Presumed Domestication? Evidence for Wild Rice Cultivation and Domestication in the Fifth Millennium BC of the Lower Yangtze Region

Article excerpt

Presumed domestication

In some legal traditions, people are presumed innocent until proven guilty. In the study of agricultural origins it is perhaps prudent to presume plants are wild until evidence can be found to indicate domestication. This has not, however, been the convention in the archaeology of East Asia, where domestication is taken for granted, unproven and unquestioned. The recent research report by Jiang and Liu (Antiquity 80: 355-61), unfortunately continues this tradition. It seems a curious fact that little discussion has ever been devoted to wild rice foraging in Asia, which logically must have preceded agriculture, or how this might appear archaeologically. We believe that the rice chaff that was used to temper Shangshan pottery will most likely turn out to be a product of foraging, and we would like to take this opportunity to consider a larger body of available evidence from the Lower Yangtze area (Figure 1) that suggests the process of rice domestication came to an end (full domestication) closer to 4000 BC after an extended period, of a millennium or more, of pre-domestication cultivation and presumably a much longer period of wild rice use by foragers.


Writings on rice in the Lower Yangtze contain a paradox. It has been suggested that rice assemblages contain a mixture of indica and japonica cultivars, or just indica or just japonica, or intermediate 'ancient' forms of rice (e.g. You 1976; Zhou 1981; 2003; Li 1985; Zhao & Wu 1987; Oka 1988; Bellwood 1997: 206; Zhang 2002), but such claims are riddled with contradictions between different scholars looking at the same material, and all such claims are predicated on a now disproven theory of rice origins. In the 1980s prominent botanists, especially Oka (1988) but also T. Chang (1989), favoured a single origin for rice followed by differentiation into indica and japonica subspecies under cultivation. This scenario, however, is no longer tenable as data accumulated through newer genetic techniques indicates that indica and japonica are phylogenetically distinct, and represent separate domestication events from distinct progenitor species, Oryza nivara for indica and Oryza rufipogon (sensu stricto) for japonica (e.g. Sato et al. 1990; Sano & Morishima 1992; Chen et al. 1993; Wan & Ikehashi 1997; Cheng et al. 2003; Vaughan et al. 2003; Li et al. 2004; for nomenclature, see Vaughan 1994). Indeed, the most comprehensive study to date indicates two separate domestication events for japonica rice, in the South China region, and two domestications of indica rices in South Asia or western Southeast Asia (Londo et al. 2006). This evidence has only begun to be considered in the archaeological literature recently (e.g. Crawford & Shen 1998; Jones & Brown 2001; Fuller 2002: 297; Sato 2002). It should be noted that this genetic evidence overturns the assumption of a single Asian rice origin, which remains prominent in many textbooks (e.g. Bellwood 1997; Higham 2005). While this alone should call into question older archaeobotanical descriptions, since the dominance of 'indica' in early China should be impossible, in fact the attribution of these early finds to indica or japonica domesticated rice is not supported by available morphometric evidence.

Traditional identification ratios do not work unless the presence of wild species can be excluded. Attributions of ancient rice material has been made on the basis of grain or spikelet length-to-width ratios, with ratios of greater than 2.5 attributed to indica and ratios of less than 2.3 attributed to japonica. Recent morphometric data collected on modern rice species indicates that this does not work if wild species, including both the wild progenitors and other Oryza spp. are included (Figure 2). There is much variation in the size and proportions of domesticated and wild rice species today, making it difficult to assign one or a few grains to any given population (Thompson 1996: 176; Harvey 2006). …