SO llth African Jovmal of Botany 2001 67 65 - 68 Copynght © NISC Ply Ltd Pnnted m South Afnca - All fights reSBfYed SOUTH AFRICAN JOU n NAL Of BOTANY /SSN 0254-6299

Observations on hairs in the capitula of some southern African genera

PPJ Herman

National Botanical Institute, Private Bag X101, Pretoria 0001, South Africa e-mail: [email protected]

Received 29 May 2000, accep ted in revised form 26 September 2000

Septate hairs are recorded and described for the first the hairs have perpendicular cell walls. In sub­ time from the capitula of various southern African aster· Sphenogyne the septate hairs also have oblique aceous genera . Such hairs are present on the paJeae of cell walls and, in addition, spiral wall thickenings. Eriocephalus L. , the cypselas of Tarchonanthus L. , Multicellular, biseriate hairs mixed with some septate Lasiospermum Lag., Arctotis L. and Arctotheca JC hairs were observed in Arctotheca. The basal coma of Wendl. and the receptacle of Lidbeckia PJ Bergius. The the cypselas of Arctotis and Less. con­ basal coma of hairs on the cypselas of Ursinia Gaertn. sists of twin hairs. Septate hairs can be considered a subgenus Sphenogyne (R.Br.) Prassler also consists of generic characteristic of the genera Tarchonanthus and septate hairs. The hairs in Eriocepha/us, Arctotheca and possibly of the genera Eriocephalus, Tarchonanthus, Lasiospermum, Arctotis and Lasiospermum, Lidbeckia and Ursinia subgenus Arctotheca have obliqu.e cell walls while in Lidbeckia Sphenogyne.

Introduction

The cypselas of all the species of Lasiospermum Lag. and cypselas. The cypselas of the genus Arctotheca contain sim­ Tarchonanthus L. are enveloped by long , silky hairs, while ilar pilose hairs as Lasiospermum. Ursinia , Arctotis L. an d similar hairs are present on the paleae of all species of Haplocarpha Less. have a basal coma of hairs on their Eriocephalus L. The genera Eriocephalus and cypselas, but they belong to different tribes ( Lasiospermum were revised by MOil er (1988) at the Nation al and Arctoteae respectively). Almost all th e genera belonging Herbarium of Namibia (WIND) but the results of his study to the tribe Arctoteae have hairY cypselas (Bremer 1994). It had not been published by the time he died tragically in a was decid ed to investig ate these hairs in more detail in motor ca r accident during April1 gg7. Permission was grant­ some southern African genera from these tribes. ed by the University of Stellenbosch for publication of the taxonomic part of his thesis by Ms H. Kolberg (WIND) and Materials and Methods myself as it was considered to be a valuable contri bution to in so uthern Africa (MOi ler et al. in press) . Capitu la an d/or cypselas were obtained from herbariu m The Tarchonanthus camphoratlls L. complex in southern specimens housed in PRE (Table 1). A full list of taxa and Africa is currently being revised by myself (Herman in prep.). speci mens studied is available from the author. For quick While identifying and sorting the ta xa of these three genera, sca nning , the material was rehydrated in 'Windolene' and it was observed th at the hairs on the cypselas of the hairs temporarily mounted on microslides for Lasiospermum and Tarchonanthus species and on the microscopy. For photomicrography the material was soaked paJeae of Eriocephalus species were septate. The presence in boiling water, dissected and hairs from around the cypse­ of these outgrowths raised th e following questions: (1) Can las (Lasiospermum , Tarchonanthus and Arctotheca) , from septate hairs be considered a generic character in the basal com a (Haplocarpha and Ursinia) , the receptacle Eriocephalus, Lasiospermum and Tarchonanthus L.? (2)Are (Udbeckia) or from the outer paleae (Eriocephalus) were the long hairs on cypselas of other southern African genera, stained in toluidene blue for about 1 to 2 minutes. The e.g. Ursinia Gaertn. and Arctotheca J.CWendl. also sep­ preparations were washed in water and temporaril y mount­ tate? ed in a drop of water. Bl ack-and -white photom icrographs Apart from Er;ocephalus and Las;ospermum, the southern were taken with an Ol ympus-Va nox-S microscope using African genera Ursinia, Thaminophyfium Harv., Artemisia L" IIford Panf film with ASA 50. Inezia E Phillips, Udbeckia PJ Bergius and Cotula L. , in the It must be stre ssed that the ontogeny of th e hairs trtbe Anthemideae, were listed by Bremer and Humphries described here was not stUdied. (1 993) and Bremer (1994) as ha ving pilose receptacles or 66 Herman

Table 1: List of genera and number of taxa examined

Genus No. of taxa No. of taxa No. of specimens Literature reference for number of taxa in genus examined examined In southern Africa or subgen us Arctotheca JC Wendl 4 4 24 Bremer (1994) A/clotis L 64 3 11 Herman et 81 (1997) Eriocephafus L 26(37) (33) 230 Bremer (1 994). Muller el al. (in press) Haplocarpha Less 5 3 25 Herman et al. (1997) Lasiospermum Lag 4 2 20 Bremer (1994), Muller et al. (In press) Udbeckl8 PJ Berglus 2 1 3 Sremer (1994) Ta rc /Jonanthus l. 3(7) 3(7) 261 Paiva (1 972), Herman (in prep ) Ursmia Gaertn 34 20 108 Prassler (1967) subgenus Sphenogyne (R Sr.) Prassler

Obse rvations hairs , where the upper cell dominates the lower cell . His the­ ory did not receive much attentio n as it was not reported In all the inves tigated species of Eriocephalus, afterwards in literature dealing with Lasiospermum . Lasiospermum , Tarchonanthus, Ursinia and Aretatl7eca the Beauverd (1915) described spiral wall thickenings in hairs hairs consisted of a long , uniseriate row of cells (Figure 1A). of Ursinia and Lasiospermum , but mentioned nothing about The cell walls between cells were very oblique (Figure 1B) septa , Faint but dense spiral wall thickenings were some­ and sometimes even 'footlike' (Figure 1C) . Similar septate times present in the septate hairs of Lasiospermum. Spiral hairs were observed in Udbeckia quinqueloba (U.) Casso wall thickenings in the coma of hairs at the base of the but the transverse cell walls were perpendicular to the outer cypselas in Ursinia species were reported by Hess (1938) walls of the hairs (Figure 1D). In Ursinia spiral wall thicken­ for one species (u. anthemoides (L.) PO ir.), but no septate ings also occurred (Figure 1 E), while in Lasiospermum faint hairs were recorded. Prassler (1967) noted the coma of but dense spiral wall thickenings were sometimes present hairs at the base of the cypselas in the subgenus (Figure 1F) . In Haplocarpha twin hairs were observed Sphenogyne of the genus Ursinia but did not mention septa (Figure 1G ) but no septate hairs, whereas in Arctotis both or spiral wall thickenings in these hairs. Robinson and twin hairs and some septate hairs were seen . In Arctotheca Brettell (1973) described the pappus of Ursinia as having calendula (L.) Levyns numerous, multicellular, biseriate spiral wall thickenings. This must be a mistake as the pap ­ hairs were present together with the septate hairs . In these pus of Ursinia consists of scales (Prassler 1967). True pap­ hairs the cell apices diverged slightly so that the next two pus bristles never occu r in the tribe Anthem ideae as in other cells arose between them (Figure 1 H). In the other tribes (Bremer and Humphries 1993). It could be that Arctotheca species studied , these hairs were fewer and Robinson and Brettell actually referred to the basal coma of mixed with some septate ha irs . hairs . Bremer and Humphries (1993) described 'rather stiff. unbranched hairs of 3-8 cells with spiral wall thickenings' Discussion from the cypselas of Microcepha/a Pobed ., a genus from Asia, Iran, Afghanistan and Pakistan. This could perhaps be The tribe Anthemideae (s ubfamily As teroideae) similar to the hairs from the basal coma of cypselas from The genera Eriocephalus, Lasiospermum, Ursinia and Ursinia described here, but no mention was made about the Udbeckia belong to the tribe Anthemideae of the subfamily kind of cell walls and no illustration was provided. (Bremer and Humphries 1993, Bremer 1994). In Lidbeckia septate hairs are rare, very few having been Most species of these genera are southern African found on the receptacle of only one species, L. quinque/oba endemics (Bremer and Humphries 1993). (3 specimens). The receptacie of the other species, L. pecti­ MOiler (1988), in his unpublished version of the rev ision of nata PJ Bergius , was found to be glabrous. The perpendi­ the genera Eriocephalus and Lasiospermum , did not refer to cular cell walls of the septate hairs in L. quinqueloba could septate hairs or spiral wall thickenings in the capitula of be used in distingu ish ing between genera and/or species. these genera, although much attention was given to the The other southern African genera in the tri be ontogeny of the capitula. Hess (1938) referred to the septate Anthemideae described by Bremer and Humphries (1993) hairs of Lasiospermum radiatum Trevir. as modified twin and Bremer (1994) as having pilose receptacles or cypselas

Figure 1: Light micrographs of septate hairs (A) Hair consisting of rang. uniseriate row of cells from a cypsela of Tarchonanthus tnlobus DC. var. galpinii (Hutch . & E.Phillips) Paiva (Smuts 13, PRE), Scale = 1001Jm ; (8) Oblique cell wall In a hair from a cypseJa of Tarchonanthus cam­ phora/us L (Acocks 2352, PRE) Scale = 10J.lm; (C) Foollike cell wall tn a hair from a cypsela of a Tarchonanthus sp_ (Pole-Evans 2495(42), PRE), Scale = 10IJm (O) Perpendicular cell wall in a hair from th e receptacle of Udbeckia quinque/oba (L.t.) Casso(G alptn 10578, PRE). Scale = 10IJm (E) Spiral wall thickenings and septum in a hair from the basal coma of hairs of a cypsela of Ursinia heterodonta (DC.) N.E.Br. (KurzwetI1785, PRE), Scale = 10J.lm (F) Faint but dense spiral wall th ickenings and oblique cell wall in a hair from a cypsela of Lasiospermum brachyglossum DC. (Burger and Louw 299, PRE), Scale = lO~m , (G) Twin hairs from the basal coma of hatrs of a cypsela of Haplocarpha scaposa Harv. (Turner 837, PRE), Scale = 1 OO~m (H) Multicellula r, biseriate hair with two cell apices diverging slightly and the next two cells arising between them, from a cypsela of Arctot/Jeca calendula (L.) Levyns (Meyer 7372, PRE), Scale = 100J.lm South Afrtcan Journal of Botany 2001 . 67 65·68 67

I ! ,.1. .,

c

B o

. 1 .; !" ~ . i'i" ,i ,: F \) 'i I i • I I' ~il I I'J . I ! I E · H I/ I • 68 Herman

are Thaminophyllum, Artemisia, Inezia and Cotula. Brusse References (1989) described the hairs on the corolla and cypselas of Inezia. They are not septate as described here for the other Beauverd G (1915) Contribution a l'etude des Composees. Bulletin genera. Due to in sufficient herbarium material in PRE . de la Societe Botanique de Geneve (second series) 7' 21 - 56 T/laminophyllum was not studied . The South African repre­ Beentje HJ (1999) Th e genus Tarchonanthus (Compositae­ sentative of Artemisia, A. afra Jacq. ex Willd. has glabrous Mulisleae) Kew Bulletin 54: 81-95 cypselas and was therefore not included in this study. Th e Bremer K (1994) Asteraceae, cladis tics and claSSification. Timber Press, Portland. Oregon . ISBN 0-88 192-275-7 genus Cotula was also not included in this study. Bremer K, Humphnes C (1993) Generic monograph of the Asteraceae-Anthemldeae Bulletin of the Natural History Museum Tribe or tribe (subfamily (Botany Seri es) 23 ) Brusse F (1989) Asteraceae. A new species of Inezia (Anthemideae) from the north-eastern Transvaal. Bothalia 19' The genus Tarchonanlhus belongs to the tribe Mutisieae 27-29 (Bremer 1994) or the tribe Tarchonantheae (Keeley and Herman PPJ (in prep.) The Tarchonanthus camphoralus complex Jansen 1991) of the subfamily Cichorioideae (Bremer 1994). (Asteraceae-Tarchonanlheae) in southern Africa. revised In his revision of Tarchonanthus, Paiva (1972) did not Herman PPJ, Weiman W, Retief E, Koekemoer M (1997) mention anything about septate hairs in the capitula, as Asteraceae. In . Meyer NL, Messmer M, Smith GF (eds) described here; neither did Pope (1992) or Beentje (1999) in Taxonomic literature of southern African . StreJilzia 5: their treatments of the genus. 50-60. Nationa l Botanlcallnslitule, Pretoria ISBN 1-919795-51- In Karis et al. (1992) simple, multicellular hairs with o obli que cell walls, similar to those observed during this Hess R (1938) Vergleichende Untersuchungen uber die Zwillingshaare der Compositen. Botanische JahrbOcher 68: study, were illustrated and mentioned for a number of gen­ 435-496 era in the tribe Mutisieae of the subfamily Cichorioideae, but Karis PO, KaliersJo M, Bremer K (1992) Phylogenetic analysis of the not for Tarchonanthus as observed during this study. This Cichorioideae (Asteraceae), with emphasis on the Mutisieae. seems to be the first report of septate hairs in the genus Annals of the Missouri Botanical Garden 79: 416-427 Tarchonanthus. Keeley SC , Jansen RK (1991) Evidence from chloroplast DNA for the recognition of a new tribe, the Tarchonantheae, and the tribal Tribe Arctoteae (subfamily eichariaideae) placement of Pluchea (Asteraceae). Systematic Botany 16: The genera Arctoris, Arctotheca and Haplocarpha belong to 173-181 the tribe Arctoteae of the subfamily Cichorioideae (Bremer MOiler MAN (1988) 'n Mortologiese en taksonomiese studie van die 1994). genusse Lasiospermum Lag. en Eriocephalus L. (Asteraceae) in In Haplocarpha the basal coma of hairs on the cypselas suidellke Afrika Unpublished PhD Ihesis. University of consists of the well documented twin hairs (Hess 1938, Stell en bosch, Stellenbosch Bremer 1994) but no septate hairs, as is the case in Ursinia Mu ller MAN, Herman PPJ , Kolberg HH (i n press) Eriocepha/us and of the tribe Anthemldeae (subfamily Asteroideae). In the Lasiospermum Flora of southern Africa 33 Paiva JAR (1972) New and httle known species from the Flora Arctotis species studied, the basal coma of hairs also con· zambesiaca area. 22. Notes on Inuleae. Boletim da Sociedade sisted of twin hairs but the hairs from the cypselas were sep­ Bro\eriana 46 (2nd series). 355-381 tate . In Arctotheca the septate hairs from around the cypse­ Pope GV (1992) 97. Compositae. Flora zambesiaca 6: 9- 11 las were mixed with multicellular, biseriate hairs. These hairs Prassler M (1 967) Revision der Gattung Ursinia. Mitteitungen der were not seen in any of the other preparations. Botanischen Staatssammlung MOnchen 6. 363-478 Rob inson H, Brettell RD (1973) Tribal revisions in the Asteraceae. Acknowledgements - Dr E. Steyn is thanked for assistance with Vl11 A new tribe, Ursinieae Phylologia 26: 76-85 procedures and valuable comments on the manuscript and Mrs A Romanowski and Mrs S. Turck for technical assistance.

Edited by GF Smith