Flotation samples analyzed from the 1986 excavations represent the processing of 500 liters of feature fill. A total of 504.62 g of plant remains was recovered from these samples, including 396.52 g of wood and stem charcoal and 88.06 g of plant food remains. Carbonized fragments of root or rhizome, one unidentified bud, one pedicel or peduncle, and other unidentified fragments also were found in the flotation samples (Table 38).
In general, analysis of plant remains from the 1986 season at Fredricks confirmed earlier interpretations of plant use at the site. The following discussion will, for the most part, depend upon the cumulative findings of four field seasons. This comprehensive data set is more useful as a basis for interpretation of plant use since its division into sets by date of excavation is an artificial one imposed by researchers and has no direct relevance to the activities of site occupants. However, several types of plant remains not recovered in previous years deserve special mention.
The most important of these is watermelon (Citrullus vulgaris Schrader ex Ecklon and Zeyher). One watermelon seed was found in a flotation sample from Feature 45 (Table 42) and an additional seed was sorted from a waterscreened sample from Zone 2 of Feature 47. Like peach, watermelon was introduced to North America by Europeans. It was apparently first grown in North America by Spanish colonists in the coastal Southeast as early as the late sixteenth century. Watermelon seems to have reached the Atlantic Coast colonies somewhat later by way of the West Indies (Blake 1981:194). Unlike peach, which originated in Asia, watermelon is thought to be native to Africa. Its adoption by post-contact Native American groups of the Eastern Woodlands was no doubt facilitated by the aboriginal practice of cultivation of New World cucurbits. The only other occurrence of watermelon from North Carolina is from Upper Saratown (31Sk1a), a site roughly contemporaneous with Fredricks on the Dan River in Stokes County, from which a single seed was reported (Wilson 1977). It is difficult to assess the extent of use of watermelon at Fredricks. Because of its thin rind and the fact that the fruit would not have been dried for storage or cooked for consumption, watermelon remains are likely to be underrepresented archaeologically relative to the frequency of its use.
Other plant taxa not recovered during previous seasons at Fredricks are strawberry (Fragaria virginiana Duchesne) and amaranth (Amaranthus sp.). Strawberry seeds are minute and have not been recovered frequently from Eastern archaeological sites. Strawberry plants probably grew near the village in old fields and woods edges, which are the species's preferred habitat today (Radford et al. 1968:533). The food value of this species is somewhat limited by the size of the fruits, but strawberry could have been a useful diet supplement or "snack food" in the spring and early summer, when most other fruit-producing species are still maturing.
Species of amaranth are common weeds in fields and disturbed ground. Although cultivated varieties were developed in Mesoamerica and are still used there today, amaranth is seldom reported from North American sites. The presence of a single seed in Feature 51 at Fredricks is best explained by the plants having grown near the site and their seeds having been carried into a fire by wind or, unintentionally, by people. Although amaranth could have been used as spring greens, the seeds would have been present in the fall.
Most types of plant food remains recovered during 1986 had been recovered in previous seasons at the site and occurred in similar proportions. Hickory (Carya sp.) nutshell was the most abundant nutshell type by weight at 97.1%, followed by acorn (Quercus sp., 1.7%), walnut (Juglans nigra L., 0.9%), and Juglandaceae (the family including both walnut and hickory, 0.3%) (Table 43). In addition, acorn meat was found in Feature 53, including one whole carbonized acorn.
Comparison of weights of the inedible portion of different nut types (the shell) may, however, be misleading because of different ratios of "meat" (that is, the edible portion comprised of embryo and cotyledons) to shell. The difference in meat-to-shell ratio between acorn and hickory in particular can be quite large. Studies have shown that a given quantity of acorn shell can represent anywhere from five to 200 times as much food as an equal quantity of hickory nutshell (Lopinot 1983). If the total quantity of acorn shell is multiplied by 50 (a factor suggested for general use by Yarnell and Black [1985]) and divided by the quantity of hickory shell, an estimated ratio of acorn to hickory meat of 0.87 is obtained. In other words, the representation of edible quantities of acorn and hickory is similar for the 1986 sample.
Site totals to date, however, yield an acorn-to-hickory ratio of 1.67, although percentages of nutshell types are similar to those for the 1986 season alone (Table 43). Despite the difficulties of interpretation involved, it can be stated confidently that acorn and hickory were used extensively by the Fredricks site people, with a possible bias in favor of acorn. Walnut was of only minor importance, perhaps because of the comparatively high effort required in processing it compared to hickory (Talalay et al. 1984) and/or its scattered distribution in the Piedmont (Radford et al. 1968).
Ubiquity values for 1983-1986 seasons (as percentage of features in which a taxon is represented) rank hickory first and acorn third (after maize) (Table 44). The values are, however, quite close and probably also reflect acorn shell's lower preservability. A ranking of nutshell occurrences by ubiquity results in the same relative order as ranking by weight (hickory first, acorn second, and walnut third). Considering acorn's high meat-to-shell ratio and lower shell preservability relative to hickory, the interpretation that acorn was perhaps a larger dietary component than hickory nuts at Fredricks and probably of approximately equal importance is a reasonable one. Walnut was only a minor food.
The relative subsistence importance of these three nut types is relatively easy to determine because we have some knowledge of the meat-to-shell ratios of two of the genera and also an understanding of preservation factors that might influence representation of nutshell types. Trying to compare nuts to cultigens or different types of cultigens to each other, on the other hand, is subject to considerable difficulties. Hickory nutshell is dense and durable and likely to be preserved through carbonization, unlike more fragile remains such as Cucurbita rind. Maize cupules and rachis fragments are fairly dense, whereas the kernels are starchy and less durable. Common bean is more likely to be prepared by boiling than by parching or roasting, so the seeds are less likely to become carbonized. The three Mesoamerican cultigens found at the Fredricks site (maize, Zea mays L.; pepo squash, Cucurbita pepo L.; and common bean, Phaseolus vulgaris L.) therefore produce archaeological remains that can be difficult to interpret.
However, of the three Mesoamerican cultigens, maize was certainly the most important. Although the percentage of maize remains is quite low compared to most other types of plant foods from the 1986 season at 4.7% (Table 40), maize cupules and kernels are still much more abundant than common bean cotyledons or cucurbit rind (both less than 0.05%). Cucurbit seed weights are included in totals for all seeds, but are neither numerous nor heavy and would not affect this ranking. Maize quantities are higher for site totals (Table 45) at 28.4% of plant food remains for all feature types, compared to less than 0.05% for Cucurbita rind and 0.3% for common bean.
The relative prominence of maize remains by weight can be accounted for in part by the fact that the maize cob (botanically, the rachis) and cupules are durable and useful as fuel, contributing to the likelihood of their carbonization. However, kernels alone still comprise 2.5% of plant food remains and make up more than half of the total maize recovered in the 1986 sample. Site seed and fruit totals strengthen the interpretation that maize was the most important of the Mesoamerican cultigens at Fredricks, since maize kernels comprise 53.8% of total identified seeds and fruits (Table 46). Thus maize far outranks any other seed type found at the site, with grape a distant second at 7.7%. The best way to compare plant food remains of different types, such as seeds and nutshell, is to consider ubiquity. Site totals for Fredricks (Table 44) indicate that maize remains occurred in 86.5% of features sampled and was exceeded only by hickory, which occurred in 92.3% of features. In comparison, common bean and pepo squash rank tenth and twelfth, respectively.
The importance of maize to the Fredricks site population as a staple is confirmed using several methods of comparison and quantification. What quantities of common bean and pepo squash were used by this group is impossible to tell. Lawson (Lefler 1967:82-83, 182) refers to several types of legumes and squashes grown by Indians in the Coastal Plain and Piedmont, so presumably both were of some dietary importance during the Historic period. Preservation and depositional factors have probably resulted in the underrepresentation of these cultigens relative to their actual importance. Although maize was more important than common bean or pepo squash (probably used in quantities similar to acorn and hickory nuts), the magnitude of difference in importance is impossible to assess.
Only two Old World domesticates, watermelon and peach (Prunus persica L.), were found at Fredricks. Watermelon is discussed above along with other taxa first discovered in the 1986 sample. Like peach, watermelon is somewhat weedy, being capable of colonizing highly disturbed habitats. Watermelon occurs today as a waif in waste places (Radford et al. 1968:999); although it can germinate successfully without human aid, it requires some husbandry in order to maintain a population. Watermelon has growing requirements similar to those for other cucurbits, and was probably easily incorporated into the aboriginal gardening system. Peach is frequently found today as an escape from cultivation (Radford et al. 1968:566). Native to Asia, the peach was first imported into the New World by the Spanish as a mission crop in the sixteenth century (Sheldon 1978). The English brought peach pits to the Massachussetts Bay Colony as early as 1669 (Hedrick 1972:463). In part because of its weedy properties, peach may have been dispersed somewhat independently of direct aboriginal/European contact. A number of European observers (Salley 1911; Hedrick 1972:463) noted peach trees growing "wild" in the Southeast. However, they may not have recognized signs of limited husbandry.
Although peach trees can grow and produce fruit without human intervention, they were probably tended to some extent at the Fredricks site, at least through removal of plants competing for light and nutrients and perhaps through planting as well. Fredricks site inhabitants may have tended, planted, or otherwise protected native fruit trees such as persimmon before contact, although there is no direct evidence for such practices. Peach trees produce fruit in three to five years after germination (Sheldon 1978) and relatively little investment of time or energy would have yielded large amounts of palatable fruit (reported by Lawson to have been dried and made into cakes for later consumption [Lefler 1967:217]). The stony endocarp, or pit, of the peach fruit is quite amenable to carbonization and comprised 1.3% of plant food remains in the 1986 sample (Table 40) and 3.4% of the total plant food remains from the site (Table 45). Although it was probably an important fruit crop, peach was a dietary supplement and not a staple. Considering the weight and durability of the pit, greater representation would be expected if peach were as important as a food such as hickory nut. Preparation of the fruit by drying also should make peach pit overrepresented relative to some other types of plant food remains. However, peach pit ranks high by ubiquity (ranking fourth at 48.1% of features, Table 44), which indicates that it was probably important relative to other (wild or semi-cultigen) fleshy fruit types. High preservability probably increases the contrast in representation between peach and other fleshy fruits.
The indigenous fleshy fruits recovered from the Fredricks site are mostly heliophilic species or genera that favor disturbed ground or edges between wooded and open areas. All of them generally produce greater quantities of fruit in these kinds of habitats than in closed-canopy situations. Taxa found at Fredricks such as bramble (Rubus sp.), sumac (Rhus sp.), strawberry, elderberry (Sambucus canadensis L.), hawthorn (Crataegus sp.), persimmon (Diospyros virginiana L.), blueberry (Vaccinium sp.), viburnum (Viburnum sp.), and grape (Vitis sp.) all indicate some degree of forest opening (Yarnell 1982:5) because of their preference for gaps in the forest canopy. Thus there is strong evidence for a symbiotic relationship between humans and these taxa, probably something on the order of Rindos's (1984) incidental or specialized domestication, in which humans increase habitat areas for useful plants and disperse their seeds incidentally to consumption of the fruits.
For some taxa there is a stronger case for prehistoric domesticatory relationships. Maypops (Passiflora incarnata L.) has been considered a quasi-cultigen because of its close association with humans in eastern North America prehistorically (Yarnell 1987). Even tree fruits such as persimmon have had similar long-standing relationships with human groups. It is possible that some management of fruit trees such as persimmon was practiced prehistorically and helped facilitate the adoption of peach as a tree crop. The quantities of persimmon seeds found at Fredricks (Table 46) indicate that use of this species was common, although its seed's durability may skew its apparent importance relative to other fruit types. However, no direct evidence for management of fruit trees has been found. An expected morphological criterion for domestication of edible fruits is increased fruit size, a change that cannot be studied at most open sites using archaeological evidence since fleshy fruit parts are ususally destroyed when burned. Management of some kind does seem likely given the long-standing relationship between populations of humans and persimmon trees in the East, but may have been somewhat casual by European criteria.
Numbers of fleshy fruit seeds recovered during the 1986 season appear in Table 42. Of fleshy fruits, grape comprises the greates percentage of total identified seeds and fruits (7.7%), followed by maypops (6.4%) and persimmon (3.7%). Persimmon has from three to eight seeds per fruit and grape one to four, whereas maypops has many. Although "minimum number of individuals" or some similar measure has not been calculated for seed and fruit types, number of seeds per fruit is a factor that should be considered. Therefore, maypops may be overrepresented relative to persimmon and grape. However, calculation of ubiquity ranks these fruit types similarly (Table 44) with maypops and persimmon reversed in rank order (but with very similar values). Thus it appears that grape, persimmon, and maypops were the most commonly used indigenous fleshy fruits at Fredricks. Other taxa that rank relatively high in numbers and ubiquity include bramble, groundcherry (Physalis sp.), and sumac. Other taxa that occurred in smaller quantities at Fredricks include viburnum, nightshade (Solanum sp.), blueberry, elderberry, strawberry, huckleberry (Gaylussacia sp.), and hawthorn.
Grain or weed seeds found at Fredricks include knotweed (Polygonum sp.), amaranth (discussed above), and chenopod (Chenopodium sp., found in previous seasons). All three of these genera include species that have been cultivated in North America prehistorically (although cultigen amaranth has not been recovered north or east of the Ozarks). The numbers of seeds of these taxa are quite low, and all occur as weeds today on disturbed ground. Thus there is no reason to assume that they were cultivated (or used) at Fredricks. The same can be said of poke (Phytolacca americana L.), another weed used prehistorically as a source of greens (Yarnell 1983).
Several seed types are included in the "Miscellaneous" category. Most occurred in small numbers and probably represent incidental inclusions in cultural deposits. Lespedeza sp., wood sorrel (Oxalis sp.), unidentified Type B (listed in previous reports as possible henbit, Lamium sp.), bearsfoot (Polymnia uvedalia L.), beggar's lice (Desmodium sp.), spurge (Euphorbia sp.), morning glory (Ipomoea sp.), and Nightshade family (Solanaceae), have been discussed in Gremillion (1986, 1987). Horse gentian (Triosteum sp.) seeds have been identified in the 1985 samples since publication of Gremillion (1986). This is an herbaceous genus in the Honeysuckle family (Caprifoliaceae) that grows in woods and openings on neutral or basic soils. T. perfoliatum L. was used as a coffee substitute by Germans in Pennsylvania (Hedrick 1972:576) but its use by aboriginal groups, if any, is not known. Bedstraw (Galium sp.) was used as a coffee substitute and as bedding in northern Europe (Hedrick 1972:285; Uphof 1968:236). Aboriginally some bedstraw species have ethnographically documented medicinal uses among some North American groups (Moerman 1986). It is not known how bedstraw was used at Fredricks, but it was found in relatively large quantities there (5.6% of total identified seeds, ubiquity 28.8% of features). Use of the vegetative parts of the plant as bedding would explain the presence of large numbers of seeds of this genus, which usually grows in wooded rather than open habitats.
Summary of Plant Food Remains
The paleoethnobotanical data from the 1986 field season at Fredricks support previous interpretations of plant use at the site (see Gremillion 1986). Maize was the most important crop, and common bean and pepo squash also were grown. Hickory and acorn seem to have been staples, although their contribution to the diet relative to that of maize has not been assessed. Since there was presumably a long tradition of human use of nut-producing trees in the Piedmont as elsewhere in the East, nut trees were possibly managed in some way, if only indirectly through protection and culling of competing species. In addition to the tropical Mesoamerican cultigens, the Fredricks site people had close relationships with various herbaceous and woody fruit-producing taxa growing in anthropogenic habitats. Management of such species probably spanned a continuum from toleration and unintentional habitat enrichment to protection and perhaps propogation as well. The only Old World domesticates grown at Fredricks, peach and watermelon, were both fleshy fruit crops.
Excavations to date have revealed no evidence of cultivation or consumption of indigenous starchy or oily grains such as sumpweed, maygrass, chenopod, or knotweed. The only occurrences of such grains at Fredricks are in such small quantities that there is no compelling reason to assume that they are anything other than weed seeds. In many parts of the East, grains like chenopod and maygrass declined from a utilization peak during Woodland times as maize became more important (Yarnell and Black 1985, Asch and Asch 1986, Fritz 1986). However, except for sumpweed, these grain crops (possibly only quasi-cultigens) continued to be used in Historic times in the East, at least at Cherokee sites in the Little Tennessee River valley (Chapman and Shea 1981).
Although indigenous grains are poorly represented at Fredricks, maygrass was found in large quantities at the Mitchum site on the Haw River, an Historic period site occupied slightly earlier than Fredricks (Gremillion 1987). Paleoethnobotanical data from prehistoric sites in the Piedmont will be needed to determine whether cultivation of these indegenous grain-producers was ever a Piedmont tradition as it was elsewhere in the Eastern Woodlands. Perhaps the Mitchum site maygrass represents the persistence of such a tradition into historic times. The Fredricks site people, if they ever had similar traditions (an important question since ethnic relationships between groups occupying sites like Mitchum and Fredricks are unclear), either abandoned them for reasons as yet unknown or carried them out in localities away from the village on the Eno represented by the Fredricks site. In any case, only further excavation can help answer these important questions about Historic period aboriginal subsistence.