Published in PNAS last week was another major genetic study rice genetic diveristy, phylogenetics and genome structure. The study lead of Ken McNally at IRRI, entitled "Genomewide SNP variation reveals relationships among landraces and modern varieties of rice" once again confirmed the wide divergence between indica and japonica, as well as the rather distinct and intermediate position of the aus group of rices, adpated to dry cultivation Bangaldesh and thereabouts. Their summary phylogeny illustrated this clearly:
The dataset set size, in terms of numbers of accessions is quite small (20 selected varieties), but that it made of for this is the large coverage of the genome and the detail of genome structure revealed by its focus on single point mutations (Single Neuclotide Polymorphisms or SNPs). They covered 3.6 million base-pairs of sequence and discovered nearly 160,000 SNPs! In all of this the deep divergence of indica and japonica, which must precede the beginnings of cultivation was indicated. Even if morphological domestication traits are shared (such as the reduced shattereing of the sh4 mutation, or the Prog1 gene for erect growth) this must be attributed to much more recent hybridization probable since the start of management/cultivation. I favour a hypothesis of the introduction of introduce domesticated japonica to the Ganges or Indus where hybridization took place with unimproved proto-indica cultivars. (A key distinction here is between cultivation and domestication). Such a hypothesis, which can explain the genetics of domesticated rices is at odds with claims for an early Holocene rice domestication in the Ganges, e.g. at Lahuradewa, which does not fit with the genetic make-up of modern indica. But such early and important hybridization events were probably not the last.
The key conclusion of this study is that this is evidence for a large degree of hybridization between the different domestic rice clades, i.e, between indica and japonica, and between these and aus. This is an important conclusion, as it means we can not consider the long-term history of these rice lineages as separate but rather we must see them as entangled. With these lineages crossed (intentionally or unintentionally) over the millennia of cultivation to produce a greater range of varieties and adaptations. This may seem obvious but many (most?) prehistories of rice paint extremely simplistic (and unrealistic) pictures, whether one takes common archaeological accounts such as single origins linked to Austronesian or Austroasiatic expansions, or one takes some recent geneticist just-so stories of decreasing 'primitiveness' of rice varieties as one move north or west from Island SE Asia as indicative of an early (and stealth) origin somewhere in the Islands from which Chinese rice derived (e.g. as postulated in the recent Izawa et al. paper).
We can look forward to some future unravelling of the complex, entangled history of rice spreads and hybridizations.
This study got some on-line publicity, e.g. on ScienceDaily, but which seems to have missed the key point about how major hybridization has been in the evolutionary history of rice!