ANR MICA The full project

Ce carnet de recherche a vocation à présenter au public les recherches et les activités scientifiques du projet de recherche MICA « Modéliser les Interactions entre changement Climatique et Agriculture dans l’occident ancien », ainsi qu’à servir de liaison entre les membres et partenaires du projet.

MICA est un Projet de Recherche Collaborative (PRC) financé par l’Agence Nationale de la Recherche (MICA Project – ANR-22-CE27-0026) pour une durée de 4 ans (2023-2027).

Son objectif est d’évaluer l’impact des variations climatiques sur les rendements des cultures (blé, orge, millet, vigne, olivier) et leur influence, en interaction avec les facteurs sociaux, sur les transformations des économies agraires et de l’agrobiodiversité en Europe et Méditerranée occidentales, de l’Age du Bronze à la fin de la période romaine (2000 BCE – 600 CE).

MICA réunit 6 partenaires institutionnels et une équipe interdisciplinaire de plus de 40 membres intégrant les champs disciplinaires de l’archéobotanique, la paléogénomique, l’archéologie spatiale, l’agronomie, la paléoclimatologie et la modélisation.

Proposal’s context, positionning and objective(s)

Objectives and research hypothesis

Since the Neolithic beginnings of agriculture, in one of the warmest periods of the Holocene, Europe experienced numerous climate variations (Magny 2004) with potentially large impact on its food security, as crops largely depend on water supply and temperature conditions for their growth. Periods of warming are often perceived as favorable to agricultural productivity while cooling periods are associated with a greater frequency of crop failure (Büntgen et al. 2011). Understanding the impacts of climatic variations on past crop productivity and on the transformation of agriculture needs to take in account the complexity of the relationship between historical, socio-economic and environmental factors. The underlying objective of MICA is also to decipher the different and complex pathways of adaptation of human societies to climate change.
MICA focuses on a multi-millennial period, from the Bronze Age to the end of Roman times (2000 BCE – c. 700 CE), during which agriculture and crop sets underwent major transformations, most notably with the spread of millets and later the onset of arboriculture (Bouby 2014, Pérez-Jordà et al. 2021), while important climatic fluctuations are recorded (Büntgen et al. 2011). Concomitantly, economic systems, demographic and environmental pressures changed drastically and entered a new scale.


The main objectives of MICA are:

1) to model spatio-temporal variations in the potential yields of the major cash and subsistence crops using a process-based dynamic vegetation model and paleoclimatic reconstructions based on both pollen data and data-model assimilations. We have singled out three annual crops, bread wheat, hulled barley, common millet, and two perennial plants, grapevine and olive,

2) to reconstruct ancient agrosystems and agrobiodiversities using archaeological and archaeobotanical data. Morphometric and paleogenomic approaches will be employed to characterize the intraspecific diversity of grapevine and olive,

3) to assess how societies have responded to climatic variations, in terms of adapting their agrarian strategies together with other socio-economic factors.

Our research hypotheses are that:

1) Specialized commercial farming systems that developed during the Iron Age and Roman period took advantage of favorable climatic conditions. But when, where and why did these forms of agriculture arise? What is the part of climatic and socio-economic factors in these changes?

2) Crop diversification is a way to cope with harsher conditions. Did the spread of millets and the rise of hulled barley play a key role in this process in the Bronze and Early Iron Ages? Does a new agricultural diversification coincide with phases of climate cooling, such as that occurring during Late Antiquity?

3) Adaptations in crop and varietal diversity can favor the sustainability of farming systems. Have farmers growing olives and grapevine been able to secure their crops or extend them to different territories using new or better adapted varieties?

In order to assess the major agricultural changes in response to diverse environmental and socio-cultural situations, MICA will focus on four areas of interest (AOI): (A) Paris basin, (B) Mediterranean France (Languedoc, Provence), (C) Southeastern Spain (Andalusia, Valencia) and (D) Northern Morocco.
The project proposes to come up to these issues by an innovative and broadly interdisciplinary approach bringing together cutting-edge methods, in order to tackle the complexity of societies and to avoid simplistic deterministic scenarios. This implies a variety of analytical and modelling methods and interconnected geographical scales, with a transfer of concepts and methods across disciplines.
The choice of the grapevine and the olive tree as models for the study of intraspecific diversity is motivated by 1) their importance for speculative agricultures that flourished during the Iron Age and Antiquity, 2) pre-existence of innovative methods to reach intraspecific accuracy, 3) impact of temperature and water supply variations on grape and olive productivity and fruit quality, which can be countered by changes in varieties (De Ollas et al. 2019).

Position of the project as it relates to the state of the art

State of the art

A better understanding of the impact of climate changes on past agriculture requires a detailed assessment of climate variations, changes in agrobiodiversity and farming systems (crop systems, distribution of farms and agrarian installations) through time and space. Extensive data on paleoclimates, archaeobotany or spatial archaeology are now available in certain regions.

Numerous works offers today very precise past climatic reconstructions. Among the multiple proxies, pollen records offer the advantage of being extensive, thus providing access to a consistent network across time and space. Pollen data combined with vegetation modelling have allowed to reconstruct variations in precipitations and temperatures over the Mediterranean and Europe in the last 10 000 years (Guiot & Kaniewski 2015). In parallel, the strong development of archaeobotanical studies has provided major insights in the identification of cultivated plants and crop changes since the beginnings of agriculture. The implementation of databases makes it possible to monitor detailed variations at regional level, in relation to social and geographical patterns. Depending on the geographical area, archaeobotanical research is more or less advanced. Hundreds of sites have already been studied in the Paris Basin, Mediterranean France and Southern Spain. Studies supported by proper databases or large datasets have already been published in these regions (Bouby et al. 2017, Zech-Matterne et al. 2017, Pérez-Jordà et al. 2021).

The number of sites investigated is much less important in Morocco, but the rapid development of archaeobotanical research is now under way (Ruas 2018). Crop weeds provide evidence of changes in agricultural practices and in the characteristics of cultivated soils, some of which may be driven by climate variations (e.g. soil moisture).

Such changes are occasionally analyzed in relation to climatic indicators obtained directly from isotope analyses of crop remains (Aguilera et al. 2012, Riehl 2012). Recent methodological advances now provide keys to characterize intraspecific crop diversity. Geometric morphometric methods (GMM) allow for a larger-scale and much more detailed description of modern and past agrobiodiversities than traditional morphometrics, based on modern and archaeobotanical seeds (e.g. Terral et al. 2004). At the same time, the advent of high-throughput sequencing methods (NGS) has truly launched paleogenetic research in archaeobotany by being capable of studying ancient DNA in a much more systematic and powerful way than previous techniques (Wales et al. 2014). On the other hand, methods of spatial archaeology have developed widely in recent decades (Rodier et al. 2011), together with systematic prospecting, preventive archaeology and databases.

These methods, based on spatial analyses under GIS make it possible to analyze the choices of location of human occupations according to landscape and environmental conditions. Very recent work applied to the restitution of vineyards in the Roman Empire allows to glimpse more generally the potential of these methods to evaluate the extent of specialized crop areas, such as those identified from bioarchaeological data in eastern Spain and Gaul (Stubert et al. 2020, Bernigaud et al. 2021). Beyond these spatial restitutions, the evolution of societies in interaction with their environment can now be studied using Agent Based Modelling (ABM). ABMs make it possible to test multiple alternative hypotheses and scenarios in order to assess the share of historical, socio-economic and environmental factors in the fluctuating trajectory of ancient societies. ABMs have been applied to European archaeology for less than a decade around issues related to demography and exploitation of resources (e.g. Bertoncello et al. 2018, Joyce 2019).

A number of large-scale studies based on archaeobotanical and/or archaeological databases suggested that changes in agricultural strategies, demographic trajectories or land use patterns should be connected with climatic changes (e.g. Bevan et al. 2017). In order to go beyond the stage of making comparisons that give rise to hypotheses, it is now necessary, and possible, to integrate the results produced by different approaches in a more systematic way, and test different hypotheses and scenarios based on modelling. In addition, the intraspecific diversity of cultivated plants has never been taken into account when considering the adaptation of agriculture to climate change.

A project following on the results of two previous projects:

Two recent projects have advanced the use of modelling and intraspecific diversity study methods, providing significant amount of data and cutting-edge technologies that pave the way for the MICA project:

RDMed project (2018-2021)

RDMed, coordinated by J. Guiot, funded by the “Interdisciplinarity 2016” call of AMIDEX foundation, brought together historians, archaeologists, paleoclimatologists, paleoenvironment specialists, agronomists and economists from diverse laboratories and universities. By combining climate reconstructions over the past millennia with archaeological, historical and environmental data and an original modelling approach, the project assessed the impact of climate variability on societies and their environments. The aim was to understand the interactions between climate, environmental factors, crop yields, human settlements and the conditions for trade over several millennia. The project focused on olive, grapevine and wheat in Southern France from the Iron Age to Late Antiquity (6th century BCE-7th century CE). RDMed helped to adapt agroecosystemic modelling (Fader et al. 2015) to simulate the impact of past climate change on potential yields of different crop species (Contreras et al. 2019, Bernigaud et al. 2021). RDMed led to the creation of a functional agricultural societies simulator, which will be further developed in MICA (Bernigaud et al. submitted). A method of data assimilation has also been designed based on paleoclimate modelling inter-comparison project (PMIP) (Guiot et al. submitted).

VINICULTURE ANR project (2017-2021)

VINICULTURE was an ANR PRC program, supervised by L. Bouby, that brought together several of the partners (ISEM, AASPE, ASM, CAGT) and members of the new MICA project. Its aim was to identify the characteristics and diversity of grapevines and wines in France, from the end of prehistory to Modern times. It developed an integrative archaeological approach based on archaeobotany, geometric morphometrics, archaeogenomics, biochemistry and experimental archaeology. Several of its results give the key to the research we seek to undertake in the framework of MICA. The study of seed morphology of hundreds of modern varieties using GMM provided the means to reconstruct domestication traits of grapevines (berry size, wine/table use, origin) (Bonhomme et al. 2020, 2021). In parallel, approximately 20,000 archaeological seeds were processed in VINICULTURE. At the same time, the study of ancient DNA has made it possible to identify for the first time close kinship relations between modern grape varieties and archaeological pips from Iron Age, Roman and Medieval sites (Ramos-Madrigal et al. 2019).

Innovative nature and originality of the objectives and methods of MICA

The interlocking use of several types of modelling from different branches of research (ABM, spatial analysis, agro-ecosystem modelling) is particularly original. Agro-ecosystem modelling is an innovative method for agriculture studies and ABM is a key-technique for crossing human sciences and environmental studies. Although spatial analysis methods have developed in archaeology, they are less so in archaeobotany. Moreover, far from considering agricultural productivity as a single variable, MICA proposes a new strategy, modelling a range of complementary cultivated plants using archaeobotany and up-to-date developments to integrate intraspecific diversity. Finally, while archaeological and archaeobotanical syntheses are most often carried out in a regional context, the geographical scale considered in MICA gives it an ambitious international dimension that changes the scope of interpretations.

The MICA project aims to assemble an interdisciplinary community around the study of socio-ecosystems at different time scales (also known as the study of complex systems). It involves not just dividing the work and then sharing the results, but, and above all, emphasizes that, the specialists from various disciplines should define together the object of study (selection of archaeological sites and periods of climate change, refinement of the key scientific questions, definition of the model parameters related to agriculture, human communities, trade, demography …), and should work together during the whole duration of the project. The socio-ecosystems to be studied require integrative modelling of the endogenous interactions between the agents, interactions with external agents, and interactions between agents and their environments. An additional dimension, Time, is also an integrative factor.

The approach is transferable to other regions and other resources, opening the way to a more global vision of the adaptation and acclimatization of societies to environmental changes. Modelling across multiple spatial and temporal scales to consider the human consequences of climatic changes has been identified as a pressing need for studying human-environment interactions over the long-term (Kintigh et al. 2014), and such a long-term perspective is necessary for establishing an integrated modelling approach that can be used for future projections.

Methodology and risk management

MICA is scheduled to last 48 months. The complexity of the questions that carry this project requires the development of interdisciplinary research, involving mainly archaeobotanists, archaeologists, historians, agronomists, paleoclimatologists, genomicists and modellers, organized in five tasks (Fig. 1).

Our investigations will focus on four geographical areas of interest (AOI): Paris basin, Mediterranean France (Languedoc and Provence), Southeastern Spain (Andalusia and Valencia) and Northern Morocco. These AOI spread along a north-south gradient in western Europe-Mediterranean and document climatically and environmentally varied conditions. The AOI were also chosen because they represent contrasted situations concerning phases of major changes both in agriculture in general and in the development of farming economies decisively oriented towards speculation and trade. The Paris Basin was strongly impacted by Middle and Late Bronze Age agricultural innovations and the spread of new crops, including millets, which are, to a large extent, the result of Central European dynamics (Bouby et al. 2017). From the 2nd century BCE onwards, agriculture has specialized, with the increase of naked wheat in particular, in relation with the emergence of cities and the needs of their supply (Zech-Matterne et al. 2014).

In northern France this tendency was consolidated during the Roman period but with a zonation in the agricultural specialization probably determined by environmental factors. In the Mediterranean regions of France and Spain, the innovations of the Bronze Age occurred later and in a more mitigated way than in the Parisian Basin. On the other hand, the specialization of agriculture in relation to the development of trade occurred earlier, in association with the encounters with Mediterranean cultures (Greeks, Phoenicians). In Southern France, specialization on naked wheat started during the 5th c. BCE, somehow in line with the beginnings of viticulture (Bouby 2014). The integration in the Roman Empire led to the development of speculative viticulture and, to a lesser extent, olive growing, in the 1st – 2nd c. CE. In Southern Spain, arboriculture started by the beginnings of the 1st millennium BCE, in an area experiencing Phoenician influences, and gradually expanded northwards (Pérez-Jordà et al. 2021). Morocco had a highly specialized agriculture in the Roman period, but the proportions of olive growing and viticulture remain difficult to assess (Carrato et al. 2020). The organization and management of the project must allow for the incorporation of a wide variety of environmental and chrono-cultural situations, the collection of significant quantities for data group (Fig.2 & 3), and the combining and integration of these data using geo-statistical and modelling tools. Spatial information can be easily shared online. Linking the open data is a realistic task, that will offer the possibility to maintain, share and reproduce the work of the consortium.

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search