The Origins of Plant Cultivation and Domestication in the New World Tropics Patterns, Process, and New Developments
Abstract
The New World tropical forest is now considered to be an early and independent cradle of agriculture. As in other areas of the world, our understanding of this issue has been significantly advanced by a steady stream of archaeobotanical, paleoecological, and molecular/genetic data. Also importantly, a renewed focus on formulating testable theories and explanations for the transition from foraging to food production has led to applications from subdisciplines of ecology, economy, and evolution not previously applied to agricultural origins. Most recently, the integration of formerly separated disciplines, such as developmental and evolutionary biology, is causing reconsiderations of how novel phenotypes, including domesticated species, originate and the influence of artificial selection on the domestication process. It is becoming clear that the more interesting question may be the origins of plant cultivation rather than the origins of agriculture. This paper reviews this body of evidence and assesses current views about how and why domestication and plant food production arose.
Introduction
It has long been recognized that numerous New World plant domesticates—more than half, in fact, of all American crops and many of the staples that supported indigenous peoples when Europeans arrived—originated in Neotropical forests (e.g., Harris 1972; Sauer 1950). In the past few decades, a large corpus of archaeobotanical, paleoecological, and molecular/genetic information has become available from Central and South America that has led to a significantly increased understanding of the geography and chronology of tropical food production. This information has established the lowland Neotropical forest as an early and independent cradle of agricultural origins. In a volume published in 1998, Deborah Pearsall and I reviewed and synthesized the known archaeological, paleocological, ethnographic, and molecular/genetic evidence bearing on the prehistory of Neotropical agriculture (Piperno and Pearsall 1998).
Since then, much more evidence has been generated from all the contributing disciplines. New archaeological sites have been discovered, excavated, and analyzed using the full complement of now-available archaeobotanical techniques. Important early sites that were first identified and studied more than 30 years ago and for which little to no botanical information was available have seen reexcavation and applications of microfossil research. Phytoliths and in some cases macrofossils from other early human occupations have been revisited during the past 10 years with the use of expanded and improved modern reference collections, and some of these remains have been directly dated. Starch grain data from stone tools and human teeth are becoming available from a growing number of sites, and construction of large modern reference collections has allowed identification of some domesticated root and seed crops. Major collecting efforts of important economic taxa such as the Cucurbitaceae have been undertaken, adding significant knowledge about modern landrace diversity in Cucurbita and Lagenaria, along with distributional ranges of wild species. Last but not least, newer molecular and other information has refined and in some cases revised the geography of plant domestication and also caused us to reconsider how we should view and come to an understanding of the domestication process. In this paper, I review the presently available information and assess current views of how and why plant food production and domestication arose.1
The Geography of Domestication
Agriculture may originate in discrete centers or evolve over vast areas without definable centers.
(Jack R. Harlan 1971:468)It is useful to begin by assessing what is currently known about the geography of Neotropical plant domestication (for a summary, see fig. 1). Scholars such as David Harris, Carl Sauer, Jack Harlan, and others of their generations knew that tropical forest agriculture probably developed independently long before European arrival, and they for the most part understood whether Central America (largely Mesoamerica) or South America was the original home of the major crops. More precise area origins for New World crops, and whether single or multiple domestications or even hybrid origins occurred in some, was often less clear for a number of reasons. Wild congenerics that potentially could be progenitor species on the basis of shared morphological attributes or apparent lack of hybridization barriers were often broadly distributed, and they sometimes occurred in both Central America and South America (e.g., Manihot [manioc]; Cucurbita [squashes and gourds]; Zea [teosinte], the ancestor of maize; Ipomoea [sweet potato]; Xanthossoma [malanga or cocoyam, New World representatives of the taro family]; Dioscorea [yams]; Phaseolus lunatus [sieva beans]; Gossypium [cotton]; many tree crops; Piperno and Pearsall 1998).
Postulated domestication areas for various tropical crops in Central and South America. Open circles are archaeological and paleoecological sites with early domesticated crop remains. The numbers in parentheses after a taxon indicate that more than one independent domestication occurred. The possible area of origin for the sieva bean extends into the Pacific lowlands north of the oval area. The oval here and those in B labeled D1–D4 designate areas where it appears that more than one or two important crops may have originated. Arrows point to approximate areas and are not meant to denote specific domestication locales (for more details and sources used in the figure, see Piperno and Pearsall 1998; Piperno 2006a; see also Cross, Saade, and Motley 2006; Hughes et al. 2007; Matsuoka et al. 2002; Miller and Schaal 2005; Motta-Aldana et al. forthcoming; Olsen and Schaal 1999; Plowman 1984; Rodrigues, Filho, and Clement 2004; Sanjur et al. 2002; Wessel-Beaver 2000; Westengen, Huamán, and Heun 2005). Modern vegetation zone guides are (a) 1, tropical evergreen forest; 2, tropical semievergreen forest; 3, tropical deciduous forest; 4, savanna; 5, low scrub/grass/desert; 6, mostly cactus scrub and desert; and (b) 1, tropical evergreen forest (TEF); 2, tropical semievergreen forest (TSEF); 3, tropical deciduous forest (TDF); 4, mixtures of TEF, TSEF, and TDF; 5, mainly semievergreen forest and drier types of evergreen forest; 6, savanna; 7, thorn scrub; 8, caatinga; 9, cerrado; 10, desert.
Postulated domestication areas for various tropical crops in Central and South America. Open circles are archaeological and paleoecological sites with early domesticated crop remains. The numbers in parentheses after a taxon indicate that more than one independent domestication occurred. The possible area of origin for the sieva bean extends into the Pacific lowlands north of the oval area. The oval here and those in B labeled D1–D4 designate areas where it appears that more than one or two important crops may have originated. Arrows point to approximate areas and are not meant to denote specific domestication locales (for more details and sources used in the figure, see Piperno and Pearsall 1998; Piperno 2006a; see also Cross, Saade, and Motley 2006; Hughes et al. 2007; Matsuoka et al. 2002; Miller and Schaal 2005; Motta-Aldana et al. forthcoming; Olsen and Schaal 1999; Plowman 1984; Rodrigues, Filho, and Clement 2004; Sanjur et al. 2002; Wessel-Beaver 2000; Westengen, Huamán, and Heun 2005). Modern vegetation zone guides are (a) 1, tropical evergreen forest; 2, tropical semievergreen forest; 3, tropical deciduous forest; 4, savanna; 5, low scrub/grass/desert; 6, mostly cactus scrub and desert; and (b) 1, tropical evergreen forest (TEF); 2, tropical semievergreen forest (TSEF); 3, tropical deciduous forest (TDF); 4, mixtures of TEF, TSEF, and TDF; 5, mainly semievergreen forest and drier types of evergreen forest; 6, savanna; 7, thorn scrub; 8, caatinga; 9, cerrado; 10, desert.
(Continued)
The analysis of protein and DNA-based molecular markers has significantly elucidated these issues, most recently for some major root, seed, and tree crops such as manioc, various species of Cucurbita, chayote (Sechium edule), P. lunatus (sieva beans), Leucaena spp. (guaje), Spondias purpurea (jocote or Mexican plum), Bactris gasipaes (pejibaye or peach palm), peanuts (Arachis hypogaea), South American cotton, and others (e.g., Bertoti de Cunha et al. 2008; Cross, Saade, and Motley 2006; Hughes et al. 2007; Léotard et al., forthcoming; Milla, Isleib, and Stalker 2005; Miller and Schaal 2005; Motta-Aldana et al., forthcoming; Olsen 2002; Olsen and Schaal 1999; Robledo, Lavia, and Seijo 2009; Rodrigues, Filho, and Clement 2004; Sanjur et al. 2002; Westengen, Huamán, and Heun 2005; fig. 1). Furthermore, it has been conclusively shown that the wild ancestor of maize is not native to the semiarid Mexican highlands but rather to lower, warmer, and moister habitats of Guerrero and Michoacán states, where seasonal tropical forest is the native vegetation (e.g., Doebley 2004; Matsuoka et al. 2002; van Heerwaarden et al. 2011).
In many cases studied where multiple domestications were discussed, single domestications are now indicated (e.g., manioc, maize, the important pejibaye palm; Léotard et al., forthcoming; Matsuoka et al. 2002; Olsen 2002; Olsen and Schaal 1999; Rodrigues, Filho, and Clement 2004). On the other hand, a double domestication of the lima bean, one event in the southern Ecuador/northern Peru Andes leading to the large-size lima and another in the humid tropical lowlands leading to the small-seeded sieva bean, is again indicated by the newest molecular data (Motta-Aldana et al., forthcoming). Furthermore, lowland western Mexico is definitively revealed as an origin area for the sieva, and it is possible that another independent domestication event occurred in the lowlands of Central or South America (Motta-Aldana et al., forthcoming). More than one domestication of an important Central American tree crop, S. purpurea, also appears likely (Miller and Schaal 2005).
An Old World plant, the bottle gourd, was transported to and well used throughout tropical America during the pre-Columbian era. Erickson et al. (2005), working with modern and ancient DNA, established conclusively that prehistoric plants originated in Asia, not Africa, as was previously assumed. The archaeological rinds they studied bore the mark of domesticated plants. The investigators concluded that humans, not ocean currents, probably carried the domesticated plant from Asia during the initial colonization of the New World.
Importantly, molecular data have also elucidated in more detail some of the different processes involved in the domestication of plants (see also “The How of Plant Domestication”). For example, interesting cases of hybrid origins of Mesoamerican polyploid tree crops were confirmed in the important genus Leucaena, whereby human-mediated sympatry in “backyard gardens” of previously separated wild taxa, extensive predomestication cultivation, and subsequent spontaneous hybridization led to the emergence of different domesticated species in the genus (Hughes et al. 2007). The most widely cultivated Mesoamerican cactus, Opuntia ficus-indica, probably arose by the same means (Griffith 2004; Hughes et al. 2007). The importance of these processes and backyard, also called “dooryard,” cultivation in contributing to the stock of plants domesticated prehistorically was proposed long ago by Edgar Anderson (1954). We should remember that other major American crops, such as peanuts, are products of the hybridization of two different species and the how, when, and where of the predomestication hybridization events that led to their initial emergence are not well understood.
Figure 1 shows known or postulated geographic zones of domestication for some Neotropical crops on the basis of current molecular, archaeological, and ecological evidence (for a more complete list of crops native to the Neotropics, see Harlan 1992). In some cases, designations of the circled and other localities on the maps as origin areas for particular crops should be treated as hypotheses that require testing with additional molecular and archaeological data. For example, more precise domestication localities for yam Dioscorea trifida, leren Calathea allouia, cocoyam Xanthossoma sagittifolium, and achira Canna edulis within their likely origin areas of northern South America may be forthcoming once relevant molecular studies have been undertaken. There is a major root crop for which neither Central American nor South American origins have been conclusively demonstrated. Sweet potato’s presumed wild ancestor is Ipomoea trifida, a poorly demarcated taxon naturally distributed from Mexico through northern South America (Piperno and Pearsall 1998). Some molecular work based on amounts of diversity in modern landraces suggests that it may be Central American in origin (Zhang et al. 2000), although the oldest archaeological evidence for it by far derives from western South America (Piperno and Pearsall 1998; Shady, Haas, and Creamer 2001). Centers of present diversity do not always accurately pinpoint a center of origin, and additional molecular work focusing on the construction of a rigorous phylogeny of wild and domesticated sweet potato is needed to resolve the issue.
Even with the existing uncertainties, it is clear that crop origins were spatially diffuse. Using the conventional definition of a center—a circumscribed region where agriculture began and out of where it spread—Mesoamerica, more precisely Mexico, may still qualify, although it is clear now that maize, the common bean (Kwak, Kami, and Gepts 2009) and other Phaseolus species (the tepary bean), various tree crops, chile peppers, cotton, chayote, and others were domesticated in different and sometimes ecologically dissimilar regions of the country (see also Piperno, forthcoming). In South America, an attempt to designate a single circumscribed center or core area of agriculture would clearly not be supported. The origins of major and now minor crops are spread from the northern parts to the southern parts of the continent, west and east of the Andes, mostly in seasonal types of lowland tropical forest for major root and seed crops but also in lowland wet forests and midelevation moist forest habitats.
Patterns indicating multiple areas of domestication become even more accentuated when highland crops such as Phaseolus, Amaranthus spp., Chenopodium spp., and the various Andean tubers are added to the picture (the Andean complex also includes additional species of Cucurbita and Capsicum; Piperno, forthcoming; Piperno and Dillehay, forthcoming). Crops that would eventually become major staple foods or condiments and that were commonly grown together after food production was established—such as maize, manioc, various squashes, chile peppers, sweet potato, and Phaseolus and Canavalia beans—had spatially disparate areas of origin and did not at first spread together. In fact, some of the earliest and most widespread anchors of lowland food production are now minor, even disappearing, foods (below). Early patterns of dispersals probably did not involve significant population movements or diffusion of crops in packages.
Of course, knowing that crop plant origins were geographically widespread does not necessarily lead to a conclusion that food production was independently developed wherever a particular crop originated. Rather, the spread of a crop or crops into new regions may have inspired receiving cultures to grow their own native plants. This issue is further complicated by the fact that we do not know when some of the crops displayed in figure 1 were initially brought under cultivation. How many truly independent developments of lowland food production were there? Given that lowland northern and southern South American domestication zones are separated by large distances and that a number of plants native to these areas were taken under cultivation and domesticated by the middle of the eighth millennium BP, it is difficult to see how the northern and southern lowland regions do not form at least two to three independent areas of food production (e.g., D1, D3, and D4 in fig. 1B). These questions will be further illuminated in the near future as more data are accumulated.
In sum, Harlan’s (1971) idea that peoples over a wide geographic area were simultaneously engaging in early cultivation and domesticatory relationships with plants and considerably influenced the early development of some domesticates after the plants left their native areas seems particularly relevant in the light of current data. Opportunities for such kinds of processes to occur would have been even greater if, as in the Old World (e.g., Fuller 2007; Jones and Brown 2007; Weiss, Kislev, and Hartmann 2006; Willcox, Fortnite, and Herveux 2008), protracted periods of predomestication cultivation and/or spread of predomesticated crops occurred in the Neotropics. This issue is discussed in more detail below.
Finally, molecular and botanical studies together with an increasing amount of archaeobotanical data, summarized below, tell us that the wild ancestors of many important crop plants are native to the seasonal tropical forest, those formations that receive a 4–7-month-long period every year during which little to no rain falls. Annual precipitation in these areas averages about 1,200–1,800 mm a year, soils are less weathered and thus more highly fertile than in ever-wet (aseasonal) forest, and the prolonged dry season enabled early farmers to efficiently clear vegetation and prepare plots for planting with the simple use of fire. Seasonally dry tropical forests do not carry the distinction of their rain forest relatives, but their prominent position in Neotropical agricultural origins is clear.
Early Food Production and Its Cultural and Ecological Contexts
The New World was colonized before 15,000 BP, by which time human populations had traversed most of the western South American landmass, probably by moving along the Pacific Coast, and found themselves in what is now southern Chile (Dillehay et al. 2008). (All dates given in the text are in calibrated 14C years; uncalibrated and calibrated determinations for specific dates from sites discussed are found in table 1.) The earliest indisputable evidence of human occupation in what is now the lowland Neotropical forest comes from sites distributed from Belize to eastern Brazil that were first occupied at about 13,000 BP (Piperno and Pearsall 1998; Ranere and Cooke 2003). Beginning soon after the Pleistocene ended at about 11,400 BP, human settlement of the Neotropics began to change from these sparsely distributed and short-term occupations. People “settled into” their landscapes, staying for longer and/or more frequently returning to specific locations, and they frequently manipulated and altered their environments by creating clearings in forests and/or burning them. They developed tool kits indicating that for the first time the exploitation of plants was as important an economic strategy as hunting had been (e.g., Gnecco and Aceituno 2006; Mora 2003; Ranere and Cooke 2003; Ranere et al. 2009). Archaeobotanical information indicates that food production began in a number of localities in tropical Central and South America during the early Holocene (between 11,000 and 7600 BP), not long after the Neotropical climate and vegetation underwent profound changes associated with the end of the Pleistocene (discussed in more detail below).
Crop plant occurrence and chronology
| Site | Age | Crop plants |
|---|---|---|
| Mexico: | ||
| Guerrero State: | ||
| Xihuatoxtla Shelter | By 7920 ± 40 BP (by 8960–8940, 8850–8840, and 8780–8630 cal BP) | Maize (Phy and SG-GS), Cucurbita (Phy) |
| Tabasco State: | ||
| San Andrés | 6208 ± 47 BP (7204–6904 cal BP) | Maize (Phy, Po) |
| Panama: | ||
| Central Pacific Panama: | ||
| Aguadulce Rock Shelter | By ca. 8600 cal BP | Cucurbita moschata, leren, bottle gourd (Phy), Arrowroot (Phy, SG-GS) |
| 6910 ± 60 BP (7740–7640 cal BP) | Maize, manioc (SG-GS), | |
| 7061 ± 81 BP (7922–7754 cal BP) | Maize (Phy) | |
| By ca. 5700 cal BP | Dioscorea trifida (SG-GS) | |
| Cueva de los Ladrones | 6860 ± 90 BP (7804–7631 cal BP) | Maize (SG-GS, Phy, Po) |
| Cerro Mangote | 6810 ± 110 (7779–7584 cal BP) | Maize (SG-GS) |
| Western Panama: | ||
| Chiriqui Rock Shelters | 6560 ± 120 BP (7554–7381 cal BP) | Arrowroot, maize (SG-GS) |
| Ca. 5600 cal BP | Manioc (SG-GS) | |
| Hornito | 6270 ± 270 BP (7779–7584 cal BP) | Maize (SG-GS) |
| Colombia: | ||
| Middle Porce Valley | Between 6280 ± 120 and 5880 ± 80 BP (between 7321–7032 and 6799–6597 cal BP) | Maize (SG, Phy-GS; Po) |
| Middle Cauca Valley: | ||
| El Jazmin | 7590 ± 90 BP (8493–8313 cal BP) | Dioscorea (SG-GS [D. trifida?]) |
| Between ca. 7000 and 5000 BP (ca. 8000–6000 cal BP) | Maize (Po) | |
| Middle Cauca Valley, Calima Region: | ||
| El Recreo | 7980 ± 120 and 7830 ± 140 BP | C. cf. moschata (Phy); Persea americana (M, [Cul?]); |
| (9001–8674 and 8903–8508 cal BP) | Cucurbitaceae (M) | |
| Hacienda Lusitania | >5150 ± 180 BP (>6138–5721 cal BP) | Maize (Po) |
| Hacienda El Dorado | 6680 ± 230 BP (7771–7349 cal BP) | Maize (Po) |
| Upper Cauca Valley: | ||
| San Isidro | 9530 ± 100 BP (11,058–10,706 cal BP) | Bottle gourd (M, Phy), Cucurbita (Phy [Cul.?]), P. americana (M), Maranta cf. arundinacea (SG-GS [Cul.?]) |
| Colombian Amazon: | ||
| Middle Caquetá Region: | ||
| Peña Roja | 8090 ± 60 BP (9107–8884 cal BP) | Cucurbita, leren, bottle gourd (Phy) |
| Abeja | >4695 ± 40 BP (>5539–5351 cal BP) | Maize, manioc (Po) |
| Southwestern Ecuador: | ||
| Las Vegas Sites: | ||
| OGSE-80 and OGSE-67 | Between 10,130 ± 40 and 9320 ± 250 BP (11,750–10,220 cal BP) | Cucurbita ecuadorensis (Phy) |
| 9320 ± 250 BP (11,060–10,950, 10,780–10,220 cal BP) | Leren, bottle Gourd (Phy) | |
| 7170 ± 60 BP (8015–7945 cal BP) | Maize (Phy) | |
| >5820 ± 180 BP (6850–6810 cal BP) | Maize (Phy) | |
| Valdivia Sites: | ||
| Real Alto | Ca. 4300 BP (ca. 5000 cal BP) | Leren, achira, arrowroot, maize, manioc (SG, Phy-GS; Phy), jack bean, cotton (M) |
| Loma Alta | 4470 ± 40 BP (5260–5000 cal BP) | Arrowroot, maize, manioc, jack bean, Capsicum (SG-Cer) |
| Ca. 5500–4400 BP (ca. 6500–5200 cal BP) | Maize, Cucurbita, Achira (Phy) | |
| Ecuadorian Amazon: | ||
| Ayauchi | Ca. 5300 BP (ca. 6000 cal BP) | Maize (Po, Phy-Lake sediments) |
| Eastern Amazon: | ||
| Geral, Brazil | 5760 ± 90 BP (6662–6464 cal BP) | Slash-and-burn cultivation (?) |
| Ca. 3350 BP (ca. 3800 BP) | Maize (Po, Phy-Lake sediments) | |
| Northern Peru: | ||
| Zaña Valley | 9240 ± 50 BP (10402–10253 cal BP) | C. moschata (M) |
| 7840 ± 40 BP (8630–8580 cal BP) | Arachis sp. (M) | |
| 5490 ± 60 BP (6301–6133 cal BP) | Cotton (M) | |
| Ca. 7500 BP (ca. 8500 cal BP) | Manioc (M) | |
| 8210 ± 180 BP (9403–8784 cal BP) | C. moschata, Arachis, Phaseolus, Inga feuillei (SG-HT) | |
| 7120 ± 50 BP (7950 cal BP) | Coca (Erythroxylum novagranatense var truxillense; M) | |
| Siches | 9533 ± 65 BP (11015–10885 cal BP; BGS 2426) and 9222 ± 60 BP (10243–10306 cal BP; BGS 2475) | Cucurbita (Phy) |
| Southern Coastal Peru: | ||
| Paloma | Ca. 7800 BP (ca. 8800 cal BP) | Bottle gourd (M) |
| 5070 ± 40 BP (5900–5740 cal BP) | Cucurbita ficifolia (M) | |
| By 5300–4700 BP (6500–5700 cal BP) | Phaseolus lunatus, Cucurbita spp., guava (Psidium guajava; M) | |
| Chilca 1 | 5616 ± 57 BP (6440–6310 cal BP) | P. lunatus (M) |
| By 4400 BP (5400 cal BP) | Cucurbita, Achira, Jicama (Pachyrhizus ahipa), jack bean | |
| Quebrada Jaguay | 7660 ± 50 BP (8445–8395 cal BP) | Bottle gourd (M) |
| Southeastern Uruguay: | ||
| Los Ajos | 4190 ± 40 BP (4800–4540 cal BP) | Maize, Phaseolus (SG-GS); maize, Cucurbita (Phy) |
The best evidence currently comes from the Central Balsas Valley of southwestern Mexico (Piperno et al. 2009; Ranere et al. 2009), central Pacific and western Panama (Dickau 2010; Dickau, Ranere, and Cooke 2007; Piperno 2006c, 2009; Piperno et al. 2000), the sub-Andean and premontane zones (elevation between 1,000 and 1,600 m) of the Cauca and Porce valleys in Colombia (e.g., Aceituno and Castillo 2005; Bray et al. 1987; Gnecco and Aceituno 2006), the Colombian Amazon (Cavelier et al. 1995; Mora 2003; Piperno and Pearsall 1998), southwestern Ecuador (Pearsall 2003; Pearsall, Chandler-Ezell, and Chandler-Ezell 2003; Pearsall, Chandler-Ezell, and Zeidler 2004; Piperno 2009; Piperno and Stothert 2003; Zarillo et al. 2008), and the Zaña Valley of northern Peru (Dillehay et al. 2007; Piperno and Dillehay 2008). The vegetation of all of these areas was humid tropical forest with the exception of the Vegas, Ecuador, sites located at an ecotone between forest and scrub vegetation. Table 1 contains detailed information on crop plant occurrence and chronology. Associated published references and other information about the sites involved can be found in CA+ supplement A. Figure 2 displays site locations. The reader should refer to the citations given above and in CA+ supplement A in discussions that follow below.
Location of sites with early food production in Central and South America discussed in the text. The sites are shown with open circles. Sites with an open triangle are important examples of pre-6000 BP crop diffusion into southern coastal Peru (Chilca 1, Paloma, and Quebrada Jaguay) and regions of South America where agriculture as a whole had been poorly documented (e.g., Geral, located in the interior of the Amazon, and possibly Los Ajos, Uruguay, by ca. 4000 BP). They are not discussed in the text but are listed in table 1.
The relevant archaeobotanical data are in the main part from microfossils, namely, starch grains recovered from numerous securely dated stone tools and human teeth and phytoliths retrieved from the same stone tools and/or closely associated sediments. In several cases, phytoliths and starch grains have been directly dated. Macrobotanical information is available from Colombia and northern Peru, and paleoecological efforts allied with archaeological work provide pollen, phytolith, and charcoal data indicating crop presence and/or substantial human vegetational modification near sites. Table 1 also includes early macrobotanical data from the arid coast of Peru, where crops grown were largely from elsewhere and their first appearance provides a minimum date for their domestication.
The earliest crops were Calathea allouia (leren) and Maranta arundinacea (arrowroot), both grown for their tubers; Cucurbita moschata, Cucurbita ecuadorensis, and possibly Cucurbita argyrosperma squash; bottle gourd; maize; manioc; peanuts; avocado; and pacay (Inga feullei), another tree crop. The first three—leren, arrowroot, and C. moschata squash—along with the bottle gourd are persistently present in northern South America and Panama between 10,200 and 7600 BP, underlining their probable northern South American origins. In many cases, different kinds of archaeobotanical remains provide mutually supporting evidence for the same crop species from the same site and specific context (table 1). It is obvious now that the earliest crop complexes were neither seed, tree, nor root crop based but rather mixtures of these different elements.
During the middle eighth millennium BP, the first signs of major crop movements northward from their area(s) of origin in southern South American can be seen. Macrofossil and starch grain data show that peanuts moved into the Zaña Valley of northern Peru by 8500 BP. Macrofossils of manioc (identification confirmed by starch grains isolated directly from root remains) also occur there by about 8500 BP, and starch grains from the plant are present in central Panama at 7600 BP. Pollen evidence from the Colombian Amazon indicates that manioc arrived there before 5800 BP. It is possible that peanuts and manioc moved north together from a common area of origin (fig. 1B). Chile peppers were well dispersed in southern Central America and South America by 6000–5000 BP (Perry et al. 2007).
Recent studies in the Central Balsas River Valley of Mexico, maize’s postulated cradle of origin, document the presence of maize phytoliths and starch grains at 8700 BP, the earliest date recorded for the crop (Piperno et al. 2009; Ranere et al. 2009). A large corpus of data indicates that it was dispersed into lower Central America by 7600 BP and had moved into the inter-Andean valleys of Colombia between 7000 and 6000 BP. Given the number of Cauca Valley, Colombia, sites that demonstrate early maize, it is likely that the inter-Andean valleys were a major dispersal route for the crop after it entered South America (table 1; fig. 2). Furthermore, directly dated starch grains from food residues in early Valdivia ceramics once again affirm maize and other crop presence in domestic contexts in these Early Formative (6000–5000 BP) occupations (Zarillo et al. 2008; see table 1 for other Valdivia crop plant occurrences).
An abundance of artifactual evidence also speaks to early traditions of dedicated plant exploitation and cultivation. Stone implements used for plant processing (“edge-ground cobbles,” quern stone bases, hand milling stones) are common in sites with early food-producing activities, and they are often present from the earliest Holocene onward in sites that were occupied at that time. Residues studied from these tools have commonly produced starch grains and phytoliths from a variety of cultigens (table 1). In the Porce and Middle Cauca valleys of Colombia, the Colombian Amazon, and the Zaña Valley in Peru, finely made stone hoes firmly dated to from 9500–7300, 8000, and 7000 BP, respectively, occur in sites with evidence of early food production (table 1) and in others for which botanical data are not available (Aceituno and Castillo 2005; Gnecco and Aceituno 2006; Herrera et al. 1992; Mora 2003; Salgado 1995). The Cauca region Colombian examples, which by 7300 BP are beautifully polished, are among the most typical implements documented in early preceramic occupations there. Stone axes presumably used for creating openings in the forest are also found in the Cauca, Porce Valley, and Colombian Amazon occupations at the same time.
In the way of other economic remains documented, a number of wild plant taxa, including a variety of palms and other tree fruits (e.g., Spondias, Erythrina, Byrsonima), yams other than Dioscorea trifida, Calathea species other than C. allouia, and Zamia have been identified. It is possible that some of them, particularly a few of the palm genera, were cultivated, but this cannot be empirically demonstrated with either the macrobotanical data or the microbotanical data. Current information similarly cannot be employed to adequately quantify changes in the proportions of wild and cultivated plant foods as early periods of cultivation ensued. Bone often does not survive in the humid tropics, but in the few sites with faunal records (e.g., Las Vegas and the Zaña Valley), both large and small animals were hunted. The totality of evidence indicates broad-spectrum subsistence orientations shortly before and at the beginning of food production.
With evidence accumulating rapidly now, other questions such as how early cultivated and domesticated plants moved (through movements of people, objects, or cultural knowledge) will take on increasing importance. This issue cannot be discussed in any detail here, but considering the present evidence, simple down-the-line forms of exchange that did not involve significant population shifts or movements of material culture may best account for early crop plant diffusion. It should be stressed that different scenarios may have been true for later cases of crop diffusion through the Neotropics not discussed here, when Formative-period societies were established and population numbers were much higher than they were during the early and early middle Holocene.
It is clear that unlike in the Near East and China (see papers this issue), Neotropical food production did not originate and take hold in the context of large or fairly large permanent and nucleated villages situated in major river valleys. Rather, intensive foot surveys and excavations in the Balsas region of southwestern Mexico, central and western Panama, southwestern Ecuador, the Cauca and Porce regions, Colombia, and northern Peru show that between 11,000 and 7000 BP, sites are typically rock shelters and/or limited clusters of small open-air occupations that were located beside secondary watercourses and seasonal streams whose small stretches of alluvium likely were used for planting gardens. Settlements were typically less than 1 ha in size, and many may have been occupied seasonally. Settlement organization was similar to modern tropical hamlets and hamlet clusters, where one to a few nuclear families composed the residential community.
Early expressions of permanent settlements and seminucleated communities are found in the Zaña Valley, Peru, where occupations between 10,000 and 7600 BP are small circular houses located 200–400 m apart with stone foundations and stone-lined storage pits. The deep and dense Las Vegas, Ecuador, middens, situated near the Pacific coast, where a wide variety of marine and estuary resources (e.g., mollusks, fish, and crabs) as well as plants and terrestrial animals were exploited, may also have been occupied on a permanent basis.
It was not until substantially later in time that settlements in these and other regions were positioned to exploit the rich bottomland of significant river courses. The earliest such evidence comes from the Valdivia culture of southwest Ecuador and dates to 5500 BP (e.g., Pearsall, Chandler-Ezell, and Chandler-Ezell 2003; Raymond 2008). In other regions, this development took place at about 4000–3400 BP. Another related and often-discussed facet of early farming is the view that it should originate in zones of plentiful wild food resources (see papers this issue). For a number of reasons, the high biodiversity of tropical forests does not translate into habitats with abundant and stable wild resources for hunters and gatherers. Useful calorie-rich plant species occur in low number and are widely dispersed in space, and animal protein is similarly far from plentiful, with the effect that natural resources would generally be expected to support small groups of mobile foragers (Piperno and Pearsall 1998:52–78). Resource abundance would have been better, at least seasonally, in localities of early food production where permanent lakes occurred, such as in an area of the Central Balsas, Mexico, studied recently (Piperno et al. 2007, 2009; Ranere et al. 2009). However, most early food producers did not have access to lakes. Piperno and Pearsall (1998:323) concluded that in the Neotropics, early farming occurred “in the most optimal zones of the most optimal types of forests and, in this sense, they represent resource abundance,” adding that “nonetheless, as long as people are not starving or otherwise experiencing frequent and severe shortfalls of food, food abundance per se may have little relevance.” There is no reason to alter this view.
When Did Neotropical Farming Become a Significant Subsistence Practice?
The evidence increasingly indicates that it would be a mistake to assume that Neotropical food production during the entire 11,000–7000-BP period was a casual undertaking practiced by people who are better called foragers than farmers (see also Iriarte 2007, 2009 for discussions of this issue). Multifaceted archaeobotanical data—now including starch grain and phytolith evidence from the calculus of human teeth—along with settlement pattern, landscape modification (in some cases resulting from slash-and-burn cultivation), and artifactual information indicate that by 8800–7600 BP in the Zaña Valley, Peru; central Panama; probably the Central Balsas, Mexico; Tabasco, Mexico; and possibly other regions, a significant number of dietary calories and nutrients came from crop plants. For example, more than 70% of the starch grains recovered from the teeth of nine different Zaña Valley individuals dated from 8800–7700 BP were from four crops: Phaseolus, Cucurbita moschata, peanuts, and Inga feuillea, a tree (Piperno and Dillehay 2008). In Panama, the Colombian Amazon, and southwestern Ecuador, root crops such as leren and arrowroot that were significant and reliable sources of carbohydrates are ubiquitous components of early crop plant assemblages. Cucurbita, another ubiquitous early plant, provided high-quality proteins and fats. Furthermore, the starch, phytolith, and macrobotanical evidence from a number of regions makes it clear that squashes (including the fruit flesh in the Zaña Valley) were routinely consumed, had undergone artificial selection for different traits related to fruit edibility, and thus were not primarily used as little-modified nondietary plants.
In the Central Balsas and San Andrés regions of Mexico and in central Panama, forest clearance from slash-and-burn cultivation is documented by pollen, phytolith, and charcoal records from lake sediment cores beginning at 7600–7200 BP (Piperno et al. 2007; Pohl et al. 2007; Pope et al. 2001). Small irrigation canals are present in the Zaña Valley beginning at 7700 BP (Dillehay, Eling, and Rossen 2005). In a number of regions, then, agricultural intensification began before 7000 BP. In central Panama; the Zaña Valley, Peru; and the Cauca and Porce regions, Colombia, it is also evident that site number and artifact density significantly increased around 7600–7000 BP. These trends are likely the result of an increase in human carrying capacity made possible by effective systematic food production.
In summary, these people appear to have been committed horticulturists and slash-and-burn cultivators who, while still integrating planting with collecting and hunting, were taking important steps along the pathway to full-scale agriculture. The appearance of large sedentary and nucleated villages, which postdates 6000 BP throughout the Americas, should no longer be considered a necessary backdrop for the occurrence or recognition of effective and productive agriculture in the Americas. This is all the more true when it is considered that indigenous Neotropical farmers still live in—and more commonly did so in the recent past—small seminucleated and shifting communities. The idea that the appearance of sedentary village life and all of its trappings should be the measure of when farming began in the Americas emerged from faulty comparisons with Near Eastern pre- and postagricultural trajectories, which were a product of ecological and demographic circumstances very different from those associated with the beginnings of Neotropical food production (Piperno and Pearsall 1998). As also stressed by Iriarte (2007), Vrydaghs and Denham (2007), and Denham (2011), agricultural origins in different parts of the world should be studied using the totality of evidence relevant to the region being investigated.
The How of Plant Domestication
Phenotypic innovation depends on developmental innovation, and developmental innovation spans a broader field of possibilities than does mutation alone.
(Mary Jane West-Eberhard 2003:144)Allowing for cryptic variants and novel phenotypes from new epistatic combinations to arise during domestication, it is easy to imagine that maize was domesticated from teosinte.
(John Doebley 2004:56)Evolutionary Development, Gene Expression, and Developmental Plasticity
Recent developments in evolutionary biology and archaeobotanical data gathering and interpretation indicate the need to reassess how plant domestication occurred. Fundamental new insights about the origin of novel traits and adaptive evolution are arising from the fields of evolutionary developmental biology (evo-devo), molecular biology, and developmental plasticity. In addition, a constant stream of detailed archaeobotanical investigations has altered our views about the trajectory of plant domestication following the advent of systematic cultivation (defined here as the cycle of planting and harvesting in plots prepared for this purpose). I start first with new insights from the nonarchaeological spheres of research.
Gene regulation/expression acting during an organism’s development and phenotypic (developmental) plasticity are now widely viewed as significant forces, indeed by some investigators as the primary forces, in evolutionary diversification and the origin of novel traits (e.g., Carroll 2005; Kruglyak and Stern 2007; Pigliucci 2005; West Eberhard 2003). There is good reason to believe that these research foci should become standard elements of domestication studies (Gremillion and Piperno 2009a). Regulatory genes are not protein-coding or functional genes but rather act to switch other genes on or off or change when and where in an organism they are active or increase/decrease their effects, usually during early ontogeny. This is the process of gene expression (e.g., Kruglyak and Stern 2007). Novel phenotypes may then often result from reorganizations of existing genomes, not by the spread or appearance of mutations, and new phenotypic variation can rapidly arise in populations without a corresponding genetic change.
Developmental plasticity, recently given a complete treatment in a major book (West-Eberhard 2003), refers to the inherent capacity of organisms to rapidly produce novel heritable phenotypes through one of several available developmental pathways in direct response to changes in their environments. This capacity should be particularly important in plants, which cannot simply get up and move to places more to their liking when physical and biotic conditions change and become less favorable. Gene expression during plant development orchestrates this process, resulting in different phenotypic pathways to adulthood. The new phenotypes have the potential to become fixed (stabilized) by genetic assimilation/accommodation if the new ecological conditions are maintained over multiple generations (West-Eberhard 2003). An important source of genetic variation that may enable the creation of novelties through developmental-mediated mechanisms is “cryptic genetic variation,” so named because it is unexpressed in the phenotype of standing populations and thus normally hidden from selection (e.g., Gibson and Dworkin 2004; Lauter and Doebley 2002). However, it may be set in motion or “released” by interactions between genes (epistasis) or by perturbations from the external environment, rapidly resulting in new phenotypes. Lauter and Doebley (2002) have shown that maize’s wild ancestor possesses a significant degree of cryptic variation that is associated with important traits such as polystichous (many-rowed) cobs and nonshattering ears. Studies of this kind are needed in many other crop plants.
Crop evolution has usually been characterized using the assumption that phenotypic change is driven by selection for rare mutants that are deleterious in wild plants or by selection for new random mutations that appeared after cultivation began. Associated single trait–single gene models are also deeply rooted in domestication studies. However, it is increasingly recognized that (1) inheritance of domesticated traits is often complex (simple Mendelian segregation is not indicated) and does not involve straightforward human selection on mutations that occurred as rare variants in wild ancestral or early cultivated populations; (2) some “domestication genes” controlling crucial phenotypic attributes are in fact regulatory genes, for example, in maize (tb1 and tga1, controlling plant architecture and the formation of naked grains, respectively), rice (sh4 and qSH1, controlling shattering), and tomato (fw2.2, underwriting fruit size increase); (3) these and other identified genes control developmental pathways in specific immature tissues and organs that lead to adult phenotypic variability; (4) phenotypic outcomes are significantly influenced by the external environment as well as by regulatory and other gene-gene interactions; and (5) considerable cryptic variation occurs in at least one crop progenitor, teosinte, that provides preexisting genetic material that can translate into multiple developmental trajectories (e.g., Doebley 2004, 2006; Jones and Brown 2007; Lauter and Doebley 2002).
Doebley (2004) describes what happens in crosses between teosinte and maize as a result of gene actions that alter the trajectories of plant development: “The dichotomies of single versus paired spikelets, shattering versus nonshattering ears, soft versus hard glumes, and two- versus multiranked ears are striking when one compares maize and teosinte. However, in F2 families, these discrete classes blur into a continuum of phenotypes, and novel phenotypes and interactions appear” (54). All of this means that the domestication process involved simple unconscious/conscious selection targeted at specific mutations controlling discrete (single) traits and more complex plant-people-genetic-environmental interactions that often played out during plant ontogeny and involved humans “reconfiguring” (Lauter and Doebley 2002:341) the preexisting pool of variability into new combinations. Present evidence suggests that in some crops, the latter may have been more important than the former. Gene expression appears to have been very important, and it probably gave early cultivators a great deal of phenotypic variation to work with in some crops and traits, especially as hybridization between crops and wild ancestors was initially common. This may have either slowed down or speeded up the domestication process. Furthermore, preexisting variation for plasticity would have enabled crops to more quickly adapt to new ecological circumstances as they were brought from their natural habitats into cultivated fields and dispersed out of their areas of origin, and it likely in part determined the amount of landrace diversification that was possible after domestication. Maize is a classic example of a domesticated species that was capable of adapting successfully to many different environmental circumstances, and prehistoric peoples across the New World created many different races from it, neither surprising in light of how much variation and plasticity are present in its wild ancestor.
Additional research will provide more detailed information about how gene expression and multiple possible developmental trajectories influenced the creation of domesticated phenotypes, and this type of work will be important for animal domestication as well (e.g., Dobney and Larson 2006). What has so far been largely neglected in domestication studies and should be an important part of future investigations is the role of what West-Eberhard (2003) calls “environmental induction” in creating new phenotypes from ancestral populations (Gremillion and Piperno 2009a). This is a crux of the phenotypic plasticity argument as developed by West-Eberhard—that alternative phenotypes can be induced directly by environmental change and subsequently genetically assimilated/accommodated (genetically stabilized) under the proper conditions. It is known, for example, that maizelike phenotypes in plant architecture can be induced in teosinte by environmental stresses such as lowered temperatures and light intensity (West-Eberhard 2003). Given that crop domestication first occurred shortly after the environmental perturbations that marked the end of the Pleistocene, human environmental modification was significant starting during the early Holocene, and human care for/selection of novel phenotypes in secure ecological niches—cultivated fields—would lead to the genetic assimilation necessary to stabilize the new phenotypes, the role of natural- and human-induced environmental induction as a substitute for and/or complement to artificial selection in engineering some of the major phenotypic steps in crop evolution should be investigated (see Gremillion and Piperno 2009a for possible examples relating to Chenopodium seed attributes and maize branching and inflorescence architecture).
Predomestication and Nondomestication Cultivation
Newer and more highly detailed archeobotanical information indicates that rather than a rapid appearance of domesticated plants shortly after cultivation began, protracted periods of predomestication cultivation (PDC) and even instances of nondomestication cultivation (NDC) of cereals and pulses occurred in the Old World. PDC goes hand in hand with what appears to have been a nonsimultaneous development of the suite of traits that make up the domestication syndrome (e.g., Cohen 2011; Dillehay et al. 2007; Fuller 2007; Piperno and Dillehay 2008; Weiss, Kislev, and Hartmann 2006; Weiss and Zohary 2011; Willcox, Fortnite, and Herveux 2008; Zeder 2011). Some important implications from the accumulated data are obvious. Domestication took longer after cultivation was initiated and was a more complex process than was thought, involving different kinds and thresholds of (often competing) natural and artificial selection pressures from region to region and cultivar to cultivar. The cultivation of major crop species may have arisen independently more than once. Productive and stable food production could be based on morphologically wild plants. And if, as seems likely, PDC plants spread, defining a single geographically localized area as the cradle of origin for the domesticated plant would prove difficult because multiregional processes involving gene flow were at work (Allaby, Fuller, and Brown 2008; Brown et al. 2009; Jones and Brown 2007).
I previously suggested that the commonplace focus on domestication as the preeminent event in human/plant relationships is perhaps misplaced and that the crucial shift we may want to understand is the origins of plant cultivation, when people began to repeatedly sow and harvest plants in plots prepared for this purpose (see Piperno 2006a for a more detailed discussion). Given the evidence brought to light more recently, I see no reason to change this view. Less attention has been paid to PDC and NDC issues in the New World, but some highland and lowland data already suggest that they were components of early food production. At Guilá Naquitz Cave, highland Mexico, a type of morphologically wild runner bean (Phaseolus spp.) was present between ca. 10,600 and 8500 BP in quantities that suggested to the investigators that people were artificially increasing their density by cultivating them (Flannery 1986). The plant was never domesticated. In the Zaña Valley, Peru, hulls of peanuts (the nuts themselves did not survive) dated to 8500 BP do not exhibit some features of modern domesticated plants, nor do they conform to a known wild species (Dillehay et al. 2007). Predomestication cultivation of this plant and its transport out of its area of origin in southern South America before it acquired the full package of domesticated characteristics is thus indicated. Peanut nut starch grains recovered from human teeth, on the other hand, are identical to those from modern varieties of Arachis hypogaea and are unlike starch in modern wild peanuts closely related to A. hypogaea analyzed so far (Piperno and Dillehay 2008). More work is needed on wild peanut starch, but the suite of phenotypic traits that characterize domesticated peanuts may not have developed all at once.
In fact, as discussed in detail elsewhere (Piperno 2006a:162–163; Piperno and Pearsall 1998), PDC and NDC scenarios for agriculture, even if they were not called that, have long seemed to a number of investigators to be particularly well suited to the Neotropics (e.g., Harlan 1992; Harris 1989). Even today, many horticultural plots still contain morphologically wild plants that do not change their phenotypic characteristics even after many years of persistent cultivation. Crops grown for their belowground organs (e.g., Calathea and Dioscorea spp.) and many palms and other fruit trees are among the most stubborn taxa. Instances of PDC and NDC should be sought for more widely in the New World by using multifaceted archaeobotanical data sets and allied paleoecological evidence whenever possible.
Coming full circle back to the discussion that initiated this section, the complex inheritance of and multifactorial influences on domestication traits may have played roles in slowing domestication. Furthermore, the effects of gene expression and possible environment-on-phenotype influences at the onset of plant cultivation and after may have led to pools of interesting phenotypic diversity in early crops that people picked through and discouraged or encouraged over a protracted period before the domesticated examples that we identify archaeologically emerged.
The Why of Food Production
Human Behavioral Ecology and Agricultural Origins
Piperno and Pearsall (1998) and Piperno (2006a) offered an account of why and when Neotropical agricultural origins occurred that is rooted in human behavioral ecology (HBE), especially optimal foraging theory, and they suggested that HBE models had potential relevance in other regions of the world. Explanations such as population pressure and changing social relationships (e.g., competitive feasting) used by some investigators to account for agricultural beginnings elsewhere (Piperno and Pearsall 1998:10–18) did not and still do not appear to be important in the Neotropics. Before discussing the relationships between agricultural origins, HBE, and other explanations for the transition to food production, I should reiterate my belief that the near synchroneity of food production origins in at least seven widely dispersed and ecologically disparate regions of the world when or shortly after the world’s fauna and flora were experiencing profound shifts driven by the end-Pleistocene ecological perturbations should cause us to look for common underlying processes that may have been influential in the transition wherever it occurred. Put at its simplest, it is difficult to believe that near-synchronous origins of such a major economic transformation were a coincidence, an accidental convergence of disparate regionally specific underlying factors in widely dispersed and different social systems. This is not to say that identifying specific aspects of Neolithic developments on a region-by-region basis is not important, only that we should in a deliberate way look for underlying commonalities rather than immediately emphasizing putatively unique, supposedly causal local or regional variables that may turn out to be more apparent than real.
The basic concepts and uses of HBE in archaeology are by now well known and discussed and will not be reviewed here in any detail (for recent literature reviews, see Bird and O’Connell 2006; Lupo 2007). In brief, HBE is concerned with the adaptive plasticity of the human phenotype in response to variation in its particular ecological and social environment. It assumes that behavior is shaped by evolutionary forces and asks “why certain patterns of behavior have emerged and continue to persist” (Bird and O’Connell 2006:144). The emphasis on a flexible phenotype means that HBE’s engine for change resides in human decision making. HBE is embedded in archaeological theory and empirical testing to a degree such that it is viewed along with dual inheritance theory as a primary and distinct research tradition in evolutionary archaeology (e.g., Shennen 2008). As such, it has been and will continue to be a focus of discussions, both incisive and off the mark. Examples of the latter (Smith 2009; Zeder and Smith 2009) make faulty claims that explanations derived from HBE are analogous to covering law models or that HBE ignores the active roles of foragers in altering environments (see Gremillion and Piperno 2009b for additional comments). Investigators in fact are drawn to HBE by its focused, hypothesis-driven, problem-solving approaches that, as mentioned, incorporate human agency and dynamic human/environmental relationships. These elements constitute good science and are productive means for “eliminating problematic answers and identifying and pursuing more promising ones” (Bird and O’Connell 2006:171).
Although HBE has been used to study a broad array of economic and social issues, “foraging theory” applications that address subsistence decisions are most common to this point. At the heart of foraging theory is the assumption that, all things being equal, more efficient food procurement strategies should be favored by natural selection over those less efficient, a largely unquestioned premise in nonhuman animal studies. The simplest version of foraging models is called the “optimal diet” model, or the “diet breadth” model (DBM). It employs a straightforward currency—energy—to measure the costs and benefits of alternative resource sets and assumes that humans will have a goal of optimizing the energetic returns of their subsistence labor. Recent research indicating that spatial memory in women is preferentially engaged and most accurately expressed for calorie-rich foods further buttresses the likelihood that natural selection shaped a human proclivity for efficient plant foraging (New et al. 2007). As reviewed in detail elsewhere (Piperno 2006a; Piperno and Pearsall 1998:15–18), there are a number of good reasons why the DBM can productively be applied to agricultural origins and dispersals. For one thing, its sometimes counterintuitive predictions regarding resource choice bring new insight to issues such as the onset of “broad-spectrum” subsistence strategies and the links between diet breadth, technology, and resource intensification. Another strength of the model is its ability to elucidate from patterns in the archaeological record important processual questions relating to human economic decision making and its determinants vis-à-vis the focal point of energetic constraints. Importantly, in this regard, paleoecological data can serve as objective informants on past shifts in the availability of different subsistence resources and as proxies of changing returns to labor from foraging, adding valuable information that can be used in conjunction with or in the absence of archaeological subsistence records.
A range of ethnographic information indicates that in the Neotropical forest, as elsewhere, energetic efficiency is a major influence on food procurement decisions (reviewed in Piperno and Pearsall 1998). The following lines of evidence were used to argue that optimal foraging and the relative energetic efficiency of resource sets available to foragers during the late Pleistocene and early Holocene played a major role in Neotropical agricultural origins (details in Piperno 2006a; Piperno and Pearsall 1998). First, a large set of paleoecological data ranging now from the Central Balsas region of southwest Mexico to Bolivia (e.g., Burbridge, Mayle, and Killeen 2004; Piperno et al. 2007) indicates that the shift from foraging to food production began within contexts of rapid and significant changes of climate, vegetation, and fauna occurring at the close of the Pleistocene. Large increases in temperature, precipitation, and atmospheric CO2 levels resulted in vegetational transitions from savanna-like/thorny scrub growth to seasonal types of tropical forest across the Neotropics. Many now-extinct megafauna were replaced by the smaller, fewer, and more dispersed tropical forest fauna found in modern environments.
Second, on the basis of robust sets of ethnographic and ecological data, these environmental perturbations would have significantly lowered the overall efficiency of food procurement for hunters and gatherers in those zones where open-land types of vegetation gave way to forest when the Pleistocene ended. For example, the big game and open-land plants that disappeared were almost certainly higher-ranked resources compared with those offered by the tropical forest, in which animals were far fewer and smaller, carbohydrates were limited and spatially dispersed, and many plants were toxic and required extensive processing before they were consumed. Decreasing foraging efficiency as dietary breadth expanded to incorporate these low-ranked resources was probably an important selection pressure acting on human food procurement strategies during the early Holocene.
Third, on the basis of these factors, along with the costs of plant cultivation estimated from modern small-scale tropical farming, it appears that early Neotropical food-producing strategies probably were more energetically efficient, not more costly, than full-time foraging (see Piperno 2006a for a detailed discussion of this issue for the tropical forest). Finally, following the DBM, people would have initiated the cultivation of some plants when the net return from this strategy exceeded the return from full-time hunting and gathering. In light of shifts in vegetational formations and available resources indicated by paleoecological records, the ca. 11,000–9000-BP time period should have been highly relevant for the initiation of food production. It should be noted that all of the above points are especially relevant to areas that today support or would support in the absence of human disturbance highly seasonal types of tropical forest, where the end-Pleistocene environmental perturbations would have impacted foraging return rates most strongly.
Comparing these points with archaeological data (see above; Piperno and Pearsall 1998), it appears that predictions of optimal foraging and the DBM are well supported and effectively account for when and where food production emerged in the lowland Neotropics. As has been discussed in detail elsewhere, other issues such as why some plants and not others out of the many available were singled out for human manipulation can be elucidated using HBE (Piperno 2006a; see below). The point made above about early farmers’ energetic efficiency should be stressed. Although many investigators assume that early farmers everywhere experienced diminishing returns to their labor, studies in the southwestern and eastern United States and the Near East have also failed to support that notion when it was properly tested using appropriate measures of resource costs and benefits (e.g., Barlow 2002; Gremillion 2004; Simms and Russell 1997). The energetic efficiency of early farming compared with the last hunting and gathering should be reevaluated in other regions of the world. Especially in the cases where foraging theory appears to effectively account for the initiation of food production, it would not be surprising to see a revision of previously assumed cost/benefit equations for foraging versus farming.
The DBM or combinations of DBM and other HBE theories have recently been productively applied to predict under what circumstances and how intensively foragers should become farmers in a number of world regions (for other examples, see, e.g., Barlow 2002; Bird and O’Connell 2006; Gremillion 2004; Kennett and Winterhalder 2006). In the eastern United States (Gremillion 2004), it was found that some of the predictions of the DBM (e.g., that immediate resource returns provide the best measure of their utility) did not fit the archaeological data. In this case, the research led to a better grasp of the variables most important in agricultural transitions (e.g., risk reduction, energetic considerations of processing certain key resources with regard to the seasonal availability of others), and refinements in how HBE models should be applied in this regional circumstance. Students of HBE see this as one of its strongest features. If a good fit between theory and data is not evident, then one moves on to find a better-supported theoretical approach.
HBE and Risk
HBE is a highly flexible research program by no means limited to simple energetic efficiency models and currencies. For example, risk sensitivity, an often-used explanation for why foragers turned to farming, is increasingly being incorporated into HBE research (see papers in Kennett and Winterhalder 2006). It is evident that under a number of circumstances, energetically efficient diets are also risk-minimizing diets and that attention to risk avoidance does not necessarily override attention to energetic efficiency. It is worth repeating here that there is also considerable ethnographic evidence that among tropical foragers and horticulturists, food sharing in the context of extensive household exchange is the most important strategy for mitigating risk (see Piperno and Pearsall 1998:239–241 for a detailed discussion of these issues).
HBE and Coevolution
There is little disagreement that coevolutionary forces and food production origins were entwined. As the general process unfolded, dependency of humans on certain plants increased and vice versa, and both the plants and people involved experienced fitness increases. However, when coevolution is operationalized at more detailed levels of analysis, discordance between the model and empirical archaeobotanical data becomes evident. This was initially discussed by Piperno (2006a), and more examples can now be added.
The Cucurbitaceae and Marantaceae, each with important examples of early domesticated plants, are large families with many edible species that respond to a human presence by quickly colonizing disturbed areas near living places. This makes them particularly suitable for testing the most complete and prominent explication of how agricultural transitions exemplify coevolution, that of Rindos (1984). The fruit rinds of most Cucurbitaceae genera produce high numbers of recognizable phytoliths, and Marantaceae seeds and tubers are similarly rich in diagnostic phytoliths. These durable microfossils provide a record of exploitation unbiased by the preservation factor. Marantaceae tubers also produce many diagnostic starch grains, which would be expected to occur on stone tools that were used to process them. The sites of Xihuatoxtla, Mexico; San Isidro, El Recreo, and Peña Roja, Colombia; Siches, Ecuador; and Zaña Valley, Peru, can be added to those from central Panama and Vegas, Ecuador, as examples of where only Cucurbita (often along with bottle gourd) is recorded among plants from this family. In the sequences from Tehuacan and Guilá Naquitz, Mexico, where macrobotanical preservation was excellent, one additional genus of the Cucurbitaceae, Apodanthera, occurred with Cucurbita and bottle gourd, but rarely, and it was considered a possible modern intrusive. Students of foraging theory would immediately notice that the fruits and seeds of wild Cucurbita are among the largest in the family and therefore would likely provide higher return rates relative to other cucurbit taxa.
A similar example of highly selective exploitation is found with regard to the Marantaceae. Maranta arundinacea and a few species of Calathea, usually solely Calathea allouia, are the only Marantaceae taxa represented in the tropical sites. There are numerous plants from a variety of other families with edible seeds and underground organs that should be visible in phytolith and starch grain records had they been exploited early on with any persistency, but they are not recorded. Moreover, just a few species of wild grasses out of the hundreds that were available to human foragers were manipulated in pre-Columbian Mexico: two species of Setaria (Austin 2006) and Zea mays ssp. parviglumis (Balsas teosinte). Only the latter was demonstrably domesticated. A correlation between grass domestication and seed size is also evident, as teosinte has the largest grains of any annual Mexican grass.
Therefore, Neotropical archaeobotanical data better support predictions from foraging theory that stress the importance of deliberate directed human choice rather than the slowly unwinding reciprocal plant/human interactions involving experimentation with numerous different taxa that would be characteristic of coevolution. Efficiently procuring food, especially in the tropical forest, would necessarily entail paying serious attention to and deriving the most plant calories, proteins, and fats from a relatively small set of the available wild species. Long periods of experimentation and mutualistic relationships with a high diversity of species do not appear to have been associated with squash, Marantaceae, and grass domestication in part because energetic efficiency was probably a major determinant of subsistence decisions.
Conclusions
Discussions regarding the origins and dispersals of agriculture are increasingly based on robust empirical information as new and more precise data from all the contributing disciplines of study become available. With gene expression, phenotypic plasticity, and HBE among the important additions to the how and why of food production, the level of analysis has been raised to incorporate the phenotype of both the plants and humans involved. Given the immense importance that gene expression and developmental plasticity are finding in evolutionary biology, it is likely that they will contribute much to our understanding of how people domesticated plants. Human behavioral ecology similarly is likely to have growing importance in agricultural origin studies. Microfossil data have been shown to be essential lines of evidence in documenting the transition from foraging to farming in the Neotropics. As new sites are discovered and older ones reexamined, much more information can be expected on early agriculture in the New World’s tropical regions.
I thank Doug Price and Ofer Bar-Yosef for inviting me to attend the Temozon conference. I learned much from the stimulating discussions and papers presented there.
Supplement A. Additional Notes for Table 1
Other General Notes
Descriptions and explanations of phytolith and starch grain analytical techniques can be found in the following: Chandler-Ezell, Pearsall, and Zeidler (2006); Dickau, Ranere, and Cooke (2007); Holst, Moreno, and Piperno (2007); Pearsall (2000); Pearsall, Chandler-Ezell, and Chandler-Ezell (2003); Pearsall, Chandler-Ezell, and Zeidler (2004); Perry et al. (2007); Piperno (1988, 2006a, 2006b, 2006c, 2009, forthcoming); Piperno et al. (2000); Zarillo et al. (2008). Maize phytoliths are derived from the glume and/or cupules of the cob and/or the leaves (often both cob and leaf phytoliths were present during the same time periods), while maize starch derives exclusively from the kernels. All arrowroot phytoliths identified have a morphology indicating a domesticated species. All Cucurbita phytoliths are derived from the rind of the fruit.
Site References and Data Notes for Table 1
Central Panama
Aguadulce Rock Shelter (Piperno 1988, 2006c, 2009; Piperno et al. 2000).
Cueva de los Ladrones (Dickau, Ranere, and Cooke 2007; Piperno 2006c; Piperno et al. 1985).
Cerro Mangote (D. R. Piperno and I. Holst, unpublished data). This site is a preceramic shell midden.
Colombia
Middle Porce Sites (Aceituno and Castillo 2005; Castillo and Aceituno 2006).
Middle Cauca, Calima Region (Bray et al. 1987; Herrera et al. 1992; Salgado 1995; D. R. Piperno, unpublished data).
El Recreo is a single-component early preceramic occupation with the radiocarbon dates listed above. The Cucurbita phytoliths have a size indicative of a domesticated species and morphology indistinguishable from Cucurbita moschata (D. R. Piperno, unpublished data). Macrobotanical remains of Cucurbitaceae rinds have been identified from the site and also from the nearby early preceramic site Sauzalito (the latter with associated dates of 
and 
BP) and will be further evaluated to determine whether they are from Lagenaria and/or Cucurbita (Heli Gaspar Morcote Rios, personal communication). At the Hacienda Lusitania, a sediment core sequence, maize pollen first occurs 15 cm below a 5-cm-thick peaty soil dated to 
BP. At the Hacienda El Dorado, another sediment core sequence near Hacienda Lusitania, maize pollen first occurs in a level dated to 
BP. These sites are located near El Recreo and Sauzalito.
Middle Cauca, El Jazmin (Aceituno and Castillo 2005).
Upper Cauca, San Isidro (Gnecco and Aceituno 2006; Piperno and Pearsall 1998; D. R. Piperno, unpublished data). This site is a single-component early preceramic occupation with 14C dates of 
, 
, and 
BP. Evaluation of the avocado seeds indicates they are substantially larger than in wild species and are likely to be cultigens (Gnecco and Aceituno 2006).
Colombian Amazon, Peña Roja (Cavelier et al. 1995; Mora 2003; Piperno and Pearsall 1998). The site is an early preceramic occupation and is the earliest known evidence for human settlement of the western Amazonian rain forest. In addition to the phytolith date of 8090 BP determined on the uppermost stratigraphic assemblage containing the cultivar phytoliths, it has 14C determinations of 
, 
, and 
BP, the latter determined on a carbonized palm seed. Cucurbita phytoliths are of both a size (Piperno and Pearsall 1998) and a morphology (D. R. Piperno, unpublished data) that identify them as derived from domesticated plants, probably C. moschata, and phytoliths from the tubers of leren as well as from its seeds occur (D. R. Piperno, unpublished data). The bottle gourd phytoliths are of a size typical of modern domesticated specimens from Colombia (Piperno and Pearsall 1998).
Colombian Amazon, Abeja (Mora et al. 1991).
Ecuador
Las Vegas (Piperno 2009; Piperno and Pearsall 1998; Piperno and Stothert 2003; Stothert 1985; Stothert, Piperno, and Andres 2003). At Las Vegas site OGSE-67, the first appearance of maize phytoliths is in a 10-cm level bracketed by dates directly determined on phytoliths of 
BP (9020–8440 cal BP; Beta-225645) on a 10-cm level immediately below it without maize and 
BP (6810–6410 cal BP; Beta-225643) on a 10-cm level immediately above it with maize (D. R. Piperno, unpublished data). This chronology is in good agreement with that determined from the Vegas typesite OGSE-80, where maize phytoliths first occur at the end of the Vegas occupation. The earliest 10-cm level with maize at OGSE-67 recently yielded a phytolith date of 
BP (12700–12400 cal BP; Beta-264193), which is anomalously old and rejected; further analysis will evaluate the age of the sample. The 7170 BP date determined from OGSE-80 on the earliest maize phytoliths there may be ca. 500 years too old because the dated phytolith extraction contained both fine and coarse silt-sized phytoliths. The dated phytolith extractions from OGSE-67 contained only fine silt-sized phytoliths and are likely to be more accurate ages (see Piperno 2006c:129–131 for a detailed explanation), excluding the rejected outlier determination. Phytolith dating in addition to detailed evaluations and counts of vertical distributions of different types of phytoliths (D. R. Piperno, unpublished data) confirm again that downward vertical movement is inconsequential and has not skewed phytolith indications of agriculture.
Real Alto (Chandler-Ezell, Pearsall, and Zeidler 2006; Pearsall 1978; Pearsall, Chandler-Ezell, and Chander-Ezell 2003; Pearsall, Chandler-Ezell, and Zeidler 2004).
Brazil
Geral (Bush et al. 2000; Piperno and Pearsall 1998). At this lake, pollen and phytolith disturbance indicators and charcoal frequencies suggesting slash-and-burn cultivation appear at 6800 cal BP. Maize and other cultivar pollen and phytoliths are absent at this time and appear at 3800 cal BP. The characteristics of the earlier disturbance horizon may suggest a cultivation system using crops other than maize.
Peru
Zaña Valley (Dillehay, Netherly, and Rossen 1989; Dillehay et al. 2007, 2010; Piperno and Dillehay 2008). The Cucurbita moschata seeds are matches for those in modern traditional landraces of C. moschata from northern South America in size, morphology, and color and are unlike those in wild or other domesticated species (e.g., Rosas, Andres, and Nee 2004; Wessel-Beaver 2000). Macrobotanical information shows that coca (Erythroxylum novagranatense) was in use at 7950 cal BP (Dillehay et al. 2010).
Siches (D. R. Piperno, unpublished data). Siches was identified as an early preceramic occupation by James Richardson (1973). Reexcavations of the site were carried out by Daniel Sandweiss (2003) and James Richardson in 1995 and 2001. Sediments obtained during the 2001 work were analyzed by the author and yielded the Cucurbita phytolith information. The radiocarbon dates associated with the Cucurbita remains are unpublished and were generously provided by Daniel Sandweiss (personal communication, 2009). The Cucurbita phytoliths do not have characteristics of any wild taxon, of Cucurbita ficifolia, or of semidomesticated Cucurbita ecuadorensis. They are matches for phytoliths found only in modern Cucurbita maxima thus far, but more reference work is needed to establish that the same types of phytoliths do not also occur in C. moschata. Thomas Andres recently discovered Cucurbita maxima ssp. andreana, the wild ancestor of C. maxima, growing in the Peruvian Amazon just east of the Andes (Rosas, Andres, and Nee 2004). Previously, this wild species was known only from Bolivia and Argentina. Molecular work is needed to determine which populations of C. andreana were likely to have given rise to C. maxima, but the presence of the wild taxon in more northerly locations of South America opens the possibility that it was domesticated there or was domesticated twice. Further collecting work will better establish the distribution of this taxon.
Paloma (Pearsall 2003; Weir and Dering 1986; T. Andres and D. R. Piperno, unpublished data). I directly dated a seed recently identified as C. ficifolia by Thomas Andres, and it yielded a determination of 
BP (5900–5740 cal BP; Beta-263860). The seed measured 13 mm long × 9 mm wide. Cucurbita ficifolia was domesticated somewhere in the Andes. Analysis of other plant remains is still in progress by D. M. Pearsall.
Chilca 1 (Kaplan and Lynch 1999; Pearsall 2003).
Quebrada Jaguay (Erickson et al. 2005). Paloma, Chilca, and Quebrada Jaguay are located on the arid coast of Peru and provide macrobotanical evidence for various crops (e.g., lima bean, C. ficifolia, jicama, guava) not presently available from their areas of origin.
Uruguay
Los Ajos (Iriarte et al. 2004).
Notes
Dolores R. Piperno is Senior Scientist and Curator of Archaeobotany and South American Archaeology in the Program for Human Ecology and Archaeology in the Department of Anthropology at the Smithsonian National Museum of Natural History (10th Street and Constitution Avenue, Washington, DC 20560, U.S.A. [[email protected]]).
1. The following definitions are used in this paper. “Cultivation” and “farming” refer to the preparation of plots specified for plant propagation and repeated planting and harvesting in such plots. A “cultivated plant” or “cultivar” refers to those that are planted and harvested, regardless of their domesticated status. “Domesticated species” are those that have been genetically altered through artificial selection such that phenotypic characteristics distinguish them from wild progenitors.
References Cited
Aceituno, F. J., and N. Castillo. 2005. Mobility strategies in Colombia’s middle mountain range between the early and middle Holocene. Before Farming 2005(2), art. 2. Allaby, R., D. Q. Fuller, and T. A. Brown. 2008. The genetic expectations of a protracted model for the origins of domesticated crops. Proceedings of the National Academy of Sciences of the USA 105:13982–13986. Anderson, E. 1954. Plants, man, and life. London: Melrose. Austin, D. F. 2006. Fox-tail millets ( Setaria: Poaceae): abandoned food in two hemispheres. Economic Botany 60:143–158. Barlow, R. K. 2002. Predicting maize agriculture among the Fremont: an economic comparison of farming and foraging in the American Southwest. American Antiquity 67:65–88. Bertoti de Cunha, F., P. M. Nobile, A. A. Hoshino, A. Moretzsohn, M. de Carvalho, C. A. Lopes, R. Catalina, and M. A. Gimenes. 2008. Genetic relationships among Arachis hypogaea L. (AABB) and diploid Arachis species with AA and BB genomes. Genetic Resources and Crop Evolution 55:15–20. Bird, D. W., and J. F. O’Connell. 2006. Behavioral ecology and archaeology. Journal of Archaeological Research 14:143–188. Bray, W. B., L. Herrera, M. Cardale Schrimpff, P. Botero, and J. G. Monsalve. 1987. The ancient agricultural landscape of Calima, Colombia. In Pre-Hispanic agricultural fields in the Andean region. W. M. Denevan, K. Mathewson, and G. Knapp, eds. Pp. 443–481. British Archaeological Reports International 359. Manchester: British Archaeological Reports International. Brown, T. A., M. K. Jones, W. Powell, and R. G. Allaby. 2009. The complex origins of domesticated crops in the Fertile Crescent. Trends in Ecology & Evolution 24:103–109. Burbridge, R. E., F. E. Mayle, and T. J. Killeen. 2004. Fifty-thousand-year vegetation and climate history of Noel Kempff Mercado National Park, Bolivian Amazon. Quaternary Research 61:215–230. Bush, M. B., M. C. Miller, P. E. DeOlivera, and P. A. Colinvaux. 2000. Two histories of environmental change and human disturbance in eastern lowland Amazonia. Holocene 10:543–553. Bush, M. B., D. R. Piperno, and P. A. Colinvaux. 1989. A 6,000 year history of Amazonian maize cultivation. Nature 340:303–305. Carroll, S. B. 2005. Endless forms most beautiful: the new science of evo devo. New York: Norton. Castillo, N., and F. J. Aceituno. 2006. El bosque domesticado, el bosque cultivado: un proceso milenario en el valle medio del Río Porce en el Noroccidente colombiano. Latin American Antiquity 17:561–578. Cavalier, I., C. Rodríguez, L. F. Herrera, G. Morcote, and S. Mora. 1995. No solo de caza vive el hombre: ocupación del bosque Amazónico, holoceno temprano. In Ambito y ocupaciones tempranas de la América tropical. I. Cavalier and S. Mora, eds. Pp. 27–44. Bogotá, Colombia: Instituto Colombiano de Antropología, Fundación Erigaie. Chandler-Ezell, K., D. M. Pearsall, and J. A. Zeidler. 2006. Root and tuber phytoliths and starch grains document manioc ( Manihot esculenta), arrowroot ( Maranta arundinacea), and lleren ( Calathea sp.) at the Real Alto site, Ecuador. Economic Botany 60:103–120. Cohen, D. J. 2011. The beginnings of agriculture in China: a multiregional view. Current Anthropology 52(suppl. 4):S273–S293. Cross, H., R. Lira Saade, and T. J. Motley. 2006. Origin and diversification of chayote. In Darwin’s harvest. T. J. Motley, N. Zerega, and H. Cross, eds. Pp. 171–194. New York: Columbia University Press. Denham, T. 2011. Early agriculture and plant domestication in New Guinea and Island Southeast Asia. Current Anthropology 52(suppl. 4):S379–S395. Dickau, R. 2010. Comparing subsistence and the transition to agriculture in western and central Panama through the integration of macrobotanical, microbotanical, and paleoenvironmental information. In Integrating zooarchaeology and paleoethnobotany: a consideration of issues, methods, and cases. A. M. VanDerwarker and T. M. Peres, eds. Pp. 99–134. New York: Springer. Dickau, R., A. J. Ranere, and R. G. Cooke. 2007. Starch grain evidence for the preceramic dispersals of maize and root crops into tropical dry and humid forests of Panama. Proceedings of the National Academy of Sciences of the USA 104:3651–3656. Dillehay, T. D., H. H. Eling Jr., and J. Rossen. 2005. Irrigation canals in the Peruvian Andes. Proceedings of the National Academy of Sciences of the USA 102:17241–17244. Dillehay, T. D., P. J. Netherly, and J. Rossen. 1989. Middle preceramic public and residential sites on the forested slope of the Western Andes, Northern Peru. American Antiquity 54:733–759. Dillehay, T. D., C. Ramirez, M. Pino, M. B. Collins, J. Rossen, and J. D. Pino-Navarro. 2008. Monte Verde: seaweed, food, medicine, and the peopling of South America. Science 320:784–789. Dillehay, T. D., J. Rossen, T. C. Andres, and D. E. Williams. 2007. Preceramic adoption of peanut, squash, and cotton in northern Peru. Science 316:1890–1893. Dillehay, T. D., J. Rossen, P. Netherly, and A. Karathanasis. 2010. Early Holocene coca production in northern Peru. Antiquity 84:939–953. Dobney, K., and G. Larson. 2006. Genetics and animal domestication: new windows on an elusive process. Journal of Zoology 269:261–271. Doebley, J. 2004. The genetics of maize evolution. Annual Review of Genetics 38:37–59. ———. 2006. Unfallen grains: how ancient farmers turned weeds into crops. Science 312:1318–1319. Erickson, D. L., B. D. Smith, A. C. Clarke, D. H. Sandweiss, and N. Tuross. 2005. An Asian origin for a 10,000 year-old domesticated plant in the Americas. Proceedings of the National Academy of Sciences of the USA 102:18315–18320. Flannery, K. V. 1986. The research problem. In Guilá Naquitz: archaic foraging and early agriculture in Oaxaca, Mexico. K. V. Flannery, ed. Pp. 3–18. Orlando, FL: Academic Press. Fuller, D. 2007. Contrasting patterns in crop domestication and domestication rates: recent archaeological insights from the Old World. Annals of Botany 100:903–924. Gibson, G., and I. Dworkin. 2004. Uncovering cryptic genetic variation. Nature Reviews Genetics 5:681–690. Gnecco, C., and J. Aceituno. 2006. Early humanized landscapes of northern South America. In Paleoindian archaeology: a hemispheric perspective. J. E. Morrow and C. Gnecco, eds. Pp. 86–104. Gainesville: University of Florida Press. Gremillion, K. J. 2004. Seed processing and the origins of food production in eastern North America. American Antiquity 69:215–233. Gremillion, K. J., and D. R. Piperno. 2009 a. Human behavioral ecology, phenotypic (developmental) plasticity, and agricultural origins: insights from the emerging evolutionary synthesis. In Rethinking the origins of agriculture. M. N. Cohen, ed. Special issue, Current Anthropology 50:615–619. ———. 2009 b. Parallel causation does not make a prime mover and historical particulars do not make an explanation: response. In Rethinking the origins of agriculture. M. N. Cohen, ed. Special issue, Current Anthropology 50:709–711. Griffith, M. P. 2004. The origins of an important cactus crop, Opuntia ficus-indica (Cactaceae): new molecular evidence. American Journal of Botany 91:1915–1921. Harlan, J. R. 1971. Agricultural origins: centers and non-centers. Science 174:468–474. ———. 1992. Crops and man. 2nd edition. Madison, WI: American Society of Agronomy and Crop Science Society of America. Harris, D. R. 1972. The origins of agriculture in the tropics. American Scientist 60:180–193. ———. 1989. An evolutionary continuum of people-plant interaction. In Foraging and farming: the evolution of plant exploitation. D. R. Harris and G. Hillman, eds. Pp. 11–26. London: Unwin Hyman. Herrera, L., W. Bray, M. Cardale de Schrimpff, and P. Botero. 1992. Nuevas fechas de radiocarbono para el preceramico en la cordillera occidental de Colombia. In Archaeology and environment in Latin America. O. R. Ortiz-Troncoso and T. Van Der Hammen, eds. Pp. 145–164. Amsterdam: Universiteit van Amsterdam. Holst, I., J. E. Moreno, and D. R. Piperno. 2007. Identification of teosinte, maize, and Tripsacum in Mesoamerica by using pollen, starch grains, and phytoliths. Proceedings of the National Academy of Sciences of the USA 104:17608–17613. Hughes, C. E., R. Govindarajulu, A. Robertson, D. L. Filer, S. A. Harris, and C. Donovan Bailey. 2007. Serendipitous backyard hybridization and the origin of crops. Proceedings of the National Academy of Sciences of the USA 104:14389–14394. Iriarte, J. 2007. New perspectives on plant domestication and the development of agriculture in the New World. In Rethinking agriculture. T. Denham, J. Iriarte, and L. Vrydaghs, eds. Pp. 167–188. Walnut Creek, CA: Left Coast. ———. 2009. Narrowing the gap: exploring the diversity of early food-production economies in the Americas. In Rethinking the origins of agriculture. M. N. Cohen, ed. Special issue, Current Anthropology 50:677–680. Iriarte, J., I. Holst, O. Marozzi, C. Listopad, E. Alonso, A. Rinderknecht, and J. Montaña. 2004. Evidence for cultivar adoption and emerging complexity during the mid-Holocene in the La Plata Basin, Uruguay. Nature 432:614–617. Jones, M., and T. Brown. 2007. Selection, cultivation, and reproductive isolation. In Rethinking agriculture. T. Denham, J. Iriarte, and L. Vrydaghs, eds. Pp. 36–49. Walnut Creek, CA: Left Coast. Kaplan, L., and T. F. Lynch. 1999. Phaseolus (Fabaceae) in archaeology: AMS radiocarbon dates and their significance for pre-Columbian agriculture. Economic Botany 53:262–272. Kennett, D. J., and B. Winterhalder, eds. 2006. Behavioral ecology and the transition to agriculture. Berkeley: University of California Press. Kruglyak, L., and D. L. Stern. 2007. An embarrassment of switches. Science 317:758–759. Kwak, M., J. A. Kami, and P. Gepts. 2009. The putative Mesoamerican domestication center of Phaseolus vulgaris is located in the Lerma-Santiago Basin of Mexico. Crop Science 49:554–563. Lauter, N., and J. Doebley. 2002. Genetic variation for phenotypically invariant traits detected in teosinte: implications for the evolution of novel forms. Genetics 160:333–342. Léotard, G., A. Duputié, F. Kjellberg, E. J. P. Douzery, C. Debain, J.-J. de Granville, and D. McKey. Forthcoming. Phylogeography and the origin of cassava: new insights from the northern rim of the Amazon basin. Molecular Phylogenetics and Evolution. Lupo, K. D. 2007. Evolutionary foraging models in zooarchaeological analysis: recent applications and future challenges. Journal of Archaeological Research 15:143–189. Matsuoka, Y., Y. Vigouroux, M. M. Goodman, J. Sanchez, E. Buckler, and J. Doebley. 2002. A single domestication for maize shown by multilocus microsatellite genotyping. Proceedings of the National Academy of Sciences of the USA 99:6080–6084. Milla, S. R., T. G. Isleib, and H. T. Stalker. 2005. Taxonomic relationships among Arachis sect. Arachis species as revealed by AFLP markers. Genome 48:1–11. Miller, A., and B. Schaal. 2005. Domestication of Mesoamerican cultivated fruit tree, Spondias purpurea. Proceedings of the National Academy of Sciences of the USA 102:12801–12806. Mora, S. 2003. Early inhabitants of the Amazonian tropical rain forest: a study of humans and environmental dynamics. Latin American Archaeology Reports, vol. 3. Pittsburgh: University of Pittsburgh. Mora, S., L. F. Herrera, I. Cavelier, and C. Rodríguez. 1991. Cultivars, anthropic soils and stability. Latin American Archaeology Reports, vol. 3. Pittsburgh, PA: University of Pittsburgh. Motta-Aldana, J. R., M. L. Serrano-Serrano, J. H. Torres, G. Castillo-Villamizar, D. G. Debouck, and M. I. Chacón. Forthcoming. Domestication patterns in wild lima beans ( Phaseolus lunatus L.) from the Americas. In Harlan II: biodiversity in agriculture: domestication, evolution, and sustainability. A. Damania and P. Gepts, eds. Cambridge: Cambridge University Press. New, J., M. M. Krasnow, D. Truxaw, and S. J. C. Gaulin. 2007. Spatial adaptations for foraging: women excel and calories count. Proceedings of the Royal Society B 274:2679–2684. Olsen, K. M. 2002. Population history of Manihot esculenta (Euphorbiaceae) inferred from nuclear DNA sequences. Molecular Ecology 11:901–911. Olsen, K. M., and B. Schaal. 1999. Evidence on the origin of cassava: phylogeography of Manihot esculenta. Proceedings of the National Academy of Sciences of the USA 96:5586–5591. Pearsall, D. M. 1978. Phytolith analysis of archaeological soils: evidence for maize cultivation in Formative Ecuador. Science 199:177–178. ———. 2000. Paleoethnobotany: a handbook of procedures. 2nd edition. San Diego, CA: Academic Press. ———. 2003. Plant food resources of the Ecuadorian Formative: an overview and comparison to the Central Andes. In Archaeology of Formative Ecuador. J. S. Raymond and R. L. Burger, eds. Pp. 213–257. Washington, DC: Dumbarton Oaks. Pearsall, D. M., K. Chandler-Ezell, and A. Chandler-Ezell. 2003. Identifying maize in Neotropical sediments and soils using cob phytoliths. Journal of Archaeological Science 30:611–627. Pearsall, D. M., K. Chandler-Ezell, and J. A. Zeidler. 2004. Maize in ancient Ecuador: results of residue analysis of stone tools from the Real Alto site. Journal of Archaeological Science 31:423–442. Perry, L., R. Dickau, S. Zarillo, I. Holst, D. M. Pearsall, D. R. Piperno, M. J. Berman, et al. 2007. Starch fossils and the domestication and dispersal of chile peppers ( Capsicum sp.) in the Americas. Science 315:986–988. Pigliucci, M. 2005. Expanding evolution. Nature 435:565–566. Piperno, D. R. 1988. Phytolith analysis: an archaeological and geological perspective. San Diego, CA: Academic Press. ———. 1990. Aboriginal agriculture and land usage in the Amazon Basin, Ecuador. Journal of Archaeological Science 17:665–677. ———. 2006 a. A behavioral ecological approach to the origins of plant cultivation and domestication in the seasonal tropical forests of the New World. In Foraging theory and the transition to agriculture. D. Kennett and B. Winterhalder, eds. Pp. 137–166. Berkeley: University of California Press. ———. 2006 b. Identifying manioc ( Manihot esculenta Crantz) and other crops in pre-Columbian tropical America through starch grain analysis: a case study from Panama. In Documenting domestication: new genetic and archaeological paradigms. M. Zeder, E. Emschwiller, D. Bradley, and B. S. Smith, eds. Pp. 46–67. Berkeley: University of California Press. ———. 2006 c. Phytoliths: a comprehensive guide for archaeologists and paleoecologists. Lanham, MD: AltaMira. ———. 2009. Identifying crop plants with phytoliths (and starch grains) in Central and South America: a review and an update of the evidence. Quaternary International 193:146–159. ———. Forthcoming. New archaeobotanical information on early cultivation and plant domestication involving microplant remains. In Harlan II: biodiversity in agriculture: domestication, evolution, and sustainability. A. Damania and P. Gepts, eds. Cambridge: Cambridge University Press. Piperno, D. R., K. Clary, R. G. Cooke, A. J. Ranere, and D. Weiland. 1985. Preceramic maize in central Panama: phytolith and pollen evidence. American Anthropologist 87:871–878. Piperno, D. R., and T. D. Dillehay. 2008. Starch grains on human teeth reveal early broad crop diet in northern Peru. Proceedings of the National Academy of Sciences of the USA 105:19622–19627. ———. Forthcoming. Agricultural origins and their social implications in South America. In The Cambridge prehistory. C. Renfrew and P. Bahn, eds. Cambridge: Cambridge University Press. Piperno, D. R., J. E. Moreno, I. Holst, M. Lachniet, J. G. Jones, A. J. Ranere, and R. Castanzo. 2007. Late Pleistocene and Holocene environmental history of the Iguala Valley, central Balsas watershed of Mexico. Proceedings of the National Academy of Sciences of the USA 104:11874–11881. Piperno, D. R., and D. M. Pearsall. 1998. The origins of agriculture in the lowland Neotropics. San Diego, CA: Academic Press. Piperno, D. R., A. J. Ranere, I. Holst, and P. Hansell. 2000. Starch grains reveal early root crop horticulture in the Panamanian tropical forest. Nature 407:894–897. Piperno, D. R., A. J. Ranere, I. Holst, J. Iriarte, and R. Dickau. 2009. Starch grain and phytolith evidence for early ninth millennium BP maize from the Central Balsas River Valley, Mexico. Proceedings of the National Academy of Sciences of the USA 106:5019–5024. Piperno, D. R., and K. Stothert. 2003. Phytolith evidence for early Holocene Cucurbita domestication in southwest Ecuador. Science 299:1054–1057. Plowman, T. 1984. The origin, evolution, and diffusion of coca, Erythroxylum sp., in South and Central America. In Pre-Columbian plant migration. D. Stone, ed. Pp. 126–163. Cambridge, MA: Peabody Museum, Harvard University. Pohl, M. E. D., D. R. Piperno, K. O. Pope, and J. G. Jones. 2007. Microfossil evidence for pre-Columbian maize dispersals in the Neotropics from San Andrés, Tabasco, Mexico. Proceedings of the National Academy of Sciences of the USA 104:6870–6875. Pope, K. O., M. E. D. Pohl, J. G. Jones, D. L. Lentz, C. Von Nagy, F. J. Vega, and I. R. Quitmyer. 2001. Origin and environmental setting of ancient agriculture in the lowlands of Mesoamerica. Science 292:1370–1373. Ranere, A. J., and R. G. Cooke. 2003. Late glacial and early Holocene occupation of Central American tropical forests. In Under the canopy: the archaeology of tropical rain forests. J. Mercader, ed. Pp. 219–248. Piscataway, NJ: Rutgers University Press. Ranere, A. J., D. R. Piperno, I. Holst, R. Dickau, and J. Iriarte. 2009. Preceramic human occupation of the Central Balsas Valley, Mexico: cultural context of early domesticated maize and squash. Proceedings of the National Academy of Sciences of the USA 106:5014–5018. Raymond, J. S. 2008. The process of sedentism in northwestern South America. In Handbook of South American archaeology. H. Silverman and W. H. Isbell, eds. Pp. 79–90. New York: Springer. Richardson, J. B., III. 1973. The preceramic sequence and the Pleistocene and post-Pleistocene climate of northwest Peru. In Variations in anthropology: essays in honor of John C. McGregor. D. W. Lathrap and J. Douglas, eds. Pp. 199–211. Urbana: Illinois Archaeological Survey. Rindos, D. 1984. The origins of agriculture: an evolutionary perspective. Orlando, FL: Academic Press. Robledo, G., G. I. Lavia, and G. Seijo. 2009. Species relations among wild Arachis species with the A genome as revealed by FISH mapping of rDNA loci and heterochromatin detection. Theoretical and Applied Genetics 118:1295–1307. Rodrigues, D. P., S. A. Filho, and C. R. Clement. 2004. Molecular marker-mediated validation of morphologically defined landraces of Pejibaye ( Bactris gasipaes) and their phylogenetic relationships. Genetic Resources and Crop Evolution 51:871–882. Rosas, R. V., T. Andres, and M. Nee. 2004. The Goldman cucurbit collecting expedition in Peru, 2004. http://www.cucurbit.org/Peru/PeruReport.html .Salgado, H. 1995. El precerámico en el cañón del río Calima, cordillera occidental. In Ámbito y ocupaciones tempranas de la América Tropical. I. Cavelier and S. Mora, eds. Pp. 73–82. Bogotá: Fundación Erigaie, Instituto Colombiano de Antropología. Sandweiss, D. H. 2003. Terminal Pleistocene through mid-Holocene archaeological sites as paleoclimatic archives for the Peruvian coast. Palaeogeography, Palaeoclimatology, Palaeoecology 194:23–40. Sanjur, O., D. R. Piperno, T. C. Andres, and L. Wessell-Beaver. 2002. Phylogenetic relationships among domesticated and wild species of Cucurbita (Cucurbitaceae) inferred from a mitochondrial gene: implications for crop plant evolution and areas of origin. Proceedings of the National Academy of Sciences of the USA 99:535–540. Sauer, C. O. 1950. Cultivated plants of South and Central America. In Handbook of South American Indians. J. Steward, ed. Pp. 487–543. Bureau of American Ethnology, Bulletin 143, vol. 6. Washington, DC: U.S. Government Printing Office. Shady Solis, R., J. Haas, and W. Creamer. 2001. Dating Caral, a preceramic site in the Supe Valley on the central coast of Peru. Science 292:723–726. Shennan, S. 2008. Evolution in archaeology. Annual Review of Anthropology 37:75–91. Simms, S., and K. W. Russell. 1997. Bedouin hand harvesting of wheat and barley: implications for early cultivation in Southwestern Asia. Current Anthropology 38:696–702. Smith, B. D. 2009. Core conceptual flaws in human behavioral ecology. Communicative and Integrative Biology 2:533–534. Stothert, K. 1985. The preceramic Las Vegas culture of coastal Ecuador. American Antiquity 50:613–637. Stothert, K., D. R. Piperno, and T. C. Andres. 2003. Terminal Pleistocene/early Holocene human adaptation in coastal Ecuador: the Las Vegas evidence. Quaternary International 109–110:23–43. van Heerwaarden, J., J. Doebley, W. H. Briggs, J. C. Glaubitz, M. M. Goodman, J. J. Sánchez González, J. Ross-Ibarra, B. S. Gaut, and L. Eguiarte. 2011. Genetic signals of origin, spread, and introgression in a large sample of maize landraces. Proceedings of the National Academy of Sciences of the USA 108:1088–1092. Vrydaghs, L., and T. Denham. 2007. Rethinking agriculture: introductory thoughts. In Rethinking agriculture. T. Denham, J. Iriarte, and L. Vrydaghs, eds. Pp. 1–15. Walnut Creek, CA: Left Coast. Weir, G. H., and J. P. Dering. 1986. The lomas of Paloma: human-environment relations in a central Peruvian fog oasis: archaeobotany and palynology. In Andean archaeology: papers in memory of Clifford Evans. R. Matos, S. A. Turpin, and H. H. Eling Jr., eds. Pp. 18–44. Los Angeles: Institute of Archaeology, University of California. Weiss, E., M. E. Kislev, and A. Hartmann. 2006. Autonomous cultivation before domestication. Science 312:1608–1610. Weiss, E., and D. Zohary. 2011. The Neolithic Southwest Asian founder crops: their biology and archaeobotany. Current Anthropology 52(suppl. 4):S237–S254. Wessel-Beaver, L. 2000. Evidence for the center of diversity of Cucurbita moschata in Colombia. Cucurbit Genetics Cooperative Report 23:54–55. West-Eberhard, M. J. 2003. Developmental plasticity and evolution. Oxford: Oxford University Press. Westengen, O. T., Z. Huamán, and M. Heun. 2005. Genetic diversity and geographic pattern in early South American cotton domestication. Theoretical and Applied Genetics 110:392–402. Willcox, G., S. Fornite, and L. Herveux. 2008. Early Holocene cultivation before domestication in northern Syria. Vegetational History and Archaeobotany 17:313–325. Zarillo, S., D. M. Pearsall, J. S. Raymond, M. A. Tisdale, and D. J. Quon. 2008. Directly dated starch residues document Early Formative maize ( Zea mays L.) in tropical Ecuador. Proceedings of the National Academy of Sciences of the USA 105:5005. Zeder, M. A. 2011. The origins of agriculture in the Near East. Current Anthropology 52(suppl. 4):S221–S235. Zeder, M. A., and B. D. Smith. 2009. A conversation on agricultural origins: talking past each other in a crowded room. Current Anthropology 50:681–691. Zhang, D. P., J. Cervantes, Z. Huamán, E. Carey, and M. Ghislain. 2000. Assessing genetic diversity of sweet potato ( Ipomoea batatas (L.) Lam.) using cultivars from tropical America using AFLP. Genetic Resources and Crop Evolution 47:659–665.
