Demographic interactions between the last hunter-gatherers and the first farmers

Abstract

Demographic interaction processes play a pivotal role during episodes of cultural diffusion between different populations, particularly when these episodes can lead to competition for the same resources and geographic space. The diffusion of farming is one prototypical case within this broader scenario, where groups of incumbent hunter-gatherers occupied a space which would later be claimed by expanding farmers. In this work, we tackle such processes through a two-population mathematical model, where farmers and foragers compete and interact in the same geographic space. We present this work as a conceptual approach where, first, we assess the implications of our theoretical model and its general applicability and, second, we empirically test it on three case studies: Denmark, Eastern Iberia, and the island of Kyushu (Japan). While these regional case studies do not encompass the full range of processes observed in the interaction between migrant farmers and incumbent hunter-gatherers they provide reasonable variation to illustrate how our model can be fitted to a diverse range of empirical data and provide insights into these demographic processes. In particular, our theoretical model and case studies illustrate how endogenous interaction processes alone can explain the demographic fluctuations observed in the archaeological record during this transition, highlighting how these should be accounted for before invoking external forces as primary drivers.

Keywords: demographic interaction; dynamic modelling; farming expansion; group competition; population dynamics.

Fig. 1.

Map of the selected regions (Denmark, Iberia, and Kyushu). Red dots refer to sites associated with farming, blue dots are associated with hunter-gatherers, and yellow dots show dates for both of them.

Fig. 2.

Experiment 1. Event 2. Time taken for the farmers to reach their carrying capacity under the full range of the interspecific mortality parameters. Growth rate parameter is set to 0.015 for the hunter-gatherers and 0.02 for the farmers. Assimilation parameter and initial population ratio follow the specifications of Table 1. Magenta indicates the parametric region where the farmers will be the only surviving population, yellow area indicates coexistence, in the blue area only hunter-gatherers will survive and in the gray area only one population survives, and which one does depends on the initial conditions. Isolines are set for every 100 y.

Fig. 3.

Experiment 2. Event 3. Time taken for the hunter-gatherers to completely disappear under the full range of the growth parameters. Interspecific mortality parameter is set to 0.01 for both populations. Assimilation parameter and initial population ratio follow the specifications of Table 1. Magenta indicates the parametric area where the farmers will be the only surviving population and yellow area indicates coexistence. Isolines are set for every 100 y.

Fig. 4.

Population dynamics for the three areas considered. The left column shows the fitted SPDs produced by the model and the right column shows the actual population dynamics produced. On the fitted SPDs, lines represent the observed population and lighter bands represent the 95% HPDI. The right hand figures represent the accepted normalized population curves produced by the model. Red color represents the hunter-gatherers and green color the farmers.

Fig. 5.

Posterior distribution of the parameters in the observed data per area. Bars indicate the 95% HPDI and dashed line represents the median. Y axes indicate the range of the uniform prior.

Fig. 6.

Time taken for different events reflecting the substitution of hunter-gatherers by farmers in each region. Darker areas indicate the 95% HPDI and dashed line represents the median.