In underground hibernacula temperate northern hemisphere bats are exposed to Pseudogymnoascus destructans, the fungal agent of white-nose syndrome. While pathological and epidemiological data suggest that Palearctic bats tolerate this infection, we lack knowledge about bat health under pathogen pressure. Here we report blood profiles, along with body mass index (BMI), infection intensity and hibernation temperature, in greater mouse-eared bats (Myotis myotis). We sampled three European hibernacula that differ in geomorphology and microclimatic conditions. Skin lesion counts differed between contralateral wings of a bat, suggesting variable exposure to the fungus. Analysis of blood parameters suggests a threshold of ca. 300 skin lesions on both wings, combined with poor hibernation conditions, may distinguish healthy bats from those with homeostatic disruption. Physiological effects manifested as mild metabolic acidosis, decreased glucose and peripheral blood eosinophilia which were strongly locality-dependent. Hibernating bats displaying blood homeostasis disruption had 2 °C lower body surface temperatures. A shallow BMI loss slope with increasing pathogen load suggested a high degree of infection tolerance. European greater mouse-eared bats generally survive P. destructans invasion, despite some health deterioration at higher infection intensities (dependant on hibernation conditions). Conservation measures should minimise additional stressors to conserve constrained body reserves of bats during hibernation.
Life history theory suggests that organisms optimise their defences against pathogens by differential allocation of resources to support different physiological functions1,2,3,4. There is a trade-off mechanism applied to modulate investment into individual life history components5,6. Homeostasis and survival of hosts challenged with exposure to a pathogenic agent could be considered a physiological measure of health2. Host physiological status under pathogen pressure will be impacted by standard components of the ‘disease triangle’, i.e., host susceptibility, virulence of the infectious agent and environmental determinants. In general, hibernation, a slow life history strategy, is linked to higher survival rates7. Successful hibernation of mammals is constrained by their energy reserves, suitable microhabitat availability8 and thermoregulatory behaviour9. Additional stressors may deplete the animal’s resources, resulting in adverse consequences.
Skin, the largest organ of the body, acts as a barrier between the animal and its environment while providing multiple anatomic and physiological functions. A bat’s membranes are essential for flight, increasing the ratio of body mass to body surface. Moreover, naked flight membranes have a surface area eight times greater than that of fur-coated skin. This increases the area of potential exposure to dermatopathogens. In bats, healthy skin is essential for maintaining physiological homeostasis10.
In the northern temperate zone, hibernating bats are exposed to a non-systemic fungal infection that mainly affects the areas of skin without fur. Over the last decade, the fungus Pseudogymnoascus destructans has caused a devastating decline in North American bat populations11,12,13,14,15,16. During this time, there have been only sporadic cases of mortality in Eurasia17,18,19,20,21. In contrast to standard cutaneous dermatomycoses, the so-called white-nose syndrome (WNS) fungus invades living layers of skin10,22,23.
Despite considerable advances in our understanding of molecular pathogenesis and factors affecting the virulence of P. destructans infection24,25,26,27, the fundamental pathophysiological mechanisms of mortality associated with WNS remain unconfirmed10,28,29. Adverse effects increase with the extent of wing membrane pathology. While early stages of skin infection induce a two-fold increase in fat energy utilisation29, late-stage infected bats have altered torpor-arousal cycles, abnormal hibernation behaviour as well as emaciation and increased mortality30. Several studies of clinical blood parameters (e.g. electrolytes, acid-base balance, hydration status, haematology) in little brown bats, Myotis lucifugus, reveal that WNS disrupts blood homeostasis29,31,32. Interestingly, European P. destructans isolates are virulent and produce WNS in this North American bat species33. Infected bats have histopathology identical to skin lesions in Palearctic bat species18,19,20,21,22. P. destructans occurred in Europe before the outbreak of the Nearctic epidemic34,35. This, together with phylogenetic studies36,37, indicates that North American species of bats might be naïve hosts to a fungal pathogen originating in the Palearctic region. Intercontinental and interspecies comparisons may provide greater insights into variation of host responses to fungal infection.
Two defence mechanisms can evolve from host-pathogen interactions: resistance and tolerance38,39,40. Resistance protects the host by reducing the pathogen burden. As a consequence, prevalence of the agent in the host population decreases. In comparison, tolerant hosts limit the damage caused by the pathogen and remain healthy without mounting sterilising immunity41, though prevalence remains high or even increases within the susceptible population. Host cost/benefit trade-offs from its response to infection should favour tolerance when disease severity allows survival and host adaptation42. As European bats infected with P. destructans display no population-level effects, they are thought to tolerate infection, despite high fungal loads and almost 100% prevalence17,20,21. Host tolerance and/or disease resistance can be measured as a regression slope between health and pathogen load38,42.
Here, we report on host-pathogen interactions in the greater mouse-eared bat (Myotis myotis), the European species showing highest skin infection intensity, based on blood parameters. We hypothesise that hibernating European bats are unable to maintain blood parameters within the normal physiological ranges found in healthy bats when exposed to the WNS fungal agent, and that blood homeostasis disruption could be related to infection intensity and hibernation temperature. We predict 1) a skin lesion threshold distinguishing healthy and diseased bats, 2) blood acidosis and a decrease of blood glucose in bats with high P. destructans infection intensity, and 3) a lower body mass index (BMI) in bats with blood homeostasis disruption. We also predict a reduced rate of BMI loss with increasing infection intensity, indicative of disease tolerance. An improved understanding of how hibernating bats optimise their health to survive pathogenic pressure will have positive ramifications for wildlife and conservation medicine.
Each bat was handled in such a way as to minimise sampling distress and was released at the hibernaculum one hour after capture. Fieldwork and bat sampling was performed in accordance with Czech Law No. 114/1992 on Nature and Landscape Protection, based on permits 1662/MK/2012S/00775/MK/2012, 866/JS/2012 and 00356/KK/2008/AOPK issued by the Agency for Nature Conservation and Landscape Protection of the Czech Republic. Experimental procedures were approved by the Ethical Committee of the Czech Academy of Sciences (No. 169/2011). Sampling at the Nietoperek Natura 2000 site (Poland) was approved by the II Local Ethical Commission in Wrocław (No. 45/2015) and the Regional Nature Conservancy Management in Gorzów Wielkopolski (WPN-I-6205.10.2015.AI and WPN-I-6205.20.2016.AI). The authors were authorised to handle free-living bats under Czech Certificate of Competency No. CZ01341 (§17, Act No. 246/1992) and Polish Certificate of Competency in Experimental Procedures on Animals (Polish Laboratory Animal Science Association, Certificate No. 2413/2015).
Seventy-nine bats were sampled at three important European hibernacula, the Nietoperek bunker (NIE; Poland), the Šimon and Juda mines (SJM; Czech Republic) and the Sloupsko-Šošůvské caves (SSC; Czech Republic), during the late hibernation period in 2015. Control sampling was undertaken at the Nietoperek bunker during March 2016 to evaluate infection dynamics. All three localities differ in geomorphology and microclimate conditions. No mass mortalities have been reported from any of the sites43,44,45,46 and numbers of hibernating M. myotis have remained stable, or have increased slightly over recent years.
The Nietoperek bat reserve lies in the underground corridors of an abandoned German military fortification from the central sector of the Międzyrzecz Fortified Front in western Poland (52°25′N, 15°32′E). The aboveground bunkers are connected by 3–4.5 m high and 2.5–4 m wide underground railway corridors. Sites preferred by hibernating M. myotis have a median temperature of 8.7 °C (min-max 6.1–9.9 °C), 100% relative humidity (min-max 77.5–100.0%) and 9 g/m3 absolute humidity (min-max 6–9 g/m3)46.
The Šimon and Juda mines comprise two gallery systems, the entrances of which open into a 10 m deep iron ore quarry. The mines were closed in 1870 and most galleries were flooded after World War I. While the galleries were drained between 1956 and 1957, no more mining took place. The lower gallery system has four horizontal storeys, each 2–2.5 m high, the upper system comprising an irregular labyrinth of galleries and chambers. Differences in geomorphology mean that each system has a different microclimate, the lower being colder, with temperatures rarely exceeding 5 °C (relative humidity close to 100%; absolute humidity 7–8 g/m3) and the upper having temperatures around 7 °C in most parts, though dropping close to the main entrance47.
The Sloupsko-Šošůvské caves comprise a natural karst system with 7 km of chasms, domes and corridors. The 8 m high and 20 m wide main entrance is located in the northern part of the cave. Due to their complicated geomorphology, microclimatic conditions vary widely. Hibernating bats mainly use those parts close to the main entrance (e.g. the Nicová and Eliška cave45) with mean annual temperatures fluctuating between 5.5 and 7.5 °C and an absolute humidity of 7–8 g/m3 48.
Measurements of bat health
The body surface temperature of each hibernating bat was measured using a Ryatek contactless laser thermometer (Total Temperature Instrumentation Inc.) prior to its removal from the hibernaculum wall. Each bat was then sexed and its age estimated based on epiphyseal ossification of the thoracic limb fingers and tooth abrasion49. Callipers were used to measure forearm length and body mass was determined using a portable top-loading balance. BMI was calculated as body mass (g) divided by left forearm length (mm)50.
After a re-warming period of 60 minutes, the skin was disinfected with alcohol and a blood sample (100 µl) taken from the uropatagial vessel using a heparinised tube51. An i-STAT portable clinical analyser (EC8+ diagnostic cartridge, Abaxis, Union City, CA, USA) was used to measure sodium (Na, mmol/L), potassium (K, mmol/L), chloride (Cl, mmol/L), total dissolved carbon dioxide (tCO2, mmol/L), blood urea nitrogen (BUN, mmol/L), glucose (GLU, mmol/L), haematocrit (Hct, L/L), pH, partial dissolved carbon dioxide (pCO2, kPa), bicarbonate (HCO3, mmol/L), base excess (BE, mmol/L), anion gap (AnGap, mmol/L) and haemoglobin (Hb, g/L).
A subsample was used to prepare a blood smear, which was then treated with Romanowsky stain. Differential white blood cell counts were determined by counting 100 leukocytes under oil immersion magnification and calculating the relative number of lymphocytes, monocytes, neutrophils, basophils and eosinophils.
Measurement of infection intensity
Immediately following capture, the surface of the left wing was swabbed (FLOQ Swabs, Copan Flock Technologies srl, Brescia, Italy) in a standardised manner to collect fungal biomass. Fungal load was calculated using quantitative polymerase chain reaction (qPCR) and the QIAamp DNA Mini Kit (Qiagen, Halden, Germany) was used to isolate fungal DNA from the wing swabs. A dual-probe TaqMan (Life Technologies, Foster City, CA, USA) was used to quantify P. destructans DNA (ng per left wing area; triplicate samples) using a previously described protocol employing positive and negative controls and a dilution series calibration curve from a positive control19,21,52,53. Suspected fungal growths from other parts of the body (e.g. ears, muzzle) were collected for laboratory culture examination12. Skin lesions were enumerated by photographing both wings over a 368 nm ultra-violet (UV) lamp19,21,54. A 4 mm punch biopsy, centred over the lesion, was collected from each bat to confirm P. destructans infection on histopathology22,23.
Normality of variable distribution was tested using the Shapiro-Wilk test. Non-normally distributed variables were log transformed and rechecked. All parameters were normally distributed after transformation, with the exception of body surface temperature (W = 0.936, p = 0.005), haematocrit (W = 0.896, p < 0.001), haemoglobin (W = 0.896, p < 0.001) and percentage of eosinophils, monophils and basophils (Shapiro-Wilk tests, p < 0.001). In these cases, statistical analysis was conducted using non-parametric tests, i.e. the Kruskal-Wallis test, the Mann-Whitney U test and Spearman’s correlation. The slopes and intercepts of linear regressions were compared using the Student’s t-test.
Haematological parameters did not differ between age classes (adult vs. sub-adult) or sexes (ANOVA and t-test); hence, the data were pooled for subsequent analyses. We scored the level of wing damage based on the total number of UV-fluorescing skin lesions on both wings as: 1 = 0 to 50 lesions, 2 = 51 to 250 lesions, 3 = 251 to 500 lesions, 4 = 501 to 1000 lesions and 5 = more than 1000 lesions. Effect of locality and wing-lesion score on blood parameters was tested using general linear mixed models (GLMM) with locality (hibernaculum) set as a random effect. As blood parameters were highly inter-correlated, we used principal component analysis (PCA) to evaluate inter-individual differences in distribution along axes linked with severity of skin infection, i.e. number of UV-fluorescing skin lesions.
Data on fungal load, number of UV-fluorescing skin lesions and blood parameters from Nietoperek (Poland), the Šimon and Juda mines (Czech Republic) and the Sloupsko-šošůvské caves (Czech Republic) are presented in Supplementary Table S1.
Relationship between number of skin lesions, body surface temperature, BMI and P. destructans load
The number of skin lesions fluctuated between 0 and 3782. One wing for a given individual always had significantly more lesions (Wilcoxon Matched Pairs Test; Z = 2.497, p = 0.013; Fig. 1), though the number of lesions on each wing was correlated (Spearman rank order correlation rs = 0.857, p < 0.05). Consequently, we used the sum of UV-fluorescing lesions from both wings as it more precisely expressed total skin infection severity. Number of lesions was positively correlated with P. destructans load and negatively with body surface temperature (Table 1). The decline in BMI with higher P. destructans load was the same at all localities, though the regression curve for the Sloupsko-Šošůvské caves did not differ significantly at its lower intercept (SSC vs. NIE t = −0.303, DF = 53, p = 0.382; SSC vs. SJM t = −0.141, DF = 32, p = 0.445) due to the lowered BMI of hibernating bats (Figs 2 and 3). When material from all localities and both years of sampling was pooled, the decline in BMI with increasing P. destructans load and number of skin lesions became statistically significant (P. destructans load F(74) = −2.955, p = 0.004; UV-fluorescing skin lesions F(75) = −2.461, p = 0.016).
GLMM confirmed both the differences in P. destructans load (F = 8.106, p < 0.001) and BMI (F = 7.785, p < 0.001). While post-hoc univariate tests indicated locality as the main effect for P. destructans load; BMI was significantly influenced by both locality and wing-lesion score (Fig. 3). Highest BMI scores were recorded at lowest infection severity (score 1). Kruskal-Wallis tests confirmed a difference in body surface temperature for both locality (H(2, N = 57) = 49.826, p < 0.001) and wing-lesion score (H(4, N = 57) = 16.773, p = 0.002), hibernating bats with lowest body surface temperatures showing increased infection severity (Fig. 4).
Relationship of skin infection level to blood chemistry and haematology profile
Nine blood parameters were significantly affected by locality and wing-lesion score (Tables 2 and 3), the GLMM model explaining between 14.3 and 37.7% of variability. We used seven continuous blood parameters (excluding percentage of neutrophils and lymphocytes) for PCA analysis of samples from 2015, the first three components of which explained 92.5% of total variability. Acid-base variables (tCO2, pH, HCO3 and BE) displayed a strong negative correlation with the first component, electrolytes (Na and Cl) correlated positively with the second component and glucose negatively with the third component. The space defined by the first and third components provided the best separation between individuals (Fig. 5). Highest principal component values were observed in a single healthy individual (without UV-fluorescing skin lesions), its position subsequently being considered a new midpoint for the principal component axes. All individuals (n = 18) located in the upper right space were diagnosed with homeostasis disruption associated with skin infection, the three worst cases (top right position in Fig. 5) displaying infection intensities of 2394, 876 and 1490 lesions. As all bats from Nietoperek proved healthy, we took control samples during winter 2016 and repeated the PCA analysis with a larger sample size. While there was no difference in BMI and skin infection intensity in 2015 and 2016 (T-test; t = −1.598, p = 0.118, and t = 1.154, p = 0.256); P. destructans load was higher in 2015 than 2016 (T-test; t = 3.176, p = 0.003). PCA added seven new cases to those diagnosed with homeostasis disruption, including two from Nietoperek. Average number of skin lesions, P. destructans load, BMI and body surface temperature differed significantly between healthy and diseased bats (Table 4). We defined a theoretical breaking point (skin lesion threshold) for the manifestation of skin infection in disruption of blood parameters, based on the median between the 95% confidence intervals of the two groups (i.e. 328.5 skin lesions; Fig. 6). Six blood parameters differed between the groups defined by PCA; however, only glucose, skin lesion number, log P. destructans load and body surface temperature differed between the groups defined by the skin lesion threshold (Table 4).
The continent-wide colonisation of Palearctic underground hibernacula and ongoing spread of the fungal pathogen in North America makes exposure of bats hibernating in the temperate zone of the northern hemisphere to P. destructans highly probable11,16,17,20,21,55,56,57,58. The wide distribution of both the host and its pathogen results in spatial and temporal variation in host-pathogen population interactions, allowing performance studies into host health and pathogen virulence under differing environmental conditions. Here we show that even early-stage fungal damage of bat wing membranes may negatively impact physiological status, dependent on infection intensity and hibernation conditions. Our data suggest that the pattern of disease impact can vary between sites.
We chose European M. myotis as a model species to examine naturally occurring host-pathogen interactions with P. destructans as they have a higher survival capacity than the North American little brown bat (M. lucifugus) and big brown bat (Eptesicus fuscus)59. Nevertheless, sporadic mortalities associated with P. destructans infection have been reported for M. myotis, documenting that infection intensity can range from mild to severe18. Recently, a model taking account of temperature, humidity-dependent fungal growth and bat energetics during hibernation was devised, which predicted that the likelihood of surviving P. destructans infection increases with increasing body size and drier and/or colder hibernation sites59. While M. myotis is one of the largest European bat species, highest infection intensity has been found in those hibernating at low hibernation body temperatures, contrary to our prediction. Three hypotheses may explain this paradox. First, bats select lower hibernation temperatures as an adaptation to conserve energy when dealing with high infection intensity. Second, bats selecting low hibernation temperatures develop increased infection intensities as a consequence of a reduced ability to up-regulate immune functions and clear the infection24,60,61. Third, under conditions of natural infection, P. destructans growth and virulence is stronger in those bats hibernating at low temperatures, despite laboratory studies suggesting a temperature optimum of between 12.5 and 15.8 °C62. Unfortunately, we lack data on hibernation temperature history, arousal frequency, infection dose and duration of infection for each bat, which would allow these hypotheses to be tested explicitly. In the present study, bats were only sampled once at the end of the hibernation period. Interestingly, skin lesion number differed between the left and right wings, suggesting differing exposure to infection and uneven spread of the fungal agent across the body surface during cleaning when aroused. A field experiment analysing fungal load dynamics in relation to wing membrane pathology and bat body temperature during hibernation is needed to provide greater insight into such host-pathogen interactions under natural conditions.
BMI loss in diseased bats indicative of disease tolerance
Our data provide further evidence for tolerance of Palearctic bat species to the P. destructans fungus21. The regression lines obtained by plotting host BMI against pathogen burden (skin fungal load) at each hibernaculum display a shallow BMI loss slope as infection intensity increases (Fig. 2). Owing to the differences in origin of the hibernacula used in this study, host-pathogen population interactions will have undergone different evolutionary histories, ranging from tens, to hundreds or even thousands of years. Cave-hibernating bats, with their distinctly lower BMI (Fig. 2), showed lowest tolerance capacity to infection. Transition to euthermia in the early post-hibernation period allows bats to mount an effective immune response against the fungal pathogen and clear any skin infection19,63. Later, the bat’s defence strategy turns from tolerance to resistance42. Very little is known about the health of P. destructans infected hosts in the period following emergence from hibernation. The dichotomy of disease outcome in euthermic bats results in either healing63 or mortality due to immunopathology64. In M. myotis, healing may occur within two weeks, during which the diagnostic UV-fluorescence disappears and a scab develops over the previously infected skin19. The costs of neutrophilic inflammation and wing membrane tissue remodelling are hard to estimate. Likewise, we lack detailed quantification of physiological costs associated with flight performance and changes in foraging efficiency in bats recovering from P. destructans infection63,65,66,67. Upon arousal, early euthermic females may also face a trade-off between mounting an immune response and the energetic investment needed to initiate gestation68,69. Higher cortisol levels, indicative of chronic stress, have been recorded in bats surviving exposure to P. destructans, and this may have adverse effects on reproductive success70. Further, North American species recovering from P. destructans infection have shown shifts in pregnancy and lactation, suggestive of reproductive fitness consequences71. Similar studies on reproductive fitness consequences have yet to be performed on European bats facing fungal pathogen pressure.
Alterations in blood homeostasis in diseased bats
While infectious diseases are commonly thought to induce biochemical responses that differ between species72, our data showed hibernation site-specific differences. Haematology and blood chemistry reflect body and tissue status. A range of mechanisms maintain blood parameters within a narrow range; blood pH, for example, being maintained through respiratory system and kidney function. Hibernation, on the other hand, represents a specific physiological state resulting in changes to metabolic and biochemical pathways73. Heart rate, cardiac output and respiration are greatly reduced during deep hibernation, and these changes lead to a drop in pH and marked acidosis.
Studies of blood homeostasis in M. lucifugus indicate a pattern of changes dependent on WNS intensity29,31,32,74. While such studies have sampled blood by decapitation, we used non-lethal vessel puncturing to study M. myotis, a strictly protected European bat. Cryan et al.31, using data on P. destructans infection in both captive and wild hibernating bats, noted that electrolyte depletion increased with increasing wing damage severity. Furthermore, measurements of urine-specific gravity suggested that bats underwent hypotonic dehydration. In a second study, captive hibernation of M. lucifugus following experimental inoculation with P. destructans complicated blood sampling, allowing analysis of only eight infected bats32. The addition of data from two follow-up captive inoculation experiments, however, showed no difference between the infected and control groups68. While data obtained by Warnecke et al.32 were only suggestive of metabolic acidosis, Verant et al.29 observed chronic respiratory acidosis with metabolic compensation in bats at an early stage of the disease.
A skin lesion threshold distinguishing healthy and diseased bats
As all bats in our study were naturally infected in their hibernacula and confirmed positive for P. destructans (with the exception of one individual from Nietoperek), it was not possible to compare host physiological responses to the fungus against a non-infected control group. Nevertheless, our non-diseased and diseased groups, as defined by PCA, differed in blood pH, tCO2, bicarbonate, base excess/deficit and anion gap. These acid-base parameters shifted to mild metabolic acidosis in the diseased group (Table 4). The diseased group displayed higher infection intensity, distinguished by both fungal load and UV-fluorescing skin lesions. Bats defined as diseased (i.e. with blood parameters showing homeostasis disruption) had a hibernating body surface temperature around 2 °C lower than non-diseased individuals. The diseased group also displayed significantly decreased glucose concentrations and BMI. Contrary to blood biochemistry results for M. lucifugus, we observed no differences in electrolytes between diseased and non-diseased M. myotis, suggesting that the acid-base disruption was due to increased energy utilisation associated with infection. The increase in differential neutrophil count was non-significant, probably because the white blood cells migrated from blood to the infected sites19. Interestingly, significant eosinophilia was observed in diseased M. myotis. Peripheral blood eosinophilia is commonly associated with chronic, parasitic and fungal infections75. As eosinophilia is also associated with hypersensitivity reactions, however, our findings may support the hypothesis that immunopathology plays a role in post-emergent WNS mortality64.
While exposure of bats to multiple natural and/or anthropogenic stressors is a realistic environmental scenario76,77,78,79, sub-lethal adverse effects are mostly underreported. Disease pathogenesis and the action of multiple stressors during hibernation are not yet fully understood; however, different stressors may well combine to exert synergistic effects80. Importantly, disturbance by human activities, such as tourism, caving or research, could also threaten hibernating bats by increasing energy expenditure81.
Following the emergence of WNS and recognition of its impact on North American bat populations in 2006, chiropterologists concerned with European bat conservation have asked one essential question: are Palearctic bat populations and communities threatened by this fungal disease? Up to now, there have been no functional studies addressing host-pathogen interactions in relation to WNS. However, there is mounting evidence for virulent skin invasion and pathognomonic lesions in many hibernating Eurasian bat species. As these findings have not been associated with mass mortalities and/or population declines, research should be directed toward examining health consequences in terms of trade-off mechanisms modulating investment into host response to infection.
In this study, we were able to show variation in fungal pathogen pressure in relation to hibernaculum-dependent physiological effects of P. destructans infection. We conclude that European M. myotis survive P. destructans invasion, despite showing deterioration in health, with infection intensity dependent on hibernation conditions. Disruption in blood homeostasis was observed in bats, even with a low threshold number of skin lesions on both wings. We argue that overwintering in underground hibernacula colonised by this virulent pathogen is associated with health-related costs for European bats. Further research should aim to quantify levels of homeostasis disruption in terms of constrained energy reserves and compatibility for survival.
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This study was supported through Czech Science Foundation Grant No. 17-20286S and Grant No. 221/2016/FVHE of the Internal Grant Agency of the University of Veterinary and Pharmaceutical Sciences Brno. We thank Natália Martínková for measuring the fungal load of infected bats and for valuable comments on the original manuscript. We are grateful to Dr. Kevin Roche for correcting and improving the English text.
Electronic supplementary material
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.