19951 BOYDAK AND BARROW: IN

Principes,39(3), I995, pp. LI7 -122 A New Localityfor Phoenixin Turkey: G6lkOy-BOdrum

MBun Boyner Professor of Siloicuhure, Faculty of Forestry, Uniuersity of , BiiyiikdereJstanbul, Turhey

SesseBennow The Herbarium, Royal Botanic Gard.ens,Kew, Richmond, Surrey TVg 3AB, U.K.

ABSTRAcT Phoenix theophrasti was describedby Greuter Until recently, the two native stands of palms in Turkey n 1967, from the famous grove at Vai in . were known to be of P. theophrasti Greuter in the Datga It is now known from nine coastal localities on Peninsula, and Kumluca-Karaiiz in Bay. A new locality that island (Turland et al. 1993), where it can be fot Phoenix in Turkey is now reported at Giilkiiy, north of found growing along moist valley floors, stream Bodrum. In certain of its characteristics this new population banks, by springs, on differs from P. theophrasti.lnthis article the identity of Giilkiiy coastal rocks and cliffs, in palm and its conservation shtus are discussed. all casesby the sea,from 0-230 m altitude. Greu- "Cretan ter (1967) noted that the " had been known since Classicaltimes. when it was Natural standsof PDoenix theophrasti Greuter recorded in the writings of Theophratus. Further were first discoveredin Turkey in l9B2 (Boydak records were made of the palm by travellers to 1983, Boydak and Yaka 1983, Boydak I9B5), the regionin later times(e.g. Leake 1835, Kirch- in the south-west of the country in the DatEa ner IB75). Many consideredit to be an escaped Peninsula. Boydak predicted that ftrrther popu- cultivar of the Date Palm, P. dactylifera L. (Lan- lations of Phoenix would be found in Turkey, and geron 1927). However, Greuter (L967) consid- sure enough three years later a second locality ered the Cretan palm to be a distinct species,and wasrecorded, a long way from Datga, at Kumluca- he named it P. theophrasfi in honor of the Greek Karaiiz in Finike Bay (Boydak 1986, 1987). A botanist-philosopher. third Turkish population of Phoenix has recently Greuter's formal taxonomic description of P. been found at the village of Giilkiiy, north of theophrasti did not end the confusion between Bodrum. this species and the cultivated Date Palm. The The Phoenix population has been known to the two speciescertainly can appear similar, especially inhabitants of Gtilkiiy for hundreds of years. In when P. dactylifera has been left untended, and 1989 A. Bayraktar and I. Aslanboga(Professors when there are no flowers or present. This of LandscapeArchitecture) visited the Giilkiiy palm confusionprobably accountsfor reports of P. theo- population and consideredit to be representative phrasti on other Aegean Islands, sirch as Kalim- of P. theophro,sti.ln the early summer of 1990 nos, Nisiros and Simi. These palms are likely to one of the authors, M. Boydak, made the first of be P. dactylifera. several trips to Giilkiiy and immediately noticed P. theophra.s/i is distinguished from P. d.ac- several distinct differences between the Giilktiy tylifera by its upright fruit clusters, and small palms, P. theophrasti and P. dactylifera. Sam- inediblefruits (Greuter 1967, Anon. 1983). Tur- ples were sent to the Royal Botanic Garden, Edin- land et al. (1993) note P. theophrasti to have burgh (December 1990) and Royal Botanic Gar- leaveswhich are smaller, shorter and sharper than dens,Kew (July 1993). In April 1994 both authors those of P. dactylifera.The Phoenix populations visited the Giilkiiy population and the two native of Datga and Kumlaca-Kara6z match the descrip- stands of P. theophrasfi in Datga and Kumluca- tion of P. theophrasfi of Crete, and so they are Karaiiz in Finike Bay and collected further her- included under that name. Certain morphological barium material. characteristics of the Giilktiy palm make it less n8 PRINCIPES [Vor.. 39

...-"--{""''-'u" 'q^av.--t'.--.--' )

EGE DENiZi (Eogen Seo) I'hnostrr Hr. ".kq Bog oreo -':i \ ,"p.r&i^. \

l A map to show the locality of the palm Srove at Giilktiy, near Bodrum, Turkey'

easy to name than the other Turkish palm pop- grove. There are reasonsto believe that it could ulations. This new population of.Phoenix is found be native. Firstly, the palms yield small fruit that growing on boggy ground on the boundaries of are slightly sweet,but scarcely fleshy, and it seems the fast-growing village of Gtilkiiy, between patches very unlikely that they were planted as a fruit of Pinus brutia Ten. and the sea. Several palms crop. Alternatively, it might have been the are now included within the gardens of newly-built or leaf-basefiber that was the desired crop, but houses.The main stand of palms covers an area no local records support this. Secondly, the pop- of approximately one hectare bordering the vil- ulation is well establishedand is regeneratingsuc- lage. However, palms can be found in smaller cessfully by both suckers and seedlings. outlying groups or individual clumps over an area from the village across roughly six times this size, The ldentity of the GdlkdY Palm the flat land to the sea (700-800 m away). On first seeing the population, so close to the The Giilkiiy palm differs from P. theophrasti village, one cannot help but ask if it is natural, as in terms of its fruiting stalk length, fruit size and opposedto being the remnantsof a cultivated palm seedshape. Gtilkiiy fruit stalks arc 0.6-2 m long, r99sl BOYDAKAND BARROW:PHOENIX IN TURKEY 1I9

/ /,i i11

2. A view of the Giilkiiy Phoenix palm population growing amongst piles of bricks: evidence of the expanding tourist village close by. 3. A cluster of scorched palm trunks at Giilkiiy. The suckers at the trunk base are an indication that the palms are recoverins well from the fire of 1993. PRINCIPES lVoL. 39

4. P. theophrasti growing along remote beachesand in rocky gullies of the Datga peninsula, Turkey. 5. P. theophrasti in full staminate flower, Datqa peninsula, Turkey. r99sl BOYDAKAND BARROW:PHOENIX IN TURKEY whereasthose ofP. theophrastirarely exceed0.3 a three year study ofthe whole genusby S. Barrow m, and they hold fruit that is marginally larger. at the Royal Botanic Gardens, Kew. It is hoped Although the Giilktiy palms grow on a fertile plain that the study of the chloroplast DNA will help to site, their height, at 4-B m, is generally shorter provide answersto the many questionsabout rela- than that of P. theophrasti of Datga and Kum- tionships within the genus. However, even if dif- laca-Kara6zwhere some palms up to 17 m can ferences are discoveredbetween the DNA of the be seen.The scanty fleshof the of the Giilkiiy three Turkish Phoenix populations, there is still palm have a slightly sweet taste, reminiscent of the problem of how to interpret them. It is impor- the cultivated date. It seemsthen that there are tant that they are interpreted against the bach, diferences betweenthe Giilktiy palm and P. theo- ground of variation n Phoenix as a whole. New' phrasti, the issue to be determined now relates speciesor varietiesof Phoenix cannotbe described to how we shouldinterpret and acknowledgethese without reference to the rest of the genus, and differences.Do they indicate that the Giilkiiy palm therefore the Gtilktiy grove cannot be formally is a new species,or a variety of P, theophrasti named until all Phoenix specieshave been studied or P. d,actylifera, or is there some other expla- to the same level. For example, P. reclinata ut nation? Africa showswide morphologicalvariation and yet The characterization of the various speciesof is currently held under a single name on account Phoenix is difficult at the best of times, due to of its pointed male flowers. If the Giilkiiy grove is the distinct lack of obvious differentiating char- describedas a new variety, and the DNA variation acters,and the ready hybridizationwith each other. betweenthe various forms of P. reclinata is found Throughout the natural range of Phoenix and its to be as great or greater than that between the range of cultivation, intermediates between the Giilkiiy grove and P. theophra*i, there would be speciescan be found. It is often a difficult, and strong grounds supporting the description of sev- perhapsimpossible, task to discover their parent- eral new varieties of African Phoenix. Now we age. It cannot be ruled out that hybridization has are entering the shaky ground of determining the played a role in the origin of the Giilkiiy palm. If conceptofa speciesin Phoenix, a subjectperhaps it is of hybrid stock, then what are the possible best left for another day! parents? P. theophrasli is a sure contender for Whatever decisionsare made about the history, one parent, since it shares the characters of a nature and name of the Giilkiiy Phoenix, there is clustering habit and upright fruit clusters with the no doubt that the populationis ofgreat importance Giilkiiy palm. The sweetnessof the Gtilkiiy fruits as a natural palm grove in Turkey, and therefore could point to P. dactylifera, which has been it is imperative that positive action is taken to cultivated in the Aegean region for many years, ensure its conservation. as the second parent. A preliminary anatomical study of the leaflets Conservationof the G6lk6y of the two Turkish populationsof P. theophrasti, Grove the Giilkiiy palm and samplesof non-Turkish P. There is considerablelocal interest, and a cer- dactylifera have provided some useful and inter- tain amount of pride, in the Giilkiiy palms, but estingresults. All three palmsare groupedtogether unfortunately this may not be enough to protect by thg presenceof large bundles of non-vascular them againstthe two major threats of water drain- fibers in the mesophyll layer of the leaflets. Very age and fire. Tourism has become an important small groups of fibers are occasionally found rn source of income for the local people, and there the leaf mesophyll of other speciesin the genus, is growing pressureto expand the village with new but they are never so abundantor in large bundles. summer housingand tourist facilities (Fig. 2). The Anatomical data clearly indicates a closerelation- Giilkiiy palm grove occurs within a nature con- ship between P. dactylifera, P. theophrasti and servation area and therefore any activities that the Gtilkiiy palm, though further sampling is might alter the natural balance of the ecosystem required before the exact nature of their relation- are forbidden. Despitethis, plans are afoot to drain ships can be understooi. the bog area, in which the palms grow to accom- At present therefore, there is no answer to the modate the inevitable golf course. A drainage Giilktiy problem, but further study of the mor- trench surroundingthe whole area is now in place. phology, anatomy and DNA of P. dactylifera, P. Before its construction, the boggy ground was theophrasti and the Gtilkiiy palm is included within completely flooded for several months each year. PRINCIPES lVoL. 39

Since the construction ofthe drainage trench cer- not only becauseit representsone of only three tain parts of the palm grove are inundated with native palm stands in Turkey, but also because water for only 2-3 months annually. The water- the identity ofthe palm is not yet clear. Therefore, relations of Phoenix are little understood, but it in t993 M. Boydak applied to local and city is clear that many of the speciesrequire an envi- governors requesting effective protection mea- ronment that is either permanently or seasonaliy sures for the Giilkiiy grove' including a ban on flooded, and therefore the drying out of the bog the water drainage scheme and expansion of the could be disastrousfor the Giilkiiy grove. village into the palm population. Unfortunately, "lake Giilktiy, meaning village," is named after the necessarymeasures have not yet been imple,- the lake that borders the palm grove and village' mented, but the attempt to realize positive action Thankfully, several abundant springs feed the lake, to ensure the conservation of the Giilkiiy grove and therefore it shouldbe impossiblefor the drain- continues. age ditch to dry out the bog completely. However, the sinking of the water-tablewill have an adverse Acknowledgments and this, in turn, impact on the local ecosystem We are grateful to ProfessorDr. F. Yaltirik, could affect the well-being of the palm grove. Mr. I. J. Hedge and Miss J. Lammond, and Dr. As to the other threat, in June 1993, Turkish J. Dransfield for their assistancein the attempt to to clear land T.V. reported thai a fire, purpose-lit identify the Giilkiiy palm. To Professors Dr. A. control and spread to for development,got out of Bayraktar and Dr. I. Aslandoga for the contri- grove (Fig. 3). Initially, a part of the Giilktiy palm bution of initial information on the palm grove. palms had been it was suggested that many Sincere thanks also go to the Turkish Forest Ser- blackened stumps destroyed, and certainly many vice-Regional Directorate of Mugla, especially palms. For- can be found amongst the surviving Mr. Behig Guyen and Mr. Osman Ceylan and show the grove tunately, our recent observations Forest Service Officials of Bodrum Forest Con- even to have to have recoveredwell, and perhaps servancy,and Giilkiiy's Muktar, Mr. Fetan Ozbek, in the main benefited, from the fire. All palms and Mr. Fevzi Ertugrul who contributed to the trunks, but most stand have charred, scorched various expeditions. show a healthy abundance of suckers sprouting at the base. Many of the palms were flowering LtreRerunn Ctrno during our visit, and throughout the grove, - lings can be found, evidence of successfulfertil- ANoN. I 983. Phoenix theophrasti Greuter' Red Data Sheet ization of the female flowers over the last few (1983 version). Kew, IUCN Threatened Commit' living uP to its name- ree, yearc. Phoenir is obviously BoyDAK, M. 1983. iikemizin nadide bir dogal tiirii Datga sake: rising from the fire! Hurmasi (Phoenix theophrasti Greuter). Qevre Koruma If fire was the only threat facing the palms, lB: 20-21. Istanbul. then the grove might be consideredrelatively safe. BoyDAK,M. ANDM. YAKA. I983. DatgaHurmasi(Phoenix Yarimadasindasaptanan dogal yay' However, the combined threats of fire and drain- theophrasti) veDatga ilisi. I. U. Orman FakiiltesiDergisi. A 33-t: 73-92. age of the bog convey an increasingly vulnerable BoYDAK,M. 1985. The distributionof Phoenix theophrasti conservationstatus upon the grove, since there is in the Datga Peninsula, Turkey. Biological Conservation little real control over the rate of expansion and 32: 129-135. developmentof the village. Certain outlying parts BoYDAK, M. 1986. Kumlaca-Karaiiz'de SaPtanan.Yeni bir (Phoenix theophrasti) yayilisi. I. U' Orman Bay Dogal Palmiye of the P. theophrasti population at Finike FakiiltesiDergisi, A 36-l: l-13. (Kumluca-Karaiiz) is similarly threatened by the BoYDAK,M. 1987. A New Occurrenceof Phoenix theo- growing tourist village nearby, though fortunately phrastiin Kumluca-Kara6z,Turkey. Prinicpes3l(2): 89- the bulk ofthis population,growing on top of steep' vJ. Beitrlge zur Flora der Siiddglis B-9. calcareouscliffs, and down their sidesto the sea, GRnurtn, W. 1967. Bauhinia3:243-250. occurs in areas unsuitable for housing develop- LeNcrnon, M. 1927. Annotaiions et additions i la flore ment. The various stands of the species in the cr6toise.Bull. Soc. Bot. Fr. ?4: 1.30-139. Datga Peninsula are, at presentoalso safe due to LEAKE,W. M. 1835. Travels in .Vol. 3' their occurrence in either steep, rocky gullies, or KIRcHNER,O. 18?5. Jahrb.Class. Philolog., Suppl' 3:451- 539. beaches. along remote, uninhabited TuRLAND.N. J., L. CnIlrou, eno J. R. Pness. 1993. Flora It is our opinion that conservationaction must of the Cretan area: an annotated checklist and atlas. be focusedsoon upon the Giilkiiy Phoenix grove, HMSO, London.