Origins of Agriculture: an Ecological Perspective on Crop Domestication

Lead Research Organisation: University of Sheffield
Department Name: Animal and Plant Sciences

Abstract

The proposed research aims to develop a new ecological model for crop domestication, integrating the roles of environmental change, plant traits, and human agency, under the constraints of the archaeological record. It addresses the idea that natural selection and human agency played a critical role at different stages in the emergence of agriculture, focusing on the interactions between plants, humans and environment during the period preceding fully agricultural societies. We propose a research programme with parallel experimental and archaeobotanical work packages that will consolidate the evidence necessary to develop and refine this new model. Our ecological model is formulated within the archaeologically documented framework of a gradual, widespread origin of agriculture, and is based on the proposition that different elements of the 'domestication syndrome' arose independently during different stages in the transition from gathering to farming. In particular it distinguishes the two archaeologically visible domestication traits: larger grain size and seed indehiscence, and sees these as consequences of different selective pressures operating at different stages in the domestication process. Our focus is on the former, and we argue that specialisation on a limited range of large-seeded species, and selection for larger seed size within these species, were both driven by an interaction between human diet choice and ecological processes. Our central thesis is that seed size correlates with a suite of functional traits which, through ecological processes, favour some species as crops over others and, through evolutionary processes, select for large-seeded genotypes of these crop species. We advance the model through the discussion of four hypothetical phases along the path towards greater sedentism and agriculture. Previous research has demonstrated that at least two of these phases can be recognized in the archaeobotanical record via the presence of non-food species representing either the wild communities from which seeds were gathered, or weed assemblages. Empirical archaeobotanical evidence, in the form of the plant spectrum, grain size and ecological conditions in different time periods will be used to assess the feasibility of the proposed model for explaining observed changes. The construction of an archaeobotanical database of pre-agricultural and early agricultural sites, and quantitative analysis of these records, will establish (a) the range of species likely to have been deliberately collected as food plants and provide the ecological context of gathering, (b) the extent, geographical locations, and date of the narrowing of the plant spectrum as the crop progenitor species came to prominence, and (c) the appearance of potential weed communities. Grain size measurements of both wild (crop progenitor and other grain species) and domesticated cereal and pulse crops will be analysed to establish when (in the case of crops) and whether (in the case of wild grains) seed size increased, and its timing in relation to changes in the associated non-food plant species assemblages, in particular in relation to changes in CO2 levels or climate during the period before potential weed assemblages are first recognized. At the same time, ecological experiments will assess the validity of the mechanisms of change proposed in the model by testing the underlying hypotheses concerning the differential responses of wild crop progenitors and other wild species (and of large and small-seeded genotypes of crop progenitors) to changing ecological conditions. Ecological experiments will determine the relationship of plant species and genotypes to increasing CO2 levels comparable to those occurring at the end of the last ice age, and to human-generated microenvironments with greater levels of disturbance and higher fertility.
 
Description The emergence of domesticated crops was a crucial stage in the Neolithic development of agriculture. However, prior to the earliest evidence of plant cultivation 10,000 years ago, there is substantial evidence of wild plant exploitation, including both crop progenitors and small-grained species that were never domesticated. Domestication traits evolved in the crop species over millennia, through a hypothesised process of 'unconscious selection' exerted by cultivation and harvesting practices. These traits included greater seed mass, and the loss of dormancy and natural dispersal mechanisms. In this project, we use experiments to compare the growth and harvest characteristics of wild grain species, crop progenitors and domesticated crops, which archaeobotanical evidence suggests were exploited in the Fertile Crescent during the Neolithic transition to agriculture. Our central hypothesis is that seed mass is positively correlated with a suite of functional traits including plant size, seed yield and growth rate, which favour some species as crops over others. Findings from this work will have important implications for our understanding of how agriculture began. Our results have been surprising. Firstly, crop progenitors are no more productive than other wild species when grown as isolated plants. However, grain is packaged into larger ears, and the plants are more productive when grown as stands. Secondly, domestication has led to important increases in yield mediated via the size of plants, allocation to grain, growth rate and duration. Finally, we have some evidence that competition was involved in the domestication process, such that crop progenitors may have outcompeted other species in anthropogenic environments during the transition to agriculture.
Exploitation Route The main beneficiaries of knowledge arising from this research are anticipated to be members of the archaeological community who are engaged in research on the transition to farming. The availability of a detailed, centralised database of all primary archaeobotanical data records for the Epi-palaeolithic and early Neolithic periods, combined with the most up- to-date radiocarbon dates for the material, will provide a valuable public resource for this community that will last well beyond the duration of the project.

The "big science question" tackled by this project, and its multi-disciplinary approach, also mean that the work is likely to attract interest from a number of other scientific communities. First, the global change community will be interested in the knowledge created about the responses of wild food plants to sub-ambient CO2 concentrations. This information will enhance current understanding of how global change has impacted on human societies in the past, and how atmospheric change may drive ecological and evolutionary processes. Secondly, the work will be of interest to the ecological community, since the question of which plant traits are associated with seed size is a classic question in plant ecology. Finally, the general science community is likely to be interested in the new perspective we propose to bring to the question of how crops were domesticated. Our current crisis in food security makes the fundamental plant traits involved in selecting our crop species from the pool of available plant species of high topical interest.
Sectors Agriculture, Food and Drink,Environment