Historical realism* in Tiyambe Zeleza's Smouldering Charcoal ...
Charcoal Stratigraphies for Kaua'i and the Timing of Human Arrival1
16
Charcoal Stratigraphies for Kaua‘i and the Timing of Human Arrival 1 Lida Pigott Burney 2,3 and David A. Burney 3 Abstract: Evidence from microscopic charcoal particle stratigraphy is presented from nine locations distributed throughout Kaua‘i in the Hawaiian Islands, in- cluding windward and leeward coastal sites and interior bogs at elevations ranging up to 1220 m. The overall trends are comparable with those reported for other mesic tropical island areas lacking strong seasonality, beginning with a general dearth of charcoal in sediments that predate evidence for humans on the island, followed by an increase of an order of magnitude or more at a time that probably represents first human presence at the site. In most cases, this initial peak or plateau of increased charcoal from presumably anthropogenic sources is followed by a prehistoric decrease and a second peak after European contact. Charcoal evidence presented here suggests a human presence in leeward coastal areas beginning ca. 830 G 50 yr b.p. (1050–1095, 1140–1280 cal yr a.d.). One windward site, Limahuli Bog, may show charcoal evidence for humans as early as 1470 G 60 yr b.p. (440–670 cal yr a.d.), but resolution is poor in the upper part of that core. Charcoal and sedimentological evidence suggests that Hawai- ians were constructing fishponds as early as about eight centuries ago and that the massive stoneworks forming the Alekoko or Menehune Fishpond, probably the largest prehistoric stone structure in the Hawaiian Islands, may have been completed by 580 G 30 yr b.p. (1305–1420 cal yr a.d.). Charcoal peaks in pre- human times, particularly at 3800 G 40 yr b.p. (4080–4290 cal yr b.p.), may be associated with prolonged drought conditions. Charcoal particles are virtually absent from the late Pleistocene sediments collected from interior bogs. I n evaluating possible causes for late prehistoric extinctions on any given landmass, perhaps no single data point is more impor- tant than the inferred date of human arrival. Scientific controversies regarding this critical point in the Americas (Martin and Steadman 1999) and Australia (Flannery 1994) make it clear that this is no easy determination at the continental level. In contrast, studies from the large remote island of Madagascar (Burney 1999) definitively show human arrival ca. two millennia ago, with good agreement between independent data sets, but estimates derived by various methods for first human arrival in many other areas show large discrepancies. For instance, estimates for human arrival in the Hawaiian Islands from various archae- ological sites range from the first century b.c. (Beggerly 1990) to a.d. 800 (Athens 1997). Although adherents of both the ‘‘long chro- nology’’ and the ‘‘short chronology’’ continue to hold forth, newer interpretations of gene- alogical information and reexamination of earlier radiocarbon dates mostly support the shorter chronology (Masse and Tuggle 1998). It would clearly be useful to the testing of hypotheses for a human role in late prehis- toric extinctions in the Hawaiian Islands (Olson and James 1982) to know more cer- tainly when Polynesian settlement began in the mid-Pacific. Stratigraphic charcoal parti- cle analysis is an approach to the question 1 This research was supported by NOAA Human Dimensions of Global Change grant no. NA46GP0465, NSF grant DEB-9707260, National Geographic Society grant 7072-01, and funding from the Smithsonian Insti- tution, Kaua‘i Community College, Kı ¯lauea Point Natu- ral History Association, and the Faculty Research Grants and Faculty Fellowship programs of Fordham University. Manuscript accepted 27 June 2002. 2 The Louis Calder Center–Biological Field Station, P.O. Box 887, Armonk, NY 10504. 3 Department of Biological Sciences, Fordham Uni- versity, Bronx, New York 10458. Pacific Science (2003), vol. 57, no. 2:211–226 : 2003 by University of Hawai‘i Press All rights reserved 211
-
Upload
independent -
Category
Documents
-
view
0 -
download
0
Transcript of Charcoal Stratigraphies for Kaua'i and the Timing of Human Arrival1
Charcoal Stratigraphies for Kaua‘i and the Timing of Human Arrival1
Lida Pigott Burney2,3 and David A. Burney3
Abstract: Evidence from microscopic charcoal particle stratigraphy is presentedfrom nine locations distributed throughout Kaua‘i in the Hawaiian Islands, in-cluding windward and leeward coastal sites and interior bogs at elevationsranging up to 1220 m. The overall trends are comparable with those reportedfor other mesic tropical island areas lacking strong seasonality, beginning with ageneral dearth of charcoal in sediments that predate evidence for humans on theisland, followed by an increase of an order of magnitude or more at a time thatprobably represents first human presence at the site. In most cases, this initialpeak or plateau of increased charcoal from presumably anthropogenic sources isfollowed by a prehistoric decrease and a second peak after European contact.Charcoal evidence presented here suggests a human presence in leeward coastalareas beginning ca. 830G 50 yr b.p. (1050–1095, 1140–1280 cal yr a.d.). Onewindward site, Limahuli Bog, may show charcoal evidence for humans as earlyas 1470G 60 yr b.p. (440–670 cal yr a.d.), but resolution is poor in the upperpart of that core. Charcoal and sedimentological evidence suggests that Hawai-ians were constructing fishponds as early as about eight centuries ago and thatthe massive stoneworks forming the Alekoko or Menehune Fishpond, probablythe largest prehistoric stone structure in the Hawaiian Islands, may have beencompleted by 580G 30 yr b.p. (1305–1420 cal yr a.d.). Charcoal peaks in pre-human times, particularly at 3800G 40 yr b.p. (4080–4290 cal yr b.p.), may beassociated with prolonged drought conditions. Charcoal particles are virtuallyabsent from the late Pleistocene sediments collected from interior bogs.
In evaluating possible causes for lateprehistoric extinctions on any given landmass,perhaps no single data point is more impor-tant than the inferred date of human arrival.Scientific controversies regarding this criticalpoint in the Americas (Martin and Steadman1999) and Australia (Flannery 1994) make itclear that this is no easy determination at the
continental level. In contrast, studies from thelarge remote island of Madagascar (Burney1999) definitively show human arrival ca. twomillennia ago, with good agreement betweenindependent data sets, but estimates derivedby various methods for first human arrival inmany other areas show large discrepancies.
For instance, estimates for human arrivalin the Hawaiian Islands from various archae-ological sites range from the first century b.c.(Beggerly 1990) to a.d. 800 (Athens 1997).Although adherents of both the ‘‘long chro-nology’’ and the ‘‘short chronology’’ continueto hold forth, newer interpretations of gene-alogical information and reexamination ofearlier radiocarbon dates mostly support theshorter chronology (Masse and Tuggle 1998).It would clearly be useful to the testing ofhypotheses for a human role in late prehis-toric extinctions in the Hawaiian Islands(Olson and James 1982) to know more cer-tainly when Polynesian settlement began inthe mid-Pacific. Stratigraphic charcoal parti-cle analysis is an approach to the question
1 This research was supported by NOAA HumanDimensions of Global Change grant no. NA46GP0465,NSF grant DEB-9707260, National Geographic Societygrant 7072-01, and funding from the Smithsonian Insti-tution, Kaua‘i Community College, Kılauea Point Natu-ral History Association, and the Faculty Research Grantsand Faculty Fellowship programs of Fordham University.Manuscript accepted 27 June 2002.
2 The Louis Calder Center–Biological Field Station,P.O. Box 887, Armonk, NY 10504.
3 Department of Biological Sciences, Fordham Uni-versity, Bronx, New York 10458.
Pacific Science (2003), vol. 57, no. 2:211–226: 2003 by University of Hawai‘i PressAll rights reserved
211
jgh
of when humans arrived that is independentof the site-discovery constraints inherent toconventional archaeology. Detecting a sud-den, prolonged spike, generally better de-scribed as a plateau, of microscopic charcoalparticles in sediments of lakes, bogs, andcaves has been used effectively on many is-lands and some continental areas to elucidatehuman arrival (see Grayson 2001 for a recentsummary).
Evidence for human activity from charcoalparticles has been reported for high-elevationinterior Maui (Burney et al. 1995) and severalcoastal sites on O‘ahu (summarized in Athens1997). The former study showed that anactive volcano, and possibly droughts, led in-termittently to substantial Holocene charcoaldeposition before humans, but that valueswere generally low in the late Holocene untilEuropean contact. This site is extremely re-mote on the upper flank of Haleakala Vol-cano at 2270 m. Along the coasts of O‘ahu,Athens and his collaborators have foundcharcoal evidence for a human presence be-ginning ca. 800 cal yr a.d., increasing until ca.1200 cal yr a.d., then decreasing to low levelsuntil European arrival, which resulted in asecond increase (Athens 1997).
Kirch (1985) summarized the earlier con-sensus among Hawaiian archaeologists thatthe wetter leeward valleys of the HawaiianIslands would have been settled first, perhapsby a.d. 400 in his estimation (but there iscurrently disagreement on this—see Athens1997). Kirch further speculated that, by ca.a.d. 1000, the dryer leeward valleys of theIslands, presumably less desirable for the tra-ditional Polynesian styles of wetland tarocultivation, would have begun to be settled,perhaps at least partially in response to pop-ulation pressure.
If this scenario is true, the expectations interms of a charcoal stratigraphy for a dry lee-ward site might be: (1) very low backgroundlevels of charcoal in times before human ar-rival, except in the vicinity of active volca-noes; (2) a large (e.g., one to three orders ofmagnitude) increase in microscopic charcoalwith human settlement in the vicinity (i.e.,perhaps within a few kilometers—see Burney1987, Clark 1988, Burney et al. 1994) by ca.
1000 cal yr a.d.; (3) after one to a few cen-turies, a decrease in charcoal values, based oncharcoal results from other islands, but not tothe very low values typical of times beforehuman arrival; and (4) a second peak of an-thropogenic charcoal after European contactand through the subsequent two centuries.
We report here on nine sites, arrayedalong the dry leeward south coast of Kaua‘i,on the wetter windward north side, and atvarious elevations in the interior. Each hasbeen cored or excavated for sedimentologicalanalyses and 14C dating (Burney 2002), andone, the caves and sinkhole at Maha‘ulepuon the south coast, has been studied in detailfor a wide range of stratigraphic indicatorsof environmental change, including artifacts,bird bones, invertebrate shells, diatoms, pol-len, and plant macrofossils (Burney et al.2001).
materials and methods
For coring and excavation methods em-ployed, see Burney et al. (2001) and Burney(2002). Sediment subsamples of 0.5 ml wereprocessed for charcoal analysis using concen-trated HNO3 to remove organic matter, as inSwain (1973). Hydrofluoric acid was used toremove clay and sand, and microslides wereprepared from this residue. Samples werespiked with a known quantity of Lycopodiumspores and counted in liquid-glycerinemounts at 400� using a 10-mm grid reticle.As in previous studies (e.g., Burney 1987,1993, Burney et al. 1994), microscopic pieceswere classified as ‘‘graminoid’’ (recognizableas derived from cuticles of grasses and sedges)or ‘‘other’’ (mostly wood and dicot leaf ma-terial, often amorphous). The projected areaof each piece on the microslide was measuredby counting grid squares and placed in sizecategories of 0.5–1, 1–2, 2–4, 4–8, 8–16,16–32, 32–64, 64–128, and >128 squares.These data can then be converted to mm2/cm3 of raw sediment as in most stratigraphiccharcoal particle studies (e.g., Swain 1973,Paterson et al. 1987). Pieces in the entiresubsample residue larger than 275 mm in thesmallest dimension are caught on a sieve ofthat mesh size after chemical treatment, so
PACIFIC SCIENCE . April 2003212
they are individually counted, measured, andshown separately on the diagrams as ‘‘macro-scopic charcoal.’’ Microscopic charcoal par-ticles, depending on size distribution in thesample and taphonomic processes (Clark1988), are likely to be derived from regional,extralocal, and local fires. Macroscopic char-coal is almost exclusively derived from localsources.
In cores that showed sedimentologicalevidence for good temporal resolution in thelate Holocene (e.g., preservation of fine-scalesediment structures) an effort was made toobtain 14C dates on the level at which char-coal first shows a large increase that is likelyto be an anthropogenic signal. Whereverpossible, plant macrofossils were used for thisdating. They were treated with hot HCl toremove carbonates, followed by NaOH toremove secondary organic acids, and a secondHCl treatment. If a plant macrofossil was
not available, bulk sediments were picked forrootlets and given the same pretreatments.This is the ‘‘humin’’ fraction, primarily pol-len, spores, charcoal, and plant fibers. Alldates are corrected for isotopic fractionationand reported in radiocarbon years beforepresentG 1 s (yr b.p.), followed by the cali-brated range at 2 s (cal yr b.p. or cal yr a.d.),based on INTCAL98 (Stuiver et al. 1998).
results and discussion
Figure 1 shows the location of each site.Table 1 contains additional informationabout each location. Sites analyzed for char-coal include estuaries from the dry (a1300mm/yr) leeward southern coast and the wet(1600–2000 mm/yr) windward northerncoast. Interior peat bogs and a crater swampprovide information on sites ranging from169 to 1220 m in elevation. Increasing site
Figure 1. Location of sites on Kaua‘i used for charcoal particle stratigraphy analysis.
Kaua‘i Charcoal Stratigraphy . Burney and Burney 213
elevation correlates with increasing isolationfrom the coastal lowlands, where humanpopulation is concentrated today and pre-sumably in prehistoric times as well (Kirch1985). The interior sites all receive morerainfall than the leeward lowlands. It is likelythat all of the coastal sites were utilized forprehistoric aquaculture, and this is confirmedfor all but Limahuli Bog (Burney 2002).
Maha‘ulepu Cave System
Five years of work by a multidisciplinary teamof scientists has produced a monograph(Burney et al. 2001) on the sinkhole paleolakesite at Maha‘ulepu on the leeward southcoast. Results from land snails, bird bones,pollen, plant macrofossils, and artifacts showstratigraphic evidence for human occupation
TABLE 1
Sites for Stratigraphic Charcoal Particle Analysis from Kaua‘i
Site Latitude/Longitude/Elevation/TypeSummary of Charcoal Particle Record and Key
14C Datesa
Maha‘ulepu Caves 21� 53 0 30 00 N/159� 25 0 17 00 W2 mCave system/paleolake
Cores and excavations span the entire Holocene.Charcoal scarce in the prehuman Holocene,except at ca. 3800G 40 yr b.p. (4080–4290 calyr b.p.). First evidence for humans at 822G 60yr b.p. (1039–1241 cal yr a.d.). Large increase400 yr b.p., then declines, increases again intwentieth century.
Alekoko Fishpond 21� 57 0 07 00 N/159� 22 0 34 00 W0 mPrehistoric fishpond/estuaryb
Core begins at ca. 3200 yr b.p. Charcoalincreases markedly at 580G 30 yr b.p. (1305–1420 cal yr a.d.), when sedimentologicalevidence indicates pond construction. Afterthat time, decreases until nineteenth century.
Kekupua Fishpond 21� 56 0 46 00 N/159� 39 0 14 00 W2 mPrehistoric fishpond
Organic sedimentation begins ca. one millenniumago. At 830G 50 yr b.p. (1050–1095, 1140–1280 cal yr a.d.), large charcoal increase noted.Charcoal decreases over subsequent centuries.
Lawa‘i Kai Fishpond 21� 53 0 35 00 N/159� 30 0 26 00 W1 mPrehistoric fishpond/estuaryb
Estuary forms ca. 6000–7000 cal yr b.p. Charcoalvery scarce from that time until level reachedto which dredging was maintained. Pond filledin by marine sediments in twentieth century.
Limahuli Bog 22� 13 0 31 00 N/159� 34 0 37 00 W2 mFlood plain
Throughout prehuman Holocene, charcoal isvery scarce except ca. 3800 yr b.p. Disturbedlayer above contains much charcoal; wooddates to 1470G 60 yr b.p. (440–670 cal yra.d.); sediment to 1780G 40 yr b.p. (130–370cal yr a.d.).
Silver Falls 22� 10 0 37 00 N/159� 25 0 21 00 W169 mMontane bog
Charcoal scarce and intermittent in Holocene,increases near bog surface.
Kilohana Crater 22� 00 0 02 00 N/159� 25 0 42 00 W243 mCrater swamp
Charcoal very scarce in late Pleistocene andHolocene, increases near swamp surface.
Wahiawa Bog 21� 58 0 35 00 N/159� 30 0 35 00 W665 mMontane bog
Charcoal very scarce in late Pleistocene andHolocene, except at layer dated to 3800 yr b.p.No increase near bog surface.
Alaka‘i Swamp 22� 09 0 28 00 N/159� 35 0 56 00 W1220 mMontane bog
Charcoal very scarce in late Pleistocene andHolocene. No increase near bog surface.
a Dates given in radiocarbon years (yr b.p.)G 1 s, followed by calibrated range at 2 s (cal yr b.p. or a.d.). Calibrations, whereavailable, from Stuiver et al. (1998).
b Sites designated ‘‘prehistoric fishpond/estuary’’ appear from sedimentary and geomorphological contexts to be natural estuariesmodified by humans.
PACIFIC SCIENCE . April 2003214
in the vicinity in late prehistoric times. Theoldest dated evidence for a human presence isan accelerator mass spectrometry determina-tion on the gelatin fraction from the pelvis ofthe Polynesian-introduced Pacific rat (Rattusexulans) at 822G 60 yr b.p. (1039–1241 cal yra.d.), from a depth of 305–315 cm below sur-face and below datum in the lake sediments.
Unfortunately, sediments around this levelare highly disturbed by an inferred tsunamievent that occurred at the site ca. 400 yr b.p.(Burney et al. 2001). In any case, a large in-crease of charcoal particle deposition occursin the cores and excavations in the site fromca. 400 yr b.p. (i.e., as soon as normal sedi-mentation resumes after the tsunami [Figure2]). Thus it is not feasible to conclude fromthe charcoal stratigraphy at this site that therewas a local human presence in the vicinitybefore ca. four centuries ago, although thepresence of the Pacific rat eight centuries agoconfirms that people were on the island.
The sediments in this site are virtuallycharcoal-free in the levels predating the be-ginning of the second millennium a.d., with anotable exception. Core 6, which is the mostcomplete Holocene section from the site, is a9.5-m piston core from near the center of thepaleolake. In this core, there is a very dis-tinctive prehuman charcoal peak that beginsat ca. 475 cm below surface (560 cm belowdatum) and continues up through about 50cm of sediment. This plateau, composed ofroughly one-third graminoid charcoal, beginsabout 3800G 40 yr b.p. (4290–4080 cal yrb.p.). It is not surprising that the other twocores from the site analyzed for charcoalfailed to detect this event, because sedi-mentological evidence suggests that only thecenter of the lake, in the vicinity of the core 6location, was receiving sediment at that time,probably as a result of low-water conditions.The pollen evidence from the core (Burney etal. 2001) also indicated aridification at thattime. Athens (1997) noted prehuman pollenevidence for aridity on O‘ahu in the mid- tolate Holocene.
Charcoal also reaches a high level at725 cm below surface in core 6 and moder-ate levels intermittently between these twospikes. The source of this charcoal is un-
known, but the presence of particles in thelarger size classes, including macroscopiccharcoal, indicates that the source was proba-bly on Kaua‘i rather than, for instance, a ma-jor volcanic eruption on Maui or the BigIsland of Hawai‘i, 200–400 km leeward ofKaua‘i.
Alekoko (Menehune) Fishpond
A registered National Historical Landmark,the fishpond originally known as Alekoko(commonly referred to as the MenehuneFishpond in tourist guidebooks) is ideallysituated to provide information regardingchanges in the lowlands of the southeasterncorner of Kaua‘i. Its aspect is transitionalbetween windward and leeward, with meanannual precipitation of ca. 1300 mm. It is aplace of great traditional, archaeological, andhistorical interest as one of the finest exam-ples in the entire archipelago of prehistoricstonework and fishpond construction (Ki-kuchi 1976). Said to have been built over-night by the legendary menehune, it has alsobeen suggested that this pond was built in thenineteenth century by Chinese plantation la-borers. Early historical records and photosshow that the pond predates the arrival ofChinese laborers (Ching et al. 1973). Thecurrent consensus among archaeologists isthat most fishponds in the Hawaiian Islandswere built by Polynesians well before Euro-pean contact to practice advanced aquacultureand wetland agriculture (Kikuchi 1976, Kirch1985).
A 4-m core collected from the pond(Burney 2002) showed that the estuary hasdeposited sediments there at least since3230G 70 yr b.p. (3620–3340 cal yr b.p.). Aunique layer of woody detritus marks a fun-damental change to finer sediments at 580G30 yr b.p. (1305–1420 cal yr a.d.). Charcoalparticles begin increasing ca. 15 cm below thewoody layer. This material lacks a graminoidcomponent (Figure 3). At and above thewoody detritus layer, both graminoid andother charcoal reach values more than 10times any previous value in the core. Thisincrease is followed, after ca. 20 cm ofcharcoal-laden sediment, with a marked de-
Kaua‘i Charcoal Stratigraphy . Burney and Burney 215
Figure
2.T
hre
ech
arco
alp
rofi
les
fro
mth
esi
nkh
ole
pal
eola
kein
the
Mah
a‘u
lep
uC
ave
syst
emo
nth
eso
uth
coas
to
fK
aua‘
i.T
he
Eas
tP
itp
rofi
leis
fro
mth
eea
ster
ned
geo
fth
esi
nkh
ole
;co
re20
isfr
om
nea
rth
en
ort
her
ned
ge;c
ore
6is
atth
eap
pro
xim
ate
geo
grap
hic
cen
ter
of
the
sin
kho
le.R
om
ann
um
eral
sin
firs
tco
lum
nin
dic
ate
stra
tigr
aph
icu
nit
s:(I
)a
pre
vio
us
inte
rgla
cial
;(I
I)ca
.95
00ca
lyr
b.p
.;(I
II)
ca.
7000
cal
yrb.p
.;(I
V)
mid
-to
late
Ho
loce
ne;
(V)
late
Ho
loce
ne;
(VI)
1430
–16
65ca
lyr
a.d
.;(V
II)
late
eigh
teen
thce
ntu
ryto
nin
etee
nth
cen
tury
;(V
III)
nin
etee
nth
toea
rly
twen
tiet
hce
ntu
ry;(
IX)
mid
-tw
enti
eth
cen
tury
top
rese
nt.
See
Bu
rney
etal
.(20
01)
and
Bu
rney
(200
2)fo
rd
atin
gb
asis
.
crease to levels slightly above the backgroundvalues in previous millennia. Charcoal par-ticles increase again in sediments of theuppermost meter of the core, inferred to
be post-European contact from sedimentarycharacteristics (Burney 2002).
We interpret this charcoal record to sug-gest that human settlement at the site extends
Figure 3. Charcoal stratigraphy from the Alekoko (Menehune) Fishpond near Lıhu‘e on the southeastern coast ofKaua‘i.
Kaua‘i Charcoal Stratigraphy . Burney and Burney 217
back at least seven centuries and that duringmuch of the last few centuries of Polynesianoccupation before the arrival of CaptainCook in 1778, relatively little burning of grassor trees occurred in the immediate vicinity. Itis not known whether this is because the landnearby was nearly all covered with intensiveagricultural and residential use soon after thefishpond was constructed or because of someinherent tendency of fire-ravaged landscapesto generate less charcoal over time after initialhuman settlement (see Burney et al. 1994 fora discussion of this pattern in the West In-dies, Burney 1996 for Madagascar).
Kekupua Fishpond
Between Makaweli and Waimea on thesouthwestern coast of Kaua‘i there is a small,shallow body of fresh water at the Kapala-wai location known as the Kekupua Fish-pond, said to be a prehistorically constructed,spring-fed reservoir. Four cores have beencollected from the pond. The longest of thesehas been dated and described (Burney 2002).
This 220-cm core contains a record of ca.one millennium. At 170 cm, a level dated to980G 80 yr b.p. (895–1225 cal yr a.d.) on afruit of hala or screw pine (Pandanus tectorius),soft sediments overlie gravel. These sedi-ments are low in charcoal particles (Figure 4).At 140 cm, dated on the humin fraction ofbulk sediments to 830G 50 yr b.p. (1050–1095, 1140–1280 cal yr a.d.) both graminoidand other charcoal levels are quite high, andthis sediment is interpreted as the beginningof deposition after the pond was constructed(Burney 2002). Low to moderate charcoallevels have been recorded in the sedimentsbetween 105 cm and the surface.
Although the inferred date for initial fish-pond construction is surprisingly early, this isthe most parsimonious explanation for thechange in sedimentation regime at that time,from coarse clastics to fine organic detritus.Although the difficulties of dating the pond’sestablishment are acknowledged in Burney(2002), firmer conclusions await completionof studies of additional cores taken from thesite. To date, no oral traditions have been
recorded in the vicinity that provide preciseclues to the age of the pond, beyond con-firming that it long predates European con-tact.
Lawa‘i Kai
At Lawa‘i Kai, roughly midway betweenMaha‘ulepu and Kekupua Fishpond on thesouth coast (Figure 1), there is another fish-pond believed to have been constructed inprehistoric times. This is a small naturalestuary on the property of the NationalTropical Botanical Garden that has been im-pounded by stone walls and packed earthalong its west side to form a fishpond nowentirely filled with marine sediments de-posited by historically known tsunamis andhurricanes (Burney 2002). The longer of twocores collected from the site was a 654-cmrecord that shows a good stratigraphic cross-correlation with the Maha‘ulepu sedimentaryrecord (Burney et al. 2001, Burney 2002). Inboth sites, sea level approached the base levelof the basin (ca. 5 m below current sea level)ca. 6000–7000 cal yr b.p. Brackish or fresh-water ponding followed, with an accumula-tion of fine sediments interrupted by periodiccoarse-clastic influx that may represent hur-ricanes or tsunamis. The natural sedimentaryrecord at Lawa‘i Kai appears to be truncated,perhaps by dredging, at ca. 180 cm belowsurface. At that level, large amounts of‘‘other’’ charcoal and a trace of graminoidsappear for the first time in the record (Figure5). Above that level, the sediments are marinein origin, probably deposited by the 1946tsunami and Hurricanes ‘Iwa (1982) and ‘Iniki(1992). This reconstruction is consistent withlocal accounts (R. Gage, pers. comm., 1998)that the pond was maintained by dredging ata sufficient depth to raise mullet (Mugil ceph-alus) and perhaps other fish until the site wasdevastated by the 1946 tsunami overwash andfurther choked with sediments from the twohurricane storm surges.
Limahuli Bog
There is a small fenlike wetland seaward ofthe entrance to Limahuli Gardens on Kaua‘i’s
PACIFIC SCIENCE . April 2003218
north shore (Figure 1) that has receivedterrestrial sediments at least intermittentlythroughout the Holocene (Burney 2002). A6-m core from the site contains very littlecharcoal until a point 15 cm above a leveldated 3910G 60 yr b.p. (4515–4465, 4450–4155 cal yr b.p.). A dense layer of wood pre-vented sediment recovery in much of lateprehuman times, until a disturbed layer isreached in which large tree trunks are mixedwith coarse and fine sediments (Figure 6). Atthat level, a piece of the outermost 2 cm of aburied log yielded an age of 1470G 60 yr b.p.
(440–670 cal yr a.d.). Sediments above thiswood show a large (10 times) increase incharcoal. A bulk date on the humin fractionof associated sediments from a depth of 229–230 cm was 1780G 40 yr b.p. (130–370 cal yra.d.). It is unfortunate that the sediments inthis key part of the record are disturbed(possibly by a prehistoric tsunami—see Bur-ney 2002). The charcoal plateau in this core isnot firmly dated, but it raises the possibilitythat settlement on the wet northern coastpredates the south coast settlements by ca.five centuries.
Figure 4. Charcoal stratigraphy from Kekupua Fishpond near Makaweli on the southwestern coast of Kaua‘i.
Kaua‘i Charcoal Stratigraphy . Burney and Burney 219
Silver Falls
Also on the north side of the island is a sitethat is the lowest in elevation of the interiorbogs examined. This interfluvial swamp islocated inshore from the Kılauea Point areaat an elevation of 169 m (Figure 1). A 3.5-mrecord from the two adjacent cores collectedat this site begins at a level dated on the hu-min fraction of bulk sediments to 8160G 40yr b.p. (Figure 7). The sediments mostlyconsist of clays laid down during the subse-quent approximately two millennia of theearly Holocene. At a level interpolated to ca.seven millennia ago, there is a small peak of‘‘other’’ charcoal. Charcoal appears again atca. 70 cm below surface, but no large accu-
mulation was detected. Dating of this layerwas not attempted because of the abundanceof fresh roots down to this zone and the lackof plant macrofossils in the sandy silty clay.Charred logs are present on and near thesurface, evidence for twentieth-century firesremembered by elderly people in the area(U. Ratchener, pers. comm., 1999).
Kilohana Crater
The heavy clays on the floor of the largeswamp forest at the bottom of Kilohana Cra-ter in Kaua‘i’s southeastern interior (Figure1) are underlain by layers of peaty sedimentsfrom the late Pleistocene. Coarse peat from
Figure 5. Charcoal stratigraphy from the prehistoric fishpond at Lawa‘i Kai, on the south central coast of Kaua‘i.
PACIFIC SCIENCE . April 2003220
170 cm below surface in a 3-m profile yielded26,020G 230 yr b.p. (Figure 8).
Resolution of Holocene events at this siteis poor, as in some other interior bogs of thearchipelago (e.g., Hotchkiss 1998). It shouldbe noted, however, that charcoal particles areessentially absent from the late Pleistocenesediments and only begin to appear ca. 25 cmbelow the surface, in the humic silty clay thatforms the uppermost sedimentary unit.
Wahiawa Bog
Also known as Kanaele Swamp, WahiawaBog is a large, remote montane bog in thesouth central interior of the island at 665 m(Figure 1). A 4.9-m core obtained here (Bur-ney 2002) yielded a basal age of 23,090G 180yr b.p. from the humin fraction of bulk sedi-ment (Figure 9). It is not clear why the twoHolocene dates near the top of the core are
Figure 6. Charcoal stratigraphy from a bog on the north coast of Kaua‘i at Limahuli Gardens. White space in thelithology diagram indicates a gap in core recovery due to woody nature of the layer.
Kaua‘i Charcoal Stratigraphy . Burney and Burney 221
stratigraphically reversed, but the lower ofthe two could be contaminated somehow bymaterial from above, or the upper by re-deposited material eroding elsewhere in theswamp. Either is possible. The presence of atrace of charcoal around the level dated to3800G 40 yr b.p. is intriguing, because this isthe date of the prehuman charcoal peak atMaha‘ulepu. In any case, it is apparent thatcharcoal is otherwise very scarce at this site
in the late Pleistocene and throughout theHolocene.
Alaka‘i Swamp
The Kilohana region of Alaka‘i Swamp (notto be confused with Kilohana Crater else-where on the island, see earlier in Results) isan open grass-sedge bog area of northwesternKaua‘i at ca. 1220 m (Figure 1). The vicinity
Figure 7. Charcoal stratigraphy from a bog at 169 m elevation on the Silver Falls Ranch on the north side of Kaua‘i.
PACIFIC SCIENCE . April 2003222
was an important coring site for the palyno-logical studies of Selling (1948), which un-fortunately predated the radiocarbon method.Subsequently, we have dated a core from ap-proximately the same location (Burney 2002)with a basal age of 15,100G 80 yr b.p.(18,525–17,623 cal yr b.p.), and a date of5000G 50 yr b.p. (5895–5615 cal yr b.p.)from 70 cm below surface.
No substantial amount of charcoal appearsin this core at any level examined. Sampleswere processed and counted at 15-cm inter-vals, but no diagram was prepared because
only a few very small particles were detectedat any level and none at most levels. BothHolocene and late Pleistocene sediments areessentially charcoal-free.
conclusions
Three interior sites contained sediments de-posited in the late Pleistocene. None of thesesediments contains more than a trace of verysmall charcoal particles. Fire occurrence mayhave been low at that time as a result of evenless lightning activity than today on Kaua‘i.
Figure 8. Charcoal stratigraphy from the swamp forest on the floor of Kilohana Crater at 243 m elevation in theinterior of Kaua‘i.
Kaua‘i Charcoal Stratigraphy . Burney and Burney 223
The early Holocene was a time of low fireoccurrence as well, although more charcoalwas recorded at times on both coasts and inthe interior than in the Pleistocene sedimentsfrom these sites. By contrast, the charcoalrecord from Flat Top Bog on the flank ofHaleakala Volcano on Maui (Burney et al.1995) shows early and mid-Holocene char-coal values that were quite high, probably as-sociated with the activity of the volcano.The sudden increase of charcoal at three ormore Kaua‘i sites (Maha‘ulepu, Limahuli,Wahiawa, and perhaps Silver Falls) around3800 yr b.p. (4000–4300 cal yr b.p.), followedby a decline to low background values, isintriguing. A multidecadal drought could ex-
plain the occurrence of fires at that time, es-pecially if combined with an increase in drylightning occurrence. Although an extremedrought event, perhaps lasting three cen-turies, is believed to have caused the collapseof Mesopotamian civilization at precisely thattime (reviewed in deMenocal 2001), it is notknown whether similar conditions prevailedin the mid-Pacific. Pollen and sedimento-logical evidence from Maha‘ulepu (Burneyet al. 2001) is also consistent with droughtconditions at that time.
Charcoal evidence for human arrival con-firms that people were widespread in theKaua‘i lowlands, including the dry leewardvalleys, by ca. a.d. 1100. Evidence that hu-
Figure 9. Charcoal stratigraphy from Wahiawa Bog at 665 m in the interior of Kaua‘i. White space in the lithologydiagram indicates a gap in core recovery due to the presence of an indurated layer.
PACIFIC SCIENCE . April 2003224
mans were at Limahuli on the north coast ca.500 yr earlier is equivocal. The pattern ofinitial peaking of presumably anthropogeniccharcoal particles, with a plateau of highcharcoal values for a few centuries, followedby a drastic decline until European times isdocumented here from Maha‘ulepu, Lima-huli, Alekoko, and Kekupua. Interior sitesshow that charcoal, and therefore human-caused burning, was not prevalent in thehigher interior areas in prehistoric Polynesiantimes, but European-period fires, probablyfrom historically documented activities suchas forest clearance and sugarcane burning, areclearly registered near the top of some cores.Flat Top Bog in the high interior of Mauishows the same pattern, with no discerniblePolynesian charcoal signal but a substantialrise after European-introduced pollen typesappear in the sediments (Burney et al. 1995).
Perhaps the Polynesian land-managementstrategy in prehistoric times was to clearlowland areas with fire, but it may be thatsettlement was too dense in many low areasby ca. a.d. 1600 or before to allow muchcontinued biomass burning. Macrofossil andpollen data from Maha‘ulepu (Burney et al.2001) show that some indigenous trees werestill present at the site about the time of Eu-ropean arrival or just before (although none ispresent today), but it is not known how muchactual forest remained or to what extent firewas employed as a landscape-managementtool at this late stage in prehistoric agricul-tural development of the lowlands.
As for the interior, it appears from thecharcoal signals from Silver Falls and Kilo-hana Crater that elevations up to ca. 300 mwere occupied in Polynesian times, buthigher sites such as Wahiawa Bog at 665 mand Alaka‘i Swamp at 1220 m (and Flat TopBog on Maui at 2270 m [Burney et al. 1995])either were not heavily settled by humansor else were settled without much biomassburning. Substantial charcoal at Limahuli inearly Polynesian times (or just before) sug-gests that coastal windward areas were subjectto burning despite the high rainfall. The un-likelihood of prehuman fires under such wetconditions with no lengthy dry season (seeBurney 1996, 1999 for a discussion of a re-lated situation in the rain forests of Mada-
gascar) makes it more plausible that thecharcoal plateau initiated at Limahuli at ca.1470 yr b.p. is anthropogenic. The dating ofthis event is problematic, however, and thedate from bulk sediments suggests that thepeak could be even older (1780 yr b.p.). Ifthese dates are discounted due to the evi-dence for tsunami disturbance and sedimentmixing, then no evidence for humans onKaua‘i presented here predates the beginningof the second millennium a.d. This is close tothe estimate of Athens (1997) for late humanarrival in the Hawaiian Islands (ca. a.d. 800).
What is clearly needed to further elucidatethe fire history of Kaua‘i and make confidentcomparisons with dated evidence for last oc-currence of extinct organisms is a sedimen-tary record from a site with fine-scale (e.g.,decadal to annual) resolution of the Holo-cene. To our knowledge, such a site has notbeen found so far.
acknowledgments
Coring efforts were assisted by Adam As-quith, Alec Burney, Reginald Gage II, DonHeacock, Helen James, Pila and Dolly Ki-kuchi, Kristina Kikuchi-Palenapa, LorrinMano‘i, Ka‘ohulani McGuire, Cameron Mc-Neil, Bob Nishek, Storrs Olson, Mary L.O’Rourke, and Ed Sills. Thanks also to thelandowners, public and private, who per-mitted us to work on their property. LaFrance Kapaka and Nancy McMahon assistedwith official permissions.
Literature Cited
Athens, J. S. 1997. Hawaiian native lowlandvegetation in prehistory. Pages 248–270 inP. V. Kirch and T. L. Hunt, eds. Histori-cal ecology in the Pacific Islands. YaleUniversity Press, New Haven, Connecti-cut.
Beggerly, P. E. P. 1990. Kahana Valley, Ha-waii, a geomorphic artifact. Ph.D. diss.,University of Hawai‘i at Manoa, Hono-lulu.
Burney, D. A. 1987. Late Quaternary strati-graphic charcoal records from Madagascar.Quat. Res. 28:274–280.
———. 1993. Late Holocene environmental
Kaua‘i Charcoal Stratigraphy . Burney and Burney 225
changes in arid southwestern Madagascar.Quat. Res. 40:98–106.
———. 1996. Climate change and fire ecol-ogy as factors in the Quaternary biogeog-raphy of Madagascar. Pages 49–58 in W.Lourenco, ed. Biogeographie de Madagas-car. ORSTOM, Paris.
———. 1999. Rates, patterns, and processesof landscape transformation and extinc-tion in Madagascar. Pages 145–164 in R.D. E. MacPhee, ed. Extinctions in neartime: Causes, contexts, and consequences.Plenum/Kluwer, New York.
———. 2002. Late Quaternary chronologyand stratigraphy of twelve sites on Kaua‘i.Radiocarbon 44:13–44.
Burney, D. A., L. P. Burney, and R. D. E.MacPhee. 1994. Holocene charcoal stra-tigraphy from Laguna Tortuguero, PuertoRico, and the timing of human arrival onthe island. J. Archaeol. Sci. 21:273–281.
Burney, D. A., R. V. DeCandido, L. P. Bur-ney, F. N. Kostel-Hughes, T. W. StaffordJr., and H. F. James. 1995. A Holocenerecord of climate change, fire ecology andhuman activity from montane Flat TopBog, Maui. J. Paleolimnol. 13:209–217.
Burney, D. A., H. F. James, L. P. Burney, S.L. Olson, W. Kikuchi, W. L. Wagner, M.Burney, D. McCloskey, D. Kikuchi, F. V.Grady, R. Gage, and R. Nishek. 2001.Fossil evidence for a diverse biota fromKaua‘i and its transformation since humanarrival. Ecol. Monogr. 71:615–641.
Ching, F. K. W., P. B. Griffin, W. K. Ki-kuchi, W. A. Albrecht, J. C. Belshe, and C.Stauder. 1973. The archaeology of Puna,Kaua‘i. Archaeological Research CenterHawaii, Lıhu‘e, Kaua‘i.
Clark, J. S. 1988. Particle motion and thetheory of charcoal analysis: Source area,transport, deposition, and sampling. Quat.Res. 30:67–80.
deMenocal, P. B. 2001. Cultural responses toclimate change during the late Holocene.Science (Washington, D.C.) 292:667–673.
Flannery, Tim. 1994. The future eaters.George Braziller, New York.
Grayson, D. K. 2001. The archaeological
record of human impacts on animal pop-ulations. J. World Prehist. 13:1–68.
Hotchkiss, S. C. 1998. Quaternary vegetationand climate of Hawai‘i. Ph.D. diss., Uni-versity of Minnesota, St. Paul.
Kikuchi, W. K. 1976. Prehistoric Hawaiianfishponds. Science (Washington, D.C.)193:295–299.
Kirch, P. V. 1985. Feathered gods and fish-hooks. University of Hawai‘i Press, Ho-nolulu.
Martin, P. S., and D. W. Steadman. 1999.Prehistoric extinctions on islands and con-tinents. Pages 17–55 in R. D. E. MacPhee,ed. Extinctions in near time: Causes, con-texts, and consequences. Plenum/Kluwer,New York.
Masse, W. B., and H. D. Tuggle. 1998. Thedate of Hawaiian colonization. Pages 229–235 in C. M. Stevenson, G. Lee, and F. J.Morin, eds. Easter Island in Pacific con-text: South Seas symposium: Proceedingsof the 4th International Conference onEaster Island and East Polynesia. EasterIsland Foundation Occasional Paper 4.Bearsville and Cloud Mountain Presses,Los Osos, California.
Olson, S. L., and H. F. James. 1982. Fossilbirds from the Hawaiian Islands: Evidencefor wholesale extinction by man beforeWestern contact. Science (Washington,D.C.) 217:633–635.
Paterson, W. A., K. J. Edwards, and D. J.MacGuire. 1987. Microscopic charcoal asa fossil indicator of fire. Quat. Sci. Rev.6:3–23.
Selling, O. 1948. Studies in Hawaiian pollenstatistics: Part III. On the Late Quaternaryhistory of the Hawaiian vegetation. Ber-nice P. Bishop Mus. Spec. Publ. 39.
Stuiver, M., P. J. Reimer, E. Bard, J. W.Beck, G. S. Burr, K. A. Hughen, B.Kromer, F. G. McCormac, J. v. d. Plicht,and M. Spurk. 1998. INTCAL98 14C. Ra-diocarbon 40:1041–1083.
Swain, A. M. 1973. A history of fire andvegetation in north-eastern Minnesota, asrecorded in lake sediments. Quat. Res.3:383–396.
PACIFIC SCIENCE . April 2003226
