Rice cultivation is believed to have originated with the Neolithic Shangshan (上山) Culture in the Early Holocene of the Lower Yangtze Valley Basin, between about 10 000 and 8200 years before the present. The Neolithic Rice-farming tradition of this region persisted for several millennia, culminating in the Liangzhu (良渚) Culture of about 5300-4400 years ago. Archaeologists have tried to study this process in a variety of ways, although most investigations have concentrated on changes to Rice Plants over this time, or else changes in cultivation practices and their impact on the landscape.
In a paper published in the journal PLoS one on 7 December 2022, Jiajing Wang of the Department of Anthropology at Dartmouth College, Jiangping Zhu of the Pujiang Museum, Dongrong Lei of the Longyou Museum, and Leping Jiang of the Zhejiang Provincial Institute of Cultural Relics and Archaeology, and the Department of History of Science and Scientific Archaeology at the University of Science and Technology of China, present the results of a study of early Rice cultivation in the Lower Yangtze Valley that concentrated on the stone tools used by these early farmers to harvest Rice.
Harvesting techniques in themselves apply selective pressure to a cultivated crop. Most Cereal crops have lost the ability to shatter their own seed cases, effectively making them dependent on Human intervention. This development in early cereal crops has been widely linked to the use of sickles as a harvesting tool, as this selects for seeds with a tougher rachis (section of the stem to which the seed is attached), something only found in a small proportion of wild Cereal Plants. This fits well with the development of Cereal crops in the Fertile Crescent, where the onset of Cereal cropping is marked by the appearance of hundreds of flint sickles. However, these tools are absent at the Neolithic sites of the Lower Yangtze where Rice cultivation originated. This has raised the possibility that the initial cultivation of Rice was marked by a different form of harvesting, such as see picking or plucking and beating. Evidence to support such theories, however, is lacking.
Excavations of Shangshen Culture (10 000-8200 years before present) and Kuahuqiao (跨湖桥) Culture (8000-7000 years before present) culture sites in the Lower Yangtze River Basin have produced hundreds of crudely made flaked stone tools in the core-and-flake tradition widely found in South China during this interval. These have been recorded variously as scrapers, burins, or drills, based upon their shape, but little thought has been applied to their purpose. Wang et al. observe that many of these blades have sharp edges which would be suitable for harvesting crops, and, in order to test this hypothesis, analysed use-wear marks upon 52 of these blades, as well as examining them for phytolith residues.
Phytoliths are produced by plants as a way of handling silica absorbed with water from the soil. Most plants produce phytoliths to some extent, but Monocotyledons, such as Grasses and Palms, which utilise phytoliths both as structural support and a defence against herbivores (silica phytoliths quickly wear down the teeth of animals which lack specific adaptations to deal with them), produce phytoliths which can often be used to identify the maker to genus or even species level. Rice Plants produce three highly distinctive types of phytolith, double-peak husk cells, Oryza-type bulliform leaf cells, and scooped parallel bilobates, which can make tools which have been used to work Rice easy to identify.
Cereal harvesting produces both distinctive wear traces upon a tool, and leaves a residue of plant tissues. Examination of stone tools that have been used to cut Poaceaen Plants (Grasses, Reeds, etc.) reveal that this activity produces a particularly distinct pattern of striations and polished areas, and that this remains constant across a variety of stone types that would have been available to Neolithic farmers in East Asia.
The Shangshan Culture are thought to have been the first people to have cultivated Rice, and to have built permanent settlements in the Lower Yangtze River Valley. To date, Archaeologists have uncovered 19 settlements associated with the Shangshan Culture. The Kuahuqiao Culture dates from slightly later in the same area, and used many of the same technologies, but also engaged in more intense landscape modification with specialised Earth-working tools.
Wang et al. obtained stone tools from two sites in Zhejiang Province; Shangshan, the type location of the Shangshan Culture, and Hehuashan, which has both Shangshan and Kuahuqiao layers. The Shangshan settlement covered an area of about 30 000 m², and includes dwellings, storage and midden pits, and possible burials. It has produced a diverse stone tool assemblage, including flaked stone tools (82% of the material), grinding stones (15%) and polished stone tools (3%). Hehuashan is located about 80 km to the southwest of Shangshan, on a small hill, and is divided into two areas, the East and West. East Hehuashan has an (excavated) area of about 600 m², and includes houses, pits and Shangshan-style stone tools and ceramics, while West Hehuashan has an excavated area of about 150 m², including houses, pits, ditches, and Kuahuqiao-style stone tools and ceramics.
At both sites, chipped flake tools form a significant part of the total lithic assemblage (82% at Shangshan and 57% at Hehuashan). At both sites these tools seem to have been made when needed, and discarded after use, with attempts at retouching being rare. None of these blades resemble the sickle blades developed in the Neolithic of the Fertile Crescent, or the denticulate sickles of the Huai River Valley of China (developed by the Peiligang (裴李崗) Culture of Henan Province in central China between 9000 and 7000 years before present, and thought to have been used to harvest Grasses and Rushes). These tools are made from locally available rock, including vitric tuff, river pebble, porphyry, and sandstone, with the blades made by the two cultures essentially identical in form.
The tools analysed in the tests came from the collections of the Pujiang and Longyou museums. Twenty tools were selected from the early phase of the Shangshan Culture (10 000 to 9000 years before present), eighteen from the late phase of the Shangshan Culture (9000 to 8200 years before present), and fourteen from the Kuahuqiao Culture (8000 to 7000 years before present). These averaged 42 mm in length and 42.8 mm in width, suggesting that they were hand held tools.
Use wear analysis of these tools suggested that they had been used for five different tasks; harvesting Siliceous Plants (i.e, Grasses or Reeds) (30 tools), cutting Animal tissue (7 tools), processing hard materials (10 tools), scraping woody material (6 tools), and an unidentified function (13 tools).
Tools used for cutting Siliceous Plants generally have an uneven finish, with fine and uneven striations, and and extensive patches of polishing, on both raised and scored areas. Of the 30 tools in this category, 22 have striations perpendicular or diagonal to the cutting edge of the tool, suggested as being indicative of cutting or scraping, while 14 have striations parallel to the cutting edge, probably caused by slicing, and six have both.
Tools used to cut the soft tissues of Animals tend to develop patches of highly polished surface, but lack the striations produced by cutting Plant matter or hard material.
Stone tools used to scrape woody material develop a rough polish on sinuous domed areas, with occasional deep, tapering striations perpendicular to the cutting edge.
Scraping hard materials such as Animal bones produces a level topography and rough polish on a stone tool, with deep, even striations. The polished areas are concentrated on the highest parts of the tool, with reticulated areas between.
Interestingly, while the tools had all been identified as simple, single edged, single function tools, analysis of the striations identified secondary cutting edges on thirteen of them. These typically had a principle cutting edge with a low-angled concave surface, and a more steeply convex secondary cutting edge. In these tools the concave surface showed signs of having used to process Grasses, while the convex surface appeared to have been used to cut Animal tissues.
Fifty of the fifty two tools analysed yielded phytoliths, the vast majority of which came from types of Grass. These included silate or sinuate elongates from Grass leaves, stems, and inflorescences. More specific material included ouble-peak husk cells, Oryza-type bulliform leaf cells, and scooped parallel bilobates, all diagnostic of Rice, as well as Reed-type bulliforms, Sedge achenes, and articulated quadrilobates, which are associated with Panicoid Grasses (the group which includes Sugarcaine, Maize, Sorghum, and Switchgrass). This suggests strongly that a high proportion of the tools were used in the harvesting of Grassy crops, including Rice.
The composition of the phytolith assemblage does, however, change significantly between the Shangshan and Kuahuqiao tools. In the Shangshan material, double-peak Rice husk phytoliths are found on 71% of the tool assemblage, while in the Kuahuqiao material this falls to 7%. The reverse trend for Rice leaf/stem phytoliths, which were recovered from 34% of the Shangshan tools, but 57% of the Kuahuqiao tools. The phytoliths of Reeds and Sedges also increased, with both being recovered from 8% of the Shangshan tools, while Reed phytoliths were found on 43% of the Kuahuqiao tools and Segde phytoliths on 29%. These changes probably reflect increased landscape management and a change in harvesting technique.
Wang et al. found a significant correlation between the abundance of phytoliths and wear patterns on blades, with blades showing wear patterns associated with Grass processing yielding a significantly higher proportion of Grass phytoliths. Furthermore, of the 30 blades interpreted as having been used in harvesting activities, 28 yielded Rice phytoliths. Comparing the phytolith and wear-use data suggests that the blades were used in two ways, as finger knives used to reap the panicles (seed clusters) at the top of the Plant, and as sickles used to cut the stems at their base.
The best example of a blade used to cut Rice using the finger knife technique is specimen SS-10, a flaked blade made from a river pebble, with a slightly convex working edge 58 mm in length. This tool has numerous fine striations perpendicular to the blade edge, interpreted as indicative of cutting or scraping Grass. The blade edge of this tool yielded a large number of phytoliths of the Rice glume double-peak type, suggesting that it has been in direct contact with Rice seeds. In combination, this is interpreted as evidence of being used to cut the panicles at the top of the Rice Plant during harvesting.
In contrast, specimen HHS 38 is interpreted as having been used in sickle harvesting. The blade of this tool has a convex surface about 92 mm in length, with many fine striations parallel to its cutting edge, thought to be indicative of slicing or sawing through plant matter. This tool yielded no phytoliths associated with Rice husks, but numerous phytoliths with morphotypes associated with Grass leaves and stems, including Oryza-type bulliforms and scooped parallel bilobates from Rice leaves. This combination of sawing-type striations and leaf and stem phytoliths is considered to be indicative of cutting Rice stems close to the ground with a sickle-action.
Wang et al.'s findings strongly support the hypothesis that these flaked stone tools were used in the harvesting of Rice. Furthermore, they demonstrate a shift in the way in which Rice was harvested which had not previously been suspected. In the Shangshan Culture Rice appears to have predominantly been harvested using a finger-knife technique to remove seeds from the top of the Plant, while the use of sickles to cut the stems was rare, while in the Kuahuqiao Culture the use of sickles appears to have become the predominant method of harvesting Rice.
Although the materials from which tools are made have changed, both finger-knife and sickle harvesting of Rice have persisted into modern times. The finger-knife technique allows the farmer to repeatedly revisit the same Plant, which seems unduly arduous, but can be useful under conditions where the seeds of the Plant do not all ripen at the same time. This technique reduces the number of panickles (seed husks) that shatter during harvesting, thereby maximising the amount of Rice recovered from each Plant. Sickle harvesting enables the farmer to quickly cut through bunches of stems, gathering the entire harvest rapidly, with the seeds subsequently separated from the stems and leaves, which can be used for fuel or Animal feed. The choice of tool can be driven by spiritual beliefs as well as practical considerations, with many Southeast Asian cultures having resisted the introduction of the sickle for a long time, as sickles were perceived as harming the soul of the Rice, and therefore being less pleasing to the Rice Goddess.
Little is known (or is likely to be known) about the spiritual beliefs of the earliest Rice farmers of the Lower Yangtze Basin, but it is possible to make a comparison between the change in Rice harvesting technique and the process of domestication. The wild progenitor of modern East Asian Rice, Oryza rufipogen, grows in swampy wetlands, where its seeds ripen unevenly, and are shed into the (often deep muddy water. Trying to harvest this Plant with sickles has been shown to be very difficult experimentally, while a better yield can be achieved with finger knives. This would presumably also have been the case in the Early Holocene, when the process of Rice domestication began. Samples of Rice spikelets from Huxi, a late Shangshan Culture site, dated to between 9000 and 8400 years before the present, included only 8.7% non-shattering husks, which are predominant in modern Domestic Rice. This presumably indicates that the Rice fields of the Shangshan people were dominated by a type of Rice with unevern ripening and self-shattering husks, better harvested with the more selective finger knife technique.
The succeeding Kuahuqiao people had a much more developed farming culture, originating about 8000 years ago, in which Rice was cultivated in managed Grass-Reed swamps, where water flow was controlled by irrigation systems and periodic burning was used. This system is thought to have been used to cultivate multiple wetland and forest Plants together, and to have promoted more synchronous Rice growth, making the harvesting of bundles of stems with a sickle a more viable option. At the Kuahuqiao type site (which gives its name to the culture), one of the key finds was a bunch of Rice stalks cut squarely at the base, something most likely to have been achieved with a sickle technique.
This adoption of sickle harvesting appears to have been roughly synchronous across several cultures in Ancient China about 8000 years ago. At the Peiligang Culture Jiahu Site in Henan Province, denticulate sickles begin to appear in large numbers during Phase III, which has been dated to between 8000 and 7500 years before the present. Wear use analyses carried out on these blades has suggested that they were used to harvest Grasses, and phytolith analysis of deposits from the site's residential areas suggests that Rice stalks were being brought into the settlement.
While finger knife harvesting requires more time within the Rice field, the product it produces is relatively 'clean' and requires little post-harvest processing. Sickle harvesting on the other hand, is likely to introduce more 'weeds' into the yield, which need to be removed later. The Kuahuqiao blades yield significantly more Sedge phytoliths than the Shangshan blades, which is probably evidence of a common weed being harvested alongside the Rice. There is also an increase in the number of Reed phytoliths at this time, but this may reflect the use of the new sickle technology to harvest an Reed stems as an additional crop. Items made from Reed stems have not, to date been discovered at any Kuahuqiao sites, but the use of Reeds in the Chinese neolithic is well documented at later Hemudu (河姆渡) Culture sites such as Tianluoshan, dated to about 7000 years ago, and the Hemudu type-site, dated to about 6000 years ago.
Both finger knife and sickle harvesting would produce selective pressure for non-shattering husks in Rice. With finger-knife harvesting, the Plants would be revisited numerous times, with the last crop obtained each year presumably contributing most to the seed crop saved for the following year. Since the final crop will include the highest proportion of non-shattering genotypes (which ripen slower), this would lead to a higher proportion these genotypes in each crop. With the introduction of sickle-harvesting, seeds with self-shattering husks would be more likely to be lost between the harvesting site and the processing site, effectively removing themselves from the domestic Rice genepool, applying another selective pressure in favour of non-shattering husks.
Given these pressures, it is somewhat surprising that the domestication process for Rice took as long as it did (about 5000 years). This probably relates to the knowledge level of the farmers themselves, who would not have considered they were involved in a multi-generation domestication process, but rather have been concentrating on each years crop. Such farmers would not have intentionally selected non-shattering husked Rice as a seed crop, and may not always have selected their seed crop from the last harvest of the year. Furthermore, they are highly unlikely to have cleared a virgin plot to farm each year, so that self-seeded Rice, with self-shattering husks, would have been present in each year's crop.
For the Shangshan and Kuahuqiao cultures, Rice was probably only a small proportion of the diet, supplementing wild-gathered foods such as Acorns. This process of domestication proceeded through the following Hemudu Culture, with Rice probably not becoming fully domestic until the onset of the terminal-Neolithic Liangzhu Culture, considered to have been the first culture in southern China to form centralised states, about 5300 years ago.
See also...
