Early Rice Project Aims

Understanding the development, diversification and spread of rice agriculture is central not only to our understanding of the processes of human population growth, dispersal and formation of civilizations in Asia, it is also central to reconstructing how past agricultural activities might have impacted global climate through methane emissions and deforestation.

Current Aims & Hypotheses: The impact of intensification and deintensification of Asian rice production: transitions between wet and dry ecologies (NERC Grant NE/N010957/1)

1.  Our overall aim is to reconstruct rice cultivation across a range of sub regions, environments and periods in South China, mainland Southeast Asia, Northeast and Peninsular India, and to assess the extent to which rainfed and wet irrigated systems were practiced, and whether different regional patterns or a single evolutionary trajectory can be reconstructed.

Ø Hypothesis 1A. Early Chinese rice was wet and evolved into dry ecologies with its dispersal to the southern provinces (e.g. Neolithic Guangdong) and Southeast Asia, while wet ecologies evolved or were introduced secondarily in mainland Southeast Asia after 2000 BP.

Ø Hypothesis 1B. Early rice in Southwest China (Sichuan/ Yunnan) was predominantly wet since the Neolithic (with dry locales used for millet cultivation).

Ø Hypothesis 1C. Early rice (>2500 BP) in Northeast India (Assam/Meghalaya) was initially dry shifting cultivation, and wet systems introduced first along the Ganges/Brahmaputra after ca. 2500 BP to Bangladesh and Assam.

Ø Hypothesis 1D. Early rice in western India (Gujarat/Rajasthan) was irrigated, due to rainfall too low to support dry rice, in contrast to many parts of South India. This irrigated rice was associated with urbanisation after 2500 BP, and earlier reports of rice were intrusive or imported.

1. We will also determine the key regions and periods when transitions from wet to dry took place and to test our current hypotheses about the timing and cultural historical context of these transitions in Southern China and mainland Southeast Asia (Thailand).

Ø Hypothesis 2A. A key transition from dry to wet rice took place in the plains of Thailand 2000-1500 BP and subsequent wet rice systems dominated lowland Southeast Asia.

Ø Hypothesis 2B. Rainfed rice or semi-wet low intensity systems dominated Cambodia up until historical (Angkorian) period (ca. 1300 BP).

Ø Hypothesis 2C. Early urbanism in central Myanmar was associated with the arrival of wet (irrigated) rice, around 1750 BP, representing a major expansion of wet rice farming.

Ø Hypothesis 2D. In Guangdong the  transition to wet rice was ca. 2500-2000 BP

2. To further refine our existing analytical weed flora analogues, so that we can divide past rice systems into four grades along the spectrum from dry upland to fully flooded wet rice. We will do this through further analyses of additional modern analogue samples that have already been collected from the uplands of Thailand and Laos, and through some additional targeted analogue sampling in northeast India/Bangladesh (for high rainfall dry rice and low elevation flooded rice).

Ø Hypothesis 3A. Phytolith assemblages from upland shifting cultivation can be distinguished from lowland dry rice by higher phytolith diversity levels of dicots.

Ø Hypothesis 3B. Well-watered wet rice will have reduced dicots and sedges with a dominance of bulliforms, and other water-sensitive grass morphotypes.

Fixed vs Sensitive grass phytolith ratios from Weisskopf et al Antiquity 2015

Ø Hypothesis 3C. The relative levels of bulliforms, sedges and other water-sensitive grass morphotypes will need to be corrected for inter-regional climatic variation to allow direct comparison of sequences from China with those from Southeast Asia or India.

3. To use our improved understanding of how rice was cultivated in different times and periods to produce improved geospatial models of dispersal of dry and wet rice in time and space, incorporating, and testing for, hypotheses under 1 and 2 above.

4. To use our empirical database and model of the past distribution of wet rice ecologies to make improved estimates of methane emissions from rice over time, as an improved test of the early anthropogenic methane hypothesis.

Ø Hypothesis 5A. The greatest increase in rice-derived methane can be reconstructed for the First Millennium AD, when there was both a large increase in wet rice areas closer to the tropics (India, Sri Lanka, Southeast Asia) and the higher temperatures in these areas

Ø Hypothesis 5B. The expansion of wet rice will alone only be able to explain part, perhaps around half, of the anomalous later Holocene methane trend, highlighting the importance of other factors (cattle, forest burning, etc.).

5. To use our research results and methods as a basis for training and knowledge exchange with colleagues and students in major rice growing countries (specifically China, Thailand and India) where indigenous science-based archaeology can be encouraged. This is being pursue through planned workshops in Bangkok, Chengdu and Vadodara.

Previous Grant Aims: The impact of evolving of rice systems from China to Southeast Asia. (NERC Grant NE/K0034021)

Archaeobotanical evidence offers a powerful set of tools for not only documenting where and when rice was cultivated in the past, but how it was cultivated through the analysis of ecology of associated weed flora in macro-remains assemblages and phytolith assemblages. We have pioneered the study archaeological rice weed flora and the combination of archaeological plant macro-remains and phytoliths in our recent NERC-supported research in parts of India, Sri Lanka and China (NE/G005540/1). We propose to roll out this method over a wider geographical and cultural area, as well refining the approach through some additional modern analogues. Because current evidence already provides an empirical framework for the early development of rice cultivation systems in the Yangtze (especially the Lower Yangtze), between 5000 and 2000 BC, and a firm basis for the later intensification of rice agriculture in the plains of northern and eastern India (2000-500 BC), we propose to focus our work on the less known parts of Asia, especially mainland Southeast Asia and the southern parts of China, as well as further work in the eastern parts of India. These regions are central to hypotheses on the dispersal of rice cultivation, including models linking the spread of rice to major language families such as Austroasiatic and Austronesian,and yet a lack systematically-studied evidence for rice cultivation itself, or evidence as to whether early rice represented an extension of the alluvial wetland cultivation systems like those of the Neolithic Yangtze (early subspecies japonica, typical of many modern temperate japonica) or the development of upland rainfed systems (the latter typical of many modern tropical japonica rices), with a secondary later parallel evolution of irrigated wet rice systems amongst indica rices. It is also hypothesized that irrigated rice in mainland southeast Asia was a later introduced from Indian irrigated traditions different from the upland rice systems that had diffused from China in the Neolithic.

Our aims are therefore to:

  1. Reconstruct the earliest rice cultivation systems (Neolithic-Bronze Age) along the three hypothesized trajectories of rice diffusion southwards from the Yangtze basin towards southeast Asia, namely in Yunnan, Guangdong, and Fujian.
  2. To reconstruct rice cultivation across a range of subregions, environments and periods in mainland Southeast Asia (mainly in Thailand and Vietnam) to assess the extent to which rainfed and wet irrigated systems were practiced, and whether different regional patterns or a single evolutionary trajectory can be reconstructed.
  3. To test the validity of our existing analytical weed flora analogues, developing largely on our modern studies in India and to a lesser extent China, through additional analogue sampling in Southeastern Asia, including Thailand and Yunnan.
  4. On the basis of our reconstructed rice arable systems and weed flora assemblages to assess the likelihood of single or multiple pathways for the spread of rice into Southeast Asia, by analysing the geographical and chronological patterns of weed flora.
  5. To use our improved understanding of how rice was cultivated in different times and periods of southern China and Southeastern Asia to produce improved models of past wetland rice area and linked methane emissions over time, grounded in the empirical evidence for past rice cultivation.
  6. To use the archaeological history of rice and its varied forms of cultivation for a knowledge exchange and public engagement program on the importance of long-term history and science-based archaeology to understanding agriculture's role in human impact on the environment over the long-term (see pathways to impact plan).