(1) Taiwan rice cultivation

As mainland Han immigrants supposedly introduced rice to Taiwan, its Gaoshan natives used slash & burn cultivation for miscellaneous grain, yam and millet, but rice was also planted (Laichuan 1954:55-64). Did Malayan rice cultivation slash & burn rice cultivation mentioned in Part 1 occur in Taiwan?.

The Study of Taiwan's Fan People, vol. 1, says millet was the crucial Gaoshan cereal (Taiwan Fan People Research Group 1976:448), with rice introduced much later. Mountain Gaoshan ate various grains but mostly millet, some adding rice, while flatland people ate rice. As all generally depended on slash & burn cultivated millet, dry rice, sweet potato, etc., I exemplify E coastal Amei paddy rice cultivation.

The Amei usually planted millet (crucial), dry rice, sweet potato, dasheen and legumes on slash & burnt land for 3-4 years, then fallowed (Investigation of Old Taiwan Social Customs 1913; 1914; 1956b). Millet and dry rice cultivation and harvest are alike. Historic Amei daily fare was millet and sweet potato, later importing Mao rice from the Bima of E coast Taiwan, japonica and water buffalo from China, and dry rice considerably later.

Brushhooklike tools, axes and spades (lalalu) were used to prepare slash & burnt land for cultivating dry rice. Rows were grooved, seeded and covered like strip-sowing, or broadcast without covering and harvested with bamboo and other small knives (op.cit.1915b:32, 175). While simple millet broadcasting differs, harvesting is the same, using fingers or bamboo knives to collect rice ears to bind them for storage. Only harvesting is identical to the Malayan technique, but some ear processing is common; e.g., the Malayan tool resembles the Amei's Nanshi or Qimi tribal rice processors (papokupokupokan or doudan; op.cit.1913:36; 1914:52) in shape and use, like the SE Asian islands. The Qimi also cook rice with a bamboo tube but only on a special day: "leaving the river, people prepare their meal" by putting water and rice in a bamboo tube and hanging it above the fire to make it delicious (op.cit.1914:50). They boil fish with salt in a betel tree bowl with hot stones, a method displaying close ties with the SE Asian islands.

The mid-N mountain Taiya use a dry rice cultivation technique like the Amei and SE Asian islanders, totally depending on slash & burn using a spade (paietsu or pajie) and 80 cm wood and iron dibble (kekifui) to prepare the field and a bamboo tool (rohe) or small knife to pick rice ears. Ears are dropped in a winnow basket for treading to remove the grain (op.cit.1915a:127-130). Although the Taiwanese accepted paddy rice cultivation after Han immigration, was it all Han? More research is needed because the Malayan technique may have entered Taiwan by sea via the Philippines in ancient time (Luye 1946:35-55), and examples exist in Amei rice cultivation.

A Taiwan Old-Custom Investigation Society report (1915b:31-32) said the Hualian Port Amei cultivated rice using a tabe to plow and water buffalo to pull the kuwabe rake. Re-raking followed irrigation, then a karutsu for planting. 40-day seedlings are transplanted and weeded after 20 and 50 days. Ears are sickled 5" above the root, immediately beaten in a barrel to extract rice, which is dried and stored in an ariri.

Except for an Amei change from long to short plow in 1890-1900, Matsuyama (1985:48) reported few rice cultivation changes in Fengxi village, Hualian County. This resembles Chinese rice cultivation, but is worth noting the Amei retain hoof-trampling. He (op.cit.:39) said farmers "tied 5-6 water buffalo with a rope to bury weeds and rice roots in the knee-deep paddy, using a maguwa to rake the field 1-2 times; then a lattakk (like the Yaeyaha's kurubashia) or samolihi (see aforementioned supitsu) for levelling."

Hoof-trampling is mainly in Jiqi and Xinshe Villages near Fengbin Village. We don't know if the Han brought it, but Matsuyama lists two possibilities: (1) paddy rice cultivation was imported from China and then SE Asia; and (2) Taiwan rice cultivation was introduced from China's Fujian Province, and impacted by SE Asia. He is vague on the "origin of hoof-trampling, the base of SE Asian technique" (op.cit.:49), but why would it be used in Chinese rice cultivation involving water buffalo, transplanting and harvesting? As historic records in Taiwan or Fujian are lacking, we associate it with SE Asian islands, where it is widespread.

Much research centers on cultural similarities, especially agricultural, between Taiwan hill and SE Asian island people, some stressing ties between slash & burn and dryland cultivation, others linguistic bonds (Miyamoto 1954). While hoof-trampling and cultural ties must be explained, the fact some Malayan rice cultivation traits are in Taiwan is important to study expansion because Taiwan connects SW and SE island rice cultivation.

(2) SW island rice cultivation

Taiwan mountain farmers did not accept paddy rice cultivation until recently, but SW islanders adopted it in late 15th century (Chenzong Dawang Records, vol.150 - Li Dynasty; Yibo 1927:1172-1212; Li Xiyong 1972:453-467). This book states it was common on Yonaguni, Iriomote and Okinawa islands, uncommon on Xincheng, Irabu and Miyako islands, and undeveloped on Hateruma, Kuroshima and Tarama islands; i.e., the rice cultivation center was in SW island's Bachongshan region and Okinawa (Sasaki 1984:30-31). Rice cultivation16 began with Yonaguni islanders who "did not use plows but a small spade to prepare and weed, buffalo trampling paddies in December before seeding and transplanting in January". Early and late rice was harvested in April and May, straightening rice roots for continuous growth, and re-harvesting in July or August. Harvesters "bound rice with straw before storing, flailed it with bamboo and shelled it with mortar and pestle." Iriomote islanders put harvested rice near the paddy. While harvesting is vague, ear picking occurred only on the islands. The book also records some harvest taboos, like scrutinizing behavior so nobody shouts or whistles. After Okinawans sow rice in winter, harvest in May and buffalo-trample before reseeding, they transplant in July and reharvest in winter", suggesting dual-cropping.

These records show SW island rice cultivation has Malayan traits, but to show their expansion the Yaeyaha had ancient hoof-plowing, which persisted in SW islands until recently on Yanaguni, Iriomote, Hateruma, Bama, Ishigaki, Miyako, Okinawa, Yoron, Xijie, Amami, Satsunan and Tanegashima islands, and S Kyu Syu (Tanaka & Furukawa 1982:24-30; Tanaka 1983:319-322). While ties between modern amd Amei hoof-trampling are vague, it was clearly imported from SE islands where Malayan rice cultivation was widely adopted.

It is now necessary to explore factors beyond the SW and SE island hoof-trampling by noting ties. As Jinguan (1955:107-144) and Guofen (1970:121-193) studied the cultural impact of SE islands on SW islands, I introduce recent research showing N Malayan rice cultivation expansion.

As Sato (1990) said some researchers thought Japanese rice comprised several species without elaborating, he assumed rice cultivation expanded from China and tropical islands. Shouyi explains the mutual support of genes Hwc-1 & 2 in hybrid weakening. As Japanese Hwc-2 frequency is 93%, it may be from China, but 7% cannot. Sato shows the mainland China Hwc-2 gene is mostly Indian type (Fig. 7), while the same gene in Japanese rice occurs only between tropical and SW islands, plus Bhutan to N Burma, Thailand and Yunnan Province. Its analysis and control in early or late rice shows: (1) Japanese rice began with China's weakening gene Hwc-2 and long PSP with short temperate-zone Japanese BVP; and (2) it originated (or was imported from) tropical islands with opposing tropical Japanese genes"17. Sato advocated "introducing Yanagida Kunio's southern theory."

As Sato thinks the tropical Japanese type used to be classified as javanica (Part One), a species related to javanica is in original Japanese rice. In fact, Okinawa and Amami island species have higher frequencies of Hwc-2 than Japanese (up to 30%). As Japanese species with Hwc-2 resemble Philippine and Indonesian types (op.cit.:2-4), we may infer rice cultivation expansion from tropical to SW islands.

Taiwan and SW islands are the main regions of imported techniques of Chinese (and later Japanese) rice, but the ancient Malayan technique might have expanded there too, its route N seen in Taiwan slash & burn, hoof-trampling in Taiwan and SW islands, and the existence of temperate-zone and tropical island species.