In co-infected populations, the way that parasites interact within hosts influences parasite prevalence. However, within-host interactions are not homogenous across hosts- rather they can be determined by the order in which parasites infect hosts, a.k.a. priority effects. Priority effects are important because they can create feedback loops between within-host processes and between-host processes. This is because the relative prevalence of co-infecting parasites in a host population influences the order of arrival of parasites in an individual host. This then determines within-host interactions, which then influence relative prevalence of parasites. These feedback loops may create systematic changes in the course of multi-parasite epidemics. Thus, we asked (1) do priority effects create feedback loops between within host and between host processes? and (2) how do these feedback loops alter multi-parasite epidemics? We answered these questions experimentally using Daphnia dentifera co-infected by Pasteuria ramosa and Metschnikowia bicuspidata. To answer question 1, we first tested how changes in relative spore density in the environment changed the order of arrival of P. ramosa and M. bicuspidata. We then measured how per capita spore transmission changed due to priority effects. To answer question 2, we compared a multi-pathogen epidemic model parameterized with and without priority effects.
Results/Conclusions
In experiment one, we found that increasing the relative abundance of parasites in the environment increased the likelihood that a parasite would arrive first in a host. For experiment two, we found that P. ramosa always had lower transmission from co-infected hosts than from singly infected hosts, but that it had a higher transmission from co-infected hosts when it was the second parasite to arrive vs. the first parasite to arrive. We found that M. bicuspidata had a higher transmission rate from co-infected hosts than singly infected hosts if it was the second parasite to arrive in a host, but a lower transmission rate from co-infected hosts than singly infected hosts if it was the first parasite to arrive. Combined, these experiments suggest that parasite transmission increases at low relative prevalence, creating negative feedback loops that may mediate the negative impact that parasites have on one another. However, our co-infection epidemic model shows that M. bicuspidata is relatively unaffected by priority effects, but that priority effects decrease the size of P. ramosa epidemics. Overall, our results show that priority effects can systematically change the course of epidemics, and should be taken into consideration in future disease modelling of co-infected populations.