Agent-based modelling, molluscan population dynamics, and archaeomalacology

Agent-based modelling (ABM) is an emerging archaeological tool that offers insights into processes which are archaeologically invisible or difficult to detect. Here we illustrate the potential of ABM for archaeomalacology, posing two research questions and comparing ABM results with Pacific archaeological sequences. The first analysis considers how molluscan energetic return rates (ERR) and age of reproductive maturity (ARM), singularly or in combination, influence prey population resilience. The second analysis assesses how prey spatial structure affects foraging efficiency and prey susceptibility to resource depression. Consistent with expectations from evolutionary ecology and life history theory, the ABM results demonstrate that both ERR and ARM influence prey population resilience (or vulnerability). However, the analysis also demonstrates that ARM is the more important variable and taxa with high ERR (i.e., large-bodied) are disproportionately affected by human harvesting. Not only are efficient foragers more likely to target high ERR taxa, but these prey often have delayed ARM and un-foraged individuals are more likely to be smaller and immature, with disadvantages for population stability and recovery. In short, early-maturing taxa are highly resilient, while late-maturing organisms are more vulnerable; these outcomes also are observed archaeologically. The ABM analyses also demonstrate the effects of prey spatial structure on molluscan susceptibility to resource depression. High prey aggregation initially allows for high foraging efficiency, but prey abundance and encounter rates often rapidly decline. In contrast, when prey are dispersed, search time is greater, leading to lower encounter rates and reduced foraging efficiency, but greater prey population stability. Our ABM and archaeological examples further illustrate that while general principles can be derived, the resilience and spatial structure of specific prey populations, as well as foraging outcomes, are context dependent and continuously evolving. Finally, we note that model departures from theoretical expectations serve to stimulate further research, including use of additional parameters, consideration of novel contextual evidence, and/or investigation of social, technological or environmental hypotheses.