Pacific Science, vol. 54, no. 2: 149-156© 2000 by University of Hawai'i Press. All rights reserved

Prehistoric Giant Swamp Taro (Cyrtosperma chamissonis)from Henderson Island, Southeast Polynesia1

JON G. HATHER2 AND MARSHALL I. WEISLER3

ABSTRACT: Subfossilleaf fragments of giant swamp taro (Cyrtosperma cha-missonis) were recovered from archaeological contexts dating as early as A.D.1451 (mean date) on Henderson Island (24 0 22' S, 128 0 19' W), Pitcairngroup-a raised limestone (makatea) island isolated at the extreme margin ofsoutheastern Polynesia and the Indo-West Pacific biotic province. Comparisonof subfossil specimens and modern reference material from a range of knowncultigens under scanning electron microscopy confirms the identification. Aperiod of active interarchipelago voyaging between A.D. 1000 and A.D. 1500 isknown from recent summaries of the geochemical analysis of exotic fine-grained basalt artifacts from archaeological sites throughout Polynesia. If notan initial colonization, it is during this time that Cyrtosperma should havebeen introduced prehistorically to most, if not all, of the inhabitable islands ofthe region, especially those island groups lying to the west of Henderson. Inves-tigation of subfossil plant remains adds another dimension to understandingplant distributions, prehistoric crop use, and subsistence practices in the Indo-Pacific region.

INVESTIGATION OF THE plants associated withpast diet and subsistence economy in the Pa-cific by the direct analysis of the archaeolog-ical remains of plant tissues is limited com-pared with that in other parts of the Tropics(for example, Central and South Americaand northern Africa). This is largely the re-sult of the importance of perishable root andtuber foods in Pacific subsistence over that ofseeds that often preserve well. Methodologiesfor the archaeobotanical analysis of partiallyor wholly seed-based economies are well es-tablished and have met with considerable

1 Fieldwork on Henderson Island was supported bythe Wenner-Gren Foundation for Anthropological Re-search (additional sponsors are acknowledged in Weisler-1995). Funding to J.G.H. by the Science and Engineer-ing Research Council (SERC), U.K. This is contribution62 of the Sir Peter Scott Commemorative Expedition tothe Pitcairn Islands. Manuscript accepted I August 1999.

2Institute of Archaeology, University College Lon-don, 31-34 Gordon Square, London WCIH OPY, U.K.

3 Department of Anthropology, University of Otago,P.O. Box 56, Dunedin, New Zealand (to whom corre-spondence should be addressed).

success. Comparable analysis of tropical ar-chaeological root and tuber-based econo-mies has, until relatively recently, lacked awell-defined methodology and so, apart froma few isolated important identifications (seeHather 1992 for a general review), has re-ceived less attention. Consequently, the plantremains associated with past root and tuber-based economies have also been left un-analyzed. In this paper we report the identi-fication of the leaf remains of an importantPacific cultigen, Cyrtosperma chamissonis(Schott) Merr, from a prehistoric con;6xt anddiscuss the significance of this idyntificationin terms of its past and presel)Ydistributionand its contribution to Pacific archaeobotany.

Cyrtosperma chamissonis, the giant swamptaro, is an aroid o( considerable dimensions,cultlvatecCfor· its starchy corm (Figure 1).Petioles are often spiny and arise directlyfrom the corm at ground level and may reachover 2 m long. The leaf blades, up to 1 mlong, are sagittate with pointed inferior lobes,with the midrib on a continuous line with thepetiole. In this character they are similar toAlocasia but differ from Colocasia. Varieties

149

150 PACIFIC SCIENCE, Volume 54, April 2000

FIGURE I. Giant swamp taro, Cyrtosperma chamissonis, growing in a large excavated pit on Ebon Atoll, southernMarshall Islands. (photo by M. I. Weisler)

differ in overall size as well as leaf shape andthe spininess of the petiole. Corms commonlymature after 2 to 4 yr, when they may weighup to 15 or 20 kg. If left for 10 or more years,

corms may reach 80 kg in weight (Barrau1961:43, Small 1972:66).

The nomenclature of the domesticatedCyrtosperma is a little confused. Arisaconthis

Prehistoric Giant Swamp Taro-RATHER AND WEISLER 151

chamissonis Schott and Cyrtosperma eduleSchott are now both considered to be synon-ymous with C. chamissonis. Cyrtospermamerkusii is considered by some also to besynonymous with C. chamissonis but byothers to be a separate and nondomesticatedpossible ancestor of the domesticate (Thomp-son 1982). Others view C. merkusii as a do-mesticate itself and a variety of C. chamisso-nis in the western Indo-Pacific (purseglove1972). It has been considered that C. cha-missonis was domesticated in Indonesia(Massal and Barrau 1955), and so the twopossibilities of C. merkusii being either anancestor or an early variety of C. chamissonisare not far apart. In this paper, the plant re-mains identified as Cyrtosperma sp. are con-sidered to be C. chamissonis.

Cyrtosperma chamissonis is generally cul-tivated on the low Pacific atolls in conjunc-tion with coconut, breadfruit, arrowroot, andpandanus. On such islands C. chamissonis isfavored for its ability to tolerate both occa-sional drought and the swampy saline con-ditions that characterize the only fertile re-gions of such islands. The herbaceous andrelatively low habitat of aroids is also fa-vored on islands with frequent high winds.The most important product of C. chamisso-nis is the starchy corm, which on some is-lands holds a position of considerable prom-inence in the diet. A secondary use of thecrop is, however, the leaves, important inwrapping food for cooking in the earth ovenand for covering the earth oven duringcooking (Massal and Barrau 1955).

Thompson (1982) described the moderndistribution of C. chamissonis in a paper thatrelates the economic importance of the aroidsin the Pacific to both cultural preference andecology. Thompson reported that by far themain concentration of C. chamissonis as amajor crop is in Micronesia, whereas in bothMelanesia arid .Polynesia,' wnere C. cha-missonis is reported, it is of minor or onlyoccasional use. This, Thompson suggested, isculturally significant, citing Luomala's workin Kiribati (1953, 1974) as an important ex-ample, but is largely determined by environ-mental conditions. The identification ofCyrtosperma, reported here, suggests that the

distribution of this domesticate in the Pacificdeserves reexamination. Brief review of theevidence for prehistoric long-distance inter-archipelago interaction (so-called trade andexchange) within Polynesia strongly suggeststhat Cyrtosperma should have been an ab-original introduction throughout the region.This additional archaeological information isespecially important because Whistler (1991),in his recent summary of Polynesian plantintroductions, reported Cyrtosperma as ab-originally introduced only as far east as theCook Islands (see also Peters 1994) and thatit was a modern introduction to the Mar-quesas and the Society Islands (Whistler1991: 45, 59).

Archaeological Context

The subfossil plant remains were collectedfrom the raised limestone (makatea) islandof Henderson, Pitcairn group (24 0 22' S,128 0 19' W), isolated near the southeasternmargin of Polynesia and the Indo-West Pa-cific biotic province (Kay 1984, Benton andSpencer 1995). A 5-month archaeologicalprogram during 1991-1992 was designed toinvestigate how small human groups survivedon isolated landfalls with impoverished ma-rine and terrestrial biota, nutrient-poor soils,unpredictable rainfall, and, perhaps mostimportant, isolation from high volcanic is-lands with correspondingly richer and morediverse resources (Weisler 1994, 1997a, Wal-dren et al. 1999). Twenty-eight sites were re-corded, consisting mostly of rockshelters andcaves, with a few gardening areas (marked byremnant stands of Cordyline fruticosa, levelsoil pockets, and stone mounds) and a largecoastal midden (Weisler 1998a). Measuringup to 225 m2 in floor area, all caves androckshelters were situated within 50 m of thehigh-tide mark. The deep. recesses of theshelters and the dry, calcareous sandy depos-its provided ideal conditions for the preser-vation of plant macrofossils. Thirty-one ra-diocarbon age determinations from 12 sitessuggest an occupation sequence spanningabout 600 yr beginning A.D. 1000 (Weisler1995). All sites from which plant macrofossilswere recovered are clearly prehistoric.

152

The archaeological deposits were ex-cavated by trowel and sediments sievedthrough 6.4-rnm mesh (other samples wereprocessed with 3.2- and 1.6-rnm screens).Both leaf samples were found in situ. Thesample from site HEN-6 was found 19 cmbelow the surface, just above an earth oven(in a layer with dense combustion features)and associated with abundant prehistoric ar-tifacts, shellfish, and faunal remains includ-ing bones of turtle, fish, bird, and rat. Woodcharcoal from deposits immediately belowwere dated to 380 ± 50 B.P. (before present[laboratory sample number Beta-59009]) andcalibrated to A.D. 1451-1650 (conventionalage -40 yr for Southern Hemispheric sam-ples, then calibrated at 2 (J after Stuiver andReimer 1993). From HEN-7, the leaf sample(11 cm below the surface) was associatedwith an earth oven in a layer with prehistoricartifacts, shellfish, and bone midden. A woodcharcoal sample (Beta-62941) provided aconventional age of 270 ± 50 B.P. calibratedto A.D. 1522-1955. Although the latter end ofthe calibration is modern, the archaeologicalcontext is clearly prehistoric where, on Hen-derson, this predates the early seventeenthcentury.

MATERIALS AND METHODS

Plant remains were systematically col-lected for archaeobotanical analysis from theHenderson excavations. Leaf fragments froma variety of contexts were collected by handand carefully packed for detailed laboratoryanalysis. The leaf fragments were examinedunder low-power incident light microscopy,and small fragments of both the abaxial andadaxial surfaces were mounted for examina-tion by scanning electron microscopy. Stan-dardbotanical texts were used to comple-ment data g-athered froill inodem- ieTerencematerial in identifying the archaeological re-mains. The modern reference material com-prised the following taxa: Alocasia macro-rrhiza, Artocarpus altilis, Asplenium nidus,Cocos nucifera, Colocasia esculenta, Cordy-line fruticosa, Cucuma longa, Cyrtospermachamissonis, Hibiscus tiliaceus, Morinda cit-

PACIFIC SCIENCE, Volume 54, April 2000

rifolia, Musa paradisiaca, Pandanus tectorius,and Zingiber officinalis, collected from avariety of locations in the Pacific. Leaves ofmodern material were air dried and thenplaced in a drying oven at 100°C.

RESULTS

Much of the archaeological leaf materialseemed well preserved under low-power inci-dent light microscopy. Several individualfragments were up to 6 cm across and,though thin and fragile, were pliable. Fewfragments had the original leaf margin, andthe important character of the marginal vein,used to differentiate some taxa, was notavailable. Under scanning electron micros-copy the leaf surfaces appeared considerablyabraded, and trichomes (hairs), where pres-ent, were visible only as disrupted basal cells.Stomata, however, were well preserved andprovided an important character upon whichto identify the taxa present. Cyrtospermachamissonis was identified from two samples.Figure 2 shows the similarities in leaf textureand stomata between the modern referencespecimen and the archaeological sample.

DISCUSSION

The distribution of any plant has a tem-poral dimension that may be of considerableimportance to anyone studying the evolutionof a specific plant or flora. To those studyingthe origin and dispersal of plants used bypeople, the temporal dimension is of particu-lar importance and forms the basis of manytheoretical discussions of archaeological top-ics such as economy, migration, sedentism,and agricultural origins. Cyrtosperma is notconsidered to be a crop of widespread im-

-- portance; indeed, if -fenos·fo . be·-grown as amajor staple only in regions where there isenvironmental discrimination against othercrops, with perhaps cultural preference afactor of varying significance. The impor-tance of the archaeological remains of Cyr-tosperma is in indicating a marginal agrarianenvironment where conditions were such that

'''~~''''''r " ...... " ..... ,..... ~ ~, ..",,..., '..~ ". ''''_' ,.,.,,,.-~ ... ,.. ."".........

Prehistoric Giant Swamp Taro-HATHER AND WEISLER 153

FIGURE 2. Scanning electron micrographs of Cyrtosperma chamissonis. (1) Surface of modern leaf (scale bar = 75Ilm); (2) detail of stomata of modern leaf (scale bar = 10 Ilm); (3) surface of archaeological leaf (scale bar = 75 Ilm);(4) detail of stomata of archaeological leaf (scale bar = 10 Ilffi). (photos by J. G. Hather)

other crops could not be grown. The practiceof using Cyrtosperma as a crop in such con-ditions, so as to make productive otherwisemarginal islands, is uncommon in Polynesia,although well known in Micronesia (Weisler1999). As reported by Thompson (1982 : 193),the modern distribution of Cyrtospermaplaces its farthest east occurrence aboll! 2000kin northwest of Henderson (128 0 18' W,24 0 22' S) on Makatea Island, Tuamotu Ar-chipelago (148°l5'W, 150 50' S). As such,the presence of Cyrtosperma on Hendersongives, perhaps, an indication of agrarianpractices in the Pacific that may now be onlylocally isolated, but may have been of con-siderable importance in the past.

Archaeological evidence for long-distance,postcolonization voyaging within Polynesiahas been summarized recently (Weisler 1997b,1998b) and adds an important independentline of data for interpreting aboriginal plantdispersals in the region. Figure 3 presents theprehistoric interaction spheres as inferredfrom. the distr!}mtion of. exotic fine-grainedbasalt adze material. The largest solid linedelimits the area with artifacts originatingfrom Tutuila, Samoa-an interaction sphereoperating perhaps as early as 2000 yr ago(Best et al. 1992, Clark et al. 1997: 79). Eastof the Cook Islands, all interaction spheresdate to between A.D. 1000 and A.D. 1500 andsuggest a time of prolonged long-distance

154 PACIFIC SCIENCE, Volume 54, April 2000

0'

20'

40'

110'

100llkm--===

130"150'E•

Mariana Is.

Yap Is...

/

Prehistoric Giant Swamp Taro-HATHER AND WEISLER 155

Their particular use in cooking-to wrapfood for placing in an earth oven or for stor-age, to line or cover an earth oven, or forserving cooked food-positions leaf materialin a broad range of preservational contexts.Both leaf remains described here were foundassociated with earth ovens, lending weightto this argument. The types of remains pres-ent in the agrarian past of Europe, SouthwestAsia, North Africa, and Mesoamerica are soconsiderably different from those in the Indo-Pacific region that many of the tried andtested techniques and methods of interpreta-tion are not applicable. We should not let thisfact delay our pursuit of understanding pastplant uses in the Indo-Pacific region.

ACKNOWLEDGMENTS

M.I.W. thanks Steve Christian and familyof Pitcairn Island for their wonderful hospi-tality and Roger Green for supplying copiesof an Auckland University Ph.D. thesis.Funding to J.G.H. by the Science and En-gineering Research Council (SERC), U.K. isalso gratefully acknowledged.

LITERATURE CITED

BARRAu, J. 1961. Subsistence agriculture inPolynesia and Micronesia. Bernice P.Bishop Mus. Bull. 223 : 1-94.

BENTON, T., AND T. SPENCER, EDS. 1995. ThePitcairn Islands: Biogeography, ecologyand prehistory. Academic Press, London.

BEST, S., P. SHEPPARD, R. C. GREEN, and R.PARKER. 1992. Necromancing the stone:Archaeologists and adzes in AmericanSamoa. J. Polynesian Soc. 101 :45-85.

BROWN, F. B. H. 1931. Flora of southeasternPolynesia. I, Monocotyledons. Bernice P.Bishop Mus. Bull. 84: 1-194.

CLARK, J. T., E. WRIGHT, and D. J. HER-DRICH. 1997. Interactions within and be-yond the Samoan archipelago: Evidencefrom basaltic rock geochemistry. Pages68-84 in M. I. Weisler, ed. Prehistoriclong-distance interaction in Oceania: An

interdisciplinary approach. N. Z. Ar-chaeol. Assoc. Monogr. 21.

HATHER, J. G. 1991. The identification ofcharred archaeological remains of vegeta-tive parenchymous tissue. J. Archaeol. Sci.18: 661-675.

---, ED. 1992. Tropical archaeobotany:Applications and new developments.Routledge, London.

HATHER, J. G., and J. C. ELLISON. 1995.Plant macrofossil analysis from lake cores:An example from Polynesia. Pages 191-202 in A. J. Barham and R. I. Macphail,eds. Archaeological sediments and soils:Analysis, interpretation and management.Institute of Archaeology, University Col-lege, London.

HATHER, J. G., and P. V. KIRCH. 1991. Pre-historic sweet potato (Ipomoea hatatas)from Mangaia Island, Central Polynesia.Antiquity 65: 887-893.

KAY, A. 1984. Patterns of speciation in theIndo-West Pacific. Pages 15-31 in F. J.Radovsky, P. H. Raven, and S. H. Somer,eds. Biogeography of the tropical Pacific:Proceedings of a symposium. Bernice P.Bishop Mus. Spec. Publ. 72.

LUOMALA, K. 1953. Ethnobotany of the Gil-bert Islands. Bernice P. Bishop Mus. Bull.213: 1-129.

---. 1974. The Cyrtosperma systemicpattern: Aspects of production in the Gil-bert Islands. J. Polynesian Soc. 83: 14-34.

MASSAL, E., and J. BARRAu. 1955. Pacificsubsistence crops-taros. South Pac.Comm. Q. Bull. April 1955: 17-22.

PETERS, C. 1994. Human settlement andlandscape change on Rarotonga, SouthernCook Islands. Ph.D. thesis, University ofAuckland.

PuRsEGLOVE, J. W. 1972. Tropical crops:Monocotyledons I and II. Longman,London.

SMALL, C. A. 1972. Atoll agriculture in theGilberts and Ellice Islands. Department ofAgriculture, Tarawa.

STUIVER, M., and P. J. REIMER. 1993. Ex-tended 14C data base and revised Calib 3.014C age calibration program. Radiocar-bon 35: 215-230.

156

THOMPSON, S. 1982. Cyrtosperma chamissonis(Araceae): Ecology, distribution, and eco-nomic importance in the South Pacific. J.Agric. Trop. Bot. Appl. 29: 185-203.

WALDREN, S., M. 1. WEISLER, J. G. HATHER,and D. MORROW. 1999. The non-nativeplant species of Henderson Island, South-central Pacific Ocean. Atoll. Res. Bull.463.

WEISLER, M. 1. 1994. The settlement of mar-ginal Polynesia: New evidence from Hen-derson Island. J. Field Archaeol. 21: 83-102.

---. 1995. Henderson Island prehistory:Colonization and extinction on a remotePolynesian island. BioI. J. Linn. Soc.56: 377-404.

---. 1997a. Prehistoric long-distance in-teraction at the margins of Polynesia.Pages 149-172 in M. 1. Weisler, ed. Pre-historic long-distance interaction inOceania: An interdisciplinary approach.N. Z. Archaeol. Assoc. Monogr. 21.

PACIFIC SCIENCE, Volume 54, April 2000

---, ED. 1997b. Prehistoric long-distanceinteraction in Oceania: An interdisciplin-ary approach. N. Z. Archaeol. Assoc.Manogr. 21.

---. 1998a. Issues in the colonization andsettlement of Polynesian islands. Pages79-94 in P. Vargas, ed. Easter Island andEast Polynesian prehistory. Easter IslandStudies Institute, FAU, University ofChile, Santiago.

---. 1998b. Hard evidence for prehistoricinteraction in Polynesia. Curf. Anthropol.39: 521-532.

---. 1999. The antiquity of aroid pitagriculture and significance of buried Ahorizons on Pacific atolls. Geoarchae-ology 14 (7): (in press).

WmsTLER, W. A. 1991. Polynesian plant in-troductions. Pages 41-66 in P. A. Cox andS. A. Banack, eds. Islands, plants, andPolynesians: An introduction to Polyne-sian ethnobotany. Dioscorides Press,Portland, Oregon.