Contrasting patterns in crop domestication and domestication rates: recent archaeobotanical insights from the Old World

Abstract

Background: Archaeobotany, the study of plant remains from sites of ancient human activity, provides data for studying the initial evolution of domesticated plants. An important background to this is defining the domestication syndrome, those traits by which domesticated plants differ from wild relatives. These traits include features that have been selected under the conditions of cultivation. From archaeological remains the easiest traits to study are seed size and in cereal crops the loss of natural seed dispersal.

Scope: The rate at which these features evolved and the ordering in which they evolved can now be documented for a few crops of Asia and Africa. This paper explores this in einkorn wheat (Triticum monococcum) and barley (Hordeum vulgare) from the Near East, rice (Oryza sativa) from China, mung (Vigna radiata) and urd (Vigna mungo) beans from India, and pearl millet (Pennisetum glaucum) from west Africa. Brief reference is made to similar data on lentils (Lens culinaris), peas (Pisum sativum), soybean (Glycine max) and adzuki bean (Vigna angularis). Available quantitative data from archaeological finds are compiled to explore changes with domestication. The disjunction in cereals between seed size increase and dispersal is explored, and rates at which these features evolved are estimated from archaeobotanical data. Contrasts between crops, especially between cereals and pulses, are examined.

Conclusions: These data suggest that in domesticated grasses, changes in grain size and shape evolved prior to non-shattering ears or panicles. Initial grain size increases may have evolved during the first centuries of cultivation, within perhaps 500-1000 years. Non-shattering infructescences were much slower, becoming fixed about 1000-2000 years later. This suggests a need to reconsider the role of sickle harvesting in domestication. Pulses, by contrast, do not show evidence for seed size increase in relation to the earliest cultivation, and seed size increase may be delayed by 2000-4000 years. This implies that conditions that were sufficient to select for larger seed size in Poaceae were not sufficient in Fabaceae. It is proposed that animal-drawn ploughs (or ards) provided the selection pressure for larger seeds in legumes. This implies different thresholds of selective pressure, for example in relation to differing seed ontogenetics and underlying genetic architecture in these families. Pearl millet (Pennisetum glaucum) may show some similarities to the pulses in terms of a lag-time before truly larger-grained forms evolved.

Fig. 1.

An evolutionary model from foraging to agriculture, with archaeobotanical expectations indicated at the bottom (modified from Harris, 1989). The stages of pre-domestication cultivation are shaded. In this version, domestication is represented as a process of gradual frequency change, with an earlier, more rapid ‘semi-domestication’ and a later, slower fixation of full domestication. The gap in time elapsed between these two can be taken as a minimal estimate of domestication rate (d.r.).

Fig. 2.

Map of south-west Asia, showing the locations of sites with archaeobotanical evidence that contributes to understanding the origins and spread of agriculture. Sites are differentiated on the basis of whether they provide evidence for pre-domestication cultivation, enlarged grains, mixed or predominantly domestic-type rachis data. Note that these sites represent a range of periods, and many sites have multiple phases of use, in which case the earliest phase with significant archaeobotanical data is represented. Shaded areas indicate the general distribution of wild progenitors (based on Zohary and Hopf, 2000, with some refinements from Willcox, 2005). It should be noted that wild emmer (Triticum dicoccoides) occurs over a sub-set of the wild barley zone, and mainly in the western part of the crescent.

Fig. 3.

Pre-Pottery Neolithic B wheat grain measurements from the Jordan Valley (after Colledge, 2001). This indicates the predominance of the larger domesticated-type grains. The gap between the two groups is comparable with that in modern reference material. These sites are dominated by wild-type barley rachis (see below).

Fig. 4.

Scatter-plots of archaeological grain measurements showing the increase in grain size under early pre-domestication cultivation (after Willcox, 2004). (A) Barley grain measurements, comparing early Pre-Pottery Neolithic A Jerf el Ahamr with the much later domesticated material from Kosak Shimali. (B) Comparing early and late Jerf el Ahmar, indicating that shift towards larger grain size had already occurred. (C) Similar comparison of einkorn grains (probably including some rye grains) at early Jerf el Ahmar and Kosak Shimali. (D) Trend towards larger grain sizes over the course of Jerf el Ahmar occupation.

Fig. 5.

Proportions of wild and domesticated barley and einkorn rachis/spikelet remains on early Near Eastern sites, arranged chronologically from left to right, and grouped into broader phases. (A) Barley rachis types, including domesticated (tough), wild (shattering) and uncertain (but more likely wild). (B) Proportions of einkorn spikelet forks/glume bases, including domesticated (tough), wild (shattering) and uncertain (but more likely domesticated). Sites, approximate ages and data sources: Ohalo 2, 21 000–18 500 BC (Kislev et al., 1992); Wadi Hammeh, approx. 12 000 BC (Colledge, 2001); Mureybit, 10 500–9500 BC (Van Zeist and Bakker Heeres, 1986); Iraq-ed-Dubb, approx. 9300 BC (Colledge, 2001); Jerf el Ahmar (early, with two-grained einkorn), 9700–9300 BC (Willcox, 1999, 2002); Wadi Jilat 7, 8800–8300 BC, Wadi Jilat 13, 7000–6500 BC (Colledge, 2001); Aswad, 8700–8000 BC (Van Zeist and Bakker Heeres, 1985; Tanno and Willcox, 2006a); Azraq 31, 7500–7000 BC (Colledge, 2001); Wadi Fidan A, 7500–7000 BC, Wadi Fidan C, 7000–6500 BC (Colledge, 2001); El Kowm, 7500–6800 BC (De Moulins, 1997); Catal Hoyuk, 7400–6800 BC (Fairbairn et al., 2002); Ramad, 7500–6500 BC (Van Zeist and Bakker Heeres, 1985; Tanno and Willcox, 2006a); Magzaliyeh, 7100–6400 BC (Willcox, 2006); Tell el Kherkh, 8600–8300 BC (Tanno and Willcox 2006a, b); Nevali Cori, 8500–8000 BC (Tanno and Willcox, 2006a); Qaramel, approx. 10000 BC (Tanno and Willcox, 2006a); Netiv Hagdud, 9500–9000 BC (Kislev, 1997); Cafer Hoyuk, (IX–XIII) 8300–7700 BC, (III–VIII) 7500–7000 BC (De Moulins, 1997); Kosak Shamali, approx. 5000 BC (Tanno and Willcox, 2006a).

Fig. 6.

Domestication rates in barley and einkorn modelled from archaeobotanical data (based on Fig. 5). Proportion of domesticated type for each site is plotted by a box against a median estimate of site age. A margin of error is indicated by the line which connects the sum of domesticated and uncertain types (indicated by a cross or x). Trend lines are shown based on the lower estimate. (A) Barley domestication rate model, on which period averages are also plotted for the PPNA, Early PPNB and Late PPNB, in which the diamond indicates the proportion of domesticated types and the circle the sum of domesticated and uncertain types. (B) Einkorn domestication rate model; the much later Kosak Shamali has been excluded.

Fig. 7.

Map of East Asia indicating early millet and rice sites, with inset of Yangtze region, showing archaeological sites mentioned in this article: 1, Hemudu; 2, Tianluoshan; 3, Kuahuqiao; 4, Shangshan; 5, Liangzhu; 6, Majiabang area, including Nanzhuangqiao, Luojiajiao and Pu'anqiao; 7, Nanhebang; 8, Maqiao; 9, Songze; 10, Xujiawan; 11, Chuodun; 12, Weidun; 13, Longnan and Caoxieshan; 14, Qiucheng; 15, Longqiuzhuang; 16, Sanxingcun; 17, Lingjiatan; 18, Jiahu; 19, Bashidang; 20, Pengtoushan.

Fig. 8.

(A) Scatter plot of length and width of grains measured in modern populations (15 grains measured from 72 populations; Fuller et al., 2007). (B) Grain measurements from selected Neolithic sites, showing maximum and minimum measured ranges with solid lines and statistical standard deviations with dashed lines (as reported). Note that grains from Kuahuqiao, Bashidang and the lower (Majiabang period) levels (8–6) at Longqiuzhuang fall largely or entirely in the expected immature grain proportions, while the latest grains from Longqiuzhuang, Songze period (level 4), indicate a clear shift towards longer and fatter grains that can be regarded as fully mature, and thus domesticated. Chouden (Late Majiabang) also indicates a shift towards mature japonica-type grains, but suggests local population differences from the domesticated rice at Longqiuzhuang. The small grains from Jiahu are suggestive of wild rice not in the sativa complex, such as O. officinalis. Sources: Kuahuqiao (Zheng et al., 2004b), Longqiuzhuang (Huang and Zhang, 2000), Jiahu (Henan Provincial Institute of Archaeology, 1999), Chuodun (Tang, 2003) and Middle Yangzte Bashidang (Pei, 1998).

Fig. 9.

(A) Graph indicating the expected relative frequency of grains reaching maturity during eight stages, of 2 d each, on an individual rice plant (based on anthesis data of a modern japonica cultivar, from Hoshikawa, 1993). (B) A graph converting this data into potential grain yields to a hunter-gather if this were a morphologically wild plant. This indicates the proportion of grains more than 6 d immature which would differ in grain proportions from mature grains. Although this represents an individual plant it must be assumed that a population of wild rice has individuals that begin this process at different times over a period of weeks. With cultivation there should be a tendency for plants to become synchronous.

Fig. 10.

Size increase in Lower Yangzi rice phytoliths. (A) Measured horizontal length and vertical length of rice bulliform phytoliths from Majiabang period samples (M); (B) measurements from samples of the subsequent Songze (S) and Liangzhu (L) phases. The dashed oval represents the distribution of the Majiabang measurements. Data re-plotted from Zheng et al. (1994, , b) and Wang and Ding (2000).

Fig. 11.

A map of the wild progenitors of Vigna radiata and V. mungo in India in relation to the moist deciduous forests and the region with extensive wild rice populations (based on Tomooka et al., 2003; Fuller and Harvey, 2006). Archaeobotanical finds of the Vigna pulses are indicated which include secure species-level identifications. Sites are numbered: 1, Semthan; 2, Hund; 3, Balu; 4, Kunal; 5, Burthana Tigrana; 6, Mitithal; 7, Hulas; 8, Hulaskhera; 9, Charda; 10, Imlidh-Kurd; 11, Narhan; 12, Khairadih; 13, Malhar; 14, Senuwar; 15, Tokwa; 16, Mahagara; 17, Koldihwa; 18, Balathal; 19, Babar Kot; 20, Rojdi (two phases); 21, Oriyo Timbo; 22, Kaothe; 23, Tuljapur Garhi; 24, Paithan; 25, Apegaon; 26, Bhokardan; 27, Nevasa; 28, Inamgaon; 29, Terr; 30, Golabai Sassan; 31, Piklihal; 32, Hallur; 33, Tekkalakota, Kurugodu, Sanganakallu and Hiregudda; 34, Hattibelagallu; 35, Sanyasula Cave; 36, Veerapuram; 37, Rupanagudi; 38, Ramapuram, Hanumantaraopeta and Peddamudiyam; 39, Kodumanal; 40, Perur. For details of primary sources, see Fuller and Harvey (2006).

Fig. 12.

Metrical data for Indian Vigna pulses indicating no size increase with domestication. (A) Modern length–width measurements in Vigna radiata, V. mungo and their wild progenitors. The separation between them has been adjusted for a high estimate of 20% shrinkage with carbonization and shown on the other plots as a dashed line. (B) Neolithic Vigna finds from South India (before 1400 BC) (data from Fuller and Harvey. 2006).

Fig. 13.

(A) Late Chalcolithic (1400–900 BC), Iron Age (900–200 BC) and Early Historic (200 BC – 400 AD) Vigna finds from South India, indicating size increase. (B) Archaeological Vigna measurements from northern India, including Ganges valley Neolithic, Iron Age and Harappan Bronze Age civilization finds, suggesting a correlation between larger size and plough agriculture (data from Fuller and Harvey, 2006).

Fig. 14.

A synoptic geography of early agricultural developments and precursors in Africa. Shown are the modern distributions of wild Sorghum bicolor and Pennisetum glaucum with genetic connections to the domesticates (after Harlan, 1971; Tostian, 1992). The previously wetter conditions imply a northward shift in the Sahara–Sahel transition (see Gasse, 2000; Marshall and Hildebrand, 2002). Early Holocene ceramic-using forager sites based on Jesse (2003), and mid-Holocene pastoral sites based on Jousse (2004). Early evidence for wild sorghum gathering is indicated (based on Stemler, 1990; Barakat and Fahmy, 1999; Wasylikowa and Dahlberg, 1999). The spread of Near Eastern crops is indicated in the Nile Valley vis-à-vis the pre-ceramic Neolithic distribution in the Eastern Mediterranean. Sites with early pearl millet are numbered: 1, Dhar Tichitt sites (cited in D'Andrea and Casey, 2002); 2, Dhar Oualata sites (Amblard and Pernes, 1989); 3, Djiganyai (MacDonald et al., 2003); 4, Winde Koroji (MacDonald, 1996); 5, Karkarichinkat (cited in D'Andrea and Casey, 2002); 6, Ti-n-Akof (cited in D'Andrea and Casey, 2002); 7, Oursi (cited in D'Andrea and Casey, 2002); 8, Birimi (D'Andrea et al., 2001); 9, Ganjigana (Klee et al., 2004); 10, Kursakata (Zach and Klee, 2003). Historical sites with pearl millet metrical data: 11, Arondo (cited in Zach and Klee, 2003); 12, Jarma (Pelling, 2005); 13, Qasr Ibrim (Steele and Bunting, 1982).

Fig. 15.

Scatter plots of pearl millet (Pennisetum glaucum) grain width vs. thickness. (A) Modern population averages and minima of domesticated populations reduced by 10% to account for shrinkage, compared with modern wild population averages with maxima, reduced by 10%. Dashed line indicates expected separation between wild and domesticated forms. Sources: Brunken et al. (1977) and Zach and Klee (2003). (B) Plots of archaeological site averages and ranges. Early West Africa averages (Birimi, 1700–1500 BCE; Kursakata, 1500–800 BCE) fall in the wild zone although ranges extend into the larger domesticated zone. The earliest finds in India (Surkotada, approx. 1700 BCE) are close to these as are Early Historic (200 BCE–300 CE) Nevasa in southern India. North Indian Narhan (1400–800 BCE) shows a marked shift towards larger sizes comparable with modern domesticates, as does early medieval Qasr Ibrim (Egypt, approx. 450 CE: this find is preserved by dessication and has been reduced to be comparable with carbonized material). Jarma in south-west Libya may show an apparent shift towards somewhat larger grains during the early first millennium CE, comparable with the size found in medieval Senegal at Arundo. Later Medieval Jarma has shifted back towards to near wild size range. Sources: Kajale (1977), Steele and Bunting (1982), Chanchala (1995), Saraswat et al. (1994), D'Andrea et al. (2001) and Zach and Klee (2003) (Jarma data from Ruth Pelling, personal communication).