Coastal subsistence and settlement in the early Ceramic


Material Information

Coastal subsistence and settlement in the early Ceramic a zooarchaeological study from central Pacific Panama
Physical Description:
xi, 204 leaves : ill. ; 29 cm.
Peres, Tanya Michelle, 1973-
Publication Date:


Subjects / Keywords:
Animal remains (Archaeology) -- Panama   ( lcsh )
Kitchen-middens -- Panama   ( lcsh )
Indians of Central America -- Antiquities -- Panama   ( lcsh )
Anthropology thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Anthropology -- UF   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 2001.
Includes bibliographical references (leaves 159-176).
General Note:
General Note:
Statement of Responsibility:
by Tanya Michelle Peres.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 027750788
oclc - 48555447
System ID:

Full Text







Dedicated to my real and fictive kin.


This dissertation could not have been completed without the assistance of many people and the funding and cooperation of several institutions. Funding for this research was provided by Sigma Y; a Short-Term Fellowship from the Smithsonian Tropical Research Institute, Panama. Funding was also provided by several on-campus sources, including the Department of Anthropology, the Latin American Studies Program (through a Field Research Grant), the College of Liberal Arts and Sciences, the Office of Research and Graduate Education, the Graduate Student Council, and a Charles H. Fairbanks Award.

I would like to thank all of the members of my committee. Without their

dedication, understanding, and patience, this work would not have been completed within the time constraints I have been working under. I am grateful to Lynette Norr, as my advisor and friend, she stimulated my interest in Central American Archaeology, and lent a tremendous amount of support to me in all of my endeavors. The fact that she has had so much faith in me and my capabilities, was at times, what kept me going. The project she has undertaken at Zapotal is both exciting and puzzling, and I am glad for the opportunity to have been involved with it.

Richard Cooke has been a trusted friend and advisor. I would never have

undertaken such a project if he had not been available for guidance, suggestions, and his

infamous comparative collections. Richard is a walking library of Latin American prehistory; I am grateful for the opportunity I have had to work with him.

The 1999 Zapotal Field School participants are very much a part of this work. Without their hard work during long, hot days, and sometimes long, snake-infested, nights, much of this assemblage would have not been excavated or analyzed. Their enthusiasm and humor made the '99 field season enjoyable, memorable, and successful.

My colleagues in the Environmental Archaeology Lab at the Florida Museum of Natural History (FLMNH) were a tremendous help during my project. T thank Irvy Quitmyer for his assistance with the Osteichthyes' atlas measurements; Sylvia Scudder for access to the comparative collections; Donna Ruhi for verifying plant names, and all of the folks in Anthropology for their support, encouragement, and interest. John Slapcinsky, of the Malacology Range of the FLMNI*4, helped to verify mollusc identifications, and I thank him for his time and enthusiasm.

To my long-time mentor and friend, Rochelle Marrinan, whose pep talks and words of wisdom have sustained me through the rough periods, I extend my gratitude. Many thanks to my friends and colleagues, Scott Fisher, Liz de Grummond, Greg Heide, Erin Kennedy, Lesley Martin, Wendy Nettles, Sharyn and Pat O'day, Theresa Schober, Samantha Stytzer, and others I have not mentioned by name here. Their emotional support, "crisis counseling," phone calls, and emails kept me motivated and sane. My utmost appreciation (and a few dinners) goes to Greg Heide for producing first-rate maps for this project. John Krigbaum has opened his office and home to me in the last few months of teaching and editing. Without his hospitality, I am not sure where I would be.


My family has been overwhelming in their support and understanding of my

choice of careers, but especially during this time of self-imposed seclusion. I could not have finished without "being left alone," when it was the most needed. All of the encouragement that my siblings have showered upon me has been very appreciated. My grandmother has been the most patient, waiting four years to attend the last of my graduation ceremonies, all along asking how "the book" was coming.

Most of all, I would like to thank my friend, confidante, and partner, John Lemons. Without his understanding, encouragement, and patience, I would not have had the motivation to begin and complete this project. The last few months have been trying for a number of reasons, and without his strength and love I would have been over the brink several times.



L IST O F T A B L E S ............................................................................... viii

L IST O F F IG U R E S .............................................................................. ix




W hy Study Subsistence? ..................................................................... I
Models of Subsistence and Settlement ..................................................... 3
Subsistence and Settlement in Central Pacific Panama .................................. 8


C ultural G eography ........... ................................................................. 13
C entral Pacific Panam a ....... ................................................................ 21
Summary of Central Pacific Panama Prehistory ............................................ 65

3 THE ZAPOTAL SITE (He- 15) .............................................................. 66

Modem and Prehistoric Environments..... '' , '' * ........ *** ...... 66
Zapotal (He- 15) Background Considerations ............................................. 72
T he Study A ssem blage ......................................................................... 75
S u m m ary ........................................................................................ 8 0

4 T H E SA M P L E ................................................................................. 82

The 1999 Faunal Assemblage .......................................................... ...... 82
Methodology of Sample Analysis ............................................................. 83
R esults of Sam ple A nalysis ................................................................... 93
S u m m ary ... .................................................................................... 1 13

5 SAMPLE INTERPRETATION .............................................................. 115

Ecological Zones Adjacent to Zapotal ................................................... 116
Subsistence Practices and Resource-Use ................................................. 120
Regional Patterns of Group Mobility and Settlement ...................... I ............ 132
S u m m ary .................................... ................................................. 13 6


LATIN AMERICA ......................................................... 137

Pacific Latin America in the Archaic (7,000 4,500 B.P.) ....................... 139
Pacific Latin America in the Formative (5,500 3,000 B.P.) ... ................. 144
Trends in Subsistence Economies .............................................. 148


Environment and Resource-Use................................................151
Settlement Patterns in Central Pacific Panama...................................152
Pacific Latin America and Resource-Use........................................ 154



AS SEMBLAGE PRIMARY DATA .........................................178

ASSEMBLAGE SECONDARY DATA ..................................... 191




Table page
2.1 Cultural Chronology of Central Pacific Panama and Associated
Archaeological Sites............-................................... 24

2.2 Archaeological Research Projects Conducted in the Central Region of Panama
Since 1970.................................................... .......... 25

2.3 Identified Faunal Remains from Preceramic and Ceramic Sites Located in
Central Pacific Panama ......................................... ........ 28

2.4 Radiocarbon Dates from Cerro Mangote.............................. 42

2.5 Radiocarbon Dates from Skeletal Remains at Cerro Mangote ........44 2.6 Radiocarbon Dates from Monagrillo .........................48

2.7 Radiocarbon Dates from Zapotal ........................... 52

3.1 Description of Natural Strata (A-J) from Unit 45, Sondeo ............80

4.1 NISP Per Taxon From Two Screen Sizes ......................92

4.2 Taxa Identified in 1999 from Unit 45, Sondeo ..................76

4.3 Primary Data, Biomass, and MMI for All Taxa, Unit 45 Sondeo,
All Levels..................................... ................. 97

4.4. Composite of Vertebrate and Invertebrate MMI from Unit 45 Sondec,
All Levels........................................................ 81

4.5 Atlas Measurements and Weight Estimates............................... 113

5.1 Animals Identified from Archaeological Samples at Zapotal and
Their Preferred Habitats ......................................... 121



2.1 Map of Lower Central America......................................... 14

2.2 Map of the Intermediate Area ........................................... 16

2.3 Map of central Pacific Panama indicating locations of
Late Preceramic sites ........................................... 27

2.4 Map of central Pacific Panama indicating locations of
early Ceramic-A sites ........................................... 47

2.5 Map of Zapotal site plan indicating site boundaries and location
of ancient coastline............................................... 51

2.6 Map of central Pacific Panama indicating locations of
early Ceramic-B sites ........................................... 58

3.1 Map of central Pacific Panama indicating location of
Zapotal (He-15) ................................................ 67

3.2 Plan view of Zapotal indicating location of Block 47,
Unit 45 sondeo ............................................ .... 76

3.3 Profile map of Unit 45, sondeo, East Wall.............................. 79

6.1 Map of Pacific Latin America.......................................... 138


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy


Tanya Michelle Peres

August 2001

Chair: Lynette Norr
Major Department: Anthropology

The Zapotal Site (He- 15), is a an early Ceramic-A, coastal shell midden located in the Azuero Peninsula of central Pacific Panama. During the 1999 field season, under the direction of Dr. Lynette Norr, crewmembers excavated a 50cm x 50cm sondeo, that captured the cultural sequence deposited at the southern end of the site. It is from this context that the faunal assemblage for this project was recovered. The purpose of this project is to study the relationships between resource-use in a subsistence-based economy, habitat exploitation, procurement technologies, and patterns of group mobility and settlement.

Zapotal is a suitable choice for asking and answering questions related to subsistence because of the excellent preservation and quantity of faunal remains recovered, as well as methods used to recover both vertebrate and invertebrate food


remains. I addressed some specific research questions with the zooarchaeological assemblage from Zapotal are:

1 What ecological niches were the inhabitants of Zapotal using in relationship to
their home base?
2. Was long-range planning involved in choosing specific resources for
3. What do the represented taxa from the archaeological record tell us about
residential mobility and settlement patterns in the area?

The data show that during the deposition of the shell midden, there was an

increase in the exploitation of local resources and deposition of those remains at the site. This conclusion is based on density of the deposits, and the large sample size recovered from the midden period. The people who used Zapotal to extract and process aquatic vertebrates and invertebrates favored species that regularly occurred in the estuaries or in the shallow near-shore waters of Parita Bay. The site's location was advantageous because of its proximity to critical resources like potable water, firewood, and food. The occupations may have been short-term fishing forays by specialized detachments of collectors that belonged to a larger group of agriculturalists that lived at more inland locales. These smaller collecting groups may have fished at the coast during the dry season when the agricultural plots were laying fallow.



This dissertation presents new data on the subsistence strategies of the prehistoric inhabitants of central Pacific Panama. A number of issues related to the human inhabitants of this area are addressed, including the relationship among subsistence strategies and habitat exploitation, patterns of residential mobility, procurement technologies, and seasonal scheduling of resources. The subsistence economy of an archaeological population is reconstructed using environmental archaeological methods, which incorporate zooarchaeological and paleoenvironmental data. Analyses of faunal remains from archaeological sites allow for an informed discussion of a group's approach to the food quest. Knowing the subsistence base allows for further exploration of anthropological questions crucial to the study of culture change, human ecology, and human adaptability.

Why Study Subsistence?

Knowledge of a group's subsistence is key in understanding the relationships between people and their environments; the technologies they create and use to exploit and modify their environments; and relationships among the people themselves. Different subsistence strategies result in highly variable responses to human/environment interactions and human/human interactions. The archaeological record allows us to know what prehistoric peoples were choosing to eat, but not necessarily why. It is accepted



among archaeologists that humans do not make use of every available edible resource in their environment, but that food choices are made based on cognitive cultural beliefs. These cognitive choices are visible in the archaeological record.

The suite of taxa that are represented in the subsistence record can inform us

about habitat exploitation. What ecological niches are favored, and which are forsaken? Again, not every single available food source will be exploited, thus one habitat may be chosen for exploitation over another. The locations of archaeological sites can aid in interpreting the importance of resources to past human populations. Large sites located immediately adjacent to rivers and estuaries are better positioned to exploit these resources than a site that is located 15 krn from the same habitats. The represented taxa, site location, and duration of occupation can further inform about the scheduling of seasonal resources. Procurement technologies, such as the use of fishing tackle, digging sticks, and storage items, may be inferred from the resources (represented taxa, quantity, and size) that were being exploited.

Populations settling along the coastlines of lower Central America could have potentially exploited a number of resources and habitats, on a gradient from exclusively terrestrial environments to exclusively marine environments. These choices were interdependent on environmental conditions, available technologies, social structure and group size, inter- and intra-group political interactions, and cultural traditions. We consider several research questions of primary importance in coastal Central America What protein resources were being exploited by the inhabitants of lower Central
America between 5,000 B.P. and 2,000 B.P.?
" What ecological niches were being exploited, and with what methods? How did the subsistence base of this area affect residential mobility and settlement


Zooarchaeological remains can aid interpretations of ancient resource choices,

technological adaptations, cultural continuity, and settlement patterns. Thorough studies must use multiple lines of evidence, the basis of environmental archaeology. Through the study of specialized and utilitarian artifact assemblages, site locations and catchment areas, paleoethnobotanical remains, and stable isotope analysis of human skeletal remains, additional information can be obtained to strengthen or alter these interpretations. For instance, the use of stable isotope analysis in bone collagen allows for determination of the origin of the protein resources eaten by an archaeological population. They also give information about continuity and variation in consumed resources through time, between populations, and within a population (Norr 1990; Pate 1992; Schoeninger 1986; Schoeninger and Moore 1992; Schwarcz, 1991; Tieszen 1991; van der Merwe 1982).

Models of Subsistence and Settlement

A number of models in the anthropological literature attempt to explain the

relationship between foraging subsistence practices and settlement patterns. Two models that help to explain these relationships are presented here. While these models were developed several decades ago they are still useful for the present discussion of foraging behavior. One is a model of time and energy conservation based on long-range planning (Jochim 1976, 198 1), and the other is a model for planned, logistically-oriented foraging practices (Binford 1980). We know that the people inhabiting Central America, and Panama specifically were involved in large scale food production by 7,000 B. P. (Piperno


and Pearsall 1998:4). 1 present a summary of the evidence for origins of food production in this area, and these data relate to the area under study.

Long-Range Planning

The first model views human behavior as a result of problem-solving or decisionmaking. The model proposed by Jochim (1976, 198 1) is based on the observed crosscultural regularities in economic goals and behaviors based on a survey of the ethnographic literature of hunter-gatherers (Jochim 1976:xiii). The model draws from general economic and optimal foraging theories, assuming that there were time and energy choices associated with resource use. These choices include the proportional amount of a resource used, time spent procuring/using these resources, and the demographic and spatial patterns chosen to effectively exploit these resources (Jochim 1976:4).

Optimal foraging theory contributes to the idea that people seek resources that increase their efficiency and security, and that these decisions are not rash, but part of long-range planning. Human groups are flexible in their resource choices, such that if an opportunity arises that deviates from the local general pattern, the group will amend their behavior until the short-term opportunity is no longer profitable, at which time they will return to their "normal" pattern of resource exploitation. Under this same model, it is assumed that people strive for security in their food resources. Different subsistence strategies are used until the most successful (time and energy-efficient) are chosen. Faunal data from early Ceramic period archaeological sites in central Pacific Panama are used to determine the extent of long-range planning by human groups.


Settlement Patterns

The Forager-Collector model proposed by Binford (1980) attempts to organize hunter-gatherer groups by resource-use. Foragers have high residential mobility and regular daily food-procurement strategies (Binford 1980:9). Two types of sites are associated with foragers: residential bases and locations for food-gathering purposes. Residential bases are the hub of subsistence activities, while locations are sites where extractive tasks are carried out (Binford 1980:5). Foragers spend their days gathering food items, which are taken back to the residential base every evening. Storage is not practiced, and residential groups move seasonally among resource "patches." Thus they may not be particularly visible in the archaeological record. Foragers are viewed as 44mapping onto" resources through residential moves and adjustments to group size.

The other group outlined by Binford is that of Collectors. These groups are

logistically organized, practice food storage for at least part of the year, move resources to consumers, and have a greater degree of visibility in the archaeological record. Collectors organize themselves on the landscape to maximize critical-resource use. This logistical strategy requires that a residential group positions itself near the critical resource that has the largest bulk demand (i.e., lithic source). From the residential camp, specialized task groups forage in their catchment area and procure other critical resources (which are less bulky, or more easily transported) and bring them back to the consumers. In addition to the residential base, collectors are associated with field camps, stations, and caches (Binford 1980:12). Field camps are the places where food procurement operations are planned and executed, and some degree of food processing may take place to expedite transporting of resources to consumers. Stations are where special-purpose


task groups gather to exchange information on local resources. Caches are used to temporarily store foodstuffs awaiting transport to the consumers at the residential base.

Foragers and Collectors should not be thought of as polar opposites, but as a

graded series from simple to complex (Binford 1980:15). There are infinite combinations of these two types of groups, making these strategies environment- and demographicdependent. For instance, in equatorial environments, while there are marked wet and dry seasons that dictate breeding and fruiting cycles, animals and plants produce year-round, thus there are continuously available food resources. In this type of environment, highlymobile foraging strategies work well. In temperate environments, where marked seasonal cycles dictate the scheduling of breeding patterns in animals, and the growing seasons of plants, a collecting strategy based on semi- or permanent-sedentism, that facilitates the storage of resources is the most useful. This relieves the problem of "temporal incongruity" (Binford 1980:15).

Food Production in Central America

Piperno and Pearsall (1998:4) propose that food production in the tropical

lowlands of Central and northern South America emerged around the same time as it did in the Near East. They believe that the beginnings of food production in the tropics were common in small-scale house gardens adjacent to residential structures by 8,000-9,000 years ago in Panama, Peru, Ecuador, and Colombia (Piperno and Pearsall 1998:4). The plants that were involved in this food production were a mix of "wild" species and those that can be considered "domesticated" (i.e., larger seed sizes and change in phytolith sizes) (Harris 1989; Piperno and Pearsall 1998:4, 8). Food production on a larger-scale is


evident in the archaeological record by 7,000 B. P. House garden plots were abandoned in favor of larger fields located away from the residence. The introduction of cultivated slash-and-burn plots led to deforestation, and is visible in the paleoecological. records (Piperno and Pearsall 1998:4).

1 follow Piperno and Pearsall's (1998:6) definitions of the following terms related to food production:

" cultivation-all human activities involving the preparation of plots used for plant
propagation; the repeated of use (planting and harvesting) of plants in these plots horticulture-activities at one end of an evolutionary continuum; small-scale
cultivation (house gardens); may make use of both "wild" and domesticated plants agriculture-activities at the other end of the evolutionary continuum; larger scale
cultivation and fields; domesticated plants are common and often dominate as the
staple crops

There are several explanations of why food production arose. Among them are:

* food production is a result of human innovation (Sauer 1952; Braidwood 1960)

* changes in climate are the cause for food production (Childe 1952; Wright 1993) population growth and other demographic factors force people to produce food to
support an ever increasing population (Cohen 1977; Smith and Young 1972)

" a Darwinian view that plants and people share a continuous evolutionary mutually
beneficial relationship (Rindos 1984)

* subsistence decisions are made based on relative return rates, and food production has
the highest yield per time or energy invested (Gremillion 1996; Hawkes and
O'Connell 1992; Kaplan and Hill 1992; Keegan 1986; Piperno 1997)

* social factors such as prestige, competitive feasting, and ideology are viewed as the
main mechanism in starting food production (Bender 1978; Hayden 1995; Price


Piperno and Pearsall (1998:16-18) frame their research within a model based on behavioral ecology, optimal foraging theory, and "the diet breadth model." The ideas are similar to Jochim's view of humans as making rational decisions about food choices. In the optimal foraging theory and diet breadth models synthesized by Pipemo and Pearsall (1998:17), humans must continually meet the subsistence needs of the group, and will try various foraging strategies. The strategy or strategies that are the most successful (based on relative return rates) will be repeated. Foods enter the diet because higherranked choices may decline, and foragers spend less time in searching for food by expanding their selection of food items, The reduction in search time results in a broader diet, greater investments in storage and processing of food items, and lead to smaller foraging areas. This broader diet and reduction in time spent searching for food may lead to demographic changes in human populations (i.e., less mobility, more sedentism, and population size).

Overall, food production in the area under study arose earlier than once thought

(by ca. 9,000 years ago), and was a result of changing resources sought after by

human foragers. Food production was a successful human adaptation to the changing

environments of the tropical lowland forests.

Subsistence and Settlement in Central Pacific Panama

Previous research in the central region of Panama indicates that for the time

period under study, populations were living in small settlements scattered in the foothills and along the coast. These populations practiced a mixed economy of plant cultivation, wild plant gathering, hunting, and fishing (Cooke and Ranere 1984). This pattern is


believed to be continuous from 7,000 B.P. to ca. 2,500 B. P., when agriculture focusing on non-indigenous cultigens such as maize, squash, and root crops became the primary economic activity (Piperno, 1989). This abrupt change is believed to have been related to a rapid growth in population, to nucleation, and to larger more permanent settlements (Cooke and Ranere 1992a, b; Piperno 1989; Piperno et al. 1991a, b, 1992;- Piperno and Pearsall 1998).

Three scales of analyses are used in this study. First, a small-scale of analysis is conducted at the local level. This is accomplished by incorporating data from identified and analyzed vertebrate and invertebrate faunal remains from Zapotal to address sitespecific questions relating to subsistence and human-environment interactions. Jochim' s (1976, 1981) long-range planning model is used to determine the settlement choices being made regarding food resources.

The second scale of analysis attempts to understand changes in subsistence and settlement patterns in the Santa Maria River and Parita Bay region. By incorporating the local-level data with paleoethnobotanical and faunal data from other locales within the region, applying Binford's (1980) Forager-Collector model, and using Pipemno and Pearsall's (1998) ideas of tropical food production, inferences can be made about these subsistence-related issues at a larger scale. The regional level includes other sites in central Pacific Panama dating from the Preceramic through Middle Ceramic periods. The third scale of analysis is more geographically encompassing than the regional scale. Studies from other parts of middle and Central America are compiled to present a view of culture change and continuity over a much larger area during the same time periods.


The Case Study-Zapotal

Zapotal is an excellent choice for asking and answering subsistence-related

questions because of the excellent preservation and quantity of faunal remains recovered, and because of the methods used to recover both vertebrate and invertebrate food remains. Specific research questions were addressed with the zooarchaeological assemblage at Zapotal:

1 Was there a discernible shift in subsistence strategies during the time period of

2. What ecological niche(s) were the inhabitants of Zapotal using in relationship to
their home base?

3. Was long-range planning involved in choosing specific resources for

4. What do the represented taxa. from the archaeological record tell us about
residential mobility and site seasonality?

5. What inferences can be made about procurement technologies using middle-range
research and ethnographic studies?

While it is necessary to analyze data at the local (site) level, it is also good

practice to incorporate these data into the regional picture. Therefore a second set of research questions can be answered using multiple lines of evidence from several sites within central Pacific Panama: Was there a diachronic shift in subsistence strategies in this region? If so, what was the shift? Were these shifts, if present, a result of cultural factors (i.e., population expansion) or environmental factors (i.e., sea-level rise/fall), or both? What was the extent of population movement in the region during the time period from 5,000 B.P. to 2,500 B.P.?


This study has an environmental archaeology focus (Dincauze 2000). This means that the study is concerned with the interactions among humans and their environment, the influence that humans had on the environment in which they inhabited, and in turn, the environment's impact on the human experience. Environmental archaeology views humans as living within environments, manipulating and modifying them in the search to fulfill basic human needs. Just as aspects of a particular environment (water sources, species diversity/density) can determine subsistence choices and settlement patterns among humans, these same choices made by people (subsistence, living conditions) can in turn alter the environment (e.g., deforestation and overharvesting of fish and shellfish).

Applying this mindset to the available archaeological data allows for a greater understanding of the interplay among people and the world they lived in. Daily choices of resource exploitation and technological innovation were based on these interactions. The distance to food resources (among other raw materials needed to satisfy the quest for food, shelter, and reproduction) and the availability of these resources (either seasonally or in abundance) are parameters that can be outlined using environmental archaeology methods. Zooarchaeology is a large part of this subdiscipline. The methods used in this study to collect data on archaeological faunal remains are considered standard zooarchaeological practice (Reitz and Wing 1999). Organization of this Study

Chapter 2 presents an overview of Panama: its prehistory and modem

environmental setting, as well as published reconstructions of past environments in the study area. Chapter 3 details all excavations at Zapotal from 1952 to the present. This


chapter also describes the sources of archaeological data used in the site study of Zapotal: provenience, methods and techniques used in excavation, and the justification for choosing this particular assemblage.

The vertebrate and invertebrate faunal sample is described, as are zooarchaeological methods. Results of the analysis are given in Chapter 4. Interpretations of the Zapotal sample, including subsistence, favored resource habitats, procurement technologies, and discernible patterns of group mobility are discussed in Chapter 5. Chapter 6 presents data from central Pacific Panama and other coastal locations in Middle, Central, and South America to further interpret the subsistence and settlement strategies for this time period.

Chapter 7 provides a synthesis of the proceeding chapters, and makes recommendations for future research in this area.


Archaeologists have assigned Panama to several classification schemes. Some have emphasized cultural geography, others linguistics and subsistence economy as the main characteristics for division. Based on these different criteria, Panama has been grouped with different countries in different schemes. The area presently known as the Isthmus of can be characterized by several geographical classification schemes. The term Central America is used to denote the modem countries of Guatemala, El Salvador, Honduras, Nicaragua, Costa Rica, and Panama. Nicaragua, Costa Rica and Panama can be further subdivided into the area known as Lower Central America (Lange and Stone 1984:3; Sheets 1992:17) (Figure 2. 1). Lower Central America has been characterized as being diverse in geography and human cultures, leading to a range of cultural adaptations and patterns. Johnson (1948:43) defines Central America as a cultural region that shares a basic cultural tradition. This "cultural substructure" may have been derived from South American cultures, and influenced by Mesoamerican cultures to the north (Johnson 1948:43). Panama has been included in the prehispanic cultural region known as the Intermediate Area (Willey 197 1).

There is another way of looking at Panama as part of "interaction spheres" or "historical regions" that fink this country's archaeology to northern South America. This




0 500 1000

Kilometers Caribbean Sea

Fiue21 a fLower Central America Gahcb rgHie


is more satisfying because historical linguistics show that modem lower Central American Indians are related to speakers of Chibchan and Chocoan languages from Colombia and Ecuador (Greenberg 1987a, b).

Cooke and Ranere (1992b) offer a summary of data from historical linguistics and human population genetics as a framework in which to view social change. These data support the hypothesis that the present distribution of genetic and language diversity across Panama and lower Central America has roots in a homogenous ancestral population. This founder population became fragmented over time due to internal pressures, and diversified in situ (Cooke and Ranere 1992b:290). These data show that tribal groups were hierarchically organized, with a general east-west pattern of grouping, and the closest affmities were between neighbors (Barrantes et a]. 1990:63).

Willey (1971:61, 69) assigns Panama to the Northwest South American Littoral Tradition (NWSALT), which may have had roots in the Andean Hunting-Collecting Tradition or the East Brazilian Upland Tradition. The NWSALT tradition is an adaptive approach that is based on Pacific littoral resources. Willey believes the tradition began as early as 5,000 BC in some areas, and included coastal peoples from Ecuador, Colombia, Lower Central America, Venezuela and the West Indies, all living in similar environmental settings: tropical forest, mangrove estuaries or savanna-like areas (Willey 1971: 65-66). Archaeological data point to a variety of coastal adaptations within the littoral tradition, but all include fishing and mollusk collecting. The earliest material evidence of the NWSALT hes in the artifact inventory. This littoral tradition is characterized by rough chipped-stone tools (scrapers, choppers, cutting edges, absence of projectile points) and some shell and bone artifacts that were used for cutting/scraping



w E


0 500 1000
Belize M%

Guatemala Caribbean Sea El Salvador

Costa Rica

Pacific Ocean

Colombia The Intermediate Area



Figure 2.2. Map of the Intermediate Area. (Graphic by Greg Heide)


fish, opening shellfish, processing plants, and making wooden implements (Willey 1971:364). Later in time ground stone tools were added to this repertoire, and eventually replaced chipped-stone implements. Willey (1971:66,494) suggests the introduction of pottery (Valdivia tradition) into the NWS ALT marks the transition into the Intermediate Area agricultural tradition, but more recent data indicate agricultural practices predate pottery (Piperno and Pearsall 1998).

The Intermediate Area

Willey (1971:254) originally defines the Intermediate Area as the geographic

area, "between the Mexican-Guatemalan and Peruvian centers." This delineation is truly the intermediate area between Mesoamerica, to the north, and the Andean culture region, to the south. Lange (1992:1-7) presents a review of the definitions of this area, and proposes that the Intermediate Area includes: eastern Honduras, Nicaragua, Costa Rica, Panama, northern Ecuador, Colombia, and western Venezuela (Figure 2.2). The Intermediate Area's perimeters do not follow exact modem political, geographical, topographical, or physiological boundaries. Rather, it is an area that was linked in prehistory by a number of shared cultural traits. According to Lange (1992:4), these regional cultural patterns can be have been identified in the archaeological record as early as 3,500 B.P., and are believed to have continued through Spanish contact.

In contrast to Willey's NWSALT, Stothert (198 5) and others (Cooke 1992c; Ranere and Cooke 1996) argue that central Panama and more southern regions of the Intermediate Area had developed relationships earlier in time. Stothert (1985) suggests that the Vegas Culture of Ecuador and the Late Preceramic site of Cerro Mangote,


Panama, were "historically linked" because of the commonalties between burial practices and stone tool types. She believes these cultures belonged to a Preceramic interaction sphere that covered the Parita Bay region of Panama south to northern Peru. The adaptive strategies of this interaction sphere were not based on a Pacific littoral tradition, but rather they were based on tropical forest lifeways (Cooke and Ranere 1992b; Ranere 1980a, b).

Willey (1971:277-278) outlines twelve features that he sees as prominent markers of the Intermediate Area. The first feature is plant cultivation. The principal crops for this area were manioc and maize. Although they were important to most inhabitants of the area, there were during different times depending on the region under discussion. Modes of plant cultivation included slash-and-burn, (more-permanent) deforestation for agricultural fields, garden plots/house gardens, and irrigation; the last two were used to a lesser degree than the first two.

The second, third, and fourth features address settlements. In this area small

hamlets and villages are seen throughout, but some areas boasted towns and small cities. Ethnohistorical accounts document small independent communities and moderate territorial states. Ceremonial centers were common throughout this area. Some of these centers were located within towns, while others were central to a number of villages.

Burial treatments are the fifth feature that Willey addresses. No single burial

treatment exists that is common or abundant throughout the area. Urn burials and shaft tombs are known for parts of Ecuador and Lower Central America. In later times, funerary objects are viewed by archaeologists as evidence for wealth, prestige, and hierarchy.


Ceramics in the Intermediate Area are high in quality and quantity. Forms that are repeated throughout the area consist of pedestal, annular-based, and tripod-based bowls and plates. Decorative techniques include negative painting and applique decorations. Other pottery forms include figurines, spindle whorls, and musical instruments.

Lithic tools recovered from this area are of both the polished/ground and chipped stone varieties. Polished stone celts and small ornaments are common throughout; chipped stone tools are not common after 8,000 B.P. Bifacial fluted points are found throughout Central America, some as isolated finds, and others from excavated and unexcavated sites (Ranere and Cooke 1991). In early Preceramic times (10,000 7,000 B .P.) archaeological sites in central Panama are identified by the presence of stone tools (Cooke and Ranere 1992b:252). At L4 Mula-West, recovered graving/boring and scraping tools are similar to artifacts from other early human sites in the Americas (Cooke and Ranere 1992b: 253), namely those of the Clovis culture, dated to ca. 11,200 10,900 B.P. in North America (H-aynes 1991).

Similarities in tool types may be due to similarities in subsistence practices among these groups. The people that made bifacial projectile points, along with various other implements, were probably hunting extinct megaherbivores and extant mammals. Several environmental zones were exploited during the Preceraniic period, including forested and open vegetation areas (Cooke 1998).

Recent archaeological evidence from Cerro Juan Diaz (Cooke and S~.nchez 1997; Cooke et al. 2000) indicates that metallurgy appeared on the lower isthmus by AD 200. In the Gran CocI6 culture area, metal artifacts are associated with particular pottery types


through time (i.e., Tonosi, Cubiti). Metallurgical techniques included casting, gilding, soldering, alloying, and "lost-wax". The metals that were worked were gold, silver, copper and alloys of these. The objects that were made from metals consist of figurine pendants in zoomorphic forms (birds and other animals), beads, rings, nose clips, and small hammered discs and plaques (Cooke et al. 2000:171).

Monumental stonework (i.e., slab carvings, column statues) is found throughout the Intermediate Area. While styles vary, one commonality of most of them is the trend towards angular treatment of life forms.

Language is important in defining cultural boundaries, as previously discussed. Willey identifies three principal language families in the Intermediate Area: Chibchan, Paezan, and Macro-Cariban. Wheeler (1972) -identifies a single linguistic phylum, Macro-Chibchan, as the dominant language from Nicaragua through Panama, coastal Colombia, Venezuela and northern Ecuador. Greenberg (1987a, b) classified the native languages of lower Central America and northern South America as "nuclear Chibchan" in the "Chibchan-Paezan" branch or subgroup of Amerind. All three of these sources group native inhabitants of the Intermediate Area into similar linguistic groupings.

Due to the Intermediate Area's proximity to the two well-known and well-studied culture areas, the cultures of this region have often been compared to their northern and southern neighbors. It has been the custom to view the Intermediate Area as a cultural periphery, unlike the Peruvian Andes or Mesoamerica, both which are considered cultural cores (Lange and Stone 1984; Linares 1979:38; Sheets 1992). Cultural characteristics of the Intermediate Area that contrast with those of Mesoamerica and the Andean area


include long-term cultural stability, absence of extensive cultural horizons, and the absence of state-level societies (Lange 1992:3; Sheets 1992).

Sheets (1992:20) outlines a number of achievements made by the peoples of the Intermediate Area and directly correlates these to local environments, demographics, and subsistence-not to external influences. These accomplishments include emergence of an early and efficient agricultural and village lifestyle; increased societal stability related to "avoidance of the state"; polities that encompassed smaller geographical locations and had smaller population sizes, continuation of egalitarian and simple ranked societies; localized economic systems that emphasized independence; and artistic achievements in certain categories (i.e., goldworking).

The next section discusses cultural features, accomplishments, and traits in the context of a specific region of the Intermediate Area-Central Pacific Panama.

Central Pacific Panama

Ever since Preceramic and early Ceramic occupations were discovered on the central Pacific coast of Panama in the 1950s, it has been apparent that this region has been inhabited for long periods of time by cultural groups that show a strong degree of cultural continuity. This is reflected in the material culture record and is also apparent in lake cores, and from studies of historical linguistics and population genetics.

My study site, Zapotal, is located in that part of Panama where a distinctive art style-the "Gran CocI6 semiotic tradition" developed after about 300 BC (Cooke et al. 2000). Lothrop (1942) referred to this area as the "CocI6 Culture" whose boundaries were defined by the geographical distribution of polychrome pottery, which he identified


at the site of Sitio Conte. Since Lothrop's fieldwork, more data about the distribution of this art style have been acquired and it has become clear that the "CocI6 culture" was spread over a larger area of the isthmus than Lothrop originally thought, including some areas of the Caribbean watershed (Griggs 1995, 1998, 2000).

The major population centers of the tradition were located around the Gulf of Montijo, down the eastern side of the Azuero peninsula and in the coastal plains and foothills of Veraguas and CocI6 provinces. Its boundaries with contemporary but different traditions to the east and west are not well defined archaeologically and seem to have waxed and waned in time in response to still poorly understood social and economic factors (Cooke 1998; Cooke et al. 2000; SAnchez and Cooke 2000).

Currently, artifact distributions suggest that three major traditions can be grouped into "western", "central" and "eastern" areas (Cooke 1976, 1984; Cooke and Ranere

1 992b) to which the epithets "Gran Chiriqui" (west), "Gran Cocl6" (central) and "Gran Darikn (east) have recently been applied (Cooke et al. 2000). Differences among these areas are clear in the material culture record as far back as Preceramic times, and become accentuated through time as each area developed its own traditions. But archaeologists do not know whether each semiotic tradition reflects linguistic and population homogeneity or not; ethnohistoric sources suggest a high degree of linguistic heterogeneity in this area in the sixteenth century.

During Zapotal's major occupation (2500-1200 cal BC), similarities in material culture are observed in the area stippled on Figure 2.4 with important sites identified.


Cultural Chronology

The central region of Panama has witnessed more archaeological research projects than any other. These research projects resulted in a solid cultural chronology that is based on changes in technologies and subsistence activities. The chronology, pertinent dates, and important sites are listed in Table 2. 1. The Paleomndian and Late Ceramic periods are not discussed in the text because the late Preceramic and early Ceramic periods are the focus of this study. These cultural divisions are based on previous and recent archaeological work (Richard Cooke, personal communication 2001), and discussions presented in the literature by Cooke (1984, 1995), Cooke and Ranere (1 992b, c), Pipemno and Pearsall (1998), and Willey and McGimsey (1954), among others.

The Proyecto Santa Maria

Although archaeological research in central Pacific Panama has been conducted since the first half of the 2e~ century, the largest interdisciplinary project in the region was not undertaken until the latter half of the 1900s. This project-the Proyecto Santa Maria (PSM), ran from 1981 to 1985, under the direction of Richard Cooke and Anthony Ranere (Cooke and Ranere 1984, 1 992a, 1 992b). The PSM conducted surveys and test excavations in the Santa Maria River drainage. The project had several multipleworking hypotheses that focused on identifying the first human groups in the tropical forests, and the short- and long-term impacts of human occupation on these forests. Attention was paid to sites that were occupied during time periods ranging from


Table 2.1. Cultural Chronology of Central Pacific Panama and Associated
Archaeological Sites

Period Description Calibrated BC/AD Important Sites
I Paleoindian Terminal Pleistocene La Mula-Sarigua, Llano Hato
IIA Early Preceramic 8000-5000 BC Aguadulce, Carabali,
La Mula-Sarigua,
Cueva de los Vampiros IIB Late Pre-Ceramic 5000-2500 BC Cerro Mangote, Ladrones,
Aguadulce, Carabali
IIIA Early Ceramic 2500-1200 BC Ladrones, Aguadulce,
Carabali, Monagrillo,
Zapotal, La Mula-Sarigua, Cueva de los Vampiros IIIB 1200-500 BC Ladrones, Aguadulce,
La Mula-Sarigua
IVA Middle Ceramic 500-250 BC La Mula-Sarigua, Sitio Sierra
IVB 250 BC AD 200 La Mula-Sarigua,
Sitio Sierra, Cerro Juan Diaz, La India, Bdcaro
IVC AD 200-500 Cerro Juan Diaz, Sitio Sierra,
Sitio Conte, Cerro Gir6n, El Indio, La Cmnaza
IVC AD 500-750 Cerro Juan Diaz, Sitio Sierra,
Cerro Gir6n
VA Late Ceramic AD 750-1050
VB AD 1050-1200
VC AD 1200-1400
VD AD 1400-1520

the Paleoindian (ca. 10,000 BP) to the Middle ceramic, which was then believed to have

begun about A.D. 500. The following discussion of Central Pacific Panama prehistory

comes from data generated by a number of interdisciplinary research projects that were

conducted in this area since the 1970s (Table 2.2).


Table 2.2. Archaeological Research Projects Conducted in the Central Region of Panama
Since 1970

Years Investigators Site(s) Publications
1970-1975 Cooke Sitio Sierra, Cooke 1979, Bird & Cooke
Ladrones 1978, Cooke 1984, 1995, Cooke
& Ranere 1992c, 1999

1973, 1975, Ranere, Hansell, Aguadulce Shelter Ranere & Hansell, 1978, Hansell 1998 Piperno, Cooke 1979, Piperno et al. 2000, Cooke
& Ranere 1992c

1969-1972 French

1981-1985 Proyecto Santa Corona, Carabali, Cooke 1995; Cooke and Ranere
Maria Vampiros, Zapotal 1984; Hansell 1988;
and others Weiland 1984

1996,1997, Norr Zapotal Norr 1997b

1992-2001 Cooke Cerro Juan Diaz Cooke and Sanchez 1997; Cooke
et al. 1998; Cooke et al. 2000;
Diaz 1999;

Period H A and B The Preceramic

Technological change, incorporation of non-local plants into the diet, and

increases in human populations are hallmarks of this time period in the central region of Panama. Prior to this time period, settlement and subsistence patterns revolved mainly around the neo-tropical forests in which small groups practiced a mixed economy of hunting and gathering with simple cultivation of local plants and tubers (Cooke and Ranere 1992b, c; Piperno and Pearsall 1998). Around 7,000 B.P., the world's coastlines began to stabilize. This resulted in the formation of mangrove-fringed estuaries for the Paritsa Bay area (Clary et al. 1984). Although there are a limited number of buried


Preceramic sites along the coastline, the importance of marine and estuarine resources is shown by their presence at sites that are located a great distance from the coast.

PSM surveys located several surface scatters of stone artifacts along the coast that are probably Preceramic in age (Hansell 1988, Weiland 1984). But the only buried Preceramic deposit that was in a truly shoreline situation is Cerro Mangote (McGimsey 1956, 1957, McGimsey et al. 1986/7). Several rockshelters located inland from the coast have Preceramic deposits and contain some materials of coastal origin: Carabali (55 km), Corona (35 km), Aguadulce Shelter (11-22 km) and Cueva de los Ladrones (11-22 km).

Subsistence and settlement in the Preceramic-B are known from three welldocumented sites: Cueva de los Ladrones, Aguadulce Shelter, and Cerro Mangote (Figure 2.3). Location, material culture, subsistence evidence, and other relevant information, when available, are presented for each site in the following discussion.

Cueva de los Ladrones. This cave site is located at an elevation of 300 m above sea level, and during the time of occupation the site was ca. 11-22 km from Parita Bay (Cooke and Ranere 1992b:263, 268). The available floral data for the Preceramic-B from Ladrones include pollen and phytoliths of maize (Zea mays) and phytoliths Heliconia sp.(Piperno 1988:188). These data also suggest a decline in the use of arrowroot, possibly due to the incorporation of more productive starch sources (i.e., maize and manioc) (Piperno and Pearsall 1998:291). The corresponding faunal data (Table 2.3) recovered from the Preceramic-B deposits at Ladrones clearly show the use of marine and estuarine fish, crabs, and mollusks (Cooke 1995, 2001; Cooke and Ranere 1992b:268).


Enh"d Area Caribbean Sea

P a c f O c e a nt

0 50 100


Shell Mounds

Rockshelters Panama C

Cueva de los La

A ulce

erro Mang Parita Bay

Azuero Peninsula

Figure 2.3. Map of central Pacific Panama indicating locations of Late Preceramic
sites. (Graphic by Greg Heide).




Ux X

~t3 o t3
v~ ~ Q j8 gZ




xd x xx xx



Q,, to

_____C ~Z



x0 X




00 4



NX x xxxxxx x Xxx

oxx XXX x xxx xx xxxxxxx


x x x x x x x x x x


x x x X

X x X x XXXXXXXX x X xxx

x x x x x x x X x X X

v L. L, I %)

;I ;x A A N ;x ;s l t


Wxx Xx xxx x X XXXXX x

xx xxxxx XX x x x XXXXXXX

X x x

x X


x x x X X x x X

2 2
a! W t:o bo
00 t:3 t3
3t 5t 4 Q tj z
Q, Q ts C4. Q

43 -S -t C% z
zs %


x Xx xxxxxxx xx

V"XXXXX x xx xxxxxxx x xxx xx

C-i X

X x

t3 t3
tr ;%
ts q3 3


XX x

I-s 3z A


w z 2


~~X X X~ x x


x0 X0xX xXX

'~ 9 9

Eli- Ou












t3 t3 Q U

0: It !
L ~ 00 000R 0


bix x X<


~~X X

~4 o1
C3tosC, R (

H3t 3 R

I- ~?<~ <~< z

;3z z sz9 41



XX x X x

x x x x x X x X x x X

ts g E 9 .9
z z
4 ul, uk ul, u, k u,
(Z) u t!, 'al .2




X- Xt x



4z Z3

Q c, c s C


40 wj

wx xx it



Aguadulce Shelter. This rockshelter would also have been located ca. 11-22 km from the sea during Preceramic times (Cooke and Ranere 1992b:263). A Preceramic-age edge-ground cobble yielded a starch grain characteristic of maize (Piperno and Pearsall 1998:219). A second edge-ground cobble from the same context yielded two starch grains of a probable native species of yam (Dioscorea spp.) and two starch grains identical in size and shape to domesticated manioc (Piperno and Pearsall 1998:220). A Preceramic grinding base produced five starch grains identified to modern domesticated manioc. This same grinding base yielded an additional ten phytoliths from maize glumes (Piperno and Pearsall 1993), phytoliths that are probable arrowroot tubers (Piperno and Hoist 1997), and other unidentified legumes (Piperno and Pearsall 1998:221). Manioclike starch grains and maze glume phytoliths are the most common residues from grinding bases and edge-ground cobbles, suggesting these two taxa were the most common used, or the ones used on a regular basis (Piperno and Pearsall 1998:221).

In the late Preceramic at Aguadulce, edge-ground cobbles become the dominant plant-grinding tool in the lithic assemblage (Ranere and Cooke 1996). The botanical records show a decline in the importance of arrowroot at Aguadulce (Piperno 1988, 1995) Macrobotanical evidence suggests the use of oil palms (Elaeis sp.), prickly palm (Acrocomia mexicana), nance (Byrsonima sp.), hogplum (Spondias sp.), and (Hymenaea sp.) (Piperno and Pearsall 1998:291). The faunal data (Table 2.3) indicate the use of estuarine and mangrove fish, crabs, and mollusks (Cooke and Ranere 1989; 1992b:268, c:30-31).


Cerro Mangote. The site of Cerro Mangote is one of the oldest coastal sites in central Pacific Panama (Cooke and Ranere 1992b, 1999). The uncalibrated radiocarbon dates for this site range from 6810 +/- 110 to 5055 +/- 150 (Table 2.4). The site is a small habitation zone recognized by large quantities of shell, crab exoskeleton fragments, vertebrate faunal remains, and lithic artifacts. During the time of occupation it was located approximately 1.4-5.5 km from the coast. This site is not believed to have been a specific-use site (i.e., fishing station) (Cooke and Ranere 1992b; Piperno and Pearsall 1998:293). Vertebrate and invertebrate food remains include deer, raccoon, iguana, small reptiles, mangrove, shoreline and scrub-forest birds, crabs, and shows an emphasis on fishing and shellfish harvesting (Cooke and Ranere 1992b, 1992c). The quantity of deer and raccoon at Cerro Mangote has been suggested to be an indication of long-term use of the site (Cooke and Ranere 1992b:268). Refer to Table 2.3 for a complete list of identified taxa.

Table 2.4. Radiocarbon Dates from Cerro Mangote (after Cooke and Ranere 1992a: 120)

Material C14 Date (BP) Lab No.
charcoal 6810 +/- 110 Y-458D
shell* 6710+/- 170 TEM-206
charcoal 6670 +/-215 Beta-1219
shell 5990 +/- 180 TEM-174
shell 5440 +/- 130 TEM-176
shell 5140+/- 120 TEM-175
shell* 5055 +/- 150 TEM-207

Samples that were corrected for 12C/13C fractionation.


Presently the site is located 8 km from the sea along the Santa Maria river. The river runs fresh most of the year in this area. Archaeologically, only one freshwater taxon (represented by a single specimen) has been identified (Table 2.3) (Richard Cooke, personal communication 2001). The identification of specific fish taxa suggest a landbased (vs. watercraft) fishing strategy that focuses on taxa that are common to abundant in turbid inshore waters, such as mangrove fringes, river mouths and the oligohaline stretches of eastern tropical Pacific (ETP) rivers. This strategy focused on two sizeclasses offish: small (<.15 kg) individuals and large (.3-15 kg) individuals. The dominant small fish taxon is sleeper (Dormitator sp.), of which the ETP species, Pacific fat sleeper (D. latifrons), is abundant today in tidal and alvina/tidal pools (Cooke and Ranere 1992b:268). The fish species that attain much larger sizes, such as brown (Selenaspis dowi) and kessler's catfishes (Arius kessleri), some snook (Centropomus sp.), and snapper (Lutjanidae) live in the in-shore waters and tidal rivers while the most common corvina species recorded, the seatrout (Cynoscion albus), often moves into the upper estuaries (Jimfnez and Cooke in press). They are easily captured with simple weirs made from sticks used to block off mangrove channels (Cooke and Ranere 1992b), although no evidence for fishing gear has been recovered. In the 1956 excavations by McGimsey and his colleagues (McGimsey et al. 1986-87) two manatee (Trichechus manatus) rib fragments were recovered. Interestingly, manatees are only known from the Caribbean side of Panama, thus these were not locally secured items. Additional taxa recovered from Cerro Mangote are reported in Cooke and Ranere (1992c:36-37) and in my Table 2.3.


An important aspect of the Cerro Mangote site not observed in the other sites discussed was the presence of human burials. There were 90 individuals recorded by McGimsey et al. (1986-87). These skeletal remains were analyzed by Lynette Norr (1995) for their stable carbon and nitrogen isotopic compositions. Norr's analysis showed that the people at Cerro Mangote ate primarily terrestrial protein sources, with the occasional inclusion of estuarine, mangrove, and inshore marine protein resources (Norr 1995:222). The dates from the human skeletal remains (Table 2.5) and those taken from the midden (Table 2.4) do not match up. There are two possibilities for this: (1) the skeletons are not contemporary with the underlying midden, but are intrusive; or (2) the bone may be contaminated, thus making the dates more recent (R. Cooke, personal communication 2001). Another piece of evidence that points to the burials being of Preceramic age is that very few artifacts were found with the ca. 90 burials. What was found was shell, not ceramic. One would expect to find ceramics in burials of the ceramic period. Further dating will help to answer the questions that have arisen with these samples.

Table 2.5 Radiocarbon Dates on Skeletal Remains from Cerro Mangote (Richard Cooke, personal communication 2001)

Element Date
fibula 2630+/-60
femur 2320+/-50
intercostal bone 2200 +/-50
intercostal bone 1850 +/-45
tibia-fibula 2220 +/-45
intercostal bone 1970 +/-60
intercostal bone 2013 +/-50
intercostal bone 2260 +/-50



This discussion of the known Preceramic-B sites from the central Pacific area of Panama demonstrates that people were staying for long periods of time at Ladrones, Aguadulce and Cerro Mangote. This is suggested by the variety of the faunal resources and the distances that were being traveled to acquire them. The extent of the regional exchange system that allowed for the movement of foodstuffs (i.e., salted and dried fish) and animal-derived goods (i.e., manatee rib bones) across the landscape is not known (Cooke and Ranere 1992:268). Further research at these sites and the location of other Preceramic-B sites in this area may help to further illuminate the subsistence and possible exchange economies present during this time. Period HII A The Early Ceramic-A (4,500-3,200 B.P.)

Both similarities and differences exist between the Preceramic-B and the early Ceramic-A periods in central Pacific Panama. The lithic technologies reflected in both chipped and ground stone show no change during this time. Major developments present in the early Ceramic-A period, but not seen in the late Preceramic-B are (Cooke and Ranere 1992b; Piperno and Pearsall 1998:293)

1. The first appearance of pottery in the archaeological record of this area.

2. An increase in the number and size of sites in the Parita Bay region,

3. An apparent increase in the exploitation of marine/estuarine resources

4. Evidence in the paleobotanical record of new crop plants that were not recorded


There are no known cemetery sites for the early Ceramic-A period, thus data on paleopathologies, diseases, and isotopic composition of bone are absent.

The Monagrillo Ceramic Complex

The earliest ceramics in Panama are part of the Monagrillo Ceramic Complex first described by Willey and McGimsey (1954), and appear in the archaeological record by 5,000 -4,500 B. P. (Cooke 1995). Characteristics of Monagrillo ceramics include: a sand- and grit-tempered paste, vessels of simple, globular forms, and decorative techniques that include excising and (rarely) incising (Willey and McGimsey 1954; Cooke 1995). Monagrillo ceramics are friable, and prolonged exposure to the elements quickly erodes them eventually causing complete disintegration. This style of pottery has been recovered from five shell mound sites, and from post-Preceramic deposits at seven rockshelters in the region.

There are a number of sites in central Pacific Panama that contain these

Monagrillo ceramics (Figure 2.4): 1) the type site, Monagrillo (He-5); 2) Zapotal (He15); 3) La Mula-Sarigua 4) He-12; 5) He-18; and 6) Ag-66; 7) Cueva de los Vampiros; 8) Aguadulce; 9) Los Santanas; 10) Cueva de los Ladrones; 11) Carabali; 12) Corona; and 14) Rio Cobre (Cooke 1995; Cooke and Ranere 1992b). Of the open shell mounds, Monagrillo and Zapotal are the largest and have the most data recorded, while Cueva de los Ladrones and Aguadulce have provided the clearest evidence for plant use in the area, and also have identified vertebrate fauna assemblages.


Enlarged Area Caribbean Sea


Pacifc Ocean

0 50 100


Shell Mounds

Rockshelters Panama C

Cueva de los
Los Santanas Ladr g

A ulce

AfaRz'aParita Bay

Zapotal a Mula-Sarigua


Azuero Peninsula

Figure 2.4. Map of central Pacific Panama indicating locations of early Ceramic
sites. (Graphic by Greg Heide).


Shell mounds

The Monagrillo site (He-5) (4,400-2,200 B.P.). The Monagrillo shell mound was first recorded by Willey and McGimsey (1954) during their initial survey of the Parita Bay region in 1948. They returned in 1952 to conduct further excavations at the site. Their intention was to construct a much-needed artifact chronology with both temporal and spatial data. They determined the site to be approximately 1.4 ha in area, with cultural deposits averaging 3m in depth. There are nine radiocarbon dates from Monagrillo (Table 2.6). These dates range from 4,405 +/- 75 radiocarbon years to 3,245 +/- 100 radiocarbon years (Cooke and Ranere 1992a, b:270).

Ranere and Linares revisited Monagrillo in 1975 and excavated to two test pits (4m x 4m and 1.5m x 1.5m) (Cooke 1995; Hansell 1979). They used fine-mesh

Table 2.6. Radiocarbon Dates from Monagrillo (after Cooke and Ranere 1992a: 120-121)

Material C14 Date (BP) Lab No.
charcoal 4405 +/- 75 SI-2842
shell* 4350 +/- 160 TEM-208
charcoal 4135 +/- 80 SE-2844
charcoal 4090 +/- 70 Y-585
charcoal 3615 +/- 80 SI-2840
charcoal 3485 +/- 100 SI-2839
charcoal 3385 +/- 75 1-2838
charcoal 3325 +/- 85 1-9384
charcoal 3245 +/- 100 SI-2843
Indicates samples that have been corrected for 12C/13C fractionation.


screens to recover vertebrate and invertebrate faunal materials (Cooke 1995:171), and a column sample to completely recover molluscan remains, which were later analyzed by Hansell (1979). They confirmed Willey and McGimsey's reconstruction of the site's environmental history, and added ten additional radiocarbon dates (Cooke 1995).

Anthony Ranere revisited Monagrillo in 1975, and conducted excavations (Cooke and Ranere 1984, 1992a, b; Ranere and Hansell 1978). Using improved recovery techniques for faunal remains (1/8" mesh), numerous vertebrate food remains were recovered. Cooke (1995) has presented a brief analysis of this vertebrate assemblage, which will be summarized here. The sample that was analyzed yielded 8,500 fragments of vertebrate fauna remains from a minimum of 525 individuals (Cooke 1995:179). The taxa represented are marine in-shore species that include: deepbody thread herring (Opisthonema libertate), Pacific lookdown (Selene peruviana), Pacific bumper (Chloroscombrus orqueta), yellow bobo (Polydactylus opercualris), longfin herring (Ilishafurthii), and congo catfish (Cathoropsfuerthii). Cooke (1995:179) interpreted the remains as showing that the Monagrillo occupants were exploiting mainly shallow, lowturbidity waters. Based on the fish processing practices in use today in the area, the absence of burnt and charred fish bones suggests that the fish were salted, sun-dried and transported inland (Cooke 1995:179).

Charcoal recovered from the site contains wood fragments, but pollen and

phytoliths were not preserved (Cooke 1995:178; Piperno and Holst 1998:772). However, an edge-ground cobble recovered from just beneath the surface of the site yielded starch grains from its grinding facet. These grains most closely identified with those of manioc and maize. A second edge-ground cobble was recovered from 20-30 cm below the


surface and yielded phytoliths of palm and a starch grain that has characteristics consistent with those of maize (Piperno and Hoist 1998:772).

Unfortunately, the data presented here, and any remaining unanalyzed materials previously recovered from Monagrillo, are all that are available about the subsistence economy of this important early ceramic-A site. The Parita Bay region of Panama has undergone tremendous industrial and urban growth in the past 20-30 years. I visited the Monagrillo site in June 1999, and while the site is intact, all that were visible of the shell mound were a few shells observed under residential homes. Cooke (1984) recommended that this site, and the Aguadulce Shelter, be declared national monuments, but his words were not heeded.

The Zapotal site (He-15) (4,000-3,000 B.P.). The Zapotal shell mound, like

Monagrillo, was first recorded by Willey and McGimsey (1954) during their work in the Parita Bay area. They believed the site to be less than 1 ha, while Cooke (1995), on the basis of a field map drawn by Monique Giausserand, later estimated the site's size at approximately 3.1 ha. Further testing by Norr (1997) has shown the actual shell mound component of the site to be smaller in size, approximately 250 m in long by 50 m wide, with its long axis running parallel to the ancient coastline (Norr 1997) (Figure 2.5). The cultural deposits here are not as deep as those at Monagrillo (1.5m at Zapotal vs. 3m at Monagrillo). There are five radiocarbon dates from Zapotal that range from 4,010 +/- 100 and 3,500 +/- 80 radiocarbon years (Table 2.7) (Cooke 1984; Cooke and Ranere 1992a:121; 1992b).


Santum I

Grid Blocks

14 ISSo00 W60 8100 Wo q

Positive Probe "
a Negaive Probe 9 20 21

2 23 24 2

26 2 2& 29

31 33
S200 WA,100 .0S20o WO
34 35 $6 37 38 39 4 41

$250 1 o- Datum
22550 W

5100 W100


Figure 2.5. Map of Zapotal site plan indicating site boundaries and location
of ancient coastline. (Courtesy of Norr 1997)


Table 2.7. Radiocarbon Dates from Zapotal (after Cooke and Ranere 1992a: 121).

Material C14 Date (BP) Lab No.
shell* 4010 +/- 100 Beta-21,389
shell* 3850 +/- 70 Beta-20,849
shell* 3610 +/-70 Beta-21,388
shell* 3520 +/- 80 Beta-20,850
shell* 3500 +/-80 Beta-9574

Indicates samples that have been corrected for '2C/13C fractionation.

The Proyecto Santa Maria archaeologists opened a I m x 2m test unit in 1984

(Cooke 1995). They, like Willey and McGimsey, found Monagrillo ceramics scattered through the three very clear natural strata identified. They also employed fine-screen techniques in the recovery of vertebrate faunal remains, and found that the site matrix, "1... contained large amounts of shell and animal bone" (Cooke 1995:173). A preliminary analysis of fish remains from the 1984 cut was undertaken by Cooke and found evidence for a heavy reliance on small (50-100 g) and very small (<50 g) fish specimens. The represented taxa were predominantly grunts (Haemulidae) and herrings (Clupeidae) (Cooke and Ranere 1992b:273). Cooke (Cooke and Ranere 1992b:273) suggests that the Zapotal fisherfolk may have been using fine-mesh gill nets and watercraft, although no material evidence of fishing implements has been found to date.

In 1987, Yale University graduate student, Monique Giausserand, undertook

excavations at Zapotal (Cooke 1995). She opened out several trenches and began to strip off natural strata but unfortunately did not return to the site and never studied the materials she recovered. Giausserand recovered one possible residential structure from the Monagrillo period. Cooke observed this structure in 1987 and took photographs of it.


The structure was believed to be oval, with a central hearth (Cooke and Ranere 1992b:273). Alongside this structure was a cache of cobbles, some of which were edgeground (Cooke 1995). Giausserand also recovered approximately 1,200 Monagrillo sherds, and one possible ceramic figurine (Cooke 1995). Unfortunately, Giausserand' s analysis and excavation notes were never published, as she returned to France with the artifacts in her possession, believed to be held privately (Cooke 1995; Cooke personal communication, 1998).

Lynette Norr, of the University of Florida, began a multi-year project of surveying and testing at Zapotal in 1996. Norr returned to Zapotal in 1997 and 1999 for further excavations. Her research aimed at answering two main questions: 1) was Zapotal a permanent village or a series of seasonal occupations?; and 2) was the subsistence economy at Zapotal dependent on fishing, agriculture, or both (Norr 1997)? Her research design reflected these questions by including sub-surface probes; sondeos to discover features, that could be further investigated with larger units; and column sampling (Norr 1997). Excavating in a shell midden is not an easy task. The sondeos proved useful in identifying the locations of features, as many were found, including a hard-packed floor that is associated with the Monagrillo occupation of the site; a burned surface of an activity area, believed to be historic; post holes and a possible roof drip-line both feature types are probably associated with a later occupation (Norr 1997).

The 1999 field season is the focus of this study, and will be elaborated on in the following chapters. As seen at the Monagrillo site, this area of Panama is experiencing a rapid growth in urbanization and much of this area's land has been turned into cattle pasture. The owners of the land that Zapotal is on have turned the area covering the site


into unimproved pasture land for their cattle ranch. The major impact to the site is a crushed-stone road that runs through the middle of the site along its long axis.

La Mula-Sarigua. It is believed that La Mula-Sarigua was sparsely or

infrequently populated during early Ceramic-A times. Monagrillo ceramics are found in two areas of the site, mainly in surface scatters, covering an area of 1.3 ha (Cooke and Ranere 1992b:272-273; Hansell 1988). The ceramics only constitute 0.1% of the site's entire ceramic assemblage. The occupation of La Mula-Sarigua increased in size and intensity during the Middle Ceramic period, discussed in a later section.


Seven rockshelters on the Pacific slopes of central Panama and two on the Caribbean contain Monagrillo ceramics. On the Pacific these sites are: Aguadulce, Cueva de los Ladrones, Cueva de los Vampiros, Los Santanas, Carabali, Corona, and Rio Cobre (Cooke and Ranere 1992b:270). While all of these rockshelters contain Monagrillo ceramics, they vary greatly in their distance from the coast. Figure 2.4, shows the location of each site. Of these seven rockshelters, Aguadulce, Ladrones, Vampiros, and Los Santanas will be discussed further.

Aguadulce Rockshelter. This rockshelter was also described for the Preceramic period of central Pacific Panama. Ceramics and grinding stones that are similar to those recovered from the Monagrillo type site (He-5) were recovered from deposits that were stratified over late Preceramic deposits (Ranere and McCarty 1973). Paleoethnobotanical data for the early Ceramic period at this site include pollen grains of sweet potato and


maize (Piperno and Pearsall 1998:293). These pollen grains were present in the same layers as early Monagrillo ceramics (Piperno and Pearsall 1998:294). A date of 4,250 +/- 60 B.P., was obtained from phytoliths that were contextually associated with this early ceramic layer (Piperno and Pearsall 1998:294). Also within the Monagrillo layer is the first appearance of Curatella americana, an indicator taxon of major land degradation (Piperno 1998). Identified vertebrate taxa that represent the protein portion of the diet are listed in Table 2.3. Ranere (Ranere and McCarty 1973) proposed that Aguadulce was not a specialized coastal activity center, but a habitation site used by fishing-hunting-gathering peoples and, later in time, by farmers.

Cueva de los Ladrones. Ladrones is a 30 x 15 m rockshelter located on the southern flanks of a prominent hill. It was known to have been occupied in the Preceramic period as well as the early Ceramic-A. Bird and Cooke (1978) recovered 3,470 Monagrillo sherds from deposits stratified over these Preceramic levels. All of the ceramic sherds recovered from Ladrones were of the Monagrillo Plain variety (Cooke and Ranere 1992). Radiocarbon dates at Ladrones range from 4,800 +/- 100 radiocarbon years (4850 BP) to 3,770 +/- 80 radiocarbon years (3820 BP) (Cooke and Ranere 1992b). These dates place Ladrones in the early portion of the early-Ceramic A period.

Cueva de los Vampiros. Vampiros, also identified as Ag-145 during the PSM survey, is a rockshelter located 2.5 km from the coast. It has early Preceramic through Middle Ceramic (Period IV) components (Piperno and Clary 1984:86). The earliest occupation is represented by approximately one meter of sediments stratified below a sterile clay layer. This is the first contextual data recorded for Panama that represents the early Preceramic (Cooke and Ranere 1984:7). Estuarine mollusks and fish were


recovered from these deposits. The early Ceramic-A occupation of this site began by 3,800 +/- 120 radiocarbon years before present (Beta-5870) (Cooke 1995; Cooke and Ranere 1992b:273). Cooke believes that this site was predominantly used as a fishing station. Of the recovered vertebrate faunal assemblage, 99 percent of the specimens were fish remains. The fish taxa represented in this assemblage include sierra (Scomberomorus sp.), needlefish (Belonidae), and green jack (Caranx caballus) (Cooke 1988; Cooke and Ranere 1984). Vampiros was damaged by shrimp farm operations in 1999, and will be re-excavated in 2001 by Cooke, Carvajal and Pearson (Cooke personal communication 2001).

Los Santanas. This rockshelter was located during the Proyecto Santa Maria. It is situated at an elevation of 800 m, and is ca. 70 km from Parita Bay. Monagrillo ceramics were recovered from this site, stratified over Late Preceramic deposits (Cooke and Ranere 1992b).

Summary. The early Ceramic-A period shows an increase or intensification of trends first seen in the Preceramic period Stone tool technologies are similar in both periods, as both chipped and ground stone implements have been recovered. Pottery is introduced into the archaeological record ca. 4,400 B.P. With this new technology comes an increase in the numbers of coastal sites, like Monagrillo and Zapotal, as well as the continued use of rockshelters and more inland locales. At some coastal sites (i.e., Cerro Mangote, Vampiros, Mongarillo) the types of fish that are present in these deposits are different than those from earlier time periods, of coastal sites brings an increase in the exploitation of marine and estuarine fish and shellfish. At Cerro Mangote paleoethnobotanical remains suggest the continued use of locally available plants and


cultivated crops. The lack of confirmed domestic structures and cemeteries make population estimates for these sites difficult at best.

Period IV Middle Ceramic (3,000 B.P. A.D. 500)

The Middle Ceramic is best characterized by sites without four color polychrome pottery. The lifestyle reconstructed for this period is drastically different that those of the Preceramic and early Ceramic periods. Evidence from La Mula-Sarigua and Cerro Juan Diaz indicate that most of the characteristics of the Gran Cocl6 culture are evident by this time. Large (50 ha) settled villages that focus on the cultivation, processing, and consumption of crop plants are evident in the archaeological record. Maize increased in importance in the diet. Marine fish, shellfish, and white-tailed deer are still important contributors to the diet. The incorporation of new technologies, such as legged met ates and cylindrical manos, as well as stone tools for the production and processing of crop plants, are evident. Ceramics become more refined with the introduction of new decorative techniques and vessel forms, although four color polychrome pottery is absent in this period. Several sites have been documented for Period IV (Figure 2.6): La MulaSarigua, Sitio Sierra, Gir6n, along the Santa Maria River, and Cerro Juan Diaz, along the La Villa River. Available data on these sites will be discussed.

Tool production during this period involved an introduction of new tool forms as well as changes in existing tool function. Ground and polished stone tools were used for forest clearing, and plant-grinding implements such as manos and metates dominate the


Enlarg~ed Arm Caribean Sea
w *E

Paii OCCU Colmbi

0 50 100


~Panamna C'

-'i'rr 0 id irra Parita Bay

3tLa Muria arigua Cerro Juan Di

Azuero Peninsula

Figure 2.6. Map of central Pacific Panama indicating locations Middle Ceramic sites.
(Graphic by Greg Heide).


assemblages of this period (Piperno and Pearsall 1998). An increase in types of ceramic vessel forms, as well as a presumed increase in time and effort involved in their production were recovered. The agricultural system in place during the Middle Ceramic emphasized the cultivation of maize and additional crops on a non-shifting basis along the major coastal plain rivers and streams (Piperno and Pearsall 1998:295).

La Mula-Sarigua. Located on a ridge overlooking the mouth of the Parita River, La Mula-Sarigua is considered the earliest nucleated village in the region (Cooke and Ranere 1992c; Hansell 1987, 1988). During the thousand years of documented occupancy, the site grew from 8.5 ha to 58 ha in size, suggesting a rapid population growth, combined with nucleation (Cooke and Ranere 1992b, c; Hansell 1987). A 5-ha store of chalcedony cobbles identified at the center of the site was an important source of raw material for tool production (Cooke and Ranere 1992b). Tools used in food production that were recovered from the site include: legless slab metates, pear-shaped polished celts; and small, hafted, unifacial points (Cooke and Ranere 1984; Hansell 1998). New ceramic vessel forms have been identified from La Mula-Sarigua, and include: vessels with collars; tall necks; thickened exterior and interior rims; straight thick walls; and the addition of small handles. Larger-sized vessels were also recovered (Cooke and Ranere 1992; Hansell 1988).

The paleoethnobotanical data from La Mula-Sarigua include starch grains, but no macrobotanical remains (Piperno and Holst 1997). The starch grains were removed from a legless metate and suggest that a variety of crop plants were being prepared. The identified taxa are represented by maize, manioc, a type of yam, and a probable Calathea species (Piperno and Holst 1997). Other dietary lines of evidence include bone isotope


analysis and vertebrate and invertebrate faunal remains. The a 3C and a'5N values for three human skeletons suggest that the diet consisted of C4 plants (i.e., maize, setarias, amaranths, chenopods). These values are greater than those from the Preceramic site of Cerro Mangote (Norr 1990). The vertebrate faunal remains are dominated by marine fish taxa such as, catfish, Pacific bumper, and deepbody thread herring, as well as shark and ray. Other vertebrates include deer, armadillo, toad, sea turtle/and or crocodile, bird, rodent, lizard and mud turtle (Cooke and Ranere 1992c). Invertebrate faunal remains from La Mula-Sarigua include several species of ark, carditids, venus clams, oyster, pitar, slippersnail, ceriths, tulip, moonsnail, sharkeye, rocksnail, and other bivalves and gastropods (Hansell 1979).

Sitio Sierra (A. D. 250 A.D. 750). Evidence of a sedentary maize-farming community has been identified from the site of Sitio Sierra (Cooke 1979, 1984; Cooke and Ranere 1979, 1992c, 1999). Remains of domestic structures, and their contents, have been recovered, offering insight into the subsistence economy and activities practiced at Sitio Sierra. The following discussion of artifacts and features recovered from Sitio Sierra is summarized from Cooke (1984:284-285). Features related to several possible living structures include: living floors; hearths; post holes; hour-glass-shaped pits with sand linings and stones in the centers; and elongated pits at the structures' edges (Cooke 1984:284). Additionally, wasp nests with leaf impressions were found along the periphery of the structures. These impressions were of palm leaves and grasses, indicating their use in roofs and walls (Cooke 1984:284).

The material culture that was recovered from the floors of these structures suggest on-site agricultural activities. These items include: partial-legged grinding tables;


elongated, cylindrical manos; large quantities of carbonized maize; polishing pebbles for axes; and rectangular blades(Cooke 1984:284). The presence of these indicates that the inhabitants of Sitio Sierra were maize agriculturists, who lived in houses with cane or palm walls and thatched roofs. The hour-glass-shaped pits may have been used for steaming, smoking, or roasting of foods, while the elongated pits may have been used to store vegetables (Cooke 1984).

Other subsistence data are in the form of faunal remains and dietary

reconstructions based on bone stable isotope composition of human remains. The discussion of the vertebrate faunal assemblage from Sitio Sierra is summarized from Cooke 1984, 1992; Cooke et al. 1996; and Cooke and Ranere 1992c, 1999.

The faunal assemblage from Sitio Sierra is well-preserved and contains a large number of specimens, but has not been completely analyzed. The fish remains from one unit reveal that ca. 30% of the represented taxa are freshwater (Cooke and Ranere 1999). Amphibians and reptiles such as toads and turtles were identified in large quantities (Cooke 1984b, 1989). Abundant remains of birds and mammals were identified from this site. The avian faunas are represented by herons, egrets, ducks, bobwhite quail, whitetailed nightjar, doves and falcons. The mammals are represented by paca, nine-banded armadillo, rodents, and white-tailed deer. The most abundant taxa (in terms of Minimum Number of Individuals) identified from Sitio Sierra is the white-tailed deer, suggesting that it was an important game animal for the occupants of Sitio Sierra.

Human burials recovered from Sitio Sierra represent two components of the site: ca. 2200 B.P. and A. D. 1 100. Interestingly, both groups have isotopic values that were quite similar (Norr 1995). The stable isotope values for carbon and nitrogen from Sitio


Sierra reveal a diet that was highly mixed; both C3 (i.e., temperate grasses, trees, fruits, tubers) and C4 plants were used, and both terrestrial and marine proteins were consumed (Noff 1990, 1995).

Gir6n (Co-2). The village site of Gir6n was excavated in 1952, during Willey and McGimsey's survey of the area (Ladd 1964; Willey and Stoddard 1954). Much of this discussion of the site is summarized from Ladd (1964:154- 184). The site is located on the north bank of the Santa Maria River, approximately 10 kilometers from the river's delta. The main features of this site consisted of three middens around the base of a low hill. A total of two 3mi x 3 m units, and one 1im x Ilm test pit were excavated at the site. Artifacts recovered from this site included 3,675 identified ceramic sherds of varying types (Escotd Polychrome, Gir6n Polychrome, and Red Daubed Variety); manos and metates; celts; jasper flakes; two cut bone tubes; and numerous shellfish remains.

In later excavations (McGimsey 1959), three human skeletons were recovered

from the lowermost deposits. Two of these individuals, one adult and one juvenile, were buried in an extended position, while the third was flexed. The cultural associations of the burials are not known.

Based on the artifact inventory of the site, Gir6n is believed to have been an

agricultural village dependent upon the fertile soils of the alluvial plain for successful plant cultivation. The close proximity to the river, estuaries, and coastline made this site an ideal spot to exploit fish and shellfish. While no vertebrate fauna remains were mentioned in the Gir6n accounts, based on research for sites occupied in earlier time periods in this region (i.e., Cerro Mangote, Monagrillo, Zapotal), it can be assumed that teleosts as well as invertebrates contributed large portions of the dietary protein.


Cerro Juan Diaz (2,200 300 B.P.). The site is situated on top of a hill that

overlooks the La Villa River and its adjacent estuary. The river is the modern boundary between the Herrera and Los Santos Provinces. Cerro Juan Diaz is located along the eastern shore of the La Villa River, in Los Santos Province. Publications to date from this project have dealt with flanerary artifacts (Cooke and Sinchez 1998; Cooke et al 1998; Cooke et al. 2000; Cooke in press; Cooke and Sinchez in press), vertebrate faunal remains (Jim~nez 1999; Jim~nez and Cooke in press), and invertebrate faunal remains (Carvajal 1998). A detailed description of the paleopathology and physical characteristics of 115 skeletons from a late Ceramic A burial ground is included in Diaz (1999).

Excavations at the site have been on-going since 1992, under the direction of Richard G. Cooke. This site has a long history of occupation (ca. 2,000 years). Due to this long occupation sequence, the site has numerous features, some disturbed and redeposited by subsequent inhabitants, others by modern looters. The features and artifacts from Cerro Juan Diaz correspond to several types of activities associated with both a village and a cemetery.

Numerous human burials have been recovered, belonging to four different time periods. Many are associated with funerary objects. The known presence of these funerary objects have made the site susceptible to heavy looting activity. Funerary objects include, but are not limited to: gold ornaments, polychrome pottery, pearls, jade ornaments, metates, and shell beads. Assessments of skeletal remains mortuary customs have been conducted and are ongoing (Diaz 1999; Sdnchez 1995; Cooke and Sinchez 1998; Cooke et al. 1998; Cooke et al. 2000).



Many lines of evidence suggest that there was an important and apparently quite rapid shift in settlement location during the Middle Ceramic in the central region of Panama. These processes seem to be in place by a bout 250 cal BC, but their origins are surely older. Whether this change was essentially endogenous in nature, being related to natural population growth and changes in the subsistence structure, or whether populations from outside the area were involved, remains moot. These problems have been discussed for the western region (Gran Chiriqui) (Linares 1977; Linares et al. 1975; Linares et al. 1980).

Stone tools include a number of implements used in plant cultivation (elongated blades, polished axes) as well as items for plant processing (legged metates, cylindrical manos). Ceramics become more varied, with the introduction of multiple color motifs and plastic decorations (scarified jars, etc.) (Cooke 1984). Funerary objects exhibit the craftsmanship of specialists working with gold, jade, and other precious media. Other objects of personal adornment are common at cemetery sites. Subsistence data indicate a diet that incorporated both terrestrial and marine protein resources, as well as C3 and C4 plants. Further analyses of the Cerro Juan Diaz individuals will add to the current view on subsistence for this time period. During this time period there is little evidence for wealth differentiation among individuals, but this type of evidence has been recovered from Sitio Conte post-A.D. 750 (Briggs 1986, 1989, 1992, 1993).


Summary of Central Pacific Panama Prehistory

This overview of the prehistory of central Pacific Panama from 7,000 B.P. to

AD. 750 is intended to feature the main sequence of cultural and economic change and continuity. The cultural time periods follow data that have accrued from several research projects undertaken in the area since 1970 (Tables 2.1 and 2.2). Within each period, important sites have been discussed including relevant subsistence, technological, and settlement data. The reader should have a sense of the long-term continuity in subsistence and settlement practices, and how the slowly emerging trends of plant cultivation brought about a change in stone tool technologies in Middle Ceramic and also a shift away from hillsope hamlets towards more nucleated settlements on the coast. The appearance of pottery in the early Ceramic-A is well-documented, and has been found clearly stratified over Preceramic layers at some sites. The addition of pottery to the artifact assemblage does little to change stone tool manufacture. The pottery is very poorly made and fired, suggesting it was a local invention, although the stimulus may have come from elsewhere. The abundance of ceramics tends to be at sites nearer the coast rather than away from it. It appears to have been an expedient technology that probably was a response, in part, to the fact that as populations grew and migrated it became more and more difficult to provide plant containers for every day life.

While many questions have been answered about the subsistence economies and settlement practices of the earliest peoples of central Pacific Panama, as with most aspects of archaeology, more questions have arisen that can only be answered with further site surveys, excavations, and analyses.


The Zapotal site is located in the Santa Maria River watershed in central Pacific Panama. This watershed encompasses a 3,500 km2 area to the south of the Continental Divide (Figure 3. 1). Over the last three decades researchers have studied this area's geological, hydrological, vegetational, and faunal histories, and have applied these data to questions concerning cultural adaptations and culture/environment interactions (Barber 198 1; Clary et al. 1984; Cooke 1992; Cooke et al. 1996; Cooke and Olson 1984; Cooke and Ranere 1992a, c; Dere 1981; Golik 1968; Piperno et al. 1991b; Piperno and Pearsall 1998; and Stehli and Webb 1985). A review of some of these data will be presented to place the Zapotal site within its local environmental setting.

Modern and Prehistoric Environments The Modern Environmental Setting

To flfly describe the environment of a specific locale, multiple lines of evidence must be gathered and integrated. The modem environmental data sets that will be discussed are rainfall, ecological zones, geological formations, ocean currents, and plant communities.



Ealrgme Area Caribbean Sea E


Pacsse Ocean

0 50 100


Panama C

Parita Bay


Azuero Peninsula

Figure 3. 1. Map of central Pacific Panama indicating location of Zapotal (He-15).


The Santa Maria River. The nearest freshwater source to the Zapotal site was and still is, the Santa Maria River located in the lower coastal plain, the largest of four rivers in this area. The floodplain of the Santa Maria is wide, with numerous oxbow lakes. Presently the river is cutting into non-lithified sediments, such as silt, sand, and clay, creating steep banks (Clary et al. 1984:58).

Rainfall. The study area is located in the driest part of Panama. There are two seasons in this area; the wet or rainy season (May to November), and the dry season (December to April), when little to no rainfall occurs. The average annual rainfall of this area is 1,000 1,400 mm (Cooke 1984:282).

Ecological zones. The coastline of Parita Bay consists of a system of estuaries,

mangrove swamps, and salt flats (alvinas) (Clary et al. 1984). Based on data recovered by a system of cores taken from both sides of the Santa Maria River the alvinas have been interpreted to be of relatively recent origin (ca. 2,000 years old) (Clary et al. 1984; Cooke and Ranere 1984). Additionally, Barber (198 1) has indicated that the coast has been prograding in the area of the Santa Maria delta at the rate of 1 km per thousand years.

The coastal area has been divided into four geomorphological zones:

1) supratidal; 2) high-tidal (alvinas); 3) mid-tidal; and 4) low-tidal (Clary et al. 1984; Dere 198 1). The following discussion of the four zones specific to the Santa Maria delta is summarized from Clary et al. (1984:). The supratidal zone is transitional between the mainland and the high-tidal zone. Cores taken in this area show that the sediment is a silty clay. This zone is sparsely inhabited by small trees and cacti, as well as grassy vegetation. Often there are standing and fallen dead trees in this zone.

Moving towards the Parita Bay from the supratidal zone, the surface begins to flatten and gradually slope towards the water. The area where the surface begins to flatten is the high-tidal zone, primarily made up of the alvinas. The high-tidal zone is a small area of transition, where the mainland begins to grade into the bare alvinas.


The mid-tidal zone contains bare alvinas that support algal growth as well as

mangroves along the more seaward edge. These scrubby mangroves (Avicennia sp.) give way to stands of large red mangroves (Rhizophora sp.), which can be quite dense. This dense vegetation grades into the strand-line where very tall red mangroves grow in soils that are typically inundated. Along the coastline of Parita Bay, mangroves cover an area estimated at 11,553 ha (Jim~nez 1992:260).

The strand-line grades into the low-tidal salt flats which are exposed by the diurnal tides. These tidal flats are marked by sand ripples resulting from low-wave action, fluvial sources, and longshore currents. Vegetation does not grow along the surface of these ripples, but organic debris such as leaves, shell fragments, polychaete worm tubes, "sand-dollars," crab tracks, and animal burrows, does accumulate here.

Ocean currents. The western shore of the Parita Bay is a low energy coastline. The highly variable diurnal tidal phases can range up to 6 m. Longshore currents from the northern and southwestern sides of the bay converge onto this area, depositing sediments on the river delta, causing the local river valleys to become silted-up (Clary et al. 1984:59). During the dry season, strong northerly trade winds displace surface waters in Parita Bay, causing an upwelling and nutrient enrichment of the bay waters (Ranere and Hansell 1978). This upwelling results in an increase in phytoplankton production, seasonal influxes of shellfish, crustaceans, schools of shallow-water fish, and flocks of sea birds (Glynn 1972).

Sea-level rise. Researchers agree that since the last ice age sea-levels have risen globally in response to the melting ice glaciers. Golik (1968) gathered data from Parita Bay to interpret the rate and timing of sea-level rise. He has indicated that there was a post-glacial rise in sea-level of 6.6 mn per thousand years until 7,000 B. P., in Parita Bay. Adams et al. (1982 in Clary et al. 1984) have gathered data from sediment cores and propose a local sea-level rise for Parita Bay of 0.3 m per thousand years for the last 5,000 years.


Erosional processes. The bareness of the alvinas allow for wind-transport of
sand and silt, which also contributes to silting-up the river. Where the Santa Maria River crosses the alvina, sand dunes build up and cover the aerial root systems of the mangroves, causing them to die. This loss of the mangroves and their root system results in further aeolian erosion (Clary et al. 1984).

The Paleoenvironmental Setting

Much of the information outlined in the preceding section is also pertinent to a discussion of the paleoenvironmental setting, but some changes in the environment have also occurred since 7,000 B.P., due to sea-level rise, coastal progradation, erosion, and human actions, among others. This section will outline what the ecology of the Zapotal site would have been like during the time it was inhabited, and how Zapotal's environment differed from that of the nearby Monagrillo.

Geological formations. Deposits of quartz, chalcedony, and jasper are numerous within the basement geology of the Santa Maria River basin (Clary et al. 1984). Often these deposits are exposed at the ground surface and in stream beds. Prehistorically these exposed deposits would have been important sources of raw lithic materials for inhabitants of the river basin (Cooke and Ranere 1984).

Paleoethnobotanical evidence. Piperno and Pearsall (1998) used

paleoethnobotanical data recovered through archaeological excavations and surveys and geological cores to outline the botanical communities of the Santa Maria watershed of central Panama. Their findings from these different lines of evidence show that by 2,000 B.P., prehistoric peoples fully participated in agriculture, emphasizing land use in the alluvial bottomlands of the coastal plain (Piperno and Pearsall 1998:295). Through their analysis of lake sediments for this time period, Piperno and Pearsall (1998) noted an absence of primary trees and the presence of grasses that reflect human modified


environments. They suggest that deforestation of the uplands resulted in people moving into and resettling in areas with more fertile soils (Piperno and Pearsall 1998:295).

Coastal landscape. Short-term landscape features made the Parita Bay coastline favorable for occupation during the time that Zapotal and Monagrillo were inhabited. Monagrillo was located near a shallow-water coastal embayment (Willey and McGimsey 1954); Zapotal was part of the active river delta (Cooke and Ranere 1992a). According to Willey and McGimsey's (1954) interpretations, the first occupation of Monagrillo was on a sandbar that was subjected to wave action. Continued occupation was intermittent until the formation of a shallow lagoon around the site, after which the occupation became more permanent. The site was in a favorable location for people to fish and collect shellfish until about 3,500 B.P. At that time, the alvinas expanded to such an extent that they made the site unattractive for subsistence purposes. Ranere (1980; Cooke 1984) believes that Monagrillo was occupied intermittently as early as 4,400 B.P., but the ideal conditions for the exploitation of clams did not stabilize until 3,700 B.P. The conditions that favored clam harvesting were related to a nearby high-energy sandy beach. The formation of a mangrove swamp by 2,950 B.P., led to the site being abandoned as conditions were no longer favorable for shellfish collecting. Ranere (1980) based his conclusions about site formation on data recovered from the PSM drilling project, as well as radiometric dates (Cooke 1984).

For a time, Zapotal' s location along the river delta and ancient coastline made it extremely favorable for exploitation of estuarine fish and shellfish. Occupation, which was slightly later than Monagrillo, coincided with a particular developmental stage in the Parita Bay coastline that favored exploiting a variety of local resources (Cooke and Ranere 1 992b). The site lies parallel to the ancient coastline and perpendicular to the course of the Santa Maria River (Figure 2.6). This location granted inhabitants of the site easy access to a variety of food resources.


Zapotal (He-15) Background Considerations Site Location

Zapotal is located on the south bank of the Santa Maria River, approximately 6

km from the river's delta, along the prehistoric coastline (Figure 3.1). The shell mound is estimated to be approximately 250m x 50m in size (Norr 1997). Between Zapotal and the Santa Maria River to the north is apotrero, and an alvina lies to the east. Apotrero (also called a pasture) is barren grassland, similar to an alvina, but in contrast it is flooded and silted by the fresh water from the river. The potrero in this vicinity, like the alvina, is believed to be of recent geological origin (Willey and McGimsey 1954).

Early Excavations Willey and McGimsey 1952

The first excavations at Zapotal were initiated by Willey and McGimsey (1954) in March 1952. They excavated six 3m x 3m units in three different locations across the site, recovering "several hundred sherds" of Monagrillo ceramics (Plain, Red), as well as sherds of Zapotal Plain, that they believed to be a variant of Monagrillo Plain (Willey and McGimsey 1954:94). Other artifacts included three triangular pebble choppers, four small ball-like pebble hammers, and faunal remains. They note that the "only identifiable bone encountered was in the 0-10 cm levels. Freshwater turtle, bird, cow, and peccary were represented (the cow would have been deposited historically, during the time the land was used as a corral) (Willey and McGimsey 1954:96). It was noted that shellfish from the lagoon were utilized, and their shells were found throughout the site's matrix. The vertebrates recovered from Zapotal in 1952 were identified as Protothaca grata, Anadara grandis, and Naticia unifasciata (Willey and McGimsey 1954:93).


The Proyecto Santa Maria (PSM)-1984

The archaeologists of the Proyecto Santa Maria (PSM) revisited Zapotal in 1984, during their survey of the Santa Maria River drainage. A Im x 2m test unit was excavated and materials were fine-screened to recover data relevant to interpreting subsistence and settlement patterns in the Santa Maria watershed. The PSM archaeologists identified Mongarillo ceramics, and obtained a single 14C date of 1,190 B.C. [BETA-9574] from the lowest clearly identified natural stratum (Cooke 1995:173). As described in Chapter 2, the vertebrate faunal assemblage recovered in 1984, indicated a heavy reliance on taxa of small and very small fishes, represented by grunts and herrings. This is in contrast to the faunal assemblage recovered by Willey and McGimsey (1954) in 1952, that did not contain any bony fish remains.

Monique Giausserand's Excavations 1987

In 1987 Monique Giausserand obtained funds to conduct large-scale excavations. Her research objectives were to gather data related to the site's subsistence economy and to assess the ceramic assemblage. She intended to accomplish this by opening a large portion of the site horizontally. Unfortunately, Giausserand's never completed her field program, her trenches had to be back-filled and the work she did has never been reported. However, Cooke (1995:173) describes a partial oval-shaped structure with a central hearth that was uncovered during Giausserand's excavations. This is the first possible structure identified for the early Ceramic-A period. The deposits immediately below the floor produced a "shell date of 1,660 (1,430) 1,220 B.C." (Cooke 1995:173). Giausserand also recovered approximately 1,200 Monagrillo sherds and one possible ceramic figurine during her excavations at the site (Cooke 1995).


The Zapotal Archaeology Project 1996, 1997, 1999

Research and sampling strategy. As described in Chapter 2, Lynette Norr, of the University of Florida, began a multi-year project at Zapotal in 1996. Norr's sampling strategy superimposed a horizontal grid system of 20m x 20m block across the site. Each block was numbered consecutively from 1 to 5 1, and further divided into one hundred 2m x 2m excavation units, which were numbered sequentially (Norr 1997) (Figure 2.6). Twelve of the 2m x 2m excavation units were randomly selected for immediate excavation. Each of the excavation units was first explored by a I m x 50 cm test pit to expose the site's stratigraphy, midden deposits, and features. If features were encountered, then the entire 2m x 2m excavation unit was opened. Excavated sediments were screened through 1/4" screen and 1/8" mesh sequentially to recover artifacts, vertebrate and invertebrate faunal remains, as well as macrobotanical remains.

This sampling strategy also called for four column samples from each block to be removed in 5 cm arbitrary levels within the natural stratigraphy. This would allow for total recovery of all artifacts, and the matrix to be processed later at the field lab. Lab processing consisted of dry screening the column samples through 1/4", 1/8" and 1/16" mesh, then water-screening with 1/32" mesh to allow for the recovery of microbotanical remains. Preliminary processing of all recovered materials was performed at the field laboratory. This allowed for ease in transporting materials to the University of Florida campus in Gainesville.

The 1999 field season. The field crew for the 1999 season was comprised of Lynette Non-, Project Director; Tanya Peres, Field and Teaching Assistant; Ruben Enriquez, an inspector for the Patrimonio division of the National Institute of Culture; and four UF undergraduate Anthropology students: Gurudev "Dave" Allin, Eric Merges,


Ben Pariser, and Denise Warner. Their participation in the field season afforded them the opportunity to learn field excavation and laboratory techniques. Beginning work at the site on June 22, 1999, the crew re-opened three 2m x 2m excavation units in Blocks (Units 47, 56, 65), begun in 1997. Merges opened a new I m x 50cm. sondeo, Unit 45 (Figure 3.2).

Progress on the three excavation units was slow due to the enormous amount of shell in the matrix. All excavated sediments were screened through 1/8" mesh at the site, which took an extremely long time. Not one of the three 2m x 2m units was excavated to sterile soil in the 1999 field season. The sondeo from Unit 45, however, was completed.

The Study Assemblage


A number of factors were considered in choosing an assemblage from Zapotal for further zooarchaeological analysis. These factors included:

1. the unit had to have been completely excavated to sterile soil;
2. all sediments were screened through at least 1/8" mesh to insure recovery of a more
complete size range of vertebrate faunal remains; but this technique can not ensure
the recovery of the entire size range;
3. a complete sample of invertebrate remains that correlated with the vertebrates was
4. the excavated remains were preferably from the whole-shell shell midden;
5. the specimens needed to be easily and quickly transported back to the UF campus

In years prior to 1999, not all of these criteria had been met. Not all of the units had been completely excavated to sterile soil; nor were sediments screened with at least 1/8" mesh. Often representative "grab" samples were taken for invertebrates, thus limiting their usefulness in interpretations of species abundance and diversity; and not all materials were sent back to the UF campus.


Zapotal -1 2 "Grid Blocks

Excavation Block 47 :- = 'Unit 45


Site map courtesy
of Norr 1997.


0 2 4

X = Location from which faunal sample was taken. = Positive probe.

Figure 3.2. Plan view of Zapotal indicating location of Block 47, Unit 45 sondeo.
(Graphic by Greg Heide)


The Unit 45 Sondeo

The sondeo from Unit 45 from the 1999 field season was chosen for this study because it fit all of these criteria. The invertebrate remains were not transported back to the UF campus because they were completely identified, analyzed, and quantified (with the exception of weights) at the field laboratory (Merges 1999b). This is important because more significant interpretations can be made regarding species richness and diversity, contributions to the diet, site function and seasonality, and habitat exploitation.

Excavation. The sondeo with Unit 45 was chosen for excavation and assigned to Merges. Excavations began by taking standard datum measurements, mapping, and noterecording. The research strategy called for sondeos to be excavated in 5 cm arbitrary levels within natural strata. During the course of excavations, ten natural strata were identified. Within the natural strata twelve 5 cm arbitrary levels were excavated as they cross-cut the natural levels. The entire sondeo was excavated to sterile soil, approximately 70 cm below datum (cmbd) (Figure 3.3).

Artifacts recovered from the sondeo are common to the entire Zapotal

assemblage: Monagrillo ceramics, chipped stone, fire-cracked rock, macrobotanical remains (e.g., carbonized wood), and faunal remains.

Sondeo soil matrix. The strata of Unit 45 clearly defines the midden deposition, as well as post-midden deposits. Table 3. 1, offers a description of the sondeo 's natural strata from the top of the excavation unit (ground surface) to the base of the shell midden. This description includes: depth in centimeters below datum (cmbd); strata designation (A, B, C, etc.); Munsell soil chart color designation and description; the sedimentary composition of the soil; and the types of inclusions within each stratum.


As can be seen in Table 3. 1, and in Figure 3.3, the whole-shell component of the shell midden is first encountered at 25.5 cmbd, at the base of Stratum C/top of Stratum D, and ends at a depth of 65.5 cmbd, in Stratum JI No cultural features were encountered in the sondeo. While carbonized botanical remains were recovered for future radiocarbon dating, no large concentrations of carbon were uncovered that are suggestive of any type of activity area. Since no evidence of structures was recovered during this excavation, this area can be considered to be entirely within the midden. The lack of features in the midden zones suggests this part of the site has not incurred post-depositional disturbances. The integrity of the deposits is high, and the Unit 45 sondec can be considered to be intact, primary deposition, midden.

Stratum A, 3-8 cmbd, is part of the modem ground surface, and several historic artifacts were recovered (e.g., barbed wire, glass). Strata B and C are deposits of sand, clay, and silt, common soil components for this area. Crushed shell inclusions and pebbles are a result of deposition directly on top of the shell midden. The vertebrate and invertebrate faunal remains from the shell midden recovered from this unit have been analyzed, and will be discussed further in Chapter 4.


Unit 45 East Wall




0.5 .

0 .4 .. .... ... ... .. .. . . . . . .

...........An.d.. .. .. .. .. .d.. .. ... . .
........... ........ .. ..b .

0 . . . . . . . . .
0 . .1 . . . . 0 .2I .3. 0...4 .5

... ... ... .. ... ... .. Z.. n .. .. w it qndden

Figure........... 3..Prflem pofU it4,.od.,EstW .. (Graphic. by. Greg... i. e


Table 3.1. Description of Natural Strata (A-J) from Unit 45, Sondeo.

Color Soil
Depth Strata Munsell Description Composition Inclusions
3 8 cmbd A 7.5 YR 2.5/1 black sandy- none noted
8 20 cmbd B 5 YR 3/1 very dark sandy- pebbles, broken
gray clayey-silt shell
20 25.5 cmbd C 7.5 YR 2.5/1 black clayey-silt fine-crushed shell,
25.5 42 cmbd D 5 YR 2.5/1 black sandy- shell midden
42 53 cmbd E 7.5 YR 3/1 very dark sandy- shell midden,
gray clayey-silt small balls of clay 55 60 cmbd F 10 YR 3/2 very dark Some clay, shell midden
grayish brown sandy-silt 60 60.5 cmbd G mottled 10 YR very dark clayey- pebbles, crushed 2/2, 3/2 brown, very sandy-silt and whole shell dark grayish
60.5 65.5 H mottled 10 YR very dark clayey- less shell than
cmbd 2/2, 3/2 brown, very sandy-silt Stratum F
dark grayish
65.5 70 cmbd I mottled 10 YR very dark clayey-silt no inclusions 3/2, 2/1, 3/3 grayish
brown, black,
dark brown
65.5 70 cmbd J mottled 10 YR very dark sandy- base of shell
3/2, 3/3, 2/1 grayish clayey-silt midden
brown, dark
brown, black


Work conducted in the early and mid-1980s produced a number of data sets

relevant to the understanding and reconstruction of the prehistoric environment of the

Santa Maria River drainage basin and the central region of Panama. Sites like

Monagrillo and Zapotal occupied specific niches along the ancient coastline of Parita Bay


for sustained periods of time. The data sets recovered from Monagrillo during the excavations by Willey and McGimsey (1954) and the Proyecto Santa Maria (Cooke and Ranere 1984) are all that will ever be available from that site. A number of archaeologists have worked at Zapotal over the past 50 years. Zapotal was inhabited during the early Ceramic-A period and Monagrillo ceramics are the diagnostic artifact Fish and shellfish were important resources exploited by the site's inhabitants, however, there are still many unanswered questions. Questions concerning the human use of certain environments, the importance of particular animal species in the diet, procurement techniques, and the dietary contributions of aquatic protein resources, are now taking precedence over cultural chronology and ceramic typologies.

Chapter 4 addresses the vertebrate and invertebrate faunal assemblage recovered from Unit 4 during the 1999 field season at Zapotal. Methods used for sample processing, identification, and analysis will be discussed and zooarchaeological data generated about the faunal assemblage will be presented. Further discussion and interpretations of the data are included in Chapter 5.


The 1999 Faunal Assemblage

The vertebrate and invertebrate faunal remains analyzed for this study were

recovered during the 1999 field season at Zapotal. Zapotal is primarily a shell midden, defined as, "a cultural deposit of which the principal visible constituent is shell" (Waselkov 1987:95). The assemblage is from a 50cm x 50cm sondeo (test unit) that was part of a l-m-x-50-cm sondeo/column sample located in the northwest corner of a 2m x 2m excavation unit, Block 47 (S260 W60), Unit 45 (S268 W70). The sondeo comprised the west half of the Inm x 50Ocm area. (Figure 3.2); the remaining portion of the unit was left unexcavated. This unit had a total of twelve 5cm, arbitrary levels within ten natural strata. Each arbitrary level was assigned a distinct Provenience Designation (PD) number. Strata were recognized by the characteristics of the soil matrix, including soil color and mottling (using a Munsell color chart) as well as soil texture and inclusions (i.e., shells, pebbles, clay) (Table 3. 1). All materials were screened at the site using 1/4" and 1/8" mesh.

The faunal assemblage that forms the basis of this dissertation (n = 22,432) was chosen because it contains both vertebrate and invertebrate faunal remains recovered using the same technique. During the 1999 field season, the invertebrate remains from this unit were identified, analyzed and quantified by Eric Merges (I 999b) with assistance from the field crew. Many invertebrate remains were identified (n = 11,033), and it took



the greater part of the field season to complete. These invertebrate data will complement the data generated by my analysis of the vertebrate remains.

Methodology of Sample Analysis

Sample Processing

Vertebrate faunal assemblage

Several studies (Gordon 1993; Shaffer 1992; Wing and Quitmyer 1985) have

shown that assemblages screened with larger mesh sizes (1/2" or 1/4") are biased towards large animals (i.e., mammals), and give a skewed picture of the relative abundance and importance of one class of animals compared to another. The use of 1/8" and 1/16" meshes allows for a more complete recovery of delicate plant and animal remains (i.e., small fishes, shrimp mandibles, and plant remains), while using finer sieves insures the recovery of many types of seeds. While using 1/8" mesh at a shell midden site increases the time needed for screening, the results are worth the effort. Based on Cooke's experiences (Cooke and Ranere 1999:105), faunal remains recovered with mesh sizes smaller than 1/8" tend to be comprised of smaller individuals of the taxa identified with larger meshes, and are difficult to identify with a high degree of accuracy. Not using 1/ 16" mesh does not appear to affect species richness and diversity.

Primary data. Materials were sorted and bagged in the field by artifact class (i.e., vertebrate fauna, invertebrate fauna, ceramics, etc.). Sorting the artifacts in this manner allowed for less processing time in the lab and kept potentially fragile remains (e.g., fish bones, charred macro-plant remains) from being damaged by less fragile artifacts (e.g., marine shells, lithics). A detailed analysis of the vertebrate faunal


assemblage was performed by the author at the Smithsonian Tropical Research Institute (STRI) and the University of Florida. A number of Osteichthyes specimens were identified by Richard Cooke and Mdximo Jim~nez, both of STRI. Standard zooarchaeological procedures were used in this study following Reitz and Wing (1999). The first visual sort of the remains entailed dividing them into classes within each level. Some elements were identified to the appropriate taxa using the comparative collection at STRI, and the remainders were identified with a comparative collection on loan from STRI, as well as the Zooarchaeological Comparative Collection located at the Florida Museum of Natural History.

All vertebrate specimens were identified to genus and species when possible.

Identified elements were sided (i.e. left, right, axial) where appropriate. Any evidence of use-wear, thermal alteration or butchering was noted. Weights and Number of Individual Specimens (NISP) were recorded for all vertebrate specimens. All primary and secondary data were entered into EXCEL spreadsheets and are presented in Appendix A.

Secondary data. Sample biomass estimates were calculated for this assemblage using the archaeological specimen weights using the regression formulae described below. Sample biomass refers to the estimated total weight represented by the archaeological specimen (Reitz and Wing 1999). Calculating the biomass of an animal requires data on the correlations between skeletal weight and total weight (Casteel 1974; Hale et al. 1987; Reitz et al. 1987; Reitz and Wing 1999). These data are collected on modern specimens for application to biomass estimates. At this time, the comparative skeletal collection housed at STRI is not ontogenetically complete, thus many speciesspecific applications are not possible (Cooke and Ranere 1999:112). However, biomass


was calculated using values at the family or class level. These data are important for preliminary interpretations.

For the vertebrate portion of the study assemblage, biomass was estimated using specimen weight in the following allometric formula (Reitz and Wing 1999:224):

Y= e


Y logoo a +b logoo X)


Y = the estimated sample biomass (kg) contributed by the archaeological
specimen for a taxon
X = specimen weight of the archaeological specimens for a taxon
a = the Y-intercept of the linear regression line
b = slope of the regression line

Biomass for each vertebrate taxon was calculated using values from Reitz and Wing (1999:72) and Wing (2001). General class and/or family values were used in cases where values for specific taxa were not available.

Measuring relative abundance is the zooarchaeologist' s principal objective in the collection and quantification of faunal remains. Arguments have been made both for and against particular quantification tools, with a common consensus that there is no one perfect strategy (Grayson 1984; Jackson 1989; Nichol and Wild 1984; Reitz and Wing 1999). Data should be quantified using tools that will yield the most information from the assemblage. Counts, weights (where available), biomass, and MINI estimates for this assemblage will be presented.


The Minimum Number of Individuals (I) was determined using standard

procedure: the most abundant diagnostic element of each taxon was counted as the MINI (Grayson 1984; Reitz and Wing 1999). If this element was a paired element (left and right), then the higher count of the two was used. Size differences were also taken into account when appropriate. MINI was determined for each taxon within each level and recalculated for each stratum.

Invertebrate faunal assemblage

The invertebrate assemblage from the Unit 45 sondec was also recovered using 1/8" and 1/4" mesh. Specimens that were either whole shells (gastropods and bivalves), hinges (bivalves) or apices (gastropods) were collected, and all other shell fragments were discarded. The hinges and apices are the most diagnostic element of invertebrates, and thus can be included in MMI counts. In previous field seasons, specimens from sondeos and excavation units were recovered using 1/4" screen, with "representative grab samples" of shell being collected by the excavator based on recognizable specimens. While this allowed for identification of invertebrate remains from these proveniences, discussion was limited to the presence or absence of taxa. The recovery methods employed in the 1999 field season allow for a discussion of the relative abundance of mollusks as well as a more informed interpretation of seasonality and environmental conditions.

Once recovered from the sondeo, the invertebrates were transported to the field laboratory where they were washed and air-dried on window screens. Field school student Eric Merges (I 999a, b) was in charge of the invertebrate processing. At first,


each specimen was washed with water and brushed with a soft toothbrush. We quickly realized that this method was time consuming as it took four people three hours to process a single bag of specimens (approximately one liter in volume) (Merges 1 999b). It was decided that the shells did not need to be thoroughly washed for identification purposes, so a mass-washing technique was employed. Specimens within a single level were placed in a bucket of water, and water from a garden hose was used to agitate the water in the bucket, removing soil from the insides of the shells. The shells were then drained into a 1/8" mesh box and rinsed with running water to remove loose soil. Finally, the shells were placed on a 1/ 16" window screen to be air-dried.

Primary and secondary data. Once dry, the shells were sorted by class (Gastropoda, Bivalvia) and then identified to genus and species when appropriate. Bivalves were further separated into left hinge and right hinge categories. The specimens were identified by Merges, who had previous experience with the Zapotal invertebrates in the laboratory at the University of Florida (Merges 1999a). A modem comparative collection of live invertebrates of more than 400 individuals from 113 species was created in 1996 for the Zapotal project by Marcos Alvarez (L. Norr personal communication, 2001). The collection was designed to include the species listed from the Monagrillo site (Greengo 1954), and from comparisons with the Zapotal midden (Norr 1998). The taxonomic nomenclature was verified using Keen's (1971) Sea Shells of Tropical West America: Marine Mollusks from Baja California to Peru, as a reference. This reference was also used in the field in 1999 as an identification aid. The nomenclature in this study follows Turgeon et al. (1998) and Keen (1971) when taxa are not listed in the former reference. A more recent publication (Skoglund 2001) on


Panamic bivalves and chitons has become available since this study was done, but the overall nomenclature for the invertebrates has not changed.

The NISP were recorded for each taxon, but not the weight. Weights were not recorded for several reasons: 1) the field scale was not appropriate for shell weights; 2) only diagnostic elements of mollusks (i.e., hinges, apices) were collected from the unit (versus complete recovery of all invertebrate fragments); 3) comparisons of relative abundance of invertebrates and vertebrates using weight will give skewed results; 4) MNI estimates for invertebrates and vertebrates allow for a more informed discussion of relative abundance in shell middens. The invertebrates are being stored at the field house in Paris de Parita, Panama, sorted by provenience designation and taxon.

The invertebrate MNI were determined using standard acceptable procedures: for gastropods, diagnostic elements include apices and columella; for bivalves, hinges were separated into left and right elements with the higher count of the two used as the MNI. Size differences were also taken into account during MNI estimations, when appropriate. MNI was determined for each taxon within each stratigraphic level.

Sources of Bias

Faunal assemblages that are recovered for studies do not include all of the materials that were originally deposited. Zooarchaeologists must identify possible sources of bias in potential studies. There are two types of biases common to (zoo)archaeological samples; those introduced as a result of taphonomic history and biases inadvertently introduced by the excavator and zooarchaeologist. These biases can be viewed as a continuum along the life-span of an archaeological assemblage, from


selection and deposition of food items by the prehistoric consumers to recovery of archaeological remains by the modem-day archaeologist.

Taphonomic processes. The taphonomic history, i.e., the sum of all conditions acting upon the remains of a dead animal, determines the extent of preservation of that animal in the archaeological record. Zooarchaeologists look to taphonomic processes to understand what has aided or inhibited a particular assemblage's preservation, and to gain a perception of what may have been lost. Taphonomic processes that can affect bone and shell assemblages include: differential preservation, weathering, site inundation, erosional forces, redeposition, trampling, scavenging, human actions, soil pH, and plant intrusion (Davis 1987; Klein and Cruz-Uribe 1984; Lyman 1994; Reitz and Wing 1999).

Probably the single most important taphonomic process that operates on faunal

assemblages is differential preservation. In many archaeological sites, faunal remains are highly degraded or not recovered at all. Faunal remains can be well preserved, poorly preserved, or only slightly altered depending on their specific osteological characteristics and the conditions of the surrounding environment. Osteological characteristics can include: chemical composition (bone vs. shell), relative maturity and size of the individual, diagnostic landmarks, density and friability. Environmental conditions that affect preservation are soil acidity, climate, geographical location, and the matrix from which the remains were recovered.

The type of deposit and the geographical location of the deposit will determine which taphonomic processes will be most destructive. In the neo-tropical lowlands of central Pacific Panama, taphonomic processes that must be considered include: soil pH, erosion, weathering, and disturbance/dispersal by non-human scavengers. At Zapotal,