Using multi‐response models to investigate pathogen coinfections across scales: Insights from emerging diseases of amphibians

Abstract

Summary Associations among parasites affect many aspects of host-parasite dynamics, but a lack of analytical tools has limited investigations of parasite correlations in observational data that are often nested across spatial and biological scales.Here we illustrate how hierarchical, multiresponse modeling can characterize parasite associations by allowing for hierarchical structuring, offering estimates of uncertainty, and incorporating correlational model structures. After introducing the general approach, we apply this framework to investigate coinfections among four amphibian parasites (the trematodes Ribeiroia ondatrae and Echinostoma spp., the chytrid fungus Batrachochytrium dendrobatidis, and ranaviruses) and among >2000 individual hosts, 90 study sites, and five amphibian host species.Ninety-two percent of sites and 80% of hosts supported two or more pathogen species. Our results revealed strong correlations between parasite pairs that varied by scale (from among hosts to among sites) and classification (microparasite versus macroparasite), but were broadly consistent across taxonomically diverse host species. At the host-scale, infection by the trematode R. ondatrae correlated positively with the microparasites, B. dendrobatidis and ranavirus, which were themselves positively associated. However, infection by a second trematode (Echinostoma spp.) correlated negatively with B. dendrobatidis and ranavirus, both at the host- and site-level scales, highlighting the importance of differential relationships between micro- and macroparasites.Given the extensive number of coinfecting symbiont combinations inherent to natural systems, particularly across multiple host species, multiresponse modeling of cross-sectional field data offers a valuable tool to identify a tractable number of hypothesized interactions for experimental testing while accounting for uncertainty and potential sources of co-exposure. For amphibians specifically, the high frequency of co-occurrence and coinfection among these pathogens – each of which is known to impair host fitness or survival – highlights the urgency of understanding parasite associations for conservation and disease management.

The application of community ecology theory to co-infections in wildlife hosts

“…Given that immune responses are costly and can be traded off with other demanding processes, such as reproduction and growth (Lochmiller and Deerenberg 2000), mounting an immune response sufficient to combat both of these parasites could have been beyond the resource capacity of our hosts. Field studies have found that Ranavirus and Bd are not associated with each other at the site (Olori et al 2018, Stutz et al 2018) or individual level, but there are positive associations between these parasites at the species level for some of the sampled amphibian species (Stutz et al 2018). Despite this, co-infections with Bd and Ranavirus, are common in nature (Whitfield et al 2013, Warne et al 2016, Watters et al 2018) and our findings suggest that these co-infections can be facilitative and thus appear to put host populations at greater risk than single infections.…”

“…This nematode is a common parasite of amphibians in the Southeastern US and Latin America (Baker 1987, Vhora and Bolek 2013, Ortega et al 2015). Importantly, parasitic worms, Bd, and Ranaviruses are known to co-occur in amphibian hosts in the field (Hoverman et al 2012, Stutz et al 2018, Watters et al 2018). In the Peruvian Andes, an amphibian biodiversity hotspot, up to 50% of frogs were coinfected with Bd and Ranavirus (Warne et al 2016).…”

Guidance for Developing Amphibian Population Models for Ecological Risk Assessment

“…Other landscape-level influences on amphibian population dynamics include habitat degradation, pathogen presence, and chemical exposure. Interactions between parasite occurrence, habitat degradation, and chemical exposure introduce further complications in assessment of vulnerability (Koprivinkar and Redfern 2012;Stutz et al 2017) with the potential for agrochemicals to reduce immunity to parasite infection (Kiesecker 2002;Rohr et al 2008) and decrease parasite survival (Hua et al 2016). Multiple parasites do not always produce synergistic effects; compensatory effects of coinfection are also possible (Wuerthner et al 2017).…”

Endemic infection can shape exposure to novel pathogens: Pathogen co‐occurrence networks in the Serengeti lions

“…Detecting associations between pathogens is also likely to depend on taxonomic and spatial scales that are seldom considered (Araújo & Rozenfeld ; Stutz et al . ). Studies commonly aggregate pathogen data to genus level, but associations between pathogens can be subtype or genotype‐specific (e.g.…”