Content uploaded by Margaret J. Schoeninger
Author content
All content in this area was uploaded by Margaret J. Schoeninger on Oct 18, 2016
Content may be subject to copyright.
Society for American Archaeology
The Late Woodland Diet on Nantucket Island and the Problem of Maize in Coastal New England
Author(s): Elizabeth A. Little and Margaret J. Schoeninger
Source:
American Antiquity,
Vol. 60, No. 2 (Apr., 1995), pp. 351-368
Published by: Society for American Archaeology
Stable URL: http://www.jstor.org/stable/282145
Accessed: 02/10/2008 12:19
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=sam.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the
scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that
promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.
Society for American Archaeology is collaborating with JSTOR to digitize, preserve and extend access to
American Antiquity.
http://www.jstor.org
THE LATE WOODLAND DIET ON NANTUCKET ISLAND AND THE
PROBLEM OF MAIZE IN COASTAL NEW ENGLAND
Elizabeth A. Little and Margaret J. Schoeninger
Carbon and nitrogen isotope ratios of () bone collagen from six burials of the Late Woodland Period at Nantucket
Island, Massachusetts, and (2) a wide range of potential dietary materials provide data for evaluating coastal diets.
Archaeological and historical data give evidence for the availability and use of dietary items. The bases of the food
chains and trophic levels define the possible food groups: terrestrial C3 and C4 plants and their consumers, marine
C3 or C4-like plants and their consumers, and marine carnivores. From these data, computer analysis of multiple
linear mixing equations relating isotope ratios of human bone collagen to those of dietaryfood groups shows allowable
ranges of these food groups in the diet. The results argue for a diet of 40-65 percent oceanic animals, with the rest
consisting of substantial amounts of animals from salt marsh and eelgrass meadows or of maize, and minor amounts
of C3 plants and their consumers.
Las proporciones de isotopos de carbono y nitrogeno tomadas de (1) coldgeno de hueso, sacados de seis entierros del
periodo Woodland Tardio en la Isla Nantucket, Massachusetts; y (2) una gran variedad de materiales alimenticios
posibles, proveen datos para valorar las dietas costenas. Los datos historicos y arqueologicos establecen evidencia
sobre la utilizacion y disponibilidad de los consumibles. Las bases de la sucesion de comestibles y los niveles troficos
determinan los grupos alimenticios posibles: plantas terrestres C3 y C4
y sus consumidores; plantas marinas parecidas
a las plantas C3 y C4 y sus consumidores, y carnivoros marinos. Un andlisis por computadora de ecuaciones mez-
cladores multi-lineares que relaciona razones isotopicos de coldgeno de hueso humano a las de los grupos alimenticios,
demuestra limites permisibles de estos grupos alimenticios en el regimen. Los resultados indican una dieta de 40-
65 por ciento de animales ocednicos y lo demds es compuesto de cantidades sustanciosas de animales del saladar y
de las praderas de Zostera marina, o de maiz, y cantidades pequenas de plantas C3 y sus consumidores.
A major anthropological issue in coastal
New England and New York concerns
variations in the intensity of use of maize in
the Late Woodland period (about cal A.D.
1000 to 16001). Beginning about cal A.D.
1000 in Illinois, West Virginia, and Ohio,
carbon stable isotope ratios in human bone
collagen indicate a marked shift toward de-
pendence on maize (van der Merwe and Vo-
gel 1978). Archaeological evidence from
coastal southern New England-in 1991 a to-
tal of only 12 maize kernels dating to the Late
Woodland period (Bernstein 1992)-has sug-
gested that such a dietary shift occurred on
the coast somewhat later (Bendremer et al.
1991; Ceci 1990). Distinguishing between the
presence of maize and the intensive use of
maize (McBride and Dewar 1987), a rise in
herb pollen such as would result from land
clearing for horticulture does not occur until
after cal A.D. 1280-1400 in Rhode Island,
and not until the early 1600s on Cape Cod,
Massachusetts (Winkler 1985). A single maize
kernel was dated at Martha's Vineyard Island
at cal A.D. 1192-1290 (charcoal; Ritchie 1969:
52), while at nearby Nantucket Island several
maize kernels in the historical association
collections lack a securely dated context, and
there is no evidence for land clearing until
about A.D. 1660 (Dunwiddie 1990). Al-
though maize is verified in the coastal zone
from the twelfth century on, there is no ev-
idence that people of the southern New En-
gland coast participated in large-scale or in-
Elizabeth A. Little * Nantucket Historical Association, 37 Conant Road, Lincoln, MA 01773
Margaret J. Schoeninger * Department of Anthropology, University of Wisconsin, Madison, WI 53706-1393
American Antiquity, 60(2), 1995, 351-368.
Copyright ? 1995 by the Society for American Archaeology
351
AMERICAN ANTIQUITY
tensive horticulture like that found up the
Connecticut River (Bendremer et al. 1991)
until well after cal A.D. 1300, if at all. Be-
cause we recognize the difficulty of making
such an argument on negative evidence, we
undertook the project described in the fol-
lowing pages.
Previous studies (reviewed in Schoeninger
and Moore 1992) have demonstrated the use-
fulness of stable carbon isotope ratios, rep-
resented as $13C values in standard per mil
notation,2 in distinguishing between two sets
of dietary variables. First, in areas that lack
marine foods, it is often possible to distin-
guish the relative amount of C4 plants (chiefly
maize but including CAM plants with C4 sig-
natures) and C3 plants3 in the food web. Sec-
ond, in areas with no maize, it is often pos-
sible to distinguish the relative amount of
marine and C3 foods. Because marine foods
have 613C values that fall between the end
points described by C4 and C3 plants, the use
of bl3C values alone cannot distinguish be-
tween a marine and a terrestrial diet. For this
reason, recent coastal studies (Keegan and
DeNiro 1988; Schoeninger et al. 1990) have
explored the use of both carbon and nitrogen
stable isotope ratios, represented as b'sN val-
ues in standard per mil notation. With both
613C and e35lN
measurements, it becomes pos-
sible to distinguish three dietary variables:
(1) plankton-based marine fauna, (2) terres-
trial C3 plants and consumers, and (3) C4
plants and their consumers. Here we focus
on Nantucket Island, Massachusetts, because
it provides both Late Woodland human bur-
ials and a wide variety of potential food items
identified from historic records of the earliest
European contact in the area (see Medaglia
et al. 1990). Analyses using b13C and b'sN
permit the formulation of linear mixing equa-
tions relating the bl3C and 6b5N values of the
foods to the values in human collagen. So-
lutions of the equations give allowable ranges
for each food group.
Background
Nantucket Island is a low, sandy island about
130 km2 in area, with rich marine, salt marsh,
and seagrass meadow resources, that lies 40
km south of Cape Cod at 41? N (Figure 1).
The temperature-moderating effect of the
ocean provides warmer winters and cooler
summers than on the mainland and a 165-
day frost-free growing season, May 7 to Oc-
tober 17 on average, more than adequate for
maize horticulture. Even so, the soil on gla-
cial outwash deposits is acidic, and most of
the land is too porous for agriculture (Lan-
glois 1977). Today's list of native mammals
includes only deer, mice, voles, and bats, but
shell deposits (middens) contain remains of
deer, dog, raccoon, fox, muskrat, vole, and
mouse. Marine resources are readily avail-
able and plentiful throughout the year. Fish
may be caught in ponds, along shores, and
from boats. Gray seals breed off the west end
of Nantucket, as many as 1,000 harbor seals
winter in Nantucket Sound, and whales and
waterfowl migrate seasonally near the island
(Little and Andrews 1982). Shellfish habitat
and associated historic and Late Woodland
shell middens are found along the coast of
Massachusetts at the edge of harbors, estu-
aries, and tidal creeks (Little 1988; Mc-
Manamon et al. 1986).
Archaeological Studies
Archaeological studies at Nantucket, includ-
ing a 1978 survey funded by the Department
of the Interior through the Massachusetts
Historical Commission, provide a substan-
tial database for research (Little 1985a;
Luedtke 1980a). Most northeastern projectile
point styles of the past 11,000 years are found
here, and include diagnostic artifacts of the
Paleoindian, Archaic, and Woodland peri-
ods. The local glacial drift contains most of
the lithic materials used by prehistoric in-
habitants, except soapstone (steatite), New
York cherts, and Pennsylvania jasper
(Luedtke 1987). Since Nantucket became an
island after about 5000 B.P. (Oldale 1986),
these lithics provide evidence for coastal nav-
igation.
Although Late Woodland middens are
dominated by shellfish remains, they often
preserve remains of deer, fish, waterfowl,
[Vol. 60, No. 2, 1995]
352
REPORTS
El Principal Shellfish Habitat
-- South Bound Shell Midden Zone
* Burial Sites
intucket km
Figure 1. Map of Nantucket Is-
land showing shellfish habitat,
adjacent shell midden zone, and
approximate sites of six Late
Woodland Nantucket burials. In-
set shows coastal Massachusetts
with shellfish habitat (dotted),
which correlates with major har-
bors, modern urban centers, and
Late Woodland archaeological
sites (Little 1988).
crustaceans, turtles, and, rarely, seal and whale
(Little 1985b; Pretola and Little 1988). Bar-
bara E. Luedtke (personal communication
1989) analyzed a pit at Quaise, Nantucket,
containing bone that was, by weight, 63 per-
cent mammal (76 percent deer), 37 percent
fish (24 percent sturgeon), and 1 percent
bird-a pattern found also at Nauset on Cape
Cod dating between cal A.D. 970 and 1340
(Borstel 1984; Fitzgerald 1984). It is not pos-
sible to estimate the actual contribution of
each of these categories to prehistoric human
diet, since the overwhelming weight contri-
bution by shells and terrestrial mammal bone
may reflect preservation and discard patterns
rather than actual use.
Other areas of coastal southeastern New
England have a greater range of food items
than have been recovered archaeologically on
Nantucket (Lavin 1988; Ritchie 1969). Pre-
historic extinction of most of the island's ter-
restrial mammals can account for this defi-
ciency, although inadequate recovery tech-
niques of the past are a factor. Large marine
mammal bones would have been left at the
beach, where the next storm tide would have
swept them into the sea. Finally, the locations
of sites of landing fish, of cleaning fish, and
of disposal of fishbones (Hayden et al. 1987),
may not have been at shell middens.
Reports of Early European Explorers and
Settlers
The first real description of the people and
foods of coastal southeastern New England
comes from Champlain, who visited Cape
353
AMERICAN ANTIQUITY
Cod in July 1605 and October 1606, some
75 or more years following the first European
contact in the region. He reported people who
"are not so much great hunters as good fish-
ermen and tillers of the land" (Champlain
1880:124), with gardens of maize near each
dwelling (Little 1988) and diets supplement-
ed with Brazilian beans, squashes, tobacco,
Jerusalem artichokes, as well as oysters and
fish. Similar early lists of native foods exist
for the Elizabeth Islands (Brereton 1602),
Plymouth (Mourt 1986 [1622]), Lynn (Wood
1865 [1635]), Narragansett Bay (Williams
1936 [1643]), eastern Massachusetts (Gookin
1970 [1674]), and Nantucket (Crevecoeur
1971 [1782]). Like Cape Cod at this time,
Nantucket had few large mammals, but on
the Massachusetts mainland:
Their food is generally boiled maize, or Indian corn,
mixed with kidney-beans, or sometimes without. Also
they frequently boil in this pottage fish and flesh of all
sorts, either new taken or dried, as shads, eels, alewives
or a kind of herring, or any other sort of fish ... ven-
ison, beaver, bears flesh, moose, otters, rackoons, or
any kind they take in hunting.... Also they mix with
the said pottage several sorts of roots; as Jerusalem
artichokes, and ground nuts, and other roots, and
pompions, and squashes, and also several sorts of nuts
or masts, as oak-acorns, chestnuts, walnuts; these
husked and dried, and powdered, they thicken their
pottage therewith. (Gookin 1970 [1674])
After 1659 about 34 Englishmen pur-
chased land on Nantucket from the local "sa-
chems," native rulers in southeastern Mas-
sachusetts (Starbuck 1924). From the Nan-
tucket Proprietors' Records (NPR) and Nan-
tucket County Deeds (NCD) of the
seventeenth and eighteenth centuries at the
Nantucket County Courthouse, Nantucket,
Massachusetts, we learn that the historic na-
tives of Nantucket kept dogs (NPR 1:27),
burned their planting fields in April (NPR
1:5), planted corn and "ponckence" (Star-
buck 1924:147) after May 1 (NCD 2:7; Star-
buck 1924:127), caught fish in weirs (NPR
1:6), and owned bows (NCD 2b:3) and ca-
noes, probably dugouts, for water transport
and fishing (NCD Grantee 4:19, 21). In 1782
Crevecoeur reported that, "Before the arrival
of the Europeans, [the Nantucket aborigines]
lived on the fish of their shores" (1971:103).
According to a report of the Reverend James
Freeman (1807), the original inhabitants were
acquainted with boiling but not with roasting,
and cultivated only maize, beans, squashes,
and tobacco. "They could now and then kill
a bird; and there were a few deer .... Fish
could be obtained in the harbours, and on
the coast; and shell fish were abundant. Dur-
ing winter however, they frequently suffered
the extremities of famine. Their clothes were
sometimes skins, but for the most part coarse
mats, made of grass" (Freeman 1807:35).
"The natives of Nantucket... at their feasts
... had several sorts of good food, and very
good strong beer" (Macy 1792). "[Some] ...
occupied the land around their dwellings with
gardens, wherein they raised corn and vege-
tables of various kinds, some of which they
sold to the English. They frequently had fruit
trees in their gardens and near their houses"
(O. Macy 1842 in Starbuck 1924:612). These
dietary patterns, noted long after the original
contact by Europeans in southern New En-
gland, may not reflect the diets of earlier in-
habitants (see Bernstein 1992), and the tim-
ing of maize introduction into the area re-
mains an open question.
Nutritional Analysis
Because there is little archaeological evidence
supporting an assumption that the pattern of
Late Woodland maize utilization west of New
England applies to coastal New England and
Long Island, New York, alternative means of
recovering prehistoric subsistence strategies
are necessary. One approach involves the use
of analogy with extant human foragers, as-
sessing their diet and nutritional status and
also the strategies they use for obtaining and
storing food items. A recent estimate for the
diet of hunter-gatherers gives about 3,000
calories per day, of which 35 percent of the
total weight derives from animal products
and 65 percent is from vegetable products
(Eaton and Konner 1985). Such a diet pro-
vides in percentages of the total energy,4 251
g of protein (34 percent), 71 g of fat (43 per-
cent), and 333 g of carbohydrates (23 per-
354 [Vol. 60, No. 2, 1995]
REPORTS
cent). It has more protein, less carbohydrates,
and about the same amount of fat as the av-
erage diet in the United States in 1978 (USDA
1978:56).
The percentage of protein remains about
the same for several other hunter-gatherer
diets in different environments. For example,
in the premodern Eskimo diet (2,500 calories
per day) consisting of almost 100 percent an-
imal products (Draper 1977), the sources of
the total daily energy are 200 g of protein (32
percent), 185 g of fat (66 percent), and 10 g
of carbohydrates (2 percent). The diets of the
Dogrib Indians of the Northwest Territories
in Canada in 1985 (Szathmary et al. 1987)
and salmon fishers on the Northwest Coast
(Chisholm et al. 1982) consist of about the
same percentage of protein. High fat and/or
carbohydrate intakes characteristic of north-
ern and temperate coastal hunter-gatherers
appear to offset the negative consequences of
ingesting too much protein (Speth and Spiel-
mann 1983; Eaton and Konner 1985; Noli
and Avery 1988).
Based on theoretical expectations (Perl-
man 1980), archaeological data, and histor-
ical sources, a proposed diet of the Late
Woodland period for coastal New England
in general, and Nantucket specifically, might
have consisted primarily of deer, fish, marine
mammals, waterfowl, shellfish, crustaceans,
and native plants, with possibly some maize.
The early historic diet probably consisted of
the same species, with a possible lack of deer
due to overhunting, and with an increased
reliance on the cultigens maize, beans, squash,
and Jerusalem artichokes. Archaeological ev-
idence has not been found for Jerusalem ar-
tichokes (Helianthus tuberosus), prickly pear
cactus (Opuntia humifusa), and edible sea-
weeds such as irish moss (Chondrus crispus),
dulse, and kelp (Fernald 1970).
Fish, marine mammals, shellfish, crusta-
ceans, and deer could have provided protein.
Marine foods such as seals and whales, fish
(eels, spiny dogfish, herring, scup), and wa-
terfowl are very high in fat and oils. Quahogs
and oysters are relatively high in carbohy-
drates on a seasonal basis (Noli and Avery
Table 1. Predicted Diet from Historical,
Archaeological, and Nutritional Records.
Fish
striped bass * * alewives * A* shad *,
herring OA* eels OA* bluefish A smelt 0
sturgeon 0 white perch * scup * m
Mammals
deer ** seal *A+ whale OA
Shellfish
oysters *? quahogs ** softshell clams -
m* scallops * surf clams whelk
Crustaceans
lobsters 0* *
Birds
ducks OA+ geese OA+ pigeons *+ swans 0
Nuts and seeds
hazelnut A* beechnut A* hickory nuts OA
*, oak acorns beach peas
Roots
cattail sassafras ground nuts
Berries
blueberries currants strawberries 0
bunchberries cherries beach plums grapes
elderberries
Note: Based on (0) abundance and use (Noblick 1977;
see text for historic and archaeological sources), (m) re-
port of preservation, (A) high energy (Watt and Merrill
1975), (*) high returns/effort (Perlman 1980).
1988). Certain wild plant foods available to
the inhabitants of coastal New England are
quite high in fat and protein as well as in
carbohydrates (e.g., hickory, beech, and hazel
nuts, and pokeberry). Lobsters, striped bass,
white perch, clams, and nuts were preserved
by sun drying and smoking, thereby extend-
ing the period of time when they would have
been available (Table 1).
Late Woodland Human Remains
Samples of bone from six Late Woodland
humans were used for this study. In 1623 in
Plymouth, Indians who were not sachems
were buried in or near their wigwams, which
were then abandoned (Winslow 1841:363
[1624]). Lacking associated grave goods, few
of the isolated human burials in southeastern
Massachusetts have been radiocarbon dated
or attributed unambiguously to the Late
Woodland (Little 1985b; but see Mc-
Manamon et al. 1986). Conventional radio-
355
AMERICAN
ANTIQUITY
carbon ages were obtained on bone collagen
from six prehistoric Nantucket burials (Fig-
ure 1): 610?90 B.P. (Beta-21916-AMS),
610?80 B.P. (GX-14301-G), 650? 105 B.P.
(GX-15353-G), 820? 105 B.P. (GX-15351-
G), 940?105 B.P. (Beta-18835-AMS), and
970? 165 B.P. (GX-15352). Calibrated with
atmospheric tables, all the dates would range
(with 68 percent probability) between cal A.D.
902-1405, and with a correction for 100 per-
cent of marine carbon in the protein diet, the
range would be cal A.D. 1193-1690. In either
case, these burials took place during the Late
Woodland period, the interval in New En-
gland from shortly before the introduction of
maize until European settlement.
The Use of Stable Isotopes as Diet Indicators
Animal tissues have 613C and b15N values that
reflect those of their diets (DeNiro and Ep-
stein 1978, 1981). In North America the most
common archaeological application of this
phenomenon has been to use b~3C
values in
human bone collagen to determine the extent
of dependence on the introduced cultigen,
maize (a tropical C4 plant with a high b33C
value), relative to most native plants (with
C3 pathways and low b13C
values). However,
as mentioned previously, in an area like
coastal New England, humans eating a diet
containing a mixture of foods in C3 and C4
food webs could have b13C values in bone
collagen that are equivalent to those of hu-
mans who obtained most of their food from
marine sources.
Because b5 N values are higher in marine
vertebrates in a plankton-based food chain
than in terrestrial animals (Schoeninger and
DeNiro 1984), they can serve to identify hu-
man diets that include plankton-based ma-
rine foods. Recent studies have explored the
use of both 6b5N
and b13C values to estimate
the marine, terrestrial C3, and maize com-
ponents of human diets on the East Coast of
North America (e.g., Bourque and Kreuger
1994; Keegan and DeNiro 1988; Schoeninger
et al. 1990; Tuross et al. 1994; Yesner 1988).
Some important qualifications should be
mentioned. Certain marine plantlike organ-
isms and plants, i.e. blue-green algae, corals,
and Spartina alterniflora, a salt marsh cordg-
rass, obtain their nitrogen by fixation directly
from the atmosphere (Capone and Carpenter
1982). Since atmospheric 615N = 0 per mil,
consumers (including humans) in food chains
based on such plants have lower 355N values
than those in food chains based on plankton,
for which 615N = 8.6 per mil (Peterson et al.
1985; Schoeninger et al. 1983). From the sim-
ilarity of the 613C and 6b5N values of reef fish
and maize, Keegan et al. (1988) show the
resulting indeterminancy between these foods
in the diet of prehistoric humans from the
Bahamas. The prickly pear, Opuntia (a CAM
plant native to Nantucket), uses a C4
pathway
and has a high 613C value (O'Leary 1988),
which makes a distinction between maize and
this cactus difficult. Carbon and nitrogen iso-
tope studies at coasts with extensive salt
marsh and seagrass meadows, in which plants
have b13C values similar to those of maize
(Fry and Sherr 1984), may show a lack of
distinction between maize and inshore ma-
rine fauna. For dietary analysis the local food
chains must be a part of the study.
Although it is accepted that dietary 613C
and '5
N values are passed on to consumers,
the differences or offsets (As) between the 6b3C
and b'5N values of the human diet and those
of human bone collagen have not been di-
rectly measured and remain a matter of much
debate. Even so, several investigators have
used b13C and b'5N measurements with as-
sumed end values and bone/diet offsets to
deduce the relative amounts of C3 and C4
foods in nonmarine diets (Schwarcz et al.
1985; Spielmann et al. 1990), or the relative
amounts of plankton-based marine and ter-
restrial foods in nonmaize coastal diets (Chis-
holm et al. 1982; Walker and DeNiro 1986),
with results that suggest the considerable val-
ue of the approach.
Isotope Measurements
Because of the complicating factors for both
carbon and nitrogen isotope values in esti-
mating coastal human diets, we undertook
the collection and the analysis of a large num-
356 [Vol. 60, No. 2, 1995]
REPORTS
ber of marine organisms from near shore and
open ocean in addition to terrestrial plants
and animals to determine the distribution of
stable isotope values in potential food items.
Almost all the items were collected at Nan-
tucket Island (Table 3:Note a). Six human
bone samples and over 40 samples of plant,
fish flesh, bird flesh, terrestrial mammal flesh,
and flesh from one harbor seal were prepared
for stable isotope analysis.5 The results are
presented in Tables 2 and 3.
Similar tissues of organisms show frac-
tionation or increases in 613C and 615N with
increasing trophic level in the same environ-
ment (Schoeninger and DeNiro 1984). The
Nantucket foods were grouped into catego-
ries of similar environments and trophic lev-
els (Figure 2), which clarifies the increase in
6'5N values with trophic level for plants, her-
bivores and carnivores. The wide range of
613Cs and 6l5Ns of the four
plants
at the bases
of the food webs provide a great variety of
bs higher up in food webs. Spartina alterni-
flora of the salt marsh (a C4
plant) and upland
C3 plants exhibit the low 615N values of ni-
trogen-fixing plants in nutrient-poor salt
marsh and upland sandy soils (Peterson et al.
1985). The marine plants, eelgrass (high 613C
value), and phytoplankton (613C = -21.3 per
mil [Peterson et al. 1985]), show middle rang-
es of 6'5N values.
Graphical Analysis of Results
As a result of the rich mixture of foods with
a variety of 6s at Nantucket, we now have far
too many items for analysis. As a first ap-
proximation, we aggregate foods with similar
bs, which yields 6 estimates for seven groups
(Table 4; Figure 3): (1) upland C3 plants; (2)
deer and freshwater fish; (3) oceanic (plank-
ton-based) carnivores; (4) nearshore marine
omnivores; (5) nearshore herbivores; (6)
nearshore carnivores; (7) C4 plants. Deer and
freshwater fish can be grouped together be-
cause they have similar 613C and 6b5N values,
while C3 terrestrial plants are very low in 15N
and require a separate group. We consider
freshwater plants as possible members of
group 2 and prickly pear (C4-pathway) as a
Table 2. 613C and 6'N Values for
Nantucket Human Bone Collagen.
Site Sample b13C l15N
Number Number (%oo,
PDB) (0o, AIR)
19NT-154 MS3199 -10.3 15.3
19NT-108 MS3197 -10.4 15.5
19NT-130 MS3738 -10.6 15.1
19NT-134-3 MS3736 -10.4 16.7
19NT-153 MS3198 -11.0 14.1
19NT-134-1 MS3735 -9.6 15.3
Average -10.4 15.3
Standard deviation ?.2 +.3
possible member of Group 7, groups within
which prickly pear and freshwater plants are
indistinguishable isotopically.
Analyses of modem maize have produced
6b5N values that range from 2.1 to 8.6 per
mil (DeNiro and Hastorf 1985). The varia-
tion reflects the source of nitrogen available
to the plants, with the lower values resulting
from the use of synthetic fertilizers with val-
ues close to 0 per mil. We chose to use a
value of 8.7 per mil estimated from the iso-
tope data for prehistoric maize-eating hu-
mans in southern Ontario (Schwarcz et al.
1985), rather than the 2.2 per mil of Table
3. However, a study of6'5N values for coastal
maize fertilized by fish or beans (Ceci 1975)
is needed.
The polygon in Figure 3 is formed by lines
between the estimated food group end points,
each of which represents the 6 value of a diet
of 100 percent of that food group. These lines
encompass the entire set of diets that can be
formed from the food groups 1-7. Consumer
diet 613C and 6'5N values are negatively offset
from the values of consumer bone collagen.
The range of modem human diet to archae-
ological human bone offsets, A/s, derived
from the literature is: A613C
= 5.1 to 6.5 per
mil; A6'5N = 2.4 to 4 per mil (Chisholm et
al. 1982; DeNiro and Epstein 1981; Keegan
and DeNiro 1988:329; Schwarcz et al. 1985;
van der Merwe and Vogel 1978). Including
all six human diets within the polygon re-
quires the diet to bone offsets to be at least
Ab'3C = 6 per mil and A615N = 3 per mil (as
shown in Figure 3). That our AL13C is larger
than the usual 5 per mil results from the com-
357
AMERICAN
ANTIQUITY
Table 3. 613C and b5N Values of Nantucket Flora and Fauna.
bl3C b65N
Species (common name) Diet %o %O
C3 Terrestrial and Freshwater
Odocoileus virginianus (white-tail deer, muscle)
Salicornia virginica (samphire)
Atriplex patula var. hastata (orach)
Cakile edentula (sea rocket)
Lathyrus japonicus (beach pea)
Apios americana (groundnut)a
Phaseolus vulgarus (common bean)a
Amelanchier sp. (shadberry)
Rubus hispidus (blackberry)
Sambucus canadensis (elderberry)
Vaccinium angustifolium (blueberry)
Corylus americana (hazelnut)
Helianthus tuberosus (Jerusalem artichoke)
Morone americana (white perch)
Typha angustifolia (cattail)
Acorus calamus (sweet flag)
Oceanic carnivores
Pomatomus saltatrix (bluefish)
Roccus saxatilis (striped bass)
Phoca vitulina (harbor seal, muscle)a
Hippoglossus hippoglossus (halibut)
Nearshore fauna
Mytilus edulis (blue mussel)
Crassostrea virginica (oyster)
Mercenaria mercenaria (quahog)
Mya arenaria (clam)
Geukensia demissa (ribbed mussel)
Argopecten irradians (bay scallop)
Branta canadensis (Canada goose)
Pseudopleuronectes americanus
(winter flounder)
Cancer borealis (Jonah crab)
Stenotomus chrysops (scup)
Busycon canaliculatum (channelled whelk)
Lunata heros (moon shell snail)
Homarus americanus (lobster)
Tautogolabrus adspersus (cunner)
Anguilla rostrata (eel)
C4 (& CAM) Terrestrial flora
Zea mays (modern 8-rowed northern flint corn)a
Opuntia humifusa (prickly pear) CAM
Miscellaneous
Odocoileus virginianus (deer bone)
0. virginianus (bone >400 years old)
Ammodytes americanus (sand eel)b
C4 Spartina alterniflora (saltmarsh cordgrass)b
C4-like Zostera marina var. stenophylla (eelgrass)b
0. virginianus (deer fat)
P. vitulina (seal fat)a
Crassostrea virginica (internal organs)
Mercenaria mercenaria (internal organs)
Mya arenaria (internal organs)
Average C3 upland terrestrial plantsc
C3 plants
freshwater fish, plants
fish
fish
fish, invertebrates
fish, invertebrates
-24.1
-24.8
-27.2
-27.8
-23.8
-28.1
-27.9
-26.7
-25.0
-26.5
-25.8
-26.9
-25.9
-23.9
-28.1
-26.3
-18.7
-18.3
-17.2
-17.8
mixed C3, C4 marine, terrestrial diets
detritus, plankton -21.0
detritus, plankton - 18.6
detritus, plankton -18.5
detritus, plankton -16.1
detritus, plankton -15.1
detritus, plankton -14.0
Spartina, etc. - 13.6
invertebrates, etc.
invertebrates, etc.
invertebrates, etc.
invertebrates, etc.
invertebrates, etc.
invertebrates, etc.
omnivores, eelgrass
invertebrates, etc.
-17.5
-17.0
-16.6
-16.1
-15.9
-13.7
-13.5
-13.4
-10.7 2.2
-14.1 -.7
C3 plants
C3 plants
plankton, invertebrates, etc.
detritus, plankton
detritus, plankton
detritus, plankton
-21.6
-20.2
-19.3
-12.0
-6.8
-30.3
-23.9
-20.5
-22.2
-19.1
-26.6
Note: Analyses made on moder and edible portions unless noted. Species names and diets from Bigelow and
Schroeder (1953), Fernald (1950).
4.2
4.7
2.2
1.5
-.4
-.8
.2
.6
2.2
3.3
2.3
6.7
10.3
8.1
5.8
16.2
16.6
14.7
14.5
7.2
2.8
5.3
7.2
4.1
6.3
6.8
10.7
10.6
12.7
11.1
9.3
11.8
10.6
9.8
3.8
2.0
11.7
1.0
5.4
6.8
6.3
1.2
358 [Vol. 60, No. 2, 1995]
REPORTS
parison of prehistoric human bone with mod-
em foods. A fossil fuel effect on atmospheric
613C has changed the 613C of maize grown
after A.D. 1800 by about - 1.5 per mil (Tiesz-
en and Fagre 1993), and affects all of today's
foods (see deer bone in Table 3; fish bone in
Keegan and DeNiro 1988; and Chisholm et
al. 1982). Reasonable estimates are A6l3C =
6 per mil and A6b5N = 3 per mil.
From our estimations in Figure 3, we may
draw some fairly strong conclusions:
(1) To fulfill the high 6b5N requirements of
the Nantucket protein diet, the largest part
of the protein diet must consist of oceanic
fish and sea mammals in planktonic food
chains.
(2) To fulfill the high 613C requirements, a
large part of the diet must be eelgrass carni-
vores, such as lobsters and eels, or maize,
with a small amount of nearshore shellfish,
fish, crabs, and waterfowl.
(3) Because of their low 613C values, fresh-
water fish and deer contribute only a small
amount to the diet.
Analysis of Results Using a Linear Model
A number of authors have used a linear mod-
el for the analysis of diets with two variables6
(Chisholm et al. 1982; Schwarcz et al. 1985;
van der Merwe and Vogel 1978; Walker and
DeNiro al. 1986; Yesner 1988). To study the
Nantucket Late Woodland diet, however, we
propose a model with three simultaneous lin-
ear equations describing the mixing of bl3C
and 615N values of food groups into the b13C
and 6b'N values of the diet for the graphical
model shown in Figure 3. The first equation
simply says the food group proportions must
add to one. The second describes the mixing
of foods of different 613C
values to create the
overall diet value of 613C. The third does the
same for 615N. Let Di be the proportion of
kilocalories of a given food group, i, in the
diet. If 6i13C and bi15N are the isotope values
of those groups, the three equations have the
following form:
Equation 1. Total diet = 2 Di = 1, where
0 < Di < 1
Equation 2. z2 bi 13C Di = die 13C
Equation 3. 2 bi 15N Di = 6diet 15N
Studies have shown that we must make an
adjustment for the overrepresentation of diet
protein in bone collagen of consumers of high
protein diets (Ambrose and Norr 1993; see
also Chisholm et al. 1982). The 615N values
of consumer bone reflect only protein, since
carbohydrates and fat contain no nitrogen.
Although the offsets between consumer bone
and dietary protein are not yet well known,
protein weighting of food components (Spiel-
mann et al. 1990) can approximate the con-
tribution of proteins to the total isotope value
of the diet. We multiply each Di in Equation
2 and 3 by pi/P, where Pi is the amount of
protein (g/100 kcal) in food group i and P is
the total protein, P = 2 pi Di (g/100 kcal).
Equations 1, 2, and 3, weighted for protein,
become:
Equation 1: Z Di = 1, where 0 < Di < 1
Equation 2p: 2 (pi/P) bi 13C Di = bdiet 13C,
or 2 pi (6diet 13C - i, 13C) Di = 0, and
Equation 3p: 2 (Pi/P) bi '5N Di = 6diet 15N,
or 2 Pi (6diet 15N - 6i 15N) Di = 0.
In estimating the sizes of the nutritional
components (Table 5), extra fat was added to
group 3, oceanic carnivores, to compensate
for the modem viewpoint of the nutritional
tables. This reduced the size of group 3's pro-
tein component. Marine mammals of tem-
perate and cold climates carry layers of blub-
ber for insulation, as well as a great deal of
fat in their flesh, and, as mentioned earlier,
a number of fish species from coastal New
England are very oily. All these sources would
have provided the fat that nutritional studies
indicate is necessary for humans to tolerate
a diet high in meat.
For three equations involving three vari-
a Samples from Massachusetts.
b Not known dietary items.
c Does not include Jerusalem artichoke or samphire.
359
AMERICAN
ANTIQUITY
OCEANIC -
CARNIVORES
a i
key:
o terrestrial flora (C3)
* ** fauna
& freshwater flora
* ., fauna
, plant bases of food webs
0 sand eel
FRESHWATER
C3
0
TERRESTRIAL
C3
o
CO
o typnd
0
-25
sand eel 4> ? NEARSHORE
0 CARNIVORES
0 C
+
plankton
+ + +
+ NEARSHORE
OMNIVORES
+ &HERBIVORES
P maiTe
C4 &CAM slr a
PLANTS sltrshpcordgros
*prickly pear
-20 -15
ables, each the proportion of one food group,
we can usually solve exactly for the three di-
etary proportions. For more than three food
groups an indeterminancy will usually result.
However, linear programming methods make
it possible to set limits on the proportions of
each food group. For each group we deter-
mine the largest range of values that are con-
sistent with all the bone and food group iso-
tope data.
Applying these methods here (John D. C.
Little, personal communication 1993), we find
the solution to Equations 1, 2p, and 3p, with
seven variables (Di); coefficients from Tables
4 and 5, and the proposed values, bdiet 13C =
-16.4 per mil and diet 1
5N = 12.3 per mil.
The linear programming solutions indicate
allowable Di (in percentages of total calories):
D1: 0-15; D2: 0-7; D3: 39-66; D4: 0-15;
-10
Figure 2. Values of 513C
and
eeigress 15N for dietary items, the bases
of food webs and other constitu-
ents of Nantucket ecosystems
(Table 3). Plankton and its 5 val-
ues are from Falmouth, Massa-
Ss chusetts (Peterson et al. 1985).
Sand eel plays an important role
6 'e o/ in the plankton/oceanic carnivore
5 food chain.
D5: 0-34; D6: 0-40; D7: 0-57. The sizable
ranges result from the large number of food
groups that can create diets constrained by
the three equations. Even so, this solution
establishes a dietary minimum of 40 percent
oceanic seafood (group 3), and a low amount
of C3 flora and fauna (groups 1 and 2), which
we would have expected from the graphic
analysis of Figure 3.
We can narrow the ranges by reducing the
number of groups being considered. Drop-
ping group 1 (its low 613C
indicates it could
not have been a large factor in the diet), and
merging groups 4 and 5 (which aggregates a
great many different foods of the harbor)
leaves five groups. The large triangle in Figure
4 now displays the 6 values for three diet
groups, 2, 3 and X, with three constituents
for X: 4 and 5, 6, and 7. With protein weight-
Table 4. Mean and Standard Deviation of the Mean for 613C and b'N Values of Food Groups.
Group Foods in Diet (no. in group) 613 C /o il5N %OX
1 Upland C3 plants (10) -26.6 ? .4 1.2 ? .5
2 Deer & freshwater fish (2) -24.0 ? .1 7.3 ? 2.5
3 Oceanic carnivores (4, seal, bluefish, etc.) -18.0 ? .3 15.5 ? .5
4 Nearshore marine fauna, moderate 613C (11, bi- -16.9 ? .6 7.9 ? 1.0
valves, fish, crabs, gastropods)
5 Nearshore herbivore, high 6'3C (1, Canada goose) -13.6 6.8
6 Nearshore carnivores, high il3C (3, lobster, cunner, -13.5 ? .1 10.7 ? .6
eel)
7 C4 plant (1, maize) -10.7 8.7a
a Estimated (see text).
6 'N%o
15-
10
5
0
360 [Vol. 60, No. 2, 1995]
REPORTS
Figure 3. Estimated 513C and
515N values for seven groups of
Nantucket dietary items (Table
4) and Late Woodland Nantuck-
et human diets (Table 2; see text
for conversion from bone colla-
gen to diet). The polygon drawn
between estimated end points en-
closes all possible consumer diets
that may be formed from
mixtures of these dietary groups.
ing, the three diet equations for i = 2, 3, X
become
D2 + D3 + Dx= 1
2 pi Di (Odiet 13C - (3i
'3C) = 0
2 Pi Di (3diet 15N - i 15N) = 0
The 6x values must lie within the small tri-
angle and we estimate /5xl3C = -13.5 per mil
and /X15 N = 9.8 per mil. Three equations
with three variables may be solved algebra-
ically and exactly (see Note 6 for the basic
method). If we solve them for the diet of each
of the six Nantucket humans, we get six sets
of food group proportions. The ranges of the
percentages of each of these food groups that
will produce the 6 values of our six Nantucket
diets are: D2 = 1-12 percent, D3 = 71-38
percent, Dx = 29-50 percent.
Discussion of Results of Dietary Analyses
The numerical results from the linear model
support the graphical results of Figure 3, al-
though protein weighting introduces some
nonlinearity not reflected in the 6 graphs. The
model, like the graph, requires a small amount
of terrestrial C3 fauna, a large amount of oce-
anic seafood (group 3), and a substantial
amount of group X, which includes maize,
eelgrass carnivores, and other nearshore om-
nivores and herbivores.
Since maize, lobsters, and eels (and any-
thing else in the eelgrass food web) have like
6 values, we cannot, with only 6l3C and (15N
measurements, determine the relative im-
portance of maize compared with the fauna
of seagrass meadows in the diet. A case might
be made that the Late Woodland diets at
Nantucket with the highest /513C values in-
cluded some maize, and that lobsters and eels
were staples. That corn was used from the
twelfth century on but not intensively culti-
Table 5. Estimated Nutritional
Components of Nantucket Food Groups.
Ci,
Car- Pi,
F,, bohy- Pro-
Group Foods in Diet Fat drates tein
1 Upland C3 plants (including
nuts) 4.5 9.8 3.9
2 Deer and freshwater fish 3.3 0 16.6
3 Oceanic carnivores (includ-
ing liver & blubber) 6.8 0 8.0
4 Nearshore marine fauna 1.5 3.6 16.5
5 Nearshore herbivores 6.8 0 9.1
6 Nearshore carnivores high
bl3C 5.0 0.5 12.7
7 C4 plant 1.0 22.0 3.1
Note: All measurements in g/ 100 kcal. Values estimated
from representative food items in Watts and Merrill
(1975).
361
AMERICAN
ANTIQUITY
Marine
6'5N%o
15-
10-
Nearshore
-20 -15
Eels
)Maize Figure 4. Estimated stable iso-
tope values for dietary groups:
(2) terrestrial C3 fauna, (3) oce-
anic fauna, and (X) consisting of
nearshore fauna (4, 5, 6), and/or
(7) maize, showing the indetermi-
-1o 613C
%o nancy of group X.
vated on the coast is currently validated by
archaeological recoveries of maize kernels,
and soil and climate studies (Demeritt 1991).
The case for lobsters and eels being used
as staples by Nantucketers also appears to be
strong, although we lack archaeological evi-
dence here. Lobsters are often mentioned in
the literature of the early explorers and set-
tlers of New England. "Lobsters are there in-
finite in store in all parts of the land, and very
excellent" (Brereton 1602:226). "Our bay af-
fording many lobsters, [the Indians] resort
every spring-tide thither" (Mourt 1986:62
[1622]). "The Indians get many of them [lob-
sters] every day for to baite their hooks with-
all, and to eate when they can get no Basse"
(Wood 1865:37, 100 [1635]). "In the summer
these Indian women when Lobsters be in their
plenty and prime, they drie them to keepe
for Winter, erecting scaffolds in the hot sun-
shine, making fires likewise underneath them,
by whose smoake the flies are expelled, till
the substance remain hard and drie" (Wood
1865:101 [1635]). Old timers at Nantucket
told J. Clinton Andrews (personal commu-
nication 1991) that before the 100-year-old
rock jetties were built, lobsters lived in holes
in eelgrass meadows on sandy shoals. Eels,
too, were a reliable delicacy. "[I]n September
we can take a hogshead of eels in a night with
small labor, and can dig them out of their
beds all the winter" (Mourt 1986:84 [1622]).
"[March 24] Squanto went at noon to fish for
eels; at night he came home with as many as
he could well lift in one hand ... .They were
fat and sweet" (Mourt 1986:59 [1622]). Note
that lobsters are available inshore chiefly in
summer, and eels were especially available
between September and March; these two do
not make such an odd couple after all-they
are seasonally complementary.
Have we found all of the high 63'C foods?
The inhabitants of Baja California regularly
harvested and ate eelgrass seeds (Felger and
Moser 1973), but at Nantucket the plant does
not produce enough seeds to be a significant
food source (J. Clinton Andrews, personal
communication 1989). From eelgrass found
in middens in New Brunswick, archaeologists
have inferred its use in cooking (as in New
England clambakes; Rau 1884:222), but we
have no evidence for human dietary use of
eelgrass, seaweeds, or Spartina roots in Mas-
sachusetts. Similarly, prickly pear, "Indian
fig" (Fernald 1970), was eaten in the southern
United States, but at its northeastern limits
we have no ethnographic or archaeological
evidence for its consumption.
The Native American coastal diet fre-
quently included dogs, and many prehistoric
coastal middens or ceremonial graves in
Massachusetts contain their remains (i.e.,
Ritchie 1969). For bone collagen of a Boston
Harbor dog with a conventional radiocarbon
age of 1710?70 B.P., Nelson (1989) reported
b'3C(collagen) = - 13.1 per mil, which sug-
Terrestrial
-25
362 [Vol. 60, No. 2, 1995]
REPORTS
-20
O
NWCoast marine hunters
O California fishermen
OMaine fishermen
Q Nantucket
Q
Danish fishermen
O S.Ontario
fishermen S.Ontario
fish + maize
Figure 5. Average 1'3C
and
615N values for bone collagen
showing the difference between
marine (mammal) hunters, ma-
rine fishermen, freshwater fisher-
men, and C3 and C4 horticultur-
alists (Schoeninger et al. 1983;
Schwarcz et al. 1985; Walker and
DeNiro 1986; Yesner 1988; Ta-
ble 2).
gests that this dog also ate
amount of seafood.
Staple Isotopes and Health
-10 Danish
farmers (C3)
-20
a substantial
A reasonable diet from the middle of the Di
ranges of the general solution for seven diet
groups, with an emphasis on seafood and
normalized to I Di = 100 percent, would be,
in percentages: DI = 4, D2 = 2, D3 = 55, D4
= 4, D5 = 15, D6 = 16, D7 = 4. With these
estimates, the percentages contributed by
protein, carbohydrates, and fat to a daily diet
of 2,500 kcal (Table 5) would be: by energy,
36 percent protein, 5 percent carbohydrate,
59 percent fat; and by weight, 226 g protein,
34 g carbohydrate, 147 g fat. If a fourth equa-
tion is added to equations 1, 2p, and 3p with
the constraint that the total daily protein be
less than or equal to 250 g, there is a small
(<2 percent) reduction in the allowed upper
limits of protein-rich food groups, D2, D4,
and D6. In terms of protein values, the diet
lies between Draper's (1977) Eskimo diet of
200 g and Eaton and Konner's (1985) world-
wide average hunter-gatherer diet of 251 g.
This diet is high in animal products (93 per-
cent), specifically seafood, but its 226 g of
protein represents only 36 percent of daily
energy requirements. As noted by Draper
New Mexico
farmers (C4)
-15 -10 -5
(1977), fish and marine mammal meat, oils,
fat, viscera (especially livers), as well as parts
of bones, provide the necessary energy, vi-
tamins, and minerals for the diet of coastal
hunter-gatherers. At the same time, the poly-
unsaturated fatty acids of marine oils con-
tribute to the low cholesterol levels charac-
teristic of Eskimos on a native diet (Draper
1977). In this context, 56 Late Woodland re-
mains from an ossuary at Cape Cod, and one
from Nantucket, examined in detail, dem-
onstrate general good health as well as a low
percentage of teeth with caries (Little 1985b;
McManamon et al. 1986).
Accuracy
Testing the sensitivity of the model by vary-
ing the parameters showed that changing
A613C by eight percent, from 6.0 to 6.6 per
mil, would only allow an additional two per-
cent of D2. There are, however, a number of
sources of uncertainty that need to be rec-
ognized. First, we have a small number of
human samples for good statistics, but their
range in bs is small, too, considering the 500-
year span. DeNiro and Schoeninger (1983)
have shown that for 15 animals on a mo-
notonous diet, collagen b13C and '5
N values
for one individual will lie within 1 per mil of
6 15N %
-15
363
AMERICAN
ANTIQUITY
the mean. For 20 southern Ontario humans
with maize-free diets, the greatest 6 difference
from the mean is 2.5 per mil, and for 20 with
maize diets, 2.8 per mil (Schwarcz et al. 1985).
Similar greatest difference data for (13'C and
b5SN
values from 12 Late Woodland Maine
fishermen is 1.3 and 1.0 per mil (Yesner 1988).
The Nantucket sample of six had a maximum
difference from the mean 13'C and bl5N val-
ues of .8 and 1.4 per mil, a relatively narrow
range of values like those of the Maine sam-
ple. All of these diets include many choices
of foods with different 6 values, the avail-
ability of which may have varied with time.
Confining the data to Nantucket samples,
where possible, has greatly increased the rel-
evance of the measurements.
Our sometime crude estimates have led to
a fairly robust model, which can be tested
and refined on a variety of diets. Additional
linear equations that quantitatively relate dif-
ferent nutritional components of diet with
properties of consumers could reduce inde-
terminancy with greater precision (DeNiro
and Epstein 1981; Peterson et al. 1985).
Conclusions
Our results, taken with other lines of evidence
such as archaeological finds and historic re-
cords, can most reasonably be interpreted to
support current suggestions of an early arrival
of maize along the coast, and a mixed record
of adoption of intensive maize use with time
throughout the Northeast (Demerrit 1991).
From stable isotope measurements, we con-
clude that the diet of the inhabitants of Nan-
tucket Island during the Late Woodland pe-
riod included substantial amounts of oceanic
and nearshore fauna, a limited amount of C3
plants and their consumers, and an unknown
amount of C4 plants like maize. Because lob-
sters, eels, and maize have similar 613 C and
(35 N values, we cannot determine the relative
importance of maize compared with the fau-
na of seagrass meadows and salt marshes us-
ing only these isotope measurements.
In terms of nutrition, either maize or near-
shore carnivores or some of each are accept-
able since the diet as it stands is well within
the parameters described for present day
hunter-gatherers.
When the archaeological record is consid-
ered, however, a diet containing a significant
amount of maize becomes less likely. The
modest recoveries of maize kernels and cobs,
and lack of evidence for land clearing in the
Late Woodland, argue that any maize that
was eaten was probably grown in small gar-
dens near houses with southern exposures
(Little 1985a, 1988) or traded in from else-
where (Luedtke 1980b). While the latter pos-
sibility cannot be completely dismissed, de-
pendence on trade is a recognized unstable
adaptation.
The 6(13C and (5 N values of other marine-
adapted humans-such as Mesolithic Danish
and Moshier Island, Maine, fishermen (Fig-
ure 5)-are similar to those measured for Late
Woodland Nantucket inhabitants. The diets
of Maine fishermen (970+145 B.P., human
bone, GX-7061 G, conventional, uncalibrat-
ed [Yesner 1988]) and California fishermen
(about A.D. 1150-1800 [Walker and DeNiro
1986]) reflect a greater use of plankton-based
marine food, and a lesser use of foods from
eelgrass meadows and salt marshes than the
diet at the island of Nantucket, an unsur-
prising finding given the paucity of the latter
habitats in those areas. Maize might have
been a factor in diets of some inhabitants of
North Haven Island, Maine, with bl3C
values
almost as high as those at Nantucket, but with
inconclusive dates (Bourque and Kreuger
1994).
Our results suggest a greater importance of
marine foods in the diets at Nantucket than
the dominance of mammal bone in the ar-
chaeological shell middens indicates. The rel-
ative lack of fish bones in shell middens, not
uncommon, needs further exploration.
The analysis of the stable carbon and ni-
trogen isotopes at Nantucket provides a firm
base for future studies of human diets on the
East Coast. With these data as a starting point,
we see the next challenge as the discovery of
some means of distinguishing between maize
and marine resources of salt marsh and sea-
grass meadows in coastal diets.
364 [Vol. 60, No. 2, 1995]
REPORTS
Acknowledgments. Support for this project was pro-
vided by the Massachusetts Historical Commission, the
Massachusetts Commission on Indian Affairs, the Nan-
tucket Historical Association, the Nantucket Maria
Mitchell Association, Harvard University, and National
Science Foundation grants BNS8607659 and
BNS9004063 to M. J. Schoeninger. Christian C. Med-
aglia, John Blitz, Katherine Moore (now of Bentley Col-
lege), Mark Schurr (now of Notre Dame University),
Isabel Treichel (University of Wisconsin), and Renee
Robinson (Harvard University) helped in sample prep-
aration and mass spectrometry, and Urszula Iwaniec
(University of Wisconsin) with the Kjeldahls. J. Clinton
Andrews, Shirley Blancke, Robert Hasenstab, Tonya B.
Largy, Timothy Lepore, Eleanor Lucas, Greg Early (New
England Aquarium), and Wesley N. Tiffney (University
of Massachusetts Field Station) contributed samples. John
D. C. Little (Massachusetts Institute of Technology) pro-
vided the linear programming. Barbara E. Luedtke (Uni-
versity of Massachusetts) has commented on the regional
maize discussion, and two reviewers have made very
constructive criticisms. Elizabeth A. Dice translated the
abstract into Spanish.
References Cited
Ambrose, S., and L. Norr
1993 Experimental Evidence for the Relationship of
the Carbon Isotope Ratios of Whole Diet and Di-
etary Protein to Those of Bone Collagen and Car-
bonate. In Molecular Archaeology of Prehistoric Hu-
man Bone, edited by J. Lambert and G. Grupe.
Springer, Berlin.
Bendremer, J. C., E. A. Kellogg and T. B. Largy
1991 A Grass-Lined Maize Storage Pit and Early
Maize Horticulture in Central Connecticut. North
American Archaeologist 12:325-349.
Bernstein, D. J.
1992 Prehistoric Use of Plant Foods in the Narra-
gansett Bay Region. Man in the Northeast 44:1-13.
Bigelow, H. B., and W. C. Schroeder
1953 Fishes of the Gulf of Maine. 1st rev. Fishery
Bulletin of the Fish and Wildlife Service Vol. 53,
No. 74. U.S. Government Printing Office, Wash-
ington D.C.
Borstel, C.
1984 Prehistoric Site Chronology: A Preliminary
Report. In Chapters in the Archaeology of Cape Cod,
I, vol. 1, edited by F. P. McManamon, pp. 231-
288. Cultural Resources Management Study No. 8,
National Park Service, Boston.
Bourque, B. J., and H. W. Kreuger
1994 Dietary Reconstruction from Human Bone
Isotopes for Five Coastal New England Populations.
In Paleonutrition: The Diet and Health of Prehistoric
Americans, edited by K. D. Sobolik, pp. 195-209.
Occasional Paper No. 22, Center for Archaeological
Investigations, Southern Illinois University, Car-
bondale.
Brereton, J.
1602 A Brief and True Relation .... George Bishop,
London.
Capone, D. G., and E. J. Carpenter
1982 Nitrogen Fixation in the Marine Environment.
Science 217:1140-1142.
Ceci, L.
1975 Fish Fertilizer: A Native North American
Practice? Science 118:26-30.
1990 Radiocarbon Dating "Village" Sites in Coastal
New York: Settlement Pattern Change in the Middle
to Late Woodland. Man in the Northeast 39:1-28.
Champlain, S. de
1880-1882 Voyages, vol. 2. Prince Society, Boston.
Chisholm, B. S., D. E. Nelson, and H. P. Schwarcz
1982 Stable Carbon Isotope Ratios as a Measure of
Marine Versus Terrestrial Protein in Ancient Diets.
Science 216:1131-1132.
Crevecoeur, H. St. J. de
1971 [1782] Letters from An American Farmer. Ev-
eryman's Library, New York.
Demeritt, D.
1991 Agriculture, Climate, and Cultural Adaptation
in the Prehistoric Northeast. Archaeology of Eastern
North America 19:183-202.
DeNiro, M. J., and S. Epstein
1978 Influence of Diet on the Distribution of Carbon
Isotopes in Animals. Geochimica et Cosmochimica
Acta 42:495-506.
1981 Influence of Diet on the Distribution of Bitrogen
Isotopes in Animals. Geochimica et Cosmochimica
Acta 45:341-51.
DeNiro, M. J., and C. A. Hastorf
1985 Alteration of '5N/'4N and '3C/12C Ratios of
Plant Matter during the Initial Stages of Diagenesis:
Studies Utilizing Archaeological Specimens from
Peru. Geochimica et Cosmochimica Acta 49:97-115.
DeNiro, M. J., and M. J. Schoeninger
1983 Stable Carbon and Nitrogen Isotope ratios of
Bone Collagen: Variation within Individuals, be-
tween Sexes, and within Populations Raised on Mo-
notonous Diets. Journal of Archaeological Science
10:199-203.
Draper, H. H.
1977 The Aboriginal Eskimo Diet in Modern Per-
spective. American Anthropologist 79:309-316.
Dunwiddie, P. W.
1990 Postglacial History of Coastal Islands in South-
eastern New England. National Geographic Re-
search 6(2):178-195.
Eaton, S. B., and M. Konner
1985 Paleolithic Nutrition: A Consideration of Its
Nature and Current Implications. The New England
Journal of Medicine 312:283-289.
Felger, R., and M. B. Moser
1973 Eelgrass (Zostera marina L.) in the Gulf of Cal-
ifornia: Discovery of Its Nutritional Value by the
Seri Indians. Science 181:355-356.
Fernald, M. L.
1970 Gray's Manual of Botany. 8th ed. Van Nos-
trand Reinhold, New York.
Fitzgerald, J.
1984 Floral and Faunal Archaeological Remains. In
Chapters in the Archeology of Cape Cod, I, vol. 2,
edited by F. P. McManamon, pp. 43-81. Cultural
Resources Management Study No. 8. National Park
Service, Boston.
365
AMERICAN
ANTIQUITY
Freeman, Rev. J.
1846 [1815] Notes on Nantucket, August 1st, 1807.
Collections of the Massachusetts Historical Society,
2d ser., 3:19-38.
Fry, B., and E. B. Sherr
1984 b13C Measurements as Indicators of Carbon
Flow in Marine and Freshwater Ecosystems. Con-
tributions in Marine Science 27:13-47.
Gookin, D.
1970 [1674] Historical Collections of the Indians in
New England, edited by J. H. Fiske. Towtaid,
Worcester. Hayden, B., B. Chisholm, and H. P.
Schwarcz
1987 Fishing and Foraging: Marine Resources in the
Upper Paleolithic of France. In The Pleistocene Old
World: Regional Perspectives, edited by 0. Soffer,
pp. 279-291. Interdisciplinary Contributions to Ar-
chaeology, M. Jochim, editor. Plenum Press, New
York City.
Keegan, W. F., and M. J. DeNiro
1988 Stable Carbon- and Nitrogen-Isotope Ratios of
Bone Collagen Used to Study Coral Reef and Ter-
restrial Components of Prehistoric Bahamian Diet.
American Antiquity 53:320-336.
Langlois, K. H., Jr.
1977 Soil Survey of Nantucket County, Massachu-
setts. USDA Soil Conservation Service in cooper-
ation with Massachusetts Agricultural Experiment
Station. U.S. Government Printing Office, Wash-
inton, D.C.
Lavin, L.
1988 Coastal Adaptations in Southern New England
and Southern New York. Archaeology of Eastern
North America 16:101-120.
Little, E. A.
1985a Prevailing Winds and Site Aspects: Testable
Hypotheses about the Seasonality of Prehistoric Shell
Middens at Nantucket, Massachusetts. Man in the
Northeast 29:15-27.
1985b Prehistoric Diet at Nantucket Island, Massa-
chusetts. Nantucket Archaeological Study No. 6.
Nantucket Historical Association, Nantucket, Mas-
sachusetts.
1988 Where Are the Woodland Villages on Cape
Cod and the Islands? Bulletin of the Massachusetts
Archaeological Society 49:72-82.
1993 Radiocarbon Age Calibration at Archaeologi-
cal Sites of Coastal Massachusetts and Vicinity.
Journal of Archaeological Science 20:457-471.
Little, E. A., and J. C. Andrews
1982 Drift Whales at Nantucket: The Kindness of
Moshup. Man in the Northeast 23:17-38.
Luedtke, B. E.
1980a Survey of the University of Massachusetts,
Nantucket Field Station. In Widening Horizons, ed-
ited by C. Hoffman, pp. 95-129. Massachusetts Ar-
chaeological Society, Middleboro.
1980b The Calf Island Site and the Late Prehistoric
Period in Boston Harbor. Man in the Northeast 20:
25-76.
1987 Jasper and the Pennsylvania Connection: Jas-
per at Massachusetts Sites. Bulletin of the Massa-
chusetts Archaeological Society 48:37-47.
McBride, K. A., and R. E. Dewar
1987 Agriculture and Cultural Evolution: Causes and
Effects in the Lower Connecticut River Valley. In
Emergent Horticultural Economies of the Eastern
Woodlands, edited by W. F. Keegan, pp. 305-328.
Occasional Paper No. 7. Center for Archaeological
Investigations, Southern Illinois University, Car-
bondale.
McManamon, F. P., J. W. Bradley, and A. L. Magennis
1986 The Indian Neck Ossuary. In Chapters in the
Archeology of Cape Cod, V, edited by F. P. Mc-
Manamon. Cultural Resources Management Study
No. 17. National Park Service, Boston.
Macy, Z.
1810 [1794] A Short Journal of the First Settling of
the Island of Nantucket (1792). Collections of the
Massachusetts Historical Society 3:155-160.
Medaglia, C. C., E. A. Little, and M. J. Schoeninger
1990 Late Woodland Diet on Nantucket Island: A
Study Using Stable Isotope Ratios. Bulletin of the
Massachusetts Archaeological Society 51: 49-60.
Mourt, G.
1986 [1622] Mourt s Relation, edited by D. B. Heath.
Applewood Books, Cambridge.
Nelson, C. M.
1989 Radiocarbon Age of the Dog Burial from
Squantum, Massachusetts. Bulletin of the Massa-
chusetts Archaeological Society 50:29.
Noli, D., and G. Avery
1988 Protein Poisoning and Coastal Subsistence.
Journal of A
rchaeological Science 15:39 5-401.
Oldale, R. N.
1986 Late-Glacial and Postglacial Sea-Level History
of New England: A Review of Available Sea Level
Curves. Archaeology of Eastern North America 14:
89-100.
O'Leary, M. H.
1988 Carbon Isotopes in Photosynthesis. BioScience
38:228-335.
Perlman, S. M.
1980 An Optimum Diet Model, Coastal Variability,
and Hunter-Gatherer Behavior. In Advances in Ar-
chaeological Method and Theory 3:257-310.
Peterson, B. J., R. W. Howarth, and R. H. Garritt
1985 Multiple Stable Isotopes Used to Trace the Flow
of Organic Matter in Estuarine Food Webs. Science
227:1361-1363.
Pretola, J., and E. A. Little
1988 An Archaeological Record from the Far Island.
Bulletin of the Archaeological Society of Connecticut
51:47-68.
Rau, C.
1885 Prehistoric Fishing in Europe and North Amer-
ica. Smithsonian Contribution to Knowledge No.
25. Smithsonian Institution, Washington D.C.
Ritchie, W. A.
1969 The Archaeology of Martha's Vineyard. Nat-
ural History Press, Garden City, New York.
Schoeninger, M. J., and M. J. DeNiro
1984 Nitrogen and Carbon Isotopic Composition of
Bone Collagen from Marine and Terrestrial Ani-
mals. Geochimica and Cosmochimica Acta 48:625-
639.
Schoeninger, M. J., M. J. DeNiro, and H. Tauber
1983 Stable Nitrogen Isotope Ratios of Bone Col-
lagen Reflect Marine and Terrestrial Components
of Prehistoric Human Diet. Science 220:1381-1383.
[Vol. 60, No. 2, 1995]
366
REPORTS
Schoeninger, M. J., and K. Moore
1992 Bone Stable Isotope Studies in Archaeology.
Journal of World Prehistory 6:247-296.
Schoeninger, M. J., and N. J. van der Merwe, K. Moore,
J. Lee-Thorpe, C. S. Larsen
1990 Decrease in Diet Quality between the Prehis-
toric and Contact Periods. In The Archaeology of
Mission Santa Catalina De Guale: 2. Biocultural
Interpretations of a Population in Transition, edited
by C. S. Larsen, pp. 78-94. Anthropological Paper
No. 68. American Museum of Natural History, New
York.
Schwarcz, H. P., J. Melbye, M. A. Katzenberg, and M.
Knyf
1985 Stable Isotopes in Human Skeletons of Southern
Ontario: Reconstructing Palaeodiet. Journal of Ar-
chaeological Science 12:187-206.
Speth, J. D., and K. A. Spielmann
1983 Energy Source, Protein Metabolism, and Hun-
ter/Gatherer Subsistence Strategies. Journal of An-
thropological Archaeology 2:1-31.
Spielmann, K. A., M. J. Schoeninger, and K. Moore
1990 Plains-Pueblo Interdependence and Human
Diet at Pecos Pueblo, New Mexico. American An-
tiquity 55:745-765.
Starbuck, A.
1924 The History of Nantucket. C. E. Goodspeed,
Boston.
Stuiver, M., and P. J. Reimer
1993 Extended 14C Data Base and Revised CALIB
3.0 14C Age Calibration Program [Rev. 3.03]. Ra-
diocarbon 35:215-230.
Szathmary, E. J. E., C. Ritenbaugh, and C-S. M. Goodby
1987 Dietary Change and Plasma Glucose Levels in
an Amerindian Population Undergoing Cultural
Transition. Social Science and Medicine 24:791-
804.
Tieszen, L. L., and T. Fagre
1993 Carbon Isotope Variability in Modem and Ar-
chaeological Maize. Journal of Archaeological Sci-
ence 20:25-40.
Tuross, N., M. L. Fogel, L. Newsom, and G. H. Doran
1994 Subsistence in the Florida Archaic: The Stable-
Isotope and Archaeobotanical Evidence from the
Windover Site. American Antiquity 59: 288-303.
U.S. Department of Agriculture
1978 Handbook of Agriculture Charts. Agriculture
Handbook No. 551. Government Printing Office,
Washington, D.C.
van der Merwe, N. J., and J. C. Vogel
1978 13C
Content of Human Collagen as a Measure
of Prehistoric Diet in Woodland North America.
Nature 276:815-816.
Walker, P. L., and M. J. DeNiro
1986 Stable Nitrogen and Carbon Isotope Ratios in
Bone Collagen as Indices of Prehistoric Dietary De-
pendence on Marine and Terrestrial Resources in
Southern California. American Journal of Physical
Anthropology 71:51-61.
Watt, B. K., and A. L. Merrill
1975 [1963] Handbook of the Nutritional Contents
of Foods. Handbook No. 8, U.S. Department of
Agriculture. Dover, New York.
Williams, R.
1936 [1643] A Key into the Language ofAmerica. 5th
ed. Reprinted for the Tercentenary Committee,
Providence, Rhode Island.
Winkler, M. J.
1985 A 12,000-Year History of Vegetation and Cli-
mate for Cape Cod, Massachusetts. Quaternary Re-
search 23:301-312.
Winslow, E.
1841 [1624] Good News from New England.... In
Chronicles of the Pilgrim Fathers of the Colony of
Plymouth, from 1602-1625, edited by A. Young, pp.
354-367. Little and Brown, Boston.
Wood, W.
1865 [1635] New England's Prospect, edited by C.
Deane. Prince Society, Boston.
Yesner, D. R.
1988 Subsistence and Diet in North-Temperate
Coastal Hunter-Gatherers: Evidence from the
Moshier Island Burial Site, Southwestern Maine. In
Diet and Subsistence: Current Archaeological Per-
spectives, edited by B. Kennedy and G. LeMoine,
pp. 207-226. Proceedings of the 19th Annual Con-
ference of the Archaeological Association of the
University of Calgary, Alberta.
Notes
Radiocarbon ages will be specified as B.P. (14C years
before 1950, conventional, 613C corrected; ?+), or as
calibrated ages, cal A.D. (the mean value of a calibrated
age in parentheses between the plus and minus sigma
ages, or the plus and minus sigma age range only). Cal-
ibrated ages are derived using CALIB 3.03 (Stuiver and
Reimer 1993), with AR = -95 ? 43 14C years for marine
materials (Little 1993:465).
2 Expressed in the delta (6) notation, the '3C/'2C ratio
(R) is: 6'3C = (Rsample/RPDB standard - 1) x 1,000%o (per
thousand or per mil), relative to the PDB (Peedee bel-
emnite marine carbonate) standard. Similarly, the '5N/
14N ratio (R) is expressed as b'5N = (Rsample/RAiR
- 1)
x 1,000%oo,
relative to AIR (ambient inhalable reservoir).
3 Photosynthesis occurs in plants by C3, C4, or CAM
metabolic pathways (O'Leary 1988). Most temperate zone
plants are of the C3 type, with low 613C values; C4 plants,
with high 613C values, include tropical grasses such as
maize, sugarcane, etc. The CAM pathway is found in
desert plants.
4 Note that percentages by weight and energy of nutri-
tional components are not identical. To convert P (pro-
tein), C (carbohydrates), F (fat) from g/100 kcal to kcal,
1
g P = 4 kcal, 1 g C = 4 kcal, and 1 g F = 10 kcal (Watt
and Merrill 1975).
5 The bone samples were cleaned mechanically of surface
debris and then ultrasonically in double distilled water.
Pieces of bone were demineralized in 1 M HC1, rinsed
to neutrality in double distilled water, soaked for 20
hours in .1 M NaOH, rinsed to neutrality in double
distilled water, and lyophilized to dryness. The flesh
samples were separated mechanically into fat and meat.
The meat was lyophilized to dryness. Mollusc meat was
367
AMERICAN ANTIQUITY
separated from stomach portions; both were lyophylized
to dryness. Plant material was separated mechanically
in order to analyze only those portions that would have
been eaten. These were ground at liquid nitrogen tem-
peratures and then lyophilized to dryness. Prior to the
final drying step, all samples were placed in acetone for
24 hours for lipid extraction.
In all cases, dried organic material was weighed into
clean quartz tubing with excess cupric oxide, elemental
copper, and elemental silver. The tubing was sealed un-
der vacuum, combusted at 800?C and allowed to cool
to room temperature. The resulting H20O,
CO2, and N2
were separated cryogenically in a glass vacuum line. For
the plant samples, only the CO2 was collected. Kjeldahl
digestions were performed on all plant samples and the
N2 collected cryogenically using liquid nitrogen and mo-
leular sieve. The CO2 and N2 were analyzed mass spec-
trometrically. The N2 was prepared in M. Schoeninger's
laboratories at Harvard University and the University
of Wisconsin. It was measured in James McCarthy's
laboratory at Harvard University on a V.G.602E and in
Schoeninger's laboratory at the University of Wisconsin
using a Finnegan MAT 251. To assess precision between
the different mass spectrometers and laboratory prepa-
rations, duplicate preparations were made on 25 samples
(four of these prepared in triplicate); the average differ-
ence was .4 per mil; the total range of differences between
any two duplicate preparations on one sample was 1.0
per mil. This variation includes those duplicates made
using the Kjeldahl preparation. The CO2 was prepared
and measured in Schoeninger's laboratories at Harvard
University and at the University of Wisconsin, and mea-
sured at John Hayes's laboratory at Indiana University
on a Finnegan MAT Delta E and at Schoeninger's lab-
oratory at the University of Wisconsin on a Finnegan
MAT Delta E. To assess precision between the different
mass spectrometers and laboratory preparations, 30
samples were prepared in duplicate (seven of these in
triplicate). The average difference between duplicates was
.3 per mil with the greatest difference between duplicates
being 1.5 per mil. In the case of duplicate preparations,
the average value is given. Repeated analyses on a glycine
lab standard gave a precision (calculated as the standard
deviation of the measurements) of .2 per mil for 613C
values and .3 per mil for bS'N values.
6 To derive a simple linear mixing equation, consider
two dietary components A and B with 3b3CA = -28 per
mil and S3CB = -12 per mil, which are mixed into a
diet, D, with 6b3CD = -15 per mil. We wish to know
what percentages of D are A or B. If the following re-
lations hold: (1) A + B = 1 and (2) A6b3CA + B6b3CB =
6b3CD, then, substituting A = 1 - B from (1) into (2),
we derive the following: B = [613CD - 613CA] / [6b3CB -
63CA], for which the solution, with the parameter values
given, is: B = 80 percent, and A = 20 percent.
Received June 23, 1993; accepted December 29, 1994.
Data Collection ....... ; ;
: .-
i
X :a
from Human Skelet:al !0 . ; E
R
mnIne
m.. .
i. ....
::..
.
Remains
Edi
...Box 1249
e AR 72702-1249
fax 501-575-5453
368 [Vol. 60, No. 2, 1995]
