132 AROIDEANA, Vol. 29 Arum-A Decade of Change Peter C. Boyce Malesiana Tropicals Suite 4, Level 9, Tun Jugah Tower No. 18JaIan Abdul Rahman 93000 Kuching, Malaysia
ABSTRACT cense, Arum italicum ssp. albispathum, Arum italicum ssp. canariense, Arum it When published in 1993 The Genus alicum ssp. majoricense, Arum italicum Arum (Boyce, 1993) presented for the first ssp. neglectum, Arum longispathum, time in 70 years, and ever in English, a tool Arum lucanum, Arum maculatum, Arum to identify with some degree of confi nigrum, Arum orientale, Arum pictum, dence all of the 25 Arum species and their Arum pictum ssp. sagittijolium, Arum subordinate taxa then recognized. Inevi rupicola, Arum sintenisii, Croatia, Cyprus, tably since publication there have been polymorphic species, Turkey changes in species delimitation as well as new discoveries, such that the species tally for Arum now stands at 28. Changes of TAXONOMIC UPDATES note since 1993 include the recognition of Arum longispathum two species treated then at subspecific rank within A. orientale Bieb. [A. longis Resurrection of A. longispathum at spe pathum Reich. & A. sintenisii (Eng!.) P. C. cies rank is supported following observa Boyce] while a further species [A. alpinar tions made of living material from Croatia iae (K. Alpinar & R. R. Mill) P. C. Boyce], to which I did not have access before pub treated then as a subspecies of A. elon lication of The Genus Arum. It is now clear gatum Steven, is formally raised to the lev that A. longispathum is distinct from A. or el of species in this paper. Additional study ientale, indeed from any other Arum, by has resolved a number of issues including the inflorescence carried on a long pedun application of names for the species hith cle clear of the foliage in the manner of A. erto called A. alpinum Schott & Kotschy rupicola Boiss., but strongly odorous of and the status of subordinal taxa in A. it cattle dung. Discounting the long pedun alicum Miller. These are all detailed below cle, the general aspect of the inflorescence together with sundry other observations. is that of A. elongatum, with which A.lon While much work remains to be done gispathum shares a penetrating smell at on Arum it seems a worthwhile exercise anthesis but differs by the spathe being to present an overview of these changes pale purple with a still paler central por wrought in the past decade and to cast tion and the spadix appendix dull lilac; the some pointers for work that still remains entire spathe and spadix of A. elongatum to be tackled. is deep purple. Sterile flowers of A. lon gispathum are filiform and not thickened KEYWORDS as in A. elongatum. The information (Boy ce, 1993) concerning the restricted natural Arum alpinariae, Arum alpinum, Arum range (endemic to Croatia) of A. longis apulum, Arum besserianum, Arum cylin pathum remains valid. draceum, Arum dioscoridis, Arum dios coridisvar. cyprium, Arum dioscoridisvar. Arum sintenisii and Cypriot Arum syriacum, Arum dioscoridis var. philis taeum, Arum elongatum, Arum elonga The change of status of this Cypriot en tum ssp. alpinariae, Arum hygrophilum, demic was published in the Annales of the Arum italicum, Arum italicum f. majori- Goulandris Museum 0994 pub!. 1995). PETER C. BOYCE, 2006 133
However, this is not a well-known journal the sense of tuber elongation in, e.g. A. and so it seems worth repeating here the maculatum L. and its kin, in which the tu salient points. Arum sintenisii differs from bers are rhizome-like with the tuber grow A. orientale by the long-peduncled inflo ing laterally through the soil. The tubers of rescences carried at or slightly above the A. orientale are better described as being level of the leaves and the spadix appen obliquely or asymmetrically globular in dix producing a strong fruity smell at an such that the new tuber does not form di thesis (inflorescence carried below the fo rectly over the old tuber, as is the case liage and smelling of cattle dung/urine in with A. besserianum, but rather slightly A. orientale). In general aspect A. sinten offset to one side so that observing tubers isii rather closely resembles A. hygrophil of several years growth a somewhat elon um Boiss., especially as herbarium mate gated tuber does result, but not in the rial, although it differs in the spathe colour manner in which such rhizome-like tubers and spadix odour (A. hygrophilum has the occur in A. maculatum, etc, where new spathe pale green with a very narrow pur growth is produced laterally to the tuber ple margin and a spadix that is odourless axis and a long, sausage-like tuber results. at all times). It has further become appar The synonymy cited for A. besserianum ent that all the vouchered records of A. (Boyce, 1993) is still valid. hygrophilum from Cyprus are attributable to A. sintenisii and that Arum hygrophil Arum alpinarlae um appears to be absent from Cyprus. A plant of Cypriot origin that I errone Treated as a subspecies of A. elongatum ously reported as a putative hybrid be (Boyce, 1993), A. alpinariae (K. Alpinar & tween A. orientale and A. hygrophilum on R. R. Mill) P. C. Boyce is distinct from it by pages 179/80 of The Genus Arum is also the much shorter and considerably less referable to A. sintenisii. massive spadix appendix and most readily As an outcome of these changes, three distinguished by the startling rich crimson Arum species are currently recorded for spathe interior, quite different to the som Cyprus: bre purple spathe of A. elongatum and, indeed, unmatched by any other Arum species. These striking characteristics, to A. dioscoridis var. cyprium (Schott) Engl. gether with the so far endemic habitat (see note below under discussion of A. dioscoridis Sm.) strongly support its recognition as a dis tinct species. A. rupicola Boiss. The formal publication of its new status A. sintenisii (Engl.) P. C. Boyce is presented below: Arum besserlanum Arum alpinarlae (K. Alpinar & R. R. Mill) Arum besserianum Schott, treated as in P. C. Boyce, stat. nov. sufficiently known in 1993, is, based on Basionym: Arum elongatum ssp. alpi study of plants now in cultivation, quite nariae K. Alpinar & R. R. Mill in Fl. Turkey distinct from and indeed probably not par & E. Aegean Is., 10: 236, 219 (1988). ticular closely related to A. orientale. The Type: TURKEY: Sebsen to Bolu, Bolu spathe is deeper coloured and the tuber Yolu, Agu Dagi, Kesik yaylasi, amongJun globose and compressed, rather in the iperus, 1,600 m, 6 June 1983, K. Alpinar manner of a begonia tuber, while the ster ISTE 50605 (holotype ISTE!; isotype E!; K ile flowers are more densely arranged. photograph!). At least one nursery offering A. orientale and A. besserianum states that they may Arum cyUndraceum be separated by tuber shape, which is true, but implies that the tuber of A. orien Perhaps the most unfortunate outcome of tale is elongated. This is not strictly so in these new studies is to note that A. alpinum 134 AROIDEANA, Vol. 29
must now be called A. cylindraceum Gasp., pinum ssp. danicum (Prime) Terp6. a name treated as doubtful in 1993 but, fol A. orientale ssp. danicum (Prime) lOwing recollection of material from the type Prime locality, shown to be unequivocally identical A. transsilvanicum Cztez with the later A. alpinum. Additionally, new collections of herbar Arumpictum ium and living specimens of A. lucanum Cavara & Grande have demonstrated with Arum pictum 1. f. is a distinctive autumn out a shadow of doubt that it too is refer flowering species occurring in the Balearic able to the prioritised A. cylindraceum. islands, Corsica, Sardinia and mainland Given that A. alpinum has been so-long western Italy. Although unquestionably all obfuscated by A. maculatum and further belonging to one species, the populations burdened by a plethora of trivial taxonom in the Balearics tend to have rounded ic epithets, I feel it worthwhile presenting leaves with rather weakly defined rounded the new name for this species with its basal lobes whereas populations from Cor complete synonymy but without the long sica, Sardinia and mainland Italy have type and publication citations, for which leaves almost triangular in outline with see Boyce (1993). quite well formed basal lobes. Given this, the publication of a purportedly Majorcan Arum cylindraceum Gasp. A. macula endemic, Arum pictum ssp. sagittifolium J. tum var. cylindraceum (Gasp.) Engl. A. Rossell6 & 1. Saez (Acta Bot. Barcin., 44: A. alpinum Schott & Kotschy. A. ma 170 (1997)) is intriguing in that the speci culatum var. alpinum (Schott & Kot men that forms the type has leaves with schy) Engl. A. maculatum ssp. alpin much more the appearance of A. pictum um (Schott & Kotschy) Richter. A. from the easterly part of its range. maculatum var. angustatum subvar. alpinum (Schott & Kotschy) Engl. A. A New Approach with Polymorphic orientale ssp. alpinum (Schott & Kot Species schy) H.Riedl. I was inconsistent with my treatment of A. alpinum var. pannonicum Terp6, polymorphic species such that while I A. alpinum var. pannonicum fm javorkae treated some, such as A. maculatum as a Terp6. single highly variable entity, others, e.g. A. A. danicum (Prime) O.N. Dubovik A. italicum and A. dioscoridis I attempted to gracile Unverr. A. maculatum var. split into discrete subordinate taxa. The angustatum subvar. gracile (Unverr.) unfortunate result of this latter course of Engl. A. alpinum ssp. gracile (Un action was to leave behind a mass of verr.) Terp6 plants unplaceable within any taxon with A. intermedium Schur ex Schott. A. alpin out resorting to finer and finer degrees of um var. intermedium (Schur ex differentiation. These ultra-narrowly de Schott) Terp6 fined taxa, while arguably useful to the A. italicum var. lanceolatum Boiss. & horticulturist, are of little taxonomic value Heldr. ex Engl. and I have come to the conclusion that A. lucanum Cavara & Grande. A. orientale these highly variable species are best treat ssp. lucanum (Cavara & Grande) ed as single entities, as in the manner with Prime which I dealt with A. maculatum and that A. majoricum ssp. lucanum ('Bonafe') A. A. italicum should be treated as a highly M. Romo variable species. A. maculatum var. angustatum Engl. A. maculatum ssp. angustatum (Engl.) Arum italicum Richter A. maculatum var. attenuatum Engl. Since publication of The Genus Arum A. maculatum ssp. danicum Prime. A. al- (Boyce, 1993) I have examined in habitat a PETER C. BOYCE, 2006 135 great number of plants of two of the three state that ssp. neglectum never displayed western European subspecies of A. itali such variegation and used this as one of cum, viz. ssp. italicum and ssp. neglectum the potentially useful characters to aid (F. Towns.) Prime, with the result that I have identification. This is erroneous. Plants come to the conclusion that the two are not with typical ssp. neglectum leaf shape dis after all separable and fall under the one playing silver-grey veining are not at all name, A. italicum ssp. italicum. Despite uncommon. Interestingly some nursery this, I note with some distress that further catalogues state that ssp. neglectum differs subdivisional tinkering with this remarkably from ssp. italicum in having the leaf var variable species continues apace (e.g. A. it iegation not associated with the veins. This alicum f. majoricense (1. Chodat) M. Mus, J. is not so. A large wild population of A. J. Pericas & J. A. Rossell6 & A. italicum ssp. italicum will display a bewildering range majoricense (1. Chodat) o. de Bolos, R. M. of leaf markings associated with the ssp. Masalles & J. Vigo. neglectum leaf shape, with leaf markings The most easily observable character I ranging from silver-grey vein-associated used to distinguish between ssp. italicum variegation to clouded variegation not as and ssp. neglectum was that of leaf shape, sociated with veining while the ssp. itali in particular the relationship between the cum leaf shape also frequently displays basal and middle lobes of the leaf partic non-venation-associated variegation. ularly for the first flush of leaves to emerge The floral characters I used, mainly at the onset of growth. In 'typical' ssp. it dealing with the relative abundance of the alicum the basal lobes are long and nar sterile flowers above and below the zone row and diverge sharply giving rise to the of fertile stamens are also part of a varia familiar leaf shape of the forms of A. ital tion cline and in the intervening years it icum most often favoured in gardens has come forcibly to my attention that the whereas the leaves of 'typical' ssp. neglec abundance of these sterile flowers in di tum have shorter, broader lobes that par rectly linked to plant vigour in this and all tially overlap and diverge only slightly or other Arum and is not to be seen as a re not at all. Observation of large populations liable character upon which significant of A. italicum sensu lato in western France taxonomic decisions are made when dis and southern Spain has showed that these tinguishing between elements within the leaf forms are simply elements within a same species. However, the form and ar considerable pool of variation and that rangement of these sterile structures re where there are enough plants over a mains a powerful tool for separating be large enough area intermediates in leaf tween different species. shape are abundant. The feature of inflorescence colours in Another character I utilized in separat ssp. neglectum being more sombre than in ing the subspecies, albeit primarily with ssp. italicum is stable only for the U.K. regard to cultivated plants, is that of leaf populations where the dull ochre or pur markings. In gardens typical ssp. italicum plish ochre spadix appendix and purple is probably best recognized and certainly flushed spathe limb are diagnostic. On most favoured by gardeners in the form continental Europe the colours are vari with strikingly silver-grey main and sec able, with the sombre colours 'typical' of ondary veins. As pointed out in The Genus ssp. neglectum not infrequently occurring Arum, this leaf marking form is by no in inflorescences of typical ssp. italicum means the norm in wild populations, and vice versa. In particular, the purple where unmarked or only scantily silver staining on the inside of the convolute part veined leaves are far more abundant while of lower spathe seems to be much more forms with plain green or variegated inconsistent than I once thought. Evidence leaves with purple-black spotting are also I had to hand when working on A. itali common as are plants with leaves discrete cum for the book supported that the lower ly spotted silver-grey. Nonetheless, I did spathe of 'typical' ssp. italicum seldom if 136 AROIDEANA, Vol. 29 ever displayed such purple staining amass enough information to deal with A. whereas that of 'typical' ssp. neglectum in italicum ssp. canariense (Webb. & Berth.) variably has staining to a degree and fre P. C. Boyce in a similar critical manner; quently quite markedly so. Now that I there is also the problem of what to do have had the opportunity to examine with A. italicum ssp. albispathum (Steven spathes from hundreds of different indi ex Ledeb.) Prime, which I admit to know viduals over a wide range of habitat it is ing only through cultivated specimens, apparent that the spathe staining is unre never having seen it in habitat. I suspect liable in defining taxa within a species, al the most logical step will be resurrecting it though again it remains a useful tool at the as a species, but, for moment, I'm main species to species level with, e.g. A. ni taining the unsatisfactory status quo. grum Schott, A. apulum (Carano) P. C. Boyce and A. orientale, etc., readily sep Arum dioscoridis arable using this character. I cited a geographical basis for the sep Retaining or dispensing with taxonomic aration, with ssp. neglectum accounting subordinates within A. dioscoridis is more for the northerly part of the range of A. complex than with A. italicum not least italicum and ssp. italicum the southerly due to the importance that the spathe part. Curiously this is still a reasonably ac markings have in deciding which individ curate statement, although the remarkably uals are horticulturally desirable; thus the unvarying nature of ssp. neglectum north morphological markers that are provided of the Loire Valley is odd given the chaotic by the recognition of formal taxonomic introgression south of the Loire Valley. entities are horticulturally important. One explanation might be that the south The problem, however, is that the for ern range of ssp. neglectum is blurred by mal varietal taxa I recognized in 1993- massive hybridization with ssp. italicum. var. dioscoridis, var. cyprium (Schott) If that were so, then one would expect Engl., var. syriacum (Blume) Engl. and ssp. italicum in its strict sense to become var. philistaeum (Kotschy & Schott) Boiss., the dominant plant further south where it are still far from universally applied in hor is far removed from the supposed north ticultural circles. The greatest confusion is erly subspecies. This is not the case; the associated with var. dioscoridis, a variety range of variation remains the same as far quite clear-cut in habitat by the confluent south as the Costas (coastal towns associ purple-black spotting overlaid with vary ated with bathing beaches) of Spain and ing degrees of purple staining and the up into North Africa. Given the unvarying ap per part of the spathe, usually, but certain pearance of plants from the northerly pop ly not invariably bereft of any purple spots ulations it seems likely that at least the or stains. Such forms without any upper wild UK populations of ssp. neglectum are spathe markings are not infrequently sold for the greater part clonal or closely inbred under the varietal epithet luschanii RR such that variation is minimal; sites at Mill despite the fact that such plants ab Arundel in West Sussex and on the Isles of solutely match the illustration that forms Scilly (where A. maculatum is absent) are the type of A. dioscoridis and is thus also typical of such possible near-clonal pop the type of the typical variety, var. ulations. Although this needs to be inves dioscoridis. The forms of var. dioscoridis tigated using molecular techniques if it is in which the upper limb is variously pur to be regarded as more than just suppo ple stained have no formal epithet; if all sition the conformity of these northerly the varieties I recognized in 1993 are to be populations is striking. meaningful and retained in horticultural In conclusion for the moment A. itali usage then an epithet for these plants is cum ssp. italicum will remain as the name required! for the western continental European and You may gather from the last sentence UK plants because as yet I have still to that I am in favour of merging at least PETER C. BOYCE, 2006 137 some of the varieties I formerly recog 7. Arum besserianum Schott. Northeastern nized. The introgression between var. cy eastern Europe into southwestern CIS. prium (typically pale green spathe with 8. Arum longispathum Reich. Croatia. discrete large purple spots and no stain 9. Arum sintenisii (Engl.) P. C. Boyce. ing) in one direction into var. dioscoridis Cyprus. (confluent spots and staining) and into var. 10. Arum gratum Schott. Northeast Tur syriacum (small, scattered spots, no stain key. ing, sometimes entirely pale green) is so 11. Arum apulum (Carano) P. C. Boyce. complete that, at least in botanical terms Central Italy. these varieties have no real value. 12. Arum nigrum Schott. Western Bal The one exception to this merging of kans. minor taxa might be var. philistaeum, de 13. Arum cyrenaicum Hruby. Ubya, Crete. scribed from Gaza, a plant still very poorly 14. Arum purpureospathum P. C. Boyce. understood and, despite at least one nurs Crete. ery offering plants under this name, not 15. Arum balansanum R. R. Mill. Central presently in cultivation. The possibility west Turkey. that this plant may turn out to be a hybrid 16. Arum hainesii Agnew & Hadac ex H. between A. palaestinum Boiss and A. Riedl. Northeast Iraq. dioscoridis should not be discounted. 17. Arum elongatum Steven. Northeastern Turkey, southwestern CIS. CONCLUSIONS 18. Arum alpinariae (K. Alpinar & R. R. Mill) P. C. Boyce. Central west Turkey. The species list with distributions now 19. Arum rupicola Boiss. East Aegean is comprises: lands, Turkey, western Middle East. 20. Arum jacquemontii Blume. Southern 1. Arum maculatum L. Throughout Central Asia, Nepal, as far east as far northwest and central Europe as far western China. north as northern Germany and as far 21. Arum korolkowii Regel. Northern south as the eastern Balkans. Asian CIS. 2. Arum byzantinum Blume. Eastern cen 22. Arum euxinum R.R. Mill. Northeast tral Balkans, extreme western Turkey. Turkey. 3. Arum italicum Miller. Throughout Eu 23. Arum hygrophilum Boiss. Western rope as far north as the southernmost Middle East, northeast Morocco. UK and as far south as western North 24. Arum dioscoridis Sm. Southern Tur Africa, east to the western Balkans (typ key, eastern-most Mediterranean is ical subspecies). Two additional subspe lands, western Middle East. cies are still recognized: ssp. canariensis 25. Arum palaestinum Boiss. Western (Webb. & Berth.) P. C. Boyce (Canary Middle East Isles); ssp. albispathum (Steven ex Led 26. Arum idaeum Coust. & Gandoger. eb.) Prime (northeast Turkey, south Crete. western Cornmonwealth of Indepen 27. Arum creticum Boiss. & Heldr. South dent States (former USSR). east Aegean islands, southwest Turkey. 4. Arum concinnatum Schott. Southern 28. Arum pictum L. f. Corsica, Sardinia, most Greek mainland, southeast Ae western central Italy. gean islands, southwest Turkey. 5. Arum cylindraceum Gasp. Through UfERATURE CITED out Europe and west Asia, from Den Boyce, P. C. 1993. The Genus Arum. Royal mark to northeast Turkey and as far Botanic Gardens, Kew, HMSO, Lon south as Sicily. don. 6. Arum orientale Bieb. Central & east --. 1995, "1994". The genus Arum ern Europe, northeastern Turkey, (Araceae) in Greece and Cyprus. southwestern CIS, northwestern Iran. Ann. Musei Goulandris 9:27-38.