The alternate role of direct and environmental transmission in fungal infectious disease in wildlife: threats for biodiversity conservation

Abstract

Emerging fungal pathogens have substantial consequences for infected hosts, as revealed by the global decline of amphibian species from the chytrid fungus. According to the "curse of the Pharaoh" hypothesis, free-living infectious stages typical of fungal pathogens lengthen the timespan of transmission. Free-living infectious stages whose lifespan exceeds the infection time of their hosts are not constrained by virulence, enabling them to persist at high levels and continue transmitting to further sensitive hosts. Using the only Mesomycetozoea fungal species that can be cultured, Sphaerothecum destruens, we obtained tractable data on infectivity and pathogen life cycle for the first time. Here, based on the outcomes of a set of infectious trials and combined with an epidemiological model, we show a high level of dependence on direct transmission in crowded, confined environments and establish that incubation rate and length of infection dictate the epidemic dynamics of fungal disease. The spread of Mesomycetozoea in the wild raise ecological concerns for a range of susceptible species including birds, amphibians and mammals. Our results shed light on the risks associated with farming conditions and highlight the additional risk posed by invasive species that are highly abundant and can act as infectious reservoir hosts.

Figure 1. Lifecycle of Sphaerothecum destruens

.a) Spores multiply within host cells until cell death; b) Spores spread within the host and are released into the water through urine, bile, or gut epithelium; c) In freshwater, each spore can divide into up to 5 uniflagellate zoospores and survive for several days depending on the water temperature. Infection occurs directly or indirectly by ingesting the spores, attachment to the gills or skin, or gut penetration. Photo R. E. Gozlan

Figure 2. A visual representation of the SEIR model categories.

Susceptible individuals (S) are exposed to infection at rate β (direct transmission) + ε (environmental transmission). In the exposed state (E), individuals become infectious (I) at rate σ. Infectious individuals either die as a result of disease at rate α, or recover (R) at rate °. Infected individuals release spores (Z) at rate f, including zoosporulation, which have a collective mortality rate of μ.

Figure 3. The model output for an outbreak of Sphaerothecum destruens in juvenile bream Abramis brama, roach Rutilus rutilus and carp Cyprinus carpio.

The model output (red) is compared with the observed data (black) from Andreou et al (2012) for Abramis brama (lower), Rutilus rutilus (middle) and Cyrpinus carpio (top). The minimum and maximum values of the three replicate samples are shown as dashed lines. Images are all in the public domain as copyrights have expired: http://commons.wikimedia.org/wiki/File:Braxen,_Iduns_kokbok.jpg, http://commons.wikimedia.org/wiki/File:Rutilus_rutilus5.jpg, http://commons.wikimedia.org/wiki/File:Cyprinus_carpio3.jpg

Figure 4. Surviving population of Leucaspius delineatus when exposed to Sphaerothecum destruens.

The model output (red) has been fitted to Paley et al (2012) published data (black) and projected for 250 days, to observe how the epidemic would progress.

Figure 5. Sensitivity analysis of incubation rate and recovery parameters.

Changing the parameter values of σ (incubation) (a) and γ (recovery) (b) from the original common bream Abramis brama model values of 0.1 (shown in red) reveals how largely they affect the outcome of the Sphaerothecum destruens epidemic. For σ, values lower than 0.1 (shown are 0.01, 0.05, 0.075) indicating a longer time of incubation, tend to slow down the epidemic curve, although the same number of individuals eventually succumb to the disease. Higher values of σ (0.125, 0.2, 0.3, 0.4, 0.5) speed up the progression of the epidemic, however the epidemics plateau at the same level as other values. In contrast, lower values of γ (tested at the same levels as σ) from 0.1 (indicating longer time to recovery, here shown under the model values in red) severely reduced population survival.

Figure 6. A comparison of direct and environmental transmission values for different species.

The minimum, optimised, and maximum values for environmental transmission are plotted against the optimised direct transmission for each species. Initially it appears as though there is a significant linear relationship between direct and environmental transmission (lm: ε = 0.811β – 0.02016, p < 0.001). However, when the leverage of the residuals is examined, it is clear that this relationship is only caused by the significantly higher transmission values for bream. When these are excluded, there is no linearity observed.