Population-level outcomes of multi-parasite epidemics can depend on the stability of infection at the within-host scale. Within hosts, host immune system and resources together influence stability of co-infection. Shifts in the stability can result in coexistence, either stably or via oscillations, or exclusion via alternative stable states of parasites. A trait-based, mechanistic framework can predict this range of outcomes at the within-host scale. However, we still need better general but mechanistic, within-host co-infection models. We offer some such by first drawing on classic food web modules, intra-guild predation (IGP) and keystone predation (‘diamond’, KP). However, we illustrate their stability properties via feedback loops, a powerful tool to evaluate how and why interactions alter stability. We then draw parallels to immune system - resource models with one parasite (‘PIE’, like IGP) or two parasites (‘PIPE’, like KP). We then evaluate within-host mechanisms which govern shifts in parasite stability.
Results/Conclusions
Every model evaluated has upper and lower level feedback loops. The relative magnitude and sign of these levels of feedback govern stability. The lower level feedbacks come from (negative) density dependence of the players on themselves or each other; upper level feedback traced through 2-4 species loops. IGP and PIE models exhibited stable coexistence and oscillations (and IGP allowed alternative stable states). Stable coexistence arose when the strength of lower level exceeded upper level feedbacks; oscillations arose due to delays triggered from strong upper level ones. Allocation of energy by hosts to the immune system enhanced stability. KP and PIPE models showed only coexistence or alternative stable states. These outcomes depend upon positive vs. negative components of the fourth level of feedback. Our approach of combining loops and analogies from community ecology can be a promising way forward to better understand population-level outcomes of multi-parasite epidemics.