Quaternary International xxx (2015) 1e9

Contents lists available at ScienceDirect

Quaternary International
journal homepage: www.elsevier.com/locate/quaint

The beginnings of rice and millet agriculture in prehistoric Japan

Hiroo Nasu a, *, Arata Momohara b
a
Department of Evolutionary Studies of Biosystems, SOKENDAI: The Graduate University of Advanced Studies, Shonan Village, Hayama, Kanagawa 240-
0193, Japan
b
Department of Environmental Science and Landscape Architecture, Chiba University, 648 Matsudo, Matsudo-shi, Chiba 271-8510, Japan

a r t i c l e i n f o a b s t r a c t

Article history: This paper reviews recent archaeobotanical evidence of the beginnings of rice and millet agriculture in
Available online xxx prehistoric Japan, focusing on agricultural weed assemblages from early rice cultivation sites. In this
study we show that the most reliable dispersal timing of rice and millet cultivation was the end of the
Keywords: Final Jomon period, corresponding to the Initial and Early Yayoi period of northern Kyushu. Rice and
Agriculture millet were introduced from China via Korea probably at the same time, as a pair. Early rice cultivation
Jomon
was likely practiced, not in slash-and-burn elds but in primitive paddy elds, which did not possess
Yayoi
clearly dened paddy ridges or compartments, and were constructed in human-managed open wetlands
Rice
Millet
near forests. Millets were probably cultivated on upland farmlands close to rice paddy elds. The use of
Weed berries and nuts which was the subsistence base of Jomon period continued during this time, and rice
and millet agriculture gradually spread up until the Middle to Late Yayoi period.
2015 Elsevier Ltd and INQUA. All rights reserved.

1. Introduction transformation into an agricultural society by the introduction of

The spread of agriculture from its center of origin to marginal
areas is one of the key issues in understanding the cause of cultural
diffusion and subsistence changes in various human societies. In
East Asia, the cultivation of rice and millet was assumed to have
begun in China in the early Holocene (ca.8000e5000 BC) (Fuller,
2007; Cohen, 2011; Zhao, 2011) and subsequently propagated out
of China into the Korean Peninsula ca. 3000e1000 BC (Lee, 2011).
The timing and the process of the introduction of the earliest rice
and millet cultivation in prehistoric Japan have to date been a
controversial subject.
Jomon culture (ca.14000e1000 BC) in prehistoric Japan is
believed to have been a huntingegatheringeshing society; how-
ever, recent archaeobotanical studies revealed that the Jomon
people were not merely hunteregatherers. They probably managed
or cultivated chestnut (Castanea crenata) (Kuri), lacquer tree (Tox-
icodendron verniciuum) (Urushi), hemp (Cannabis sativa) (Asa),
and even soybean (Glycine max) (Daizu) and azuki bean (Vigna
angularis) (Azuki) (Kitagawa and Yasuda, 2008; Crawford, 2011;
Noshiro and Sasaki, 2014; Nasu et al., 2015). However, this
pattern of Jomon subsistence eventually underwent a
* Corresponding author.
E-mail address: nasu_hiroo@soken.ac.jp (H. Nasu).
irrigated paddy rice (Oryza sativa) (Ine) agriculture from the
continent in the Yayoi period (1000 BCe300 AD). Why did this
change happen? To answer this question, we need to rst nd a
clear answer as to when and how rice and millet agriculture was
introduced into the Jomon culture and how was it was adopted in
relation to social and environmental change.
The transition from the Jomon to the Yayoi period in prehistoric
Japan was characterized by the introduction of irrigated paddy rice
agriculture. However, there are reports of the existence of
carbonized/waterlogged rice remains, rice impressions on pottery,
and rice phytoliths in the Jomon period, before the emergence of
clearly dened paddy elds. On the basis of these Jomon rice
reports, researchers have argued that the Jomon people were
farmers who cultivated rice (Sato, 2002). Concomitant occurrences
of carbonized seeds and/or impressions of foxtail millet (Setaria
italica) (Awa) and common millet (Panicum miliaceum) (Kibi) have
led some researchers to suggest that the Jomon people grew rice
and millet through an early slash-and-burn or shifting agriculture,
similar to practices shown in modern ethnological examples
(Sasaki, 2003, 2011). However, to the best of our knowledge, no
archeological evidence of a slash-and-burn agriculture in this
period (Noto, 2002) because of the difculty in detecting the re-
mains of the elds in and around archaeological sites.
One method of investigating the cultivation form of rice and
millet during the Jomon period is to focus on the weeds. Weeds are

http://dx.doi.org/10.1016/j.quaint.2015.06.043
1040-6182/ 2015 Elsevier Ltd and INQUA. All rights reserved.

Please cite this article in press as: Nasu, H., Momohara, A., The beginnings of rice and millet agriculture in prehistoric Japan, Quaternary
International (2015), http://dx.doi.org/10.1016/j.quaint.2015.06.043
2 H. Nasu, A. Momohara / Quaternary International xxx (2015) 1e9

dened as a plant population that interfere with human activities
and that grow well in the human environment, having partially the
nature of both cultivated plants and wild plants (Yamaguchi, 1997).
Since the species composition of weeds differs depending on the
degree of interference with human activities, they can be used to
estimate the degree of human interference when the site was in use.
Using this method, Kasahara (1976a, 1982) pointed out that there
was an increase in paddy weeds between the Yayoi and Jomon pe-
riods. Thus, by looking at the species composition of weeds associ-
ated with crops, it is possible to indirectly estimate the location and
form of the cultivation environment at that time (Willcox, 2012).
In this paper, we examine the manner in which rice and millet
were cultivated, by focusing on the composition of the weeds and
wild grasses that accompany these two crops. First, we will review
recent archaeobotanical records of rice and millet in Japan during
the transition from the Jomon to Yayoi period. Second, we describe
our impression of an initial form of rice cultivation using human-
managed open wetlands before the development of clearly
distinct paddy elds, based on the example of Kitashirakawa-
oiwakecho site in Kyoto Prefecture, Japan, dated to the end of the
Final Jomon period (corresponding to the Early Yayoi period of
northern Kyushu; ca. 1000 BC). This paper expands upon Nasu's
(2014) paper, written in Japanese, with additional data and ideas.
1.1. Rice and millet in the Jomon period
When do rice and millet begin to appear in the archaeological
record in Japan? The recent progress of archaeobotanical studies in
Japan represented by the newly advanced replication method of
seed impressions on pottery fragments, phytolith analysis, and
direct dating of carbonized crop grains, has made it necessary to re-
examine previous records of Jomon period rice and millet remains
(Nakazawa, 2009; Nakayama, 2010; Obata, 2011). Here, we will
present the latest reports on the identication of rice and millet, as
well as researchers' opinions, and problems in the evaluation of
these reports (Fig. 1, Table 1).
1.2. Beginnings of rice cultivation in Japan
Nakayama's (2010) report on rice remains from Jomon and Yayoi
sites, which included data on impressions, carbonized grains,
phytolith, and pollen, is currently the most methodical-body of
data on the subject. According to this report, the oldest examples of
rice remains are phytoliths which were found at Kagoshima Uni-
versity Campus site of the Middle Jomon period. Yet, because
phytoliths are small silica particles found in the soil, we cannot
completely rule out the possibility of contamination from upper
layers into the lower layers, which can occur through the distur-
bance effect of the root systems of plants and organisms in the soil.
Thus, some researchers accept this phytolith record as evidence of
rice cultivation, whereas others do not. Nakayama (2010) evaluated
the data as follows: Reliability of the data is high because quan-
titative analysis shows the phytolith uctuation in each layer, but
there is no other evidence to support the results of this analysis,
such as traces of cultivated elds and paddies, or carbonized rice
grains (p. 209). If rice was cultivated in the Middle Jomon period,

Fig. 1. Map showing Jomon and Yayoi sites mentioned in the text. 1: Kagoshima University Campus, 2: Suitenmukai, 3: Ishinomoto, 4: Minami-mizote, 5: Bunkyo, 6: Itaya III, 7:
Kazahari, 8: Nabatake, 9: Sakamoto A, 10: Kitashirakawa-Oiwakecho, 11: Ryugasaki A, 12: Yazaki, 13: Yashikidaira.

Please cite this article in press as: Nasu, H., Momohara, A., The beginnings of rice and millet agriculture in prehistoric Japan, Quaternary
International (2015), http://dx.doi.org/10.1016/j.quaint.2015.06.043
 H. Nasu, A. Momohara / Quaternary International xxx (2015) 1e9 3

Table 1
Correspondence table of pottery typology, relative age and radiocarbon age for sites mentioned in the text.
14 14
Site Region Pottery typology Relative age Range of C age (un calibrated) Range of C age (calibrated)a
14
Itaya III Chugoku Tottaimon phase I End of Final Jomon 2900e2800 C BP ca. 1200e840 cal BC
14
Nabatake Northern Kyushu Yamanodera type Initial Yayoi 2800e2700 C BP ca. 1020e800 cal BC
14
Nabatake Northern Kyushu Itazuke I type Beginning of Early Yayoi 2600e2500 C BP ca. 820e530 cal BC
14
Ryugasaki Kinki Nagahara type End of Final Jomon 2600e2500 C BP ca. 820e530 cal BC
14
Kitashirakawa Kinki Nagahara type/Funabashi type End of Final Jomon 2600e2500 C BP ca. 820e530 cal BC
a
Calibration was used Oxcal 4.2 programme (Bronk Ramsey, 2009) with IntCal 13 calibration curve (Reimer et al., 2013).

then we could expect to nd carbonized rice in hearths or pit shows that the earliest reliable rice impression is found in the
houses; however, excavations have not found such evidence. Tottaimon pottery stage I (corresponding to Maeike-shiki pottery)
Recently, many carbonized seeds of wild or cultivated soybean and at Itaya III site, Shimane Prefecture (Table 1, Fig. 1). From the Tot-
azuki bean have been found in Middle Jomon pit houses (Aida et al., taimon period, reliable rice records increase, not only of impres-
2012; Nasu et al., 2015). The sizes of these seeds are similar to that sions but also carbonized grains, phytolith, and pollen records
of rice; therefore, if rice had been used and cultivated in the Middle (Nakayama, 2010). At present, the direct dating results of carbon-
Jomon period, its carbonized grains should have been found as well. ized rice grains show 2800e2400 14C BP (ca. 1020e399 cal BC), and
Their absence suggests that rice was not cultivated in the Middle rice remains of an earlier date have not been found (Nishimoto,
Jomon period. Furthermore, even in the southern Korean peninsula, 2009). The earliest paddy elds have also been found from this
rice cultivation was transmitted around 3100 cal BP (ca. 1150 BC) period, such as Nabatake site, Saga prefecture (Kasahara, 1982) and
(Crawford and Lee, 2003; Shoda, 2010; Lee, 2011). In our opinion, Sakamoto A site, Miyazaki prefecture (Urushibata, 2006).
there is little possibility that rice cultivation could have been Thus, many researchers admit that rice was cultivated in Japan
introduced to Japan in the Middle Jomon period. from the end of the Final Jomon period (Initial to Early Yayoi period
If it is difcult to ascertain the presence of rice in the Middle in northern Kyushu), but opinions diverge on the authenticity of
Jomon period as yet, then what about the Late Jomon period? The rice remains that precede this period. More reliable evidence is
record of rice in the Late Jomon period shows an increase in needed from the Late Jomon period and the rst half of the Final
carbonized grains, phytoliths, and impressions (Fig. 1). However, Jomon period. Future rice reports should be cross-checked by ex-
according to Nakazawa (2009) and Nakayama (2010), there are perts in multiple elds, such as pottery typology and plant
reservations concerning all of these records. For example, the identication.
identication of rice impressions on Late Jomon pottery from Ish-
inomoto site in Kumamoto Prefecture has come under question
1.3. Beginnings of millet cultivation in Japan
(Nakazawa, 2009). Based on the photo in the report, we cannot
even nd a basis for determining whether it is rice. Recently
What about millet? Millet that has its origins in East Asia include
discovered rice impressions from the Suitenmukai site, Kagoshima
foxtail millet (Awa), common millet (Kibi), and barnyard millet
Prefecture, indicates the possible presence of rice only in the Late
(Echinochloa esculenta) (Hie). Of these three types, the origin of
Jomon period (Obata and Manabe, 2011). On the other hand, Miyaji
barnyard millet is still unclear, but there is some evidence of
(2013) argued that the pottery itself is not of the Late Jomon period;
cultivation, semi-cultivation, or management of wild barnyard
it is probably Tottaimon pottery, which belongs to the end of the
grass (Echinochloa sp.) during the Initial to Middle Jomon period in
Final Jomon period. The age of rice impressions from the Minami-
northern Japan. Crawford (1983) found carbonized grains (cary-
Mizote site, Okayama Prefecture, is also in doubt, based on their
opses) of barnyard grass that were somewhat large, wide and thick,
pottery type (Nakazawa, 2009). At this site, rice phytoliths have
in otation samples from the Middle Jomon sites of Hokkaido in
been found in the pottery (Fujiwara, 1995). The possibility of
northern Japan, and discovered a trend toward seed enlargement.
contamination of phytoliths in the matrix of pottery is lower than
Crawford (1983, 1992, 2011) and Sakamoto (1988) pointed out the
for phytoliths in soil. As the age of the pottery impressions has been
possibility of barnyard millet domestication in Japan. Yoshizaki
called into question, it is likely that the dating of the pottery con-
(1992) named the Echinochloa Jomon barnyard millet (Jomon-
taining rice phytoliths should be re-examined by multiple experts
Hie). However, archaeobotanical nds of barnyard millet decrease
on the ceramics of this period. Dated carbonized rice grains have
in the Final Jomon or Yayoi period; hence, Jomon barnyard millet
been reported from Kazahari site, northern Japan (D'Andrea, 1992;
cultivation/management was probably a temporal food procure-
D'Andrea et al., 1995). Two radiocarbon ages show 2810 270 14C
ment strategy.
BP (1683e382 cal BC) and 2540 240 14C BP (1271e60 cal BC) from
On the other hand, foxtail and common millets domesticated in
direct dating of carbonized rice grains. The pottery type is of the
China were introduced in Japan around the same time as rice. For
end of the Late Jomon period, however, the radiocarbon dates have
this reason, it is considered that the wild ancestors of foxtail and
a large margin of error, and the calibrated ages indicate a period
common millets were probably not distributed in Japan. Green
ranging from Late Jomon to Early Yayoi. For the Late Jomon period,
foxtail (Setaria viridis) (Enokoro-gusa), the wild ancestor of foxtail
there is a possibility of rice cultivation, but we believe the existing
millet, can be found in various parts of Japan today. There is no
evidence is insufcient.
reliable evidence of green foxtail in the Late Glacial or Jomon period
More reliable evidence of rice cultivation is the increase in rice
in Japan, although there are some records of other wild Setaria
remains found in the Final Jomon period (Nakayama, 2010).
species (S. pumila) (Kin-enokoro) (Sugiura et al., 2011). Probably,
Nakazawa (2013) stated that no evidence of rice remains can be
green foxtail is an archeophyte that was brought from China to
found from the rst half of the Late Jomon period, as all of the rice
Japan as a weed of foxtail millet, or it is possible that green foxtail
records are identied as being from the end of the Final Jomon
survived in many areas of Japan where the soil was disturbed
period because they appear on Tottaimon pottery (corresponding to
during the Jomon period and expanded its habitat after the intro-
the Initial and Early Yayoi period in the northern part of Kyushu).
duction of agriculture. The wild ancestor of common millet has not
The recent assemblage of rice impressions by Nakazawa (2013)
yet been identied, although weedy Panicum miliaceum subsp.

Please cite this article in press as: Nasu, H., Momohara, A., The beginnings of rice and millet agriculture in prehistoric Japan, Quaternary
International (2015), http://dx.doi.org/10.1016/j.quaint.2015.06.043
4 H. Nasu, A. Momohara / Quaternary International xxx (2015) 1e9
ruderale (Inu-Kibi) in Central Asia is one of the candidates (Kimata
and Negishi, 2002). There are also some wild relatives of common
millet in Japan, such as Panicum bisulcatum (Nuka-kibi), but they
are not the direct ancestors of common millet. Therefore, common
millet was not independently domesticated in Japan but introduced
from China as a domesticated plant in the same manner as rice and
foxtail millet.
Archaeobotanical remains of foxtail and common millets have
been found from the Final Jomon period (corresponding to the
Initial-Early Yayoi period of northern Kyushu) (Fig. 1, Table 1). The
earliest evidence of foxtail millet comes from Nabatake site (Initial
Yayoi) in association with rice and paddy eld remains, although
they are not dated. The direct dating results show that the earliest
radiocarbon age of carbonized common millet is 2550 25 14C BP
(801e555 cal BC) from Ryugasaki A site, Shiga Prefecture
(Matsutani, 2006; Miyata et al., 2007). A similar radiocarbon age
(2530 20 14C BP; 794e552 cal BC) was determined for carbon-
ized foxtail millet from the Kitashirakawa-Oiwakecho site, Kyoto
Prefecture (Tomii et al., 2012). These dating results show that
foxtail and common millets reached the Kinki region around the
same time. Recent progress in seed impression studies have
revealed several instances of millet impressions on pottery from
the end of the Final Jomon period, including those found at Yazaki
site in Nagano Prefecture and Yashikidaira site in Yamanashi
Prefecture in the Central Highlands of the Chubu region, as well as
from sites in the Kanto region (Sasaki et al., 2009; Endo and
Takase, 2011, 2012; Nakazawa, 2011; Nakazawa and Sasaki, 2011;
Takase et al., 2011; Endo, 2012; Nakayama and Sano, 2012;
Nakayama and Uruma, 2012). As we have seen, since there are
no reports of millet in the Middle and Late Jomon period, it can be
considered that, as in the case of rice, the earliest evidence of
foxtail and common millets dates from the end of the Final Jomon
period.
Thus, the dispersal timing of rice and millet is thought to be the
end of the Final Jomon period (corresponding to the InitialeEarly
Yayoi period of northern Kyushu), although further conrmation is
needed. With respect to the time difference between the dispersal
of rice and millets, we currently believe that rice and millet were
introduced into Japan about the same time, but that after their
arrival, they spread in different ways. Rice was probably not
actively cultivated at high altitudes in the Central Highlands of the
Chubu region at rst (Nakazawa, 2011; Nakayama and Uruma,
2012).
2. Initial cultivation forms of rice and millet
2.1. Slash-and-burn or paddy eld
Thus, rice and millet are found since the end of the Final Jomon
period. At the time, irrigated paddy elds had already been intro-
duced in northern Kyushu, but clear evidence of such elds have
not been found in many other areas. How was rice cultivated in
these areas? Sato's (2002) DNA analysis has shown that both types
of rice, temperate and tropical japonica, are represented in rice
remains of the Yayoi period. Based on these results, some re-
searchers and lay citizens believe that Japanese farmers in the Jo-
mon period cultivated dry farmland rice (tropical japonica) through
slash-and-burn agriculture. However, tropical Japonica can be
grown in wetlands and paddy elds, as well as on dry land; thus,
the presence of tropical japonica is not necessarily evidence of
slash-and-burn agriculture. Sato's analysis can be interpreted to
mean that tropical japonica was growing in Yayoi paddy elds
because the samples that he used were rice remains from Yayoi
Please cite this article in press as: Nasu, H., Momohara, A., The beginnings of rice and millet agriculture in prehistoric Japan, Quaternary
International (2015), http://dx.doi.org/10.1016/j.quaint.2015.06.043
sites that already had irrigated paddy eld systems. As previously
discussed, the timing of the increase in reliable evidence of rice and
paddy elds in Japan matches to a substantial degree; we do not
necessarily need to assume that rice was grown using slash-and-
burn agriculture techniques. Sasaki (2003) mentioned that rice
cultivation started in a primitive rain-fed paddy eld that was not
irrigated, and it is very difcult to distinguish between paddy elds,
farmland, and slash-and-burn elds.
Kasahara (1976b) deduced that slash-and-burn agriculture was
practiced in the Jomon period by comparing weeds in modern
slash-and-burn elds with archaeobotanical weed seed data. Sub-
sequently, Kasahara (1982) argued that initial rice cultivation was
not specialized between wet paddy elds and dry farmlands, citing
weed data from the Nabatake site, Saga Prefecture. Although one of
the authors of this paper had previously expressed a similar idea
(Nasu, 2004), now we have recognized that the weeds from slash-
and-burn elds, such as Rubus, Vitis, Panicum bisulcatum, Ajuga
decumbens, Cirsium, and Picris, listed by Kasahara (1976b), could be
grown on the ridges of paddy elds and farmlands, not only in
slash-and-burn elds. Our recent weed data from the Cheng-
toushan site, Hunan Province, China (ca.6400e5300 cal BP;
4450e3350 BC) shows that rice and foxtail millet were cultivated in
different elds. Rice was cultivated on the paddy eld or wetland
area around the site, and millet was grown on the farmland at the
site itself (Nasu et al., 2007, 2012). With respect to millet cultiva-
tion, there is some possibility of the use of re in preparing the land
for cultivation. When we take this into consideration, it appears
that rice cultivation on the wet paddy eld and millet cultivation on
the dry farmland were simultaneously introduced in Japan during
the end of the Final Jomon period.
2.2. Wetland rice cultivation before the development of paddy
elds: an example from the Kitashirakawa-Oiwakecho site
The author recently excavated interesting archaeobotanical
materials at a primitive rice cultivation site, Kitashirakawa-
Oiwakecho, in Kyoto Prefecture (Chiba, 2012; Tomii et al., 2012).
The site is located on the northeastern part of Kyoto basin at the
edge of the Kitashirakawa alluvial fan which extends eastward from
Hiei Mountain. Peaty organic sediments are distributed on the Final
Jomon layer of the site. The Final Jomon level is covered by thick
sand deposits as a result of multiple ooding events. Above these
sand layers are silty to clayey sediments of the Early Yayoi period.
The excavation was conducted by Kyoto University during 2009
(Tomii et al., 2012).
Rice and millet were found from the peaty wetland area of the
site, but there are no clear paddy eld remains. Radiocarbon dating
and pottery typology point to the end of the Final Jomon period.
The weed ora of the site shows in detail the specialized landscape
of the wetland. The southwest area, where uncarbonized rice chaff
were discovered, was a relatively open area dominated by wetland
herbs and grasses, such as Trapa, Persicaria thunbergii, Persicaria
pubescens, Potamogeton, Juncus, Schoenoplectiella, and Eleocharis,
and these species can still be seen in and around modern paddy
elds. In contrast, in the northwest area, where carbonized foxtail
millet grains were discovered, farmland weeds, such as Cyperus,
Fatoua villosa, and Commelina communis, were found. These results
indicate that rice was cultivated on the wetland in the southwest
area and that millet was probably cultivated upland in the north-
west area. For some reason as yet unknown, foxtail millet was
carbonized by re in upland elds and then deposited in the
wetland area together with farmland weeds.
 H. Nasu, A. Momohara / Quaternary International xxx (2015) 1e9 5

Table 2
Comparison of plant species between Kitashirakawa and Nabatake sites.

Final Jomon Initial Yayoi Early Yayoi

Kitashirakawa Nabatake Nabatake

Wetland Paddy eld Paddy eld

Cultivated plants 3 Species 3 Species 3 Species

Oryza sativa
Setaria italica
Perilla
Cucumis melo var. makuwa
Paddy eld and wetland weeds 6 Species 5 Species 17 Species
Juncus
Schoenoplectiella
Eleocharis
Potamogeton
Persicaria pubescens
Trapa japonica
Monochoria vaginalis
Persicaria hydropiper
Alismataceae
Ranunculus sceleratus
Oenanthe javanica
Murdannia keisak
Juncus prismatocarpus subsp. leschenaultii
Blyxa echinosperma
Najas gracillima
Eriocaulon miquelianum
Monochoria korsakowii
Limnophila sessiliora
Rotala indica
Cyperus difformis
Weeds common to both uplands and paddy eld 11 Species 10 Species 17 Species
Carex
Persicaria thunbergii
Echinochloa
Cyperus
Mosla
Commelina communis
Poaceae
Persicaria japonica
Apiaceae
Cirsium
Lapsanastrum humile
Haloragis micrantha
Hydrocotyle
Cyperus brevifolius var. leiolepis
Cardamine scutata
Eclipta thermalis
Cyperus iria
Rumex
Stellaria uliginosa
Cyperus compressus
Veronica peregrina
Fimbristylis littoralis
Persicaria lapathifolia
Hypericum erectum
Rorippa indica
Forest herbs 5 Species
Chrysosplenium japonicum
Chrysosplenium
Phytolacca japonica
Pilea
Pollia japonica
Upland eld and ruderal weeds 10 Species 10 Species 26 Species
Stellaria
Persicaria
Fatoua villosa
Urticaceae
Solanum nigrum
Humulus scandens
Viola
Potentilla hebiichigo
Mollugo stricta
Fabaceae
Cyperus microiria
Oxalis corniculata
(continued on next page)

Please cite this article in press as: Nasu, H., Momohara, A., The beginnings of rice and millet agriculture in prehistoric Japan, Quaternary
International (2015), http://dx.doi.org/10.1016/j.quaint.2015.06.043
6 H. Nasu, A. Momohara / Quaternary International xxx (2015) 1e9

Table 2 (continued )

Final Jomon Initial Yayoi Early Yayoi

Kitashirakawa Nabatake Nabatake

Wetland Paddy eld Paddy eld

Amaranthaceae
Portulaca oleracea
Plantago asiatica
Boehmeria nivea var. concolor
Fallopia japonica
Boehmeria japonica
Boehmeria spicata
Ajuga decumbens
Setaria viridis
Caucalis platycarpos
Sigesbeckia pubescens
Digitaria ciliaris
Chamaecrista nomame
Hemisteptia lyrata
Potentilla anemonifolia
Trees, shrubs and vines 19 Species 11 Species 10 Species
Morus
Actinidia
Rubus
Juglans mandshurica var. sachalinensis
Quercus subsp. Cyclobalanopsis
Quercus subsp. Quercus
Aesculus turbinata
Styrax obassia
Broussonetia papyrifera
Aralia elata
Ampelopsis glandulosa var. heterophylla
Sambucus racemosa subsp. sieboldiana
Pinus subgenus Pinus
Torreya nucifera
Castanea crenata
Cornus controversa
Acer
Zanthoxylum ailanthoides
Wisteria oribunda
Ericaceae
Eurya
Rubus hirsutus
Styrax japonica
Zanthoxylum schinifolium
Broussonetia
Rubus parvifolius
Cinnamomum camphora
Lycium chinense
Rubus buergeri

1e9 grains (), 10e49 grains (), 50e99 grains (), over 100 grains ().

On the other hand, seed, wood, and pollen analysis in the east
area revealed the presence of a swamp forest before rice cultiva-
tion began. The taxa of the swamp forest in the east area exhibited
differences between the lower and upper layers. In the lower
layer, Japanese chestnut (Juglans mandshurica var. sachalinensis),
horse chestnut (Aesculus turbinate), and cornel (Cornus con-
troversa) dominated the tree taxa associated with forest herbs,
such as Chrysosplenium, Pollia japonica, and Phytolacca japonica,
which usually grow in the dark swamp forest. In the upper layer,
which corresponds to the stratum in the west area in which rice
was discovered, the secondary forest taxa of deciduous oak
(Quercus sect. Quercus), Styrax obassia, Wisteria oribunda, and
Zanthoxylum ailanthoides dominated, together with ruderal herbs,
such as Persicaria longiseta, P. thunbergii, Mosla, Carex, and
Humulus scandens. These results indicate that the dark swamp
forest became an open wetland at the time of rice cultivation.
From the data collected through wood remains analysis, fallen
Quercus trees with cut marks were also found. Pollen data show
the transition of the surrounding vegetation, from an evergreen
forest to a secondary deciduous Quercus forest at the time (Tomii
et al., 2012), and probably these changes were caused by human
activity.
2.3. Comparison of the weed species composition of the initial
paddy eld at Nabatake site
What is the difference between the wetland of the
Kitashirakawa-Oiwakecho site and paddy elds? To answer this
question, we compared the weed species composition of the
Kitashirakawa-Oiwakecho site with that of the Nabatake site,
where the Initial and Early Yayoi paddy elds were found, and for
which a detailed weed seed analysis was carried out by Kasahara
(1982).
Nabatake site is located at the end of a small valley at the base of
the hill which faces the west part of the Karatsu plainin Saga pre-
fecture, of northern Kyushu Island. From the site, cultural layers
belonging to Early Jomon, Final Jomon (Initial Yayoi), Early Yayoi
and Middle Yayoi were found during its excavation in the 1980's
(Kasahara, 1982).

Please cite this article in press as: Nasu, H., Momohara, A., The beginnings of rice and millet agriculture in prehistoric Japan, Quaternary
International (2015), http://dx.doi.org/10.1016/j.quaint.2015.06.043
 H. Nasu, A. Momohara / Quaternary International xxx (2015) 1e9
The paddy elds of Nabatake site are found from two time pe-
riods. One belongs to the Initial Yayoi period (Yamanodera type),
and the other to the Early Yayoi period (Itazuke I type). Weed
species compositions from the paddy elds of the two periods are
compared with the weed composition of Kitashirakawa-Oiwakecho
site in Table 2.
First, the list of common species in the paddy elds of Naba-
take site are as follows: Schoenoplectiella, Eleocharis, and Persicaria
pubescens in the paddy eld and wetland weed category; Carex,
Persicaria thunbergii, Echinochloa, Mosla, Cyperus, and Commelina
communis in the weeds common to both uplands and paddy
elds category; Stellaria, Persicaria, Fatoua villosa, Urticaceae,
Solanum, Humulus scandens, Viola, and Potentilla hebiichigo in the
upland eld and ruderal weeds category; and Morus, Actinidia,
Rubus, and Broussonetia papyrifera in the arboreal category. Thus,
the common species in the Nabatake paddy elds were relatively
large in number. On the other hand, a major difference between
the Nabatake paddy elds and Kitashirakawa-Oiwakecho wetland
is the absence of typical paddy weeds, such as Alismataceae and
Monochoria at the latter site. Since these species are emergent
plants, this difference indicates that there was no clear compart-
ment made by paddy ridges to keep water in the Kitashirakawa-
Oiwakecho wetland. Another feature is that a large number of
woody plants and forest herbs are still present. This indicates that
there was a rich forest near the Kitashirakawa-Oiwakecho
wetland.
Next, we compared the number of each species in terms of
percentages by ecological category (Fig. 2). In Kitashirakawa-
Oiwakecho site, the number of weed species is relatively small,
at about 50% of all the plant life, whereas woody plants and forest
herbs constitute up to 40%. In contrast, in the Early Yayoi paddy
eld of Nabatake site, the percentage of woody plants and forest
herbs decreased to 10% and 0%, respectively, while the percentage
of weeds increased to 80%. In the life-form category (Fig. 3), the
percentage of annual herb species is 35% at Kitashirakawa-
Oiwakecho site, and increased to 50% in the Early Yayoi paddy
eld of Nabatake site. These results indicate the degree of
disturbance caused by human activity in the paddy eld.
Thus, a detailed weed seed analysis is a powerful tool for the
reconstruction of the state of initial rice cultivation, even in cases
where remains of clearly dened paddy ridge partitions are unable
to be found. In the example of Kitashirakawa-Oiwakecho site, the
view of rice being cultivated in the primitive paddy elds was
probably a pleasant sight, as pointed out by Tasaki (2002).
Fig. 2. Comparison of species percentage by ecological category between Kitashir-
akawa and Nabatake sites.
Please cite this article in press as: Nasu, H., Momohara, A., The beginnings of rice and millet agriculture in prehistoric Japan, Quaternary
International (2015), http://dx.doi.org/10.1016/j.quaint.2015.06.043
7
Fig. 3. Comparison of species percentage by life-form category between Kitashirakawa
and Nabatake sites.
3. Conclusion
As we have seen above, although some Japanese researchers still
believe that the beginnings of rice and millet agriculture go back as
far as the Middle to Late Jomon period, the most reliable timing of
its dispersal is the end of the Final Jomon period (corresponding to
the Initial and Early Yayoi period of northern Kyushu). This result
supports the views of English-written literature that neither rice
nor millet was an important part of Jomon subsistence (Crawford,
1992, 2008, 2011; Imamura, 1996; Habu, 2004; Matsui and
Kanehara, 2006; Bleed and Matsui, 2010).
It is assumed that rice and millet cultivation were simulta-
neously introduced as a set, part of the agricultural complex into
the existing Jomon subsistence economy; however, it seems not to
have spread linearly or rapidly. Although the Jomon people had
already managed and cultivated several varieties of crops (soybean,
azuki bean, perilla herbs, hemp, and bottle gourd) and trees
(chestnut, lacquer tree), the grace of the forest, such as berries and
acorns, was still a major food source during the Final Jomon period.
At the time, rice was probably cultivated in primitive paddy elds,
such as in man-made open wetlands with no clear paddy ridge or
compartment, and located near a forest environment. The use of
berries and nuts in the Jomon period continued, and perhaps rice
and millet agriculture gradually spread through the Middle to Late
Yayoi period. Several types of groups may have coexisted alongside
each other, perhaps with one group which did not accept rice and
millet agriculture until late, a different group that quickly adopted
cultivation of paddy rice, and yet another group being introduced to
millet farming in the mountain regions.
Acknowledgments
We thank Makoto Tomii, Naoko Sasaki, Yumiko Murakami,
Hikaru Takahara and Yuya Tsujimoto for their help in our study of
Kitashirakawa-Oiwakecho site. We are grateful to Michihiko
Nakazawa, Seiji Nakayama, Hiroki Obata, Aya Manabe and Yuichiro
Kudo for presenting us with the recent rice and millet data, chro-
nological data and critical comments. This research was supported
in part by Grant-in-Aid for Scientic Research (KAKENHI) (grant
number 25850009, 25284154).
References
Aida, S., Nakazawa, M., Nasu, H., Sasaki, Y., Yamada, T., Koshiishi, H., 2012.
Archaeobotanical investigations of charred plant remains and seed impressions
8 H. Nasu, A. Momohara / Quaternary International xxx (2015) 1e9
on pottery using the replication method from the Mekiri site, Okaya, Nagano
prefecture, central Japan. Natural Resource Environment and Humans 2, 49e64
(in Japanese with English abstract).
Bleed, P., Matsui, A., 2010. Why didn't agriculture develop in Japan? A consideration
of Jomon ecological style, niche construction, and the origins of domestication.
Journal of Archaeological method and Theory 17, 356e370.
Bronk Ramsey, C., 2009. Bayesian analysis of radiocarbon dates. Radiocarbon 51,
337e360.
Chiba, Y., 2012. Jomon World in the Kyoto Basin: Kitashirakawa Archaeological
Remains. Shinsensha, Tokyo (in Japanese).
Cohen, D.J., 2011. The beginnings of agriculture in China: a multiregional view.
Current Anthropology 52, S273eS293.
Crawford, G.W., 1983. Paleoethnobotany of the Kameda Peninsula Jomon. Anthro-
pology Papers, no. 73. Museum of Anthropology, University of Michigan, Ann
Arbor.
Crawford, G.W., 1992. The transitions to agriculture in Japan. In: Gebauer, A.B.,
Price, T.D. (Eds.), Transitions to Agriculture in Prehistory, Monographs in World
Archaeology, No. 4. Prehistory, Madison, WI, pp. 117e132.
Crawford, G.W., 2008. The Jomon in early agriculture discourse: issues arising from
Matsui, Kanehara and Pearson. World Archaeology 40, 445e465.
Crawford, G.W., 2011. Advances in understanding early agriculture in Japan. Current
Anthropology 52, S331eS345.
Crawford, G.W., Lee, G.A., 2003. Agricultural origins in Korea. Antiquity 77, 87e95.
D'Andrea, A.C., 1992. Palaeoethnobotany of Later Jomon and Yayoi Cultures of
Northeastern Aomori and Southwestern Hokkaido. PhD dissertation. University
of Toronto.
D'Andrea, A.C., Crawford, G.W., Yoshizaki, M., Kudo, T., 1995. Late Jomon cultigens in
northeastern Japan. Antiquity 69, 146e152.
Endo, E., 2012. Millet remains on the pottery cofn of the Final Jomon. Journal of
Archaeological Society of Nagano Prefecture (Nagano-ken Koukogakkai-shi)
140, 43e59 (in Japanese).
Endo, E., Takase, K., 2011. Foxtail millet (Setaria italica (L. P. Beauv.)) and broomcorn
millet (Panicum miliaceum L.) of the nal Jomon period in the Ina Basin, central
Japan. Quarterly of Archaeological Studies (Koukogaku Kenkyu) 58, 74e85 (in
Japanese with English abstract).
Endo, E., Takase, K., 2012. A study of the Yayoi period using replication method: a
case study of the Nishishiga site, Aich prefecture, central Japan. Bulletin of Meiji
University Museum 17, 13e25 (in Japanese with English abstract).
Fujiwara, H., 1995. Results of phytolith analysis of pottery soil from the Minami-
Mizote site. In: Okayama Prefectural Ancient Kibi cultural Properties Center
(Ed.), Minami-Mizote site 1. Okayama Prefecture board of Education, Okayama,
pp. 457e459 (in Japanese).
Fuller, D.Q., 2007. Contrasting patterns in crop domestication and domestication
rates: recent archaeobotanical insights from the old world. Annals of Botany
100, 903e924.
Habu, J., 2004. Ancient Jomon of Japan. Cambridge University Press, New York.
Imamura, K., 1996. Prehistoric Japan: New Perspectives of Insular East Asia. Uni-
versity of Hawaii Press, Honolulu.
Kasahara, Y., 1976a. Analytical method and examples of buried seed of the
archaeological remains. Historical Research (Kodaigaku Kenkyu) 81, 29e33 (in
Japanese).
Kasahara, Y., 1976b. Weed species of slash and burn elds and the excavated plant
seeds of Jomon. Journal of Weed Science and technology (Zassou Kenkyu)
21(additional volume), 41e43 (in Japanese).
Kasahara, Y., 1982. A study on the analysis and identication of buried seeds of
Nabatake site: reconstruction of ancient agriculture and vegetation. In: Karatsu
City board of Education (Ed.), Nabatake Site. Karatsu City board of Education,
Karatsu, pp. 354e379 (in Japanese).
Kimata, M., Negishi, M., 2002. Geographical distribution of proso millet, Panicum
miliaceum L. on lodostarch and phenol reactions: with special reference to
northern propagation route into Japanese islands. Environmental Education
Studies 12, 15e22.
Kitagawa, J., Yasuda, Y., 2008. Development and distribution of Castanea and Aes-
culus culture during the Jomon period in Japan. Quaternary International 184,
41e55.
Lee, G.A., 2011. The transition from foraging to farming in prehistoric Korea. Current
Anthropology 52, S307eS329.
Matsui, A., Kanehara, M., 2006. The question of prehistoric plant husbandry during
the Jomon period in Japan. World Archaeology 38, 259e273.
Matsutani, A., 2006. Identication of carbonized grain on the pottery from Ryu-
gasaki A site using SEM observation. In: Shiga prefecture board of Education
(Ed.), Ryugasaki A Site. Shiga Prefecture board of Education, Ohtsu, pp. 173e178
(in Japanese).
Miyaji, S., 2013. A discussion on Jomon rice: future direction of the rice impression
studies. Annals of Archaeology of Ritsumeikan University (Ritsumeikan-daigaku
Koukogaku Ronshyu) VI, 61e68 (in Japanese).
Miyata, Y., Kojima, K., Matsutani, A., Tobe, S., Nishimoto, T., 2007. Oldest common
millet from western Japan: carbonized material adhering to pottery from the
Ryugasaki A site, Shiga prefecture. Bulletin of the National Museum of Japanese
History 137, 255e265 (in Japanese).
Nakayama, S., 2010. Botanical Archaeology and Origin of Agriculture in Japan.
Douseisha Publishing, Tokyo (in Japanese).
Nakayama, S., Sano, T., 2012. Grain impressions of foxtail millet and broomcorn
millet in the terminal Final Jomon: a case study of Yashikidaira site, Yamanashi
Please cite this article in press as: Nasu, H., Momohara, A., The beginnings of rice and millet agriculture in prehistoric Japan, Quaternary
International (2015), http://dx.doi.org/10.1016/j.quaint.2015.06.043
prefecture. Journal of Archaeological Society of Yamanashi Prefecture (Yama-
nashi-ken Koukogaku Kyoukai-shi) 21, 85e97 (in Japanese).
Nakayama, S., Uruma, T., 2012. Grain impressions of foxtail millet and broomcorn
millet in the terminal Final Jomon: a case study of Nakamichi site, Yamanashi
prefecture. Bulletin of Yamanashi Prefectural Museum (Yamanashi-ken-ritsu
Hakubutsukan Kenkyu Kiyou) 6, 1e26 (in Japanese).
Nakazawa, M., 2009. A discussion on Jomon agriculture hypothesis: verication of
domesticated plant seeds. In: Shitara, H., Fujio, S., Matsugi, T. (Eds.), Archeology
of Yayoi 5: Production and Acquisition of Food. Douseisha Publishing, Tokyo,
pp. 228e246 (in Japanese).
Nakazawa, M., 2011. Toward the evaluation of foxtail millet and common millet
seed impressions on the Late and Final Jomon pottery from Koujinsawa site,
Nagano prefecture. Tonegawa 33, 16e36 (in Japanese).
Nakazawa, M., 2013. Results and the evaluation of replication method from Kinki,
Tokai, and central highland regions in Japan. In: Nakazawa, M. (Ed.), Pro-
ceedings of a Symposium Entitled What Replication Method Did Revealed?:
Practice for the Study of Spread and Acceptance of Agriculture in the Japanese
Archipelago. Meiji University, Tokyo, pp. 2e9 (in Japanese).
Nakazawa, M., Sasaki, Y., 2011. Seed impressions of common millet from the
Fusenmon type pottery of Final Jomon and pottery from Middle Yayoi period.
Natural Resource Environment and Humans 1, 113e117 (in Japanese).
Nasu, H., 2004. Archaeology of seeds and fruits. In: Yasuda, Y. (Ed.), Handbook of
Environmental Archaeology. Asakura Publishing, Tokyo, pp. 237e246 (in
Japanese).
Nasu, H., 2014. The initial form of rice and millet cultivation during the Final Jomon-
Yayoi transition era from the view of archaeobotanical weed assemblages.
Bulletin of the National Museum of Japanese History 187, 95e110 (in Japanese
with English abstract).
Nasu, H., Momohara, A., Yasuda, Y., He, J., 2007. The occurrence and identication of
Setaria italica (L.) P. Beauv. (foxtail millet) grains from the Chengtoushan site
(ca. 5800 cal B.P.) in central China, with reference to the domestication center in
Asia. Vegetation History and Archaeobotany 16, 481e494.
Nasu, H., Gu, H.-B., Momohara, A., Yasuda, Y., 2012. Land-use change for rice and
foxtail millet cultivation in the Chengtoushan site, central China, reconstructed
from weed seed assemblages. Archaeological and Anthropological Sciences 4,
1e14.
Nasu, H., Aida, S., Sasaki, Y., Nakazawa, M., Yamada, T., Koshiishi, H., 2015. Exam-
ining the utilization of pulses in the Middle Jomon period of the Suwa region
from carbonized seed remains. Natural Resource Environment and Humans 5,
37e52 (in Japanese with English abstract).
Nishimoto, T. (Ed.), 2009. Origin of Yayoi Agriculture and East Asia: Construction of
high resolution radiocarbon chronology. Report on the Grant-in-Aid for Crea-
tive Scientic Research (KAKENHI) No. 16GS0118 (in Japanese).
Noshiro, S., Sasaki, Y., 2014. Pre-agricultural management of plant resources during
the Jomon period in Japanda sophisticated subsistence system on plant re-
sources. Journal of Archaeological Science 42, 93e106.
Noto, T., 2002. Japanese farmland study by archaeology. In: Goto, T., Mogi, M. (Eds.),
East Asia and Japanese Archaeology IV Subsistence. Douseisha Publishing,
Tokyo, pp. 119e150 (in Japanese).
Obata, H., 2011. Palaeoethnobotany in Northeast Asia and Jomon Agriculture.
Douseisha Publishing, Tokyo (in Japanese).
Obata, H., Manabe, A., 2011. Seed impressions on the pottery by replication method
from Suitenmukai site and its signicance. In: Satsuma town, Kagoshima Pre-
fecture, board of Education (Ed.), Suitenmukai Site. Satsuma Town, Kagoshima
Prefecture, Board of Education, Satsuma, pp. 126e129 (in Japanese).
Reimer, P.J., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey, C.,
Buck, C.E., Cheng, H., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P.,
Haidason, H., Hajdas, I., Hatte, C., Heaton, T.J., Hoffmann, D.L., Hogg, A.G.,
Hughen, K.A., Kaiser, K.F., Kromer, B., Manning, S.W., Niu, M., Reimer, R.W.,
Richards, D.A., Scott, E.M., Southon, J.R., Staff, R.A., Turney, C.S.M.,
van der Plicht, J., 2013. IntCal13 and Marine13 radiocarbon age calibration
curves 0e50,000 years cal BP. Radiocarbon 55, 1869e1887.
Sakamoto, S., 1988. The Road of Millets: from the Ethnobotanical Study in Eurasia.
Japan Broadcast Publishing, Tokyo (in Japanese).
Sasaki, K., 2003. Japanese Culture from the South. Japan Broadcast Publishing, Tokyo
(in Japanese).
Sasaki, K., 2011. Before Rice Agriculture, Newly-revised ed. Yousensha. (in Japanese).
Sasaki, Y., Nakazawa, M., Nasu, H., Yoneda, K., Koizumi, R., 2009. Identication of
Setaria impressions on potsherds of the end of Final Jomon to the early Yayoi
periods recovered from the Naka-yashiki site, Kanagawa and Ishigo site,
Nagano. In: Proceedings of 24th Annual Meeting of Japanese Association of
Historical Botany. Kumamoto University, pp. 48e49 (in Japanese).
Sato, Y., 2002. Japanese History of Rice. Kadokawa (in Japanese).
Shoda, S., 2010. Radiocarbon and archaeology in Japan and Korea: what has
changed because of the Yayoi dating controversy? Radiocarbon 52, 421e427.
Sugiura, M., Momohara, A., Shima, M., Murakami, M., Miyake, N., Akazaki, H.,
Matsuda, K., 2011. Plant macrofossil assemblages below the Aira-Tn Tephra
dominated by herbaceous plants in Shika, Miyakonojo, Miyazaki, Japan. In:
Proceedings of 26th Annual Meeting of Japanese Association of Historical
Botany. Hirosaki University, p. 62 (in Japanese).
Tasaki, H., 2002. Paddy eld rice cultivation in Japanese archipelago: reviews from
paddy eld remains from the 1st millennium BC. In: Goto, T., Mogi, M. (Eds.),
East Asia and Japanese Archaeology IV Subsistence. Douseisha Publishing,
Tokyo, pp. 73e117 (in Japanese).
 H. Nasu, A. Momohara / Quaternary International xxx (2015) 1e9 9

Takase, K., Endo, E., Nasu, H., 2011. Plant use on remote islands in the Final Jomon
and Yayoi periods: an examination of seeds restored from potsherds in the
Tawara site, Niijima Island, Japan. Bulletin of Meiji University Museum 16, 1e18.
Tomii, M., Ito, T., Kanai, I., Sasaki, N., Sugiyama, J., Takahara, H., Tsujimoto, Y.,
Nasu, H., Fujii, H., Murakami, Y., Mechtild, M., 2012. Excavation on the area
BH31, northern campus of Kyoto University. In: Center for Cultural Heritage
Studies, Kyoto University (Ed.), The Annual Report of the Archaeological
Research on the Campus of Kyoto University 2009. Center for Cultural Heritage
Studies, Kyoto University, pp. 111e285 (in Japanese).
Urushibata, M. (Ed.), 2006. Sakamoto A and Sakamoto B Sites. Miyakonojo city,
Miyazaki Prefecture board of Education, Miyakonojo (in Japanese).
Willcox, G., 2012. Searching for the origins of arable weeds in the Near East.
Vegetation History and Archaeobotany 21, 163e167.
Yamaguchi, H., 1997. Origin and diversication of Japanese weeds. In: Yamaguchi, H.
(Ed.), Natural History of Weeds. Hokkaido University Press, pp. 3e16 (in
Japanese).
Yoshizaki, M., 1992. Detection of Ancient millets. Journal of Archaeology (Kouko-
gaku Journal) 355, 2e14 (in Japanese).
Zhao, Z., 2011. New Archaeobotanic data for the study of the origins of agriculture in
China. Current Anthropology 52, S295eS306.

Please cite this article in press as: Nasu, H., Momohara, A., The beginnings of rice and millet agriculture in prehistoric Japan, Quaternary
International (2015), http://dx.doi.org/10.1016/j.quaint.2015.06.043