DISCOVERY OF WILD RICE GRAINS (O. RUFIPOGON) FROM
HEMUDU ANCIENT CARBONIZED RICE
(China National Rice Research Institute,Hangzhou 310006)
(National Institute of Genetics,Shizuoka, Japan)
(Zhejiang Museum, Hangzhou 310006)
It is generally recognized that O. sativa L. came from O. rufipogon and that China is one
of the origins of Asian cultivated rice. But where was the place of origin of Chinese cultivated
rice? There are three opinions: south China; Yunnan and the middle-lower area of the Yangtze
river. Our project focuses on the middle-lower area of the Yangtze; then we must solve a key
problem if there was O. rufipogon in this area 7000 years ago.
In 1973-8, a lot of carbonized rice was found at the Hemudu site. Under carbon 14
dating, it was 6950+130BP. Some scientists studied the shape, grain and phytolith of carbonized
rice, and found that carbonized rice belonged to primitive hsien-keng rice. We will, however,
study whether O. rufipogon were included in this carbonized rice through observation of its
micro-structure. Lastly, we will share our opinion about the place of origin of Chinese cultivated
1. MATERIAL AND METHOD
Zhejiang Museum provided 105 grains of Hemudu carbonized rice for us. We measured
the length, width and length/width ratio of each grain and then some carbonized rice whose awn
were examined under a scanning electron microscope after vacuum plating.
2. ANALYSIS AND RESULTS
2.1 The grain, spikelet length and spikelet width of carbonized rice
There is a rather obvious variation in spikelet length and width between each variety of
carbonized rice. Length is 5.2~8.6mm, width is 2.1~3.1mm and length/width ratio is 1.7~3.2.
These carbonized rices have characters of primitive cultivated rice (Table 1). According to their
length, width and length/width ratio, these carbonized rice can be classified as keng rice and
hsien rice. Statistics show that the length of carbonized hsien-like rice is shorter than that of local
modern hsien rice, about 0.5 mm; the width of carbonized hsien-like rice is narrower than that of
local modern hsien, about 0.19 mm. The length of carbonized keng-like rice is shorter than that
of local modern keng rice, about 0.83 mm; the width of carbonized keng-like rice is narrower
than that of local modern keng rice, about 0.7 mm. All these carbonized rice are lighter than local
modern rice. The weights of these carbonized rice grains renge from 14 to 22g. There is some O.
rufipogon among the carbonized rice. We chose 81 grains of carbonized to classify according to
their awn, broken-awn and non-awn. Then, the base-point of some carbonized rice with awn was
examined under a scanning electron microscope. Four grains with awn were discovered to be O.
rufipogon with the following characters:
(1) long and dense bristles in awn (Fig. 1-A);
(2) traces of natural shedding at maturity in spikelet rachis at its base-point (Fig. 1-B);
(3) narrow grain shape (Fig. 1-C).
(1) Generally, the origin of local cultivated rice needs two conditions:
(1.1) Some O. rufipogon in this area;
(1.2) Ancient people started to collect and cultivate plants. According to archaeological
survey, the weather of the lower area of the Yangtze and Taihu lake 8000 years ago was
warmer than that of the modern Taihu area. It was like the current temperature of south
China: Guangtong, Guangxi, Hainan etc. Although some history books record there was
some natural wild rice in Yangtze river area, as yet we are not sure that these wild rices
belong to O. rufipogon. There is plenty of O. rufipogon below 28 N Lat. and a lot of
natural wild rice between Hunan and Jianxi, but we never find natural wild rice in
Zhejiang, Jiangsu, Anhui and Hubei now. The analysis of the micro-structure of Hemudu
excavated rice shows there was much O. rufipogon in Hemudu 7000 years ago.
(2) There are four grains of carbonized rice whose length/width ratio is 2.88-3.10, although
their length is not smaller, they still belong to O. rufipogon. The length/width ratio of some O.
rufipogon is 3.0 in Guanxi. Some scientists thought O. nivara was the primitive rice of southern
Asia because its length/width ratio is 3.0. Some scientists thought that keng rice and hsien rice
have a common ancestor, and Chinese keng rice was from keng-like O. rufipogon. On the other
hand, as we cannot measure ancient Taihu rice, can we infer that four grains of carbonized rice
are keng-like O. rufipogon?
(3) We found there were four grains of O. rufipogon among a lot of carbonized rice. This
showed that O. rufipogon mixed with the primitive cultivated rice 7000 years ago in Hemudu.
This phenomenon is the natural result of differentiation without human action. Evidently, the
discovery of O. rufipogon from Hemudu ancient carbonized rice provided stronger evidence
about the theory of middle-lower Yangtze river origins.
Fig. 1 The micro-structure of Hemudu excavated rice
A. Carbonized wild rice grain
B. The awn of carbonized wild rice
C. The base-point of carbonized wild rice
D. The base -- point of carbonized Hemudu cultivated rice
Table 1 Comparison of rice grains between Hemudu excavation and local varieties in Taihu
Origin N length mm width mm Ratio l/w 1000 GW
Mean Range Mean Range Mean Range Mean Range
Excavated 43 6.61 5.2-7.4 2.91 2.7-4.1 2.3 1.7-2.5 20 14-?
at Hemudu 62 7.46 6.6-8.6 2.75 2.1-4.1 2.7 2.5-3.2 20 14-?
Local 276 7.44 6.7-8.4 3.61 2.7-4.1 2.1 1.7-3.3 24.2 16.2-?
in Taihu 205 7.96 5.4-10.2 2.94 2.0-4.4 2.7 2.2-3.8 22.6 15.6
Table 2. The length and denseness on awn in Hemudu carbonized wild rice
Rice Bristle length (um) Bristle density (per mm)
Hemudu 108 120
carbonized 110 250
wild 114 140
rice 150 100
Mean Range Mean Range
O. rufipogon (n=21) 94.3 50-155 315.6 100-750
41.3 27-58 166.7 35-50