Human-Plant Coevolution: A modelling framework for theory-building on the origins of agriculture

Abstract

The domestication of plants and the origin of agricultural societies has been the focus of much theoretical discussion on why, how, when, and where these happened. The ‘when’ and ‘where’ have been substantially addressed by bioarchaeology, thanks to advances in methodology and the broadening of the geographical and chronological scope of evidence. However, the ‘why’ and ‘how’ have lagged behind, holding on to relatively old models with limited explanatory power. Armed with the evidence now available, we can return to theory by revisiting the mechanisms allegedly involved, disentangling their connection to the diversity of trajectories, and identifying the weight and role of the parameters involved. We present the Human-Plant Coevolution (HPC) model, which represents the dynamics of coevolution between a human and a plant population. The model consists of an ecological positive feedback system (mutualism), which can be reinforced by positive evolutionary feedback (coevolution). The model formulation is the result of wiring together relatively simple simulation models of population ecology and evolution, through a computational implementation in R. The HPC model captures a variety of potential scenarios, though which conditions are linked to the degree and timing of population change and the intensity of selective pressures. Our results confirm that the possible trajectories leading to neolithisation are diverse and involve multiple factors. However, simulations also show how some of those factors are entangled, what are their effects on human and plant populations under different conditions, and what might be the main causes fostering agriculture and domestication.

Human-Plant Coevolution: A modelling framework for theory-building on the origins of agriculture

“…The source files associated with the HPC model are maintained in a dedicated online repository [ 68 ]: https://github.com/Andros-Spica/hpcModel . This repository contains several additional materials, including a web application to run simulations and the full report on the sensitivity analysis.…”

“…The same procedure was applied for the dependency coefficients and timings; see section 5.2 in [ 68 ]. The assessment of parameter importance for the dependency coefficients displayed a similar pattern, only highlighting those parameters with a direct impact on the carrying capacity of the respective population (greens and blues).…”