<<

Bothalia 43,2 (2013) 211 the following herbaria for loans and access to their col- GOLDBLATT, P., PHILLIPSON, P.B. & MANNING, J.C. 2013. The lections: BOL, BR, K, NBG. We also thank Clare Archer new species Aristea farafangana () from Madagascar, biogeographic notes on Aristea in Madagascar and a revised for checking holdings at the Pretoria herbarium (PRE) key to the . Adansonia [doi: http://dx.doi.org/10.5252/ for us. a2013n1a4], sér 3,3: 47–53. HEPPER, N. 1968. Iridaceae. Pp. 138–142 in Flora of West Tropical Africa, ed. 2, 3(1). Crown Agents for Oversea Governments and REFERENCES Administrations, London. HUTCHINSON, J. 1939. XXVIII—Tropical African : XVII. Bul- BAKER, J.G. 1877. Systema Iridearum. Journal of the Linnean Society, letin of Miscellaneous Information, Royal Botanical Gardens, Botany 16: 61–180. Kew. 1939: 241–247. BAKER, J.G. 1906. II.—Diagnoses Africanae, XVI. Bulletin of Mis- HUTCHINSON, J. & DALZIEL, J.M. 1936. Iridaceae. P. 374–376 in cellaneous Information [Royal Botanic Gardens, Kew] 1906: Flora of West Tropical Africa 2,2. Crown Agents for the Colo- 15–30. nies, London. CHEVALIER, A. 1920. Exploration botanique de l’Afrique occidentale KRAUSS, F. 1845. Pflanzen des Cap- und Natal-Landes, gesammelt un française 1. Enumeration des plantes récoltées. 1. Paris, Leche- zusammengestellt von Dr. Ferdinand Krauss. Flora 28: 305–315. vallier. VINCENT, L.P.D. 1985. A partial revision of the genus Aristea (Iridace- DE WILDEMAN, E. 1913. Decades novarum specierum florae katan- ae) in South Africa, Swaziland, Lesotho, Transkei and Ciskei. gensis. 1–VII. Repertorium specierum novarum regni vegetabi- South African Journal of Botany 51: 209–252. lis 11: 501–524. WEIMARCK, H. 1940. Monograph of the genus Aristea. Lunds Uni- GEERINCK, D. 2005. Iridaceae. Flore d’Afrique Centrale. Jardin Bota- versitets Årsskrift N. F. Avd. 2, 36: 1–140. nique National de Belgique, Meise. GOLDBLATT, P. 1989. Revision of the tropical African genus Zygotri- P. GOLDBLATT1,3, J.C. MANNING2,3 tonia (Iridaceae). Bulletin du Museum d’Histoire Naturelle, 4 sér., sect. B, Adansonia 11: 199–212. 1 GOLDBLATT, P. 1993. Iridaceae. Flora Zambesiaca 12, 4. Flora B.A. Krukoff Curator of African Botany, Missouri Botanical Garden, P. Zambesiaca Managing Committee, London. O. Box 299, St. Louis, Missouri 63166, USA. E-mail: peter.goldblatt@ GOLDBLATT, P. 1996a. Iridaceae. In R.M. Polhill (ed.), Flora of Tropi- mobot.org. cal East Africa. A.A. Balkema, Rotterdam. 2Compton Herbarium, South African National Biodiversity Institute, GOLDBLATT, P. 1996b. Gladiolus in Tropical Africa. Timber Press, Private Bag X7, 7735 Claremont, Cape Town. E-mail: j.manning@ Portland, OR. sanbi.org.za. GOLDBLATT, P. 1997. Iridaceae. In S. Edwards, S. Demissew & I. Hedberg (eds), Flora of Ethiopia 6: 162–183. University of Addis 3Research Centre for Growth and Development, School of Life Ababa, Addis Ababa. Sciences, University of KwaZulu-Natal, Pietermaritzburg, Private Bag GOLDBLATT, P. 2012 [2013]. Aristea (Iridaceae: Aristeoideae), a sub- X01, Scottsville 3209. generic classification.Novon 22,4 (in press). MS. received: 2012-12-10

IRIDACEAE

A NEW SPECIES OF HESPERANTHA FROM THE OVERBERG, WESTERN CAPE, AND OBSERVATIONS ON A NOVEL MODE FOR POLLEN TRANSFER IN THE GENUS AND FAMILY BY A HESPERID BUTTERFLY

INTRODUCTION observation for pollinators have shown that the flow- ers, which begin to open at about 6:00 and close after Plants of an unknown species of Hesperantha Ker 12:30, are visited and pollinated by the small butterfly Gawl., discovered in the Rûens Hills northeast of Tsitana tulbagha (Hesperidae), not before recorded as an Bredasdorp, Western Cape, South Africa, present some agent for pollen transfer in any sub-Saharan species of unusual features in the genus. Hesperantha, a sub-Saha- Iridaceae (Goldblatt & Manning 2006). We describe the ran African genus of some 85 species of Iridaceae sub- new species here, discuss its relationships within Hes- family Burnett, extends from Western Cape perantha, and present evidence for this novel pollination to Ethiopia and Cameroon (Goldblatt 2003; Goldblatt system within Iridaceae. & Manning 2007; Goldblatt & Manning in prep.). The genus is recognized by ± woody corm tunics and a spi- cate inflorescence of radially symmetric flowers with TAXONOMY the style branches dividing close to the mouth of the perianth tube into long, spreading style branches. The Hesperantha kiaratayloriae Goldblatt & new species, H. kiaratayloriae, has diurnal, small pink, J.C.Manning, sp. nov. salver-shaped flowers that are remarkable for the erect stamens tightly enclosing the exserted style. The style TYPE.—Western Cape, 3420 (Bredasdorp): Rûens divides well above the mouth of the perianth tube near Hills, northeast of Bredasdorp, Farm Plaatjieskraal, the apices of the anthers, thus unlike any other species upper slopes of ridge in stony quartzite gravel over in the genus, and the three style branches, which are shale, (–AC), 8 Nov. 2011, Goldblatt & Porter 13729 loosely twisted or coiled around one another, are held (NBG, holo.; K, MO, PRE, iso.). well above the mouth of the perianth tube. Although the Plants mostly 150–200 mm high, rarely to 300 mm, exserted style and erect style branches are unique for the erect, unbranched. Corm obliquely bell-shaped, 6–8 genus, the bell-shaped corms, with an oblique flat base mm diam. at base, tunics ± woody, dark grey, fragment- and floral bracts united basally around the spike, are ing from lower margins into tile-like segments. Leaves consistent with sect. Radiatae Goldblatt. Three days of several, dry at flowering time, sheaths imbricate, blades 212 Bothalia 43,2 (2013) linear, mostly 10–60 × 1.0–1.5 mm, margins smooth, uppermost 2 or 3 leaves sheathing for entire length. Spike 2–4(–7)-flowered, flexuose; bracts dry and pale brown at flowering, outer mostly 12–15 mm long with margins connate near base for 1–2 mm and sheath- ing spike axis, inner slightly shorter than outer and forked apically. Flowers diurnal, pink with white throat, unscented, anthers bright yellow and prominently dis- played; perianth tube cylindric, slightly expanded at apex, 13–16 mm long, distal half exserted from bracts; tepals subequal, spreading horizontally, elliptic, ± 12 × 4–5 mm, outer slightly larger than inner. Stamens sym- metrically disposed, erect; filaments 2.5 mm long, exserted ± 1 mm from mouth of tube; anthers ± 5 mm long, dehiscing longitudinally, yellow, remaining coher- ent around style. Ovary ovoid-oblong, ± 6 mm long; style exserted and dividing just below anther apices, branches ± 5 mm long, remaining erect, slightly twisted and often interlaced, papillate and stigmatic in upper 2/3. Capsules subcylindric, 7–8 mm long. Seeds angu- lar, ± 0.7 × 0.4 mm, with membranous raphal ridge expanded into a wing at micropylar end, surface cells colliculate. Flowering time: (late October) November. A Figure 1. Distribution & ecology: known from a single south- B facing ridge in Eastern Rûens Shale Renosterveld (Mucina & Rutherford 2006), ± 40 km northeast of Bredasdorp (Figure 2), Hesperantha kiaratayloriae grows among silcrete-quartzite stones over a shale sub- strate among tufts of the graminoids Pentachistis and Merxmuellera (Poaceae) and Bobartia (Iridaceae), and the shrubs Elytropappus rhinocerotis (L.f.) Less. (Asteraceae), Amphithalea violacea (E.Mey.) Benth and Aspalathus mundiana Eckl. & Zeyh. (Fabaceae). It shares this unique habitat with several threatened quartz- specialists, including Gibbaeum haagei Schwantes (Aizoaceae: Endangered), Elegia verreauxii Mast. (Res- tionaceae: Vulnerable), Notobubon striatum (Thunb.) Magee (Apiaceae: Near Threatened), Otholobium cur- tisiae C.H.Stirt. & Muasya ined. (Fabaceae, proposed status Endangered), as well as the recently described Xiphotheca rosemarinifolia A.L.Schutte (Fabaceae: Critically Endangered), which is known only from this single locality. We estimate that the population com- prises about 90 plants of flowering size. The site was not examined for juveniles. FIGURE 1.—Hesperantha kiaratayloriae, Goldblatt & Porter 13729. Although there is thus a good probability that H. A, ; B, detail of stamens and stigma. Scale bar: A, 10 mm; B, 2 mm. Artist: John Manning. kiaratayloriae may likewise be restricted to this single locality, additional surveys are required to establish this. Silcrete-quartz outcrops are restricted to Eastern Rûens Shale Renosterveld, a Critically Endangered vegetation early summer flowering. The pink, unscented flowers type, with < 12% remaining (SANBI & DEAT 2009). last three days (determined by maintaining cut stems in Renosterveld remnants are vulnerable to mismanage- water in the laboratory), and open early in the morning ment, including inappropriate use of fire and over-graz- each day, ± 06:00, beginning to close after 12:30 and are ing and trampling by livestock (Schutte-Vlok 2011; O. fully closed by 13:30. Curtis pers. obs.). Given the evident rarity of the spe- cies and the threats from agriculture to the single known Diagnosis and relationships: a small plant, Hesper- locality, which remains unprotected, we propose a con- antha kiaratayloriae has small pink flowers with a rela- servation status of Critically Endangered (CE). To date, tively long perianth tube, 13–16 mm long and slightly longer than the tepals, erect stamens with the anthers no seeds have been collected for seed-banking. connate around the style, and an erect style that divides The leaves of Hesperantha kiaratayloriae are dry and opposite the anther apices into three erect, slightly withered at flowering, a feature not known elsewhere in twisted branches. The short, linear leaves are ± dry at the genus and evidently associated with its late spring / flowering and the floral bracts are united for 1–2mm Bothalia 43,2 (2013) 213 evidently foraging for nectar present in the lower part of the perianth tube (nectar concentration and chemis- try not examined, but we confirm the presence of nectar, sweet to the taste). Nectar in Hesperantha, as in other genera of Iridaceae: Crocoideae, is secreted from septal nectaries, but we did not confirm presence of these nec- taries in this species. During foraging activity, the frons of the insects (but not other body parts) became visibly covered with a dense accumulation of bright orange pol- len from contact with the anthers. Butterflies were seen flying from one open flower to another and stigmas that were devoid of pollen before visits by T. cf. tulbagha were seen to bear heavy deposits of pollen after visits by the butterfly. The butterfly was identified by entomolo- gists, M. Picker and D. Edge, from photographs. No insects were trapped and killed for vouchers. Our observations show that this butterfly is an active and effective agent for pollen transfer for Hesperantha kiaratayloriae. We also note that we observed no other insect visitors to the flowers either in the early morning FIGURE 2.—Distribution of Hesperantha kiaratayloriae. or later (three to four hours of observation on different days). From this we then infer that Tsitana cf. tulba- gha is at least one, but possibly the sole, pollinator of around the spike axis. The general aspect of H. kiara- the species at this locality. The possibility of autogamy tayloriae is that of sect. Radiata, diagnosed by the flat- in this species can be eliminated because the flowers are based corms with hard tunics fragmented at the base and herkogamous, the pollen and stigmatic part of the style the partial union of the outer floral bract margins around branches are physically separated from one another; the the spike axis (Goldblatt 2003). A well developed peri- latter held about 1.5 mm above the anthers and exposed anth tube is common in the genus and not unusual in pollen. The species may also be self-incompatible, but sect. Radiata, but pink flowers are otherwise known in this was not determined for our study. the section only in H. elsiae Goldblatt from the south- ern Cedarberg, which in other features, including a long Butterfly pollination is evidently rare among south- perianth tube, is very different to H. kiaratayloriae ern African Iridaceae apart from the guild of species (Goldblatt 1984). The southern Cape H. muirii (L.Bolus) with large red flowers, mostly species of Gladiolus and G.J.Lewis is most like H. kiaratayloriae in the twisted Tritoniopsis, pollinated predominantly by Meneris tul- style branches (although spreading in the former) and baghia (Satyridae) (Goldblatt & Manning 2002, 2006), we see the two as a species pair within the section. The a very different system to that reported here. Pollina- relatively long style enclosed by the erect stamens and tion by the painted lady, Cynthia cardui (Pieridae), has dividing well above the mouth of the perianth tube close been observed in yellow-flowered Ixia acaulis Gold- to the anther apices significantly expands the circum- blatt & J.C.Manning and Nivenia parviflora Goldblatt scription of Hesperantha. & J.C.Manning, and this butterfly is the only recorded visitor to I. acaulis (Goldblatt & Manning 1993, 2006, The species is named for Kiara May-Leen Taylor in 2011). Several other species of Iridaceae are visited by honour of the generous donation to renosterveld conser- butterflies as part of a generalist pollination system also vation by her father, Oren Taylor. involving nectar-feeding bees and sometimes hopliine beetles (Sarabaeidae: Hopliini). These include I. orienta- lis L.Bolus, reported as pollinated by the butterfly Colias POLLINATION electo (Pieridae) (Goldblatt et al. 2000), and Micranthus Only a single insect species, the skipper butterfly Tsi- spp., visited by Cynthia cardui, Pieris helice (Pieridae) tana, cf. T. tulbagha (Hesperidae), a relatively common and Colias electo, but these plants are generalists and are species of southern Africa, was seen visiting flowers of also visited and pollinated by hopliine beetles and large Hesperantha kiaratayloriae over three days of observa- anthophorine bees (Goldblatt & Manning 2006). Colias tions, for three to four hours each day. The bright pink electo has also been captured while visiting flowers of flowers evidently lack floral odour as determined from Geissorhiza foliosa Baker and G. heterostyla L.Bolus, open flowers held indoors in a warm, still room. Sev- but its role in the pollination of these relatively short- eral (< 10 individuals) Tsitana butterflies (sexes not tubed species is uncertain (Goldblatt & Manning 2009). established) were observed visiting the flowers after Other Hesperantha species are pollinated by settling 6:30, soon after the perianth expanded fully, moving moths (mostly Noctuidae), long-proboscid flies (mostly directly from one opened flower to another. On alight- species of Prosoeca) or large-bodied bees and Apis ing on a flower, the frons of the insect first contacted the (Goldblatt et al. 2004). Just one species is known to be style branches, the upper two thirds of which are ciliate, pollinated by hopliine beetles. Pollination of H. kiara- sticky and stigmatic throughout the lifespan of a flower. tayloriae thus represents a novel pollination strategy in The insect then inserted its proboscis into the floral tube, the genus. 214 Bothalia 43,2 (2013)

ACKNOWLEDGEMENTS fall zone, nomenclatural changes, range extensions and notes on pollen morphology and floral ecology. Bothalia 39: 123–152. We extend our thanks to Lendon Porter for his assist- GOLDBLATT, P. & MANNING, J.C. 2011. Systematics of the southern ance and companionship in the field and Cameron and African genus Ixia (Iridaceae): 3. Sections Hyalis and Morphix- ia. Bothalia 41: 83–134. Rhoda McMaster for helping in the collection of Hes- GOLDBLATT, P., MANNING, J.C. & BERNHARDT, P. 2000. Adap- perantha kiaratayloriae at our study site in the Rûens tive radiation of pollination mechanisms in Ixia (Iridaceae: Cro- Shale Renosterveld. Mike Picker, University of Cape coideae). Annals of the Missouri Botanical Garden 87: 564–577. Town, and David Edge provided the insect identifica- GOLDBLATT, P., NÄNNI, I., BERNHARDT, P. & MANNING, J.C. tion. We also recognize the generous donation to the 2004. Floral biology of Hesperantha (Iridaceae: Crocoideae): how minor shifts in floral presentation change the pollination Overberg Lowlands Conservation Trust for renosterveld system. Annals of the Missouri Botanical Garden 91: 186–206. conservation made by Oren Taylor, London, United GOLDBLATT, P., MANNING, J.C. & GEREAU, R.E. 2013. The Cape Kingdom, facilitated by Fauna and Flora International genus Micranthus (Iridaceae: Crocoideae): nomenclature and (United Kingdom). Collecting permits were provided taxonomy. Bothalia 43,2: 127–144. MANNING, J.C. & GOLDBLATT, P. 2005. Radiation of pollination by the Nature Conservation authorities of Western Cape systems in the Cape genus Tritoniopsis (Iridaceae: Crocoideae) Province. and the development of bimodal pollination strategies. Interna- tional Journal of Plant Sciences 166: 459–474. MUCINA, L. & RUTHERFORD, M.C. (eds) 2006. The Vegetation of REFERENCES South Africa, Lesotho and Swaziland. Strelitzia 19. South Afri- can National Biodiversity Institute, Pretoria. GOLDBLATT, P. 1984. A revision of Hesperantha (Iridaceae) in the SANBI & DEAT. 2009. Threatened Ecosystems in South Africa: winter rainfall area of southern Africa. Journal of South African descriptions and maps. DRAFT for comment. South African Botany 50: 15–141. National Biodiversity Institute, Pretoria. GOLDBLATT, P. 2003. A synoptic review of the African genus Hesper- SCHUTTE-VLOK, A.L. 2011. A new species of Xiphotheca from the antha (Iridaceae: Crocoideae). Annals of the Missouri Botanical Western Cape, South Africa. Bothalia 41: 298–300. Garden 90: 390–443.

GOLDBLATT, P. & MANNING, J.C. 1993. Ixia acaulis, a new acau- 1,2 2,3 4 lescent species of Iridaceae: Ixioideae from the Knersvlakte, P. GOLDBLATT , J.C. MANNING and O.E. CURTIS Namaqualand, South Africa. Novon 3: 148–153. GOLDBLATT, P. & MANNING, J.C. 2000. The long-proboscid fly pol- 1 B.A. Krukoff Curator of African Botany, Missouri Botanical Garden, lination system in southern Africa. Annals of the Missouri Botani- P. O. Box 299, St. Louis, Missouri 63166, USA. E-mail: peter.gold- cal Garden 87: 146–170. [email protected]. GOLDBLATT, P. & MANNING, J.C. 2002. Evidence for moth and but- 2 terfly pollination inGladiolus (Iridaceae: Crocoideae). Annals of Research Centre for Plant Growth and Development, School of Life the Missouri Botanical Garden 89: 110–124. Sciences, University of KwaZulu-Natal, Pietermaritzburg, Private Bag GOLDBLATT, P. & MANNING, J.C. 2006. Radiation of pollination X01, Scottsville 3209, South Africa. systems in the Iridaceae of sub-Saharan Africa. Annals of Botany 3 Compton Herbarium, South African National Biodiversity Institute, (London) 97: 317–344. Private Bag X7, 7735 Claremont, Cape Town. E-mail: j.manning@ GOLDBLATT, P. & MANNING, J.C. 2007. New species and notes on sanbi.org.za. Hesperantha (Iridaceae) in southern Africa. Bothalia 37: 167– 182. 4 Botany Department, University of Cape Town, Rondebosch 7700, GOLDBLATT, P. & MANNING, J.C. 2009. New species of Geissorhiza South Africa. E-mail: [email protected]. (Iridaceae: Crocoideae) from the southern African winter-rain- MS. received: 2012-12-27

SANTALACEAE

THE MINOR GENERA KUNKELIELLA AND THESIDIUM INCLUDED IN THESIUM

Santalaceae (sensu APG III 2009: including Vis- and the large genus Thesium (± 340 spp: www.parasitic- caceae) comprise ± 1 100 species of hemiparasites or plants.siu.edu) distributed widely through the Old World aerial parasites assigned to ± 47 genera distributed but most diverse in southern Africa and with two species worldwide (Der & Nickrent 2008). The southern African in South America. Previously segregated as the genus representatives include ± 300 species in six genera, of Austroamericium Heyndrich (1963), the South American which Thesium L. (± 180 spp.) and Viscum L. (17 spp.) species were included in Thesium by Der & Nickrent are the largest (Germishuizen 2000; Jordaan & Bur- (2008). goyne 2000). Phylogenetic relationships in the family were analysed by Der & Nickrent (2008) using a multi- A subsequent molecular phylogenetic analysis gene data set. Eight clades were identified with strong focussed on investigating the phylogenetic relation- support. One of these, the Thesium clade, included the ships within Thesium and allied species (Moore et al. genus Osyridocarpos A.DC. as sister to a clade com- 2010). It confirmed that the segregate Austroamericium prising Kunkeliellia W.T.Stearn, Thesidium Sond. and was deeply embedded in the tropical clade of Thesium, Thesium L. This clade, treated as the segregate family affirming its inclusion in Thesium accepted earlier by Thesiaceae by Nickrent et al. (2010), is predominantly Der & Nickrent (2008). More relevantly in the southern African, with the monotypic Osyridocarpos widespread African context, Thesidium was confirmed as mono- through the continent; Kunkeliella, with just four spe- phyletic but also nested within the genus Thesium as sis- cies, endemic to the Canary Islands; Thesidium, with ter to the Eurasian species, with the remaining species of eight species, endemic to the Western and Eastern Cape; Thesium forming a clade sister to this. In order to render