Coinfection of Chytrid Fungi in Urodeles during an Outbreak of Chytridiomycosis in Spain

Amphibian chytridiomycosis, caused by Batrachochytrium dendrobatidis (Bd) and Batrachochytrium salamandrivorans (Bsal) fungi, is responsible for the greatest loss of biodiversity due to disease ever recorded (Fisher and Garner 2020). Both pathogens probably originated in East Asia and have been introduced into naïve amphibian populations through human activities and international animal trade (Hanlon et al. 2018). Currently Bd is widespread globally, being able to infect over 1,000 amphibian species (Fisher and Garner 2020). Conversely, Bsal may be more host specific, with infection and disease more frequently reported in urodele species, whereas many European anurans appear tolerant or resistant to infection (Martel et al. 2014). Recently, Bsal has been introduced to and spreading in Western Europe, contributing to the near extirpation of Salamandra salamandra populations (Martel et al. 2013). In 2018, Bsal was also associated with a mortality event in Triturus marmoratus in Spain (Martel et al. 2020).

As Bsal continues to expand into areas where Bd is present, coinfections will be increasingly common in amphibian communities. Indeed, chytrid fungal coinfection has already been reported in S. salamandra from Germany (Lötters et al. 2018). However, the potential interactions among coinfecting chytrid fungi and the consequences for amphibian hosts are still largely unknown. Interactions among coinfecting parasites can be synergistic or antagonistic and can have important consequences on host susceptibility, transmission risk, infection duration, and clinical outcome (Pedersen and Fenton 2007). Synergism in chytrid fungi could result in increased prevalence and severity of disease, leading to further population declines and amphibian extinctions. Conversely, antagonism between chytrid fungi could be protective for susceptible species and even aid the recovery of declining populations. Therefore, analyzing the patterns of coinfection is crucial to understand disease dynamics and to design effective management strategies. We analyzed Bd-Bsal coinfections in two species of free-ranging urodeles in a focal area having an outbreak of chytridiomycosis. Our goals were to assess 1) the probability of co-occurrence of both chytrid species and 2) the correlation of the pathogen loads of the two parasites in coinfected hosts.

In March 2018, Bsal was detected in a small reservoir in Montnegre i el Corredor Natural Park in Catalonia (northeast Spain), associated with an outbreak of mortality in T. marmoratus (Martel et al. 2020). As part of the mitigation strategies, 38 native T. marmoratus and 16 introduced Ichthyosaura alpestris were culled soon after the detection of the index case (Martel et al. 2020; and see Supplementary Material). Toe-clip samples were collected and carcasses were fixed in 70% ethanol until histopathologic analysis. We extracted DNA from toe-clip samples using DNeasy blood and tissue kit (Qiagen, Germany). Quantitative PCR assays for Bd (Boyle et al. 2004) and Bsal (Blooi et al. 2013) were run in duplicate for samples and positive and negative controls. We considered samples positive when both duplicates revealed infection loads >0.1 zoospore genomic equivalents. We performed histopathologic analysis in seven well-preserved T. marmoratus and three I. alpestris to confirm the disease. Cross-sections from each individual were obtained after soaking the carcass in a quick softener solution for 120 min. Histologic examination of tissues was done using microscopic inspection of paraffin-embedded 4-µm sections stained with H&E. Because of overlap between Bd and Bsal infection patterns, we did not attempt to differentiate between the chytrids on the basis of histologic examination. Prevalences with Wilson score confidence intervals of 95% were calculated using the epiR package 2.0.19 (Stevenson et al. 2021). Pathogen prevalences and log-transformed pathogen loads were compared among species using Fisher exact tests and Student's t-tests, respectively. We also explored whether Bd and Bsal were positively, negatively, or randomly associated using the co-occur package 1.3 (Griffith et al. 2016). Finally, we assessed parasite interaction using a regression analysis of pathogen loads (zeros excluded). All statistical analyses were performed using R 4.0.3 (R Core Team 2021).

Prevalences of Bd and Bsal infections and coinfections were significantly higher in T. marmoratus than in I. alpestris (P<0.05; Table 1). Pathogen loads were not significantly different among urodele species for Bd (t=–1.0455, df=4.3849, P=0.35) but were significantly higher in T. marmoratus for Bsal (t=–7.1336, df=26.909, P<0.05; Table 2). Our co-occurrence analysis showed a random association between Bd and Bsal infections in both newt species (p_l=1, p_gt=1 for T. marmoratus, and p_lt=0.82, p_gt=0.63 for I. alpestris). In T. marmoratus, when coinfections occurred in a single host, pathogen loads were positively correlated (n=31, R²=0.16, F(1, 29)=5.36, P=0.028; Fig. 1). Correlation analysis was not performed in I. alpestris as only three individuals were coinfected. Chytridiomycosis was confirmed histopathologically in all seven T. marmoratus, with multifocal areas of epithelial necrosis associated with the presence of sporangia (Fig. 2). Different degrees of inflammatory infiltrates associated with chytrid colonization in the skin were seen in six individuals (Supplementary Material). Of interest, no inflammation was seen in the most severely affected individual (almost continuous colonization of the skin), which also had the highest pathogen loads. The most intense inflammatory changes were seen in the newt with the lowest burden. No fungal colonization or dermal inflammation was detected in the three I. alpestris.

Coinfection with Bd and Bsal has not previously been reported in these species; indeed, this is only the second detection of natural Bd-Bsal coinfections globally (Lötters et al. 2018). Given the ongoing loss of biodiversity caused by Bd, its endemic occurrence in many areas, and the emergence of Bsal, understanding and predicting the outcomes of coinfections is critical for amphibian conservation. This is of particular concern in urodeles, since many European species are highly susceptible to Bsal and coinfections may influence disease outcomes (Martel et al. 2014; Gilbert et al. 2020).

We found a positive relationship between Bd and Bsal loads in T. marmoratus, suggesting a synergistic interaction, where one parasite promotes the development of the other. Parasite synergism is often mediated by the host immune system, through pathogen-induced immunosuppression, host tissue disruption, or resource depletion (Pedersen and Fenton 2007). There is evidence that Bd induces immune evasion, immunosuppression, and immunopathology in some species (Grogan et al. 2018), and Bsal probably exhibits similar mechanisms (Farrer et al. 2017). A recent study described severely dysregulated immune responses in Bd-Bsal coinfected newts, showing that compounded immunopathology and immunosuppression could lead to more severe disease outcomes (McDonald et al. 2020). Nevertheless, few studies have evaluated the clinical course of coinfections and mixed results have been obtained thus far (Longo et al. 2019; Greener et al. 2020). We found that most infected T. marmoratus had dermal inflammatory infiltrates consistent with an adaptive immune response (Grogan et al. 2018). Although it is not possible to determine the outcome of this inflammatory reaction (clearance or immunopathology) with only seven newts, the apparent effective induced defense response warrants further immunologic studies in this species. In contrast, T. marmoratus experienced disease-induced mortality at the study site (Martel et al. 2020), suggesting that the existing immune response might not be entirely effective against chytrid fungi.

We also showed that Bd and Bsal infections co-occur randomly in the species of urodeles studied. A random association between coinfecting chytrid fungi may be observed when constitutive defenses against first invasion remain unaltered. Despite both parasites depending on the same host resources and niche (i.e., epidermal cells), none of our analyses showed antagonistic interactions among chytrid fungi. These results suggest that resource limitation is not a constraint for fungal invasion and growth. Differences in chytrid fungus prevalence, load, and pathology between T. marmoratus and I. alpestris probably reflect differences in species susceptibility as demonstrated experimentally (Martel et al. 2014; Fernández-Beaskoetxea et al. 2016). The lower pathogen prevalences, reduced Bsal loads, and the lack of histologic lesions in I. alpestris are indicative of increased resistance and tolerance to chytrid fungus infection. Indeed, this species is considered a potential reservoir for Bsal in European amphibian communities (Gilbert et al. 2020).

In conclusion, our study has shown that co-occurring Bd and Bsal elicit synergistic effects, intensifying pathogen load in at least one urodele species. Chytrid fungi coinfections could, therefore, increase disease severity and have important consequences for host survival and conservation in susceptible amphibian species. A limitation of our study is the analysis of natural infections because it was not possible to determine the timing and sequence of the coinfections. Experimental infections under controlled conditions are needed to reinforce our findings. Further research on parasite interactions, immune responses, and clinical outcomes of chytrid fungi coinfections will be crucial to predict and prevent future biodiversity losses.

M.P.R. was funded through the 2021 FI Scholarship, Departament de Recerca i Universitats, Generalitat de Catalunya, Spain (FI_B 00171). E.S. was funded by the Spanish Ministerio de Economia y Competitividad through a Ramon y Cajal agreement (RYC-2016-21120). We thank the personnel from the Catalonia Reptile and Amphibian Rescue Center, Montnegre i el Corredor Natural Park (Diputació de Barcelona), and Generalitat de Catalunya for permitting the collection and analysis of the samples.

Supplementary material for this article is online at http://dx.doi.org/10.7589/JWD-D-21-00170.