https://orcid.org/0000-0002-0099-7934
Figures
Abstract
Infectious diseases are considered major threats to biodiversity, however strategies to mitigate their impacts in the natural world are scarce and largely unsuccessful. Chytridiomycosis is responsible for the decline of hundreds of amphibian species worldwide, but an effective disease management strategy that could be applied across natural habitats is still lacking. In general amphibian larvae can be easily captured, offering opportunities to ascertain the impact of altering the abundance of hosts, considered to be a key parameter affecting the severity of the disease. Here, we report the results of two experiments to investigate how altering host abundance affects infection intensity in amphibian populations of a montane area of Central Spain suffering from lethal amphibian chytridiomycosis. Our laboratory-based experiment supported the conclusion that varying density had a significant effect on infection intensity when salamander larvae were housed at low densities. Our field experiment showed that reducing the abundance of salamander larvae in the field also had a significant, but weak, impact on infection the following year, but only when removals were extreme. While this suggests adjusting host abundance as a mitigation strategy to reduce infection intensity could be useful, our evidence suggests only heavy culling efforts will succeed, which may run contrary to objectives for conservation.
Citation: Bosch J, Carrascal LM, Manica A, Garner TWJ (2020) Significant reductions of host abundance weakly impact infection intensity of Batrachochytrium dendrobatidis. PLoS ONE 15(11): e0242913. https://doi.org/10.1371/journal.pone.0242913
Editor: Emmanuel Serrano, Universitat Autonoma de Barcelona, SPAIN
Received: March 17, 2020; Accepted: November 10, 2020; Published: November 30, 2020
Copyright: © 2020 Bosch et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting information files.
Funding: This study has been supported by Fundación BBVA and Organismo Autónomo Parques Nacionales (Spain, project ref. 2399/2017; J.B.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The authors did not receive a salary from any of our funders.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Managing infectious diseases is one of the great challenges of the 21st Century. A growing array of pathogens impair human health, threaten food security and pose a risk to biodiversity conservation, however our efforts to control infections circulating in wildlife populations are largely unsuccessful (but see [1–5]). Developing and testing strategies for controlling infectious agents should therefore be a research priority, yet surprisingly few real-world trials of methods for controlling infectious disease affecting wildlife have been undertaken and subsequently published [1–4,6–9]. Among others, culling of hosts from a susceptible population is a fundamental approach in density-dependent disease mitigation based on the theory of disease ecology and epidemiology [10–15]. Note, however, that while the term culling often denotes euthanizing, it can simply be the removal of animals from the system as in this study (see below).
Most studies for controlling infectious agents are performed in humans, livestock or companion animals, whereas infections strictly affecting wildlife are largely overlooked, despite overwhelming evidence that wildlife diseases are behind a significant proportion of recent and ongoing biodiversity loss [16]. One exception is the case of the chytridiomycosis, a disease responsible for the decline of hundreds of amphibian species, with new reports accruing [17]. An enormous body of literature bears witness to the research that has revealed the ecological and evolutionary drivers of disease [18,19], but only a relatively sparse publication list provides recommendations on how to respond to the threat chytrid fungi pose to amphibian biodiversity [20–28]. Even more sparse is the collection of published studies reporting the outcomes of attempts at mitigating infections and disease in nature, none of which presents a scalable and transferable solution [5,29–31]. This is unfortunate: although a small proportion of species affected by chytridiomycosis appear to be recovering without intervention, most of them continue to decline [17]. Without methods to combat the disease in situ, it seems clear that the ‘chytrid crisis’ will continue to impair global amphibian biodiversity.
We have argued elsewhere that research should focus on approaches that are transferrable across host species and communities, easily implemented by nonspecialists and amenable to combination with other techniques [22]. The list of possible actions is short, at least in part due to the fact that techniques commonly used for controlling disease in mammals and birds, like vaccines, simply cannot be applied to most amphibian communities due to their typically large population sizes. Many species of amphibians can be easily captured or trapped, though, even when local abundance may be in the tens of thousands, which offers opportunities to ascertain the impact of altering the abundance of hosts. Several studies have argued for a potential, but as yet uncertain, role of host abundance on the prevalence and amplification of infection loads [32–34]. The latter is a key parameter affecting the severity of chytridiomycosis, as there is a strong, positive relationship between the infection intensity of Batrachochytrium dendrobatidis (Bd), one of the two pathogens causing chytridiomycosis, and the likelihood of dying from the disease [35,36].
Chytridiomycosis was first described in Europe nearly 20 years ago in the Sierra de Guadarrama mountains of Spain [37] where this study took place. At this location, Alytes obstetricans (midwife toads) were extirpated, and two other amphibian species were reported dying from chytridiomycosis and potentially declining as a result: Salamandra salamandra (fire salamanders) and Bufo spinosus (spiny common toads [38–40]). Alytes obstetricans is highly susceptible to Bd infection and was historically the most abundant species and primary source of chytridiomycosis [37,38]. Salamandra salamandra has also historically experienced lethal chytridiomycosis and sharp declines when A. obstetricans was abundant in the area, even though it is not considered to be as highly susceptible to infection and disease [39]. Larvae of both species can overwinter and are thought to act as reservoirs transmitting infections to incoming young of the year larvae [41]. The extirpation of A. obstetricans from the area means that overwintering S. salamandra larvae are considered the primary source of infection in the aquatic environment [39,42].
Because of their purported role in maintaining infections, we chose to investigate how altering host abundance of S. salamandra larvae in the most important amphibian breeding ponds in the Sierra de Guadarrama mountains affects infection intensity the following breeding season, under the assumption that disease risk is reduced at lower infection loads. Pilot experiments to establish drug concentrations to clear Bd infections of S. salamandra larvae indicated that most individuals lost infection when housed individually. Thus, our hypothesis was that, at least for fire salamander larvae that play a key role in intraspecific transmission but are weakly infected, significant alterations in host density could reduce strength of infections. We tested this through a large-scale, 2-year larval removal field trial in the permanent ponds where overwintering occurs. Additionally, we used an ex situ approach to examine how differing densities of larvae affected the intensity of infection in groups of S. salamandra and A. obstetricans larvae over time scales similar to those where infection dynamics change in the wild [41].
Materials and methods
Study area
The field experiment and the collection of salamander larvae for the lab experiment took place in the protected area of Guadarrama Mountains National Park, central Spain (408500 N, 38570 W; Fig 1). The core of the park, the Peñalara Massif, is a wetland system of about 800 ha composed of several small glacial valleys located 1800–2430 m a.s.l. The area holds around 250 fishless ponds, mostly temporary, and supports nine species of amphibian. The landscape of this alpine area consists mainly of granitic outcrops, alpine meadows, heathlands dominated by Cytissus oromediterraneus and Juniperus communis nana, and forests of Pinus sylvestris below the timber line. Local climate is characterized by very cold winters, mild summers and abundant precipitation, with snow falls usually from the end of November to the middle of May. Batrachochytrium dendrobatidis has been present in the area for at least 20 years [37], whereas neither B. salamandrivorans nor Ranavirus have not been found (Bosch, unpublished results).
The field study took place from Summer 2015 to Spring 2018 in the four permanent ponds that host the largest S. salamandra populations: Laguna Grande (elevation 2018 m, surface 6766 m2, maximum depth 2.7 m), Charca Larga (elevation 2111 m, surface 89 m2, maximum depth 0.6 m), Charca Mariposa (elevation 2136 m, surface 660 m2, maximum depth 0.7 m) and Laguna de los Pájaros (elevation 2170 m, surface 5175 m2, maximum depth 1.3 m). These ponds freeze in December and thaw in March, and they are separated by at least 0.5 km. No overwintering larvae of other amphibian species are present and only B. spinosus is relatively abundant in those ponds.
Ex situ assessment of the impact of host density on infection intensity
Naturally Bd-infected overwintering larvae of Alytes obstetricans and S. salamandra at the same stage of development were collected at the end of the winter, respectively, in the Teruel province (Aragón, Northeastern Spain) and in Laguna Grande (Guadarrama Mountains National Park). For both species, prevalence of Bd infection in larval stages is known to approach 100% during colder months at both localities [41,43]. The experiment took place in bio-secure conditions at the laboratories of the Research and Management Center ‘Puente del Perdón’, a facility of the Guadarrama Mountains National Park located a few km away of the Peñalara Massif, and where head-starting programs for A. obstetricans and Rana iberica are hosted. The larvae we used were returned to their point of capture at the end of the ex situ experiment. Initial Bd infections were confirmed following the qPCR protocol of Boyle et al. [44]. We housed indoors groups of larvae of each species in separate 5 L plastic tanks at three different densities (one, five and 15 individual each) on a 12:12-h light schedule. The densities ranged from the absolute minimum (one) up to the maximum amount of salamander larvae we could house in a single container without having them injuring each other, and while keeping densities consistent across species. Density treatments were replicated respectively seven, six and four times for S. salamandra and eight, six and four times for A. obstetricans (for densities of one, five and 15 larvae, respectively). We fed animals ad libitum with tubifex worms (salamanders) or commercial tadpole food (midwife toads), and renewed half of the water in each tub weekly. Water temperature was recorded throughout every half hour with a HOBO Water Temperature Pro v2 datalogger, and varied from 0.8 to 9.8°C. We used average water temperature during the three days prior to Bd sampling for analysis, following Fernández-Beaskoetxea et al. [43], who described short-term impacts of water-temperature on Bd load. We tracked changes in infections in one animal per tank, repeatedly sampling the same animal once at the start of the experiment and every month afterwards, for a total of five samples per animal and tank. Focal A. obstetricans tadpoles were identified using visual implant elastomers and swabbed on their keratinized mouthparts, while focal S. salamandra were identified by unique dorsal spot patterns and swabbed across the entire body surface. We quantified load of Bd infections (referred to as load hereafter) following Boyle et al. [44] by using cotton swabs (MW100 rayon tipped dry swabs from MWE Medical Wire). Swabs samples were kept at 4°C until analyses and the DNA was extracted with Prep-Man Ultra and amplified using a BIO-RAD CFX96 Real-Time PCR Detection System. Each 96-well assay plate included a negative control and four different standards containing DNA from 100, 10, 1 and 0.1 Bd zoospore equivalents (ZE). We tested all the samples (diluted 1/10), as well as the negative control and the standards, in duplicate. We considered samples with greater than 0.1 ZE in both replicates, and the expected sigmoidal shaped amplification curve, positive for Bd.
We used a general linear mixed models to analyze the within-subjects variation in Bd load. Species was treated as a categorical fixed factor, the “sampling events” (date of sampling) and the density treatment as ordered factors, and water temperature during the three days prior to sampling as a covariate in the mixed models, with tadpole identity as a random effect. Three different mixed random intercept models were built: (a) null model without any fixed effect, (b) main fixed effects model, and (c) full model containing the main effects and their interactions. Second-order Akaike’s AICc for finite sample sizes were used to compare these three models, using maximum likelihood (ML) estimation; the model with the lowest AICc figure was finally obtained by means of restricted maximum likelihood (REML) estimation, because it renders unbiased variance covariance parameters. Bd load (including zero values) was log10 transformed (y’ = log10[y+1]) prior to data analyses. The mean square (MS) and degrees of freedom (df) of the error terms were estimated following the Kenward-Roger method [45]. Homoscedasticity and normality of residuals of the final general linear mixed model were visually checked and did not show considerable deviations from the model assumptions. All analyses were carried out using R version 3.1.2 (R Core Team 2014) and the packages lme4 [46], lmerTest [47], lmtest [48], pbkrtest [49], car [50], MuMIn [51] and phia [52]. The proportion of variance explained by the model was obtained using the r.squaredGLMM function of the MuMIn package [53], while the estimations for different predictors were obtained using the SS (sums of squares) provided by the function anova.lmer of the lmerTest package (Partial-eta2 according to the ratio SSeffect/[SSeffect+SSerror]).
Field trial experiment
Four permanent ponds within the Peñalara Massif support significant larval fire salamander populations. The combination of lithology, scarcity of aquatic vegetation, low depth and crystal-clear waters is extremely favorable for larval counts and captures [38]. We removed salamander larvae from three ponds (Laguna de Pájaros, Charca Mariposa and Charca Larga) at the end of the summers of 2015 and 2016, and early spring of 2017, when no larvae of other species were present. We removed 1149 larvae from Laguna de Pájaros, 214 from Charca Mariposa and 32 from Charca Larga in 2015, 2057 larvae from Laguna de Pájaros in 2016 and 1701 larvae across all three sites in 2017. We did not remove larvae from Laguna Grande, which served as our control pond. The larvae we removed were reared in captivity in the facilities of the National Park referred to above and returned as metamorphosed juveniles to the surroundings of their pond of capture in early spring (2015 and 2016 collections) or late summer (2017 collection). Ninety-five casualties occurred as the result of removal actions, while approximately 5100 juveniles were released.
To estimate load of infection, we sampled a maximum of 20 larvae at each pond in early spring and late summer (from summer 2015 to spring 2018). We selected a sample size of 20 because our previous work in the study area and on this species showed no significant differences of load of infection among ponds within a year [41]. By definition spring samples were entirely composed of overwintering larvae, while summer samples were a mix of young of the year and overwintered animals. Again, load was quantified using qPCR as described above [44].
We estimated the removal effort (percent removed) for each removal event in each pond by counting the number of larvae before starting the removal activity and the night after. However, for statistical analyses binary values for removal activity (0: no removals, 1: removals) preceding the breeding season were assigned to each pond for each year (e.g., removal activity at the end of the summer of 2016 and at the early spring of 2017 were assigned to 2017 for analyses). Bd load (including zero values) was log-transformed to achieve normality and homoscedasticity of residuals. We fitted a general linear mixed model using the same procedures described above to detect differences in infection intensity between removal activity and among ponds, years and seasons. ‘Pond’ was considered a random effect, whereas ‘year’, ‘season’ and ‘removal activity’ were considered as fixed, categorical effects, and no interactions were entered into the model due to the lack of data involved in interactions among fixed effects.
All applicable institutional and/or national guidelines for the collection, care and use of animals were followed. Field and laboratory procedures were carried out under the permission of the competent authorities Consejería de Medio Ambiente de Madrid (permits 10/064263.9/15, 10/072229.9/16, 10/132950.9/17, 10/007480.9/18, 10/084888.9/18) and Instituto Aragonés de Gestión Ambiental (500201/24/2015/226).
Results
Ex situ experiment
The within-subjects mixed model including the full set of interaction terms was the best model (AICc = 425.0; main effects AICc = 461.6; null AICc = 592.0) and was significant (likelihood ratio test of the model compared to the null model without fixed predictors: 184.2, P≪0.001). Fixed effects alone accounted for a high proportion of variance in Bd load (87.1%). Water temperature was positively and significantly related to Bd load (Table 1) and one Celsius degree increase in average water temperature led to a x1.29 increase in Bd load (regression coefficient of temperature: 0.112, se = 0.023; 10^0.112 = 1.294). Interspecific differences in Bd load were also significant (mean ± standard error in decimal logarithm: midwife toads 4.73 ± 0.089, fire salamanders 0.79 ± 0.091). Larvae density had, on average, a subtle positive influence on Bd infection (low figure of partial-eta2 in Table 1), as the experimental group of larvae containing only one animal did generate weaker infection loads than those in the group with the greatest density. The sequential timing of “sampling events”, on average, did have any influence on Bd load. Nevertheless, the global effects of larvae density and “sampling events” were not generalizable, as all the interaction terms containing these effects were significant, with a prominent effect of the differences between species throughout the sequential sampling events. Thus, the main effects of density and time after the beginning of trials should be interpreted separately for the two species (Fig 2).
Vertical bars depict mean ± one standard error of Bd load in decimal logarithm.
The same individual per tank was sampled in five different occasions with a time lag of ca. 30 days. SS: sum of squares. partial-η: partial-eta2 estimating relative magnitude effects.
Midwife toads maintained high Bd loads throughout the experiment, with some evidence of increasing loads in some treatments (Fig 2). Fire salamander infections were weaker and decreased from the first measurement in the one and five larvae density treatments, while Bd load increased from the first to the fifth “sampling event” in the high density treatment (Fig 2). Moreover, Bd load of individually housed fire salamanders decreased until infections were almost undetectable in this treatment (Fig 2).
Field trial experiment
Bd loads of 392 salamander larvae were used to build the model (see Fig 3 for prevalence and load values by site and season), where fixed factors explained 26.2% of variance and both fixed and random effects explained 45.9%, yielding significant results for year (21.8% of the variance accounted for year accounting for the total variance observed among larvae and ponds, F3,391 = 44.8, P<<0.0001), season (5.9%, F1,395 = 36.4, P<<0.0001) and removal activity (4.0%, F1,340 = 24.3, P<<0.0001). As expected [43], Bd load was greater in spring than in summer (Fig 3). Load was weaker in ponds where larvae were removed the previous year after controlling for the other sources of variation, but only at Charca Larga and Charca Mariposa where we removed more than 80% of larvae at each removal. Conversely, and again after controlling for the other sources of variation, load of infection was even higher at Laguna de los Pájaros after larvae were removed the previous year.
Dashed vertical lines indicate date (the end of the summer of the previous year and/or the early spring of the current year) and the approximate total proportion of larvae removed. Laguna Grande acted as control pond where no larvae were removed.
Discussion
A consistent role for temperature in B. dendrobatidis infection dynamics has been reported for this study system and for both of our study species [38,43,54–56]. The result of our ex situ experiment also suggests a significant effect of environmental temperature, albeit a weak one, on increasing load. The relative weak impact of temperature on load can be attributed to the low temperatures animals experienced over the 5 months of the experiment (< 10°C): Bosch et al. [54] showed that shifts in disease in A. obstetricans attributable to changes in infection dynamics were associated with temperatures approximating optimal, ex situ growth conditions for Bd (approx. 17–24°C).
Reducing host density is frequently mentioned as a potential mitigation strategy for controlling infectious diseases of wildlife. Long-term surveys indicated that when infection loads are lower, survival and recruitment of susceptible species like S. salamandra are enough to allow population persistence [38]. Our ex situ experiment demonstrated a significant role of reduced density in Bd infection intensity at very low densities, and our field experiment showed that reducing the abundance of salamander larvae in the field had a significant impact on load of infection the following season. Recent studies indicate [57,58], self-reinfection is more important than contact with other infected hosts or the environment, however the ex situ component of our study does not support this, as the animals in the treatment entirely reliant on re-infection (individually housed animals) often were unable to maintain infections over the course of the experiment. We are aware, however, that the focal species in our field study (S. salamandra) may not play as significant role in maintaining infection than the previous reservoir in the system (A. obstetricans) played [37–39], as supported by our ex situ comparisons across species. Consequently, future research examining the effect of removals of species that function more clearly as interspecific reservoirs and super-spreaders could yield other outcomes.
Most studies of culling programs for wildlife have found that removal rates often need to be extremely high [59–62]. To control chronic wasting disease in reindeer, eradication was required [63,64]. Our field trial experiment achieved removal rates of 80–100% in just one or two sessions with less than 10 people involved and in relatively large ponds (>5000 m2). However, our results indicate that even significant reductions of host abundance produce weak impacts on load of infections. This result aligns with some recent work that illustrate how culls rarely achieved the goal of reduced prevalence, load or disease, especially for well-established enzootic diseases or when disease can persist in the environment [65–67]. In our study site, chytridiomycosis has been present for at least 20 years, whereas Bd persistence in the environment has never been recorded in these oligotrophic glacial ponds (Bosch, unpublished results). Across annual cycles, adult amphibians did not consistently sustain infections but instead gained and lost infections from year to year [68]. We also note that the two ponds where removals had the greatest impact were also the smallest in our system, and where we counted fewer than two individuals per shoreline meter during removals. In contrast counts of larvae at the larger ponds were more than twice that (4.4 individual/m). Therefore removal effort, pond size and initial density are confounded in our study and we cannot unambiguously ascribe cause to the effect of infection reductions. What is less ambiguous are year effects, which were strongest and affected all ponds, including our control pond, Laguna Grande, where larvae were not removed.
Although our ex situ experiment argues for a positive relationship between load and water temperature below 10°C, long-term monitoring data supports a negative relationship for fire salamander infections when air temperatures exceed 20°C [38]. The Peñalara Massif is experiencing atypical patterns of warming and it is plausible that the year-on-year effects that predominantly affected load in S. salamandra larvae during the course of this study were driven by continued increases in average summer temperatures [38]. These effects may be direct on the pathogen or may affect load estimates by accelerating larval development and removing older, overwintering larvae that are predominantly responsible for detectable infections in Peñalara data sets as they move from the aquatic to the terrestrial environment. Additionally, stochastic variation can play an important role in disease dynamics of wildlife [69] and may have played a role in our removal trials experiment. In conclusion, based on the weak evidence that larval removal is a successful management strategy, artificial reductions in host abundance may not be a recommendable option for mitigating chytridiomycosis in the field unless extremely large reductions are possible.
Acknowledgments
We thank J. A. Vielva, and all people working at the Sierra de Guadarrama National Park for the facilities and help provided to complete this work. We thank C. Serrano and many people involved in fieldwork. The governing department for the environment of Comunidad de Madrid extended permits for field work and approved experimental protocols.
References
- 1. Goltsman M, Kruchenkova EP, Macdonald DW. The Mednyi arctic foxes: treating a population imperilled by disease. Oryx. 1996;30:251–258.
- 2. Plowright W. The effects of rinderpest and rinderpest control on wildlife in Africa. Symp Zool Soc Lond. 1982;50:1–28.
- 3. Haydon DT, Randall DA, Matthews L, Knobel DL, Tallents LA, Gravenor MB, et al. Low-coverage vaccination strategies for the conservation of endangered species. Nature. 2006;443(7112):692–695. pmid:17036003
- 4. Freuling CM, Hampson K, Selhorst T, Schroder R, Meslin FX, Mettenleiter TC, et al. The elimination of fox rabies from Europe: determinants of success and lessons for the future. Philos Trans R Soc Lond B. 2013;3681623. pmid:23798690
- 5. Bosch J, Sanchez-Tome E, Fernandez-Loras A, Oliver JA, Fisher MC, Garner TW. Successful elimination of a lethal wildlife infectious disease in nature. Biol Lett. 2015;11(11). pmid:26582843
- 6. de Vos V, Bryden HB. Anthrax in the Kruger National Park: temporal and spatial patterns of disease occurrence. Salisbury Med Bull 1996;87:26–30.
- 7. Woodroffe R, Donnelly CA, Jenkins HE, Johnston WT, Cox DR, Bourne FJ, et al. Culling and cattle controls influence tuberculosis risk for badgers. Proc Natl Acad Sci USA. 2006;103(40):14713–14717. pmid:17015843
- 8. Doak DF, Bakker VJ, Vickers W. Using population viability criteria to assess strategies to minimize disease threats for an endangered carnivore. Conserv Biol 2013;27:303–314. pmid:23521669
- 9. Nguyen C, Gray M, Burton TA, Foy SL, Foster JR, Gendernalik AL, et al. Evaluation of a novel West Nile virus transmission control strategy that targets Culex tarsalis with endectocide-containing blood meals. PLoS Negl Trop Dis. 2019;13(3):e0007210. pmid:30845250
- 10. Anderson RM, May RM. Population biology of infectious diseases: part I. Nature 1979;280:361–367. pmid:460412
- 11. Anderson RM, May RM. Vaccination and herd immunity to infectious diseases. Nature 1985;318:323–329. pmid:3906406
- 12.
Anderson RM, Anderson B, May RM. Infectious diseases of humans: Dynamics and control. Oxford: Oxford University Press; 1992.
- 13. Van Buskirk J, Ostfeld R.S. Controlling lyme disease by modifying the density and species composition of tick hosts. Ecol Appl. 1995;5:1133–1140.
- 14. Arneberg P, Skorping A, Grenfell B, Read AF. Host densities as determinants of abundance in parasite communities. Proc R Soc B 1998;265:1283–1289.
- 15. Dobson A. Population dynamics of pathogens with multiple host species. Am Nat 2004;164:S64–S78. pmid:15540143
- 16. Fisher MC, Henk DA, Briggs CJ, Brownstein JS, Madoff LC, McCraw SL, et al. Emerging fungal threats to animal, plant and ecosystem health. Nature. 2012;484(7393):186–194. pmid:22498624
- 17. Scheele BC, Pasmans F, Skerratt LF, Berger L, Martel A, Beukema W, et al. Amphibian fungal panzootic causes catastrophic and ongoing loss of biodiversity. Science. 2019;363(6434):1459–1463. pmid:30923224
- 18. Martel A, Blooi M, Adriaensen C, Van Rooij P, Beukema W, Fisher MC, et al. Wildlife disease. Recent introduction of a chytrid fungus endangers Western Palearctic salamanders. Science. 2014;346(6209):630–631. pmid:25359973
- 19. O'Hanlon SJ, Rieux A, Farrer RA, Rosa GM, Waldman B, Bataille A, et al. Recent Asian origin of chytrid fungi causing global amphibian declines. Science. 2018;360(6389):621–627. pmid:29748278
- 20. Woodhams DC, Bosch J, Briggs CJ, Cashins S, Davis LR, Lauer A, et al. Mitigating amphibian disease: strategies to maintain wild populations and control chytridiomycosis. Front Zool. 2011;8(1):8. pmid:21496358
- 21. Scheele BC, Hunter DA, Grogan LF, Berger L, Kolby JE, McFadden MS, et al. Interventions for reducing extinction risk in chytridiomycosis-threatened amphibians. Conserv Biol. 2014;28(5):1195–1205. pmid:24975971
- 22. Garner TW, Schmidt BR, Martel A, Pasmans F, Muths E, Cunningham AA, et al. Mitigating amphibian chytridiomycoses in nature. Philos Trans R Soc Lond B Biol Sci. 2016;371(1709). pmid:28080996
- 23. Skerratt LF, Berger L, Clemann N, Hunter DA, Marantelli G, Newell DA, et al. Priorities for management of chytridiomycosis in Australia: saving frogs from extinction. Wildlife Res. 2016;43(2):105–120.
- 24. Converse SJ, Bailey LL, Mosher BA, Funk WC, Gerber BD, Muths E. A model to inform management actions as a response to chytridiomycosis-associated decline. EcoHealth. 2017;14(Suppl 1):144–155. pmid:27056609
- 25. Canessa S, Bozzuto C, Campbell Grant EH, Cruickshank SS, Fisher MC, Koella JC, et al. Decision-making for mitigating wildlife diseases: from theory to practice for an emerging fungal pathogen of amphibians. J Appl Ecol. 2018;55(4):1987–1996.
- 26. Canessa S, Bozzuto C, Pasmans F, Martel A. Quantifying the burden of managing wildlife diseases in multiple host species. Conserv Biol. 2019;33(5):1131–1140. pmid:30868671
- 27. Hettyey A, Ujszegi J, Herczeg D, Holly D, Vörös J, Schmidt BR, et al. Mitigating disease impacts in amphibian populations: capitalizing on the thermal optimum mismatch between a pathogen and its host. Front Ecol Evol. 2019;7.
- 28. Thomas V, Wang Y, Van Rooij P, Verbrugghe E, Baláž V, Bosch J, et al. Mitigating Batrachochytrium salamandrivorans in Europe. Amphibia-Reptilia. 2019;40(3):265–290.
- 29.
Vredenburg VT, Briggs CJ, Harris RN. Host pathogen dynamics of amphibian chytridiomycosis: the role of the skin microbiome in health and disease. In: Olson L, Choffnes E, Relman D, Pray L, editors. Fungal diseases: An emerging threat to human, animal, and plant health. Washington DC: National Academy Press; 2011. pp. 342–355.
- 30. Hudson MA, Young RP, Lopez J, Martin L, Fenton C, McCrea R, et al. In-situ itraconazole treatment improves survival rate during an amphibian chytridiomycosis epidemic. Biol Conserv. 2016;195:37–45.
- 31. Geiger CC, Bregnard C, Maluenda E, Voordouw MJ, Schmidt BR. Antifungal treatment of wild amphibian populations caused a transient reduction in the prevalence of the fungal pathogen, Batrachochytrium dendrobatidis. Sci Rep. 2017;7(1):5956. pmid:28729557
- 32. Rachowicz LJ, Briggs CJ. Quantifying the disease transmission function: effects of density on Batrachochytrium dendrobatidis transmission in the mountain yellow-legged frog Rana muscosa. J Anim Ecol. 2007;76(4):711–721. pmid:17584377
- 33. Mitchell KM, Churcher TS, Garner TW, Fisher MC. Persistence of the emerging pathogen Batrachochytrium dendrobatidis outside the amphibian host greatly increases the probability of host extinction. Proc R Soc B. 2008;275(1632):329–334. pmid:18048287
- 34. Briggs CJ, Knapp RA, Vredenburg VT. Enzootic and epizootic dynamics of the chytrid fungal pathogen of amphibians. Proc Natl Acad Sci U S A. 2010;107(21):9695–9700. pmid:20457916
- 35. Vredenburg VT, Knapp RA, Tunstall TS, Briggs CJ. Dynamics of an emerging disease drive large-scale amphibian population extinctions. Proc Natl Acad Sci USA. 2010;107(21):9689–9694. pmid:20457913
- 36. Stockwell MP, Garnham JI, Bower DS, Clulow J, Mahony MJ. Low disease-causing threshold in a frog species susceptible to chytridiomycosis. FEMS Microbiol Lett. 2016;363(12).
- 37. Bosch J, Martínez-Solano I, García-París M. Evidence of a chytrid fungus infection involved in the decline of the common midwife toad (Alytes obstetricans) in protected areas of Central Spain. Biol Conserv. 2001;97:331–337.
- 38. Bosch J, Fernandez-Beaskoetxea S, Garner TWJ, Carrascal LM. Long-term monitoring of an amphibian community after a climate change- and infectious disease-driven species extirpation. Glob Chang Biol. 2018;24(6):2622–2632. pmid:29446515
- 39. Bosch J, Martínez-Solano I. Chytrid fungus infection related to unusual mortalities of Salamandra salamandra and Bufo bufo in the Peñalara Natural Park (Central Spain). Oryx. 2006;40:84–89.
- 40. Garner TWJ, Walker S, Bosch J, Leech S, Marcus Rowcliffe J, Cunningham AA, et al. Life history tradeoffs influence mortality associated with the amphibian pathogen Batrachochytrium dendrobatidis. Oikos. 2009;118(5):783–791.
- 41. Medina D, Garner TW, Carrascal LM, Bosch J. Delayed metamorphosis of amphibian larvae facilitates Batrachochytrium dendrobatidis transmission and persistence. Dis Aquat Organ. 2015;117(2):85–92. pmid:26648101
- 42. Fernandez-Beaskoetxea S, Bosch J, Bielby J. Infection and transmission heterogeneity of a multi-host pathogen (Batrachochytrium dendrobatidis) within an amphibian community. Dis Aquat Organ. 2016;118(1):11–20. pmid:26865231
- 43. Fernandez-Beaskoetxea S, Carrascal LM, Fernandez-Loras A, Fisher MC, Bosch J. Short term minimum water temperatures determine levels of infection by the amphibian chytrid fungus in Alytes obstetricans tadpoles. PloS One. 2015;10(3):e0120237. pmid:25793985
- 44. Boyle DG, Boyle DB, Olsen V, Morgan JAT, Hyatt AD. Rapid quantitative detection of chytridiomycosis (Batrachochytrium dendrobatidis) in amphibian samples using real- time Taqman PCR assay. Dis Aquat Organ. 2004;60:141–148. pmid:15460858
- 45. Kenward MG, Roger JH. An improved approximation to the precision of fixed effects from restricted maximum likelihood. Comput Stat Data An. 2009;53:2583–2595.
- 46. Bates D, Maechler M, Bolker B, Walker S. Fitting linear mixed-effects models using lme4. J Stat Softw 2015;67:1–48.
- 47. Kuznetsova A, Brockhoff PB, Christensen RHB. lmerTest package: tests in linear mixed effects models. J Stat Softw. 2017;82:1–26.
- 48. Zeileis A, Hothorn T. Diagnostic checking in regression relationships. R News. 2002;2:7–10.
- 49. Halekoh U, Højsgaard S. A Kenward-Roger approximation and parametric bootstrap methods for tests in linear mixed models—the R package pbkrtest. J Stat Softw. 2014;59:1–30. pmid:26917999
- 50.
Fox J, Weisberg S. An {R} companion to applied regression, second ed. Sage, USA: Thousand Oaks; 2011.
- 51.
Barton K. MuMIn: multi-model Inference. R package version 1.43.6. 2019. https://CRAN.R-project.org/package=MuMIn.
- 52.
De Rosario-Martinez H. Phia: post-hoc interaction analysis. R package version 0.2–1. 2015. https://CRAN.R-project.org/package=phia.
- 53. Nakagawa S, Schielzeth H. A general and simple method for obtaining R2 from Generalized Linear Mixed-effects Models. Methods Ecol Evol. 2013;4:133–142.
- 54. Bosch J, Carrascal LM, Duran L, Walker S, Fisher MC. Climate change and outbreaks of amphibian chytridiomycosis in a montane area of Central Spain; is there a link? Proceedings of the Royal Society B: Biological Sciences. 2007;274(1607):253–60. pmid:17148254
- 55. Walker SF, Bosch J, Gomez V, Garner TW, Cunningham AA, Schmeller DS, et al. Factors driving pathogenicity vs. prevalence of amphibian panzootic chytridiomycosis in Iberia. Ecol Lett. 2010;13(3):372–382. pmid:20132274
- 56. Doddington B, Bosch J, Oliver J, Grassly N, García G, Benedikt R, et al. Context-dependent amphibian host population response to an invading pathogen. Ecology. 2013;98:1795–1804.
- 57. DiRenzo GV, Tunstall TS, Ibáñez R, deVries MS, Longo AV, Zamudio KR, et al. External reinfection of a fungal pathogen does not contribute to pathogen growth. EcoHealth. 2018;15:815–826. pmid:30128614
- 58. Wilber MQ, Langwig KE, Kilpatrick AM, McCallum HI, Briggs CJ. Integral projection models for host–parasite systems with an application to amphibian chytrid fungus. Methods Ecology Evol. 2016;7:1182–1194. pmid:28239442
- 59. Gross J, Miller M. Chronic wasting disease in mule deer: disease dynamics and control. J Wildlife Manage. 2001;65:205–215.
- 60. Wasserberg G, Osnas EE, Rolley RE, Samuel MD. Host culling as an adaptive management tool for chronic wasting disease in white-tailed deer: a modelling study. J Appl Ecol. 2009;46:457–466. pmid:19536340
- 61. Bradshaw CJ, McMahon CR, Miller PS, Lacy RC, Watts MJ, Verant ML, et al. Novel coupling of individual-based epidemiological and demographic models predicts realistic dynamics of tuberculosis in alien buffalo. J Appl Ecol. 2012;49:268–277.
- 62. McCallum H. Models for managing wildlife disease. Parasitology. 2016;143:805–820. pmid:26283059
- 63. Woolhous M, Donaldson A. Managing foot-and-mouth. Nature. 2001;410:515–516. pmid:11279464
- 64. Mysterud A, Rolandsen CM. A reindeer cull to prevent chronic wasting disease in Europe. Nat Ecol Evol. 2018;2:1343–1345. pmid:29988166
- 65.
Carter SP, Roy SS, Cowan DP, Massei G, Smith GC, Ji W et al. Options for the control of disease 2: targeting hosts. In: Delahay RJ, Smith GC, Hutchings MR, editors. Management of disease in wild mammals. Tokyo, Berlin, Heidelberg, New York: Springer; 2009. pp. 121–146.
- 66. Harrison A, Newey S, Gilbert L, Haydon DT, Thirgood S. Culling wildlife hosts to control disease: mountain hares, red grouse and louping ill virus. J Appl Ecol. 2010; 47:926–930.
- 67. Hallam TG, McCracken GF. Management of the panzootic white-nose syndrome through culling of bats. Conserv Biol. 2011;25:189–194. pmid:21054529
- 68. Daversa DR, Manica A, Bosch J, Jolles JW, Garner TWJ, Hawley D. Routine habitat switching alters the likelihood and persistence of infection with a pathogenic parasite. Functional Ecology. 2018;32(5):1262–1270.
- 69. Lambert S, Ezanno P, Garel M, Gilot-Fromont E. Demographic stochasticity drives epidemiological patterns in wildlife with implications for diseases and population management. Sci Rep. 2018;8,16846. pmid:30442961