Agent-based host-parasite model
For my dissertation, I used an agent-based model to investigate the effects of information state and decision strategy, not only on tick success but also on tick distribution across the simulated host surface.
One benefit of modeling the spatial distributions of ticks, in addition to predicting their success, is to connect theory with the actual distributions that researchers find on hosts in the field. Because ticks colonize and leave the host continuously, the distribution of ticks seen on hosts is simply a snapshot of that moment in time. If some areas of the host’s body allow ticks to feed more rapidly than others, there will be an observational bias towards recording those ticks in areas where they feed more slowly.
By modeling the attachment and feeding process under the threat of grooming, I demonstrate that most ticks actually feed in risky but highly rewarding areas. The ticks typically seen on hosts are those that are simply there longer and may not reflect the locations where the majority feed. This insight could inform our understanding of which hosts actually support and feed the majority of ticks in systems with vector-borne diseases.
I wrote the model in Netlogo, and the figures in R.
Project link: Github