Figure

Figure. Study sites for collection of amphibians in Western Europe, 2010–2016. Numbers correspond to field sites at which amphibians were collected and examined for Batrachochytrium salamandrivorans (Bsal) (Technical Appendix)....

During 2010–2016, we collected samples of free-living populations of newts and salamanders from 48 sites in the Netherlands, Belgium, and adjacent regions of the Eifel region in Germany (near the border with the Netherlands and Belgium) (Figure; Technical Appendix Table 1). Site selection was based on reported amphibian deaths, apparent negative amphibian population trends, preventive Bsal surveillance in susceptible populations, or geographic proximity to known outbreak sites. Samples were also collected at 6 additional sites in Germany and 1 in the Netherlands, which were located >100 km from the nearest known outbreak (Technical Appendix Table 2). Sampling was conducted by swabbing skin (7,8) of live animals and collecting skin samples from dead animals. All samples were kept frozen at −20°C until they were analyzed for the presence of Bsal DNA via real-time PCR, as described (9).

Across all 55 sites, we tested 1,019 fire salamanders (43 dead, 976 skin swab samples); at site 14, skin swab samples instead of tissue samples were collected from 16 dead salamanders. We also collected samples from 474 alpine newts (Ichthyosaura alpestris; 18 dead, 456 skin swab samples), 239 smooth newts (Lissotriton vulgaris; 2 dead, 237 skin swab samples), 80 palmate newts (Lissotriton helveticus; all skin swab samples), 79 crested newts (Triturus cristatus; all skin swab samples), and 30 Italian crested newts (Triturus carnifex; all skin swab samples). To obtain a Bayesian 95% credible interval for prevalence (online Technical Appendix), we used the computational methods of Lötters et al. (10). We ran 3 parallel Markov chains with 20,000 iterations each and discarded the first 5,000 iterations as burn-in; chains were not thinned.

Bsal was found at 14 of the 55 sites; infected amphibians were fire salamanders, alpine newts, and smooth newts. Our results demonstrate that the range of Bsal distribution may be up to ≈10,000 km² (measured as the surface of a minimum convex polygon encompassing the outermost points) across Germany, Belgium, and the Netherlands (Figure). The presence of Bsal in wild alpine newts and smooth newts shows distinct expansion of the known host range in the wild (Technical Appendix Table 1). Furthermore, we document that Bsal is present in natural fire salamander populations in Germany (confined to the Eifel region). At some sites, because of our sample sizes, the upper limit of the 95% credible interval for Bsal prevalence was as high as 0.7; therefore, we may have failed to detect Bsal at these sites (Technical Appendix Table 1). In addition, the fungus may have been present at several sites before first detection. For example, Bsal was detected at site 4, where population-monitoring efforts in the years before detection (2000–2013) showed declines in 4 newt species (http://www.ravon.nl/EID_SI_Spitzen_et_al_2016). However, because no samples were collected before 2015, we have no evidence for a causal relationship between the presence of Bsal and the declines. We have also recorded the presence of Bsal in populations with no evidence of population change so far, such as the incidental findings of dead Bsal-positive newts in fyke nets at sites 5 and 11, and the incidental findings of dead Bsal-positive fire salamanders at sites 12 and 14. Clinical signs of mycosis, such as lethargy and skin shedding (1), were observed at some Bsal-positive sites (1, 2, 7, 8, 14) but not at others.

Our study provides evidence that Bsal among wild amphibians in Europe is more widely distributed and affects a wider host range than previously known, which can either indicate recent spread of the fungus or point to historically infected sites that hitherto remained undetected. The presence of Bsal in wild populations can easily remain unnoticed because the lesions develop only near the final stage of the disease (1). This information is crucial for the design of field surveys for Bsal surveillance. Our data might be used to inform a management strategy and to implement the recommendation of the Bern Convention (11) to halt the spread of Bsal in Europe. Research to search for molecular evidence that the outbreak locations are connected is under way. Chytrid disease dynamics are affected by multiple factors (e.g., temperature regimes [1]), and yet undetermined environmental determinants might be essential for disease outbreaks (12). Untangling these factors, as well as the modes of Bsal spread and its geographic distribution, are points for further research to fully map the problem and identify populations and species at risk.