CHAPTER SEVEN IDENTIFICATION OF CHARCOAL FROM KUOLULO ROCKSHELTER by Gail M. Murakami I DENTIFICATION OF CHARCOAL from archaeological sites can add valuable information toward our understanding of the life of the ancient Hawaiians (Murakami 1983a,b). The vegetation changes which may be indicators of their relationship with the environment are often recorded in the charcoal found at such sites. The present study utilized charcoal obtained from an 80 cm column in Site D6-60 (Unit D20) in the Anahulu Valley. Analysis of this column provided an opportunity to test whether a chronology of cultural activity and vegetation change could be established at the site. METHODS The 80 cm deep column from unit D20 was collected in 5 cm layers. Each charcoal sample was sorted into groups based on the anatomy of the freshly fractured transverse section as seen under a 40X magnification dissecting microscope. After all the samples had been sorted in this manner, the groups from all samples were resorted into larger groups or Wes. Thus the macroscopic characters were used to determine the distribution of each type. Representatives of each tpe, occurring in the combined levels of 0-20, 30-40, 50-60, and 75-80 cm were then selected for embedding in Spurr's epoxy resin (Spurr 1969) in a procedure modified from Smith and Gannon (1973). Thin sections of the plastic embedded charcoal were made with a steel knife on a sliding microtome. Permanent microscope slides of the sections were preparect Identifications were made by comparing the thin sections of the charcoal with those of known woods. This reference material is part of a study on the anatomy of Hawaiian woods. Permission for its use was obtained from Dr. Charles Lamoureux of the Botany Department at the University of Hawaii. The Botany Department also provided the facilities for processing the charcoal. RESULTS The fifteen 5 cm levels in the column were sorted into 199 groups and resorted into 56 types. Fifty-seven charcoal pieces from the 31 tyes which occurred in the combined levels 3040 and 50-60 cm as well as all representatives in the 0-20 and 75-80 cm levels were embedded and sectioned. Thirteen taxa were identified (table 7.1) and are described in the review which follows. The distribution of these taxa is given in table 7.2. The unknown taxa distribution is summarized in table 7.3. Anahulu TABLE 7.1 IDENTIFIED TAXA SCIENTIFIC NAME COMMON NAME ORIGIN Amaranthaceae Nototrichium sp. kulu'i E Araliaceae Reynoldsia sp. 'ohe E Celastraceae Perrottetia sandwicense Gray olomea E Chenopodiaceae Chenopodium oahuense (Meyen) Aellen 'aheahea, 'aweoweo E Euphorbiaceae Euphorbia spp. 'akoko (E) Leguniinosae Acacia koa Gray koa E Artocarpus altilis (Parkins. ex Z) Fosb. 'ulu, (breadfruit) P Myrtacae Eugenia malaccensis L. 'ohi'a-'i P Oleaceae Osmanthus sandwicensis (Gray) Knobl. olopua E Rubiaceae Canthium odoratum (Forst. f.) Seem. aahe'e I Santalaceae Santalum sp. 'ili-ahi, (sadlwood) E Solonaceae Nothocestrum sp. 'aa E Urticaceae Pipturus sp. mnaki (E) Nomenclature follows St. John (1973). For each taxon the following information is provided: 1. Scientific name. 2. Common English name or Hawaiian name, when known. 3. Origin. The following symbols are used: E endemic to the Hawaiian Islands, i.e., occurring nowhere else in the world. I= indigenous, i.e., native to the Hawaiian Islands but also occurring naturally (without the aid of man) elsewhere. X= exotic, i.e., plants of accidental or deliberate introduction after the Western discovery of the islands (post-1778). P= Polynesian introduction, plants brought by the Polynesian immigrants prior to Captain Cook's discovery of the islands. (E)= probably endemic species although genus not restricted to the Hawaiian Islands. 104 Identfication of Charcoal e6 Co e~ 0 co -4 Con 00 Co ' -4 Co 'Vt OI V- cN * * x ~o so Cf) c'n m) t N mn N Co CID I C) es so 0 xt0000 105 0 00 0 sr in sr 0 '-4 0 kn 0 Co 0 0 04 VS 0- V ow 2 it (A Eu - w N w Ez r rA MACr1W Wo 0 Co Co 0 0A N~ 0 W-4 z oo d 0 0 oe p4 3 0 p .I 0- I, Ct * 11 t; tR in * * I .1 -1 CL 6. ?02?g w R "'2 5? :? -a I I COD I I " a rA .?s ?h :z I I :S 'u ci ? :g I CS. .1 U? "E 2 t?o .2 .4? .1.3 - ;a a. '.. ., ? ul w w 1? Al 04 r-03 Anahulu W-4 V-~~~~~~-4 cr - -10 ~ c91 0I a40 ,No C N 0 00 c --4 - ON eCIA Cl CIA ' _- Cl X - - W ~ 00 00 Cl4 Cl4CN T-I4 -4 'r, Cli C l N M It 0 0 Qot- m 0 O C4 l q " 10 0 % m a v-4-4 -4 - - --4v- -4 - W- - 106 0 00 0 UN o trn 0) 0 0 0 0 4) u 04 :U V 3* q it (A z 3 Eu ix Eu aw Ct 0 In n C4, Cfl tA~ Cl4 0 Cl4 0 Cl4 tA~ 0 0 0 Z 06 Ident#ication of Charcoal REVIEW OF IDENTIFIED TAXA Family AMARANTHACEAE Nototrichium sp. (kulu'i) There are tree species in this endemic genus. Two are shrubs several feet tall found in the lowlands and the other is a tree 15-20 feet tall from the lower forest zone between 2,000 and 3,000 feet in elevation (Rock 1913). Nototrichium is generally found in dry regions such as xerophytic forests or on lava fields. The wood which is light when fresh carbonizes to dense hard charcoal. The characteristics of the family were clearly visible under the dissecting scope in the one charcoal piece recovered from the 25-30 cm level. Family ARALIACEAE Reynoldsia sp. ('ohe) Eight endemic species are listed in St John (1973) for this Polynesian genus. Reynoldsia is a tree, 15-60 feet tall, peculiar to the very dry districts of the lowland zone (Rock 1913). Like Erythrina (wiliwili), Reynoldsia sheds its leaves in the winter to bloom on bare branches. The soft wood was used for making stilts for a game called kukuluae'o (Malo 1898). Carbonized remains of Reynoldsia were found in low numbers (2-5%) in five of ten levels between 25 and 65 cm. Family CELASTRACEAE Perrottetia sandwicensis Gray (olomea) This endemic species grows as shrubs or trees 10- 18 feet in height in dry or wet regions ranging from 1,000 to 6,000 feet in elevation (Rock 1913). The wood was used to produce fire by rubbing against the softer hau (Hibiscus tiliaceus L.) (Malo 1898). Carbonized Perrottetia was found in low numbers in several levels between 15 and 70 cm. Family CHENOPODIACEAE Chenopodium oahuense (Meyen) Aellen ('aheahea, 'aweoweo) This endemic species is usually a shrub in the coastal lowlands but is believed to have been arborescent at higher elevations. Hillebrand (1888) wrote that he saw Chenopodium "arborescent, with a woody trunk, and 12-15 feet high, in the upper woods of Mauna Kea" and recently a worm-eaten trunk 18 cm in diameter was collected at Pohakuloa at 6,200 feet elevation on Mauna Kea (Warshauer and McEldowney, FRW 3586). Despite the size that Chenopodium may have attained, its soft wood was probably not used by the ancient Hawaiians. Its leaves, however, were cooked and eaten (Hillebrand 1888; Malo 1989). Charcoal of Chenopodium was found in eight of the eleven levels between 15 and 65 cm. Family EUPHORIBLACEAE Euphorbia spp. ('akoko) There are 16 species and numerous varieties of Euphorbia endemic to Hawaii. These shrubs or small trees may be found in coastal environments as well as wet forests. The milky sap was once considered a possible source for rubber (Rock 1913). The Hawaiians valued 'akoko for firewood (Hillebrand 1888). Charcoal which resembles the various forms of Euphorbia was found in every level in the 80 cm column. Family LEGUMINOSAE Acacia koa Gray (koa) One of the largest endemic trees in Hawaii, at higher elevations koa may exceed 80 feet in height and not branch until 40 feet or more above the ground. This straight trunk was especially useful for canoes as well as paddles, surfboards, and calabashes (Malo 1898). Today, although koa forests have been narrowed from their natural range of 600-5,000 feet in elevation mainly by cattle (Rock 1913), the wood is still valued. Charcoal resembling koa was found only in the 50- 55 cm level. Family MORACEAE Artocarpus altilis (Parkins. ex Z) Fosb. ('ulu, breadfruit) Artocarpus was carried by the Polynesians in their migrations. Only one variety is found in the Hawaiian Islands, but as many as 24 subspecies or varieties are distinguished by native names in Samoa, Fiji, and Tahiti (Rock 1913). Hawaiian 'ulu seldom bears seeds and thus is found near native dwellings or in valleys of the lowlands where cultivated by suckers. The wood of the 40-60+ feet tall trees was used in construction of doors of houses and for canoe hulls (Malo 1898). The bark of the young shoot was made into rough tapa (Malo 1898). The latex was used as caulking for canoes and birdlime (Neal 1965). The fruit was baked or pounded into poi for food (Rock 1913). Charcoal which resembles 'ulu was found in eight of the thire levels between 0 and 65 cnm 107 Anahdu Family MYRTACEAE Eugenia malaccensis L. ('ohi'a-'ai) This native of Malaysia, introduced into Hawaii by early Polynesian settlers, can be found on all islands in the lowest forest zone up to 1,800 feet in elevation (Rock 1913). It is a relative of 'ohi'a-lehua (Metrosideros collina spp. polymorpha) and may attain a height of 50 feet. The Hawaiians obtained a red tapa dye from the baik and used crushed leaves and bark medicinally. Charcoal resembling 'ohi'a-'ai was found in all levels except one between 15 and 80 cm. Family OLEACEAE Osmanthus sanwicensis (Gray) Knobl. (olopua) This endemic tree which is often 60 feet tall can be found in the lower forest zone at 6004,000 feet elevation (Rock 1913). The wood from olopua was made into adze handles (Rock 1913), digging sticks (Neal 1965), house posts, and weapons (Buck 1957). Maio (1898) notes that olopua wood burned when green. Charcoal resembling olopua wood was found in five of the twelve levels between 10 and 65 cm. Family RUBIACEAE Canthium odoratum (Forst. f.) Seem. (alahe'e) This small indigenous tree which may be up to 20 feet tall is found in dry regions of the lowlands or the lower forest zone up to 2,000 feet elevation (Rock 1913). The hard wood of alahe'e was used by the Hawaiians for digging sticks (Buck 1957). Carbonized wood resembling alahee was found in the 50-55 and 75-80 cm levels. Family SANTALACEAE Santalum sp. ('ili-ahi, sandalwood) The most recent treatment of Hawaiian Santalum recognizes four endemic species and several varieties found mostly in dry or mesic regions (Stemmermann 1980a). Their habits range from low shrubs on the coast to trees 65 feet tall in upper mesic environments. These root parasites depend on other plants for nutrition until established. The fragrant heartwood was the first major export item of the Hawaiian Islands and was used by the Hawaiians to perfume tapa (Neal 1965). Charcoal resembling Santalum was found in four of nine levels in the 20 to 60 cm range. Family SOLANACEAE Nothocestrum sp. ('aiea) The three species of this endemic genus are small trees, 15 to 35 feet tall, found in dry regions to rain forests (Rock 1913). The soft wood of 'aiea was used for canoes (Malo 1898), thatching sticks, and fire making (Pukui and Elbert 1957). Charcoal of 'aiea was found in five of nine levels in the 30 to 80 cm depths of the column. Family URTICACEAE Pipturus sp. (mamaki) The thirteen endemic species of Pipturus in Hawaii are usually small shrubs and occasionally 30 feet tall trees which inhabit mesophytic forests in the 1,500 to 4,000 foot elevational range (Rock 1913). The Hawaiians used the bark for tapa makdng (Buck 1957; Malo 1898) and the fruits for medicine (Rock 1913). The single find of Pipturus charcoal was recovered from the 55-60 cm level. DISCUSSION Limitations of Study The identification of wood charcoal from archaeological sites is a relatively new technique in Hawai'i and has yet to realize its full potential. The identifications made in this study were limited by the availability of microscope slides and/or complete descriptions of known woods as well as by the unknown range of variation in wood anatomy for both the comparative and charcoal material. The Hawaiian wood collection at the University of Hawaii, although encompassing nearly half of the woody species, is mainly a survey collection rather than the basis for studies in wood variation. The few genera which have been studied for anatomical variation have shown that differences exist between and within individuals of the same species (Sastrapradja and Lamoureux 1969; Stemmennann 1980b). Identifications can be made more readily in genera for which the range of variation is known, because the anatomical features characteristic of the genus as a whole are better defined. Determining frequencies of charcoal within a level or successive levels is difficult because of the unknown degree of preservation. Soft woods such as Erythrina (wiliwili) orReynoldsia ('ohe) are very fragile when carbonized and may not survive crushing or wet environments. Incompletely carbonized woods may be subject to fungal or microbial attack. 108 Identfication of Charcoal Vegetation Patterns Idicated by Charcoal Of the 13 identified taxa, two are from dry regions (xeric environments), six range from dry regions to the lower forest zone (xeric to mesic environments), three are found only in mesic environments, and one is cultivated today mainly in mesic regions. The possible vegetation types indicated above and the possibility that twelve (perhaps all) of these taxa were used by the Hawaiians suggests that the woods were selectively collected from plants growing nearby the site in a mesic environment. Among the plants selectively collected by the Hawaiians, four were much used during the time span represented by the column. The presence of Artocarpus altilis ('ulu), Chenopodium oahuense ('aweoweo), and Eugenia malaccensis ('ohi'a-'ai) in the column is indicative of actively cultivated food sources. Although Eugenia is today naturalized in mesic forests, Artocarpus had to be cultivated by man and Chenopodium might have been encouraged to grow in fallow fields (Allen, chapter 6, this volume). Euphorbia, on the other hand, was not used for food. Its presence in every level of the column strengthens its possible importance as firewood (Hillebrand 1888). Euphorbia charcoal has been found in two other sites, Waimea-Kawaihae (Murakami 1983a) and Kaho'olawe (Murakami 1983b). In each site a large group of charcoal exhibited a range of anatomical variation beyond that of the reference material but possibly within the range expected of taxa growing in differing environments. The hypothesis that this large group represents the wide ranging and anatomically variable Euphorbia is supported by Carlquist's study of the Hawaiian, Macaronesian, and cactoid African species (1970). The variation seen in the charcoal was within the range demonstrated in Carlquist's study of Hawaiian Euphorbia. It was hoped that a column of charcoal representing a continuum of activity might be used to formulate a chronology of vegetation changes due to human activity. An increase of introduced plants in the charcoal might reflect an increase in cultural activity and the cultivation of these plants. A decreased representation of native vegetation might suggest decreased availability in the immediate vicinity. However, the relative proportions of each layer (level) remained about the same with an average of 17% from introduced, 31% from native, and 52% from unidentified plants. This suggests a consistent and perhaps constant impact on the vegetation. The present analysis was not able to detect changes in the vegetation. The answers may be locked, in part, in the identities of the unknown taxa. In addition, a vegetation survey of the study area is necessary to determine how well the vegetation has survived the impact of cultural activity indicated by the column. The data does suggest that the two introduced plants, Artocarpus altilis ('ulu) and Eugenia malaccensis ('ohi'a-ai), represented in much of the column at a constant frequency, were already well established at the time that the rockshelter was first used. The vegetation of the area may already have been modified by human activities such as the cultivation of introduced plants before the site was ever used. This chronology would then explain why no drastic vegetation changes were detected in the column. REFERENCES CITED Allen, M.S. In prep. Kaho'olawe archaeobotanical materials. Buck, P.H. (Te Rangi Hiroa). 1957. Arts and Crafts of Hawaii. B.P. Bishop Museum Special Publication 45. Bishop Museum Press. Honolulu. Carlquist, S. 1970. Wood anatomy of Hawaiian, Macaronesian, and other species of Euphorbia. Botanical Journal Linnean Society Suppl. I 63:181-93. Hillebrand, W. 1888. Flora of the Hawaiian Islands. Reprinted by Hafner Publishing Company. New York, 1965. Maio, D. 1898. Hawaiian Antiquities. B.P. Bishop Museum Special Publication 2. 2nd Edition (reprinted 1971). Bishop Museum Press. Honolulu. Murakami, GM. 1983a. Analysis of charcoal from archaeological contexts. In J.T. Clark and P.V. Kirch, eds., Archaeological Investigations of the Mudlane-Waimea-Kawaihae Road Corridor, Island of Hawai'i: An Interdisciplinary Study of an Envirionmental Transect. Dept. of Anthropology Report 83-1, pp. 514-24. B.P. Bishop Museum. Honolulu. . 1983b. Identification of charcoal from Kaho'olawe archaeological sites. In RJ. Hommon, Kaho'olawe Archaeological Excavations 1981. pp. 168-88. Prepared by Science Management Inc. for the U.S. Department of the Navy, Pacific Division, Naval Facilities Engineering Command. Pearl Harbor. Neal, M.C. 1965. In Gardens of Hawaii. B.P. Bishop Museum Special Publication 50. Bishop Museum Press. Honolulu. Pukui, M.K., and S.H. Elbert 1957. Hawaiian-English Dictionary. University of Hawaii Press. Honolulu. Rock, J.F. 1913. The Indigenous Trees of the Hawaiian Islands. Privately published. Honolulu. 109 Anahudu Sastrapradja, D.S., and C. Lamoureux. 1969. Variation in wood anatomy of Hawaiian Metrosideros (Myrtaceae). Annales Bogorienses 5(1):1-83. Smith, F.H., and B.l. Gannon. 1973. Sectioning of charcoals and dry ancient woods. Amenrican Antiquity 38(4):468-72. Spurr, AMH. 1969. A low-viscosity epoxy resin embedding medium for electron microscopy. Journal of Ultrastructural Research 26:3143. St. John, H. 1973. List of Flowering Plants in Hawaii. Pacific Tropical Botanical Garden Memoir Number 1. Honolulu. Stemmermann, L. 1980a. Observations on the genus Santalum (Santalaceae) in Hawai'i. Pacific Science 24:41-54. ____. 1980b. Vegetative anatomy of the Hawaiian species of Santalum (Santalaceae). Pacific Science 34:55-75. 110