Navigation – Plan du site

World distribution of land cover changes during Pre- and Protohistoric Times and estimation of induced carbon releases

Répartition mondiale des espaces défrichés à la Pré- et Protohistoire et estimation des rejets de carbone induits
Carsten Lemmen
p. 303-312

Résumés

Le rôle de l’Homme dans l’évolution des états de surface de la Terre au cours de la Pré- et Protohistoire doit être estimé afin de répondre à deux objectifs : 1) établir une base de données qui permette de comparer les différentes actions anthropiques sur l’environnement et 2) bien distinguer les causes « naturelles » des causes d’origine anthropique dans l’évolution des dynamiques environnementales. L’utilisation d’un logiciel intitulé « Global Land Use and technological Evolution Simulator (GLUES) » a permis de modéliser les changements démographiques, technologiques et économiques antérieurs à l’Âge du Bronze. La démarche a consisté à estimer le besoin en terres pour répondre à la croissance continue de mises en culture en se basant sur les données de la « History Database of the Global Environment » et sur la modélisation des densités de population et des types de subsistance estimées d’après GLUES. Les défrichements forestiers se sont traduits par une perte de carbone estimée grâce au modèle VECODE. Le besoin en terres dans les foyers d’expansion préhistoriques - développés principalement dans des secteurs forestiers - a conduit à des déforestations massives, représentant jusqu’à 11 % de l’ensemble des forêts potentielles. Sur la période 10 000-2000 av. J.-C., ce sont au total environ 29 Gt de carbone qui ont été perdues en raison de la substitution des champs cultivés aux forêts. GLUES a permis la visualisation d’axes d’expansion agricole à une échelle régionale. La plupart des défrichements observés avant l’Age du Bronze ont affecté un vaste territoire qui s’étend depuis l’Europe centrale jusqu’à la Chine, en passant par l’Inde. Les valeurs régionales de perte en carbone sont estimées à 5 Gt en Europe et en Méditerranée, à 6 Gt en Inde, à 18 Gt en Asie du sud et du sud-est et à 2,3 Gt pour l’Afrique subsaharienne.

Haut de page

Notes de la rédaction

Article soumis le 8 juin 2009, accepté le 20 octobre 2009

Texte intégral

The author is supported by the German national science foundation’s priority program Interdynamik (DFG SPP 1266). V. Brovkin made available the VECODE model. GLUES was developed together with K. Wirtz; I thank him and two anonymous reviewers for their helpful comments on an earlier version of this manuscript.

Introduction

1Since when have humans influenced the global carbon cycle? And what is the magnitude of anthropogenic perturbation in the carbon cycle compared to the natural variability? Commonly, the beginning of the anthropocene is placed at 1750 (Crutzen and Stoermer, 2000), a date which is used to separate the preindustrial from the industrial (anthropogenically perturbed) environment. W. Ruddiman (2003) argued for a much earlier start of human interference by carbon dioxide (CO2) and methane (CH4) emissions and placed the start of the anthropocene at 8000 years ago; at this time, the first widespread agricultural areas appeared in China, the Mediterranean, and central Europe (e.g., Childe, 1936). The advent of Neolithic life style was associated with huge cultural, technological and social changes. Foremost the transition from hunting-gathering subsistence to agro-pastoral life style had a great impact on our Earth system, or as M. Zeder (2008) stated, “domesticates and the agricultural economies based on them are associated with radical restructuring of human societies, worldwide alterations in biodiversity, and significant changes in the Earth’s land forms and its atmosphere”. Only after the termination of the last glacial, a forest extent was established which we can use as the baseline for anthropogenic perturbation during the Holocene (11,600 years ago to the present); this forest was by no means pristine but had already been shaped by intentional intermittent clearing and burning by Palaeolithic and Mesolithic people (e.g., Simmons, 1996). Neolithic agropastoral subsistence for the first time required long-term removal of forest to create space for settlements, crops, or animals; additionally, timber was harvested for construction and as fuel. For a model Neolithic European village of 30 people, S. Gregg (1988) estimated that an area of 6 km2 forested area was necessary for survival. In total, about 107 km2 of forest and woodland have been lost due to agriculture and pastoralism up to date (e.g., Williams, 2000). A quick calculation assuming a value for aboveground biomass of 100 t/ha - derived for deciduous broad-leaved temperate trees by P. Curtis et al. (2002) - and neglecting the possible loss of litter and soil carbon, estimates a global carbon loss on the order of 100 Gt. This reduction amounts to 15% of today’s carbon pool in live vegetation (650 Gt, Field and Raupach, 2004), although all these estimates have to be treated with care due to the high uncertainties of estimating the carbon content of the tropical forest areas (e.g., Houghton, 2005).

2The sequence of the transition to agriculture is best visible in the agricultural records of Europe and the Mediterranean. In forested Europe, Neolithic agriculture had been introduced from the Levant region (today’s Lebanon, Israel, Syria, Palestine, Jordan and Iraq) by an expansion through the Mediterranean basin, and by land advance through Anatolia and the Balkan; this succession is clearly visible in crop ancestor species and animal bones (e.g., Flannery, 1973; Zeder, 2008). The most conclusive evidence for tracing this transition can be found in South East and Central Europe, where the appearance of the Körös and Starçevo cultures between 6500-5500 BC and the linear pottery culture between 5650-4900 BC mark the onset of intensive horticulture and the introduction of cereal agriculture (Bogaard, 2004). To fill spatial and temporal gaps in the data obtained from excavations and to provide a framework for hypothesis testing and process understanding, model studies have been employed, ranging from simple reaction-diffusion models (e.g., Fort et al., 2004) to models with transport of a farming trait (e.g., Ackland et al., 2007), more complex socio-technological development in multiple traits (Wirtz and Lemmen, 2003), up to multi-agent simulations (e.g., Axtell et al., 2002). Agent based socio-technological models have, however, only been useful on a local scale due to the very high computational demand when simulating single individuals or households. This study investigates regional socio-technological development on a global scale and for the entire period of pre-bronze age transition from foraging to farming subsistence. It connects adaptations in the socio-cultural system to the bioclimatic background and interpolates in space and time regional demographics, economics, and environmental impact, exemplified here as deforestation. Calculates are proposed from the areal demand for crops land in previously forested areas total carbon emission. More importantly than obtaining an exact value of total carbon emission, this study shows how to attribute the emission by Prehistoric deforestation to specific location and time.

Models and data

Socio-technological model GLUES

3The numerical model GLUES (Global Land Use and Technological Evolution Simulator; Wirtz and Lemmen, 2003; https://sourceforge.net/​projects/​glues/​) simulates the transition from hunting-gathering to agropastoral societies. It has been used to investigate the role of biogeographic proposition and climatic susceptibility (ibid.), and to assess the role of migration waves in the spread of agriculture (Lemmen and Wirtz, 2009). The socio-technological model comprises population, growth, and few characteristic traits whose temporal evolution is governed by marginal benefits to population growth (adaptive dynamics, e.g., Wirtz and Eckhardt, 1996; Dieckmann and Law, 1996). Three dimensionless traits were used to describe i) the efficiency gains in production, storing, and use of the food base (=technology efficiency), ii) division of labour in food production between agropastoral and hunting-gathering activities (=farmer share), and iii) economic diversity realized in different domesticated crops and animals in each regional population. The value of all traits determines demographic change; in turn, optimization to optimal growth governs adaptive trait changes. This way, and in accordance with S. Shennan (2001), the population dynamics is directly coupled to cultural evolution. The socio-technological model is embedded in a spatially explicit geographic context. Each local society is tied to a region, characterized by geographic position, vegetation, and climate fluctuations. These are based on the IIASA climatology of R. Leemans and W.P. Cramer (1991), the climate variability data base of K. Wirtz et al. (2009), and dynamics vegetation calculations with the Vegetation Continuous Description model (VECODE; Brovkin et al., 1997, 2002), assuming Prehistoric carbon dioxide levels of 280 ppm. To find an optimal parameter set, all parameters were randomly varied within confined ranges. All variation sets were scored with the temporal and spatial proximity of simulated agricultural centres to the five global independent centres of agriculture (Smith, 1997): a parameter set was chosen to generate simulations resembling most closely the observed Prehistoric global agriculturalization pattern. Following G. Ackland et al. (2007), the model predictions should be interpreted as providing a “historical null hypothesis. Its predictions can be taken as requiring no special explanation, and its failures can be taken as evidence of rare events that had significant and long-lived consequences”.

4In GLUES, a socio-technological system, or regional culture, is defined by the four state variables population density P, technological efficiency T, farmer share (or quota) Q, and economic diversity N. The characteristic traits X {T, N, Q} are subject to gradient adaptive dynamics which optimizes a population’s fitness with the evolution equation (dX/dt)=σX(dr/dX), where r=(1/P)·(dP/dt) is the specific growth rate, and σX the variability in trait X. The growth rate r depends on the traits X, on the natural resources R, and trait-dependent food procurement S; it is composed of growth and loss terms:

(1)

5where μ, γ, ω and ρ are scaling factors. This growth rate formulation includes a term for instantaneous overexploitation (R-γ√TP) of resources based on Ehrlich’s IPAT formula, a loss of the work force to administration (1-ωT), and a decrease of density-dependent mortality with technology (-(ρ/T)P). Long-term degradation of natural resources was not considered. Subsistence S is derived from either hunting and gathering or from herding and farming. The relative amount of time and energy spent on either foraging or culturing is expressed by the farmer share Q, which defines the trade-off for the system:

(2)

6where τ describes temperature limitation. 685 simulation regions where defined based on ecological homogeneity with a clustering algorithm by K. Wirtz (pers. comm. 2004) based on net primary productivity derived from the IIASA database (Leemans and Cramer, 1991). In each simulation region, a socio-technological model is started with identical initial values. Interaction of regions occurs via trade and migration. The environmental context is set both by temperature limitation τ derived from the number of growing degree days above zero in the IIASA climatology (Leemans and Cramer, 1991) and resources R, derived from the net primary productivity, calculated with VECODE from the IIASA climatology. The simulations where performed with the standard parameters set in GLUES version 1.1.4pre, available on sourceforge.net1. Knowledge loss or climate fluctuations where not considered. Initial values are P0=0.03/km2, T0=1, Q0=0.04, N0=0.25, variability is σT=0.15, and σN=1. The parameter set for this model (equation 1) setup is γ=0.01 km2, ω=0.04, μ=0.0037/a, and ρ=0.00037/a. Conversion from forest to cropland in the VECODE model (Brovkin et al., 1997) was implemented by subtracting the required area for cropping Ac from the forest part of the land Af and recalculating the live carbon compartments (leafs and structure) assuming grass specific live carbon content Clive,g for the crops:

Clive=Clive,f(Af-Ac)+Clive,g(Ag+Ac)

(3)

7Dead carbon (soil and litter) pools were reduced by reducing the specific carbon content Cd,f by 42% (Guo and Gifford, 2002) in cropped areas, thus:

Cdead=Cdead,f(Af-0.42Ac)+Cdead,gAg

(4)

8The mathematical background and an in-depth description of the model are explained in K. Wirtz and C. Lemmen (2003).

Data and configuration for deforestation estimates

9GLUES is supplemented with scaling parameters from the History Database of the Global Environment (HYDE version 3.1; Klein Goldewijk et al., 2007) and additional information on forest share and carbon pools (soil and live carbon) from VECODE. The usage and interaction of the different models and databases in the calculation of deforestation are shown in fig. 1. In addition to the model set-up described by K. Wirtz and C. Lemmen (2003), which uses growing degree days above zero degree (GDD0) and net primary production (NPP) from VECODE, I also use predicted potential forest share as a base line from which deforestation can be calculated. The mean forest share in all areas where agriculture developed before 2000 BC, was 70%, which is close to the maximum attainable value. Cropland and population densities from the HYDE data base at 5’ resolution were regridded on the 685 GLUES regions, the ratio of cropland per capita was calculated and a representative value of 0.02 km2 per person for full-scale agricultural regions (farmer share>90%) was extracted; this value describes the intensively used area in contrast to S. Gregg (1988)’s value of 0.2 km2, which characterizes the total are in use (but not necessarily converted) by Neolithic farmers. The cropland-per-person quantity was used to scale from the product of farmer ratio and population density to simulated cropland share. No representative value for the pasture demand per capita could be derived from the HYDE database; areal demand for the forest-pasture conversion was thus neglected in this study. In the conversion from forest to cropland, the live carbon in trees is replaced with live carbon representative by grassland (VECODE only distinguishes between four types of land cover: broadleaved forest, needleleaved forest, grassland, and desert). To estimate the loss of soil carbon, we follow L.B. Guo and R.M. Gifford (2002)’s meta-analysis, who find that reported field site soil carbon loss in the conversion from forest to crops land was on average 42%. For comparison to studies where only land demand and not carbon emission was published, we use the P. Curtis et al. (2002) value of 100 t/ha carbon in aboveground vegetation derived from North American deciduous broad-leaved temperate tree to convert from cropland share. This is a reasonable assumption, when predominantly forested areas are used for crop land, and because it represents a mean value between 40 t/ha found in more arid subtropical woodland to up to 180 t/ha found in wet (sub)tropical forests. For comparison with the archaeological record, we used the comprehensive data set on the neolithization of Europe by C.S.M. Turney and H. Brown (2007), which contains the dating of 735 western Eurasian sites and their attribution to 82 distinct cultural groups and sub-groups.

Fig. 1 – Workflow for dynamic hindcasting of deforestation based on databases for cropland (historical database of the Environment, HYDE) and climate variables (IIASA), the dynamic global vegetation model VECODE, and simulated population density and agropastoral subsistence fraction from GLUES.
Fig. 1 – Cadre schématique général de la modélisation de la déforestation fondé sur l’utilisation combinée des bases de données des terres cultivées (base de données historiques de l’Environnement, HYDE) et des variables climatiques (IIASA), la modélisation des dynamiques végétales est réalisée grâce à VECODE, les densités de population estimées et les secteurs de subsistence agropastorales sont déterminés avec GLUES.

Fig. 1 – Workflow for dynamic hindcasting of deforestation based on databases for cropland (historical database of the Environment, HYDE) and climate variables (IIASA), the dynamic global vegetation model VECODE, and simulated population density and agropastoral subsistence fraction from GLUES.Fig. 1 – Cadre schématique général de la modélisation de la déforestation fondé sur l’utilisation combinée des bases de données des terres cultivées (base de données historiques de l’Environnement, HYDE) et des variables climatiques (IIASA), la modélisation des dynamiques végétales est réalisée grâce à VECODE, les densités de population estimées et les secteurs de subsistence agropastorales sont déterminés avec GLUES.

Results

10The global simulation of socio-technological traits reveals at 2000 BC a broad band of technologically advanced culture ranging from central and southeast Europe over the Middle East, through India, Indochina, and China to Japan (fig. 2, left panel). High technology levels are also found in the Maghreb and intermediate technologies in Western Africa and in Southeast Africa. Not in all of these regions high population densities are found (fig. 2, right panel). At 2000 BC, the Southeast European region comprising Italy, the Balkan, Anatolia and the Levant is heavily populated with population densities of 4-6 inhab/km2. In eastern India and Vietnam, population density is around 4 inhab/km2, and in China mostly 3 inhab/km2. Again heavily populated is the upper Huang He (Yellow River) region and Japan, where population density is approximately 6 inhab/km2.

Fig. 2 – Map of simulated technological efficiency (left panel) and population density (right panel) at 2000 BC.
Fig. 2 – Carte simulant la diffusion technologique (vignette de gauche) et la densité de population (vignette de droite) à 2 000 BC.

Fig. 2 – Map of simulated technological efficiency (left panel) and population density (right panel) at 2000 BC.Fig. 2 – Carte simulant la diffusion technologique (vignette de gauche) et la densité de population (vignette de droite) à 2 000 BC.

Technology ranges between 1 (the Mesolithic level) and 8 for near Bronze-age technology, population density attains values up to 6 inhab/km2, up from the initial uniform hunter-gatherer density of 0.1 inhab/km2.
Indices de technologie entre 1 (mésolithique) et 8 (Âge du Bronze), les densités de population atteignent des valeurs proches de 6 habitants au km², alors que celles des communautés de chasseurs/cueilleurs ont des valeurs de 0,1 hab/km².

11The temporal evolution of the transition from foraging to farming is shown in six time slices from 8000 BC to 2000 BC in fig. 3. At 8000 BC, no farming is visible anywhere on Earth, however, inspection of the data reveals that small-scale farming already begins in the Levant and in China. By 7000 BC, agriculture is fully established in these founder regions and also in central Japan. The very fast transition is consistent with estimates by, e.g., M. Zvelebil and P. Dolukhanov (1991) who note that only few hundred years are required for a subsistence change from hunting and gathering to cultivating and herding. By 6000 BC, agriculture has expanded from the founding centres: to the north into southeast Europe from the Levant, and to the south from the northern Chinese centre; a new agricultural centre appears upstream of the Indus River. 1000 years later, the Indian and Chinese crop areas join. At 4000 BC, parts of Africa (the Maghreb, western Africa and southeast Africa) have converted to full-scale farming. A broad spatially coherent band of farming reaching from Spain to Japan is present.

Fig. 3 – Time slices of the farmer share simulated with GLUES.
Fig.3 – Différentes phases d’activités fermières simulées avec GLUES.

Fig. 3 – Time slices of the farmer share simulated with GLUES.Fig.3 – Différentes phases d’activités fermières simulées avec GLUES.

Colour indicates the range from initially only foraging (light) with short mixed-strategy transitions to mainly agriculture and herding (dark).
Les couleurs indiquent les différentes possibilités s’échelonnant des activités fourragères exclusives (en bleu) à une agriculture/pastoralisme principalement (en clair), en passant par une courte période de transition (en sombre).

12The accuracy of the simulated expansion of farming is tested by comparing the timing of the transition in the model to radiocarbon dates from western Eurasian Neolithic sites compiled by C.S.M. Turney and H. Brown (2007). The overall agreement of the simulated farming occurrence with the site data is good; on a per-site basis, the mean absolute difference between model and data is around 500 years (fig. 4). For western Europe, the model simulates the emergence of agriculture between 7000 and 2500 BC, but the picture is regionally diverse. Neolithic lifestyle is simulated in northern Greece before 7000 BC, and in the entire region of the Balkan and Minor Asia before 6600 BC. By 5700 BC, the western and eastern shores of the Black Sea, Italy, and Southern Spain/Northern Morocco have acquired agriculture as their dominant life style. Central Europe (Austria, Hungary, Slovakia, Southern Poland) and northern Italy show neolithization by 5000 BC; the Iberian Peninsula exhibits full neolithization by 4300 BC; Neolithic life style reaches the Baltic Sea by 4000 BC as well as Belarus. The neolithization in southern and northern Germany occurs not before 2500 BC; the simulation fails to predict the neolithization of France and Great Britain by 2000 BC. Up to 11% of the regional land surface is subject to cultivation by 2000 BC (fig. 5, left panel). The pattern of fractional crop cover follows the spatial distribution of population density in all areas with fully developed agriculture. Maximum simulated crop fraction occurs on the Balkan, in the Levant, in southern and northern China, and in Japan. For comparison, fig. 5 (right panel) shows the crop fraction interpolated from the HYDE database (Klein Goldewijk et al., 2007) onto the model regions. On this map, only the Levant and northern Italy appear as regions with a large crop fraction of up to 10%. Most of the southern Europe, the Levant, and China exhibit a crop fraction on the order of 1%. No crops are apparent in India, the Maghreb, or western and southern Africa.

Fig. 4 – Time of reaching the Neolithic stage (threshold 50% farming ratio, between 8000-4000 BC). Simulation results (background shading) compared to the data set of Neolithic sites by Turney and Brown (2007, shown as circles).
Fig. 4 – Délai pour atteindre la phase néolithique (seuil de 50 % pour les activités fermières, période 8 000-4 000 BC) simulé à partir de GLUES (arrière plan), comparaison avec la répartition des sites néolithiques (points), d’après Turney and Brown (2007).

Fig. 4 – Time of reaching the Neolithic stage (threshold 50% farming ratio, between 8000-4000 BC). Simulation results (background shading) compared to the data set of Neolithic sites by Turney and Brown (2007, shown as circles).Fig. 4 – Délai pour atteindre la phase néolithique (seuil de 50 % pour les activités fermières, période 8 000-4 000 BC) simulé à partir de GLUES (arrière plan), comparaison avec la répartition des sites néolithiques (points), d’après Turney and Brown (2007).

Fig. 5 – Map of simulated crop fraction (left panel) and crop fraction interpolated to the simulation regions from the HYDE database (right panel).
Fig. 5 – Carte stimulant la part des secteurs cultivés (vignette de gauche) et des secteurs cultivés interpolés à partir de la base de données HYDE (vignette de droite).

Fig. 5 – Map of simulated crop fraction (left panel) and crop fraction interpolated to the simulation regions from the HYDE database (right panel). Fig. 5 – Carte stimulant la part des secteurs cultivés (vignette de gauche) et des secteurs cultivés interpolés à partir de la base de données HYDE (vignette de droite).

13Using the data from northern American temperate broadleaved forest as a prototype with ≈10·103 t C/km2 leaf and stem biomass, the simulated cropland share was converted to carbon loss from deforestation. From 12000 BC to 7000 BC, a total of 10 Gt C was lost to be replaced by crops; by 2000 BC, this number increases to 27 Gt C. Potential carbon density simulated with the DGVM is shown in fig. 6 (left panel). Large carbon stock is present in the temperate forests of Europe (250 t C/ha), in subtropical east Asia (300 t C/ha), and the tropics (<350 t C/ha), while the carbon stock in semi-arid areas is below 150 t C/ha: Most regions with early agriculture are forested, except the river-based areas along Indus, Euphrates, Tigris and Nile. The carbon release from the potential forest by land use conversion to crops is shown in fig. 6 (right panel). Carbon is released mainly in southern Europe, and south and east Asia. Largest regional losses occur in Japan, northern Greece, and south China, with up to 36 t/ha removed carbon. Globally, the dynamic vegetation model simulation gives a total carbon loss of 11 Gt by 7000 BC, and 29 Gt by 2000 BC. The heavily converted landscapes of China, southeast Asia, south Asia and western Eurasia contribute around 90% of this carbon loss; the later agricultural areas of subsaharan Africa and the Americas contribute the remaining 10% (tab. 1). The attribution can be further broken down to the country size level. For example, the carbon emission due to land use conversion in the western Eurasian agricultural centre (corresponding in the model to the region of today’s Israel and Lebanon) is 160 Mt.

Fig. 6 – Map of potential carbon density (left panel) and deforested carbon density by replacement with crops, including partial soil loss (right panel).Fig. 6 – Cartes de densité potentielle de carbone (vignette de gauche) et de densité de carbone causée par la déforestation et de remplacement par les cultures incluant en partie des pertes de sols (vignette de droite).

Fig. 6 – Map of potential carbon density (left panel) and deforested carbon density by replacement with crops, including partial soil loss (right panel).Fig. 6 – Cartes de densité potentielle de carbone (vignette de gauche) et de densité de carbone causée par la déforestation et de remplacement par les cultures incluant en partie des pertes de sols (vignette de droite).

The boxes define selected subcontinental-scale regions, for which the total carbon emission is listed in tab. 1.
Les cartons définissent les régions choisies pour la présente étude et pour lesquelles les émissions de carbone sont listées sur tab. 1.

Tab. 1 – Total and regional carbon emissions due to land use. See fig. 6 for a definition of the subcontinental-scale regions.
Tab. 1 – Emissions totales et régionales de carbone liées à la mise en valeur agricole des terres. Voir fig. 6 pour la des zones régionales (sub-continentales)

Region

Carbon emission

Americas

<1 GT

Europe & Middle East

5.2 GT

Indian subcontinent

6.1 GT

Subsaharan Africa

2.3 GT

East Asia

11 GT

Southeast Asia

6.7 GT

World

29 GT

Discussion

14The simulation results shown are realistic hindcasts of population density, Prehistoric subsistence economy and resulting deforestation, at annual resolution for 685 world regions. It is important to reconsider how this realistic hindcast came into existence to assess its utility for advancing our understanding of early human-climate interaction. The model parameter space is small with only seven parameters; to find values for these parameters, the simulation was run with identical starting conditions and randomly varied parameter values 106 times: the temporal and spatial distance of the first occurrence of simulated farming societies to the undisputed five independent centres for domestication by B. Smith (1997) is used to score each simulation run; only the best scoring simulation run is selected for further dissemination. Thus, the emergence of the Levant and Chinese regions as early agricultural centres in the model is the result of the imposed data constraints, and not a model prediction. The added value provided by the model is manifold. Foremost, the i) temporal and ii) spatial interpolation delivered by the model creates a consistent development trajectory of all world regions, despite widespread lack of continuous data. iii) It allows to test the sensitivity of simulated histories to external forcing and hypothesis testing in the archaeological community: K. Wirtz and C. Lemmen (2003) and C. Lemmen and K. Wirtz (2009) employed the model to assess socio-cultural vulnerability to climate fluctuations on intercontinental and intracontinental scales. In the light of the large uncertainties with the scaling parameters used to calculate deforestation in terms of carbon loss, it is not the absolute value of global deforestation but the continuous and explicit spatio-temporal attribution of deforestation which is a novel aspect in palaeoclimate reconstruction and modelling.

15Even considering the above caveats, the simulated total deforestation of 29 Gt (equivalent to a forest/woodland loss of 2·106 km2) is in line with published estimates. The increase between 6000 BC and 2000 BC is almost linear, extrapolation to preindustrial times would roughly double the prehistoric value to 54 Gt. Already in 1983, E. Matthews presented based on satellite observations an estimate for loss of woodlands and forests since the beginning of agriculture (Matthews, 1983). She gives a number of 108 km2 of lost forest and woodland (≈100 Gt C using the scaling of this study), which fits well with the estimate presented here, when this is extrapolated to present day. K. Strassmann et al. (2008) estimated a preindustrial value of 45 Gt and related this to a maximum atmospheric CO2 mixing ratio increase of 3 ppm. A higher estimate for total preindustrial carbon release due to agriculture was obtained by J. Olofsson and T. Hickler (2008). They find with the Lund-Potsdam-Jena vegetation model an emission of 114 Gt carbon from agriculture, but include land use conversion outside forested areas in their analysis. Even with their higher estimate, the total preindustrial emissions would not be sufficient to confirm W. Ruddiman (2003)’s early anthropogenic greenhouse hypothesis. This study adds more support to disproving an early anthropogenic greenhouse, though it is not entirely independent of the study by K. Strassmann et al. (2008) based on HYDE 3.0; furthermore, we show that the direct emission from deforestation and forest soil loss was too low to have caused a significant rise of atmospheric carbon dioxide. An equally large greenhouse effect may have been caused by methane release from the conversion of wetlands and from crop emissions (Ruddiman, 2003), both of which are not considered in this study. K. Strassmann et al. (2008) found that carbon emission due to land demand for crops and for pasture is of similar magnitude: the disregard of pasture in this study indicates that the calculation likely underestimates total carbon emission due to land use. This underestimation, however, may be much less than 50%, because the conversion from forest to pasture releases less carbon than the conversion to crops (Guo and Gifford, 2002), and pasture may have been predominantly created from natural grasslands instead of forests.

16In this study:

  • Only the contribution of removing forest and associated soil loss is considered; we do not consider, e.g., the carbon release from converting peatland to agriculture, which would require a reconstruction of palaeosoils which is not yet included in GLUES.

  • Interaction of land use with the climate system does not end with the removal of the standing forest and associated belowground carbon: important additional agriculturally influenced drivers of the landscape with climate are, e.g., surface roughness, albedo and latent heat flux. A description of this interaction would require a bi-directionally coupled system consisting of a global climate model, a vegetation module and the sociotechnological simulator.

  • Prior to the calculation employing the DGVM, a simple conversion using aboveground carbon density of forest of 100 t/ha gives a reasonable value for temperate forests but can be much higher in tropical forest. Globally, a value of 120 t/ha may be realistic though even this number is not well constrained (e.g., Houghton, 2005). Surprisingly, this crude ‘back of the envelope’ calculation yielded almost identical results as my simulation employing a dynamical vegetation model and differential treatment of live carbon and soil carbon losses. The simple scaling is valid because most of the early farming areas are naturally forested areas, as was shown by the simulation of potential forest (fig. 6).

  • The disagreement between the crop fraction predicted with GLUES and the crop fraction derived from the HYDE data base in most regions can be in part traced back to the simulated population density which is an order of magnitude greater in the GLUES simulation for all those areas where farming is established. While in GLUES, the global population compares well to A. Coale (1974)'s estimate, the HYDE data base relies on interpolation of regional data from C. McEvedy and R. Jones (1978); Prehistoric population estimates differ by a factor of 10 between many studies. In view of the sound evidence for farming in many areas where the GLUES simulation shows deforestation (e.g., Gronenborn, 1999; Huang et al., 2003; Staubwasser et al., 2003; Gupta, 2004; Habu, 2004), and considering the high Levantine fraction of crop also in the HYDE data base, the HYDE crop fraction estimates for southern and eastern Eurasia are likely too low. The HYDE database was recently used by J. Kaplan et al. (2009) in a land use and land cover change model which includes technological changes. Their earliest hindcast of European deforestation is at 1000 BC; at this point in time, their calculated sensitivity of lost forest fraction to the inclusion of technological change is 20% for Europe (Kaplan et al, 2009, their fig 8). This highlights the importance of explicitly including technology (associated in the GLUES model with economic change) in the calculation of past land use and not to rely only on population estimates.

  • This study does not address the impact of Holocene climate variability on the development of early agriculture and associated deforestation. The proxy data collection by K.W. Wirtz et al. (2009) showed that climate variability increased in the second half of the Holocene in a Panamerican corridor, where agriculture-based societies developed rather late-relative to western Eurasia and east Asia-despite early domestication successes; K. Wirtz and C. Lemmen (2003) demonstrated earlier that regular climate fluctuations may have impacted American societies more than Eurasian societies. Current preliminary studies with GLUES show that susceptibility to climate changes in Europe is negligible if these are not accompanied by loss of technology.

Conclusion

17The GLUES model provides a consistent, realistic, and spatio-temporally continuous hindcast of prehistoric population density and socio-technological change between 10,000 BC and 2000 BC. In combination with potential forest and carbon pool estimates from a dynamical vegetation model, deforestation and carbon release due to land clearing for crops can be quantified on a global scale with regional resolution. In total, 29 Gt of carbon were released by anthropogenic deforestation between 10,000 BC and 2000 BC to give room to growing crops. Most of the release occurred in the heavily populated regions of southeast Europe, the Levant, south Asia and southeast Asia, China, and Japan, all of which are situated in a broad advanced technology belt ranging from central Europe via the Middle East, India and Indochina to China and Japan. In contrast to the historical cropland estimates found in the HYDE database, here croplands appear continuous in space throughout this farming belt. Small contributions to global deforestation are visible in several parts of Africa, i.e., the Maghreb, western Africa, and southeast Africa. This study adds more evidence that the consideration of economic and technological change in land use calculations is important. Despite existing uncertainties in scaling parameters, the integration of databases, potential land cover and technology modelling is a promising way to consistent estimates of land use (change) and associated emissions-from the global to the regional level.

Haut de page

Bibliographie

Ackland G., Signitzer M., Stratford K., Cohen M. (2007) – Cultural hitchhiking on the wave of advance of beneficial technologies. Proceedings of the National Academy of Sciences 104-21, 8714-8419.

Ambrose S.H. (1998) – Late Pleistocene human population bottlenecks, volcanic winter, and differentiation of modern humans. Journal of Human Evolution 34, 623-651.

Axtell R., Epstein J., Dean J., Gumerman G., Swedlund A., Harburger J., Chakravarty S., Hammond R., Parker J., Parker M. (2002) – Population growth and collapse in a multiagent model of the Kayenta Anasazi in Long House Valley. Proceedings of the National Academy of Sciences 99, 7292-7279.

Bogaard A. (2004) – The nature of early farming in central and south-east Europe. Documenta Praehistorica 31, 49-58.

Brovkin V., Ganopolski A., Svirezhev Y. (1997) – A continuous climate-vegetation classification for use in climate-biosphere studies. Ecological Modelling 101, 2-3, 251-261.

Brovkin V., Bendtsen J., Claussen M., Ganopolski A., Kubatzki C., Petoukhov V., Andreev A. (2002) – Carbon cycle, vegetation and climate dynamics in the Holocene: Experiments with the CLIMBER-2 model. Global Biogeochemical Cycles 16-4, 1139.

Childe V.G. (1936)Man makes himself. Watts, London, 275 p.

Coale, A., (1974). The history of the human population. Scientific American 231-3, 40-51.

Crutzen P.J., Stoermer E.F. (2000)The ‘Anthropocene’. IGBP Newsletter 41, W. Steffen, National Academy of Sweden, Stockholm, 17-18.

Curtis P., Hanson P., Bolstad P., Barford C., Randolph J., Schmid H., Wilson K. (2002) – Biometric and eddy-covariance based estimates of annual carbon storage in five eastern North American deciduous forests. Agricultural and Forest Meteorology 113, 1-4, 3-19.

Dieckmann U., Law R. (1996) – The dynamical theory of coevolution: a derivation from stochastic ecological processes. Journal of Mathematical Biology, 34, 579-612.

Flannery K.V. (1973) – The origins of agriculture. Annual Review of Anthropology 2-1, 271-310.

Fort J., Jana D., Humet J. (2004) – Multidelayed random walks: Theory and application to the Neolithic transition in Europe. Physical Review E70, 3, 867-870.

Gregg S. (1988)Foragers and farmers: population interaction and agricultural expansion in prehistoric Europe. University of Chicago Press, Chicago, 275 p.

Gronenborn D. (1999) – A variation on a basic theme: The transition to farming in southern central Europe. Journal of World Prehistory 13-2, 123-210.

Guo L.B., Gifford R.M. (2002) – Soil carbon stocks and land use changes: a meta analysis. Global Change Biology 8, 345-360.

Gupta A. (2004) – Origin of agriculture and domestication of plants and animals linked to early Holocene climate amelioration. Current Science 87-1, 54-59.

Habu J. (2004)Ancient Jomon of Japan. Cambridge University Press, Cambridge, 332 p.

Houghton R. (2005) – Aboveground forest biomass and the global carbon balance. Global Change Biology 11-6, 945-958.

Huang C.C., Zhao S., Pang J., Zhou Q., Chen S., Li P., Mao L., Din M. (2003) – Climatic aridity and the relocations of the Zhou culture in the southern loess plateau of China. Climatic Change 61, 361-378.

Kaplan J., Krumhardt K., Zimmerman N. (2009) – The prehistoric and preindustrial deforestation of Europe. Quaternary Science Reviews, in press.

Klein Goldewijk K., Van Drecht G., Bouwman A. (2007) – Mapping contemporary global cropland and grassland distributions on a 5 × 5 minute resolutionn. Journal of Land Use Science 2-3, 167-190.

Leemans R., Cramer W.P. (1991)The IIASA database for mean monthly values of temperature, precipitation and cloudiness of a global terrestrial grid. Research report, International Institute of Applied Systems Analyses, Laxenburg, Austria, 68 p.

Lemmen C., Wirtz K.W. (2009) – Socio-technological revolutions and migration waves re-examining early world history with a mathematical model. In Gronenborn D., Petrasch J. (Ed.): The Spread of the Neolithic to Central Europe, Mainz, 2006. RGZM Tagungen, in press.

Matthews E. (1983) – Global vegetation and land use: New high-resolution data bases for climate studies. Journal of Applied Meteorology 22-3, 474-487.

McEvedy C., Jones R. (1978)Atlas of world population history. Puffin Books, Harmondsworth, Middlesex, 368 pages.

Olofsson J., Hickler T. (2008) – Effects of human land-use on the global carbon cycle during the last 6,000 years. Vegetation History and Archaeobotany 17-5, 605-615.

Ruddiman W. (2003) – The anthropogenic greenhouse era began thousands of years ago. Climatic Change 61-3, 261-293.

Field, .C.B., Raupach, M.R. (2004)The Global Carbon Cycle. The Global Carbon Cycle: Integrating Humans, Climate, and the Natural World, Scientific Committee on Problems of the Environment Series 62, Island Press, Washington, 526 p.

Shennan S. (2001) – Demography and cultural innovation: a model and its implications for the emergence of modern human culture. Cambridge Archaeological Journal 11-1, 5-16.

Simmons I.G., (1996)The environmental impact of later Mesolithic cultures: the creation of moorland landscape in England and Wales. Edinburgh University Press, Edinburgh, 260 p.

Smith B. (1997) – The initial domestication of Cucurbita pepo in the Americas 10,000 years ago. Science 276-5314, 932-934.

Staubwasser M., Sirocko F., Grootes P., Segl M. (2003) – Climate change at the 4.2 ka BP termination of the indus valley civilization and Holocene south Asian monsoon variability. Geophysical Research Letters 30-8, 1425, doi:10.1029/2002GL016822, 2003.

Strassmann K., Joos F., Fischer G. (2008) – Simulating effects of land use changes on carbon fluxes: past contributions to atmospheric CO2 increases and future commitments due to losses of terrestrial sink capacity. Tellus, B 60-4, 583-603.

Turney C.S.M., Brown H. (2007) – Catastrophic early Holocene sea level rise, human migration and the Neolithic transition in Europe. Quaternary Science Reviews 26, 17-18, 2036-2041.

Williams M. (2000) – Dark ages and dark areas: global deforestation in the deep past. Journal of Historical Geography 26-1, 28-46.

Wirtz K.W., Eckhardt B. (1996) – Effective variables in ecosystem models with an application to phytoplankton succession. Ecological Modelling 92, 33-54.

Wirtz K., Lemmen C. (2003) – A global dynamic model for the Neolithic transition. Climatic Change, 59-3, 333-367.

Wirtz K.W., Bernhardt K., Lohmann G., Lemmen C. (2009) – Mid-Holocene regional reorganization of climate variability. Climate of the Past Discussions 5, 287-326.

Zeder M. (2008) – Domestication and early agriculture in the Mediterranean basin: Origins, diffusion, and impact. Proceedings of the National Academy of Sciences 105-33, 11597-11604.

Zvelebil M., Dolukhanov P. (1991) – The transition to farming in eastern and northern Europe. Journal of World Prehistory 5-3, 233-278.

Haut de page

Annexe

Version française abrégée

Le début de l’Anthropocène est placé par certains auteurs autour de 1750 (Crutzen et Stoermer, 2000). L’apparition de cette période marque le passage d’un environnement préindustriel à un environnement industrialisé, caractérisé par des interventions majeures d’origine anthropique sur les écosystèmes. Pour W. Ruddiman (2003), l’Anthropocène se situe bien avant le milieu du XVIIIè siècle et débute il y a 8000 ans quand la diffusion des pratiques agricoles néolithiques se généralise. En effet, après que les forêts aient connu leur extension maximale au début de la dernière période postglaciaire, les pratiques agropastorales du Néolithique ont eu pour conséquence de grandes phases de défrichement afin de disposer de surfaces cultivables. Au total, ce sont près de 107 km2de forêts qui ont été détruits au bénéfice de l’agriculture et du pastoralisme (Williams, 2000). Ces défrichements ont induit une émission de carbone de l’ordre de 100 Gt, soit l’équivalent de 15 % des émissions actuelles par la végétation de la planète. Cette étude a pour but de montrer comment il est possible de cartographier les impacts spatio-temporels de la déforestation au cours de la Pré- et Protohistoire.

La modélisation par le logiciel GLUES indique que les pratiques agropastorales de subsistance ont commencé vers 8000 av. J.-C. au Proche-Orient et en Chine. Vers 6000 av. J.-C., on observe une diffusion importante de l’agriculture depuis les foyers pionniers vers l’Europe du sud-est et à travers la Chine et le Japon, tandis qu’un nouveau foyer agricole fait son apparition dans la partie amont de la vallée de l’Indus. Vers 4000 av. J.-C., certaines régions d’Afrique (Maghreb, Afrique de l’ouest et du sud-est) se convertissent également à l’agriculture. Les espaces cultivés se généralisent durant cette période en s’étendant depuis l’Espagne jusqu’au Japon. Vers 2000 av. J.-C., les techniques agricoles se développent à partir du centre et du sud-est de l’Europe vers l’Est, l’Inde, l’Indochine, la Chine et le Japon. A cette période, l’Europe du sud-est, l’est de l’Inde, le Vietnam, le Japon et la Chine présentent les plus fortes densités de population du globe (3-6 hab/km2). La précision du modèle d’expansion des activités agricoles a été testée en étalonnant la durée de la phase de transition à l’aide de datations par le radiocarbone réalisées sur des sites néolithiques en Eurasie (partie occidentale ; Turney et Brown, 2007). Environ 11 % des terres de la planète sont cultivées vers 2000 av. J.-C.

L’extension des cultures se superpose de manière satisfaisante à la carte des densités de population des régions où les techniques agricoles sont particulièrement avancées. Les régions où les cultures sont le plus développées concernent les Balkans, le Proche-Orient, le sud et le nord de la Chine ainsi que le Japon. La plupart des régions ayant connu un développement agricole précoce étaient auparavant recouvertes de forêts ; ces mêmes espaces ont donc subi une déforestation massive du fait de la mise en culture. Les pertes estimées les plus importantes en carbone concernent le Japon, la Grèce septentrionale et le sud de la Chine, avec une valeur proche de 36 t/ha. A l’échelle de la planète, la perte en carbone est estimée à 11 Gt vers 7000 av. J.-C. et à 29 Gt vers 2000 av. J.-C. L’importante mutation des paysages de la Chine, du sud et du sud-est de l’Asie et de la partie occidentale de l’Eurasie a contribué à des pertes totales de carbone estimées à 90 % ; pour leur part, l’Afrique sub-saharienne et les Amériques n’ont participé seulement qu’à hauteur de 10 % à la perte totale de carbone depuis la période préhistorique (8000 av. J.-C.).

Haut de page

Table des illustrations

URL http://journals.openedition.org/geomorphologie/docannexe/image/7756/img-1.jpg
Fichier image/jpeg, 20k
URL http://journals.openedition.org/geomorphologie/docannexe/image/7756/img-2.jpg
Fichier image/jpeg, 20k
Titre Fig. 1 – Workflow for dynamic hindcasting of deforestation based on databases for cropland (historical database of the Environment, HYDE) and climate variables (IIASA), the dynamic global vegetation model VECODE, and simulated population density and agropastoral subsistence fraction from GLUES.Fig. 1 – Cadre schématique général de la modélisation de la déforestation fondé sur l’utilisation combinée des bases de données des terres cultivées (base de données historiques de l’Environnement, HYDE) et des variables climatiques (IIASA), la modélisation des dynamiques végétales est réalisée grâce à VECODE, les densités de population estimées et les secteurs de subsistence agropastorales sont déterminés avec GLUES.
URL http://journals.openedition.org/geomorphologie/docannexe/image/7756/img-3.jpg
Fichier image/jpeg, 292k
Titre Fig. 2 – Map of simulated technological efficiency (left panel) and population density (right panel) at 2000 BC.Fig. 2 – Carte simulant la diffusion technologique (vignette de gauche) et la densité de population (vignette de droite) à 2 000 BC.
Légende Technology ranges between 1 (the Mesolithic level) and 8 for near Bronze-age technology, population density attains values up to 6 inhab/km2, up from the initial uniform hunter-gatherer density of 0.1 inhab/km2.Indices de technologie entre 1 (mésolithique) et 8 (Âge du Bronze), les densités de population atteignent des valeurs proches de 6 habitants au km², alors que celles des communautés de chasseurs/cueilleurs ont des valeurs de 0,1 hab/km².
URL http://journals.openedition.org/geomorphologie/docannexe/image/7756/img-4.jpg
Fichier image/jpeg, 196k
Titre Fig. 3 – Time slices of the farmer share simulated with GLUES.Fig.3 – Différentes phases d’activités fermières simulées avec GLUES.
Légende Colour indicates the range from initially only foraging (light) with short mixed-strategy transitions to mainly agriculture and herding (dark).Les couleurs indiquent les différentes possibilités s’échelonnant des activités fourragères exclusives (en bleu) à une agriculture/pastoralisme principalement (en clair), en passant par une courte période de transition (en sombre).
URL http://journals.openedition.org/geomorphologie/docannexe/image/7756/img-5.jpg
Fichier image/jpeg, 332k
Titre Fig. 4 – Time of reaching the Neolithic stage (threshold 50% farming ratio, between 8000-4000 BC). Simulation results (background shading) compared to the data set of Neolithic sites by Turney and Brown (2007, shown as circles).Fig. 4 – Délai pour atteindre la phase néolithique (seuil de 50 % pour les activités fermières, période 8 000-4 000 BC) simulé à partir de GLUES (arrière plan), comparaison avec la répartition des sites néolithiques (points), d’après Turney and Brown (2007).
URL http://journals.openedition.org/geomorphologie/docannexe/image/7756/img-6.jpg
Fichier image/jpeg, 296k
Titre Fig. 5 – Map of simulated crop fraction (left panel) and crop fraction interpolated to the simulation regions from the HYDE database (right panel). Fig. 5 – Carte stimulant la part des secteurs cultivés (vignette de gauche) et des secteurs cultivés interpolés à partir de la base de données HYDE (vignette de droite).
URL http://journals.openedition.org/geomorphologie/docannexe/image/7756/img-7.jpg
Fichier image/jpeg, 216k
Titre Fig. 6 – Map of potential carbon density (left panel) and deforested carbon density by replacement with crops, including partial soil loss (right panel).Fig. 6 – Cartes de densité potentielle de carbone (vignette de gauche) et de densité de carbone causée par la déforestation et de remplacement par les cultures incluant en partie des pertes de sols (vignette de droite).
Légende The boxes define selected subcontinental-scale regions, for which the total carbon emission is listed in tab. 1.Les cartons définissent les régions choisies pour la présente étude et pour lesquelles les émissions de carbone sont listées sur tab. 1.
URL http://journals.openedition.org/geomorphologie/docannexe/image/7756/img-8.jpg
Fichier image/jpeg, 195k
Haut de page

Pour citer cet article

Référence papier

Carsten Lemmen, « World distribution of land cover changes during Pre- and Protohistoric Times and estimation of induced carbon releases », Géomorphologie : relief, processus, environnement, vol. 15 - n° 4 | 2009, 303-312.

Référence électronique

Carsten Lemmen, « World distribution of land cover changes during Pre- and Protohistoric Times and estimation of induced carbon releases », Géomorphologie : relief, processus, environnement [En ligne], vol. 15 - n° 4 | 2009, mis en ligne le 01 janvier 2012, consulté le 17 décembre 2017. URL : http://journals.openedition.org/geomorphologie/7756 ; DOI : 10.4000/geomorphologie.7756

Haut de page

Auteur

Carsten Lemmen

GKSS Forschungszentrum Geesthacht GmbH, Institut für Küstenforschung, Max-Planck Straße 1, 21501 Geesthacht, Germany (carsten.lemmen@gkss.de)

Haut de page

Droits d’auteur

© Groupe français de géomorphologie

Haut de page
  • Logo CNRS - Institut des sciences humaines et sociales
  • Logo Groupe français de géomorphologie
  • OpenEdition Journals