AFLP Analysis of a Collection of Tetraploid Wheats Indicates the Origin of Emmer and Hard Wheat Domestication in Southeast Turkey

H. O zkan 2 A. Brandolini 0 R. Schafer-Pregl 1 F. Salamini 1 0 Istituto Sperimentale per la Cerealicoltura , S. Angelo Lodigiano (LO) , Italy 1 Max-Planck-Institut fu r Zu chtungsforschung , Ko ln , Germany 2 Department of Field Crops, Faculty of Agriculture, University of Cukurova , Adana , Turkey - Western agriculture and its most important crop plants are thought to have originated about 10,000 years ago in the Fertile Crescent, a geographical region extending from modern-day Israel, Jordan, Lebanon, and western Syria into southeastern Turkey and along the Tigris and Euphrates rivers into Iraq and Iran (Smith 1995; Bar-Yosef 1998; Diamond 1998; Moore, Hillman, and Legge 2000; Zohary and Hopf 2000; Gopher, Abbo, and Lev-Yadun 2002). Two traditional lines of evidence support that view. First, the geographical distributions of wild progenitors of modern cereal species, among them wild wheats (Triticum urartu, T. boeoticum, T. dicoccoides, Aegilops tauschii), wild barley (Hordeum spontaneum), and wild rye (S. vavilovii), intersect in this region (Nesbitt and Samuel 1996; Moore, Hillman, and Legge 2000; Zohary and Hopf 2000; Gopher, Abbo, and Lev-Yadun 2002). Second, seeds of the wild species occur in early archaeological sites of the region, followed in radiocarbon age and stratigraphic succession by the remains of domesticated forms (Moore, Hillman, and Legge 2000; Zohary and Hopf 2000; Gopher, Abbo, and Lev-Yadun 2002). Recently, molecular evolutionary studies have also begun to weigh heavily on this issue. Genetic identification of the natural stands from which wild crops were domesticated addresses the question of where specifically within the Fertile Crescent humans invented agriculture. The approach involves comparing wild and domesticated populations using molecular markers, which give genome-wide estimates of genetic similarity (Heun et al. 1997; Badr et al. 2000; Martin and Salamini 2000). One of the most promising of these techniques is amplified fragment length polymorphism (AFLP), a polymerase chain reaction (PCR)-based procedure that resolves radioactively labeled electrophoretic bands (polymorphic loci) on sequencing gels. Using AFLPs, the site of domestication of einkorna diploid wheatwas identified from the analysis of 288 AFLP marker loci (Heun et al. 1997). Those results indicated that wild populations from the Karacadag Mountains of southeastern Turkey are more similar to domesticated einkorn than other wild populations are (Heun et al. 1997). Archaeobotanical remains at early settlements near Karacadag, including Cafer Hoyuk (de Moulins 1993), Cay onu (van Zeist and de Roller 19912), Nevali Cori (Pasternak 1998), and Abu Hureyra (de Moulins 2000; Hillman 2000), provided independent evidence for the domestication of einkorn near the Karacadag Range. The publication of the einkorn data (Heun et al. 1997) renewed the debate on the origin of Near East agriculture. Lev-Yadun, Gopher, and Abbo (2000), summarizing the distributions of several cereal and other crop progenitors, reported that these intersect in a small region of southeastern Turkey, circumscribing a small core area that includes Karacadag. Here we address the question of whether the core area was also the place of origin of other additional founder crops of the Fertile Crescent agriculture, using AFLP comparisons at 204 loci from 43 domesticated lines and 99 wild populations of tetraploid wheatsprogenitors of modern hexaploid wheatssampled from primary habitats at known locations. Domesticated emmer wheat, T. dicoccum, has an AABB genome and hulled seeds; a free-threshing form (one that releases seeds during threshing) exists that is called hard wheat (T. durum). These two domesticated forms have a nonbrittle rachis (the ear releases seed but stays intact during threshing), in contrast to the progenitor, T. dicoccoides (wild emmer), the ears of which fall apart at maturity and thus cannot be threshed. Emmer was the most important crop in the Fertile Crescent until the early Bronze Age (Zohary and Hopf 2000), and domesticated forms are present at several early Neolithic archaeological sites. van Zeist and Bakker-Heeres (1982, 1985) report the presence of domesticated emmer in the lowest excavated level of Tell Aswad, dated 10,800 BP (years before present), but suggest that the plant was introduced from elsewhere. Domesticated emmer archaeological remains (de Moulins 2000) are present, but not common, in layers of Abu Hureyra 2 dating 10,400 BP onward. They are preceded at Abu Hureyra 1 by wild T. dicoccoides remains (Hillman 2000). Emmer remains from Cayonu dating from 10,600 BP onward (van Zeist and de Roller 19912) suggest a diffuse cultivation of emmer during that time. Pasternak (1998) describes contemporary-like domesticated grains and spikelet forks of emmer at Nevali Cori. In later Pre-Pottery Neolithic B settlements (tables 2 and 14 in Nesbitt and Samuel, 1996, and Helmer et al. 1998, respectively), domesticated emmer is constant and abundant in presence. The dates reported here are calibrated years (BP), that is, they refer to 14C dates that were transformed into calendar years of the absolute dendrochronological record using the data provided by Zohary and Hopf (2000, p. 14) and by Moore, Hillman, and Legge (2000, pp. 130131) and were cross-checked for consistency with the data of Gopher, Abbo, and Lev-Yadun (2002) and Maier (1996). Wild emmer, T. dicoccoides, hybridizes with domesticated tetraploid wheats, and the hybrids are fertile. The species has brittle ears that shatter (disarticulate) at maturity into individual spikelets bearing relatively large seeds. It rarely colonizes secondary habitats. In primary habitats, two morphologically distinguishable types are present (Poyarkova 1988). The geographical distribution reported by Zohary and Hopf (2000; p. 45) includes the western Fertile Crescent, its central part in southeastern Turkey, and areas in eastern Iran and Iraq. Johnson (1975) reported that the species is progressively substituted in the transect from southeastern Turkey into Iran-Iraq by the wild tetraploid wheat T. araraticum. But in the same areas, occasional T. dicoccoides populations are reported to be present among stands of T. araraticum (Tanaka and Ishii 1973). This introduces a problem: T. araraticum has an AAGG genome and does not produce fertile progeny with T. dicoccoides (Maan 1973), but the two species are phenotypically indistinguishable. When sampling T. dicoccoides accessions from several gene banks, we rarely received lines collected in Iran or Iraq. This supports Johnsons (1975) conclusion: A question is whether authentic T. dicoccoides occurs in that area . . . All the tetraploids collected in the Karacadag, in south eastern Anatolia, Lebanon and Israel were T. dicoccoides. All of the tetraploids from Transcaucasia and all of those collected in Iraq and Iran, except two, showed the typical T. araraticum protein electrophoretic pattern. Inte (...truncated)