Eastern North America as an independent center of plant domestication

Abstract

The status of eastern North America as an independent center of plant domestication has recently been called into question by a number of genetic and archaeological studies, which suggest that the region may not have witnessed the independent domestication of local crop plants, but rather may have been on the receiving end of domesticated crop plants introduced from Mexico. Here, I provide a synthesis of the currently available archaeological and genetic evidence from both eastern North America and Mexico regarding the spatial and temporal context of initial domestication of the four plant species identified as potential eastern North American domesticates: marshelder (Iva annua), chenopod (Chenopodium berlandieri), squash (Cucurbita pepo), and sunflower (Helianthus annuus). Genetic and archaeological evidence provides strong support for the independent domestication of all four of these plant species in the eastern United States and reconfirms the region as one of the world’s independent centers of domestication.

Between approximately 11,000 and 5,000 years ago, human societies in many different regions of the world brought a wide range of different species of plants and animals under domestication, marking the initial emergence of food production economies and the beginning of one of the major transitions in human history. This transition, often described as the “Neolithic Revolution” or the “Origins of Agriculture” has been an enduring area of inquiry in both archaeology and biology for more than a century. By 1940, Nikolai Vavilov had identified a total of seven primary centers of plant domestication, worldwide, based on areas of maximum genetic diversity in crop plants (1). Vavilov’s seven centers of origin have been subjected to considerable rethinking as a constant stream of genetic and archaeological studies, fueled by new approaches and technological advances, have substantially expanded and refined our understanding of the spatial and temporal contexts of initial plant and animal domestication (2–4). In just the last 5–10 years, several new independent centers of plant domestication have been identified in tropical forest regions of South America and Southeast Asia (5–7) (Fig. 1). At the same time that these Southeast Asian and South American centers have been recognized, however, the status of another, previously identified independent center of plant domestication has been called into question. A number of recent genetic and archaeological studies have led to suggestions that eastern North America, identified as a center of plant domestication 20 years ago (8), may not have witnessed the independent domestication of local crop plants but rather may have been on the receiving end of domesticated crop plants introduced from Mexico (9–14).

Fig. 1.

Currently recognized independent centers of plant and animal domestication.

Here, I reconsider eastern North America’s status as an independent center of plant domestication and provide a new synthesis of the currently available archaeological and genetic evidence from both eastern North America and Mexico regarding the spatial and temporal context of initial domestication of the four plant species identified as potential eastern North American domesticates: marshelder (Iva annua), chenopod (Chenopodium berlandieri), squash (Cucurbita pepo), and sunflower (Helianthus annuus). Genetic and archaeological evidence summarized here provides strong support for the independent domestication of all four of these plant species in the eastern United States and reconfirms the region as one of the world’s independent centers of domestication.

Marshelder (Iva annua)

Although marshelder is mentioned only briefly in passing (9) or described as an “extinct minor cultigen” (14) in several recent studies suggesting that domesticates first entered eastern North America from Mexico, this crop plant in fact plays an important role in establishing whether the eastern woodlands was an independent center of plant domestication. No longer cultivated as a domesticated crop, marshelder grows today as a wild plant across much of the eastern and central United States, and its range extends south into Tamaulipas, Mexico. No remains of marshelder, however, have ever been reported from archaeological contexts in Mexico. In contrast, small achenes^† and seeds representing wild I. annua plants have been recovered in abundance from numerous archaeological contexts across the eastern United states dating back as early as ca. 8,000 B.P., indicating that wild populations of marshelder were widely harvested by human groups for several thousand years leading up to its initial domestication (8).

The earliest evidence for domesticated marshelder in the eastern United States comes from a shallow Titterington Phase pit (Feature 20) at the Napoleon Hollow site in west-central Illinois, which yielded an assemblage of 79 carbonized marshelder achenes and seeds, including 44 that were complete enough to be measured (Fig. 2) (15). When achene measurements were adjusted to correct for shrinkage due to carbonization, this Napoleon Hollow assemblage of 44 I. annua specimens exhibited a 31% increase in mean length when compared with a reference class of 11 modern wild I. annua populations, providing clear evidence that they represent domesticated plants. An indirect date obtained on dispersed charcoal within Feature 20 indicated an age of 3,920 ± 90 B.P. in radiocarbon years B.P., consistent with the age of the Titterington cultural phase as documented elsewhere in Illinois (15). A recent direct accelerator mass spectrometry (AMS) radiocarbon date obtained on one of the marshelder achenes yielded the identical age in radiocarbon years (3,920 ± 40 B.P.) (Table 1), providing confirmation that I. annua had been brought under domestication in the eastern United States by ca. 4,400 B.P. The presence in eastern North America, by 4,400 B.P., of an indigenous domesticated crop plant provides a solid temporal boundary line for the arrival of any exotic outside domesticates if they are to be cast in a causal role. If the introduction of crop plants and associated food production technologies from Mesoamerica is to be identified as the source and “trigger” for a subsequent domestication of marshelder or any other “minor indigenous crop” (14) of eastern North America, then the Mesoamerican exotics would, of necessity, have to have arrived in the eastern United States before 4,400 B.P. Each of the other three plant species identified as indigenous eastern domesticates (chenopod, squash, and sunflower) have in fact recently been proposed as representing early introductions from Mexico.

Fig. 2.

Location of archaeological sites discussed in the text and the present-day geographical range of the three wild Cucurbita gourds identified as potential progenitors of pepo squash (Cucurbita pepo ssp. ovifera).

Table 1.

Direct AMS radiocarbon age determinations for the earliest occurrence of indigenous domesticated seed crops in eastern North America

Domesticated plant species	Age, AMS-calibrated calendar years B.P.		Age, radiocarbon years B.P.	Laboratory sample no.	Archaeological site and provenience
	Intercept	1σ age range
Pepo squash (C. pepo ssp. ovifera)	5,025	5,290–4,870	4,440 ± 75	ß 47293	Phillips Spring (Unit K2)
Sunflower (H. annuus)	4,840	4,860–4,830	4,265 ± 60	ß 45050	Hayes (Level 14)
Marshelder (I. annua)	4,400	4,420–4,290	3,920 ± 40	ß 216463	Napoleon Hollow (Feature 20)
Chenopod (Ch. berlandieri)	3,700	3,900–3,490	3,450 ± 150	ß 11348	Cloudsplitter (F.S. 1361
	3,640	3,840–3,460	3,400 ± 150	ß 11347	Newt Kash (El 1114)

Calendar calibrations are based on the Pretoria Calibration Procedure and the Intcal 98 Calibration database.

Chenopod (Chenopodium berlandieri)

A number of different wild species belonging to the genus Chenopodium grow today in eastern North America and as a group are referred to by a variety of common names, including “lambsquarter” and “goosefoot.” One of these species of chenopod, Ch. berlandieri, is of interest here because it is both currently grown as a domesticated crop plant in Mexico and was present as a pre-Columbian domesticate in eastern North America (8). Wild populations of Ch. berlandieri grow today over much of the United Sates and northern Mexico, as far south as Michoacán (16). At the southern end of its geographical range, three varieties of a domesticated subspecies of this plant (Ch. berlandieri ssp. nuttalliae) are still cultivated today (Fig. 2) (16), and their small indehiscent fruits can be distinguished from those of wild plants on the basis of a substantial reduction in the thickness of their testa or seed coat (8).

Based on observed similarities in fruit morphology between two of the modern Mexican cultivated varieties of domesticated Ch. berlandieri ssp. nuttalliae on the one hand (the chia variety, with a 10- 20-μm-thick testa, and the “naked” huauzontle variety, with a 2- to 7-μm-thick testa) and archaeological specimens recovered from across eastern North America on the other, it was suggested in 1990 that Ch. berlandieri was initially domesticated in Mexico and subsequently introduced as a domesticate into eastern North America (16). The suggestion that domesticated chenopod may have been initially introduced into the eastern United States from Mexico was recently reiterated (9), even though archaeological research in the last 15 years has provided compelling evidence to the contrary.

In eastern North America, archaeological evidence indicates that Ch. berlandieri was harvested as a wild plant as early as 8,500 B.P. (15) and was an important domesticated crop plant in the region for >3,500 years (from ca. 1850 B.C. to A.D. 1750) (8), with a number of regionally distinct cultivar varieties now recognized (17). AMS radiocarbon dating of uncarbonized thin-testa Ch. berlandieri fruits from Cloudsplitter and Newt Kash rockshelters in eastern Kentucky (Fig. 2) indicated conventional radiocarbon ages of 3,450 ± 150 and 3,400 ± 150 B.P., respectively (intercepts with the calibration curve of 3,700 and 3,640) (Table 1), providing the earliest available evidence for this species being grown as a domesticated crop plant in the East (8).

The archaeological record of Mexico also offers clear evidence of human harvesting of wild stands of Ch. berlandieri, along with a variety of other plants producing similarly sized small seeds, for thousands of years before the arrival of the Spanish.^‡ Despite careful and comprehensive analysis of extant collections, however, thin-testa fruits representing domesticated chenopod have yet to be found in any archaeological contexts in Mexico, leading McClung de Tapia and Rios-Fuentes^‡ to recently conclude that Ch. berlandieri was not present as a domesticated plant in Mexico before the 16th century. Interestingly, recent research in the south-central Andes has documented the independent domestication of a species of Chenopodium closely related to Ch. berlandieri, based on a reduction in testa thickness, beginning at approximately the same time that domesticated Ch. berlandieri first appears in the archaeological record of the eastern United States (18). This documentation of a second developmental sequence of domestication for the genus Chenopodium in South America underscores the absence of any archaeological record of domesticated Ch. berlandieri in Mexico.

A recent genetic study employing random amplified polymorphic DNA fragment analysis to identify genetic relationships among six Chenopodium taxa, including the Mexican domesticate Ch. berlandieri ssp. nuttalliae and a collection of wild Ch. berlandieri from Madison, Wisconsin, suggests a considerable amount of genetic separation between the modern Mexican domesticate and eastern North American wild samples (19). Given this apparent degree of genetic separation and recent advances in ancient DNA analyses, a three-way comparison of the modern Mexican domesticate and the ancient domesticate and modern wild populations of Ch. berlandieri from eastern North American would be worthwhile. I expect that when such a comparison is carried out it will provide further evidence that Ch. berlandieri was initially domesticated in eastern North America by 3,700 B.P., with a second independent domestication occurring in Mexico in the 16th century, >3,000 years later.

Squash (Cucurbita pepo)

Of the five species of squash (Cucurbita) that were domesticated from different progenitor species of wild Cucurbita gourds throughout the Americas (12, 20), only C. pepo is known at the present time to have been independently domesticated more than once (12, 21, 22), resulting in two present-day domesticate lineages, now classed as distinct subspecies. The first of these, C. pepo ssp. pepo, which includes cultivated pumpkins and marrows, was developed from an as yet undocumented wild progenitor in Mexico 10,000 years ago (12, 23, 24). The second subspecies (C. pepo ssp. ovifera, which includes cultivated crooknecks, acorn, and scallop squashes) has long been identified as a potential eastern North American domesticate (8).

Over the past two decades, a number of genetic studies have focused on identifying the wild ancestor (and location of domestication) of this proposed “eastern” lineage of pepo squash. Of the three taxa of wild Cucurbita gourds that have been considered as possible progenitors of ssp. ovifera, two have contiguous present-day geographical ranges north of Mexico (C. pepo ssp. ovifera var. texana and C. pepo ssp. ovifera var. ozarkana), whereas the third (C. pepo ssp. fraterna) occurs today in northeastern Mexico (Fig. 2).

In an initial molecular study focused on these progenitor candidates, the isozyme alleles of the cultivar members of C. pepo ssp. ovifera (C. pepo ssp. ovifera var. ovifera) were shown to be a subset of those in var. ozarkana, whereas the other two progenitor candidates (var. texana and ssp. fraterna) were both found to lack common alleles of the domesticates. As a result, var. ozarkana was designated the best candidate to be the wild ancestor of the “eastern” C. pepo lineage (25), and eastern North America was identified as the region of domestication.

A subsequent genetic study using an intron region from the mitochondrial nad1 gene as a marker, however, found that all three of the progenitor candidates (ssp. fraterna, var. texana, and var. ozarkana) shared identical mtDNA sequences at the nad1 locus and concluded that each should therefore be considered to have equal likelihood of being the wild ancestor of the ssp. ovifera lineage of domesticated C. pepo (12). In addition, if ssp. fraterna turned out to in fact be the wild ancestor the C. pepo ssp. ovifera lineage, rather than either of the north of the border wild gourds, then initial domestication could have taken place in northeastern Mexico, with the domesticate subsequently being introduced into eastern North America.

In contrast to the mitochondrial nad1 study, which could not distinguish between the three potential progenitor gourds, a higher-resolution approach involving phenetic analysis of random amplified polymorphic DNA fragment data was able to separate ssp. fraterna from var. texana and var. ozarkana, and effectively excluded ssp. fraterna from the cluster that includes all cultivated and wild varieties of ssp. ovifera (24). Further support for excluding ssp. fraterna from contention as a potential progenitor is provided by a recent genetic study employing amplified fragment-length polymorphism, inter-simple sequence repeat, and simple sequence repeat markers, in which ssp. fraterna was again placed at a greater genetic distance from the domesticates of ssp. ovifera than the eastern North American wild gourds (26). Molecular evidence thus now supports the Ozark wild gourd (C. pepo ssp. ovifera var. ozarkana) as the best candidate to be the progenitor of the domesticated squashes of the C. pepo ssp. ovifera lineage (22, 27). In addition, the archaeological sequences of both northeastern Mexico and eastern North America also provide substantial parallel evidence that ssp. ovifera was domesticated in eastern North America from an indigenous wild Cucurbita gourd.

In the eastern United States, a detailed archaeological record exists for the initial domestication and subsequent human selection, over time, for desired traits in C. pepo. Small C. pepo seeds and thin rind fragments recovered from archaeological sites across eastern North America indicate the widespread use of wild gourds of this species as early as 8,000 B.P. The earliest evidence of domesticated C. pepo squash in the East was recovered from water-saturated habitation layers of the Phillips Spring site in south-central Missouri, within the present day geographical range of the wild gourd C. pepo ssp. ovifera var. ozarkana (Fig. 2). Along with rind fragments and fruit stems, a total of 125 whole seeds and fragments identified as C. pepo were recovered from Unit K2 (the “squash and gourd zone”) at the site. Of these, 62 provided both length and width measurements (mean length, 10.5 mm; range, 8.3–12.2 mm), and 12 exceeded the 11.0-mm seed-length ceiling established for wild Cucurbita, providing clear evidence of domestication (28). Direct AMS dating of one of the Phillips Spring squash seeds yielded a conventional radiocarbon age of 4,440 ± 75 B.P. and a calibration curve intercept of 5,025 B.P. (Table 1). In contrast to the Unit K2 C. pepo seed assemblage, however, the thin rind fragments and four fruit-end peduncle scars from the squash and gourd zone at Phillips Spring show no indication of morphological changes beyond the parameters of the wild morphotype. They compare closely with var. ozarkana in both diameter and circular outline, indicating that the Phillips Spring C. pepo assemblage was at an early stage of domestication. Over the next 4,000 years, there is a well documented gradual increase in seed size, fruit size, peduncle size, and rind thickness of C. pepo in the archaeological record of eastern North America (29), providing a clear developmental trajectory from an early, large-seeded, small-fruited domesticate to later larger-fruited forms.

By the time domesticated C. pepo first appears in the archaeological record of eastern North America 5,000 years B.P., a domesticated pepo squash had already been undergoing human selection in Mexico for five millennia. A domesticated variety of C. pepo. ssp. pepo was being grown in the south-central highlands of Mexico at 10,000 B.P. (23), and by 8,500 B.P. its seeds and peduncles were already substantially larger than those recovered from 5,000 B.P. deposits at the Phillips Spring site (28). A domesticated C. pepo squash with large fruits, large peduncles, and large seeds was also being grown as far north as Tamaulipas, in northeast Mexico, by 6,000 B.P. (30), in close proximity to the modern-day range of the wild Cucurbita gourd C. fraterna.

The presence of this large-seeded, large-fruited domesticated C. pepo squash in Tamaulipas at 6,000 B.P. provides further evidence against the domestication of the “eastern” lineage (C. pepo ssp. ovifera) from C. pepo ssp. fraterna in northeast Mexico. If this large 6,000 B.P. C. pepo squash of Tamaulipas represents domesticated C. pepo spp. ovifera locally derived from a wild ssp. fraterna progenitor, then it would have had to have undergone substantial subsequent size reduction as it diffused north to arrive in the eastern United States at 5,000 B.P. as a much-smaller-fruited, smaller-seeded domesticate. If, on the other hand, the large-seeded, large-fruited C. pepo squash represents a long-extant domesticate of the other C. pepo lineage (C. pepo ssp. pepo), then the local wild gourd, C. fraterna, would have had to have been brought under domestication in northeast Mexico even though an already long-domesticated pepo squash (as well as another domesticated squash, C. argyrosperma) was already being cultivated in the region. In addition, once domesticated, the newly domesticated fraterna-derived squash would have had to have been carried northward into the eastern United States without leaving any trace in the archaeological record of Tamaulipas, and this new domesticate would have had to have made the journey north unaccompanied by either of the other domesticated squashes that were being grown in Northeast Mexico at 6,000–5,000 B.P. C. argyrosperma, for example, does not appear in the East until after A.D. 1,000 (31). Both molecular and archaeological evidence now provides compelling support for the independent domestication of C. pepo ssp. ovifera in eastern North America from an indigenous eastern North American Cucurbita gourd.

Sunflower (Helianthus annuus)

As is the case with Ch. berlandieri, wild forms of sunflower (H. annuus) have a broad present-day distribution in North America that encompasses much of the central and western continental United States and southern Canada, along with northern Mexico. Its pre-Columbian distribution, however, is less well understood (27). The geographical range of wild sunflower probably did not extend very far to the east of the Plains before the mid-Holocene, when it appears to have experienced a bison- and/or human-mediated range extension into the eastern woodlands before being brought under domestication (8, 27). In the early 1950s, Charles Heiser (32) hypothesized that sunflower was initially domesticated in the east-central United States, and subsequent archaeological research in eastern North America has provided considerable support for his early interpretation.

The earliest evidence for domesticated sunflower in the eastern woodlands of the United States consists of six complete carbonized seeds recovered from Level 14 of the Hayes site in Tennessee (33). When established correction ratios were used to estimate uncarbonized achene length measurements from these carbonized seeds, three of the specimens, with length estimates of 7.3, 7.3, and 7.4 mm, exceeded the 7-mm individual-achene-length baseline proposed for domesticated sunflower, and the mean length of all six seeds (x = 6.9 mm) exceeded the 6-mm baseline mean for identifying domesticated sunflower seed assemblages (8). Direct AMS dating of one of these seeds yielded a conventional radiocarbon age of 4,265 ± 60 B.P. and a calibration curve intercept of 4,840 B.P. (33) (Table 1).

A recent study, however, questioned the accuracy and applicability of the correction ratios used for estimating precarbonization achene length values from carbonized sunflower seeds and argued that the Hayes-site sunflower specimens did not meet the threshold for domesticated status (9). This proposed reclassification of the Hayes-site sunflower specimens as wild rather than domesticated, however, was itself based on a misapplication of ceiling standards developed for individual seeds vs. seed assemblages (8). When compared with the modern wild reference class sunflower samples used to establish the wild–domesticated boundary, the Hayes-site specimens are clearly larger than those collected from modern wild populations.

In the same study that incorrectly questioned the domesticated status of the Hayes-site sunflower specimens, two specimens of about the same age as the Hayes-site seeds, recovered from a site on the Gulf coast of Tabasco, in Mexico, are identified as representing domesticated sunflower. A partially carbonized achene measuring 8.2 × 4.5 mm and a carbonized seed measuring 7.8 × 4.4 mm were recovered from Late Archaic horizon contexts in different excavation units of the San Andrés site (Fig. 2). Direct AMS dates obtained on the achene and seed yielded conventional radiocarbon ages of 4,498 ± 50 B.P. and 4,617 B.P., respectively (9, 10). Both specimens are larger than achenes produced by modern wild sunflowers and are presented as evidence for the earliest domesticated sunflower in the Americas.

No description of the two San Andrés specimens is provided, however, other than length and width measurements and low-resolution photographs, and unfortunately, no morphological or other basis for their identification as H. annuus is offered. Their taxonomic identification is based on expert authentication rather than on explicit recognition and documentation of any taxonomically diagnostic morphology. When the San Andrés site specimens are considered within the context of documented morphological attributes of sunflower achenes and seeds, they cannot be identified as H. annuus with any degree of confidence.

The shell or pericarp of a sunflower achene gains its mechanical structure, strength, and distinctive appearance from a sclerenchymatous endocarp, which underlies thin outer epidermis, hypodermis, and carbon layers. Rather than being continuous and uniform in structure, the endocarp layer of sunflower achenes is comprised of a large number of parallel isolated strands or bundles of sclerenchyma fibers arranged longitudinally through the pericarp and separated by parenchyma cells, forming long-axis ropelike ridges (34, 35). These parallel strands are evident even when covered by an achene’s epidermis, hydrodermis, and carbon layers, and intact sunflower achenes are often described as exhibiting longitudinal surface lines, ridges, or striations. The endocarp ridges become even more distinct if overlying layers are absent (Fig. 3). When sunflower achenes erode and fragment, they split longitudinally along the parenchyma cell boundaries between sclerenchyma bundles, and as a result this distinctive and diagnostic morphological marker of H. annuus is retained even in very fragmented and highly eroded specimens. The San Andrés achene specimen lacks this morphology, and its identification as H. annuus is not supported (Fig. 3). The identification of the San Andrés site seed as sunflower can also be called into question. Although appearing to be comparable with sunflower seeds in general outline, the San Andrés specimen exhibits evidence of edge damage, and given the absence of any detailed morphological description or high resolution images, it cannot be assigned to H. annuus without additional documentation. Domesticated sunflower may well be present in the archaeological record of Mexico, but as is the case for any discussion of early domesticates, supporting arguments should be based on description and documentation of diagnostic morphology (36).

Fig. 3.

Comparison of an archaeological sunflower achene from eastern North America with the San Andrés specimen. (Left) Scanning electron micrograph of a sunflower achene from Cloudsplitter Rockshelter in eastern Kentucky, exhibiting distinctive parallel longitudinal strands or bundles of sclerenchyma fibers (achene length, 9.2 mm). (Right) San Andrés achene (achene length, 8.2 mm). (Photograph of San Andrés achene courtesy of David Lentz, Chicago Botanic Garden, Glencoe, IL.)

Given this reassessment of the Hayes-site sunflower seed assemblage and the San Andrés-site specimens, the assertion that the San Andrés sunflower specimens represented “a challenge to the hypothesis that eastern North America was a center of domestication independent of such centers in other parts of the hemisphere” (9) cannot be sustained. At the same time that currently available archaeological evidence offers strong support for the initial domestication of sunflower in the eastern United States, recent comprehensive genetic studies of modern populations of wild and domesticated sunflower provides conclusive confirmation.

In a landmark genetic study of H. annuus involving 18 microsatellite loci distributed across the sunflower genome that included 21 present-day wild populations of wild H. annuus from throughout its geographical range, along with eight Native American landraces and two modern cultivars, Harter et al. (14) showed that all extant domesticated sunflowers resulted from a single domestication event. This single domestication event, which involved a substantial genetic bottleneck, occurred in eastern North America (14, 27), confirming Heiser’s 1951 hypothesis. As a result, if domesticated sunflower is at some point documented in the archaeological record of Mexico, it will represent either the southward dispersal of domesticated sunflower from eastern North America or a second independent domestication of sunflower in Mexico. If sunflower was in fact domesticated a second time in Mexico, it left little trace in the archaeological or early historical record of Mexico, contributed nothing to the present-day domesticated sunflower genome, and was, at most, a short-lived, peripheral, now-extinct minor cultigen.

Conclusions

As briefly outlined above, all four of the seed plants proposed two decades ago as having been initially brought under domestication in eastern North America can be confidently confirmed as indigenous eastern domesticates, based on substantial archaeological evidence and recent genetic studies. Research of the past two decades has thus substantially strengthened the case for the eastern United States being an independent center of plant domestication. Maize (Zea mays), the earliest documented Mesoamerican domesticate to reach the East, does not arrive until ca. 2,200 B.P., almost 2,500 years after the local domestication of marshelder, squash, and sunflower (8).

Domestication of crop plants and the transition to food production occurs much more recently in Eastern North America than in most other parts of the world, and the region certainly does not compare with a number of the other centers of domestication in terms of having produced any of today’s most important crop plants (Fig. 1). But the eastern woodlands of North America does represent, along with the Near East, the best-documented regional record of this major transition in human history (37).

Such well documented regional records of the timing, sequence, and spatial, temporal, cultural, and environmental contexts of initial domestication of different species are the key to gaining a better overall understanding of the transition from a hunting and gathering way of life to human reliance on food-production economies. Over the next several decades, genetic and archaeological research on the domestication of plants and animals should substantially increase the number of comparably well documented primary centers of domestication. This improved and expanded set of regional-scale developmental sequences, in turn, will allow much more informed comparative analyses and better opportunities for the recognition of general shared patterns of developmental variation and commonality in the transition to agriculture, worldwide.

Glossary

Abbreviation:

AMS

accelerator mass spectrometry.