VT-CH-94 Vermont’s Earliest Known Agricultural Experiment Station

man in the northeast 19:1980 pp. 73-82

M. Pamela Bumsted
University of Massachusetts-Amherst

The first documented prehistoric corn remains from Vermont, in association with abundant gathered foods, were recovered in 1978 from an early Late Woodland (ca., A.D. 1450) site on the Winooski River intervale in the Champlain Lowland. The presence of both cultivated and gathered foods suggests these early peoples were not relying on either a cultivated or gathered economy, but on one which combined the nutritional and energetic advantages of both. This generalist strategy seems to continue in the Northeast until European contact while other areas of North America developed a specialized corn economy, resulting in increased disease stress for the populations involved. Continued biological / archeological research focusing on diet and disease problems is necessary to explain the Northeast as a unique physical / cultural / biological environment in North America for the adoption of and adaptation to maize agriculture.

Archeological anthropologists in North America have long been interested in the transition of prehistoric peoples from a hunting and gathering economy to one of cultivated crops. Adoption of horticulture implies a change in the human environment. The physical environment must be modified for successful cultivation. The cultural environment, whatever its prior state of flux, is subsequently altered by the new regimes of seasonal activity, social organization and aggregation of labor, increased sedentism, and often a wider social interaction network with other peoples. In North America, horticulture involves maize as the pre-eminent cultigen. It is often assumed by anthropologists that because maize offers a food source that is dependable, abundant, tasty, and subject more to human control than other foods available in the environment, that horticulturalists will focus primarily on maize as a staple and that food preferences will shift to maize away from former, gathered foods. In short, the implication exists that maize cultivation is such a biological and cultural advantage over other economies that prehistoric groups will emphasize maize unless restricted by the plant’s biological requirements.

However, maize, by itself, is recognized as a nutritionally poor staple. If dependency becomes great enough, the associated biological impacts can include pellagra, retarded growth, anaemia, and an increased susceptibility to infectious disease. Furthermore, mortality in the population may increase as malnourishment reduces the individual’s ability to cope with biological stress.


Studies of Midwestern (Lallo, Armelagos, Mensforth 1977; Lallo, Armelagos, Rose 1978; Mensforth, et al. 1978) and Southwestern (El-Najjar, et al. 1976) populations by biological anthropologists conclude that a dependency on maize had arisen in those regions and that nutritional problems followed. An association of increased infectious and nutritional stress reflected in increased skeletal lesions, increased morbidity, increased mortality, and interruptions in skeletal growth has been demonstrated for the horticulturally dependent groups at Dickson Mounds, Illinois, that was not found in the earlier hunting and gathering groups.

In the Northeast, archeological evidence indicates a generalist strategy of horticulture and hunting and gathering was used at the time of European contact (Thomas 1976), while the Midwestern and Southwestern groups continued to develop a specialist strategy dependent on maize horticulture with consequent biological impacts to the populations involved. The choice of whether to follow a generalist or specialist subsistence strategy is a critical one. Any population, of whatever time period, must meet the nutritional needs of its members, needs which change with their environment and biological state. The Northeast presents a region with different environmental conditions than the Midwest or Southwest and a different mode of adaptation involving maize agriculture. The Northeast, then, is a region in contrast, which, if studied, will elucidate the nature of horticultural adaptations.

The question of the biocultural nature of this transition from a hunting and gathering to a horticultural economy and its impacts on humans may be answerable through stable carbon isotopic composition (deltaC13) of skeletal remains (Bumsted 1980b). However, there are few skeletal samples available for study from transitional Northeastern sites. This paper will present an alternative to osteological examination by the study of organic remains from the Donohue site in Vermont. The site contains the first documented evidence for prehistoric horticulture in Vermont. We believe it also represents an occupation at a time when cultigens were being incorporated into regional economies. Thus, at the Donohue site we are witnessing the prehistoric evaluation of two distinct subsistence strategies—a reliance on gathering versus one dependent on cultivation.

VT-CH-94, the Donohue site, is located on the Winooski River intervale of the Champlain Lowland near Burlington, Vermont (Fig. 1). The Burlington area is one of the fastest growing in Vermont. There has been rapid, recent development, but most of the area is still undisturbed by massive construction. The site was tested in the summer and fall of 1978 as part of a Phase II archeological study for the Vermont Agency of Transportation (Thomas and Bumsted 1979). Through successful collaboration between the Agency and the archeologists, the highway was redesigned to avoid impacts to the site.

The prehistoric occupation occurs on a fine sandy loam levee adjacent to an unnamed slough. It seems likely from our soils studies

74 VT-CH-94

VT CH 94 Fig 1
Figure 1 Location of VT-CH-94 on the Winooski River Intervale


that the slough is a former Winooski River channel. Forty-five 35cm-square shovel test pits, 2 50cm-square units, 7 1m-square units, and 1 14m-by-50cm trench were excavated. These manual excavations were supplemented by 12 backhoe trenches (86.1m total length) around the site concentration perimeter to expedite testing of the cultural horizon which lay under 30-50cm of more recent silty alluvium. Twenty-three prehistoric features (see Table 1) and a lithic workshop were identified within this 35m-square area of intensive testing.

Although carbonized butternuts have been submitted for radiocarbon dating, the results are not yet available. On the basis of artifact typology—the presence of triangular “Levanna” projectile points, but particularly the use of incised design elements on the pottery—we suggest that the site post-dates A.D. 1000. Two distinct levels of quartzite flakes within the workshop area as well as the apparent overlapping of several features indicate that this is a multi-component site. Whether the stratigraphic gap represents a sequential habitation separated by a few months with an intervening flood deposit or a longer hiatus has not been determined.

The people who settled at VT-CH-94 had a subsistence economy based on the exploitation of horticultural crops as well as wild plants and animal food resources. Carbonized remains of such foodstuffs include acorns, butternuts, grapes, fish and maize. The association of these items indicates a late summer to early winter period of occupation.

Carbonized and charred butternut shells were the most frequently recovered organic remains. Butternuts, Juglans cincerea L., or “white walnuts” prefer rich woods and river terraces, tolerating moist conditions better than other walnut species. They are shade intolerant; where growing conditions are good, they may be locally abundant. The nuts mature in September and October, but usually stay on the tree until the leaves fall (Bailey 1915:1721; Clark 1958).

Amerindian and Colonial ethnobotanical uses of the butternut tree include: sap—for syrup; nut husk—for dyes and medicines; and inner and outer bark—as a purgative, emetic, styptic, anodyne, digestive aid, tonic, or “physic” (Clark 1958; Vogel 1970:288, 289). Whatever diverse uses may have been made of the butternut tree, we infer from the numerous shells that the prime use of the butternut at VT-CH-94 was for nutmeats.

Although remains of gathered butternuts are the most numerous food substances recovered, cultivated corn was also eaten. Beans were not recovered from the site. It should be recognized that, although only three carbonized corncobs and a few kernels were recovered, preservation of such foods is generally poor at any site. It may be assumed, therefore that maize (and perhaps beans) was more significant in the diet than the evidence implies. However, the presence of projectile points (and perhaps the scrapers and bifacial cutting implements) reaffirming the importance of hunting, the

Table 1 VT-CH-94 Feature Contents
Table 1 VT CH 94


overriding presence of gathered food remains, the pedological evidence of short-term, seasonal occupation, combined with the absence of substantial house structures and specialized corn implements, such as hoes or pestles would indicate cultigens were not the primary food staple.

This juxtaposition of two food resource types (gathered and cultivated) raises important theoretical questions. What was the process of integrating traditional gathering activities with the production of horticultural crops? Were the populations at this time weighing two, mutually exclusive options— gathering versus gardening— or had they developed a third option— relying on neither strategy to the exclusion of the other by adding cultivation to the range of choices available for responses to perturbations in the cultural and physical environments? No definitive answers exist, but the solution may lie in an understanding of several interrelated issues—energy expenditure and return; crop predictability; and nutritional quality.

Any social group must obtain sufficient resources to cover the energy costs of living (obtaining food, reproduction, growth, maintaining the social unit, etc.), plus an additional amount to counterbalance unplanned stresses, such as disease or a particularly hard season. A successful group cannot, therefore, usually afford to invest in a food crop that could not pay its own way by returning at least the energy cost of investment. Nor could a group afford to commit itself to an investment in an activity that reduces its options for future action should environmental conditions alter. Thus, energy allocation is one means of measuring the adaptive success of a group.

Butternuts, while generally predictable, only offer a good yield (commercially, from 1/4 to 1 bushel of clean seed per tree) every 2-3 years, with a light crop in between (Clark 1958). Competition from wildlife might also reduce the available nut crop for human consumption, but a judicious use of children’s tree-shinnying proclivities to dislodge the nuts might subvert the competition. Since all butternut stands may not have had the same cyclical period of high yield, however, returns could be kept high by scouting for the most productive areas.

Importantly, human energy need be concentrated on nut harvesting for only 2-4 weeks per year, with substantial returns expected. Finally, the charred shells at VT-CH-94 suggest that processing problems related to butternuts had been resolved. While the butternut-meat is difficult to extract by cracking and shelling, roasting the nuts first not only husks the nut, but breaks the shell as well, thus allowing easy recovery. In the process, nutmeats are dried, affording better storage properties. They would be immediately edible, or could be ground, boiled, and skimmed into an oily paste that could also be mixed with other foods. The association of heat-decomposed dolomite with butternut remains suggests the alkaline, MgCO3 properties of the rock were used to reduce the acidic nature of the butternutmeats, add magnesium to the diet, and perhaps aid in cooking. In short, butternuts, coupled with other wild plants, should have been a significant, storable, high yield resource.

78 VT-CH-94

What of maize? The average growing season in western Chittendon County is now about 150 days (Allen 1974) (this compares to 154 days in the Connecticut Valley of northern Massachusetts (Mott and Fuller 1967)). Even with a somewhat shorter growing season, corn horticulture during the Late Woodland period should not have been seriously limited by natural conditions. Maize, then, should have been a predictable, high yield resource. However, the investment of energy in environmental modification required when undertaking horticulture must have been great. Fields had to be cleared, soil broken up, seeds planted, sprouts weeded, and crops harvested, dried, and stored. Segments of any prehistoric community would have been tied to the general area for parts of the spring, summer, and fall; perhaps only on an ad hoc basis, during critical times in the maize growth cycle.

Table 2 summarizes the nutritional values of the identified foodstuffs at VT-CH-94. (Note that Atlantic salmon, as with beans, was not identified at the site, but is included as a representative, though unusual, fish. The food value data for butternuts are incomplete, so the English walnut data have been taken as the minimum food value for comparative purposes.) It can be seen that, except for carbohydrates and vitamins A and C, the butternut per weight is a superior food to corn. Greens and berries containing vitamin C would be plentiful in a riparian environment, however, and fish would supply vitamin A. Significantly, butternuts offer a 6 1/2 times greater supply of energy (kJ) per unit than maize. Thus, even if we assume that the total corn returns were higher than what could be gathered in the wild— a fact which VT-CH-94 does not appear to confirm— would the human energy expenditure needed for farming be worth it? The answer is obviously yes at some time, or we would not have recovered the cultigens.

We do know that populations in the Champlain Lowland had evidently made a choice by Late Woodland times to add corn (and presumably beans), with its storable properties, to the diet. It is apparent, however, that the people inhabiting VT-CH-94 were not highly dependent upon horticultural resources. Thus, the site may be typical of a transitional period of experimentation during which butternuts and maize were equally important parts of the subsistence base. Butternuts, per weight, are nutritionally more sufficient than maize and the trees need not be tended. Yet, the cyclical nature of the butternut harvest and the effort to gather significant, storable quantities would have imposed problems if large groups had tried to rely on this resource as a staple. Alternatively, the cultivation of maize is a labor intensive undertaking and must be carried out over several seasons of the year. During a period of incipient cultivation, the yields may not have always been as high as expected later in prehistoric agricultural development.

Together these foods are nutritionally complementary, with corn supplying the carbohydrate deficiencies in nuts, while the nuts provided far higher amounts of protein and energy than maize. The Late Woodland Champlain peoples had greatly expanded their nutritional


Table 2a VT CH 94

Table 2b VT CH 94
Table 2
Nutritional Composition (per 100 g edible portion) Of Postulated and Recovered Foodstuffs at VT-CH-94


threshold through the use of both cultivated and gathered resources.

However, we need to know more of the early Late Woodland adaptations in the Winooski intervale. For example, was the Donohue site occupied coevally with VT-CH-95 which lies just on the other side of the slough? On the basis of soils and aerial photography, we believe occupation in this area of the Champlain Lowland extended at least 110m by 287m (see Fig. 1). Soil microanalysis, including trace elements and pollen analysis, would help identify areas of differential cultural activity, i.e. where were the gardens? One possible explanation for the differences in stratigraphy and cultural remains between the outlying areas of VT-CH-94 and the site concentration area is that the lower areas were the locations of the gardens. This would make good agronomic and biological sense. Corn, even with intertilled beans, tends to deplete soil nutrients. Annual spring alluvial deposits would restore some soil fertility. The higher, better drained, levee surfaces of the site concentration would afford more healthful living conditions.

We do not know if the subsistence emphasis changed over time in Vermont. Would an increasing dependency on maize have to develop as populations expanded? Are nutritional and infectious diseases always associated with the introduction of maize to a population or are such disease correlates unique to Midwestern and Southwestern populations? On the other hand, if the seeming retention of a generalist subsistence strategy in the Northeast is true, archeological research focusing on diet and disease problems as has been illustrated in the Winooski intervale is necessary to explain the Northeast as a unique physical/ cultural/biological environment in North America for the adoption of and adaptation to maize agriculture.

It is not necessary to rely only on skeletal remains to study problems of human adaptation and biocultural interactions (Bumsted 1980a). I make a special plea, then, for biological and archeological anthropologists to incorporate the data base of the other in their own studies of the Northeast.


The following radiocarbon dates have been obtained (uncorrected):
GX-6297 2,930 +/- 115 (wood charcoal; Feature 3, Unit 2)
GX-6298 250 +/- 115 (butternuts; Feature 3, Unit 2)
GX-6299 510 +/- 125 (butternuts; Feature 9, Unit 2).

The first date is surprisingly early (no cultigens were recovered from this unit of the multi-component feature) but does indicate a long period of prehistoric exploitation of this area of the intervale. The other dates (mid-15th century) are somewhat later than we had expected but do reinforce the conclusions of this paper regarding Late Woodland subsistence in the Champlain Lowland.


I wish to thank Elena Filios, Rebecca Huss-Ashmore, Debra Martin, Peter Thomas, and Tom Ulrich for their helpful criticisms of the initial drafts. My gratitude also goes to Herbert Vogelmann, Professor of Botany, University of Vermont and Lawrence Kaplan, Professor of Biology, University of Massachusetts, Boston, for botanical identifications.

References Cited
Allen, George W. 1974. Soil Survey of Chittendon County, Vermont. USDA: SCS.

Bailey, L.H. 1915. The Standard Cyclopedia of Horticulture. NY: Macmillan Co.

Bumsted, M. Pamela. 1980a. 31 Facts About Physical Anthropology in Northeastern Archaeology. In, Proceedings of the Conference of Northeastern Archaeologists, James A. Moore, ed. Department of Anthropology Research Reports 19:189-191. And in, Biocultural Adaptations: Comprehensive Approaches to Skeletal Analysis, Debra L. Martin and M. Pamela Bumsted, eds. Department of Anthropology Research Reports. Amherst: University of Massachusetts. In press.

Bumsted, M. Pamela. 1980b. The Potential of Stable Carbon Isotopes in Bioarchaeological Anthropology. In, Biocultural Adaptations: Comprehensive Approaches to Skeletal Analysis, Debra L. Martin and M. Pamela Bumsted, eds. Department of Anthropology Research Reports. Amherst: University of Massachusetts. In press.

Clark, Bryan F. 1958. Silvical Characteristics of Butternut. USDA: Central States Forest Experiment Station, Misc. Release #23.

Diem, K. and C. Lentner, eds. 1970. Documenta Geigy: Scientific Tables. Basle: J.R. Geigy, S.A.

El-Najjar, M.Y., D.J. Ryan, C.T. Turner, II, and B. Lozoff. 1976. The Etiology of Porotic Hyperostosis Among the Prehistoric and Historic Anasazi Indians of Southwestern United States. American Journal of Physical Anthropology 44:477-488.

Kirschmann, John D. 1975. Nutrition Almanac. NY: McGraw Hill.

Lallo, J., G.J. Armelagos, and R.P. Mensforth. 1977. The Role of Diet, Disease and Physiology in the Origin of Porotic Hyperostosis. Human Biology 49(3): 471-483.

Lallo, J., G.J. Armelagos, and J.C. Rose. 1978. Paleoepidemiology of Infectious Disease in the Dickson Mounds Population. Medical College of Virginia Quarterly 14(1): 17-22.

Mensforth, R.P., C.O. Lovejoy, J.W. Lallo, and G.J. Armelagos. 1978. The Role of Constitutional Factors, Diet and Infectious Disease in the Etiology of Porotic Hyperostosis and Periosteal Reactions in Prehistoric Infants and Children. Medical Anthropology 1:1-57.

Mott, John R. and Donald C. Fuller. 1967. Soil Survey of Franklin County, Massachusetts. USDA: SCS.

Thomas, Peter A. 1976. Contrastive Subsistence Strategies and Land Use as Factors for Understanding Indian-White Relations in New England. Ethnohistory 23:1-18.

Thomas, Peter A. and M. Pamela Bumsted. 1979. Burlington-Colchester M5000 (3) Connector: Phase II Intensive Archaeological Survey. University of Vermont: Department of Anthropology Report #6.

Vogel, Virgil T. 1970. American Indian Medicine. Norman: University of Oklahoma Press.

Social Bookmarks:

Comments are closed.