Title: Journal of Caribbean archaeology
ALL VOLUMES CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00091746/00004
 Material Information
Title: Journal of Caribbean archaeology
Series Title: Journal of Caribbean archaeology
Alternate Title: JCA
Abbreviated Title: J. Caribb. archaelo.
Physical Description: Serial
Language: English
Publisher: Christopher Ohm Clement ;
Christopher Ohm Clement
William F. Keegan
Place of Publication: Gainesville FL
Publication Date: 2003
Frequency: annual
regular
 Subjects
Subject: Archaeology -- Periodicals -- Caribbean Area   ( lcsh )
Antiquities -- Periodicals -- Caribbean Area   ( lcsh )
Genre: periodical   ( marcgt )
 Notes
System Details: Mode of access: World Wide Web.
Dates or Sequential Designation: Vol. 1 (2000)-
General Note: Title from title screen (publisher's Web site, viewed Dec. 2, 2002).
General Note: Latest issue consulted: Vol. 5 (2004).
 Record Information
Bibliographic ID: UF00091746
Volume ID: VID00004
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 003345617
oclc - 41077527
lccn - sn 99003684
issn - 1524-4776

Downloads

This item has the following downloads:

Binder1 ( PDF )


Full Text





Journal of' ----
I II76




ARCHAEOLOGICAL RECONNAISSANCE OF ANEGADA,
BRITISH VIRGIN ISLANDS

Dave Davis
Tulane University
New Orleans, LA 70118
ddavis113@cox.net

Kevin Oldfield
258 Broadway
New York, NY 10007
.cI 'ii !lficl l ,(3 Itt i lct

An initial archaeological reconnaissance of Anegada, the only low-lying coral and limestone
island in the Virgin group, yielded information about two pre-Columbian midden sites and other
anthropogenic features of probable pre-Columbian age. Of additional note is an absence of sites
along Anegada's coastline. This apparent anomaly seems to be a product of the island's highly
dynamic geomorphology. Possible reasons for pre-Columbian activity on Anegada are discussed,
and directions for future research are %,i c % t'c 1


Anegada (Figure 1), the northeasternmost of
the Virgin Islands, is a largely flat limestone and
coral atoll that lies 25 km north-northeast of
Virgin Gorda and 33 km northeast of Tortola,
the administrative center of the British Virgin
Islands. An archaeological reconnaissance of
Anegada, conducted in July and December,
2002, resulted in the discovery of two pre-
Columbian midden sites; location of a series of
distinctive if not unique pre-Columbian conch
shell platforms; identification of natural
resources that may have attracted Native
American immigrants to Anegada in prehistory;
and recognition of geomorphic processes that
may have been responsible for the limited
surface visibility of pre-Columbian sites on the
island

Anegada has been visited briefly by
archaeologists on two previous occasions. In
1937, Herbert Krieger, with an accompanying
entourage, visited the island for "a few days" to
investigate reports of "a large shell mound"
(Krieger 1938). Krieger reported that the
mound of conch shells, located "near the eastern


end of Anegada," was measured (although
measurements are not presented in his report),
and he published a photograph of it (Krieger
1938:97). Krieger stated, however, that the
mound was not excavated because it was
"devoid of any cultural material other than the
discarded conch shells" (1938:98). Krieger
(1938:98) also reported that he recovered
"pottery, shell, and polished stone implements
gathered at random from the surface" in a brief
survey of other areas of the island, but he gave
no indication of the locations of those finds, nor
did he indicate whether they came from an
identifiable site, or were isolated surface finds.

In 1974, Jeffrey Gross and Alfredo
Figueredo spent three days on Anegada. They
located and published brief descriptions of two
"shell heaps" near the eastern end of the island,
including the mound described by Krieger
(Gross 1975). Shallow test pitting of an area
near one of the shell piles yielded "several
sherds" which were generally non-diagnostic
although, in the judgment of Gross and
Figueredo, the pastes were reminiscent of


Journal of Caribbean Archaeology 4, 2003









Reconnaissance of Anegada

I- h


0 1 2 miles
0 1 2 kilometres


Flamincic
Pon ci


Figure 1. Map of Anegada.


"Elenoid series" pottery from Puerto Rico and
the Virgin Islands. Two conch shells from the
associated midden deposit were taken as
radiocarbon samples; these yielded a date
reported as AD 1245+/-80 (Gross 1975:15).
Gross and Figueredo did not find pottery in the
shell heaps themselves, and their reconnaissance
did not extend to other areas of the island.

The only other record of previous
archaeological work on Anegada takes the form
of a private collection of pre-Columbian artifacts
that was obtained some years ago by Mr.
Wilfred Creque, an Anegada resident, from an
undocumented locality in the island's remote and
unsettled East End. The several hundred artifacts
in that collection include a variety of
undecorated ceramic body sherds; two rims
bearing anthropomorphic adornos (human
faces); two conch shell adzes; and some 15
ground and polished stone adzes manufactured
on various fine-grained volcanic rocks.

With an area of 39 km2 and a population of
approximately 180, Anegada is the second
largest but the least populated and most remote
of the British Virgin Islands. With a maximum
elevation of 8.5 m, the topography of this
"drowned land," as the Spanish named it,
presents a sharp contrast with the much higher
volcanic and metamorphic islands that comprise
the rest of the Virgin group. Like the Outer


Leeward Islands to the south, Anegada's low
topography and coral/limestone bedrock derive
from its location at the northeastern edge of the
Caribbean tectonic plate (U.S. Geological
Survey n.d.).

Although Anegada is characterized by
poorly developed soils, low rainfall, and
relatively sparse (primarily scrub) vegetation
(Figure 2), its distinctive geology provides
certain resources that would have been valuable
for pre-Columbian inhabitants. Unlike many
other low-lying islands in the northeastern
Caribbean, Anegada has a readily available
supply of fresh water, which occurs in natural
wells in the exposed limestone (Schomburgk
1832). Fresh to slightly brackish (but potable)
water can also be obtained from shallow hand-
dug wells in the sandy soils near the coastline
(Schomburgk 1832) (Figure 3, this paper).
Although much of Anegada is marked by
exposed limestone or coral bedrock, or by sand,
limited areas of shallow organic soils do occur
in the eastern third of the island. Indeed, in the
early and mid 20th century, cotton, bananas, and
sugar cane were produced in that area by the
small resident population (B-V-I Guide n.d.).
The island also possesses a wealth of marine
resources in its shallow offshore waters.
Lobsters, conch, and a wide array of other
shellfish and bonefish are abundant, particularly
near Horseshoe Reef, which, with a total length


Journal of Caribbean Archaeology 4, 2003


m


Davis and Oldfield









Reconnaissance ofAnegada


Figure 2. The interior of Anegada is dominated by scrub vegetation.


of 51 km, is the world's third largest barrier reef
(British Virgin Islands n.d.).

The Reconnaissance

Our eleven-day reconnaissance was aimed at
determining whether any substantial evidence of
Journal of Caribbean Archaeology 4, 2003


pre-Columbian occupation could be found on
Anegada. In keeping with most preliminary
investigations of this kind, limitations of time
did not allow for a systematic survey of the
entire island, nor a stratified survey of selected
habitats. Instead, we focused on areas that
seemed most likely to contain evidence of


Davis and Oldfield









Reconnaissance ofAnegada
middens or other occupation refuse, covering as
many of those areas as we could reach by
pedestrian survey given the limitations of time.
During the course of the reconnaissance, we
completed a pedestrian surface survey of the
southern shore of Flamingo Pond (Figure 1);
the northern shore of the unnamed pond to the
southwest of Budrock Pond; the southeastern
shore of Budrock Pond (including the location
of Anegada I; see Figure 1 and below); the
western half of the southern coast; the western
end of the island; the coastline of Loblolly Bay
(on the north coast to the northeast of the
airport); and an area of approximately 0.5 km2
immediately to the east of The Settlement
(including the location of Anegada II; see Figure
1 and below).

Archaeological reconnaissance in the West
Indies typically is guided by the results of
previous investigations on other islands.
However, this procedure cannot be readily
applied to Anegada. Although a great deal of
archaeological research has been completed
elsewhere in the Virgin Islands and in the
northern Lesser Antilles, many of the landforms
and habitat types that are common on those
islands are absent on Anegada. Several of the
landform types that have figured prominently in
characterizations of pre-Columbian settlement
patterns on other small islands in the eastern
Caribbean -- particularly valley bottomlands
with good agricultural soils and access to fresh
water, deep bayheads, and small peninsulas (e.g.,
Davis 1988; Haviser 1990, 1997) -- are absent
on Anegada (most bays on the island are merely
slight indentations in the coastline). In short,
settlement pattern models developed on higher
islands are largely inapplicable on Anegada.

Looking somewhat farther afield, the
Bahamas, with their limestone and coral
bedrocks, low terrain, and relatively
undifferentiated interiors, would seem to offer
better guidance. Assessing the results of some
40 years of archaeological survey in the
Bahamas, Keegan stated that:

These surveys have identified a
predominantly coastal orientation of


Davis and Oldfield
settlement. More specifically, Lucayan
sites tend to be restricted to leeward
sand beaches adjacent to shallow marine
grass flats ... Because previous studies
have shown that Lucayan settlements
were usually restricted to coastal
habitats, these locations have been the
focus of most research efforts (Keegan
1992:68).

This general preference for coastal localities is
consistent with other islands in the Virgin group,
and most islands in the Lesser Antilles, where
the vast majority of pre-Columbian sites,
especially before and after the early Ceramic
Age, are located within a few hundred meters of
the coast; indeed, many sites on those islands
are situated less than 100 m from the shoreline.
On Anegada, however, our examination of
numerous areas along and near beaches yielded
no evidence of pre-Columbian human activity.
The contrast with other islands in the Virgin
group, as well as with the Bahamas, is striking.
On high islands like St. John and St. Thomas
(e.g., Bullen and Sleight 1963; Brewer and
Hammersten 1988), pre-Columbian sites can be
found on or just behind the beaches at the head
of nearly every major bay. Indeed, the
importance of the sea for transportation and as a
food source were excellent reasons for pre-
Columbian peoples to live in close proximity to
the shoreline.

The apparent absence of sites in coastal
locations on Anegada may be a product of
natural post-depositional processes rather than
of pre-Columbian settlement patterns. Even in
historic times, storm surges and wave action
have significantly altered the Anegada coastline.
Local informants report the deposition of more
than a vertical meter of sand at coastal locales by
a single hurricane. Shoreline loss has also been
extensive. Following the passage of Hurricanes
Gabrielle and Hugo in 1989, 15 horizontal
meters of beach was lost on Anegada's north
shore at Loblolly Bay, while Cow Wreck Bay,
also on the north shore, gained over 30 meters.
On the south shore, over 15 m of beach was lost
at the Anegada Reef Hotel, and over 32 m were
lost at Nutmeg Point (UNESCO n.d.).


Journal of Caribbean Archaeology 4, 2003









Reconnaissance ofAnegada
Anegada is not unique in this regard; similar
and even greater amounts of beach loss occur
during major hurricanes elsewhere in the Virgin
Islands (UNESCO n.d.). At Cinnamon Bay, St.
John, an undetermined but extensive portion of a
late pre-Columbian beachfront site has washed
into the sea as a result of scouring produced by
hurricanes (Kenneth Wild, personal
communication, July 2000). Also on St. John,
the presence of a deeply buried preceramic site
at Lameshur Bay (Wild 1989) attests to the
deposition of substantial amounts of sand on
beaches, presumably by storm-driven waves. Yet
many pre-Columbian beachfront sites elsewhere
in the Virgin Islands remain visible to surface
examination. Since Anegada lacks deeply
indented bays, it may be subject to greater
reworking of the coastline during hurricanes.
An 1832 map of the island by German-British
explorer R.H. Schomburgk (Figure 2) provides
some evidence in favor of this hypothesis.
Schomburgk's map clearly shows that the
western salt ponds, now separated from the
ocean by over 100 m, were connected to the sea
on the south shore when Schomburgk visited
the island in 1831. Schomburgk (1832:158)
reported that "there was likewise [a connection
to the sea] on the northern side, but the
hurricane of 1819 stopped its passage." He also
stated that

The western end of the island has been
covered with sand, forced forward by an
immense ground sea or surf, to which it
is still subjected from time to time, and
hence the continual change of the figure
of the bays in that part ... The whole
north side is exposed to an impetuous
sea, but mostly on the northwestern part,
where the sand has formed little hillocks
of 40 feet in height. Behind the first
range is a second and a third
(Schomburgk 1832:157).

In addition, over a hundred years earlier, Pere
Labat noted that substantial areas of Anegada
were invaded by the sea at high tide (Labat
1970:205). Finally, although direct reports from
Anegada appear to be non-existent, tsunamis
would have exerted significant impacts on such


Davis and Oldfield
a low island. For example, in 1867, an
earthquake in the Anegada Trough between St.
Croix and St. Thomas generated a tsunami that
affected the entire West Indies east and south of
Hispaniola. In Charlotte Amalie, St. Thomas, a
wave 4.5-6 m high killed 12 people and
swamped many boats. At Roadtown, Tortola,
waves some 1.5 m high destroyed houses, and
damage was experienced as far south as the
Windward Islands (Lander, Whiteside, and
Lockridge 2002). Evidence of potentially
significant loss of pre-Columbian sites through
natural post-depositional processes has been
documented on other islands (e.g., Crock 2000;
Delpuech et al. 1999). Because of Anegada's
low topography and lack of enclosed harbors
that could afford some protection from
catastrophic events like tsunamis and major
storms, the opportunities for both scouring and
burial of archaeological sites appear to have
been greater than on many other West Indian
islands.

In light of the abundance of salt that is
available from the ponds on Anegada during dry
months, and the consequent possibility that
indigenous people from other islands may have
periodically extracted this resource from
Anegada, close attention was paid to these
landforms. Settlement around salt ponds has
been reported from a number of other islands in
the West Indies (e.g., Delpuech et al. 1999;
Hofman et al. 1999; Nokkert et al. 1995; de
Waal 1999). Approximately 5.8 linear km
adjacent to salt ponds were examined on
Anegada, but there was no indication of pre-
Columbian human activity along the immediate
pond shorelines. It should be noted, however,
that both of the pre-Columbian sites that were
identified during the reconnaissance (see below)
are situated within 100 m of salt ponds; indeed,
both sites are located on the closest firm, well-
drained land to the nearby ponds. Sullivan
(1981) suggested that certain Ceramic Age sites
in the vicinity of salt ponds in the Bahamas were
salt collection stations. We are, of course, not in
a position at this stage to suggest whether either
or both of the pre-Columbian sites on Anegada
served a similar function.


Journal of Caribbean Archaeology 4, 2003









Reconnaissance ofAnegada
Sites

The first of two sites found during the
reconnaissance, Anegada I, was located with
information provided by a local resident who
reported finding "Indian artifacts" in the East
End "years ago." The informant described the
site's location as east of Budrock Pond on land
with "soil and trees." By "soil," we learned, the
informant referred to dark organic soils, which
are virtually absent on Anegada outside of
archaeological sites. Indeed, trowel tests
indicated that Anegada I consisted of 9-22
vertical cm of dark midden sitting directly on top
of limestone bedrock.

Anegada I is located at and around
18'41.855' N and 64'16.706' W. The site is
situated at an elevation of about 5 m on a low
hill that overlooks Budrock Pond (Figures 4 and
5), which is some 40 m west of the site. Midden
soil formation is adequate to support a number
of broadleaf trees, which are exceedingly rare
elsewhere on the island. The site surface is
littered with shell and a fairly low density of
potsherds, as well as fish and other bone.


Davis and Oldfield
Midden and artifacts were encountered over an
area of approximately 45 x 30 m.

In addition to shell and animal bone, 25
potsherds were recovered from Anegada I. Two
are rim sherds, and one of those is decorated
with broad arching curvilinear incision
reminiscent of late Ceramic Age ceramics from
elsewhere in the Virgin Islands and Puerto Rico.
That sherd and four of the body sherds contain
a light-colored grit temper that appears to be
plagioclase feldspar (a material that is absent in
the carbonate bedrock of Anegada). The other
rim sherd and the remaining body sherds have
no apparent temper, and are relatively thin with
homogeneous pastes.

A noteworthy artifact in the surface
collection from Anegada I is a large mid-shaft
fragment of a ground stone artifact, probably
either a pestle or an ax. The raw material is
diorite, a volcanic rock that is present in the East
End range on Virgin Gorda (University of the
West Indies n.d.) but absent on Anegada. The
specimen weighs 425 g and measures
approximately 8.3 cm x 7.4 cm x 1.9 cm.


I
'I


~;#*~ -~.


Figure 4. Budrock Pond in the East End.

Journal of Caribbean Archaeology 4, 2003


UB--


-a.. ii









Reconnaissance ofAnegada


Figure 5. Goat trail across Anegada I.


Together with the feldspar tempered pot sherds
that we collected and the ground stone adzes in
the collection of Mr. Wilfred Creque (see
above), this artifact provides evidence of pre-
Columbian ties between Anegada and islands to
the west or south.

Only three species of shell were observed at
Anegada I: Codakia orbicularis, Strombus
gigas, and Nerita tessalata. A variety of
fishbone, including several specimens of
parrotfish (Scaridae) were also found, as well as
five fragments of bird longbone that have yet to
be positively identified, but that may represent
roseate flamingo (Phoenicopterus ruber) or
another large bird.

Anegada II was found on the last day of the
reconnaissance. The site is located at and around
18'43.048' N and 64'18.799' W, near the outer
southeastern edge of The Settlement. As with
Anegada I, the site was marked by dark midden
soil that contrasted sharply with the surrounding
sands, corals, and poorly developed limestone
soils. The density of pre-Columbian artifacts at
Anegada II was extremely low; our collection
Journal of Caribbean Archaeology 4, 2003


consists of one undecorated rim sherd, three
undecorated body sherds, and an articular end of
a mammalian long bone that appears to be
manatee. Only the mammal bone was found in
a buried context. Also present on the site were
various early and middle twentieth century
artifacts that apparently were associated with an
abandoned homestead. The midden soil ranges
in depth from eight to over 30 cm.

Perhaps the most enigmatic anthropogenic
features discovered during the reconnaissance
was a series of three conch shell (Strombus
gigas) platforms (Figure 6) that are located in
the East End near White Bay and the
southeastern shore of Shell Pond, about 400 m
southwest of Anegada I. Limitations of time did
not allow full mapping of these unusual features,
each of which consists of thousands of conch
shells. Unlike the large piles of opened conch of
colonial and modern age that can be found in
nearshore waters near The Settlement, the
conchs in the platforms are heavily weathered
and oxidized to a dark blue-gray color. The
platforms are flat-topped, approximately 50 cm
high, very roughly rectangular in plan form, and


Davis and Oldfield









Reconnaissance ofAnegada


Figure 6. One of three conch shell platforms in the East End.


range in area from approximately 70 m2 to 130
m2.

Residents of Anegada today regard the
conch platforms as part of an "Indian burial
ground," and that idea itself has considerable
antiquity. Mr. Wilfred Creque of Anegada has
a map dated June 10, 1824, entitled "A Chart of
the Islands of Anegada, Together with the
Vessels Wrecked Upon Them" (produced by
Charles Noyce for the Queen's Quartermaster's
Office) bearing four triangles in the location of
the shell platforms. The triangles are labeled
"Pyramids of Conch Shell Left by [word
obscured] Indians."

Large accumulations of presumably
prehistoric conch were observed in the early
1830's by German-British naturalist R.H.
Schomburgk (Figure 7), who wrote that

Pere Labat, the only early writer who
speaks on the lesser West Indian isles,
observes, that the aborigines used


[Anegada] as an occasional rendezvous,
where they procured great quantities of
conchs (strombus gigas); and large
piles of these shells are still to be seen at
the east end of the island, but nowhere
else; which seems to prove decidedly
that it was not permanently occupied, but
merely resorted to from time to time ...
the dry shells piled up all have a hole in
the lower end of the spire, for which the
most probable reason is, that the animal
is thus most easily extracted. It appears
surprising that so much care has been
taken to pile them up, and it has been
surmised, in consequence, that these
heaps were burial-places; but several
have been taken down, and burnt for
lime -- without any trace having been
found of human bones, or other
extraneous substance. And it is more
probable that they were merely piled up
to be out of the way, the current not
being strong enough to carry them off
had they been thrown into the sea;


Journal of Caribbean Archaeology 4, 2003


Davis and Oldfield









Reconnaissance ofAnegada


Figure 7. Eastern portion of Anegada in 1832. From Schomburgk (1832).


where, had they remained, they would
have embarrassed the fishing for the
living animal1 (Schomburgk 1832:153).

The conch platforms that we visited do not
appear to be the same "shell heaps" that were
investigated by Krieger (1938), or Gross and
Figueredo (Gross 1975). The higher of the two
shell piles examined by Gross and Figueredo,
which they identified with the mound visited by
Krieger, was reportedly 2-3 m high, covering an
area of about 200 m2, making it both higher and
more extensive than any of the three platforms
we visited. Indeed, we identified a conch pile of
roughly that dimension some 300 m east of the
conch platforms, at 1841.438' N and
6416.643' W. Moreover, none of these authors
noted that the mounds or heaps they examined
were uniformly flat-topped. Further, Gross
reported that at least one of the two piles they
examined consisted of "whitened conch,"
whereas the platforms we saw were each
composed of blue-gray, highly oxidized, shell.
The color is notable; in our experience, shell
typically weathers to a bright white color, and
the blue-gray color of the shell in the platforms
suggests the possibility that they were either


burned or covered with earth for some period of
time. The functions of all the presumably pre-
Columbian conch accumulations on Anegada,
whether as substructures of some sort, adjuncts
of burial or other ritual activity, or merely as
dumps for harvested Strombus shells (however
carefully discarded), remain undetermined.

Questions and Potential for Future Research

In comparison to many other islands in the
Virgin group and the Lesser Antilles, Anegada
has thus far yielded only limited evidence of
pre-Columbian occupation. Nevertheless,
Anegada is intriguing for both substantive and
methodological reasons. While our
reconnaissance was not a comprehensive
systematic survey, we did achieve a level of
coverage that undoubtedly would have yielded
more archaeological data on any number of
other islands. These results suggest either that
(as Schomburgk [1832] speculated) Anegada
was visited only sporadically by pre-Columbian
people from other islands; or that Anegada has
experienced a high rate of archaeological site
loss and/or concealment from storm action,
normal tidal activity, and along shore currents.


Journal of Caribbean Archaeology 4, 2003


Davis and Oldfield









Reconnaissance ofAnegada
While Anegada may have been a special
procurement locality for conch, the uniform
heights of the shell platforms, and the highly
oxidized condition of the shells, suggest some
activity additional to meat extraction. This
question needs to be investigated through
disassembly and careful recording of portions
of the platforms, and through excavation of
areas beneath the shells. In addition, of course,
additional radiocarbon analyses of the shells
should be conducted to ascertain their pre-
Columbian age and the range of time that they
represent.

The small collection of artifacts and food
remains recovered from Anegada I is consistent
with a habitation site, possibly of late pre-
Columbian age. Although the site's remote
location some seven km from the nearest road
creates logistical challenges, further testing
could document the presence of pre-Columbian
settlers, as opposed to visitors. Anegada II
should also be tested with the same question in
mind.

It seems highly likely that the dynamic
coastal geomorphic environment of Anegada has
significantly affected our current view of the
island's pre-Columbian archeological record.
Deep testing of beaches and adjacent coastal
areas through an augering program is
theoretically possible, but would be highly time-
intensive with little clear prospect of success.
However, a more thorough survey of inland
areas within 100 m of salt ponds, and of areas
containing natural fresh water wells, should be
undertaken with specific attention to identifying
organic soils that may represent archaeological
middens.

Pre-Columbian inhabitants of other Virgin
Islands, especially Virgin Gorda and Tortola,
clearly were aware of the existence of Anegada,
and it is equally clear that indigenous people
visited and almost certainly settled on the island
for at least some period in prehistory. Rainfall
is low and Anegada's weakly developed soils
appear poorly suited to agriculture. Indeed, they
are so shallow that it is difficult to imagine
successful cultivation of a root crop like manioc,
resistant as it is to drought and fluctuations in
Journal of Caribbean Archaeology 4, 2003


Davis and Oldfield
rainfall. However, certain other resources are
bountiful. In addition to the island's rich fishery
and the availability of fresh water, in prehistory,
roseate flamingos and rock iguanas, severely
reduced in numbers during historic times, were
abundant, and salt has been a readily available
resource from prehistoric through modern
times. We hope that more detailed
investigations in the future will be able to
determine which of these, or other, resources
drew and sustained pre-Columbian inhabitants
to this most distinctive of the Virgin Islands.

Acknowledgments. We gratefully acknowledge the
assistance of Marcus Eberl in the production of Figure 3.
We are also pleased to recognize the Perry-Castafieda
Library Map Collection at the University of Texas,
Austin as the repository of Schomburgk's (1832) map of
Anegada (Figures 2 and 3 of this paper).

References Cited

Brewer, D.M. and S. Hammersten
1988 Archaeological Overview and Assessment,
Virgin Islands National Park, St. John, U.S. Virgin
Islands. Southeast Archaeological Center, National
Park Service, Tallahassee, Florida.
British Virgin Islands
n.d. Coral Reefs. http://www.bvi.gov.vg/template.
php?main=atbvi§ion=environment&sub=
coral_reefs. Accessed 1/08/03.
Bullen, R.P. and F.W. Sleight
1963 The Krum Bay Site: A Preceramic Site on St.
S .- United States Virgin Islands. The Wm. L.
Bryant Foundation, American Studies No. 5.
Orlando.
B-V-I Guide
n.d. Anegada -- The Drowned Island in the BVI.
http://www. b-v-i.com/anegada/default.htm.
Accessed 1/08/03.
Crock, J.G.
2000 Interisland Interaction and the Development of
Chiefdoms in the Eastern Caribbean. Ph.D.
dissertation, University of Pittsburgh. University
Microfilms, Ann Arbor.
Davis, D.D.
1988 Biogeography and Human Subsistence:
Examples from the West Indies. Archaeology in
Eastern North America 16:177-186.
Delpuech, A., C.L. Hofman, and M.L.P. Hoogland
1999 Excavations at the Site of Anse a la Gourde,










Reconnaissance ofAnegada
Guadeloupe: Organization, History, and
Environmental Setting. Proceedings of the XVIIIth
International Congress for Caribbean
Archaeology, pp. 156-161. Grenada.
Gross, J.M.
1975 The Archaeology of Anegada Island. Journal of
the Virgin Islands Archaeological Society 2:12-16.
Haviser, J.B., Jr.
1990 Geographic, Economic, and Demographic
Aspects of Amerindian Interaction Between Anguilla
and St. Martin-St. Maarten. Paper presented at the
55th Annual Meeting of the Society for American
Archaeology, Las Vegas.
Hofman, C.L., A. Delpuech, and M.L.P. Hoogland.
1999 Excavations at the Site of Anse a la Gourde,
Guadeloupe: Stratigraphy, Ceramic CI ii. ,-1i. I. '* and
Structures. Proceedings of the XVIIIth
International Congress for Caribbean
Archaeology, pp. 162-172. Grenada.
Keegan, W. F.
2000 The People Who Discovered Columbus: The
Prehistory of the Bahamas. University Press of
Florida, Gainesville.
Krieger, H.W.
1938 Archaeology of the Virgin Islands.
Explorations and Field-Work of the Smithsonian
Institution in 1937:95-102. Smithsonian
Institution, Washington, D.C.
Labat, J.B.
1970 The memoirs of Pere Labat, 1693-1705.
Translated and abridged by John Eaden; with an
introduction by Philip Gosse. F. Cass, London.
Lander, J.F., L.S. Whiteside, and P.A. Lockridge
2002 A Brief History of Tsunamis in the Caribbean
Sea. Science of Tsunami Hazards 20(2):57-94.
Nokkert, M., A. Brokke, S. Knippenberg, and T.
Hamburg.
1995 An Archaic Occupation at Norman Estate, St.
Martin. Proceedings of the XVIIth International
Congress for Caribbean Archaeology, pp. 333-
351. Grenada.


Davis and Oldfield
Schomburgk, R.H.
1832 Remarks on Anegada. The Journal of the
Royal Geographical Society 2:152-170.
UNESCO
n.d. Hurricane Impact on Beaches in the Eastern
Caribbean Islands 1989-1995. http://www.unesco.
org/csi/act/cosalc/hurl6.htm. Accessed 1/8/03.
U.S. Geological Survey
n.d. Earthquakes and Tsunamis in Puerto Rico and
the Virgin Islands. http://pubs.usgs.gov/factsheet/
fsl41-00/fsl41-00.pdf. Accessed 8/9/02.
University of the West Indies
n.d. Landslides in the U.S. Virgin Islands. Unit for
Disaster Studies, Department of Geography and
Geology, University of the West Indies, Mona,
Jamaica. http://isis.uwimona.edu.jm/uds/Land
USVI.html Accessed 7/29/02.
Waal, M.S. de
1999 The Pointe des Chateaux Survey (1998): A
Preliminary Report. Proceedings of the XVIIIth
International Congress for Caribbean
Archaeology, pp. 268-278, Grenada.
Wild, K.L., Jr.
1989 Archaeological Investigations Conducted
Along Lameshur Road, St. John, U.S. Virgin
Islands. National Park Service, Southeastern
Archaeological Center, Tallahassee, Florida.

Notes

1 Local fishers report that, if freshly cleaned conch shells
are discarded in the sea, living conch will in fact
abandon the area. This behavior may be a response to
an increase in conch predators such as horseshoe crabs
in conch discard areas (Julie Gauthier, personal
communication, December 2002).


Journal of Caribbean Archaeology 4, 2003









Journal of' ----
I II 76




Plum Piece
Evidence for Archaic Seasonal Occupation on Saba,
Northern Lesser Antilles around 3300 BP

Corinne L. Hofman

Menno L.P. Hoogland

Recent investigations on the island of Saba, northern Lesser Antilles, revealed evidence of
preceramic occupation in the northwestern part of the island at an elevation of approximately 400
m above sea level. The inland location of dense midden deposits in a tropical forest environment
makes the Plum Piece site unique for studying the preceramic occupation of the Antilles, a period
that is otherwise mainly known from coastal settings. The recovered artifacts and the
radiocarbon dates support an attribution to the Archaic period of the preceramic Age. The nature
of the tools and the restricted number of exploited food sources i, ,c t a temporary, probably
seasonal, occupation of the site for a unique activity.


Archaeological investigations on the island
of Saba, northern Lesser Antilles (Figure 1)
during the summers of 2001 and 2002 revealed
evidence of preceramic occupation at the site of
Plum Piece in the northwestern part of the island
dating from approximately 3300 BP. Prior to
these investigations a preceramic date of 3155+
65 BP had been obtained from the Fort Bay area
in the northeastern sector of Saba (Roobol and
Smith 1980). This date came from a shell adze
with a poor context. No additional proof was
ever found for preceramic occupation in that
area, or elsewhere on Saba, except for some
isolated stone tools reported from the interior of
the island and two flint blades recovered from
construction activities at The Level in
Windwardside. In contrast to the limited data
on the preceramic occupation of Saba, Ceramic
Age settlements are known from along the north
and southeastern side of the island dating
between approximately AD 400 and 1400
(Hofman 1993; Hoogland 1996).

Contemporary preceramic sites and tool
assemblages comparable to Plum Piece are
known from other northern Lesser Antilles, but
most are from coastal settings. Faunal
assemblages in these sites point to a focus on


coastal exploitation in which shellfish
predominates. The species collected are related
to the exploitation of specific coastal
environments, varying from mangroves to
shallow-water and shallow-reef habitats.

The atypical location of the site of Plum
Piece in the tropical forest area of Saba at an
elevation of 400 m above sea level provides
another picture. The dense midden deposits
consist mainly of landcrab and bird remains,
though fish and especially mollusks are present
in very small quantities. The restricted variety of
exploited food sources suggests a temporal and
seasonal occupation of the site. The tool
assemblage at Plum Piece is similar to the
Archaic preceramic complexes known from
nearby islands defined by Rouse (1986, 1992)
as the Ortoiroid series dating between 5000 and
1000 BP. Rouse suggested a northern South
American origin for the Archaic groups of the
Lesser Antilles because of the close resemblance
in toolkits between the mainland and the islands.
Tools at Plum Piece include stones for
hammering, pounding, battering and grinding,
shell adzes, and large quantities of flaked-flint
tools. Other remains are comprised of coral and
ochre.


Journal of Caribbean Archaeology 4, 2003












Plum Piece Hofman and Hoogland


Puerto Rico
Vrgin ISl3nds


-^. *' vieques
S1 Cro.'. ---


Anguilla
St Martin -. Si BarTh


SE Eusta'uu
St KtiS ".


Guadloupe La D6sirade




Dominica


Mart.n.Lue


Sr iLuca


0 100 200 300km


Figure 1. The location of Saba within the arc of the Lesser Antilles.


The remote location of this site, the restricted
exploitation of food sources, the character of the
tools and the provenance of the materials used
for their manufacture are expected to provide
interesting and new perspectives on choices of
settlement location, and on a more general level,
on the lifeways, mobility and interaction patterns
of the Archaic peoples of the Lesser Antilles.

The island of Saba: geomorphology and land
use

Saba is situated at 63' 14' W, and 17' 38' N,
approximately 50 km south of St. Martin and 30
km north-west of St. Eustatius. The island of
St. Croix, U.S. Virgin Islands, lies 80 km in a

Journal of Caribbean Archaeology 4, 2003


westerly direction.

Saba is one of the smaller islands of the
Lesser Antilles and has a surface area of only 13
km2 (Figure 2). Its small surface area, the
pronounced relief, which leads to a slightly
higher level of precipitation than the
surrounding islands, and its difficult access, give
Saba an exceptional and unique character.

The island is the upper part of an extinct
Pleistocene volcano, which rises steeply from
the seabed from a depth of 600 m. The summit,
Mount Scenery, stands 870.4 m above mean sea
level (amsl) and is often cloaked in fog. The


',-' Barbuda


St.Vincent


Barbados ,


ClGeachinei


' Genada


Plum Piece


Hofman and Hoogland


,-_ NE1I '_: Ant,qua









Plum Piece
outlines of the
volcano attest to a
complex structure.
The summit of the
island is situated in
the middle and is
enclosed by
numerous smaller
peaks. There is only
one large relatively
flat area, which is the
location of the main
village, The Bottom.
The villages of St.
John's and
Windwardside are
situated on other
relatively flat areas.
The majority of the
island has slopes of
more than 150, while
along the coast they
even exceed 459.

The landscapes of
Saba are strongly
affected by erosion.
Along the numerous
radial steep-sided
gullies or guts,
rainwater flows into


0 b 0 0.5 __g1 Him


k CoMakol laH
I-I P ift MAiaftini* ,ai
rjOOMMMAwA yitMwe"*
4~~~e ,,,Nut
uSrteManWual.


W E


Figure 2. Map of Saba with vegetation zones and known
archaeological sites.


the sea. Steep cliffs characterize the coast all
around the island. A gentler sloping coast
occurs in the few inlets of bays, like Cove Bay
and Spring Bay on the windward side of the
island. But here, heavy breakers on the boulder
beach hamper access from sea. The leeward
side of the island offers a good anchorage at
Ladder Bay and Well's Bay, but the shore is
very steep.

Due to the relief of the island the climate can
be subdivided in three meso climates. Based on
temperature and precipitation the lower
elevations (0 450 m) can be classified as a
savannah climate. Between 450 and 800 m there
is a tropical rainforest climate with a dry season.
On the higher elevations there is a tropical
rainforest climate (Augustinus et al. 1985:2).

Today the vegetation of Saba shows a
Journal of Caribbean Archaeology 4, 2003


Colonial and more recent land-use has
modified the original landscape and vegetation
of Saba in many respects. In the past, the
landscapes of Saba were mainly impacted by
activities related to agriculture and cattle grazing.
Today, the most important threats are goats and
the destruction and degradation of habitat
caused by unregulated development in general.

Historic ruins and agricultural terraces are
encountered all over the island suggesting
intensive occupation during the first half of the
20th century. The present day habitation centers
only along the western side of the island. The
eastern side has been desolate since the mid-
twentieth century when the last inhabitants of
1 Treefernbrake forms dense groves, about 4 m high,
almost solely of Cyathea arborea, Cyathea grandifolia
and sometimes also Cyathea muricata. Palmbrake
consists mainly of Euterpe globosa (Augustinus et al.
1985:61-62).


Hofman and Hoogland
zonation more or less
parallel to these climatic
zones. The vegetation of
the low areas up to 350
m can be described as
either Croton thickets,
dry evergreen woodland
or secondary woodland.
From 350 to 650 m the
vegetation consists of
secondary rainforest with
some species reflecting
the remnants of the
original forest.
Treefernbrake and
palmbrake' are the
vegetation of the higher
elevations, while elfin
woodland is
characteristic for the
summit above 825 m
(Stoffers 1956;
Augustinus et al. 1985).

In pre-Columbian
times the area between
350 and 650 m amsl was
presumably covered by
rainforest.


14









Plum Piece
Mary's Point moved to The Bottom and Hell's
Gate. Most of the present agriculture is
practised in the zone between 350 and 650 m on
scattered, small-cultivated plots all over the
island. On the eastern side such cultivated plots
constitute the only remaining activity.

Pre-Columbian occupational history and
cultural chronology

Saba's occupational history and cultural
chronology has been defined on the basis of
several archaeological investigations carried out
by Josselin de Jong in 1923 (1947), Haviser
during the mid-eighties (Haviser 1985) and
Hofman and Hoogland between 1987 and 1992
(Hofman 1993; Hoogland 1996; Hoogland and
Hofman 1993, 1999). Except for the find of
four shell tools with a preceramic date
suggesting an occupation of the island at around
1000 BC (Roobol and Smith 1980) and the
recent discovery of Plum Piece, Saba's pre-
Columbian occupational history is limited to the
ceramic period (i.e., between AD 400 and 1400).
Saba appears to have been inhabited for the first
time by Ceramic Age people during the late
phase of the Early Ceramic Age (i.e., around AD
400) as is evidenced from the sites of Spring
Bay la and Kelbey's Ridge 1. Ceramics from
this period are characterized by traits of the
Cedrosan Saladoid subseries. The major period
of Saba's pre-Columbian occupation is situated
between AD 800 and 1200 (i.e., the early phase
of the Late Ceramic Age), as is evidenced by the
The Bottom, St. John's and Spring Bay lb, 2
and 3 sites. Ceramics from these sites belong to
the Mamoran Troumassoid subseries. The sites
of Spring Bay Ic and Kelbey's Ridge 2 provide
occupation evidence for after AD 1200. In
accordance with the regional settlement pattern, a
decline in sites is noticeable during this period.
Ceramics from the Spring Bay Ic and Kelbey's
Ridge 2 show affiliations to the Chican
Ostionoid subseries of the Greater Antilles,
dating to ca. AD 1200-1500 (Hofman 1993).


Test excavations at Plum Piece

Plum Piece is situated in the northwestern
part of the island, on the leeward side (Figure 3).
Journal of Caribbean Archaeology 4, 2003


Hofman and Hoogland
This part of the island is densely vegetated today
which hampers the view and restricts the
discovery of pre-Columbian sites during field
surveys. The plot of land on which the site is
situated was cultivated during colonial times, as
is testified by the presence of terraces. The land
is currently under cultivation of root crops. The
landowner, Mr. Carl Zagers, discovered and
reported the presence of Amerindian tools while
working his land.

A survey during the summer of 2001
confirmed the presence of an Amerindian
occupation at Plum Piece through the recovery
of numerous pieces of flint and ground-stone
and shell tools from the surface. In addition, a
large in situ grinding stone was identified at the
site (Figure 4).

Two test pits measuring 1 x 1 m were dug to
investigate the presence and depth of eventual
cultural deposits. It appears that the site was
partially disturbed by hundreds of years of
cultivation and by the construction of the
terraces. Fortunately, the cultivation of root
crops did not exceed 40 cm in depth, resulting in
restricted damage to the underlying rich
archaeological deposits.

During the 2002 campaign an additional 7
m2 were excavated to obtain a better insight into
the stratigraphy of the deposits and to collect a
sample of material and faunal remains (Figure
5). The units were dug in 10 cm arbitrary levels


Figure 3. View of the Plum Piece site situated 400
m amsl. Below is Well's Bay.









Hofman and Hoogland


Figure 4. In situ grinding stone.


Figure 5. View of the excavation unit.


S0 1m

Figure 6. The stratigraphy shows a dense midden layer covered by a disturbed plough zone.
Figure 6. The stratigraphy shows a dense midden layer covered by a disturbed plough zone.


taking into account changes in the stratigraphic
layers. The material was dry sieved through a
6.4 mm (1/4") mesh. A sample from one 1 x 1
m unit was water screened through a 3.2 mm
(1/8") mesh to obtain an adequate sample for
faunal analysis and to collect flint chips.

The archaeological deposits encountered
during the two campaigns consist of dense
midden deposits composed of vast quantities of
faunal remains intermixed with a number of
flint, ground-stone and shell tools, coral
(Arcopora palmata) fragments and red ochre.2

The stratigraphy shows a dense midden
layer reaching a depth of 50 cm in some parts
covered by a disturbed humic plowzone mixed
with charcoal varying in thickness from 10 to 40
cm (Figure 6). The latter is the result of
2 A mixture of clay, silt and iron oxide of a red color
(Munsell chart color 10R 5/8). When pounded the ochre
is very suitable to be used as a pigment.
Journal of Caribbean Archaeology 4, 2003


charcoal burning at the site during colonial and
recent times. Charcoal holes with a diameter of
30 cm penetrate the midden deposits in some
areas.

Three radiocarbon dates of landcrab samples
from undisturbed midden contexts have
provided dates of 3430 30 BP (GrN-27562),
3300 30 BP (GrN-27563) and 3320 30 BP
(GrN-27564). Calibrated at a 2 sigma interval
the dates fall between 1875 and 1520 cal BC.

Pre-ceramic subsistence economy and toolkit

Preliminary results of the faunal analysis
carried out by graduate student Peter van den
Bos supervised by Dr. Thijs van Kolfschoten of
the zooarchaeological laboratory, Faculty of
Archaeology at Leiden University point to a
fairly restricted exploitation of food sources in
the vicinity of the site. The abundance of
exoskeletal fragments of landcrabs and bird
bones in the deposits indicates a heavy reliance


Plum Piece









Plum Piece
on terrestrial faunal sources. Fish and shellfish
remains are well preserved in the deposits but
are scarce and seem to have been of minor
importance in the diet of the preceramic
occupants of Plum Piece. The presence of
conch lip adzes suggest that the inhabitants of
Plum Piece likely captured and ate conch,
though they probably extracted the meat down at
the beach and did not take the shells up to the
site.

The mountain or black crab (Gecarcinus
ruricola) dominates the faunal assemblage. The
mountain crab is a frugivorous species, foraging
nocturnally and hiding under debris in the forest
and rocks during daytime. This species occurs
at high elevations on the island but is
endangered today through modern-day hunting.
Generally, they are harvested during breeding
migrations when they migrate to the seashore to
spawn. The second species of crab present in
the assemblage, though in minor quantities, is
the soldier crab (Coenobita clypeatus), a purple-
clawed hermit. This species, however, could
have been foraging on the refuse deposits and as
such be no part of the Amerinidan diet.

Next to the land crab species, there is also
evidence of a reliance on birds. Most, if not all,
of the recovered bird bones belong to the
Audubon's shearwater (Puffinus lherminieri
lherminierii), a breeding visitor to Saba (Figure
7). The bird breeds on the island during a
limited period of the year (i.e., between February
and July). The shearwater, or 'wedrego' as it is
called today, is a small, long-tailed pelagic bird,
approximately 30 cm in length. The bird is dark
blackish-brown above and white below, but with
dark undertail coverts. Today, they are found at
night in nesting colonies or during the day far
out to sea over deep water. They return to land
only for the breeding season. The adults will
leave the nests at nighttime, which is the only
time that they can be found throughout the
island. Nests are located in cliff crevices, caves,
under vegetation, under rocks, or the birds will
dig a burrow 60-90 cm in length. The lays
consist of one white egg.3
3 It is estimated that there are 3000 to 5000 pairs in the
West Indies today of which a large number, around 1000
pairs, come to breed on Saba (EPIC 2003).
Journal of Caribbean Archaeology 4, 2003


Hofman and Hoogland
Potential nesting habitats can be found
throughout Saba. These birds are known to
breed in sheltered places in the higher parts of
the island (Voous 1955). The restricted period
during which the Audubon's shearwater breeds
on the island might point to a seasonal
occupation of the site related to the breeding
season of these birds.

Tools are made of flint, other stones and
shell. The functional analysis of the different
artifact categories is being carried out in
cooperation with Dr. A. van Gijn of the Artifact
Laboratory of the Faculty of Archaeology at
Leiden University.

The flint artifacts are being studied by
graduate student Iris Briels. The assemblage
consists of more than 700 pieces of flint, of
which the majority has a Long Island, Antigua,
provenance (Knippenberg, pers. comm.). The
majority of the flint assemblage is characterized


Figure 7. Audubon's shearwater (P "'
lherminieri) (top) and its archaeological remains.









Plum Piece
by unretouched whole flakes (with strik
platform, bulb of percussion and an inta
end) of irregular sizes and most often w
cortex (Figure 8). Some of the flakes a3
and have a blade-like appearance. Prep,
blade cores are absent. There is no evid
blade production at the site. However, tl
quantity of waste-material in the form oi
small flakes and flake cores points to an
expedient flake technology.

Freehand percussion was used for tl
flaking. The bipolar technique used for
knapping at Ceramic sites on Saba was I
employed by the Archaic people. Flake
production seems to be better
controlled than during ceramic times.
Currently a use-wear analysis
program is being carried out in order
to identify traces of use and define
the possible function of the flint
artifacts. Experiments have been
carried out on wood, fish, fibers,
grasses, reed, calabash and cultigens.

Other stone artifacts include tools
for hammering, pounding, pecking
and grinding made of volcanic rocks
locally available along the shoreline
below Plum Piece (Figure 9).
Twenty Strombus gigas artifacts
were recovered from the midden
deposits, including about twenty
ground shell adzes (Figure 10a-c).
All specimens have a length of
between 15.0 and 18.9 cm. The tools
have either rounded or squarish
edges. A number of blanks and lip
fragments were recovered as well,
while other parts of the shell and
whole conch shells were absent from
the site.

Plum Piece in a regional context of
the Lesser Antilles

The preceramic Age is still not
very well known for the Lesser
Antilles and the virtual absence of I
sites from this period in the southern l
part of the Antillean archipelago still
Journal of Caribbean Archaeology 4, 2003


Hofman and Hoogland
poses many questions regarding the origin of
these peoples. Several preceramic sites are
known, however, from the Leeward Islands and
can be dated to the Archaic Age (i.e., between
2000 and 400 BC). From north to south these
are Krum Bay, St. Thomas (Figueredo 1974;
Gross 1976; Lundberg 1989); Whitehead's
Bluff, Anguilla (Crock et al. 1995), Norman
Estate (Knippenberg 1999; Nokkert et al. 1995)
and Baie Orientale (Bonnissent et al. 2001;
Serrand 2001), St. Martin, Corre Corre Bay, St.
Eustatius (Versteeg, pers. comm.), Sugar
Factory Pier, St. Kitts (Armstrong 1978, 1980;
Goodwin 1978), Hitchman's shell heap on Nevis
(Wilson 1991), several sites on Barbuda


I


Figure 8. Flint flakes (the upper right specimen is 67.5 mm in
ength). The flint was transported to the site from Long Island,
Antigua.










Plum Piece


r















K'
a





'I


Hofman and Hoogland


)
t






I
I



-j.

I



I

-4


- F


*,


f


p*
'. '. *

r ,
^ ,.,M-rr.*


ft

'; p

rV -


C- $~


Figure 9. Ground-stone artifacts (lower right
specimen is 144 mm wide).


Journal of Caribbean Archaeology 4, 2003


Figure 10. Ground shell adzes (Strombus gigas)
(bottom specimen is 179 mm in length).


~4Y
4k









Plum Piece
(Watters et al. 1992;) and Antigua (Davis 1974,
1982, 1993, 2000; Nodine 1990; Olsen 1976)
and Pointe des Pies, Guadeloupe (Richard
1994).

At this point Antigua appears to have been
the most densely occupied island in the northern
Lesser Antilles during preceramic times known
until now. The Archaic settlements are all
situated on the low-lying limestone plain along
the northeast coast of the island (Davis 2000)
and on Long Island, a tiny island situated just
off this coast (Gijn 1993; Knippenberg 1995,
1999, 2001; Verpoorte 1993). The site of Jolly
Beach, which is located on Antigua's west coast,
is an exception. Marine food resources and
easy access to the flint sources situated along
the northeastern shore and on Long Island (i.e.,
Flinty Bay) probably determined preceramic site
locations on Antigua (Davis 2000:93-94).

All of the above-mentioned sites are situated
in coastal environments and rely heavily on fish
and shellfish exploitation. Like Krum Bay,
Norman Estate also shows high percentages of
fish from reef habitats (Reitz 1989; Brokke
1999; Nokkert 1999; Nokkert et al. 1995).
Dependency on a restricted number of species is
a common feature of preceramic sites (Brokke
1999; Crock et al. 1995; Lundberg 1989). The
variability of species between sites might be
related to local availabilities (Keegan 1994:270),
focused collecting strategies and site function
(Lundberg 1991:74). According to Lundberg
(1991:74) the location of the Krum Bay site in a
small sheltered bay within good reach of fishing
grounds and pearl oyster beds can be related to
a focused collection strategy towards certain
shellfish species and reef fishes and also to the
exploitation of pearls during successive
reoccupations of the site spanning more than a
thousand years.

Archaic subsistence on Antigua was oriented
toward shallow marine resources (i.e.,
mangroves, shallow muddy and sandy bottoms,
and shallow rocky areas) (Davis 2000:91, 101).
At Jolly Beach, shellfish exploitation was
focused on about eight major species with
smaller quantities of other species. In addition
to fish from shallow marine waters and some


Hofman and Hoogland
turtle and manatee, Jolly Beach is the only site
presenting a higher reliance on terrestrial fauna
(i.e., lizards and birds) compared to other sites
on Antigua's shore. The total absence of land
crabs from the deposits at Jolly Beach is
remarkable, for they were a common food
source in all Ceramic sites on Antigua (Davis
2000).

Archaic northern Lesser Antillean
assemblages comprise a combination of flint,
ground-stone and shell work. The lithic
technology can be considered poor and often
based on flake instead of blade production. At
Krum Bay, no flint was found, but fine-grained
rock to produce flakes in a nonsystematic
manner was recovered (Lundberg 1989). The
Whitehead's Bluff site produced flint flakes
(Crock et al. 1995), as did the site of Norman
Estate. At both sites there is a total absence of
blade production. The site of Jolly Beach on
Antigua, where blade production predominates,
forms an exception in the region. The existence
of blade production at this site has been related
to easy access to the sources and abundance of
raw material on the island (Knippenberg 1995).
In contrast, the dominance of flake technology,
on islands distant from Long Island has been
related to the fact that they were located far from
the source area (Crock et al. 1995; Knippenberg
1995).

The combination of flint, ground-stone and
shell tools in the aforementioned sites fits the
general picture of Archaic peoples of the
Antilles defined by Rouse (1986, 1992) as the
Ortoiroid series. Veloz Maggiolo (1976, 1980,
1991; Veloz Maggiolo and Vega 1982)
describes these assemblages as a technological
system that comprises a hybridization of flint,
stone and shell technologies based on the
Greater Antillean development of a particular
economic adaptation. The Archaic occupants of
the Lesser Antilles are regarded as fishers and
foragers living semi-permanently in mostly
coastal settlements and lacking pottery (Boomert
2000; Davis 2000; Keegan 1994). They are
generally thought to have populated the islands
from the south (i.e., coastal Venezuela through
Trinidad and Tobago as far as eastern
Hispaniola). The Ortoiroid radiation into the


Journal of Caribbean Archaeology 4, 2003









Plum Piece
Lesser Antilles may have occurred from 5000-
4000 BC onwards (Boomert 2000:78). In
contrast, the western Greater Antillean
preceramic cultures of the Casimiroid series
were believed to have originated in a migration
from Belize also around the fifth millennium BC
(Callaghan 1990; Rouse 1992; Wilson et al.
1998). In a more recent publication Callaghan
(2003) used computer simulations to explore the
probable origins of the earliest preceramic
cultures of the Greater Antilles and to get insight
in the level of navigational skills, which are
necessary for the colonization of the islands
from South America, Central America and
Florida. He concluded that navigation from
northern Central America seems to require
foreknowledge of the islands and from Florida
the early navigators would have encountered
considerable risk. Colonization from the
mainland of South America seems to be most
likely because of the lower degree of
navigational skills involved. In addition, despite
certain resemblances in manufacturing
technology with northern Central America, many
aspects of Belizean assemblages are not found
in the Greater Antilles. The likelihood of a
South American connection is also suggested by
the results of ancient DNA analysis on skeletal
samples from Cuban Ciboney (Lalueza-Fox et
al. 2003).

Similarities between Lesser and Greater
Antillean complexes and their lithic industries,
have drawn the attention of various scholars
(Davis 1974; Fuebles Duefias 1991). As such,
it was found that the technology of the flaked
stone industry of the Jolly Beach complex on
Antigua and its related artifacts are generally
similar to those found in the contemporary
preceramic sites of Barrera-Mordan and Cayo
Redondo in the Greater Antilles (Davis
2000:99). Similarly, shell vessels found at
Whitehead's Bluff on Anguilla are correlated
with the Casimiroid complex as well (Crock et
al. 1995). However, at this point too few reliable
comparative data are available from both areas to
draw firm conclusions about a common origin,
their relatedness and inter-regional interaction
patterns (see Davis 2000).


Hofman and Hoogland
Discussion and future research

The location of Plum Piece at a considerable
height in the mountainous tropical forest region
of Saba impedes comparisons with
contemporary settlements in the Lesser Antillean
archipelago that are all situated on the coast.
For islands to the south of Antigua little
information is available on Archaic settlements
in similar remote inland locations. The majority
of the finds from islands to the south include
isolated ground-stone artifacts, which may
reflect an ample presence of Archaic peoples in
the Windward Islands that are not as yet
pinpointed (Clerc 1976; Harris 1973; Keegan
1994:267; Sutty 1991). Two small sites,
Boutbois and Le Godinot on Martinique are
probably the only readily confirmed Archaic
settlements in the interiors of the Windward
Islands (Allaire and Mattioni 1983). To the
north on the Greater Antilles, however,
occupation of inland site locations such as river
valleys and hills was common during preceramic
times (Kozlowski 1980).

Plum Piece's faunal assemblage points to a
temporal and seasonal occupation possibly
oriented toward a set of special activities. The
size of the actual habitation area cannot as yet be
exactly determined, nor are there any precise
indications of the occupation length. The depth
of the deposits, however, suggests a recurrent
occupation of the site by a single group. The
composition of the midden deposits is indicative
of a habitation by people focusing on terrestrial
food sources. Plum Piece stands out because of
the presence of huge quantities of crab remains
and bird bones, thus far lacking from other sites
in the region.

The large quantities of bones of Audubon's
shearwater (Puffinus iherminieri lherminierii),
which breeds on Saba only between February
and July, suggest that this was the season that
Plum Piece was occupied. The Gecarcinus
ruricola and probably also the Puffinus
lherminieri typically are caught at night when
both species leave their nests and forage around
the island. According to Goodwin (1978, 1979),
landcrabs must have been an abundant protein
source available at negligible risk and requiring


Journal of Caribbean Archaeology 4, 2003









Plum Piece
little energy to capture (see also Davis 2000:90).
Goodwin bases his hypothesis on the masses of
landcrab remains found at Early Ceramic Age
sites on St. Kitts.

The low numbers of fish bones and the
nearly total absence of shellfish in the Plum
Piece deposits is striking and indicates that little
effort was put into the exploitation of marine
sources and transport of fish and shellfish from
the seashore to the mountainous camp site. This
may be indicative of the fact that Plum Piece was
occupied only seasonally alternately and
complementarily with settlements in coastal
locations on one of the nearby islands. The lack
of a well-rounded suite of subsistence remains
in Archaic coastal sites has been previously
related to the fact that people might have moved
between resource concentrations during
different seasons (Lundberg 1991:76; Keegan
1994:270). Plum Piece, although differentiated
by its location, material remains and subsistence
strategies, may have belonged to one and the
same subsistence/settlement system as these
contemporary coastal sites (see also Lundberg
1991:75) and this would at the same time
emphasize the highly diversified procurement
strategies of the Archaic Amerindians (Boomert
2000:78).

The sizeable number of mortars and pestles
available from the midden deposits suggests that
the site occupants also placed an emphasis on
the processing of seeds (and berries) to
supplement their diet. Grasses and other plants
providing edible seeds and fruits of all kinds but
also fibers for the manufacture of baskets, mats
and fish pots are present in the settlement's
environment. It has also been suggested that
certain plant species might already have been
cultivated in so-called house gardens by Archaic
people in the Antilles (Davis 2000:96; Newsom
1993; Newsom and Pearsall 2002). The
preceramic peoples have long been regarded as
pre-agricultural foragers, but paleobotanical
studies have demonstrated that wild grains and
fruits were already 'managed' or even cultivated
during this period (Newsom 1993), suggesting
that a Caribbean Horticultural Complex may
have existed during the Archaic (Keegan
1994:270).
Journal of Caribbean Archaeology 4, 2003


Hofman and Hoogland
The lower slopes situated below Plum Piece
are a suitable habitat for grasses and fruit trees
such as papaya (Carica papaya), soursop
(Annona muricata) and sweetsop (Annona
reticulata), arrowroot (Maranta arundinacea)
and tuna cactus (Opuntia spp.). Keegan
(1994:270) provided a list of plants identified in
other Archaic deposits in the Caribbean
including zamia or coontie (Zamia debilis) and
cupey (Clusea rosea, seeds of the Sterculia,
sapodilla (Manilkara [zapota] sp.), wild
avocado (Persea americana) and yellow sapote
(Pouteria campechiana), primerose (Oenothera
sp.), mastic-bully (Mastichodendron
foetidissimum), trianthema (Trianthema
protulaca), and palms (Palmae) (see also Davis
2000:93; Newsom 1993; Pearsall 1989; Rouse
and Alegrfa 1990; Veloz Maggiolo and Rimoli
1976; Veloz Maggiolo and Vega 1982).

The volcanic and tropical soils in the area of
Plum Piece also provide extremely suitable
conditions for the growing of root crop
cultigens. One might think of, among others,
bitter and sweet manioc (Manihot esculenta and
utilissima), sweet potato (Ipomoea batatas),
yam (Dioscorea alata) and tannia (Xanthosoma
spp.).

Grinding stones and mortars for plant
processing as well as ground-stone tools for
other activities as hammering and pecking are
made from volcanic stones which had been
carried up the mountain from the seashore
where they are found in large quantities on the
boulder beaches. The Well's Bay is located
below Plum Piece and is characterized by such a
boulder beach. Besides the fact that this might
have been the landing place for the Plum Piece
occupants, trips to the bay were probably made
to procure these rocks but also to obtain conch
to eat and for making adzes. Conch shells can
still be found on the sea grasses situated to the
north of Well's Bay. The absence of whole
conch shells from the midden and the presence
of unfinished lips and waste-products from the
manufacturing process suggest that the meat
was extracted and the shells were pre-worked at
the shore. The lips then were taken to the site
for further manufacturing into tools. The large
in situ boulder with grinding platform might be









Plum Piece
indicative of a shell tool atelier on the site.

With regard to the type of artifacts recovered
and the site's location, woodworking for the
building of dugout canoes might be one of the
activities for which people would have chosen to
temporarily settle in this mountainous tropical
environment. It has been suggested that similar
activities would have taken place at the inland
sites of Martinique (see Boomert 2000:78). At
those sites no food refuse was found,
suggesting that they were visited only on a daily
basis.

The flaked-stone industry, characterized by
the presence of unretouched flakes at the site
does not differ from most neighboring islands.
Only the Jolly Beach complex on Antigua is
oriented towards a direct-percussion blade
industry along with flaking. The flaked flint
artifacts from Plum Piece are predominantly
made of Long Island flint. The nearly total lack
of cortex suggests the direct procurement of the
flint material and primary reduction at the
source, in this case Long Island. Prepared cores
were then transported to Saba where further
reduction took place at the site.

The picture that emerges from the
investigations at Plum Piece is that of a campsite
occupied for the performance of specific
activities (i.e., woodworking for the building of
canoes during the spring season, February to
July). This period coincides with the Audubon's
shearwater nesting season and the spawning of
the Gecaricinus ruricola. Procuring the
shearwater and land crabs would have required a
low energy input. The diet of the preceramic
inhabitants of Saba was supplemented by root
crops, fruits and grass seeds. The toolkit
comprising ground-stone and shell tools and
flaked flint artifacts, is at first glance not
different from that described for contemporary
sites in the region. None of the materials used
for the manufacture of these tools are available
in the site's environment and therefore must have
been carried up the mountain from the shore.
The procurement, and in some cases
preworking, of the raw materials elsewhere on
the island (i.e., stone from the shoreline and
shell from sea grass beds) and in the region (i.e.,
Journal of Caribbean Archaeology 4, 2003


Hofman and Hoogland
flint from Long Island) gives the impression that
the Plum Piece dwellers consciously selected
this remote location.

Future research on the site will be directed
toward questions of settlement structure, a
thorough analysis of the toolkit and study of the
technological system. An in-depth study of the
lithic technology and use-wear analysis on the
stone and shell tools is expected to yield
insights into the Archaic tool industry (including
full sequences of core reduction) and
functions) of the chipped flint and ground-
stone and shell artifacts.

Attention will also be directed toward the
identification of plant remains for the
reconstruction of a vegetable component in the
diet, which may have formed an important
supplement to the protein based diet heavily
reliant on crab and bird. Besides further
excavation of the midden deposits for these
purposes, future fieldwork will focus on the
potential presence of features such as pits and
hearths in the area next to the midden. A field
survey of the entire northern side of island is
planned in order to detect and identify possible
features and sites related to the Archaic
occupation at Plum Piece. Primary attention will
be given to plots of land cultivated in the recent
past and indications of land clearing activities,
which entailed the introduction of secondary
growth vegetation in an otherwise nearly
impenetrable rain forest.


Acknowledgements. The authors are very much
indebted to William Keegan and John Crock for their
remarks and useful editorial comments on various issues
mentioned in this paper. William Keegan, Alistair
Bright and Hylke de Jong are thanked for their
corrections of the original English text. Sebastiaan
Knippenberg and Alexander Verpoorte are acknowledged

for their comments regarding the flint tc ii n l '* Loe
Jacobs (Institute of Pottery T. nli l .. i, IAT) has carried
out the analysis of the ochre fragments. The maps in
this paper were made by Medy Oberendorff, the drawings
of the stone artefacts were made by students Iris Briels,
Tessa Muilerman, Leonardo Pedrosa and Justine Wintjes.
These drawings were finalized by Eric van Driel, who is
also responsible for the drawings of the shell artifacts.










Plum Piece
References Cited

Allaire, L., and M. Mattioni
1983 Boutbois et le Godinot: deux gisements
ac6ramiques de la Martinique. Proceedings of the
IXth International Congress for the Study of Pre-
Columbian Cultures of the Lesser Antilles (Santo
Domingo 1981), 27-38.
Armstrong, D.V.
1978 Archaic shellfish Gatherers of St. Kitts,
Leeward Islands: A case study in subsistence and
settlement patterns. MA thesis. University of
California, Los Angeles.
1980 Shellfish gatherers of St. Kitts: A study of
Archaic subsistence and settlement patterns.
Proceedings of the VIIIth International Congress
for the Study of the Pre-Columbian Cultures of the
Lesser Antilles (St. Kitts 1979), Arizona State
University Anthropological Papers, No. 22, 152-
167.
Augustinus, P.G.E.F., R.P.R. Mees and M. Prins
1985 Biotic and Abiotic Components of the
Landscapes of Saba ( . lands Antilles).
Uitgave Natuurwetenschappelijke Studiekring voor
Suriname en de Nederlandse Antillen 115, Utrecht.
Bonnissent, D, R. Bertran, A. Chancerel and Th.
Romon.
2001 Le Gisement pr6ceramique de la Baie Orientale a
Saint-Martin (Petites Antilles), r6sultats
pr61iminaires. Proceedings of the XIXth
International Congress for Caribbean Archaeology
(Aruba 2001), vol. 1, 78-88. Publications of the
Archaeological Museum Aruba 9.
Boomert, A.
2000 Trinidad, Tobago and the lower Orinoco
interaction sphere. An Archaeological/
ethnohistorical study. Ph.D. disseration, University
of Leiden. Plantijn Casparie, Heerhugowaard.
Brokke, A.
1999 Norman Estate: shell. In C.L. Hofman and
M.L.P. Hoogland (eds), Archaeological
investigations on St. Martin (Lesser Antilles). The
sites of Norman Estate, Anse des Peres and Hope
Estate. With a contribution to the 'La Hueca
problem', 47-50. Archaeological Series Leiden
University 4, Leiden.
Callaghan, R.
1990 Possible preceramic connections between
Central America and the Greater Antilles.
Proceedings of the I th International Congress for
Caribbean Archaeology (Puerto Rico 1985), 64-


Hofman and Hoogland
71. Fundaci6n Arqueol6gica, Antropol6gica e
Hist6rica de Puerto Rico, San Juan.
2003 Comments on the Mainland Origins of the
Preceramic Cultures of the Greater Antilles. Latin
American Antiquity 14(3):323-338.
Clerc, E.
1976 Possibilit6 d'un peuplement pr6ceramique en
Guadeloupe. Proceedings of the VIth International
Congress for the study of Pre-Columbian Cultures
of the Lesser Antilles (Guadeloupe 1975), 44-45,
Gainesville, Florida.
Crock, J.G., J.B. Petersen and N. Douglas
1995 Preceramic Anguilla: A View from the
Whitehead's Bluff Site. Proceedings of the XVth
International Congress of Caribbean Archaeology
(Puerto Rico 1993), 283-294. Centro de Estudios
Avanzados de Puerto Rico y el Caribe, San Juan,
Puerto Rico.
Davis, D.
1974 Some notes concerning the Archaic occupation
on Antigua. Proceedings of the Vth International
Congress for the Study of Pre-Columbian Cultures
of the Lesser Antilles (Antigua 1973), 65-71,
Gainesville, Florida.
1982 Archaic settlement and resource exploitation in
the Lesser Antilles: preliminary information from
Antigua. Caribbean Journal of Science 17(1-
4):107-121.
1993 Archaic blade production on Antigua, West
Indies. American Antiquity 58(4):688-697.
2000 Jolly Beach and the preceramic occupation of
Antigua, West Indies. Yale University Publications
in A\il|lip.__.lN 84, New Haven.
EPIC (Environmental protection in the Caribbean)
2003 Environmental protection in the Caribbean,
http://www.epicislands.org/index.htm. Viewed 16
June 2003
Figueredo, A.E.
1974 The archaic period of St. Thomas, Virgin
Islands: new evidence and interpretations. Paper
presented at the 39th International Congress of the
Society for American Archaeology (April 1974),
Washington, DC.
Fuebles Duefias, J.
1991 Estudio comparative de las industries de la
piedra taillado de Aguas Verdes (Baracoa) y Playitas
(Matanzas): probable relacion de estas industries
con otras del SE del Estado Unidos. In M.A.
Rodriguez (ed.), Arquelogia de Cuba y de Ottras
areas Antillanos, 380-462, Editorial Academia,
Havana.


Journal of Caribbean Archaeology 4, 2003










Plum Piece
Gijn, A.L. van
1993 Flint Exploitation on Long Island, Antigua,
West Indies. In C. Bakels (ed.), The end of our
third decade: Papers written on the occasion of
the -' anniversary of the Institute of Prehistory
(II), 183-197. Analecta Praehistorica Leidensia 26,
Leiden.
Goodwin, R.C.
1978 The Lesser Antilles archaic: new data from St.
Kitts. Journal of the Virgin Islands Archaeological
Society 5, 6-16.
1979 The Prehistoric Cultural Ecology of St. Kitts,
West Indies: A Case Study in Island Archaeology.
Ph.D. dissertation, Arizona State University.
University Microfilms, Ann Arbor.
Gross, J. M.
1976 The archaic period of the Virgin Islands: new
investigations. Proceedings of the VIth
International Congress for the Study of Pre-
Columbian Cultures of the Lesser Antilles
(Guadeloupe 1975), 232-238, Gainesville, Florida.
Harris, P. O'B.
1973 Preliminary report on Banwari Trace, a
preceramic site on Trinidad. Proceedings of the
IVth International Congress for the Study of Pre-
Columbian Cultures of the Lesser Antilles (St.
Lucia 1971), 115-125, Gainesville, Florida.
Haviser, J. B.
1985 An Archaeological Survey of Saba, Netherlands
Antilles. Phase I. Reports of the Archaeological
and Anthropological Institute of the Netherlands
Antilles No. 3, Curacao.
Hofman, C.L.
1993 In Search of the Native Population of Pre-
Columbian Saba (400 1450 A.D.), Part One:
Pottery Styles and their Interpretations. Ph.D.
dissertation, Leiden University, Leiden.
Hoogland, M.L.P.
1996 In search of the Native Population of pre-
Columbian Saba(400 1450 A.D.), Part two:
Natural and social environment. Ph.D. dissertation,
Leiden University, Leiden.
Hoogland, M.L.P., and C.L. Hofman
1993 Kelbey's Ridge 2. A 14th century Taino
settlement on Saba, Netherlands Antilles. In C.
Bakels (ed.), The end of our third decade: Papers
written on the occasion of the, .' anniversary of
the Institute of Prehistory (II), Analecta
Praehistorica Leidensia 26, 163-181.
1999 Expansion of the Taino Cacicazgos towards the
Lesser Antilles. The case of Saba, N.A. Journal


Hofman and Hoogland
de la Sociedt des Americanistes 85:93-113.
Josselin de Jong, J.P.B. de
1947 Archaeological material from Saba and
St.Eustatius, Lesser Antilles. Mededelingen voor
het Rijksmuseum voor Volkenkunde, No. 1,
Leiden.
Keegan, W.F.
1994 West Indian archaeology. 1. Overview and
foragers. Journal of Archaeological Research
2:255-284.
Knippenberg. S.
1995 Provenance of Flint and Chert in the Leeward
Region, West Indies. Proceedings of the XVIth
International Congress for Caribbean Archaeology
(Guadeloupe 1995), vol. 2, 261-271. Conseil
Regional de la Guadeloupe, Basse-Terre,
Guadeloupe.
1999 Norman Estate: Lithics. In C.L. Hofman and
M.L.P. Hoogland (eds), Archaeological
investigations on St. Martin (Lesser Antilles). The
sites of Norman Estate, Anse des Peres, and Hope
Estate. With a contribution to the 'La Hueca
problem', 35-46. Archaeological Studies Leiden
University 4, Leiden.
2001 Flint collecting strategies on Long Island.
Proceedings of the XIXth International Congress
for Caribbean Archaeology (Aruba 2001), vol. 2,
88-98, Publications of the Archaeological Museum
Aruba 9, Oranjestad, Aruba.
Kozlowski, J.K.
1980 In search of the Caribbean. Uniwersitetu
Jagiellonskiego, Krakow.
Lalueza-Fox, C., M.T.P. Gilbert, A.J. Martinez-
Fuentes, F. Calafell and J. Bertranpetit
2003 Mitochondrial DNA from pre-Columbian
Ciboneys from Cuba and the prehistoric
colonization of the Caribbean. American Journal
for Physical Anthropology, 121(2):97-108.
Lundberg, E.
1989 Preceramic procurement patterns at Krum Bay,
Virgin Islands. Ph.D. dissertation, University of
Illinois, Urbana-Champagne. University
microfilms, Ann Arbor, Michigan.
1991 Interrelationships among preceramic complexes
of Puerto Rico and the Virgin Islands. Proceedings
of the XIIIth International Congress for Caribbean
Archaeology (Curacao 1989), 73-85. Reports of
the Archaeological-Anthropological Institute of the
Netherlands Antilles No. 9, Willemstad, Curacao.
Newsom, L. A.
1993 Native West Indian Plant Use. Ph.D.


Journal of Caribbean Archaeology 4, 2003










Plum Piece
dissertation, Department of Ailit. p'i l 4, N,
University of Florida, Gainesville.
Newsom L.A., and D.M. Pearsall
2002 Temporal and spatial trends indicated by a
survey of archaeobotanical data from the Caribbean
Islands. In P. Minnis (ed.), People and plants in
Ancient North America. Smithsonian Institution
press, Washington D.C.
Nodine, B.K.
1990 Aceramic interactions in the Lesser Antilles:
evidence from Antigua, West Indies. Paper presented
at the 55th Annual Meeting of the Society for
American Archaeology, Las Vegas, Nevada.
Nokkert, M.
1999 Norman Estate: Faunal remains. In C.L.
Hofman and M.L.P. Hoogland (eds),
Archaeological investigations on St. Martin
(Lesser Antilles). The sites of Norman Estate,
Anse des Peres and Hope Estate.With a
contribution to the 'La Hueca problem', 51-60.
Archaeological Series Leiden University 4, Leiden.
Nokkert, M., A. Brokke, S. Knippenberg and T.
Hamburg
1995 An archaic occupation at Norman Estate, St.
Martin. Proceedings of the XVIth International
Congress for Caribbean Archaeology (Guadeloupe
1995), 333-351. Conseil R6gional de la
Guadeloupe, Basse-Terre, Guadeloupe.
Olsen, F.
1976 Preceramic findings in Antigua. Proceedings
of the first Puerto Rican symposium in Archaeology
(San Juan 1974), Fundacion Archeol6gica,
Antropol6gica y Hist6rica de Puerto Rico, 85-94.
Pearsall, D.M.
1989 Plant utilization at the Krum Bay site, St.
Thomas USVI. In E.R. Lundberg, Preceramic
procurement patterns at Krum Bay, Virgin Islands,
290-361. PhD disseration, University of Illinois,
Urbana-Champagne. University microfilms, Ann
Arbor, Michigan.
Reitz, E.J.
1989 Vertebrate fauna from Krum Bay, St. Thomas,
Virgin Islands. Appendix B in E.R. Lundberg,
Preceramic procurement patterns at Krum Bay,
Virgin Islands, 274-289. PhD disseration,
University of Illinois, Urbana-Champagne.
University microfilms, Ann Arbor, Michigan.
Richard, G.
1994 Premier indice d'une occupation pr6c6ramique en
Guadeloupe continental. Journal de la Socidte des
Americanistes 80:241-242.


Hofman and Hoogland
Roobol M.J., and A.L. Smith
1980 Archaeological implications of some
radiocarbon dating on Saba and St. Kitts.
Proceedings of the VIIIth International Congress
for the Study of the Pre-Columbian Cultures of the
Lesser Antilles (St. Kitts 1979), 168-176.
Arizona State University Anthropological
Research Papers, No. 22, Tempe.
Rouse, I.
1986 Migrations in Prehistory: I Population
Movement from Cultural Remains. Yale
University Press, New Haven.
1992 The Tainos: Rise and Decline of the People
who Greeted Columbus. Yale University Press,
New Haven.
Rouse, I.B., and R.E. Alegrfa
1990 Excavations at Maria de la Cruz Cave and
Hacienda Grande Village Site, Loiza, Puerto Rico.
Yale University Publications in Ailii., 1| l _-. No.
80, New Haven.
Serrand, N.
2001 Les restes coquilliers du site pr6ceramique de
Baie Orientale (Saint-Martin, Petites Antilles du
Nord): consommation de mollusques et production
d'outils en coquille -- r6sultats pr61iminaires.
Proceedings of the XIXth International Congress
for Caribbean Archaeology (Aruba 2001), vol. 1,
89-98. Publications of the Archaeological Museum
Aruba 9.
Stoffers, A.L.
1956 Studies on the flora of Curacao and other
Caribbean Islands, volume 1, The vegetation of the
Netherlands Antilles. Natuurwetenschappelijke
studiekring voor Suriname en de Nederlandse
Antillen, no. 15, Utrecht.
Sutty, L.
1991 Paleoecological formations in the Grenadines of
Grenada and their relationships to preceramic and
ceramic settlements: Carriacou. Proceedings of the
XIIIth International Congress for Caribbean
Archaeology (Curacao 1989), 127-147. Reports of
the Archaeological-Anthropological Institute of the
Netherlands Antilles No. 9, Willemstad, Curacao.
Veloz Maggiolo, M.I.
1976 Medioambiente y adaptacion humana en la
prehistoria de Santo Domingo. Tomo 1, colecci6n
Historia y sociedad no. 24. Editoria de la
Universidad Aut6noma, Santo Domingo.
1980 Las sociedades Arcaicos de Santo Domingo.
Editoria Taler, Santo Domingo.
1991 Panorama hist6rico del Caribe precolombino,


Journal of Caribbean Archaeology 4, 2003










Plum Piece
Banco Central de la Repuiblica Dominicana, Santo
Domingo, Dominican Republic.
Veloz Maggiolo, M.I., and R.O. Rimoli
1976 Estudio preliminary sobre restos alimenticios y
posibles medioambientes en el preceramico de Santo
Domingo. Actas del XLI Congreso
Internatiocional de Americanistas (Mexico 1974),
vol 3, 756-771, Mexico City.
Veloz Maggiolo, M.I., and B. Vega
1982 The Antillean preceramic: A new
approximation. Journal of New World
Archaeology 5(2):33-44.
Verpoorte, A.
1993 Veldverkenning van de vuursteen voorkomens
en werkplaatsen van Long Island, Antigua, West
Indies. MA thesis, Leiden University, Leiden.
Voous, K.H.
1955 The birds of St. Martin, Saba, and St.
Eustatius. Stud. Fauna Curagao 6:1-82.
Watters, D.R., J. Donahue and R. Struckenrath,
1992 Paleoshorelines and the prehistory of Barbuda,
West Indies. In L.L. Johnson (ed.),
Paleoshorelines and prehistory: an investigation
of method, 15-52, CRC press, Boca Raton, Florida.
Wilson, S.M.
1991 The Prehistoric settlement survey of Nevis:
1984-1988. Proceedings of the XIIIth Congress for
Caribbean Archaeology (Curacao 1989), 269-279.
Reports of the Archaeological-Anthropological
Institute of the Netherlands Antilles No. 9,
Willemstad, Curacao.
Wilson S.M., H.B. Iceland and T.R. Hester
1998 Preceramic Connections between Yucatan and
the Caribbean. American Antiquity 9(4):342-352.


Hofman and Hoogland


Journal of Caribbean Archaeology 4, 2003




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs