Final Report Summary - NICOSS (Niche Construction and Subsistence Shifts: Modelling coevolutionary dynamics between hunter-gatherers and vegetation communities) The NiCoSS (Niche Construction and Subsistence Shifts: Modelling coevolutionary dynamics between hunter-gatherers and vegetation communities) project’s primary goals were: 1) to develop new theories on how subsistence shifts are driven by different mechanisms of social learning; and 2) to reconstruct the co-evolutionary dynamics of human and plant communities in the Jomon hunter-gatherers of northern Japan. The first goal has been pursued through the development of a series of agent-based models apt to explore the consequences of how social learning is affected by population size and the different levels of uncertainty for the association between a cultural trait (e.g. a specific sowing technique, or the choice of a given variant of seed) and its “payoff” (e.g. the amount of crop produced in a season). Our simulation experiments showed how certain social learning rules (e.g. copying the most successful individual) provides evolutionary advantages when discerning the cause (i.e. a cultural trait) and the effect (i.e. the payoff) is trivial and population size is large. However, with increasing uncertainty in determining this association, these learning strategies offer decreasing advantages over alternative strategies, and an increase in population becomes deleterious rather than being beneficial. These theoretical insights show how population size and community structure can profoundly affect the rate of cultural evolution, especially for traits with high-payoff uncertainty such as those pertaining subsistence strategies. The results of the theoretical simulations have further highlighted the importance of demography for understanding patterns of subsistence shifts. We thus assessed its role in relation to the construction and management of anthropogenic chestnut forests among the Jomon hunter-gatherers of the Aomori prefecture in Northern Japan. Given the importance in establishing precise temporal relationships between the timing of key subsistence shifts and population fluctuations on the one hand, and potentially concurrent changes in climate on the other, a substantial effort has been dedicated on refining the existing chronological framework offered by Japanese archaeology. This was achieved by transposing the existing archaeological lines of evidence from the relative chronological framework of pottery-based cultural phases to an absolute one defined in calendar years. In particular, the Bayesian re-dating of radiocarbon samples associated to specific chronologically diagnostic elements (e.g. potsherds), enabled a more accurate estimate of the timing of subsistence shifts, as inferred by changes in the frequency of different subsistence tool-kits. Pollen data of several on-site and off-site locations have also been chronologically reassessed by means of Bayesian age-depth models, enabling to evaluate when anthropogenic forests emerged and how long they lasted. Finally, estimates on long-term population dynamics have been inferred by means of summed probability distribution of radiocarbon dates. This was the first application of the technique within the contest of Japanese archaeology, and offers an invaluable basis for future comparative research on sedentary hunter-gatherers. The development and use of a new simulation-based analysis allowed also the statistical comparison of the population dynamics of different regions in the Japanese archipelago, showing evidence of divergence that might be linked to difference in local adaptive strategies. A comparative assessment of these multiple lines of evidence showed a clear correlation between an episode of dramatic population increase which started during the second half of the 7th millennium BP and the emergence and expansion of anthropogenic chestnut forests in several key sites. While the initial stages of this demographic transition exhibit strong similarities to patterns observed in early agricultural societies, the rate of population growth revealed to be smaller, and the growing phase shorter in duration. These lines of evidence strongly suggest that the initial benefit linked to the creation of anthropogenic forests of chestnut led also to a limitation of the steady growth of Jomon communities; the physiology of these nut-bearing species in fact placed a limit in the continuous positive feedback between human and plant communities, ultimately affecting the long term trajectory of Jomon societies. Indeed, the demographic signals suggested by radiocarbon dates never reaches the same density as those observed in this initial stage, and instead exhibit repeated episodes of minor population fluctuations, with some degree of correlation to concurrent changes in subsistence, cooling events, and shifts in the composition of anthropogenic forests from chestnut to horse-chestnut. Ultimately the archaeological evidence offered by the Jomon case study showcase the strong path dependence dictated by niche construction activities, offering at the same a detailed portrait of an alternative account of prehistoric societies and their relationship to plant communities.
