THOUGHTS ABOUT THE DOMESTICATION OF RICE

Liu, Zhiyi, Zhuzhou Industry Institute, Zhuzhou, Hunan Province:PR China

(Agricultural Archaeology 2000(1):122-128. Transl. by Li Lin; ed. by B. Gordon)

1. An issue worth notice: manner of domestication  

         There is no single solution on the origin of paddy rice agriculture because it involves many factors, rice origin, and farmers and their planting, environment and period. Fortunately, researchers of rice historiography, archaeology, biology, molecular genetics, historical geography, palaeoclimatology and linguistics are consistent on common issues; e.g., cultivated rice derives from common wild rice (Oryza sativa f. spontanea Roschev), not medicinal or verrucous wild rice ((Oryxa officinalis Wall or Oryza meyeriana Baill). Its origin is not India but China on the middle Yangtze River, not Yungui Antiplano, etc.

        But many variations remain on some issues; e.g., some insist perennial common wild rice first evolved to annuals before domestication (Zhang Deci, 1981)⁽¹⁾. Others think the perennial was directly domesticated to cultivated rice, with annual wild rice a hybrid of perennial wild and cultivated rices (Gang Yanyi, 1988)⁽²⁾. Later, Y. Sano, Hiroko Morishima and Hilo-Ishi Oka discovered osculant common wild rice in Thailand midway between perennial and annual common wild rice, concluding cultivated rice derives from osculant common wild rice (Sano, 1980)⁽³⁾.

        Others have different opinions on indica and japonica origins of cultivated rice; e.g.s, (a) both are monogenetic (monophyletic); i.e., common wild rice and human habitation differ due to diverse climate and environment at different latitudes and elevations. Common wild rice does not differentiate into indica and japonica (Gang Yanyi, 1988)⁽⁴⁾ ; (2) both are genetically independent, indica from India (Second, G. 1982)⁽⁵⁾, japonica from China. But, isoenzyme and molecular genetics of Chinese perennial rice shows common wild rice has indica and japonica differentiation, but it is tiny and elementary compared to cultivated rice (Wang Xiangkun et al., 1994)⁽⁶⁾. Research on 10,800 year-old ancient Yuchanyan site cultivated rice from Dao County, Hunan Province, shows ancient cultivated rice with wild, indica and japonica traits (Zhang Wenxu & Yuan Jiarong, 1998)⁽⁷⁾.

        Some issues appear to agree but are not fully understood, e.g., change from wild to cultivated rice.

        Researchers now believe rice was first planted by direct sowing, but replaced by seedling transplant due to disease, low unstable productivity, etc. (Chinese Rice Science, ed. by Chinese Agricultural & Scientific Institute, 1986)⁽⁸⁾. When ancient rice evolved into dry and paddy rice, paddy rice was initially cultivated by direct sowing in the south and north. In the south, the ancient land was fertilized by burning straw and weeds and watering the land (You Xiuling, 1995)⁽⁹⁾. We don't know clearly why direct sowing was used rather than transplanting.

        Actually, the method of domestication is key. Common wild rice evolved from perennial into annual, and then cultivated. Later, it differentiated to indica and japonica, with direct consequences on the method of domestication. This remains unclear.

2. Does ancient rice excavation support direct sowing theory?

        In 1997-8, Chengtoushan site ancient rice in Li County, Hunan, was excavated by Hunan Institute of Archaeology. Partial observations extending timewise confirm its direct sowing⁽¹⁰⁾. Chengtoushan paddy occupied ca. 2 qiutian (90 sq. m) and included an original irrigating system of ditches and ponds.

        In 1992-1995, Caoxieshan site paddy, Suzhou, was excavated by the Nanjing Museum under the direction of Prof. Tengyuang Hongzhi, Agriculture Dept., Japanese Guanqi University, with outside help from Suzhou Museum, Wuxian Cultural Administrative Committee, Jiangsu Agricultural Scientific Institute, Zuozuo Mugaoming, Gongle Shantong, Douchu Biluzhi and Liuze Yinan.

        The 6000 year-old Caoxieshan site has 44 ancient paddyfields, 2 ponds, 6 ditches and 10 wells. The paddyfield rows are from SW to NE, with rectangular, elliptical or irregular pits 0.9-12.5 sq. m in area, but commonly 3-5 sq. m⁽¹¹⁾, with total area ca. 150 sq. m. Without observing sections, floors, footprints or seedling holes⁽¹²⁾, we cannot deduce direct seeding or transplanting. Caoxieshan and Chengtoushan planting may have been alike because the former is only 300-500 years younger and japonica was jointly planted.

        Jianxiang Gu said carbonized rice size and weight in Caoxieshan level 4 resembles modern rice, while that in levels 6-8 are smaller. Rice width in different levels is similar, but weight variance coefficient and standard deviation rises with ascending level (8 to 4), while type variance resembles weight over time; i.e., level 4 carbonized cultivated rice, while level 6-8 rice is transitional from wild to cultivated. Gu believes wild rice evolved to cultivated rice 6300-5500 years ago.

        Chengtoushan rice was in the initial transition stage of incomplete domestication, and it is hard to say of broadcast sowing was used because there are no footprints or seed holes. Although its flat excavation surface had dry and wet parts, we cannot understand their meaning. It is hard to say if it was due to a dry over a wet level.

        We suggest Chengtoushan and Caoxieshan used broadcast sowing of annual wild or cultivated rice grain, its germinating success and survival already achieving broadcast sowing needs. This happened 6500-6000 years ago, 4500-5500 years later than Yuchanyan ancient rice. It is worth careful consideration to see if broadcast sowing was used in >5000 years domestication of wild rice.

        Although common wild rice reproduced perennially, it also reproduced sexually by flowering or annually by seeding. After initial domestication, perennial common wild rice may have mutated or hybridized between perennial wild and cultivated rice. There are minor differences between bi-peaked phytoliths of Yuchanyan cultivated rice, Jiangyong common wild rice and Dao County japonica⁽¹³⁾, as well as modern cultivated rice Xiangzhong indica type II and Chaling common wild rice⁽¹⁴⁾. Both Yuchanyan ancient cultivated rice and Dao County white husk rice were domesticated from Jiangyong common wild perennial rice, while Xiangzhong indica type II was domesticated from Chaling common wild perennial rice, not from annual common wild rice, because Jiangyong and Chaling do not have annual common wild rice.

        Thus, 10,000 year-old Yuchanyan rice was not cultivated from seed, but perennially. Chengtoushan and Caoxieshan broadcast sowing displays practical agriculture, but cannot prove this technique occurred at Yuchanyan.

3. Weak germination does not support direct seeding theory

        It was difficult or impossible for fresh mature harvested perennial wild rice seed to germinate normally under deep dormancy. As it always dropped when mature, it was impossible to broadcast sow in the first domesticated stage.

         Deep dormancy and weak germination with little success was due to (1) husk blocking germination and (2) poorly or undeveloped seed.

        Unhusked perennial wild rice germination is very weak or non-existent at 9ºC. Under ideal conditions (40ºC day/30ºC night), germination is also weak, a viability of 1.3-18.3 and success of 0.7-29.3% - too low for cultivation.⁽¹⁵⁾

         Easy grain casting under irrigation makes poorly or undeveloped germinal buds, with no germination.⁽¹⁶⁾

         So, it is impossible to domesticate common wild rice by perennial seed broadcast with evolution to cultivated rice.

         What about annual wild rice? We also think it is impossible.

         Previous research shows annual wild rice evolved more than perennial, with more germination success and viability. Its reproduction relied on powerful seed, not stalks, suggesting broadcast sowing existed.

        But, China has no natural annual wild rice because it dispersed as weedy rice within cultivated rice. Another reason was the impossibility of ancient farmers collecting sufficient weedy rice seed to domesticate by broadcast sowing.

         Academics believe weedy rice did not evolve directly from perennial wild rice, but by crossbreeding perennial wild and early cultivated rice (You Xiuling, 1990)⁽¹⁷⁾. Other theories are: (1) wild or reverted type via cross or genetic recombination in cultivated rice; (2) proximate ancestral annual of cultivated rice via seed reproduction, crossbreeding and accompanying species, and (3) offspring-generation remains crossbred by wild rice and cultivated rice (Yu Cong & Wu Wanchun, 1996)⁽¹⁸⁾; i.e., annual wild rice from the crossbred of perennial wild and domesticated original or modern cultivated rice, but not through natural evolution (first original or modern cultivated rice, then annual wild rice), its traits from original or modern cultivated rice genes. Another reason is that cultivated rice did not come from seed broadcasting.

         Weedy rice seed dormancy period is strong and irregular, experiments showing Hainan weedy rice period is between indica and wild rice. Under the same situation, indica-like Birui 1d and Fenziman 5d have lower germination (only 22.5%), slowly germinating for 3 months and then decomposing. At 25ºC they germinate after 3 months and will not decompose in 2 cm water, but dormancy lengthens like wild rice (Xu Cong & Wu Wanchun, 1996), making it impossible for ancient people to use this type of seed broadcast.

         In sum, whether perennial or annual, wild rice germination was weak and not used for broadcasting the earliest domesticated wild rice.

4. Wild Rice Transplanting is the Earliest Domestication Method

        As stem preservation was its main reproduction, transplanting was the first domestication mode to convert perennial  common wild rice to annual cultivated rice. Ancient farmers had to transplant perennial common wild rice by seed stem to provide seed production and food. The main reason for transplanting this way was flooding.

         Field investigation shows transplanting was the earliest form of domestication; e.g., the Ximengwa minority in Yunnan Province "transplant wild potato by excavating the tuber, cutting it and retaining its bud, and replanting it there or a new place, with the remainder eaten. It became domesticated over lengthy transplanting in their gardens. Although its root and leaves resemble wild potato, root hairs were less."⁽¹⁹⁾

         "Ximengwa dry rice evolved directly from wild rice, not paddy, although some dry rice varieties can be planted in paddy, while the latter cannot be planted as dry rice. After (1949) liberation of Masan Stockade Village, new paddy fields were in low-lying wet places and plowed by humans or water buffalo. It was first hoed then ploughed, while stony areas were plowed twice, then dry seeded."⁽²⁰⁾

         Perennial common wild rice is a marsh plant, not dry land, while dry rice rose after lengthy paddy rice domestication and then dry land transplanting. Thus, dry rice survives in paddies, while paddy rice cannot survive without much water. That dry rice possibly evolved from wild rice is worth discussion, but the Wa minority transplant dry rice rather than broadcast seedling, which demonstrates original rice domestication.

         Transplanting traits are: (1) paddy seedlings have few roothairs like potato; (2) early stem growth is slow, late is fast, and (3) leaf emergence rate mirrors leaf size, with lengthy growth and high yield⁽²¹⁾; i.e., perennial common wild rice traits change as it was domesticated to annual rice.

         Sown rice roots are shallow, often horizontal and scattered with much root hair, while transplanted rice roots first grow horizontally, then vertically, an important reason why paddy rice initially falls easily. This phenomenon results from stem propagation; more vertical and less lateral roots in transplanted rice are used in reconstructing wild rice. The weakness of stem propagation is increased seed production. Slowly, perennial rice developed into annual rice.

         As root damage while transplanting directly affects survival, length of greening, etc., detecting the original root system of wild rice is crucial. Often, washing of the transplant causes much damage, longer greening period and less grain than direct seeding, an average production loss of 2.1-2.7 %. ⁽²²⁾ While we are discussing modern cultivated transplants, it also applies to original domesticated perennial wild rice.

         Ancient perennial common wild rice was domesticated by transplanting seedlings, ensuring higher survival, shorter greening period and more grain, resulting in a good harvest in the following year.

5. Transplanting Incentive and its Ethnological Evidence

        Why did our ancestors transplant perennial common wild rice? Mainly due to disastrous floods.

         The Rice Derivation chapter in the ancient classic Brief Description of the Origin of Species mentions five floods in the sky and five major grains in the earth. East, west and south, grain was cultivated from seedlings. First, grain was used as food, then seed, and finally in other products, which survived to the present."⁽²³⁾ The five sky floods signify disastrous flood following heavy rain as the reason to transplant perennial common wild rice. Common wild rice usually grows in shallow marshes 30-50 cm deep because depth >1 m is too great, while >2 m drowns it. As heavy flood washes away seedlings, ancient farmers transplanted common wild rice to higher ground, building irrigation works (ponds, ditches & wells) for survival.

         As the above only mentions east, west and south, we can infer the northern absence of rice.

         The above classic's Buckwheat Derivation chapter said "buckwheat preceded the five major grains, broadcasting its seed which grew luxuriantly. It was food before other grain."⁽²⁴⁾ Its Seed Origin chapter said "Gouade, the first artisan and sky god, opened a sky door and threw seed to earth, where it grew to trees atop mountains, buckwheat on flatland and hemp in the foothills.⁽²⁵⁾" Unlike buckwheat, hemp and trees which rely on seed propagation, paddy rice was cultivated by transplanting.

         Finally cultivate has an ancient Chinese meaning, its meaning in Yi dialect is "pile or roll up" (move, hole, domesticate)⁽²⁶⁾; i.e., transplanting common wild rice with earth adhering to root (pile is earth up, move or transplant and hole is the initial shape of the paddy field. Domestication is concious cultivation.

         In ancient Yi, "plant", "transplanted rice seedlings" and "paddy field" have the same pronunciation (te³³ pron. "de"), while "cutting transplant" is tshou (pron. "chou"), "broadcast" is SI⁵⁵ & Si³³ (pron. "xing" or "si"), "dig" is du²¹ (pron. "du"), "bury" is du⁵⁵ (pron. "du" like "plant"), "rice seedling transplant" and paddy field" is te³³, remote from "broadcast" SI⁵⁵ or Si³³, meaning digging or burying, not broadcasting. "Pile" is bury, "roll" is digging; transplanting rice seedlings means burying them, and plant is bury and dig (dig first and then bury).

         The same source for ancient Yu or Zhuang language says "plant" is ndaem (pron. "dai mu"). "Plant and grow" is also ndaem, "plant and cultivate" is ung or gang (ung means plant, gang means catch up) ; "transplant rice seedlings" is ndaemnaz (pron. "dai mu na"); "transplant cutting" is baek (pron. "bai"); "interrupt a conversation" is cabbak (pron. "cha ba"); "transplant" is senjndaem (pron. "shen dai mu"); "move" is daen or ndaem; senj; coek; nod; doeng; soenh; gyod, etc., and "field" is naz (pron. "na"). Thus, the earliest paddy rice cultivation was by "seedling transplant", not "broadcast". Because "scatter" is vang (pron. "wang"), "broadcast" is vangdoek (pron. "wang duo"), completely different from daem of "plant", "plant and cultivate", "transplant seedlings" and "transplant".⁽²⁷⁾

         Daem in Zhuang and te³³ in Yi mean "dig" or "bury", but "dig" in Zhuang is Vat (pron. "wa") and "bury" is man (pron. "man"), different from Yi. As daem in Zhuang came from Yi, paddy rice cultivation is prior to Yi, then spread to Yu.

         Some scholars believe early rice farmers were Yiyang speakers of the Tongtai minority of ancient Yu. ⁽²⁸⁾But according to words like "plant", "planting seedlings", "scatter", "dig" or "bury" in Yiyang (Yilao, buyang), there are too many differences with Yi; e.g., "plant seedlings" was obviously transmitted with "plant", "dig", "bury", and "transplant seedlings". ⁽²⁹⁾

        Ancient Yi differentiation is lowest, Yilao highest and Buyi midway. "Plant", "seedlings" and "cultivate" in the latter two were obviously borrowed from Yi. We deduce early rice farmers were 10,000 year-old Yi people, not ancient people from Yiyang or Yu 6,000 years ago.

References:

(1) Zhang, Deci: Morphological Research on Asian Rice and its Wild Species as well as Identification in the Australian New Community. Agricultural Science and Technology Translation Collection, 1984(2). Transl. by Xu Yunbi, Zhejiang Agriculture University. Original text: N.Q.Ng & T.T.Chang. Biological J. of the Linnean Society (1981)16.

(2)(4) Oka, H.I. 1998. Origin of Cultivated Rice, Chapter 7.

(3) Sano, Y.; H.Morishima & H.I.Oka. Intermediate Perennial - Annual Population of O. perennis found in Thailand and its Evolutionary Significance, Botany Magazine, Tokyo 93:291-305, 1980.

(5) Second (Sai, Kongde), G. 1982. Origin of the Genetic Diversity of Cultivated Rice. Study of the Polymorphism Scored at 40 Isozyme Loci. Japanese Journal of Genetics, 57:25-27.

(6) Xiangkun Wang, Hongwei Cai, Chuanqing Sun, Zhenshan Wang & Hanhua Pang: Origin of Chinese Common Wild Rice and Discussion on some Differentiation between Indica and Japonica. Originally in Chinese Rice Science 1994, 8(4):205-210, now Origin and Evolution of Chinese Cultivated Rice (ed. by Xiangkun Wang & Chuanqing Sun), Chinese Agriculture University Publishers, 1996:101-106.

(7)(13) Wenxu Zhang & Jiarong Yuan: Initial Research on Ancient Cultivated Rice in Yuchanyan, Dao County, Hunan Province. Crops Academic Journal 1998, 4:416-420.

(8)(21)(22)(27) Ed. by Chinese Agricultural Scientific Institution: Chinese Rice Culture, Agriculture Publishers, 1986:685;692-697;500;692-697.

(9) Xiuling You: Chinese Rice Culture History, Chinese Agriculture Publishers 1995:149.

(10) Hunan Archeological Research Institute: Chengtoushan Site, Li County, 1997-1998 Summary of Annual Excavation. Cultural Products 1999, 6:4-17.

(11) Jiangxiang Gu, Houben Zou, Minchang Li, Linghua Tang, Jinlong Ding & Qinde Yao: Initial Finding of Rice Crop Agriculture in Majiabing Cultural Period, Caoxieshan Site. Eastern Culture 1998, 3:15-24.

(12) Jiangxiang Gu; letter to the author, August 10, 1999.

(14) Wenxu Zhang. Observed Traits of Chaling Common Wild rice (O. rufipogon), Hunan. Papers of 3rd International Academic Conference on Agricultural Archaeology. Rice and Yen Ti Culture. 1999, Zhuzhou.

(15) Jiaqiu Chen, Zhaoxin Hou, Chugong Tan & Xiucheng Ning: Research on Improving Germination of Wild Rice (ed. by Miaole Wu). Papers of Wild Rice Resource Research, Chinese Science and Technology Publishers 1990:170-173.

(16) Shiying Zhao, Liqin Huang & Xiaofeng Mao: Research on Improving Seed Germination of Wild Rice (ed. Miaole Wu). Papers of Wild Rice Resource Research, Chinese Science and Technology Publishers, 1990:167-169.

(17) Xiuling You: Discussion of Wild Rice Recorded in Ancient Chinese Books (ed. Miaole Wu). Papers of Wild Rice Resource Research, Chinese Science and Technology Publishers, 1990:174-179.

(18) Cong Xu & Wangchun Wu: Ecological Investigation and Identification of Weedy Rice on Hainan Island. China Paddy Rice Science, 1996, 10(4):247-249.

(19)(20) Genpan Li & Xun Lu: Ximengwa Agriculture involving Knives and Hoes. Agricultural Archaeology, 1985, 1:358-370.

(23) Bijie Area Minority Affairs Committee, Guizhou Province, Translated by Yi Language Translation Group: Brief Description of the Origin of Species, Vol. 2, Sichuan Ethnic Publishers, 1991:216-17

(24) Brief Description of the Origin of Species, Vol. 3, Sichuan Ethnic Publishers, 1993:255-259.

(25) Brief Description of the Origin of Species, Vol. 1, Sichuan Ethnic Publishers, 1990:24-27.

(26) Ed. by Literature and History Research Office, Lunanyi Ethnic Autonomous County, Yunnan Province: Yi-Chinese Concise Dictionary, Yunnan Ethnic Publishers, 1984:292; 8.

(27) Ed. by Minority Language and Character Work Committee Research Office, Zhuang Ethnic Autonomous Region, Guangxi Province: Chinese-Zhuang Dictionary, original edition, Guangxi Ethnic Publishers, 1983.

(28) Jinfang Li: New Linguistic Evidence of the Origin of Chinese Rice Culture. Ethnic Languages, 1999, 3:35-41.