The eye of the storm: silent infections driving amphibian declines

Chytridiomycosis caused by the chytrid fungi Batrachochytrium dendrobatidis (Bd) and Batrachochytrium salamandrivorans (Bsal) is considered a major driver of global amphibian declines. The regions where Bd and Bsal are clearly causing amphibian declines are generally considered to have been naïve to these pathogens, with recent colonization by virulent lineages. Although research increasingly points to potentially promising applications such as vaccination, treatment/disinfection or bioaugmentation, effective in situ mitigation of the impact of amphibian chytridiomycosis remains a challenge (Garner et al., 2016). A primary concern in the epidemiology of chytridiomycosis is the occurrence of biotic and abiotic reservoirs which sustain fungal populations, facilitating amphibian extirpation and even extinction events. Reflecting the obvious and immediate impact on biodiversity, most published studies focus on the impact of disease epidemics on the most severely affected species in the community. In their study, Brannelly et al. (2017) provide empirical insights about disease reservoirs in an Australian ecosystem where Bd has become endemic. By combining analysis of Bd disease dynamics in a frog species that is not susceptible to clinical Bd infection, over an 8-year period with skeletochronology and a laboratory experiment, they demonstrate the presence of a high prevalence of persistently infected animals that are sympatric with several species known to have declined due to Bd. In general, the main conservation implication of the presence of reservoir hosts is that focusing interventions on the most vulnerable species, or those of greater conservation importance, might prove futile if other species maintain the infection. This may apply equally to in situ, treatment-based mitigation and to longer-term ex situ rescue and reintroduction strategies. Understanding Bd endemism and the role of reservoirs therein is thus of fundamental importance in designing effective mitigation strategies for several reasons. Eradication of chytrid infections is not considered a feasible option, given the global occurrence of a hypervirulent Bd lineage in amphibian communities, its rapid range expansion (although mechanisms of transmission are poorly understood) and propensity to persist. Instead, mitigation may more effectively focus on minimizing disease impact by promoting a state of host-pathogen co-existence, reducing pathogen prevalence and infection intensity. Influencing any, and preferably multiple drivers of the epidemiological triangle (host, pathogen and environmental determinants) may shift an epidemic state characterized by high pathogen prevalence and infection loads to a low-prevalence, low-infection intensity state that may preclude reaching thresholds at which population collapses of highly susceptible species and the loss of entire amphibian communities occur (Lips et al., 2006; Briggs, Knapp & Vredenburg, 2010; Vredenburg et al., 2010). Reducing the infection pressure in the environment at least in selected systems indeed seems capable of alleviating the Bd impact at population level (e.g. Schmeller et al., 2013). Reservoir species play a key role in maintaining such an endemic state, functioning as drivers of infection pressure as described in Brannelly et al. (2017). Clarifying this role may therefore open opportunities for conservation actions: for example, reintroductions may be carried out in areas where niches of reservoir and highly susceptible host species do not overlap. The extant literature on disease ecology of multiple-host systems provides both theory and guiding principles for mitigation planning, including quantitative guidelines to determine which hosts to target (Anderson & May, 1986; Dobson, 2004). On the other hand, any practical action will need to be tailored to the specific multi-host context: while reducing prevalence and infection intensity may work for some species, this may not work for others. At least for Bsal, a mere reduction in both will probably not suffice to avert population declines, with very low infectious doses (a theoretical one zoospore) being capable of producing lethal disease and quick transmission within the highly susceptible species (Stegen et al., 2017; Canessa et al., 2018). Modeling maximum infection pressures tolerated by susceptible target species (one proxy may be dose dependency of infection) may thus inform conservation decisions. Moreover, to achieve such targets in hosts with different epidemiological dynamics may require differential treatment, which may not be feasible in the field. Again, ecological theory provides useful information for scenarios in which management can or cannot be targeted at specific hosts (Keeling & Rohani, 2008). Bd endemism in reservoir species may have the obvious impact on susceptible species as continuous source of infection and disease as shown by Brannelly et al. (2017). However, it should be noted that species that serve as reservoir hosts in a community are not necessarily immune from negative impacts; indeed, subtle, long-term effects on reservoir host populations may be underestimated. Spitzen-van der Sluijs et al. (2017) recently estimated the costs incurred by subclinical infection of frogs in North-western Europe. Infection and disease dynamics were in line with those described in Brannelly et al. (2017), suggesting a role of their study species as a reservoir host in the community. However, their longitudinal study also demonstrated a hitherto underestimated fitness cost of Bd infection, which caused additional mortality in the study species and thus increased the risk of extirpation events for small population sizes. If such dynamics were more common than currently assumed, the true global impact of chytrid infections might be substantially underestimated. Such impacts could still be negligible in widespread and abundant species as that studied in Brannelly et al. (2017); in other species however, at least those with small ranges and low population densities, estimating the fitness cost of endemic Bd infections would provide valuable information when calculating minimal population sizes. Whether to inform conservation of the reservoir species itself, or of more susceptible co-occurring hosts, we stress the importance of collecting data about population dynamics of more common and less obviously susceptible host species. The markedly different host responses to chytrid infections evidenced by Brannelly et al. (2017) call for intensified studies into the mechanisms of host susceptibility. Whereas susceptibility is the outcome of a complex interaction between host factors (innate and acquired immunity, including skin microbiomes and peptides, life stage, endocrine factors etc), pathogen virulence and environment, this study reiterates that, given identical environmental and pathogen determinants, there are intrinsic host factors which strongly determine the course of a chytrid infection. Although it is unlikely to find a single Columbus’ egg that offers a broad explanation, further elucidating the mechanisms (and underpinning genetics) of susceptibility and resistance may open perspectives for designing long-term and sustainable measures, such as marker-assisted selective breeding. In this context, the Brannelly et al. (2017) study provides a good example of the usefulness of a multidisciplinary approach, combining biological and veterinary sciences. The latter allowed disregarding Bd as a cause of mortality in the Bd infected frogs by performing histopathology on the infected skin. In conclusion, understanding the complex reality of wildlife diseases and their context could offer a solid base to address the societal challenge of biodiversity loss, and calls for joining forces in ecology, pathology, microbiology, epidemiology and mathematical modeling. The study by Brannelly et al. (2017), investigating endemic Bd infections in a reservoir host species, highlights this important, yet understudied component of the epidemiology of chytridiomycosis.