New discovery of the oldest maize weevils in the world from Jomon potteries, Japan
The maize weevil (Sitophilus zeamais) and rice weevil (Sitophilus oryzae) are two of the most damaging insects for stored grains, and are characteristic species of ancient Japan. Both species and the granary weevil (Sitophilus granarius) are common elsewhere in the world, but the natural distribution of maize and rice weevils is restricted to the Old World. Japanese archaeological records contain a few maize weevil fossils after the Middle Yayoi period (ca. 2000 BP). However, since evidence of weevils was discovered as impressions in Jomon potsherds in 2004, many weevil impressions have been found. The oldest is from the Late Jomon (ca. 4000 to 3200 BP). These findings and other archaeological evidence suggest that the maize weevil invaded Japan from Korea, accompanying the spread of rice cultivation. However, in 2010 we discovered older weevil impressions dating to ca. 9000 BP. These specimens are the oldest harmful insects discovered from archaeological sites around the world. The new discovery is valuable for future entomological research because such specimens are absent from the fossil record. It is also archaeologically and culturally interesting because this provides evidence of harmful insects living in Jomon villages.
However, the new discovery raises the question of what these weevils infested: did cereal cultivation exist 9000 years ago? We have no persuasive answer, but hope one will be provided by future interdisciplinary collaborations among geneticists, entomologists, and archaeologists.
Before maize weevil impressions were discovered on Jomon potteries (in 2004), the oldest specimens in Japanese archeological records (fossils) came from a ditch deposit at the Ikegamisone site (ca. 2000 BP). The fossils of maize weevils were recovered from the Fujiwara Palace (ca. 8th century AD) and Kiyosu Castle (ca. 16th century AD). In 2004, maize weevil impressions were discovered (ca. 3500 BP) in Late Jomon pottery on Kyushu.
Weevil models were recovered from impressions in the clay using “the replication method” developed by European paleoethnobotanists and adopted by Japanese archaeological researchers in the late 1970s. This method injects silicone into a cavity, resulting in a model of the original form that created the cavity. Another recent discovery using this method revealed soybean and adzuki bean cultivation during the Jomon Period.
Subsequently, many maize weevil impressions were discovered at Late Jomon sites, especially on Kyushu. In 2007, two maize weevil specimens were discovered at a Final Jomon site in Yamanashi Prefecture. By 2008, 32 weevil impressions had been obtained from 13 sites (Fig.1, Table 1; 1-32) . Most impressions appear to be of maize weevils based on their size and anatomical characteristics. These new discoveries suggest that maize weevils might have coexisted with humans as early as the Late Jomon period. The earliest specimen was found inside a Kanegasaki 3-type deep bowl dating to ca. 4000 BP. Later impressions of maize weevils have been reported from the Late to Final Jomon periods. The number of specimens increases rapidly by the Nishibira pottery phase (ca. 3800 BP), with a peak during the last quarter of the Late Jomon(14). Subsequently, specimen rates remain stable. We discovered three new specimens at early Late Jomon sites in Kagoshima Prefecture in 2009. This finding dates the appearance of weevils 200 or 300 years earlier than in previous research (Fig. 1, Table 1; 35-37).
Old hypothesis: Maize weevils document the origins of rice cultivation in Japan
Maize weevils feed on stored grains and on fruits such as peach or apple in North America, and also inhabit forests or grasslands in southern Japan, where they feed on flowers in the spring. The adult weevils feed on 37 families and 96 species of plants, but the larvae feed on only 11 families and 31 species. To understand their biology, we fed maize weevils many different foods. The adults preferred large grains such as rice, barley, and wheat. They did not feed on adzuki bean, hemp, or unhusked rice. The weevils also infested acorns without seed coats and successfully reproduced. However, during the Jomon period, acorns were generally stored with seed coats to protect them from decay, so few or no maize weevils would have infested acorns. Although adult maize weevils feed on rice flour, they never oviposit there. Therefore, even if sufficient suitable food is available to stimulate oviposition, weevils must still be able to penetrate the seed coat to oviposit successfully.
The maize weevil should appear at roughly the same time as agriculture expanded in Japanese archeobotanical records, and indeed, archaeological artifacts suggest strong cultural relationships between the southern Korea and northwestern Kyushu increased at this time. Furthermore, rice phytoliths have been recovered from Late Jomon pottery in northwestern Kyushu. The presence of maize weevils therefore suggests the existence of rice or barley cultivation during the Late to Final Jomon periods, and that they invaded Japan from Korea accompanying rice cultivation.
New evidence: older maize weevil impressions
The new evidence contradicts the original hypothesis. In February 2010, we discovered the oldest impressions of maize weevils from early Jomon potsherds dated to ca. 9000 BP from the Sanbonmatsu Site in Kagoshima Prefecture. The site, which is on a terrace (50 m asl) on the eastern coast of Tanagashima Island, 40 km southeast of the Oshumi Peninsula, was excavated in 2007 by the Nishinoomote City Board of Education.
Researchers discovered cultural layers containing many artifacts, mainly from the Early Jomon period.
Potsherds were examined to find seed and other impressions in February 2010. We found two fragments that contained maize weevil impressions (SBM0011 and SBM0024; Fig. 2). Both came from a Yoshida-type deep bowl (one from the body, the other from the base). Both fragments are ornamented with shells, a popular ornamentation during the first half of the Early Jomon period in southern Kyushu. This cluster has 14C dates from ca. 9400 to ca. 8700 BP, suggesting that the Yoshida type dates to ca. 9000 BP. In April 2010, we recovered five additional potsherds with maize weevil impressions (SBM0060, SBM0061, SBM0062, SBM0067, and SBM0073; Fig. 2). These were also Yoshida types based on their ornamentation and morphological characteristics. One is a rim fragment, two are bottom fragments, and two are parts of the body.
Significance of maize weevils during the Early Jomon
We discovered seven maize weevil impressions. The high discovery ratio is similar to that from the Late Jomon sites on Kyushu, indicating that these weevils were already pests in the Early Jomon.
If the Late Jomon maize weevils were associated with the spread of rice or barley cultivation into Japan, what is the significance of the Early Jomon records? Were other cereals cultivated in Japan at that time?
Two hypotheses explain the diffusion of rice cultivation into Japan.
First, it may have diffused from the Shantong Peninsula through Korea ca. 4000 BP; alternatively, it may have diffused from southern China through the Ryukyu Islands ca. 6700 BP. The former is accepted by most archaeologists. Evidence from archaeological sites at those times (charred seeds and seed impressions) suggests the cultivation of cereals in the Poaceae (e.g., rice, barley), introduced to Kyushu from Korea.
The second hypothesis is not supported by archeological or archeobotanical evidence. The earliest rice cultivation on Kyushu occurred in the 10th century AD, and was it not introduced from southern China; instead, it was introduced at the time of human immigration into the region. The Early Jomon is nearly synchronous with the beginning of rice cultivation in the Lower Yangtze River Valley in China, the origin of Asian rice cultivation. The oldest evidence of barley and wheat cultivation in East Asia is younger than ca. 4000 BP. Thus, the archaeological evidence strongly suggests there were no cultivated cereals that maize weevils could infest in Japan at ca. 9000 BP. This suggests that the existence of maize weevils in the Early Jomon was not associated with rice cultivation.
Nevertheless, the high weevil density at that time suggests the weevils were related to the Jomon people and lived in Jomon houses, where they fed on stored foods. We do not know what kind of wild plant food they fed on or infested. One candidate is the seeds of bamboo Bamb spp.). These seeds can be stored for long enough periods to permit weevil development, and have sometimes been used as an emergency food source during famines. Indeed, we found some charred bamboo seeds at archaeological sites from the Late Jomon to the Ainu Culture period on Hokkaido. However, phytolith analysis suggests rapid decreases in bamboo in this region from 11,300 to 7,300 BP because of the expansion of evergreen forest. Thus, we cannot state with certainty what food maize weevils infested at that time.
The new discovery should encourage additional research on the ecology of maize weevils, and particularly, when they began infesting stored food. This will require additional maize weevil specimens from other time periods and regions. No fossils of maize weevils or their ancestors have been discovered, so the origin and history of this taxon are unclear. DNA analysis would provide important insights.
Authors: Hiroki OBATA1, Aya MANABE2, Naoko NAKAMURA3, Tomokazu ONISHI4, Yasuko SENBA5
Source: Nature.com (Permission pending)