degree of purpling (assumed to result from the accumulation of anthocyanin). The lemma and palea of some specimens have apical nerve extensions oi teeth ol considerable length (up to 1.5 mm and 1.2 mm respectively). All such projections are termed ‘setae’ here. Specimens were also classed according to plant age. This criterion was assessed on the maturity of the measured inflorescence, and divided the specimens into ‘immature (only partly exserted and non-spreading with non-senesced spikelets), ‘maturing (becoming exserted and/or spread with non-senesced spikelets), ‘mature’ (fully spread with terminal spikelets senesced) and ‘ripe (most of panicle senesced). Agrostis billardierei and A. aemula and their varieties, were determined with the use of Vickery’s taxonomic key (Vickery 1941 ) and Morris’ descriptions (Morris 1990). Only the published key characters and not the full set of measured characters were used tor Revision of Agrostis billardierei 67 determination. Vickery’s key character for growth habit was poorly specified (i.e. ‘tall and robust’ for A. billardierei var. robusta and ‘slender’ for A. billardierei var. jilifolia) and was consequently excluded from consideration. The key character of habitat (i.e. lit- toral or non-littoral) was also excluded. On examination of the data set, it was found that the key maximum limit for spikelet length of A. aemula was often exceeded and so only the key minimum limit was used for determination. Where specimens did not meet all accepted key criteria for a particular taxa, they were listed as indeterminate. STATISTICAL TREATMENT A total of 243 specimens were subjected to statistical analysis. Morphological data for analysis were either quantitative (continuous variable) or qualitative (discrete variable). Thirty six characters were assessed (including 12 discrete variables) and a further 14 characters were derived as ratios (e.g. inflorescence height:width), percentages (e.g. basal tussock height as a percent of total plant height) or sums (e.g. lemma body + lemma setae = total lemma length) from these. Qualitative data were transformed to an ordinal scale and treated as quantitative for some analyses. Ordinal scales varied from 3 to 5 levels, with an average of 3.8 (Appendix 1). A selected set of 33 characters was used for multivariate analysis, including 10 of the qualitative characters (Appendix 1). None of the derived characters were used. As multi- variate analysis cannot easily handle missing data, characters with missing data (number of visible branches in lowest inflorescence whorl, pedicel colour and spikelet colour) were also excluded. All statistics were performed using the GENSTAT 5 package (Release 3.1, Lawes Agricultural Trust, Rothamstead Experimental Station). A Symmetric Matrix (SYMMETRICMATRIX) of associations for the selected data set (using euclidean or cityblock variate type; the former for continuous variables and the latter for discrete variables), was established by forming a Similarity Matrix (FS1MI- LARITY). This matrix was then subjected to Principle Coordinates Analysis (PCO) to assess natural groupings of the specimens and to check the validity of the current flora key determinations. Separate PCO analysis was carried out on i) the above data set with- out exclusions and ii) the same data set but excluding ‘immature’ and ‘ripe’ specimens. Discriminant (DISCRIMINANT) analysis was undertaken on the selected data sets to estimate mis-determination rates for specimens and to assign indeterminate specimens to groups. Separate discriminant analysis was carried out on: i) the selected data set without exclusions; ii) the selected data set but excluding lemma surface characters (i.e. B, Rc, Rd); and iii) the selected data set but excluding the varieties of A. aemula. Correlation matrices of variates and discriminant scores enabled the identification of those variates (i.e. characters) that were most influential in separating groups (i.e. taxa). A Reduced Similarity Matrix (REDUCE) was constructed on the total (determined and reassigned) specimen data set of selected characters, using the mean similarity between groups. Hierarchical Cluster Analysis (HCLUSTER) (using the furthest neigh- bour criterion) of the Reduced Similarity Matrix, provided an association of the taxa for the purpose of assessing specific and varietal status. Furthest neighbour linkage was used to maximise space-dilation and minimise overlap between groups. Ranges, means and significant differences for all measured characters (Appendix 1 ) were assessed by using Analysis of Variance (ANOVA) and provided for: i) identification of diagnostic characters to separate particular taxa; and ii) an enhanced description of each taxon. Least significant differences (LSD) at the 5%, 1% and 0. 1 % level were used to test differences in means between any two taxa. Additional ANOVA was performed for each separate taxon to test the effect of plant age on the measured characters. Although it is recognised that this analysis was of age classes that contained specimens from varying and often different populations, it can still provide some useful overall information for each taxon. 68 A.J. Brown and N.G. Walsh EXAMINATION OF HERBARIUM SPECIMENS Approximately 280 herbarium specimens (including type specimens and 33 duplicates) from MEL. AD. HO, NSW, CANB, K. BM. WELT and CHR were examined to check conformity of: a) survey specimens to type specimens and b) other herbarium specimens to survey specimens. Measurements were not routinely undertaken on these specimens, but only for some characters where individuals required re-appraisal or confirmation of the determination. These data were not included in the statistical analysis. HABITAT MEASUREMENTS Observations were made of the habitats occupied by each taxon for the field collected specimens. Composite soil samples (0-10 cm depth) were taken from a selection of sites and analysed for moisture content, pH (1:5 soihwater) and electrical conductivity (1:5 soikwater). Sites were selected to include Agrostis populations with substantial plant numbers and to cover a wide geographic range in south-west Victoria. Fifteen plants at each ot two populations ot A. billardierei var. robusta and two populations of var ■ filifo- lia were tagged for observation of growth habit, flowering and longevity over the 1998/99 and 1999/00 growing seasons. In addition, the position of four tussocks of var. bil- lardierei at each of two sites were noted in December 1997 or 1998 and observed again two years later. Flowering time during the 1998/99 season for this taxon was noted for a further four sites. Results TAXA DETERMINATION Thirty eight specimens were classed as indeterminate on the basis of their failure to con- form completely to Vickery's key (Vickery 1941). In the majority of cases, non-con- formity was in only one character of the several provided by the key. PCO analysis ot the total data set. provided a clear separation of all taxa, except for A. billardierei var. tenuiseta and A. billardierei var. collicola which grouped with A. bil- lardierei var. billardierei (Fig. 1). The first and second vectors of the coordination accounted for 41.0% of the variation detected, with the third, fourth and fifth vectors accounting for a further 22.5%. Restriction of PCO analysis to ‘maturing’ and ‘mature’ aged specimens (i.e. exclusion of ‘immature’ and ‘ripe’ panicles) made little difference to the outcome. For the 205 determined specimens used in this study, mis-classification rate, as cal- culated through the discriminant analysis of data without exclusions, was zero and all taxa displayed clear separation (Fig. 2), including var. collicola from the type of A. bil- lardierei by the first and third vectors (the latter not shown) and var. tenuiseta from the type by the second and third vectors. The majority of reassigned (previously indetermi- nate) specimens fitted well to the taxa assigned by the analysis. Specimens conforming to A. billardierei (Vickery 1941, Morris 1990) numbered 152, with a further 20 indeterminate specimens being reassigned to var. robusta (12 speci- mens) or var .filifolia (8 specimens). Indetermination for var. robusta was caused by a less scabrid glume keel than expected or an excessive anther length, while in var .filifolia, insufficient anther length was the main criterion. In total, specimens of A. billardierei were determined as: 26 of var. billardierei (from 14 near-coastal and 2 inland sites); 75 of var. robusta (from 39 inland and 8 near-coastal sites); 52 of var .filifolia (from 17 inland sites); 1 1 of var. tenuiseta (from 4 coastal sites); and 8 of var. collicola (made up of 2, 2 and 4 specimens each from three separate herbarium collections from highland sites). Two reassigned specimens of var .filifolia from Meerlieu in central Gippsland were slightly separated from the rest of the taxon due to their smaller than average spikelets. This population was the only collection from Gippsland and may indicate a local form. Revision of Agrostis billardierei 69 0.4 -0.4 • • -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 PCO score 2 Figure 1. PCO score 1 and 2 for the set of selected morphological characters: O Agrostis aemula var. aemula. A A. aemula var setifolia. ♦ A. billardierei var. billardierei. A A. billardierei var. tenuiseta. □ A. billardierei var. robus- ta. A A. billardierei var. filifolia. • A. billardierei var. collicola , + indeter- minate. -15 . -20 -15 -10 -5 0 5 10 j Discriminant score 2 Figure 2. Discriminant scores 1 and 2 with indeterminates reassigned, for the set of selected morphological characters: O Agrostis aemula var. aemula. A A. aemula var setifolia. ♦ A. billardierei var. billardierei. A A. billardierei var. tenuiseta. □ A. billardierei var. robus- ta. A A. billardierei var. filifolia. • A. billardierei var. collicola. There were 53 specimens conforming to A. aemula (Vickery 1941). As A. aemula var. aemula appears to form intergrades with A. avenacea Gmel. (as noted by several authors e.g. Vickery 1941, Walsh & Entwisle 1994), only specimens with clear characteristics for A. aemula var. aemula (e.g. broad leaves and large spikelets) were used in this study to provide a comparison with A. aemula var. setifolia. A further 1 8 indeterminate specimens were reassigned to var. aemula (11 specimens) and var. setifolia (7 specimens). Indetermination was caused by excessive lemma length for var. aemula, and insufficient lemma length and anther length for var. setifolia. After reassignment, a total of 3 1 speci- mens of A. aemula var. aemula (from 16 inland sites) and 40 specimens of var. setifolia (from 17 inland and 1 near-coastal sites) were identified. Discriminant analysis separat- ed var. setifolia into two, near but distinct, populations. On closer examination this was found to be due to the level of hairiness of the lemma back (i.e. level 3 or 4; see Appendix 1 for descriptions). There was no differentiation in the geographic range of these specimens. The first and second vectors of the discriminant analysis accounted for 81.2% of the variation. The first vector was highly correlated with the characters for lemma surface (B. Rc, Rd at 0.97, -0.80 and -0.76 respectively) and therefore provided considerable separa- tion of the current varieties of A. aemula and A. billardierei. Other characters with high correlation to the first vector were glume and awn length (Gib. Gub, Ab and Ac at 0.73, 0.67, 0.68 and 0.58 respectively), lemma setae length (Ls at 0.57) and glume margin cil- iation (M at -0.55). The second vector was associated with leaf width and leaf roll (Llw, Lfw and Lr at -0.83, -0.74 and 0.80 respectively), awn attachment (Aa at -0.81), palea setae and body length (Ps and Pb at 0.72 and 0.54 respectively), density of lemma back scabridity (Rd at 0.58) and awn column length (Ac at 0.50). Discriminant analysis of the data after exclusion of lemma surface characters (i.e. hairiness or scabridity; B. Rc and Rd) brought A. aemula var. setifolia and A. billardierei var. filifolia together into the same grouping (indicating that only lemma surface charac- 70 A.J. Brown and N.G. Walsh Figure 3. Discriminant scores 1 and 2 with indeterminates reassigned, for the set of selected morphological characters with B, Rc and Rd exluded: O Agrostis aemula var. aemula. A A. aemula var setifolia. ♦ A. billardierei var. billardierei. ❖ A. billardierei var. tenuiseta. □ A. billardierei var. robusta. A A. billardierei var. filifolia. • A. bil- lardierei var. collicola Figure 4. Discriminant scores 1 and 2 with indeterminates reassigned, for the set of selected morphological char- acters with the varieties of Agrostis aemula excluded: ♦ A. billardierei var. billardierei. 4- A. billardierei var. tenuiseta. □ A. billardierei var. robusta. A A. billardierei var. .filifo- lia. • A. billardierei var. collicola ter differentiates them) and resulted in a slight overlap of A. aentula var. aemula and A. billardierei var. billardierei in the first and second vectors (Fig. 3). Miscalculation rate was 4.4%, resulting from reassignment of four A. billardierei var. filifolia specimens to A. aemula var. setifolia and reassignment of five specimens from the latter to the former. There was no reassignment from A. aemula var. aemula to A. billardierei var. billardierei. The first and second vectors accounted for 85.4% of the variation. Separation by the Inst vector was largely influenced by awn attachment and length, palea setae length, leaf roll and width and anther length. The second vector was influenced by leaf roll, glume length, coarseness of lemma scabrids, glume margin ciliation, panicle stiffness and awn length. Discriminant analysis of the varieties of A. billardierei alone, provided clear separa- tion of all taxa (Fig. 4). Miscalculation rate was zero and the first and second vectors accounted for 80.4% of the variation. Separation by the first vector was mainly due to awn length and attachment, leaf roll and width, density of lemma scabrids and lemma and palea setae lengths while the second vector was influenced by coarseness of lemma scabrids, glume margin ciliation, anther length and panicle stiffness. TAXA ASSOCIATION A dendrogram for the cluster analysis of the reduced similarity matrix is shown (Fig. 5). A. aemula var. aemula separates from A. billardierei var. billardierei at 75% similarity while A aemula var. setifolia. A. billardierei var. collicola and A. billardierei var. robus- ta separate from their type varieties at 79%, 78% and 81% respectively. Agrostis bil- lardierei var. filifolia is clustered with A. aemula var. aemula rather than with A. bil- lardierei var. billardierei and has 87% similarity to A. aemula var. setifolia. A. billardierei var. tenuiseta has 88% similarity to A. billardierei var. billardierei. Revision of Agrostis billardierei 71 Percent Similarity 88 87 86 85 84 83 82 81 80 ' 79 78 77 76 75 74 ‘aemula ’ ‘filifolia ’ ‘setifolia ’ ‘billardierei ’ ‘tenuiseta ’ ‘robust a ' ‘collicola ’ Figure 5. Dendrogram of hierarchical cluster analysis (furthest-neighbour criteria) for a Reduced Similarity Matrix (taxa represented by varietal epithet only). KEY CHARACTERS The results of ANOVA analysis based on currently recognised taxa are shown in Table 2. Except for inflorescence extension (Ie), inflorescence height to width ratio (Ih:Iw) and visible peduncle length (Ipd), significant differences were found for all measured char- acters at P < 0.001. Table 1 summarises significant differences for LSD at the 0.1%, 1% and 5% levels for a number of taxa comparisons. Although this analysis has identified numerous characters that are significantly different, most of these cannot be used as distinctive diagnostic characters for separating taxa because the range in their values overlap. However, a num- ber of diagnostic or nearly distinct characters can be identified for each taxon. Agrostis aemula and A. billardierei Fifty four percent of the measured or derived characters were significantly different between A. aemula var. aemula and A. billardierei var. billardierei (forty percent being highly significant for LSD at 0.1%). Eighty five percent of these characters related to spikelet characteristics and included glume, lemma and awn size (all larger in A. aemu- la 1 ). Agrostis aemula also had longer setae points to its glumes and lemmas, smoother glume margins and more uneven glume lengths (the lower being slightly longer). Tussock height was shorter for A. aemula, leaf width was narrower and leaf length was shorter. Although most character measurements overlap between these species, the generally narrower leaves, larger glumes and awns and the hairy lemmas of A. aemula var. aemu- la serve to distinguish it from A. billardierei var. billardierei. Although the mean inflo- rescence height to width ratio was the same for these species, it did not exceed a value of two for A. billardierei but was up to five for A. aemula. Agrostis billardierei var. tenuiseta Thirty six percent of characters separated this taxon from var. billardierei (twenty six per- cent for LSD at 0.1%). It was diagnostically separate by its short awns, which had a minute or absent column and a near terminal lemma attachment. The leaves of this taxon tended to be more loosely inrolled (and therefore narrower), the Rag leaves were shorter (some had withered tips), the spikelets less gaping and the 1 It is recognised that forms of A. aemula var. aemula with similar or smaller spikelets than A. billardierei var. billardierei are currently recognised within the former taxon but were not exam- ined as part of the current study. 72 A.J. Brown and N.G. Walsh Table 1. Summary of significant differences (*** = LSD at 0.1%, ** = LSD at 1%, * = LSD at 5%) between various comparisons of currently recognised taxa (represented by varietal epithet only). Character ‘ billardierei ’ 'billardierei' 'billardierei’ ‘ billardierei ' 'billardierei' ‘ aemula ' ‘ setifolia ' to ‘ aemula' to ‘ tenuiseta ’ to ‘ collicola ' to ‘robusta’ to ‘filifolia to 'setifolia' to 'filifolia' Hgt, cm *** Llh, cm *** *** * *** Llhp, % *** * ** *** Ch, cm *** Chp, % * Lr (1-5) *** *** *** *** *** Llw, mm *** *** *** *** *** *** Lfw, mm *** *** *** *** Lfh, cm ** ** *** ' * * * * Lig, mm *** *** *** ** Ie (1-4) Ipd, cm ** * lib (no.) * *** *** Ih. cm ** *** Iw, cm *** Ih:Iw la (1-4) *** Ipc (1-5) * * Sa (1-3) *** * *** Sc (1-5) *** Sg (1-4) *** ** * Git, mm *** *** ** * *** Gib. mm *** *** *** *** Gls, mm *** * *** ** Glsp. % *** *** * Gut, mm *** *** ** *** Gub, mm *** *** *** *** Gus, mm * * * * ** K (0-3) *** M (0-2) *** *** *** *** Glt:Gut *** ** *** Lt, mm *** *** * * Lb, mm ** *** *** Ls, mm *** *** ** ** Lsp, % *** *** * * * B (0-4) *** *** *** Re (0-3) *** *** *** *** *** Rd (0-3) *** * *** *** ** *** *** Pt. mm * *** *** *** Pb, mm * * * *** *** *** Ps. mm ** *** *** Psp, % ** *** *** Re, mm * * At, mm *** *** *** ** *** ** Ac, mm *** *** *** *** *** * ** Ab. mm *** *** *** *** *** * Abp. % ** *** *** * *** Aa. mm * *** *** *** Aap, % *** *** ** *** *** Revision of Agrostis billardierei 73 Table 2. Means for morphological characters of currently recognised taxa (represented by varietal epithet only); values in the same row that share the same letter are not significantly different (assessed by LSD at 0.1% level). Character billardierei tenuiseta collicola robusta filifolia setifolia aemula LSD (0.1%) Hat, cm 49 b 44 b 14 a 48 b 44 b 41 b 45 b 15 Llh. cm 33 e 29 de 8 a 21 ede 19 be 15 ab 20 bed 9.6 Llhp, % 68 c 10 c 59 be 55 be 42 ab 38 a 45 ab 16 Ch, cm 24 be 24 be la 28 c 19 be 17 b 20 be 9.5 Chp, % 48 abc 55 be 49 abc 57 c 43 a 41 a 44 ab 12 Lr (1-5) 1.0 a 1.7 be 2.1 c 4.8 e 3.9 d 3.9 d 1.2 ab 0.61 Llw. mm 4.8 d 4.0 c 1.6 b 0.6 a 0.5 a 0.5 a 3.3 c 0.75 Lfw, mm 4.1 c 2.6 b 1.1 a 0.7 a 0.4 a 0.3 a 2.6 b 0.82 Llh, cm 11.1 d 6.6 abed 3.0 a 9.1 cd 4.7 abc 3.8 ab 6.9 abed 4.7 Lig, mm 7.0 d 6.0 cd 2.6 a 4.1 abc 4.9 be 3.8 ab 5.8 bed 2.1 le (1-4) 3.5 a 3.0 a 2.9 a 3.2 a 3.3 a 3.6 a 3.5 a 1.1 Ipd. cm 6.3 a 4.1 a 1.1 a 1.9 a 6.1 a 1.2 a 6.5 a 6.4 lib (no.) 4.8 cd 5.2 d 3.3 abc 4.5 bed 3.0 ab 2.5 a 5.2 d 1.5 Ih, cm 19 b 15 b 6 a 18 b 19 b 17 b 19 b 5.0 Iw, cm 18 b 13 b 5 a 18 b 18 b 17 b 16 b 8.2 Ih:lw 1.11 a 1.13 a 1.73 a 1 .30 a 1.72 a 1.36 a 1 .36 a 1.9 la (1-4) 1 .2 ab 1.0 a 2.0 b 3.1 c 1 .5 ab 1.5 ab 1 .7 ab 0.88 Ipc (1-5) 3.8 ab 3.0 a 4.4 ab 2.7 a 5.0 b 4.9 b 4.2 ab 1.8 Sa (1-3) 1 .6 ab 1 .6 ab 2.9 c 1.5 a 2.2 be 2.6 c 1.3 a 0.81 Sc (1-5) 2.1 a 2.1 a 3.9 ab 2.4 a 4.9 b 4. 1 ab 3.8 ab 2.0 Sg (1-4) 2.7 be 1 .6 a 3.6 c 3.0 be 3.3 be 3.1 be 2.5 ab 0.97 Git, mm 5.1 be 5.2 be 3.6 a 4.5 b 5.6 cd 6.0 cl 7.1 e 0.75 Gib, mm 5.1 c 5.2 cd 3.4 a 4.2 b 5.5 cd 5.9 d 6.9 e 0.73 Gls, mm 0.07 ab 0.00 a 0.20 b 0.33 d 0. 1 1 abc 0. 1 1 abc 0.28 cd 0.18 Glsp, % 1 ab 0 a 6 cd 1 d 2 ab 2 ab 4 be 3.3 Gut, mm 4.8 be 4.8 be 3.5 a 4.3 b 4.9 be 5.2 c 6.1 d 0.66 Gub, mm 4.1 b 4.8 b 3.4 a 4.0 a 4.8 b 5.1 b 5.9 c 0.66 Gus, mm 0.07 ab 0.04 a 0. 16 ab 0.28 c 0.08 ab 0.09 ab 0.21 b 0.15 K (0-3) 1.9 b 2.2 b 2.8 a 1.9 b 1.7 b 1.9 b 1.9 b 0.69 M (0-2) 0.9 b 1.5 c 0.6 ab 1.6 c 0.1 a 0.3 ab 0.1 a 0.64 Glt:Gut 1 .07 ab 1.08 abc LOO a 1 .06 a 1.14 cd 1.15 cd 1.17 d 0.069 Lt, mm 3.2 a 3.2 a 2.8 a 3.5 ab 4.0 be 4.4 c 3.9 be 0.64 Lb. mm 2.1 a 3. 1 abc 2.1 a 2.9 ab 3.2 be 3.4 c 2.9 ab 0.45 Ls, mm 0.50 be 0.08 a 0.32 ah 0.59 be 0.76 cd 1.01 d 1.04 d 0.31 Lsp, % 15 be 3 a 11 b 17 cd 19 cd 23 de 26 e 6.0 B (0-4) 0.0 a 0.0 a 0.0 a 0.0 a 0.0 a 3.8 b 4.0 c 0.22 Rc (0-3) 2.0 d 1.9 d 0.4 b 3.0e 1.1 c 1.0 c 0.0 a 0.26 Rd (0-3) 1.4 b 1.1 b 0.3 a 2.7 c 3.0 c 1.0 b 0.0 a 0.49 Pt. mm 2.2 a 2.4 ab 2.5 ab 2.7 b 3.4 c 3.5 c 2.5 ab 0.43 Pb, mm 2.1 a 2.4 ab 2.4 ab 2.5 be 2.8 cd 3.0 d 2.4 ab 0.39 Ps, mm 0.07 ab 0.02 a 0.02 a 0.19 b 0.57 c 0.57 c 0.08 ab 0.15 Psp, % 3 ab 1 a 1 a 1 b 17 c 16c 3 ab 4.3 Re, mm 2.3 ab 2.6 be 2.1 a 2.3 abc 2.6 be 2.8 c 2.6 be 0.45 At. mm 1.0 c 0.7 a 2.3 b 5.8 c 8.9 d 10.0 de 10.7 e 1.3 Ac, mm 2.1 c 0.0 a 0.0 a 2.1 b 3.5 d 3.9 d 3.6 d 0.52 Ab, mm 4.3 c 0.1 a 23 b 3.8 c 5.4 d 6.0 d 7.1 e 1.0 Abp, % 62 ab 100 c 100 c 65 ab 60 a 60 a 66 b 4.7 Aa, mm 1.2 b 2.8 c 2.6 c 1.2 b 0.6 a 0.6 a 1.4 b 0.27 Aap, % 39 b 88 c 89 c 33 b 15 a 14 a 36 b 6.0 A, mm 0.74 ab 0.66 a 0.68 a 0.62 a 1.28 c 1.28 c 0.92 b 0.24 74 A.J. Brown and N.G. Walsh inflorescences shorter (many larger inflorescences had disarticulated at the time of col- lection). These differences are likely to be due to most of the examined specimens being mature to very mature (see discussion for Age Effects). The outer glumes had more cili- ated margins than for var. billardierei and while lemma setae were shorter, lemma bodies were longer on average. Agrostis billardierei var. collicola Sixty eight percent of characters were significantly different between this taxon and var. billardierei (with fifty percent for LSD at 0.1%). Except for the culm height to total height percent (Chp), all the vegetative characters were significantly smaller; total plant, tussock and culm heights and leaf widths being diagnostically distinct. Inflorescences and spikelets were also significantly smaller than var. billardierei, although the largest examples of var. collicola did exceed the smallest examples of the type variety. Despite the smaller spikelets, the lemmas (and paleas) of var. collicola displayed a similar size range to var. billardierei ; mean lower glume: lemma length (including setae) being 1.20 for the former and 1.62 for the latter. Apart from the diagnostically distinct, short awns and near terminal awn attachment, var. collicola generally possessed glumes with dense, finely scabrid surfaces (extending from keels to near margins) and lemmas with almost completely smooth surfaces. The glumes of var. billardierei and var. tenuiseta (apart from the keels) were generally smooth or with occasional scattered scabrid projections and the lemmas normally had at least some microscopic scabrid projections near the central nerve. The awns of var. collicola were significantly longer than for var. tenuiseta and tend to be decurrent from the lemma back, rather than having a column (even a minute one, as is often the case in var. tenuiseta). Agrostis billardierei var. robusta Fifty six percent of characters were significantly different between this taxa and var. bil- lardierei (thirty four percent being for LSD at 0.1%). As well as the diagnostic characters of narrower and inrolled leaves and distally densely coarse scaberulous lemmas, var. robusta was significantly different from var. billardierei in its shorter ligules, less stiff inforescences, shorter glumes and awns, longer paleas, longer glume and palea setae and glumes with greater ciliation of margins. However, the currently used diagnostic charac- ter of more scabrous keels of the outer glumes was not distinct between the two taxa. Contrary to previous descriptions (Vickery 1941, Walsh & Entwisle 1994) there was lit- tle evidence of scabrid projections on the sides of the glumes, except very occasionally and then distally scattered. Examination of the holotype agreed with this finding. There were no obvious character differences between the coastal and inland specimens for either taxon. Besides the generally distinct scabridity of the distal surfaces of the lemmas and the narrow leaf width and inrolling of var. robusta, a few other characters could be useful in identifying this taxon. For example, ligule length could be a guide. Sixty nine percent of var. billardierei had ligules greater than 5 mm (with 53% greater than 6 mm) compared to only 13% of var. robusta (with none greater than 6 mm). Another useful character is the length of the outer glume setae expressed as a percent of the total glume length (Glsp). Seventy three percent of var. robusta had Glsp greater than 5% compared to none for var. billardierei. Agrostis billardierei var. fdifolia This taxon was separated from var. billardierei in sixty four percent of measured charac- ters (forty eight percent being for LSD at 0.1%). The main diagnostic character which separated var. filifolia from var. billardierei and var. robusta was its near-basal awn attachment (8 to 23% of lemma length with a mean of 15%) compared to mid-back attachment (29 to 47% with a mean of 39% for var. billardierei and 23 to 45% with a Revision of Agrostis billardierei 75 mean of 33% for var. robusta). The taxon was also clearly separated from var. billardierei by its narrow leaves and longer palea setae (0.3 to 1 .2 mm compared to 0.0 to 0.2 mm). Although the ranges in palea setae length overlapped between var .filifolia and var. robus- ta, only 4% of var. robusta specimens exceeded 0.3 mm compared to 88% of war. filifo- lia specimens. The lemma surface often served to distinguish these three taxa. While the lemma of var. filifolia was uniformly minutely granular-papillose to scaberulous, the lemma of var. billardierei was sparsely and minutely granular to scaberulous on the keel and distal part of the lemma, and var. robusta was generally strongly and densely scaberu- lous in the distal part of the lemma extending down the keel (although this character was occasionally found to be less obvious in very mature or stunted specimens). In both vars. billardierei and robusta , the lemma was almost completely devoid of scabrosities on the proximal lateral surfaces. Other significantly different but non-diagnostic characters separating var. filifolia from var. billardierei were its shorter basal tussock, shorter flag leaves, shorter ligules, less branched inflorescence, more purple spikelets, larger lemmas, paleas, awns and anthers and less ciliated glume margins. In addition to its diagnostically distinct characters, var. filifolia differed significantly from var. robusta in having shorter culms, shorter flag leaves, less branched, stiffer, less crowded and more purple inflorescences and larger glumes, lemmas, paleas, awns and anthers. Setae points on glumes, lemmas and paleas were longer but glume margins were less ciliated. Although both taxa had narrow leaves, var .filifolia leaves were either flat, conduplicate or pseudo-convolute, in contrast to the strongly involute to convolute leaves of var. robusta. Agrostis aemula var. setifolia This taxon was diagnostically separable from var. aemula in its longer palea setae (0.3 to 0.9 mm compared to 0.0 to 0.3 mm) and lower awn attachment (8 to 26% compared to 25 to 46%). It was also distinct, on the basis of leaf width (0.2 to 1.0 mm compared to 2.0 to 4.5 mm). The taxon also differed from A. aemula var. aemula in its significantly shorter ligules (Lig), lesser branched panicles, less clustered spikelets, shorter glumes 2 , glume setae and awn bristles and longer lemmas, paleas and anthers. The lemmas were also slightly less hairy on average. Apart from its hairy lemmas, A. aemula var. setifolia differed statistically from A. bil- lardierei var. filifolia only in a few floret characters, including slightly longer lemma setae and awns. AGE EFFECTS Table 3 summaiises the results of ANOVA analysis for each taxon according to age class (see Methods for descriptions). The effects of plant age (as defined by inflorescence age) was greatest for A. billardierei var. filifolia , A. aemula var. setifolia and A. aemula var. aemula. Little or no effect was seen for A. billardierei var. billardierei or A. billardierei var. tenuiseta. Obviously the uneven numbers of specimens in each age class for some taxa inhibits detailed assessment but some trends are evident. Significant differences resulted from analysis of overall plant height, tussock height and culm height for a number of taxa. In general, the relative heights to overall height for tus- socks and/or culms decreased with age. For A. billardierei var. filifolia, A. aemula var. seti- folia and A. aemula var. aemula, this was the result of increasing overall plant height with age, at least to the ‘mature’ age. At the ‘ripe’ age for these taxa and for both the ‘mature’ and ‘ripe’ ages of A. billardierei var. robusta, overall height was less on average than for 2 11 is recognised that smaller spikelet forms of A. aemula var. aemula (e.g. 3.6 to 5.5 mm, as reported by Walsh and Entwisle 1994) are likely to have glume lengths similar to or shorter than those of var. setifolia. 76 A.J. Brown and N.G. Walsh Table 3. Number of specimens assessed for each age class of the currently recognised taxa (rep- resented by varietal epithet only) and measured plant characters showing significant dif- ferences (*** = P< 0.001, ** = p 0.001 to P 0.01, * = P 0.01 to P 0.05) with age. Taxon billardierei tenuiseta collicola robusta filifolia setifolia aemula Age I 1 0 2 4 9 4 5 G 4 2 2 17 16 19 6 M 12 3 4 28 17 5 5 R 9 6 0 14 2 5 4 Vegetative Hgt * ** * ** Character Llhp * * * Chp * ** Lr ** Inflorescence Iw ** *** *** ** Character la * *** I pc ** ** *** Sa ** Spikelet Sg * *** ** Character Git ** Ls ** Ps ** * At * * Aa * I = immature, G = maturing, M = mature, R = ripe younger aged plants and either reflects an earlier maturing of smaller and less robust plants or the disarticulation of early and taller inflorescences than those still present and able to be assessed by the study. Although relative tussock height was also significantly different for A. billardierei var. collicola (between taller immature plants and smaller maturing plants), the results are based on too few specimens to have any validation. Leaf roll increased with age for A. billardierei var. tenuiseta but only to a value of 2 (i.e. some leaves displayed slight inrolling). Other characters, thought to be associated with age, such as flag leaf length, inflorescence height and spikelet gape were not signif- icantly different, but this result may just reflect the small numbers of specimens assessed by ANOVA and the lack of any immature specimens. Panicles of A. billardierei var. robusta and A. billardierei var. filifolia showed signif- icant stiffening with age and. along with the varieties of A. aemula , broadened with age, as expected. Increasing age significantly increased the purpling of the panicles and/or spikelets of A. billardierei var. collicola, A. aemula var. setifolia and A. aemula var. aemula. The generally gaping glumes of ‘immature’ spikelets of A. billardierei var. robusta and the two A. aemula varieties progressively closed with age. The same trend was noted for A. billardierei var. filifolia but was not significant. The glumes, lemma setae, palea setae and awns of A. billardierei vox. filifolia showed significantly decreased sizes as age progressed. The same trends were observed in A. aemula var. setifolia for glumes and awns but only awns were significantly different. Delicate structures such as awns and setae points are likely to show reduced length with age, due to broken or withered tips, but the reduced glume length cannot be explained by the same process (no changes to glume setae length were evident). Mean glume lengths for A. billardierei war. filifolia were 6.0, 5.7, 5.3 and 5.2 mm for ‘immature’, ‘maturing’, ‘mature’ and ‘ripe’ specimens respectively, representing a 13% reduction overall. Further Revision of Agrostis billardierei 77 study would be required to ascertain whether this is a real reduction in size with age or an artefact of the collection process (i.e. the same plant populations were not represent- ed in each age class). Agrostis aemula var. aemula also shows significant differences in palea setae length, but these do not follow a trend in one direction. Awn attachment height significantly increased with age in this species, despite no change in lemma length. This again, may be an artefact of the collection process. EXAMINATION OF HERBARIUM SPECIMENS Fifty eight percent of the herbarium specimens examined had been determined as A. bil- lardierei var. billardierei. Comparison with the type confirmed 135 of these determina- tions. This included six specimens from Kangaroo Is. (Cape de Couedic and Rocky River N.P.), despite their small stunted appearance (about 15 cm tall with flat leaves, 2-3.5 mm wide), very short, often enclosed panicles and generally smaller than average spikelets. Vickery, on her determination label, regarded one of these specimens as ‘a form’ of var. billardierei (Rocky River. J B Cleland, 24 Nov. 1945, AD 96243134). Three specimens determined as A. billardierei var. billardierei were found to be A. aemula var. aemula while the reverse was true for a further six specimens. Superficially, these 2 taxa are similar in habit and leaf character. Even the type sheet has both taxa mounted on it ( R Be 6218, 1802-05, BM), a mistake referred to in the earlier revision (Vickery 1941). An additional 25 specimens were found to be incorrectly determined as A. billardierei var. billardierei. These included five specimens of A. avenacea with glabrous lemma backs (but hairy sides), three of A. venusta Trin., one each of A. capillaris L. var. arista- ta (Parnell) Druce (syn. A. castellana Boiss. & Reut.; Batson 1998) and A. billardierei var. fdifolia, three of A. billardierei var. tenuiseta and 12 of A. billardierei var. robusta. All of these taxa have lemmas with hairless backs like A. billardierei var. billardierei, and because of their comparative rarity, are likely to have been overlooked or not considered during determination. Some confusion between var. billardierei and var. robusta was evi- dent for a number of South Australian and Tasmanian specimens. Some specimens of var. billardierei (often inland collections) had distally finely scaberulous lemmas, not unlike some lesser scabrid forms of var. robusta (e.g. specimens from the Boomer Marsh/Marion Bay/Maria Is. region of Tasmania). In these cases, other characters (e.g. leaf width) were used for diagnosis. Most of the specimens determined as A. billardierei var. robusta (15 of 17) and A. aemula var. setifolia (18 of 19) were confirmed. Although the types for these taxa have contracted, just-emerging inflorescences and the basal leaves are partially senescent and therefore rather stiff, spikelet features serve as good diagnostic characters. Nevertheless, for each of these taxa, three specimens had been incorrectly determined as A. aemula var. aemula. Only 12 of 23 specimens determined as A. billardierei var. fdifolia conformed to the type. The remaining specimens were referable to A. billardierei var. robusta (5), A. aff. avenacea (2), A. venusta (2) or A. aemula var. setifolia (2). A number of stunted speci- mens from near-coastal south-east South Australia were difficult to place. They had low awn insertion points like A. billardierei var. fdifolia but rather small and crowded spikelets, scabrid lemmas and inrolled leaves like var. robusta. Vickery regarded one of these as ‘a peculiar form' of var. fdifolia (Cape Banks, J B Cleland, 27 Nov 1945 AD 97222340). CONCLUSION The results of the statistical analyses of the survey specimens and examination of herbar- ium specimens are evidence of a taxonomic position supporting all the currently recog- 78 A.J. Brown and N.G. Walsh nised taxa, but at levels and in combinations different from those currently accepted. An arrangement that is concordant with the analyses has: ( 1 ) A. aemula var. setifolia specifically distinct from A. aemula var. aemula (2) A. billardierei var. filifolia and A. aemula var. setifolia as varieties ot the same species (3) A. billardierei var. billardierei, A. billardierei var. robusta, A. billardierei var. colli- cola and a taxon including A. billardierei var. filifolia and A. aemula var. setifolia recog- nised as separate species. Agrostis billardierei var. tenuiseta has a number of distinct characters that separate it from A. billardierei var. billardierei, but PCO and cluster analysis suggest a varietal rela- tionship rather than a specific one. Although A. aemula var. aemula is superficially similar to A. billardierei var. bil- lardierei in vegetative form, there appears to be sufficient spikelet differences (apart from hairy lemmas) to maintain their separate specific status. The appropriate combinations and new species descriptions are made in the follow- ing section. Taxonomy KEY TO TAXA TREATED IN PRESENT STUDY 1 . Leaves greater than 2 mm wide and flat - E Leaves 2 mm wide or less 3 2. Awn greater than 3.5 mm long, inserted just below lemma mid-back, exceeding the glumes la. A. billardierei var. billardierei 2. Awn less than 2.0 mm long (or absent), inserted near lemma apex, not or hardly exceeding the glumes lb. A. billardierei var. tenuiseta 3. Awn greater than 3.5 mm, inserted below lemma mid-back, plants 20 cm tall or greater, leaves mainly involute or conduplicate (sometimes flat) (lowlands) 4 3. Awn less than 3.5 mm, inserted near lemma apex, plants less than 20 cm tall, leaves mainly flat (or conduplicate to slightly inrolled on drying) (highlands) 2. A. collicola 4. Awn inserted within lower 1/4 of lemma, palea setae mainly 0.4 mm long or more, anthers 0.8 mm long or more, leaves mainly conduplicate or flat 5 4. Awn inserted from lower 1/4-1/2 of lemma mid-back, palea setae mainly less than 0.4 mm, anthers 1 .0 mm long or less, leaves mainly involute 3. A. robusta 5. Lemma covered in hairs in the lower 3/4 4a. A. punicea var. punicea 5. Lemma (excluding callus) without hairs 4b. A. punicea var. filifolia TAXON DESCRIPTIONS The following descriptions are an expansion of those provided by Walsh and Entwisle (1994) and Morris (1990). A selected list of examined specimens accompany each taxon description while a full list of examined specimens is available from the senior author tor interested readers. 1 . Agrostis billardierei R. Br„ Prodr. 171(1810). Type: New South Wales, Port Jackson, 1802-05. R Brown (holotype BM). Mid to dark green (sometimes bluish-green), tufted, glabrous, perennial (may be annual under unfavourable conditions), 25—75 cm tall (including intlorescences but these some- times overtopped by leaves): culms ascending or erect. 1 5 — 40( 50) cm long. Leaf blades rather stiff, scabrous, flat (sometimes folded or loosely inrolled on drying): basal leaves (often forming a flattened tussock) 15—40 cm long, (2— )3— 7 mm wide, ligules obtuse. 3—12 mm long. Inflorescence generally a rather stiff, open panicle with erect to spread- Revision of Agrostis billardierei 79 ing branches, 10-30 cm long, 5-30 cm wide, its base enclosed by the upper leaf sheath or its lower branches becoming free with maturity; peduncle 2-15 cm long if visible; 2-7 branches in the lowest whorl; spikelets generally partly overlapping. Spikelets (3.0-)4.0-6.5(-7.0) mm long overall (excluding awn); glumes acuminate, subequal, keeled; lemma 2. 5-4. 5 mm long overall, hairless except for callus tuft and more or less shining (often with scattered, very fine to microscopic scabrid projections in the upper half and near the central nerve in the lower half but sometimes distally finely scaberu- lous), with 2-4 setae at apex; palea 1. 5-3.0 mm long overall, narrowly and very shortly bifid at the apex (points to 0.2 mm long); rachilla extension forming a plumose bristle (1 .5)— 2.0— 3.0 mm long (including hairs); anthers 0.5-1. 0 mm long. Coast Blown-grass. In. Agrostis billardierei var. billardierei Sometimes shortly rhizomatous. Flag leaves to 20(— 30) cm long, 2-5 (-8) mm wide. Panicle branches and pedicels and spikelets usually purple to dark or reddish-purple (even at inflorescence emergence) but fading to dull brown with age. Mature spikelets more or less gaping. Glume apex occasionally with a fine seta to 0.3 mm long, moder- ately scabrous along the keel and often lightly scaberulous on the sides, margins entire or with a few scattered cilia; lemma setae 0. 2-1.0 mm long; awn fine, bent, (3.5— )5.0— 9.0 mm long, well exceeding the glumes, attached 30-45% from the lemma base. Distribution-. Widely distributed along the coast in south-eastern Australia from the vicinity of Grafton in New South Wales, to at least as far west as Port Lincoln, South Australia with a few scattered inland occurences (e.g. Little Desert in Victoria), and in Tasmania occurring virtually all around the coast (including Bass Strait islands) (Fig. 6). Also occurring through coastal areas of New Zealand except the far north (Kermadec Is) and, apparently, the south-western part of the South Island. (Fig. 6) Figure 6. Distribution map of known collections of Agrostis billardierei var. billardierei in SE Australia. 80 A.J. Brown and N.G. Walsh Ecology. See general notes on habitat and phenology following taxonomic section. Selected specimens examined: South Australia: Port Elliot, 15 Jan. 1913 (AD); Harriet River. Kangaroo Is., 7 Oct. 1922, Osborn (AD); Port Lincoln, 17 Dec. 1941, Cleland (AD); 10 km west of Naracoorte, 18 Nov. 1961, Hunt (AD); Bankers Knoll, Younghusband Penin., 15 Dec. 1981. Williams 12 ISO (AD, MEL); Warooka, 16 Nov. 1989, Brown 455 (MEL, HO); near mouth of Marne River, 19 Dec. 1995, Spooner (AD). New South Wales: Port Macquarie, Nov. 1915. Boorman (NSW); Cave Beach, 4.8 km SW Jervis Bay, 12 Oct. 1971, Coveny 3683 (NSW): North Headland, Wamberal, 10 Nov. 1973, Jacobs 638 (NSW); Long Beach, Batemans Bay, 18 Nov. 1991, Crawford 1413 (NSW. MEL). Victoria: Wingan Inlet N.P. west of mouth, 23 Nov. 1969. Beauglehole and Finch 32002 (MEL, NSW); Cape Shanck, 3 Dec. 1970, Todd 27 (MEL); Point Lonsdale. 10 Dec. 1983, Albrecht 694 (MEL); Little Desert N.P, 18 Dec. 1983, Carr 7704 (MEL); Walkerville North, 5 Dec. 1994, Paget 1146 (MEL); St. Marnock’s Swamp, Crossroads, south of Eurambeen, 4 Jan. 1996, Brown 1117 (MEL). Tasmania: South Port. Jan. 1850, Stuart (MEL): Eaglehawk Neck, 15 Jan. 1949, Blake 18281 (HO); Wybalenna Is., off Flinders Is., 12 Dec. 1968. Harris (HO); Rocky Cape. 7 Jan. 1977, Mason 13249 (HO): Peron Dunes, St. Helens Point, 7 Jun. 1983, Buchanan 1196 (HO): Turua Beach. Deadmans Bay, 21 Jan. 1987, Moscal 14225 (HO, MEL); Planters Beach, Cockle Creek, 2 Feb. 1998, Buchanan 15056 (HO). 1 h. Agrostis billardierei var. tenuiseta D. Morris, Muelleria 7: 147 (1990). Type: Tasmania, Dolphin Sands, Nine Mile Beach, 10 Dec. 1984. Buchanan 4763 (holotype HO; isotype NSW ). Often rhizomatous. Flag leaves to 12 cm long, (1.5-)3-4 mm wide. Panicle branches, pedicles and spikelets mainly green tinged with purple but fading to straw with age. Mature spikelets hardly gaping. Glume apex without a fine seta or to 0.1 mm long, mod- erately to strongly scabrous along the keel and often lightly scaberulous on the sides, margins ciliated; lemma setae to 0.2 mm long or absent; awn very fine, straight or slight- ly curved, 0.5-2. 5 mm long (sometimes absent), not or hardly exceeding the glumes, attached 70-95 % from the lemma base. Distribution: Apparently confined to coastal areas in north-eastern Tasmania (includ- ing eastern Bass Strait islands). (Fig. 7) Ecology: See general notes on habitat and phenology following taxonomic section. Selected specimens examined: Tasmania: Clarkes Is., Furneaux Group, 26 Jan. 1966, Whinray 1572 (CANB), Babel Is., Furneaux Group, 22 Jan. 1967, Whinray 1764 (MEL); Whitemark, Flinders Is., Dec. 1975. Morris (HO): Passage Is., Furneaux Group, 6 Jan. 1979, Whinray (MEL); Kelvedon Beach. Great Oyster Bay. 28 Jan. 1999. Brown 1579 (MEL); Mayfield Beach, Great Oyster Bay. 28 Jan. 1999, Brown 1585 (MEL); Scamander Beach. Beaumaris. 15 Jan 2000, Brown 1595 (HO). 2. Agrostis collicola (D. Morris) A.J. Brown & N.G. Walsh, stat. nov. Agrostis bil- lardierei R. Br. var. collicola D. Morris. Muelleria 7: 147 ( 1990). Type: Tasmania, Saddle between The Hippo and Moonlight Ridge Hill 3, 10 Feb. 1985, Collier 309 (holotype HO). Mid to dark-green, tufted, glabrous, weak perennial. 10-20 cm tall (including inflores- cences); culms erect. 5-10 cm long. Leaf blades fiat to conduplicate (sometimes pseudo- convolute on drying); basal leaves (generally forming a small erect tussock or tuft) 5-15 cm long. 1-2 mm wide; flag leaves 1.5-7. 5 cm long, 0. 2-2.0 mm wide; ligules obtuse, 1 .5 — 4.0 mm long. Inflorescence an open panicle with spreading branches, 2-9 cm long, 4-9 cm wide, its base enclosed by the upper leaf sheath or its lower branches becoming free with maturity; peduncle 1-3 cm long if visible; 2—4 branches in the lowest whorl; branches and pedicels green, becoming purple with maturity; spikelets partly overlapping to not overlapping, generally gaping. Spikelets 2. 5-4.0 mm long overall (excluding awn); Revision of Agrostis billardierei 81 Figure 7. Distribution map of known collections of Agrostis billardierei var. tenuiseta in SE Australia. t 0 250 500 Km Figure 8. Distribution map of known collections of Agrostis collicola (syn. Agrostis billardierei var. collicola ) in SE Australia. 82 A.J. Brown and N.G. Walsh usually purple (green when immature); glumes acuminate and keeled (generally the keel extending to a tine seta up to 0.5 mm long), subequal, moderately to coarsely scabrous along the keel and densely and finely scaberulous on sides, margins smooth or with a few scattered cilia; lemmas 2. 5-3. 5 mm long overall, hairless except for callus tuft, smooth, sometimes purple streaked, with 2-4 setae at apex 0.2-0. 5 mm long; awn very fine, straight, decurrent from central nerve of lemma, 1.0-3. 5 mm long, attached 80-95% from the lemma base; palea 2. 0-3.0 mm long (minutely bifid at the apex if at all); rachilla extension forming a plumose bristle 1. 5-2.5 mm long (including hairs); anthers 0.4-0.8 mm long. Hill Blown-grass. Distribution : Apparently confined to mountainous areas of Tasmania, from 800-850 m altitude (Fig. 8). Ecology. See general notes on habitat and phenology following taxonomic section. Selected specimens examined: Tasmania: Lake Ewart, 7 Feb. 1987, Buchanan 10071 , (HO); Lake Will south of Barn Bluff, 15 Jan. 1989. Collier 3941 , (HO). 3. Agrostis robusta (Vickery) A.J. Brown & N.G. Walsh stat. now Agrostis billardierei R. Br. var. robusta Vickery, Contr. New South Wales Natl Herb. 1: 110 (1941). Type: Victoria, Melbourne, 17 Nov. 1853, Adamson 224 (holotype K). Mid to light-green (new shoots can be bluish-green), tufted, glabrous, annual or perenni- al, 25-75 cm tall (including inflorescences); culms ascending or erect, 10-55 cm long. Leaf blades rather stiff to lax, scabrous, convolute to strongly involute (sometimes flat- tening with age); basal leaves (sometimes forming an erect tussock) 10-50 cm long, 0.2-1 .()(— 1 .5) mm wide; flag leaves 2.5—1 5(— 35 ) cm long, 0.2—1 ,0(— 1 .5) mm wide; ligules obtuse, 2-7 mm long. Inflorescence an open panicle with spreading to lax branch- y~ •# •- • •v • 0 250 500 Km • • Figure 9. Distribution map of known collections of Agrostis robusta (syn. Agrostis billardierei var. robusta ) in SE Australia Revision of Agrostis billardierei 83 es (except at maturity), 10-25 cm long, 10-30 cm wide, its base enclosed by the upper leaf sheath or its lower branches becoming free in late maturity; peduncle 2-9 cm long if visible; 1-8 branches in the lowest whorl; branches and pedicels generally green to grey- ish-green or occasionally purplish-green; spikelets generally partly overlapping and more or less gaping. Spikelets 3. 5-5. 5 mm long overall (excluding awn), usually green to grey- ish-green; glumes acuminate and keeled (generally the keel extending to a fine seta 0.6(- 0.8) mm long), subequal or the lower slightly longer, moderately scabrous along the keel but smooth on sides (very occasionally lightly and distally scaberulous), margins with a few scattered cilia to fully ciliated; lemma 2.5— 4.5(— 5.5 ) mm long overall, hairless except for callus tuft, rather firm, strongly scaberulous in the upper half or at least on and near the setae and nerves towards the more or less fluted apex, more or less shining in the lower half and devoid of scabrid projections except for near the central nerve, with 2^1 setae at apex 0. 3-1.0 mm long; awn bent, (3.5— )4.5— 7.0(— 8.0) mm long, attached 25-45 % from the lemma base; palea 2.0-3.5(-4.0) mm long overall, narrowly and shortly bifid at the apex (points to 0.3(— 0.5) long); rachilla extension forming a plumose bristle 1 .5— 3.0(— 3.5) mm long (including hairs); anthers 0.4— 1 .0(— 1 .2) mm long. Salt Blown- grass. Distribution'. Scattered in coastal areas from near Seaspray (south-eastern Victoria) westward to the Goolwa Barrage in South Australia, extending to c. 100 km inland on saline soils. Also scattered along the Tasmanian coastline, particularly on the east coast. (Fig. 9) Ecology. See general notes on habitat and phenology following taxonomic section. Selected specimens examined'. South Australia: Eight Mile Creek, 3 Feb. 1942, Eardley (AD); Lake Bonney, Barmera, 17 Feb. 1947, Cleland (AD); Goolwa Barrage, 6 Jan. 1950. Cleland (AD); Lake Eliza, Little Dip N.R. 9 Dec. 1975, Brock 239 (AD); Alcock 33 (AD); Mongolata Hills, 27 May 1989, Bates 18551 (AD); between Gulnare and Spalding, 16 Nov. 1989, Bates (AD); Avenue, 25 Jan. 1999, Brown 1570 (MEL, AD); Naracoorte, 25 Jan. 1999, Brown 1572 (MEL, AD). Victoria: Glenthompson, 27 Jan. 1993, Heard (MEL); Skipton, 1 1 Jan. 1994, Brown 870 (MEL); Lake Corangamite, North Cundare, 16 Dec. 1994, Brown 973 (MEL); Woodbourne, 4 Jan. 1996, Brown 1130 (MEL); Ross Bridge, 8 Jan. 1996, Brown 1149 (MEL); Murtagurt Swamp, Barwon Heads, 24 Dec. 1996, Brown 1223 (MEL); Honeysuckle Estate, Seaspray, 17 Dec. 1998, Brown 1527 (MEL). Tasmania: mouth of Boomer Creek, 12 Jan. 1941, Curtis (HO, MEL); St. Helens, 24 Dec. 1959, Burns 213 (HO); Point Lesueur, Maria Is., 12 Dec. 1977, Brown 334 & 449 (HO); Bathurst Harbour, Celery Top Islands, 12 Apr. 1978, Kirkpatrick (HO); Calverts Lagoon, South Arm, 4 Feb. 1979, Morris 79104 (HO, MEL); Gull Reef, Port Davey, 10 Feb. 1980, McKendrick (HO); mouth of Harcus River, 27 Dec. 1986, Buchanan 8973 (HO); 4 km west of Harley’s Point, Cape Barren Is., 12 Dec. 1988, Buchanan 11142 (HO); Cape Portland near Cape Lagoon, 7 Jan. 1993, Steane (HO). 4. Agrostis punicea A.J. Brown & N.G. Walsh Mom. et stat. nov. Type: Tasmania, New Norfolk, 15 Nov. 1840, Ballantine 1446 (holotype K; isotype HO). Agrostis billardierei R. Br. var. setifolia Hook, f., FI. Tas., 2: 1 15 (1860). Agrostis aemula R. Br. var. setifolia (Hook.f. ) Vickery Conti : New South Wales Nat I Herb. 1 : 1 16 (1941 ). Bluish-green, tufted, glabrous, short-lived perennial , 20-65 cm tall (including inflores- cences); culms erect, 1 0— 30(— 45) cm tall. Lcq/'blades rather stiff, scabrous, conduplicate to pseudo-convolute (sometimes becoming involute on drying); basal leaves (generally forming an erect to spreading tussock) 5— 25(— 35 ) cm long, 0.2-1 .0 mm wide; flag leaves 1.5-10 cm long, 0.2-0. 5(-l .0) mm wide; ligules obtuse, 2-8 mm long. Inflorescence a fine, open panicle with erect to spreading branches, 1 0— 25(— 30) cm long, 5-30 cm wide, generally well exserted from the upper leaf sheath; peduncle 2-20 (-25) cm long; 2-5 branches in the lowest whorl; branches and pedicels very fine, often pinkish-purple to reddish-purple; spikelets rather few, generally not overlapping, widely gaping. Spikelets 84 A.J. Brown and N.G. Walsh 4.5-7. 0 mm long overall (excluding awn); usually purple to dark purple; glumes acumi- nate and keeled (occasionally the keel extending to a fine seta 0.4 mm long), subequal, moderately scabrous along the keel but smooth on sides (occasionally lightly and distal- ly scaberulous), margins entire or occasionally with a few scattered cilia; lemma 3.0— 5.0(— 5.5) mm long overall, with 2^4 setae at apex 0.5-1. 5 mm long; awn bent, attached 10-20(-25) % from the lemma base and often (particularly when immature) lying in a groove of the lemma along the central nerve; palea 2.5-4.5 mm long overall, narrowly bifid at the apex (points (0.3— )0.4— 1 .2 mm long); rachilla extension forming a plumose bristle ( 1 .5— )2.0— 3.5 mm long (including hairs); anthers 0.8-1 .7( — 2. 1 ) mm long. Purple Blown-grass. Etymology. The epithet, meaning reddish-purple in Latin, refers to the colour of the panicle branches which is conspicuous on flowering plants. The epithets filifolia and seti- folia which might otherwise have been chosen are preoccupied at species level within Agrostis. 4a. Agrostis punicea var. punicea Lemma covered in lower three-quarters with hairs, upper nerves and setae finely scaberu- lous. Awns 7.5-12.5 mm long. The sizes of glumes, lemmas, paleas and rachilla exten- sions (including hairs) do not occur in the bottom 20 % of each range for the species. Distribution: Scattered across the volcanic plain and Dundas Tableland of south-west Victoria, extending into south-east South Australia with isolated occurrences in near- coastal south Gippsland. Also scattered in the Tasmanian midlands. (Fig. 10) Ecology: See general notes on habitat and phenology following taxonomic section. Selected specimens examined: South Australia: Cooloolie, 11 Nov. 1945, Crocker (AD); Figure 10. Distribution map of known collections of Agrostis punicea var. punicea (syn. Agrostis aemula var. setifolia) in SE Australia. Revision of Agrostis billardierei 85 Marshes Swamp, between Glencoe and Mt. Burr, 19 Jan. 1969, Wilson 970 (AD); Tantanoola Forest, 23 Nov. 1991, Bates 26388 (AD). Victoria: Steep Bank Creek, 1 Nov. 1 982, Corrick 8473 (HO, MEL); Hamilton, 3 Nov. 1982, Corrick 8544 (HO, AD); Buckleys Swamp, 22 Dec. 1986, Brown 72B (MEL); Craigieburn, 10 Nov. 1989, Frood (MEL); Mullungdung State Forest, 2 Nov. 1995, Paget 1609 (MEL); 12 km north of Sale, 27 Oct. 1997. Paget 2941 (MEL); Beeac, 7 Dec, 1995, Brown 1014 (MEL); Hexham, 15 Dec. 1995, Brown 1047 ( MEL): Mt. Pollock, 3 Dec. 1997, Brown 1384 (MEL); Byaduk, 5 Dec. 1997, Brown 1438 (MEL); Ripponhurst, 5 Dec. 1997, Brown 1445 (MEL); Digby, 3 Dec. 1998, Brown 1512 (MEL); Jack Smith Lake, Darriman, 31 Dec. 1998, Brown 1543 (MEL); Coojar, 25 Jan. 1999, Brown 1575 (MEL). Tasmania: Penquite, 9 Dec. 1844, Gunn 592 (K); Uiverstone, 1912, Burbury (HO); 7 km west of Ross, 10 Dec. 1984, Fensham (HO); Epping Forest, Rowranna, 14 Dec. 1984, Fensham (HO); Shooters Hill, 22 Dec. 1984, Fensham (HO); 1.5 km north-west of Pringle Hills, 24 Dec. 1984, Fensham (HO). 4b. Agrostis punicea var. filifolia A.J. Brown & N.G. Walsh comb. nov. Agrostis bil- lardierei R.Br.var .filifolia Vickery, Conti: New South Wales Natl Herb. 1: 110 (1941). Type: Victoria, Hawkesdale, Dec. 1901, Williamson (holotype K). Lemma hairless except for callus tuft, covered with very fine to microscopic papillose- scabrid projections, upper nerves and setae finely scaberulous. Awns 6.0-10.0 mm long. The sizes of glumes, lemmas, paleas and rachilla extensions (including hairs) cover the full range for the species. Distribution : Scattered in the western half of south-west Victoria and extending into south-east South Australia with isolated occurrences at Lyndhurst and Meerlieu in Gippsland. Also two old records from near Hobart in Tasmania. (Fig. 1 1) Ecology: See general notes on habitat and phenology following taxonomic section. Figure 11. Distribution map of known collections of Agrostis punicea var. filifolia (syn. Agrostis billardierei var. filifolia) in SE Australia. 86 A.J. Brown and N.G. Walsh Selected specimens examined: South Australia: near Mt. Gambier, 1880, von Mueller (MEL); Yallum and Beachport, Oct. 1883 (AD); Wooleys Lake, Beachport (?). Oct 1883 (AD); Penola. 27 Oct. 1945. Survey team (AD). Victoria: Buckleys Swamp, 22 Dec. 1986, Brown 72A (MEL); Connover Swamp. Drik Drik, 3 Dec. 1992, Albrecht 5159 (MEL, AD, HO); 12 km east of Woorndoo, 3 Dec. 1992, Albrecht 5189 (MEL); Lyndhurst. 19 Nov. 1993, Costello s.n. (MEL); Glenthompson, 10 Jan. 1994. Brown 827 (MEL); Hexham, 15 Dec. 1995, Brown 1046 (MEL); Hadden. 4 Jan. 1996, Brown 1123 (MEL); Lake Goldsmith, Stockyard Hill, 6 Dec. 1996, Brown 1189 (MEL); north of Karabeal, 13 Dec. 1996, Brown 1204 (MEL); Lake Repose, Glenthompson, 30 Dec. 1996, Brown 1246 (MEL); Dunkeld, 30 Dec. 1996, Brown 1241 (MEL); south of Bulart, 31 Dec. 1996. Brown 1270 (MEL); Moutjup, 14 Nov. 1997, Brown 1341 (MEL); Ballyrogan, 23 Dec. 1997, Brown 1473 (MEL); Homerton, 3 Dec. 1998, Brown 1517 (MEL); Meerlieu, 31 Dec. 1998, Brown 1539 (MEL). Tasmania: Blackman’s Bay, Feb. 1929, Rodway (HO); Hobart. 1929, Rodway (HO). NOTES ON RELATED TAXA Lachnagrostis tenuis (Cheeseman) Edgar, New Zealand J. Bat. 33: 30 (1995). Deyeuxia billardierei (R. Br.) Kunth. var. tenuis Petrie ex Cheeseman, Man. N.Z. FI. 870 (1906). Type: New Zealand, Catlins River, Clutha Co., Otago, March 1896, H.J. Matthews (holo- type WELT; isotypes AK, WELT. CHR). The 'Blown-grasses’ of New Zealand have been separated from Agrostis L. into Lachnagrostis Trin. (Edgar 1995). This genus has 12 New Zealand species, with 2 of these ( Lachnagrostis billardierei (R. Br.) Trin. syn. Agrostis billardierei R. Br., and L. fil- iformis (G. Forst) Trin. syn. Agrostis avenacea J.F. Gmel.) also found in Australia. Another species, Lachnagrostis tenuis (Cheeseman) Edgar which is not known from Australia, has many morphological characters in common with Agrostis robusta. In par- ticular, it has narrow, strongly involute leaves and scaberulous lemmas. In general appear- ance, L. tenuis resembles smaller and less mature forms of A. robusta with its relatively small stature (20-30 cm tall) and short leaves (5-15 cm). It differs principally in its small- er and only partly spreading inflorescences (5-10 cm long, 5-10 cm wide) that are almost always enclosed at the base by the leaf sheath, and its shorter (1-2 mm long) ligules. Further, the lemma surface of L. tenuis is almost completely covered with strongly scaberulous projections, whereas A. robusta is mainly scaberulous in the distal half only. Lachnagrostis tenuis has only been recorded from near-coastal latitudes south of 43° (Canterbury on South Island) whereas Agrostis robusta extends from 43° (southern Tasmania) to 33° (North Lofty Region of South Australia). However, like A. robusta , it is recorded as growing in salt marshes and tidal ground. Further study is needed before a definite conclusion can be reached on the correct placement of these taxa. Research on the generic limits ot Agrostis and related genera (in particular, Deyeuxia) are in progress (S.L.W. Jacobs, New South Wales Herbarium, pers. comm.) and the results of this study may recommend the transferral of several native taxa into other genera. This work should include a reassessment of A. robusta and L. tenuis. Current knowledge suggests that these two entities could be recognised at infraspecific rank only. Specimens examined : NEW ZEALAND, South Is.: Catlins River, Otago, I). Petrie, Mar. 1896 (WELT); Fortrose, Southland, D. Petrie, 4 Jan. 1913 (WELT); Lyttleton Harbour near Teddington,, 2 Jan. 1966 (CHR); Stewart Is. Near head of South-west Arm, H.D. Wilson and C.D. Meurk, 18 Feb. 1980 (CHR). Growth Habit and Habitat GENERAL OBSERVATIONS Agrostis billardierei var. tenuiseta has a very similar growth appearance to var. billardierei except that it appears to have a greater tendency to produce rhizomes. This habit is partic- ularly noticeable where the taxon is growing on deep wind-blown sand, such as on the edge Revision of Agrostis billardierei 87 of the seaward side of a sand dune. In Tasmania, var. tenuiseta has only been found, thus far, growing in stands of the introduced Ammophila arenaria, close to the beach, either on or between the first two or three sand dunes. In contrast, var. billardierei tends to grow in more sheltered positions, such as in tall grass or shrub communities in the depressions between sand dunes, on grassy (Poa, Austrostipa spp.) fiats behind the sand dune complex or in cliff-top coastal scrub. Where it occasionally grows in more exposed positions, it is generally stunted and inflorescences ripen before exsertion from the leaf sheaths. Observations of growth habit for A. billardierei var. billardierei and A. robusta have found that contrary to published descriptions (as A. billardierei var. robusta), A. robusta is more likely to have a less rigid habit than A. billardierei. However, habit is variable, depending on habitat. Where A. billardierei is found growing in more exposed situations or on saline flats, leaves, culms and mature panicles tend to be erect and rather stiff, com- pared to the more lax leaves and ascending culms of the smaller plants growing in the shelter of Leptospermuml Acacia thickets. Agrostis robusta can form erect and dense tus- socks where soil moisture levels are more or less permanent, but generally it occurs as a weeping tussock with few culms. In both habitats, the leaves and panicles are more or less lax until the onset of senescence. Agrostis robusta is always found on saline flats or marshes and often in association with other salt-tolerant species such as Puccinellia spp., Juncus kraussi, Plantago coronopus, Triglochin striata and Sarcocornia quinqueflora. Agrostis punicea var. punicea and var .filifolia have a similar range of habitats and are sometimes found mixed together in the same population. Their appearance in the field is identical and only examination of the lemma can discern the difference. Their fine, wav- ing panicles can often be overlooked in a field of other grasses and herbs, unless they are a dominant species in the community. These taxa have been found growing in moist depressions in grasslands of Themeda, Austrostipa and Austrodanthonia spp., on the edges of non-saline swamps and on saline flats. Agrostis collicola has been collected from exposed situations with little vegetation or from herb lawns on stream banks (Morris 1990). SOIL CHARACTERISTICS Table 4 shows soil analysis data for a range of A. billardierei var. billardierei. A. robusta and A. punicea sites sampled during January 1998 or 1999. Surface soil textures vary widely for each taxon. Means of electrical conductivity (and calculated total soluble salts) and soil moisture contents are similar for A. billardierei var. billardierei and A. punicea with data for both taxa ranging from low to moderate levels. Soil salts and moisture means and ranges for A. robusta are considerably higher than for the other two taxa. However, soil pH means (moderately alkaline) and ranges for A. robusta are similar to A. billardierei, while A. punicea has a lower mean value (slightly acid). Table 4: Range in soil texture, mean and range in soil pH in water (pH w ), electrical conductivity (EC) dS/m, total soluble salts (TSS) % and moisture content on dry soil basis (M d ) % for various Agrostis spp. sites, sampled in January 1998 and 1999. Taxa No. sites Soil texture range PH W EC TSS M d A. billardierei var. billardierei 6 sandy loam to light clay 8.3 (6. 8-9. 6) 0.73 (0.15-1.7) 0.22 (0.04-0.50) 30 (11-48) A. robusta 7 fine sandy clay loam to light clay 8.0 (7.7-8. 7) 2.9 (1. 5-5.7) 0.85 (0.45-1.7) 89 (46-269) A. punicea var. filifolia 5 organic loam to light medium clay 6.5 (5. 7-8. 7) 0.74 (0.1 1-2.0) 0.22 (0.03-0.60) 38 (15-52) 88 A.J. Brown and N.G. Walsh PHENOLOGY Inflorescence emergence and maturing for all taxa varies according to environmental con- ditions but some general observations can be made. Agrostis punicea plants tend to mature earlier than those of A. billardierei var. billardierei or A. robusta but similar to A. aemula. In general, inflorescences of these taxa can start to appear during late October, are fully expanded by mid November and have mostly disarticulated by the end of December. Inflorescences of A. billardierei var. billardierei become noticeable during mid to late November and are mature by the end of December, although some younger panicles may also emerge at this time if conditions are favourable. Agrostis robusta gen- erally does not show any inflorescences until mid December with new inflorescences emerging right up to the end of March if soil moisture is adequate. Maturity generally peaks around mid February. LONGEVITY The percentage of tussocks tagged in January 1998 that survived (i.e. grew new shoots and developed new inflorescences) through to January 1999 varied from site to site. For the two sites of Agrostis robusta , results were 44% and 100% survival. Survival rates dropped to 13% and 50% respectively over the following season. The latter site borders an area inundated with water all year except for mid summer, although even at that time, the soil remains wet. The former site, while remaining moist, is drier over the summer. Results of tussock survival for A. punicea were 80% and 100% after the first season but only 47% and 15% respectively after the second season. Both these taxa appear to be short-lived perennials, given favourable conditions. A few tussocks marked at two sites of A. billardierei var. billardierei were still growing vigorously when observed two years later and attest to the perennial nature of this taxon. Acknowledgments We are grateful to Alex Buchanan and Dennis Morris (HO), John Jessop (AD) and Surrey Jacobs (NSW) for helpful discussions and assistance in the field and herbarium, to staff of all herbaria (including MEL and CANB) for access to collections and to staff from BM, CHR. K and WELT for the prompt dispatch of loan material. We also thank staff at the State Chemistry Laboratory (Agriculture Victoria, Werribee) for the soil analyses and John Reynolds (Agriculture Victoria, Attwood) for assistance and advice with statistical analysis. References Batson, M.-G. (1998). Agrostis castellana (Poaceae), dominant Agrostis species, found in bent grass pastures in south-eastern Australia. Australian Journal of Botany, 46 : 697-705. Black, J.M. (1943). Flora of South Australia. Part 1. Cyatheaceae - Orchidaceae. Government Printer, Adelaide. Curtis, W. and Morris, D. (1994). The Students Flora of Tasmania. Part 4b. Alismataceae to Burmanniaceae. St. Davids Park Publishing, Hobart. James, E.A. and Brown, A.J. (2000). Morphological and genetic variation in the endangered Victorian endemic grass Agrostis adamsonii Vickery (Poaceae). Australian Journal of Botany, 48 : 383-395. Jessop. J.P. and Toelken, H.R. (1986) eds. Flora of South Australia. Parts 1-4. The Flora and Fauna of South Australia Handbooks Committee. Adelaide. Morris, D.I. (1990). New taxa and a new combination in Tasmanian Poaceae. Muelleria 7 : 147-171. Simon, B.K. (1993). A key to Australian grasses. Queensland Department of Primary Industries, Brisbane. Vickery. J.W. ( 1941 ). A revision of the Australian Agrostis Linn. Contributions from the New South Wales National Herbarium 1 : 101-19. Walsh, N.G. and Entwisle, T.J. (1994) eds. Flora of Victoria. Volume 2. Ferns and Allied Plants, Conifers and Monocotyledons. Inkata Press, Melbourne. Revision of Agrostis billardierei 89 Appendix 1 . Morphological plant characters measured and used in statistical analysis. Plant Measured and Multivariate Variable part derived characters analysis type Total Total height (Hgt) Yes *C Basal tussock height (Llh) Yes C % basal to total height (Llhp) No c Culm height (Ch) Yes c % culm to total height (Chp) No c Leaves Inrolling or folding (Lr) Yes **D (1-5) Basal width (Llw) Yes c Flag width (Lfw) Yes c Flag length (Lfh) Yes c Ligule length (Lig) Yes c Infloresc. Extension from leaf sheath (Ie) Yes D (1-4) Length of visible peduncle (Ipd) Yes c No. of visible branches in lowest whorl (lib) No c Height from lowest whorl (Ih) Yes c Width (Iw) Yes c Height to width ratio (Ih:Iw) No c Stiffness (la) Yes D (1-4) Pedicel colour (Ipc) No D (1-5) Spikelet clustering (Sa) Yes D (1-3) Spikelet Outer glume colour (Sc) No D (1-5) Glume gape (Sg) Yes D (1-4) Lower glume Total length (Git) No C Body length (Gib) Yes C Setae length (Gls) Yes C % setae of total length (Glsp) No C Upper glume Total length (Gut) No C Body length (Gub) Yes C Setae length (Gus) Yes c Both glumes Scabridity of keels (K) Yes D (0-3) Ciliation of margins (M) Yes D (0-2) Total length ratio (Glt:Gut) No C Lemma Total length (Lt) No c Body length (Lb) Yes c Setae length (Ls) Yes c % setae of total length (Lsp) No c Density of back hairs (B) Yes D (0-4) Coarseness of scabrid surface (Rc) Yes D (0-3) Density of scabrid surface (Rd) Yes D (0-3) Palea Total length (Pt) No C Body length (Pb) Yes C Setae length (Ps) Yes c % setae of total length (Psp) No c 90 A.J. Brown and N.G. Walsh Rachilla extension Total length inch hairs (Re) Yes C Awn Total length (At) No C Column length (Ac) Yes C Bristle length (Ab) Yes C % bristle of total length (Abp) No C Awn attachment height from lemma base (Aa) Yes C % awn attachment of total length (Aap) No C Anther Total length (A) Yes C * C = continuous variable, ** D = discrete variable [ordinal scales for discrete variables: Lr: 1= flat, 2 = flat and involute (as a result of drying out), 3 = flat and conduplicate (may become pseudo-convolute and/or slight- ly involute with age), 4 = conduplicate, 5 = convolute to strongly involute; Ie: 1 = not exserted or spreading, 2 = not exserted but spreading, 3 = not fully exserted but more or less fully spread, 4 = fully exserted; la: 1 = very stiff and erect, 2 = rather stiff and more or less erect, 3 = not stiff but erect to slightly drooping, 4 = lax and weeping; Ipc and Sc: I = light green, 2 = mid-green, 3 = greyish green (green with minute purple mottling), 4 = mid-purple, 5 = dark purple to reddish purple; Sa: 1 = spikelets overlapping and obscuring individuals, 2 = spikelets partly overlapping but individuals discernible, 3 = spikelets more or less separate without overlapping; Sg: 1= nil or scarcely gaping outer glumes, 2=slightly gaping outer glumes (up to 10°), 3 = moderately gaping outer glumes (10° to 30°), 4 = widely gaping outer glumes (> 30°); K: 0 = non scabrous, 1 = slightly scabrous, 2 = moderately scabrous , 3 = strongly scabrous (often with scaberulous sides as well); M: 0 = non ciliate, I = minutely or distantly ciliate, 2 = uniformly ciliate; B: 0 = non-hairy, 1 = occasional hairs only, 2 = slightly or distantly hairy, 3 = covered in hairs but surface still discernable, 4 = covered in hairs and surface obscured; Re: 0 = no bristles (but may be villous), 1 = minutely scaberulous, 2 = finely scaberulous, 3 = strongly scaberulous; Rd: 0 = not scaberulous (but villous), 1 = lightly scaberulous (generally scattered on central nerve of lemma and setae only), 2 = moderately scaberulous (generally upper lemma back, central nerve and setae only), 3 = dense- ly scaberulous (upper lemma back, central nerve and setae to total lemma surface)]. Muelleria 14:91 (2000) I wo new species of Thelymitra (Orchidaceae) from southeastern Australia Jeffrey A. Jeanes do National Herbarium of Victoria, Birdwood Avenue, South Yarra. Victoria 3141, Australia Abstract Thelymitra atronitida and Thelymitra planicola from southeastern Australia are described as new and illustrated. Their distribution, habitat, biology and conservation status are discussed. Thelymitra atronitida is a member of the complex of taxa surrounding Thelymitra nuda R. Br., its closest congener being Thelymitra malvina M.A. Clem., D.L. Jones & Molloy. Thelymitra plani- cola bears a superficial likeness to Thelymitra pauciflora R. Br. and Thelymitra aristata Lindl., and its possible relationship to these species is discussed. A key is provided to distinguish the new taxa from other members of the Thelymitra nuda/ pauciflora complex in southeastern Australia. Introduction Thelymitra J.R. Forst. et G. Forst. is a large and complex genus consisting of about 70 described species, several described natural hybrids and an uncertain number of unde- scribed taxa. It is mainly concentrated in higher rainfall areas of temperate Australia, but a few species occur in tropical northeastern Australia, about 10 endemic species occur in New Zealand and four additional species occur in Indonesia, New Caledonia, New Guinea and the Philippines. While examining preserved and living plants as part of a revi- sion ot Thelymitra for Victoria (in preparation), it became evident that several distinct undescribed taxa exist in Victoria and at least some of these probably extend to neigh- bouring States. Due to a paucity of collections and several poor flowering seasons because of drought, some suspected undescribed taxa have not yet been satisfactorily studied and assessed, hence their status remains unresolved. However, this opportunity is taken to describe two new species that have been reasonably well collected in recent years, and are familiar to me from herbarium material as well as from living plants in situ. Taxonomy Thelymitra atronitida Jeanes, sp. nov. Thelymitra malvinae M.A. Clem.. D.L. Jones et Molloy affinis bracteis sterilibus duabus plerumque, floribus paucioribus autofertilibus facultative, lobo post-antheram atronitido plerumque, pilis caespitosis loborum lateralium albeis differt. Type: Victoria. Beside Genoa Creek Track, c. 4 km WS W of Genoa, 27. x. 1 999, J.A. Jeanes 613 (holotype MEL 2069953, isotype CANB). Glabrous terrestrial herb. Tubers not seen. Leaf linear to linear-lanceolate, 15-35 cm long, 5-12 mm wide, erect, leathery, canaliculate, dark green with a purplish base, ribbed abaxially, sheathing at base, apex acute. Scape 30-50 cm tall, 1.5-3.5 mm diarm, straight, straw-coloured to purplish. Sterile bracts usually 2, occasionally 3, linear to linear-lance- olate, 1.5-7 cm long, 3-10 mm wide, green or purplish, acute to acuminate. Fertile bracts ovate-acuminate to obovate-acuminate, 5-25 mm long, 3-8 mm wide, green or purplish, sheathing pedicels. Pedicels 5-12 mm long, slender. Ovary cylindric to narrow-obovoid' 5-12 mm long, 2-4 mm wide. Flowers 2-8, (14-)20-26 mm across, moderately dark blue with darker longitudinal veins, opening readily only in hot weather. Perianth seg- ments (7-) 10-13 mm long, 3-8 mm wide, concave, often shortly apiculate; dorsal sepal ovate to elliptic; lateral sepals elliptic to lanceolate, slightly asymmetric; petals ovate to 92 J.A. Jeanes ori?r> rri^3F<<5LO zc>€>o Figure 1. Tlielvmitra atronitida : a habit ? 0.5; b column from rear ? 6; c column from side ? 6: d column from front ? 6. Thelymitra planicola: e habit ? 0.5; f column from rear ? 6; g column from side ? 6; h column from front ? 6. New species of Thelymitra 93 elliptic; labellum elliptic to lanceolate, often narrower than other segments; column erect from the end of ovary, 5.5-7 mm long, 2. 5-3. 5 mm wide, mostly pale blue; post-anther lobe 3-4 mm long, 1.5-2. 5 mm wide, tubular, inflated, hooded, dorsally compressed, gently curved through c. 90°, apex shortly bilobed. lobes toothed, mostly glossy black, apex yellow; auxiliary lobes absent; lateral lobes 1.3-1. 5 mm long, filiform, porrect at base then bent sharply upwards at about the middle, converging, each with a terminal toothbrush-like arrangement of white trichomes, the individual trichomes 1.2-1. 6 mm long. Anther ovoid 2. 6-3. 3 mm long, 1.2-2. 2 mm wide, with a short beak 0.5-0. 8 mm long, situated towards apex of column; pollinarium 2-2.5 mm long; viscidium more or less circular, c. 0.5 mm diam.; pollinia white, friable, mealy. Stigma ovate-quadrate, 1.7-2. 5 mm long, 1.8-2. 2 mm wide, margins irregular, situated at base of column. Capsules obovoid, 12-15 mm long, 4-6 mm wide, erect, ribbed. (Fig. 1. a-d; Fig. 3) Specimens examined: Victoria: Between Jones Creek Track and private property, W of Genoa River. 7. xi. 1970, A.C. Beauglehole 34425 (MEL 652511 ); Near the intersection of East Wingan Rd and Princes Hwy, 24. x. 1999. J.A. Jeanes 643 (MEL 2069956 ); On rail reserve near Darlimurla Station, 29. x. 1999, J.A. Jeanes 636 (MEL 2069955 ); East Wingan Rd c. 500 m from Princes Hwy, 27. x. 1999, J.A. Jeanes 615 (MEL 2069959) Distribution and habitat: East Gippsland and Gippsland Highlands Natural Regions (Conn 1993). Currently known from a few sites in eastern Victoria, mostly in the Mallacoota/Genoa area of far East Gippsland, with an outlying more westerly population near Mirboo North in the Strzelecki Ranges (Fig. 2). Grows in heathy open forest, usual- ly around the margins of grasstree plains, on well-drained sand or clay loams. A range extension into nearby southeastern New South Wales is anticipated, as similar habitat is known to occur there. Altitude: 5-200 m. Conservation Status: Poorly known; suggest 3KC by criteria of Briggs & Leigh (1996). Flowering period: Late October to early November. Figure 2. Distribution of Thelymitra atronitidci Figure 3. Thelymitra atronitidci Mailacoota area 94 J.A. Jeanes Biology : This species is facultatively autogamous. Notes: Thelymitra atronitida is part of the complex assemblage of taxa surrounding Thelymitra nuclei, a species that has been variously circumscribed by modern flora writ- ers (Bernhardt 1993; Weber & Entwisle 1994) and orchid specialists (Jones 1988; Bates & Weber 1990; Backhouse & Jeanes 1995; Bishop 1996). The type of T. nudci comes from northern Tasmania (Western Arm of Port Dalrymple), and a re-examination of this material forms the basis of a narrower circumscription of the taxon by Jones & Clements ( 1 998) (see also Jones et al. 1 999). Thelymitra nuclei sensu stricto occurs on the Australian mainland (Jones et al. 1999). and I have seen plants in Victoria (eg. near Onreo) that con- form to the new circumscription of the species. However, at present the distribution of T. nudci in Victoria is unknown, due largely to confusion with other related taxa. For exam- ple, the illustration on page 354 of Backhouse & Jeanes (1995), supposedly depicting T. nuda, probably represents an undescribed taxon related to Thelymitra megcalyptra Fitzg. Thelymitra atronitida differs from T. nuda sensu stricto in its darker blue, more prominently veined, autogamous flowers, and inflated, dorsally compressed, shortly bilobed, predominantly glossy black column post-anther lobe. The two species do not grow sympatrically. Thelymitra atronitida closely resembles Thelymitra malvina, but the former usually has only two sterile bracts (usually three in T. malvina), fewer, generally smaller Bowers that are autogamous (entomophilous in T. malvina), a column post-anther lobe that is mostly glossy black (mostly brownish in T. malvina) and white hairs tufts on the lateral-lobes (mauve or pink in T. malvina). Thelymitra malvina is widely distributed from southeastern South Australia to southeastern Queensland (including Tasmania and New Zealand) and occasionally grows sympatrically with T. atronitida. Etymology : From the Latin ater, black; nitidus, shining; in reference to the column post-anther lobe that is predominantly glossy black. Thelymitra planicola Jeanes, sp. nov. Thelymitra pauciflorae R. Br. atfinis bractea sterili interna toliiformi plerumque, lobo post-antheram breviore subcylindico truncato, lobis lateralibus longioribus differt. Type: Victoria. Golden Beach. SE edge of Lake Reeve, 26. x. 1999, J.A. Jeanes 608 (holotype MEL 2069957, isotypes MEL 206995#, CANB). Glabrous, somewhat glaucous terrestrial herb. Tubers not seen. Leaf linear to linear- lanceolate, 10-30 cm long, 5-20 mm wide, erect, leathery, canaliculate, dark green with a purplish base, ribbed abaxially, sheathing at base, apex acute. Scape 22-45 cm tall. 2-5 mm diam., straight, purplish. Sterile bracts 2-4. linear to lanceolate, 2-8 cm long, 5-15 mm wide, green or purplish, lower ones often leaf-like, closely sheathing for most of length, acute to acuminate. Fertile bracts ovate-acuminate to obovate-acuminate, 1 0-20 mm long, 4-8 mm wide, green or purplish, sheathing pedicels. Pedicels 1-8 mm long, stout. Ovarv cylindric to narrow-obovoid, 3-12 mm long, 1.5-4 mm wide. Flowers 2-12, ( 1 5-)20-25(-30) mm across, medium blue with darker blue longitudinal veins, opening readily only in hot weather. Perianth segments ( 7—) 1 0— 1 3 (— 1 5 ) mm long, 4-8 mm wide, concave, often shortly apiculate; dorsal sepal ovate to obovate; lateral sepals ovate- lanceolate, slightly asymmetric; petals ovate to obovate; labellum oblanceolate to obo- vate. often narrower than other segments; column erect from the end of ovary, 5-6 mm long, 3-4 mm wide, white to pale blue; post-anther lobe 1.5-2 mm long, 1.5-2 mm wide, hooding, semi-cylindric. gently curved, apex shallowly bilobed, lobes shallowly toothed, mostly dark blackish brown with a thin blue collar, apex yellow; auxiliary lobes absent; lateral lobes 1.2-2 mm long, linear, obliquely erect, curved, converging, each with a short, terminal, toothbrush-like arrangement of white trichomes, the individual trichomes 0.9- 1.5 mm long. Anther ovoid, 2.7—4 mm long, 1.5— 2.2 mm wide, basal part obscured behind stigma, with a short terminal beak c. 0.5 mm long; pollinarium 1 .5-2.2 mm long; New species of Thelymitra 95 Figure 4. Distribution of Thelymitra plani- cola Figure 5. Thelymitra planicola, Golden Beach viscidium circular, c. 0.5 mm diam.; poll inici white, friable, mealy. Stigma more or less quadrate, 2-2.5 mm long, 2-2.5 mm wide, margins irregular, situated at base of column. Capsules obovoid, 10-16 mm long, 3.5-7 mm wide, erect, ribbed. (Fig. 1. e-h; Fig. 5) Specimens examined'. New South Wales: Cultivated CANB ex Orient Point, E of Nowra, 4 xi. 1988, D.L. Jones 3422 (CANB 8807085.1 & CANB 8807085.2). Victoria: Gippsland Lakes Coastal Park, 27. x. 1983, A.C. Beauglehole 74832 & W.R. Beauglehole (MEL 1531699 ); Gippsland Lakes Reserve, 27. x. 1984, A.C. Beauglehole 78814 & J.R. Turner (MEL 669448 ); Providence Ponds Flora and Fauna Reserve, 22. x. 1984, A.C. Beauglehole 78729 (MEL 670413)-, Gippsland Lakes Reserve, 27. x. 1984, A.C. Beauglehole 78799 & J.R. Turner (MEL 669443 & MEL 2069952)-, Lake Reeve, 1. xi. 1993, D. Rouse 4 (MEL spirit 2422A ); Golden Beach. SE edge of Lake Reeve, 4. xi. 1996, J.A. Jeanes 258 (MEL 2034950, MEL spirit 4397B): Rail Reserve WSW of Lindenow South, 27. x. 1999, J.A. Jeanes 633 (MEL 2069954). Distribution and habitat'. Eastern Victoria and the central coast of New South Wales. Most collections are from the eastern section of the Gippsland Plain Natural Region (Conn 1993), between Sale and Bairnsdale, Victoria, with a single disjunct collection from near the New South Wales coast, east of Nowra (Fig. 4). Grows in herb-rich grass- land and grassy woodland and heathland on soils ranging from sandy loams to clay loams. Altitude: 0-80 m. Conservation Status: Poorly known; suggest 3KC by criteria of Briggs & Leigh (1996). Flowering period: Late October to early November. Biology: This species is facultatively autogamous. Notes: Thelymitra planicola is a very distinctive species that is characterised by its overall glaucous appearance, moderately tall habit, large leaf-like lower sterile bracts, short pedicels, medium blue, longitudinally veined flowers mostly 20 to 25 mm across,' and the short, semi-cylindric column post-anther lobe. The structure of the column post- anther lobe is not shared with any other members of the genus Thelymitra, so identifica- tion of fresh and spirit-preserved specimens is relatively simple. In the dried and pressed state however, T. planicola may be difficult to differentiate from T. pauciflora, although the persistent glaucous appearance of the leaf, scape, bracts and ovaries, and the often leaf-like lower sterile bracts are useful distinguishing characters. The structure of the 96 J.A. Jeanes post-anther lobe of T. planicola is somewhat intermediate between those of Thelymitra aristata and T. pauciflora, but these two species have not been recorded growing sym- patrically with T. planicola, so the latter is unlikely to be of hybrid origin. Etymology : From the Latin planum, level ground, a plain; cola, inhabitant; in refer- ence to the lowland plain habitat favoured by this species. The following key is provided to distinguish the new species from other members of the Thelymitra nudal pauciflora complex in southeastern Australia. 1. Column post-anther lobe semi-cylindric, widely open on the posterior side; lower sterile bracts often large and leaf-like T. planicola 1 . Column post-anther lobe tubular, narrowly open qn posterior side; sterile bracts never large and leaf-like 2 2. Column post-anther lobe not inflated, more or less continuous with basal part of column 3 2. Column post-anther lobe inflated, discontinuous with basal part of column 4 3. Flowers 20-25 mm across; hairs on lateral lobes in a toothbrush-like arrangement.... T. nuda 3. Flowers usually <20 mm across; hairs on lateral lobes in a mop-like arrangement .... T. pauciflora 4. Plants short and stout, clump-forming; flowers vivid mauve or purple with prominent darker longitudinal veins; basaltic grasslands of western Victoria T. gregaria 4. Plants tall and slender, usually solitary; flowers pale to dark blue or pink, lacking prominent darker veins 5 5. Column post-anther lobe not dorsally compressed; perianth segments often >15 mm long; mostly from dry inland habitats T. megcalyptra 5. Column post-anther lobe dorsally compressed; perianth segments usually <15 mm long; mostly from moist near-coastal forests 6 6. Column post-anther lobe mostly glossy black with a yellow apex; hair tufts on col- umn lateral lobes white; sterile bracts usually 2 T. atronitida 6. Column post-anther lobe mostly brown with a yellow apex; hair tufts on column lat- eral lobes pink or mauve; sterile bracts usually 3 T. malvina Acknowledgments 1 am grateful to the Australian Orchid Foundation for their financial support of my work on Thelymitra during 1999. I wish to thank Dean Rouse for first bringing Thelymitra planicola to my attention in the early 1 990’s, and Allan Peisley and James Turner for sup- plying me with specimens of the new taxa. My colleague Neville Walsh (MEL) kindly wrote the Latin diagnoses and made helpful comments on the text. Thanks also to Enid Mayfield (MEL) for executing the line drawings, and the directors and curatorial staff at BRI. CANB. HO. NSW and MEL for access to specimens. References Backhouse, G.N. & Jeanes, J.A. (1995). The Orchids of Victoria. (Melbourne University Press, Carlton). Bates, R.J. & Weber, J.Z. (1990). Orchids of South Australia. (Government Printer, South Australia). Bernhardt. P. ( 1993). Thelymitra, in "Flora of New South Wales', vol. 4. ed. G.J. Harden (University of New South Wales Press, Kensington). New species of Thelymitra 97 Bishop, A (1996). Field Guide to the Orchids of New South Wales and Victoria. (University of New South Wales Press, Sydney). Briggs, J.D. & Leigh. J.H. (1996). Rare or Threatened Australian Plants. Revised Edition. (CS1R0 and Australian Nature Conservation Agency, Canberra). Conn. B.J. (1993). Natural Regions and Vegetation of Victoria, in ‘Flora of Victoria’, vol. 1. eds D.B. Foreman & N.G. Walsh. (Inkata Press. Melbourne). Jones, D.L. (1988). Native Orchids of Australia. (Reed Books, Frenchs Forest). Jones, D.L. & Clements, M.A. (1998). Contributions to Tasmanian Orchidology-8, Australian Orchid Research 3: 178-203. Jones, D.L., Wapstra. H. Tonelli. P. & Harris, S. (1999). The Orchids of Tasmania. (Melbourne University Press, Carlton). Weber, J.Z. & Entwisle, T.J. (1994). Thelymitra, in ‘Flora of Victoria’, vol. 2. eds N.G. Walsh & T.J. Entwisle. (Inkata Press, Melbourne). Muelleria 14:99 (2000) Spirogyra cardinia (Zygnemataceae, Zygnematophyceae, Chlorophyta): a new species of freshwater alga from Victoria, Australia. Simon H. Lewis 1 and Timothy J. Entwisle 2 1 National Herbarium of Victoria, Royal Botanic Gardens Melbourne, Birdwood Avenue, South Yarra 3141, Australia. 2 Royal Botanic Gardens Sydney, Mrs Macquaries Road, Sydney, 2000, Australia. Abstract Spirogyra cardinia (Zygnemataceae, Chlorophyta), a new species of freshwater alga from Victoria, Australia, in Section Conjugata (Vaucher) Hansg., Punctata- Group (Kadlubowska 1984) from the Cardinia Reservoir area in south-eastern Victoria, is described and illustrated. It is morphological- ly similar to S. bellis (Hassall) Cleve (which is in the Maxima- Group (Kadlubowska 1984)) and S. occidentalis (Transeau) Czurda (which is in the Punctata- Group (Kadlubowska 1984)) but differs in chloroplast number and vegetative filament and zygospore dimensions. Introduction Our recent census from literature records (Lewis & Entwisle 1998) accepted 31 species of Spirogyra Link from Australia, while Kadlubowska (1984), in the latest worldwide monograph of the genus, recognized 386 species in total. Since Kadlubowska’s mono- graph, more than 30 new species have been described from around the world, yet none were from Australia. This paper describes a new species discovered as part of Australia- wide collecting for the preparation of an account of Zygnemataceae for the Algae of Australia project. Methods Preparation of Material Specimens were preserved initially in 5% commercial formalin. After preparing dried specimens and permanent microscope mounts, the remaining material was transferred to 70% ethanol with 5% glycerol. For the microscope slides, specimens were stained with 1% aniline blue and placed in 10% Karo corn syrup with 0.25% phenol then mounted in 40% Karo corn syrup with 0.25% phenol. Taxonomy Spirogyra cardinia S.H. Lewis, sp. nov. Cellulae vegetativae 42-156 pm longae, 66-75 pm latae, parietes extremi plani; chloroplasti 1 (—2) anfracti 2-6-plo in cellulis; pyrenoides 5-15, 6 pm latae. Conjugatio scalariformis, canalibus ab utrinque cellulis formatis, extremitatibus tumidis. Gametangia (et cellulae sterilis aliquot) in latera conjugationis inflata (75-90 pm latae) abbreviata plerumque. Zygosporae ellipsoidea uniformes, 63-76 pm longae, 36-48 pm latae, exospora laeve incolorata, mesospora puncticulosa (punctula ad 1.5 pm lata), aurea. Type: Australia, Victoria, Narre Warren East, small pool near Kangaroo Flat Picnic Area, in Cardinia Reservoir Park. S.H. Lewis 403, 20. v. 1998 (holotype MEL 2046440: isotype NSW). Vegetative cells 42-156 pm long, 66-75 pm in diameter, length to diameter ratio = 1.15, with plane end walls; chloroplasts l(-2), making 2-6 turns; pyrenoids 5-15, up to 6 pm in diameter. Conjugation scalariform, conjugation tubes formed equally by both gametan- 100 S.H. Lewis and T.J. Entwisle gia and swollen where they meet; gametangia (and some sterile cells) inflated on the con- jugation side to 75-90 pm and often shorter than the vegetative cells. Zygospores ellip- soid, uniform, 63-76 pm long, 36^48 pm in diameter, length to diameter ratio = 1.78, exospore smooth and colourless, mesospore finely punctate and golden brown to yellow, pits to 1 .5 pm in diameter. Diagnostic Features-. Spirogyra cardinia belongs in the section Conjugata on the basis of its plane walls, and in the Punctata- Group on account of the mostly single chloroplast per cell and ornamented mesospore (Kadhibowska 1984). It differs from all other species in this group by having broader filaments (see Table 1 and Fig 1 ). In this group, S. cardinia is most similar to S. occidentalis (Transeau) Czurda in zygospore size (50-105 pm long, 36-61 pm in diameter in that species). However, S. occidentalis has much narrower vegetative cells (40-54 pm in diampter, although Czurda (1932 p. 1 83, quoted in Kadhibowska 1984, p. 342) describes ‘swollen barrel-shaped’ cells) and gametangia that are cylindric or only inflated up to 66 pm. The orientation of the zygospores in S. occidentalis is parallel to the main axis, while that of S. cardinia is at 45-90 degrees, with only some parallel. From the illustrations of S. occidentalis, the new species also appears to have considerably shorter fertile and sterile cells. Spirogyra cardinia is also comparable to S. bellis (Hassall) Cleve, from the Maxima- Group, a group comprising species with two or more chloroplasts and sculptured zygospore walls. The new species occasionally has cells with two chloroplasts, and S. bellis has similar vegetative filament dimensions. Spirogyra bellis , however, has 5-7 chloroplasts per cell and they are more spiralled (2-6 turns per cell cf. I ). It also has larg- er zygospores (64-105 pm long, 45-70 pm in diameter; Kadhibowska 1984) that are glo- bose or lenticular rather than ellipsoid, the conjugation tubes are not swollen at the ends, and the gametangia are usually swollen on both sides or enlarged. No other species in the Maxima- Group are likely to be confused with S. cardinia. Spirogyra rugulosa Iwanoff, in the Punctata- Group, is superficially very similar to the new species but the conjugation tubes are clearly formed by the male gametangia and the zygospores are broader. A recent collection {S.H. Lewis 718 (MEL)) from the type locality of Spirogyra cardinia has been referred to S. rugulosa. In this collection the wall was not as obviously punctate as that of the new species, although the vegetative fila- ments were similarly large. The conjugation tubes were, however, all definitely formed Table 1. Comparison of Spirogyra cardinia with similar taxa and groups mentioned in the text. Data from Kadhibowska 1984 and current study. Taxon Veg. cells diam. pm Zygospores diam. pm Zygospores length pm Chloroplast No. Tubes formed by S. cardinia 66-75 35 — 48 63-76 1 (—2) both gametangia / J M/ictata- Group 1 1-62 19-68 26-150 1 (—3) varies S. rugulosa Iwanoff (Kadhibowska 1984) 45-62 45-55 27-90 1 male gametangia .S', rugulosa (S.H. Lewis 718) 63-78 42-48 66-78 1 male gametangia S. occidentalis (Transeau) Czurda 40-54 36-6! 50-105 1-3 both gametangia S. bellis (Hassall) Cleve 65-80 45-70 64-105 5-6 both gametangia Spirogyra cardinia 101 Figure 1 a-c Spirogyra cardinia sp. nov.\ a. vegetative cells; b. zygospores showing mesospore ornamentation visible in transverse section (T) and surface view through exospore (S); c. conjugating cells. Scale: 40pm. by the male gametangia and therefore matched those of S. rugulosa , whereas the tubes of the new species were all obviously formed by both gametangia. Both of these populations warrant further study. Distribution and Habitat : Spirogyra cardinia is only known from the type collection. The pool was about 30 cm deep and 3 m in diameter and apparently permanent. Other collections from the same site, at the same time and at similar times in the following year, yielded a number of other species of Spirogyra, none of which resembled S. cardinia. Spirogyra cardinia was loosely associated with Callitriche stagnalis Scop., Cyperus era- grostis Lam. and C. lucidus R.Br., and with Nitella leptostachys A.Br. var. leonhardii (R.D.Wood) R.D.Wood. Etymology. The epithet refers to the location from which it was collected. Conservation Status: Using the criteria of Briggs and Leigh (1996) the Risk Code is assessed at IK. Based on current collecting it is clearly rare, however, as with all the ephemeral Zygnemataceae, it may be widespread. Making decisions regarding the con- servation status of algae in Australia is especially difficult due to a lack of distributional data, paucity of fertile (needed for identification) collections, changes to the fragile aquatic environments (Entwisle 1997) and the transient nature of Spirogyra. 102 S.H. Lewis and T.J. Entwisle Discussion Although this species is differentiated from the other species in the Punctata - Group (Kadfubowska 1984) by its vegetative cell dimensions, a character viewed as being influ- enced by polyploidy (Hoshaw et al. 1985, 1987; Hoshaw and McCourt 1988), this dif- ference appears to be of such magnitude that the entity deserves to be recognized taxo- nomically. Furthermore, it differs from the most similar species, such as Spirogyra bellis, S. rugulosa and S. occidentalis, by a number of characters apparently less influenced by polyploidy, such as spore, gametangia and tube shape, chloroplast number and mesospore ornamentation. Acknowledgments We would like to thank ABRS for funding the project, staff at MEL, Neville Walsh for correcting the Latin diagnosis and Emma Lewis for extensive support and help with field work. References Briggs, J.D., and Leigh, J.H. (1996). ‘Rare or threatened Australian plants', 1995 revised edn. (CS1RO: Collingwood.) Czurda, V. (1932). Zygnemales. In ‘Die Siisswasserflora von Mitteleuropa’ Vol. 9, (Ed A. Pascher) (Gustow-Fischer: Jena). Entwisle, T.J., (1997). Algae. In ‘A conservation overview of Australian non-marine lichens, bryophytes, algae and fungi". (G.A.M. Scott. T.J. Entwisle, T.W. May and G.N. Stevens, eds.), pp. 34-48 (Environment Australia: Canberra). Hoshaw, R.W., Wang, J.C., McCourt, R.M., and Hull, H.M. (1985). Ploidal changes in clonal cul- tures of Spirogyra communis and implications for species definitions. American Journal of Botany 72(7), 1005-11. Hoshaw, R.W., and McCourt, R.M. (1988). The Zygnemataceae (Chlorophyta) a twenty-year update of research. Phycologia 27(4), 51 1-48. Kadlubowska, J.Z. (1984). ‘Siisswasserflora von Mitteleuropa. Conjugatophyceae Chlorophyta VIII, Zygnemales' Band 16. (Eds H. Ettl. J. Gerloff, H. Heynig, D. Mollenhauer) (Gustav Fischer Verlag: Stuttgart, New York). Lewis, S.H.. and Entwisle, T.J. ( 1998). Zygnemataceae (Chlorophyta) in Australia: a Reassessment of Records and a Key to Accepted Taxa. Muelleria 1 1 , 51-93. Muelleria 14:103 (2000) Book Review Flora of Australia, volume 12, Mimosaceae (excl. Acacia ), Caesalpiniaceae. Editor P.M. McCarthy. Published by CSIRO, Melbourne, 1998; 213 pp.; hardcover edition (ISBN 0 643 06298 X) $AU 69.95; soft cover edition (ISBN 0 643 06299 8) $AU 54.95 This volume, presents all the Caesalpiniaceae and Mimosaceae, excluding Acacia, native and naturalised in Australia. The Caesalpiniaceae are represented in our flora by 22 gen- era and 127 species; the Mimosaceae by 17 genera and well over 1000 species. This lat- ter includes the most speciose of Australia’s genera, Acacia, not treated in this volume. Without Acacia, the family is but sparingly represented, with only 43 species. Most of these grow in the wet tropical north (especially Queensland). It is a disappointment to me that in this case The Flora of Australia is committed to following Cronquist’s system of classification of the legumes, splitting them into three families. This, despite the confession of two of the contributors to this volume, that most specialists who study the group would include them in one family. Simply put, there are no characters that support Cronquist’ s classification. In fact two genera, Pentaclethra and Dimorphandra, though put in two different families by Cronquist (Mimosaceae and Caesalpiniaceae respectively) are more similar to each other than they are to any other member of their respective families. Though only a flora series, I think it unfortunate in this case that it was not decided to allow the nomenclature used to reflect our increased understanding of the systematics of the group, and to include all the legumes in one fam- ily. The editors have, in effect, perpetuated what appears to this legume researcher to be an anachronistic position. Though this is an important criticism to me, it in no way dis- tracts from the practical value and presentation of this volume. Seven authors contributed treatments to this book, and it is a credit to the editorial staff that they all conform to the same standard. In an afternoon in the herbarium using the keys with herbarium specimens I discovered no errors in fact, and hardly any diffi- culty in using them. Each couplet in the keys I used worked, and in most instances the authors are to be commended for using vegetative and reproductive characters in their couplets, thereby allowing one to identify specimens that might be fruiting or sterile. Even though the descriptions are short, as is perhaps necessary in this series, in every one that I had to refer to, enough information was provided to clearly determine a speci- men not determined for certain using the key. I am sad to say that I did find some of the illustrations to be rather coarse and unap- pealling. I could not tell if this was the lack of finesse on the part of the artist, or if the line drawings were too reduced in reproduction. The line drawings are supplemented with some beautiful colour photographs (32 of mimosoids, and 32 of caesalps, each in eight plates). These photos are a wonderful addition, and bring to the fore characters that can- not be appreciated in reading the descriptions. For instance, the photo of the fruits of Archidendron lucyi (Fig. 21, p. xviii), Aa. hendersonii and ramiflorum (Fig. 19, 20, p. xvii) are stunning. Not only because of the unusual beauty of the fruit - all bright red with contrasting, dark seeds - but also because I was struck with the similarity of this fruit to some American genera of tribe Ingeae, like Pithecellobium, or Abarema ; genera with which I am much more familiar. No description had ever conveyed that to me. At first I was put off by the treatment of Senna, with its form taxa. Having known that B. Randell and B. Barlow had both worked on the groups for quite some time, I expect- ed more. In hindsight though, after curating the collections at MEL, I see my attitude was a bit naive. Though I would have theoretically admitted that there are groups so tokoge- netically complex as to defy simple treatment, I had no personal experience with one. I see now, though, that Senna is such a group, and the treatment by Randell and Barlow seems rational and practical. 104 James Grimes The contribution of the Mimosaceae (excluding Prosopis) presents Richard Cowan’s last contribution to plant systematics. I know that Dick thoroughly enjoyed his work on the Flora of Australia, and his treatments almost vibrate with the enthusiasm with which lie approached his work. The work has obviously benefited from Dick’s experience with American Mimosaceae as the groups treated here, especially those in tribe Ingeae, have close relatives in the America's, and Dick’s familiarity with those groups adds a certain soundness to his treatments. He is sorely missed. All in all. this volume is everything it should be: sound, complete, and very practical. I can see that it will be well used by me. James Grimes CONTENTS Volume 14, 2000 Contributed Papers Page The Taxonomy of Boronia anemonifolia and B. rigens ( Boronia sect. Cyanothamnus , Rutaceae) — Peter G. Neish and Marco E Duretto 3 The lichens of Nothofagus cunninghammii- dominated and Acacia melanoxylon-dominated forests in the Otways, Victoria. ^ — Sharon Ford, Maria Gibson and Geoff Duke 17 A note on Lepidium strictum (S. Watson) Rattan (Brassicaceae) in Victoria — Neville Scarlett 31 The stomata of bluegums (Eucalytpu s spp.) — Denis J. Carr 33 Identified types in the Wilhelm Hillebrand collections in the herbarium of the Royal Botanic Gardens, Melbourne — James W. Grimes 41 Additions to the Hygrocybeae of Victoria. I. — A. M. Young 51 A revision of Agrostis billcirdierei R.Br. (Poaceae) — A.J. Brown and N.G. Walsh 65 Two new species of Thelymitra (Orchidaceae) from southeastern Australia — J effrey A . Jeanes 91 Spirogyra cardinia (Zygnemataceae, Zygnematophyceae, Chlorophyta): A new species of freshwater alga from Victoria, Australia — Simon H. Lewis and Timothy J. Entwisle 99 Book Review Flora of Australia, Vol. 12 — James W. Grimes 105