Mechanistic models can reveal infection pathways from prevalence data: the mysterious case of polar bearsUrsus maritimusandTrichinella nativa

Parasites exhibit a diverse range of life history strategies. Transmission to a host is a key component of each life cycle but the difficulty of observing host-parasite contacts has often led to confusion surrounding transmission pathways. Given limited data on most host-parasite systems, flexible approaches are needed for disentangling the obscure transmission dynamics of these systems. Here, we develop a modelling framework for formally testing long-standing hypotheses regarding how the parasitic nematodeTrichinella nativais maintained at high prevalences in polar bear populations. We evaluated transmission from marine prey, from scavenging terrestrial carrion, from cannibalism and from scavenging on dead infected bears as possible pathways, and assessed their respective importance by comparing model-projected prevalences for each mechanism against observed total and age-specific population prevalences in the Southern Beaufort Sea polar bear subpopulation. Cannibalism and the scavenging on conspecifics have previously been assumed to be critical transmission pathways, but despite data scarcity, our model exposes these mechanisms as ineffective across a wide range of plausible parameter values. Instead, our analyses suggest that transmission from the consumption of infected marine prey, and in particular seals, can explain observed prevalence levels by itself, with other transmission pathways likely playing varying small contributing roles. Furthermore, our model suggests that transmission declines with bear age, perhaps due to age-dependent changes in diet or immunity. By formalising multiple transmission mechanisms in a unified, mathematical framework, we distilled several hypotheses to a likely main mode ofT. nativatransmission to polar bears. The specifics of our model are tailored towards theT. nativa-polar bear system, but the approach is easily generalized; it provides a powerful, quantitative means for ecologists to explore competing hypothesis for parasite transmission and other difficult-to-observe animal interactions even in data-poor systems.