The red sparaxis, Sparaxis pillansii, a narrow endemic of seasonally wet sites in theCalvinia district, has unusual brick red flowers with contrasting centre, signalling anadaptation to monkey beetle pollination. Photo: J. Manning

A review of a Cape genusand its biology.

by Peter Goldblatt,Missouri Botanical Garden, U.S.A.and John C. Manning,National Botanical Institute,Kirstenbosch

Several years ago Veld &' Florapublished a r:eview of theCape genus Sparaxis, which

then included seven specieswhich shared a radially symmetricperianth and distinctive dry,crinkled and long-toothed or tornfloral bracts (Veld &' Flora 65,7-9,1979). Since then our under­standing of the Iridaceae (indeedall aspects of plant classificationand pollination biology) insouthern Africa has increaseddramatically. Flowers, we nowknow, have an important adaptivefunction. They must ensure theproduction of a new generation ofplants by fertilization of ovulesand maturation of seeds. In mostplants fertilization must beachieved by persuading an insect,bird or some other agent to visitthe flower and transfer pollenfrom the anthers of one plant orflower to the stigmas of another.Animals that perform thisfunction are called pollinators andthey visit flowers for quite selfishreasons, usually to forage on floralproducts. These products includenectar and pollen, the latter mostoften collected by female bees toprovision their nests. The transferof pollen is incidental to the polli­nator's foraging activity althoughit is vital to the plant.

22

Classical Sparaxis and two­lipped flowersIn species of Sparaxis in theclassic sense, that is the commonrange of garden hybrids and wildspecies including Sparaxis tricolor,S. elegans and S. pjJJansii, theflowers fulfil their purpose in avery unusual way. Recent studiesconducted by scientists at theNational Botanical Institute andthe University of Cape Town(see Veld &' Flora 82,17-19,1996)have demonstrated that flowers ofmany plants in the winter-rainfallzone of South Africa are used by

monkey beetles as the sites forassembly, attraction of matingpartners and copulation. Thebeetles invariably become coveredin pollen as they clamber clumsilyover a flower, and pollen in turnis brushed off onto stigmas afterthey move to another flower insearch of prospective mates.The flowers that beetles favour fortheir activities are often large andbrightly coloured, have a flatsurface or shallow cup andproduce little or no nectar butmay produce unusually largeamounts of pollen. Very often

Veld & Flora March 2000

these flowers also have contrastingdark markings that, it has beensuggested, mimic the beetlesthemselves as part of a strategy ofdecoy that may increase thefrequency of beetle visits and thusenhance cross pollination. Wenow believe that the flowers ofS. tricolor, S. elegans andS. pillansii are primarily beetleflowers, although they are alsovisited by small horseflies insearch of the traces of nectar thatare produced. Some otherSparaxis species, including thefairly common south-westernCape S. grandiflora appear to havea similar pollination biology butits flowers are also visited by arange of bees that collect pollen.

Another common south-westernCape species, Sparaxis bulbiferahas a more conventional floralbiology. Its cream flowers arecupped and have a hollow floraltube that contains moderateamounts of nectar. They arevisited mostly by honey bees andfemale bees of other species andsometimes by small horseflies insearch of nectar, but may also bevisited by monkey beetles. Pollenis passively dusted onto thebodies of these insects as theyclimb into the flowers. Monkeybeetle pollination, at least in theIridaceae, is now understood to be

a specialized system and thosespecies of Sparaxis with thispollination system are specializedin the genus.

This brings us to Synnotia, asmall south-western Cape genusnow (since 1990) included inSparaxis. Synnotia has alwaysbeen thought to be allied toSparaxis for it has the samespecialized floral bracts and thesame unusual leaf texture andvenation as typical Sparaxisspecies. The only differencebetween the two genera is thatSynnotia has highly irregular, two­lipped flowers with an enlargeddorsal tepa!, a coloured lower lip,and arched and parallel stamens.Like Sparaxis bulbifera thesespecies are mostly pollinated bybees foraging for nectar. Typicalexamples of the species onceincluded in Synnotia are Sparaxisvillosa from clay soils in thewestern Cape, and a new species,S. auriculata from rockysandstone slopes in theVanrhynsdorp district. Floralsymmetry alone is not an adequatereason for the recognition of agenus and with the new insightabout the adaptive nature of theflowe..rs of some of the typicalspecies of..$paraxis we nowbelieve th~t the species withradially symmetric flowers are

actually highly specialized for anunusual pollination system.Moreover, it is apparent fromphylogenetic studies of the genesequences of the genera of thelridaceae that two-lipped,irregular flowers are the ancestral(or primitive) condition in theIxioideae, the large subfamily towhich Sparaxis belongs. Puttingthis information together, it seemsall but certain that Sparaxis in theold sense consisted of one or twolineages of highly specializedspecies derived from an ancestorwith the irregular flowers thatcharacterized Synnotia. The twogenera were united under theolder name Sparaxis in 1992,which brought the taxonomy ofSparaxis back full circle to itsoriginal circumscription asdefined by the English botanist,John Ker Gawler in 1802, who hadactually included species ofSynnotia in Sparaxis when heerected the genus.

The fairly complex, but evidentlyancestral, flower type of Sparaxisauriculata are visited by bees in search ofnectar which the flowers do not produce.The flowers resemble those of Gladiolusvenustus a favourite nectar source forsolitary bees. Photo: J. Manning

Although the flowers of Sparaxis variegata resemble those of S. auriculata, they do,however, offer ample nectar to long-tongued bees. Pholo: J. Manning

Floral diversity and pollinationToday Sparaxis includes fifteenspecies from the south-westernCape and western Karoo all withrelatively diverse shapes andpollination strategies. Thestriking, symmetrical flowers ofS. tricolor and S. elegans arerelatively well known to gardenersand we now understand that thebrilliant flower colours areactually signals to monkey beetlesto visit the flowers. The much lesswell known S. villosa, S. variegataand the new species, S. auriculatahave flowers very like those of the

24

common bee-pollinated Gladiolusvenustus of the south-westernCape and in fact they share thesame species of bees as their polli­nators. The bright yellow lowerlip is a signal to bees and is oftencalled a nectar guide, while theenlarged dorsal tepal shelters andconceals the pollen so that is itless exposed to the elements andto insects that might take pollenwithout effecting pollination.Curiously, the display provided bythe flowers of S. auriculata is purebluff. Although this species hasthe largest flowers of any sparaxis

it does not produce nectar. Thetube is tightly closed and visitingbees are consistently disappointedin their search for nectar on thisplant. However, we suspect thatbees cannot distinguish theseflowers from very similar onesthat secrete nectar and so arerepeatedly tricked into visitingand pollinating flowers ofS. auriculata.

The flowers of S. villosa doyield nectar, but this species has afailsafe mechanism built into itsbiology. The flowers have veryshort styles so that the stigmas arein contact with the pollen. Ifinsect-mediated pollen transferdoes not occur, the flowers readilyself-pollinate themselves. Thenectar then goes to waste, butproduction of a seed crop for thenext year is safeguarded.

Sparaxis metelerkampiae fromthe lower mountain slopes of theWestern Cape has perhaps themost remarkable flowers of anysparaxis. They are dark violet incolour with the lower lip streakedwith white or yellow, and have anelongated floral tube up to 30 mmlong. The only insects able toreach the nectar produced in thetubes of these flower:s are long­proboscid flies (Prosoeca spp.)and the flowers of the species areadapted to this specialized polli­nator. The flowers resembleanother violet-flowered specieswith a white-streaked lower lip,the common Laperiousia jacquinii.This Lapeirousia andS. metelerkampiae are actuallymembers of a guild of plantspecies that have darklypigmented, long-tubed flowersand are all pollinated exclusivelyby long-proboscid flies. TheseProsoeca flies are rare today in thearea south of Clanwilliam andL. jacquinii and S. meteler­kampiae successfully reproducein the part of their range south ofClanwilliam only through self­pollination or occasional visits bystray insects that fail to reach thenectar and move to the flowers ofother plants. We have yet to learnwhy Prosoeca flies have disap­peared from part of their range forthey are often common north ofClanwilliam and extend intosouthern Namibia. There, species

Veld Er Flora March 2000

Peter Goldblatt is the Curator of African Botany atthe Missouri Botanical Garden, St. Louis, Missouriwith which he has been associated since 1972,when he left a teaching position at the Universityof Cape Town, South Africa. Peter has made thestudy of the Iris family a lifetime research interestand he is one of the world's leading experts onthe Iridaceae and its close relatives. Althoughmembers of the Iridaceae are world-wide in distri­bution, the family has a marked concentration onthe African continent where over 1000 of theestimated 1800 species are native. His research onvarious genera of the Iridaceae has taken him onnumerous field trips to Africa as well as southernEurope, the Middle East, and parts of NorthAmerica. His travels in the field have beenfunded by the U. S. National Foundation and theU.S. National Geographic Society. Author of some200 scientific papers, Peter has also writtenseveral books including The Moraeas of SouthernAfrica, The Genus Watsonia, The Woodylridaceae, Gladiolus in Tropical Africa and most

pollinated exclusively by long­proboscid Prosoeca flies are often(but not invariably) self-sterile andtherefore depend entirely on oneparticular pollinator for theirsexual reproduction.

One of the least known speciesof Sparaxis is the diminutiveS. parviflora which grows insandy and granitic soils along theCape west coast from Darling toHopefield. The flowers are tiny,but are lightly scented and offerminute amounts of nectar. Theplants also grow in densecolonies, which makes it worth­while for honeybees to visit themfor both nectar and pollen. Despitehaving this common pollinator,flowers of S. parviflora have shortstigmas and are also self-fertile.Perhaps in this species the failsafereproduction strategy is simplyinherited from an ancestor with aless reliable pollination system.

Thus despite its relatively fewspecies, Sparaxis is a fairlydiverse genus in terms of itsflower shape. The differentspecies have become adapted to asurprising range of pollinators.These include the apparentlyancestral system using antho­phorine bees with nectar as areward (or with no reward at all),long-proboscid flies, also withnectar as a reward, honeybeeswith pollen as a reward, moregeneralist systems involving arange of bees, flies and hopliinebeetles, or the highly specializedmokey-beetle and horsefly system.Knowing that flowers havespecific functions andthat particular flowershapes are adaptedfor pollination bydifferent organisms,we can look atflowers from a newperspective. TheypI ase us with theircolour, shape andscent both in thewild and in gardens,but they can also beunderstood as playingvery specific roles inthe biology of theplant and in theecosystems theyinhabit.w

The highly derived flowers of Sparaxis metelerkampiae have a long perianth tube andare pollinated by a long-proboscid fly which is attracted to dark red- or violet-colouredflowers. Pholo: J. Manning

About the Authors

recently Gladiolus in Southern Africa incollaboration with John Manning.

John Manning was born in Pieter-maritzburg,and since 1989 has been a Research Scientist withthe National Botanical Institute at Kirstenbosch.Although John has studied the anatomy, embry­ology and seed development of diverse plants,including those in families such as the Fabaceae,Proteaceae and Stilbaceae he has focused hisresearch more recently on the Iridaceae, collabo­rating on several different research projects withPeter Goldblatt. Together they have investigatedthe phylogeny and pollination biology ofLopeirousia and subsequently the systematics,pollination systems and evolution of Gladiolus insouthern Africa. John is an accomplishedbotanical artist and plant photographer. Hisdrawings have been published in numerousscientific papers, flora treatments and books.His most recent work is Gladiolus in SouthernAfrica in collaboration with Peter Goldblatt.

Veld & Flora March 2000 25