Host-switching
After the first host speciation, parasite individuals in a host species can switch to another host. For each parasite individual, we randomly selected a host species, including the one in use. If the selected host species is not the original host (donor), we follow a probability function for the host-switching event. This probability of host-switching events (Phs ) decreases over time, representing the loss of opportunity and/or compatibility of parasites associated with the evolutionary history of hosts. The probability of a parasite individual successfully migrate (host-switching) from one host to another host species, in a given generation n , is defined as:
Phs( \(n\))=exp \([-r*(n-n_{s})]\), (1)
where r is a positive parameter that controls the decay of the host-switching probability, and ns is the generation that the common host ancestor had speciated (then, n-ns is how long the two host species had diverged). If r = 0, these probabilities are equal to 1 regardless of the host divergence time, meaning that there is no restriction to host-switching. As a consequence, parasite gene flow is continuous and speciation does not occur. At the other extreme, for sufficiently large r values (Phs ~0), host-switching is absent, and cospeciation between hosts and parasites is expected. For intermediary r value, some parasite individuals can eventually switch hosts. This will increase the host repertoire of the parasitic species, and also enable speciation by isolation (by host use), similar to the speciation by founder’s effect (Mayr 1999; Gavrilets & Hastings 1996). In order to better interpret the effect of parameter r on the trajectories and compare the results between the communities, we do not present our results in terms of r, but how much it changes the overall host-switching events. To obtain this overall metric, we calculated the expected percentage of parasite individuals that switch hosts over the entire simulation and we call it ashost-switching intensity .