http://orcid.org/0000-0002-1704-2720
Figures
Abstract
Batrachochytrium dendrobatidis (Bd), a causal agent of the amphibian fungal skin disease chytridiomycosis, has been implicated in the decline and extinction of over 200 species worldwide since the 1970s. Despite almost two decades of research, the history of Bd and its global spread is not well understood. However, the spread of the Global Panzootic Lineage of Bd (Bd-GPL), the lineage associated with amphibian die-offs, has been linked with the American bullfrog (Rana [Aqurana] catesbeiana) and global trade. Interestingly, R. catesbeiana is native to the eastern U.S., where no Bd-related declines have been observed despite Bd’s presence since the late 1800s. In contrast Bd has been found to have emerged in California and Mexico in the 1960s and 1970s, after which epizootics (i.e., epidemics in wildlife) ensued. We hypothesize that Bd-GPL spread from the eastern U.S. with the introduction of R. catesbeiana into the western US, resulting in epizootics and declines of native host species. Using museum records, we investigated the historical relationship between R. catesbeiana and Bd invasion in the western US and found that R. catesbeiana arrived in the same year or prior to Bd in most western watersheds that had data for both species, suggesting that Bd-GPL may have originated in the eastern US and R. catesbeiana may have facilitated Bd invasion in the western US. To predict areas with greatest suitability for Bd, we created a suitability model by integrating habitat suitability and host availability. When we incorporated invasion history with high Bd suitability, we found that watersheds with non-native R. catesbeiana in the mountain ranges of the West Coast have the highest disease risk. These findings shed light on the invasion history and disease dynamics of Bd in North America. Targeted historical surveys using archived specimens in natural history collections and present-day field surveys along with more localized, community-level studies, monitoring, and surveillance are needed to further test this hypothesis and grow our understanding of the disease ecology and host-pathogen dynamics of Bd.
Citation: Yap TA, Koo MS, Ambrose RF, Vredenburg VT (2018) Introduced bullfrog facilitates pathogen invasion in the western United States. PLoS ONE 13(4): e0188384. https://doi.org/10.1371/journal.pone.0188384
Editor: Matthew C. Fisher, Imperial College Faculty of Medicine, UNITED KINGDOM
Received: January 31, 2017; Accepted: November 6, 2017; Published: April 16, 2018
Copyright: © 2018 Yap 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: Data are available from the amphibian disease portal (amphibiandisease.org) at https://n2t.net/ark:/21547/AxB2. All other relevant data are within the paper and its Supporting Information files.
Funding: This work was supported by National Science Foundation 1258133, 1633948 (V.T.V.), and 1441652 (M.S.K.).
Competing interests: The authors have declared that no competing interests exist.
Introduction
Chytridiomycosis is an emerging infectious disease primarily caused by the fungal pathogen Batrachochytrium dendrobatidis (Bd). This pathogen has significantly affected global amphibian biodiversity, infecting over 500 species [1] and causing declines and extinctions in at least 200 species since the 1970s [2–4]. Global trade likely played a role in the current Bd pandemic by spreading non-native, infected animals worldwide and exposing naïve populations to Bd [3,5–8].
Despite almost two decades of Bd research, the history of Bd and its global spread is not well understood. However, genomic studies have led to the discovery of multiple Bd strains that range in pathogenicity [5,9,10], which provides some insight of the evolutionary history of Bd. Different Bd lineages have been identified in Brazil, Switzerland, South Africa, and South Korea, where they appear to be in an enzootic state of coexistence with native amphibian populations [5,9,11]. The Global Panzootic Lineage (Bd-GPL) is the strain that has been associated with known epizootics (i.e., epidemics in wildlife) [9]. There is evidence that hybridization between Bd strains is possible [5], and it has been suggested that hybridization may have led to the origin of Bd-GPL [9]. However, genome-wide patterns suggest that mitotic recombination via asexual reproduction is the more likely mechanism to have led to the emergence of Bd-GPL [10].
A defining feature of Bd-GPL is the presence of specific loss of heterozygosity (LOH) events that may have occurred as recently as 1,000 years ago [10]. The GPL consists of two clades, GPL-1, which is most common in North America, and GPL-2, which is most common in Central and South America [5,12]. Because GPL-2 has additional LOH events that are absent in GPL-1, James et al. [12] hypothesize that GPL-1 is the more ancestral lineage and that GPL-2 diverged some time after it was introduced to Central and South America. Therefore, they conclude that Bd-GPL or its parental strain originated in the temperate zone of North America [12].
If Bd-GPL originated in North America, then we would expect amphibian populations to be in an enzootic state throughout, as populations never exposed to Bd are more likely to be at greater risk of experiencing an epizootic [13–15] compared to those that have existed with the pathogen for an extended amount of time [16–18]. Yet host-pathogen dynamics are not uniform across the continent. The earliest known record of Bd is from Illinois in 1888 [18], and pathogen host dynamics east of the Rocky Mountains reflect an enzootic state with no known declines due to Bd [18–22]. However, Bd epizootics have been documented in the western US (California [15,23], Arizona [24,25], and Colorado [26]) and in Mexico [27]. In addition, several studies have shown that Bd invasion occurred in areas of California and Mexico in the 1960s and 1970s [15,27–30].
The eastern US is the native range of the American bullfrog (Rana [Aquarana] catesbeiana [31]), a known Bd reservoir [7,32] that is popular in pet and food trade and has been implicated as a vector in global disease spread [3,6,7,33,34]. Rödder et al. [35] showed that Bd and R. catesbeiana have high overlap with their realized niches, further highlighting a link between these two species. The ability of R. catesbeiana to tolerate Bd infection and the lack of Bd-related declines in the eastern US, while species in the western US and Mexico have experienced Bd epizootics, suggest that Bd may have an evolutionary history with the R. catesbeiana populations in the east. Thus it may be that Bd-GPL or its parental strain originated in the eastern US.
We hypothesize that Bd-GPL or its predecessor co-evolved with R. catesbeiana in the eastern US and the introduction of R. catesbeiana in the western US is an important driver of Bd spread. To test this, we investigate the historical presence of Bd and R. catesbeiana in the western US. If indeed Bd-GPL originated in the eastern US and R. catesbeiana played a role in its spread, then we would expect R. catesbeiana occurrences to coincide with or precede Bd occurrences in the western US. We then integrate Bd habitat suitability, potential host availability, and the invasion histories of Bd and R. catesbeiana to predict areas where species may have greatest disease risk from Bd in North America.
Materials and methods
Invasion history
To determine the invasion patterns of Bd and R. catesbeiana outside of R. catesbeiana’s native range, we compared the historical occurrences of the two species in the watersheds outside the native range of R. catesbeiana west of the Rocky Mountains. Watershed boundary data (hydrologic unit code 10, HUC10) were obtained from the United States Geological Survey (USGS) [36]. Using 1062 Bd positive records in the western US from survey data from the Vredenburg Lab [37] and Bd-Maps [38] we identified the earliest record of Bd in each watershed. See S1 Table for a summary of Bd-positive data from North America. We did the same for R. catesbeiana, using a total of 2597 occurrence records compiled from several sources: two online repositories that aggregate data from hundreds of natural history collections around the world, VertNet [39] and the Global Biodiversity Information Facility (GBIF [40]), the USGS [37], and any R. catesbeiana occurrences from the Bd survey data. For watersheds that had occurrence data for both species, we identified whether 1) Bd was recorded in the same year or prior to R. catesbeiana or 2) R. catesbeiana was recorded prior to Bd. We also identified watersheds that had only Bd occurrences or only R. catesbeiana occurrences.
Bd suitability and disease risk
We integrated abiotic and biotic factors to predict the areas in North America with the greatest Bd suitability. We first created a presence-only habitat suitability model (HSM) driven by climate and land use factors using Maxent version 3.3.3k [41]. We used 1775 Bd occurrence records with 988 unique localities in the North America mainland from the Vredenburg Lab [37] and Bd-Maps [38]. See S1 Table for a summary of the Bd-positive data used. Boria et al. [42] showed that applying a spatial filter to presence data reduces overfitting; therefore, we applied a 10 arc-minute (~12 km2) spatial filter by randomly choosing one Bd occurrence site from every 10 arc-minute area using R (dismo [43] and maptools [44] packages). This resulted in 746 Bd-positive sites with environmental data for model training. We further minimized sampling bias by restricting background sampling areas to a minimum convex hull around the occurrence points (S1 Fig) [45–49].
We obtained 19 bioclimatic variables from the Worldclim database (http://www.worldclim.org/bioclim), which are a set of interpolated temperature and precipitation conditions based on monthly averages measured at weather stations across the globe from the years 1950 to 2000, latitude, longitude, and elevation [50]. To incorporate land use patterns, we used the global human footprint (HF), which quantifies anthropogenic influences on the terrestrial environment based on land cover, human accessibility, land transformation, human population density, and infrastructure between the years 1993 to 2009 [51]. Some of the bioclimatic variables are highly correlated with each other; therefore, to reduce overfitting of the model due to multicollinearity of the model predictors, we calculated Spearman’s rank correlations (r) among all the variables to determine which variables were highly correlated with each other. When variable pairs had an r2 > 0.7, we chose the higher-ranking factor. We then used the resulting subset of 11 environmental variables to create the Bd HSM: mean diurnal temperature range (Bio2), temperature seasonality (Bio4), maximum temperature of the warmest month (Bio5) minimum temperature of the coldest month (Bio6), mean temperature of the wettest quarter (Bio8), precipitation of the driest quarter (Bio9), precipitation of the wettest month (Bio13), precipitation of the driest month (Bio14), precipitation of the warmest quarter (Bio18), precipitation of the coldest quarter (Bio19), and human footprint. These variables are biologically relevant to Bd and have been shown to perform well in previous Bd HSMs [6,12,52–55].
To account for variability and other uncertainties inherent in modeling, we ran 20 replicates for each species using cross-validation: occurrence data were divided into 20 equal-sized folds, or groups, and for each replicate 19 folds were used for model training and one fold was used for model testing. We used 20 replicates because ensemble forecasts have been shown to make models more reliable and robust than single model outputs by accounting for variability [56]. The results were then averaged by Maxent to produce a final probabilistic density function of potentially suitable habitat.
In addition to abiotic factors, biotic factors like host availability or the presence of a vector species are critical for disease maintenance and spread [57]. A pathogen cannot persist without sufficient abundance of susceptible hosts; therefore, the establishment of Bd in an area with suitable habitat depends on the presence and abundance of amphibian hosts. Information on host abundance at the community level for a continental assessment is presently not available; however, several studies have shown species richness can be an important factor in Bd spread, particularly where a competent reservoir is present [1,6,30,52,58–62]. Thus, in lieu of abundance data, we considered species richness in conjunction with the presence of R. catesbeiana. To refine our predictions of Bd suitability, we calculated the product of the Bd HSM and amphibian richness, which was then interpreted at a relative scale.
We estimated amphibian richness in mainland North America by overlaying unique species ranges obtained from the IUCN Red List [63] and AmphibiaWeb [64]. We created a raster dataset in which we tallied the intersecting ranges per grid cell using R (dplyr [65] and raster [66] packages) for total amphibians. Spatial data for range maps compiled by AmphibiaWeb are available at https://github.com/AmphibiaWeb. All range maps are viewable on AmphibiaWeb by species (http://amphibiaweb.org) [64].
The Maxent software calculates multiple threshold values for each model run to aid in model interpretation. Areas with values above these thresholds can be interpreted as a reasonable estimate of a species’ suitable habitat, depending on the quality of the data [67]. Because we modified the Maxent-produced Bd HSM by incorporating amphibian richness, for our analyses, we used two thresholds that were proportional to Maxent’s HSM thresholds. To identify areas with general Bd suitability (i.e., areas with any likelihood that Bd could establish there) we used the minimum training presence logistic threshold, which is the mean of the lowest probability associated with each training point (i.e., true positives) for each of the 20 replicates. To identify areas with high Bd suitability (i.e., areas where Bd is most likely to establish), we used the most conservative threshold value that gave the lowest predicted area [47], which was the mean value at which training sensitivity and specificity were equal. We then compared the watersheds with species emergence data to the Bd suitability model to identify areas with the greatest risk of Bd emergence and disease spread.
Results
Invasion history
There were 603 out of 6141 western US watersheds with Bd and/or R. catesbeiana occurrence data: 202 had only Bd records, 301 had only R. catesbeiana records, and 100 had records for both species. Records of R. catesbeiana occurrences documented the same year or prior to Bd occurrences were found in 83% (83/100) of the shared watersheds. In nine of the 17 watersheds where Bd was documented prior to R. catesbeiana, Bd records preceded R. catesbeiana by one to seven years. Most (13/17) were adjacent to watersheds where R. catesbeiana had been documented earlier than the first Bd record of the watershed in question (often by decades).
Bd suitability and disease risk
Areas predicted to have general Bd suitability encompass the eastern half of the US, the Pacific Northwest, portions of mountain ranges in the western US, including the Rocky Mountains and the Sierra Nevada Mountains, and most of Mexico (Fig 1A). The areas deemed unsuitable include the arid and semi-arid regions in the Great Plains and the desert. Approximately 95% (940/988) of Bd-positive localities and 94% (2567/2743) of R. catesbeiana localities were found to be in predicted general Bd suitability areas (Fig 1A). All but two western US watersheds that had Bd occurrence records overlapped with general Bd suitability areas (300/302) (Fig 1B). See Supplemental S2 Fig for a breakdown of the different watersheds types (i.e., where Bd was recorded before R. catesbeiana [17/17], where R. catesbeiana was recorded in the same year or before Bd [82/83], and where only Bd has been recorded [201/202]). Most of the watersheds where R. catesbeiana were recorded overlapped with general Bd suitability areas (95% [286/301] of watersheds where only R. catesbeiana were recorded and 96% [385/401] of all watersheds in the western US where R. catesbeiana has been recorded; S2 Fig).
(A) Areas in North America predicted to have Bd suitability (increased intensity in red indicates increased suitability). Blue dots are Bd localities. Black hash indicates the native range of R. catesbeiana. And (B) western US watersheds (in blue) containing Bd occurrence records.
Areas predicted to have high Bd suitability include mountain ranges along the West Coast of the US, the highlands of Central Mexico, the Coastal Plains of the Southeast US, and the Ozark Plateau (Fig 1). Of the 301 watersheds where only R. catesbeiana has been recorded, 173 (57%) overlap with high Bd suitability areas (Fig 2). These areas are predicted to have the greatest risk of Bd outbreaks.
Areas in North America predicted to have high Bd suitability (deep red) and western US watersheds containing only R. catesbeiana occurrence records (in black). Overlapping areas are predicted to have the highest risk of Bd emergence and outbreaks.
Discussion
Chytridiomycosis has severely impacted amphibian biodiversity globally for decades, with Bd infecting hundreds of amphibian species worldwide [1–4]. Although our knowledge of Bd is incomplete, our understanding of this complex host-pathogen system continues to grow through a variety of ongoing studies of its biology and disease dynamics. By investigating the historical emergence of Bd and R. catesbeiana in the western US, we uncover evidence of a historical relationship between the invasion of Bd and the introduction of R. catesbeiana. We show that R. catesbeiana occurred prior to Bd emergence in most of the shared watersheds, which supports Huss et al. [34] and suggests that the introduction of R. catesbeiana may have played a significant role in spreading Bd to naïve amphibian populations in the western US.
The Bd suitability model is generally similar to those from previous studies [6,12,52,68,69]. Our model indicates that there is Bd suitability in R. catesbeiana’s native range in the eastern US (Fig 1A); however, this likely reflects the high environmental and host suitability for Bd in a region of endemism rather than the threat of disease from Bd to naïve amphibian populations, given that no Bd-related declines have been documented in the eastern US despite Bd’s presence [18–22]. In addition, 94% of R. catesbeiana occurrence data and 96% of western watersheds where R. catesbeiana has been recorded were within general Bd suitability areas (Fig 2). This further highlights that Bd and R. catesbeiana have high niche overlap [35]. While these results are correlational, they support a growing number of studies that link Bd emergence with the presence of introduced R. catesbeiana [3,6–8,34,70–74] and are consistent with the hypothesis that Bd-GPL (or its preceding strain) may have co-evolved with R. catesbeiana in the eastern US.
If we consider the historical link between the emergence of Bd and R. catesbeiana in the western US, we can identify the areas where native species are at greatest risk of disease. Thus, species that occur in watersheds where only R. catesbeiana has been documented that overlap with areas predicted to have high Bd suitability (Fig 2) could have increased risk of Bd emergence. The presence of a reservoir host like R. catesbeiana (or others [75]) could lead to local population declines due to disease, even well beyond the initial emergence of Bd [60]. High risk areas include many watersheds in the Cascade Range in the Pacific Northwest, the California coastal ranges, and the Sierra Nevada Mountains, all areas where native amphibians have declined [15,76–78].
Improving our understanding of the distribution and spread of Bd enhances our knowledge of Bd’s history and provides vital information for amphibian conservation and management strategies. Areas identified as having the highest risk of disease should be prioritized for both historical surveys using archived specimens in natural history collections [27] and present-day field surveys to investigate our hypothesis that the introduction of R. catesbeiana facilitates Bd spread. In addition, proactive management efforts could minimize disease spread in these areas [79].
It is important to consider that Bd is not the only disease threat to amphibian diversity. Batrachochytrium salamandrivorans (Bsal), a second, more recently discovered chytrid pathogen that has not yet been found in North America [19,20], poses another serious threat to North American amphibians [80,81]. The continuous spread of Bd threatens remaining uninfected populations, and the introduction of Bsal could compound the likelihood of more amphibian epizootics.
The most disruptive effect of a single pathogen infection is death; however, surviving species could incur sub-lethal effects from infection (or from fighting off infection) that could suppress immune defenses against other stressors [82–86], such as a second deadly pathogen. This could at least partially explain the Bsal-induced mass mortalities of European fire salamanders (Salamandra salamandra) in the Netherlands, where Bd was co-existing with local amphibian populations in an enzootic state [87,88]. Thus, species surviving in areas where Bd is already present could be more vulnerable to Bsal exposure, and the introduction of Bsal could result in disease outbreaks.
Alternatively, species or populations previously exposed to a similar pathogen might benefit from direct competition between pathogens or cross immunity, in which a host acquires immunity or partial immunity to one pathogen because of previous infection from either the same pathogen or a closely related pathogen [89]. This could potentially explain the lack of Bd-related declines in areas where both an endemic strain of Bd and Bd-GPL are present, such as in Brazil, South Africa, Switzerland, and South Korea [11,17,90]. Perhaps native Bd strains are able to outcompete introduced strains like Bd-GPL, or the evolutionary history with endemic Bd lineages has primed local amphibian populations to be able to resist or tolerate infection from Bd-GPL [5,11]. If previous exposure to Bd leads to either Bd outcompeting Bsal for hosts or cross immunity to Bsal, then persisting populations in areas where Bd is in an enzootic state may be safeguarded against Bsal infections, and amphibians in areas where Bd is more recently established or where neither Bd nor Bsal currently occur may have the highest disease risk. More studies are needed to understand the potential interactions of these pathogens.
Conclusion
We investigated the invasion history of R. catesbeiana and Bd in the western US and found a pattern of Bd dynamics consistent with the hypothesis that invasion of R. catesbeiana facilitated Bd spread in western North America. This supports the hypothesis that Bd-GPL (or its preceding strain) may have originated in the eastern US. We found that the historical presence of R. catesbeiana outside its native range was highly correlated with emergence of Bd in areas predicted to be suitable for Bd. We identified areas of increased disease risk by integrating environmental suitability, host availability, and invasion history to provide guidance for conservation and management efforts. Targeted historical surveys using archived specimens in natural history collections and present day field surveys could help test our hypotheses and further our understanding of the spatiotemporal distribution of Bd. More localized, community-level studies that include monitoring and surveillance are needed to further our understanding of the disease ecology and host-pathogen dynamics of chytridiomycosis. Species susceptibility studies could lead to the identification of the potential interactions between Bd and Bsal and the discovery of potential defenses against disease.
Supporting information
S1 Fig. Training points and background sampling areas for the Bd habitat suitability model.
https://doi.org/10.1371/journal.pone.0188384.s001
(TIF)
S2 Fig. Invaded western US watersheds compared to Bd suitability.
(A) Shared watersheds where Bd was recorded prior to R. catesbeiana (blue) and where R. catesbeiana was recorded in the same year or prior to Bd (black), (B) watersheds where only Bd has been recorded, and (C) watersheds where only R. catesbeiana has been recorded.
https://doi.org/10.1371/journal.pone.0188384.s002
(TIF)
S1 Table. Summary of Bd-positive records used for the Bd HSM.
Species and decades in which Bd was detected in the wild in North America.
https://doi.org/10.1371/journal.pone.0188384.s003
(XLSX)
Acknowledgments
We thank Vredenburg Lab collaborators and Bd-Maps for georeferenced Bd occurrence data. We thank the IUCN and AmphibiaWeb for range map data. We thank the United States Geological Survey for R. catesbeiana range data and watershed boundary data. We thank Thomas Gillespie, James Lloyd-Smith, Richard Vance, and anonymous reviewers for valuable feedback and suggestions. We thank the AmphibiaWeb and Biodiversity Informatics undergraduate apprentices at the Museum of Vertebrate Zoology (University of California, Berkeley) for species range work, and we thank MVZ staff for technical support of this research. We thank Ben de Jesus for design assistance with the figures. Funding for this project was from NSF grants 1258133 and 1633948 to VTV. The spatial data was, in part, the product of NSF grant 1441652 (“VertLife Terrestrial”) to MSK.
References
- 1. Olson DH, Aanensen DM, Ronnenberg KL, Powell CI, Walker SF, Bielby J, et al. Mapping the global emergence of Batrachochytrium dendrobatidis, the amphibian chytrid fungus. PLoS One. 2013;8: e56802. pmid:23463502
- 2. Alroy J. Current extinction rates of reptiles and amphibians. Proc Natl Acad Sci. 2015;112: 13003–13008. pmid:26438855
- 3. Fisher MC, Garner TWJ. The relationship between the emergence of Batrachochytrium dendrobatidis, the international trade in amphibians and introduced amphibian species. Fungal Biol Rev. 2007;21: 2–9.
- 4. Skerratt LF, Berger L, Speare R, Cashins S, McDonald KR, Phillott AD, et al. Spread of chytridiomycosis has caused the rapid global decline and extinction of frogs. Ecohealth. 2007;4: 125–134.
- 5. Schloegel LM, Toledo LF, Longcore JE, Greenspan SE, Vieira CA, Lee M, et al. Novel, panzootic and hybrid genotypes of amphibian chytridiomycosis associated with the bullfrog trade. Mol Ecol. 2012;21: 5162–5177. pmid:22857789
- 6. Liu X, Rohr JR, Li Y. Climate, vegetation, introduced hosts and trade shape a global wildlife pandemic. Proc R Soc B Biol Sci. 2013;280: 20122506. pmid:23256195
- 7. Hanselmann R, Rodríguez A, Lampo M, Fajardo-Ramos L, Alonso Aguirre A, Marm Kilpatrick A, et al. Presence of an emerging pathogen of amphibians in introduced bullfrogs Rana catesbeiana in Venezuela. Biol Conserv. 2004;120: 115–119.
- 8. Garner TW., Perkins MW, Govindarajulu P, Seglie D, Walker S, Cunningham AA, et al. The emerging amphibian pathogen Batrachochytrium dendrobatidis globally infects introduced populations of the North American bullfrog, Rana catesbeiana. Biol Lett. 2006;2: 455–459. pmid:17148429
- 9. Farrer R, Weinert L, Bielby J, Garner T, Balloux F, Clare F, et al. Multiple emergences of genetically diverse amphibian-infecting chytrids include a globalized hypervirulent recombinant lineage. Proc Natl Acad Sci U S A. 2011;108: 18732–18736. pmid:22065772
- 10. Rosenblum EB, James TY, Zamudio KR, Poorten TJ, Ilut D, Rodriguez D, et al. Complex history of the amphibian-killing chytrid fungus revealed with genome resequencing data. Proc Natl Acad Sci U S A. 2013;110: 9385–90. pmid:23650365
- 11. Bataille A, Fong JJ, Cha M, Wogan GOU, Baek HJ, Lee H, et al. Genetic evidence for a high diversity and wide distribution of endemic strains of the pathogenic chytrid fungus Batrachochytrium dendrobatidis in wild Asian amphibians. Mol Ecol. 2013;22: 4196–209. pmid:23802586
- 12. James TY, Toledo LF, Rödder D, da Silva Leite D, Belasen AM, Betancourt-Román CM, et al. Disentangling host, pathogen, and environmental determinants of a recently emerged wildlife disease: lessons from the first 15 years of amphibian chytridiomycosis research. Ecol Evol. 2015;5: 4079–4097. pmid:26445660
- 13. Rachowicz LJ, Hero J-M, Alford R a., Taylor JW, Morgan J a. T, Vredenburg VT, et al. The novel and endemic pathogen hypotheses: competing explanations for the origin of emerging infectious diseases of wildlife. Conserv Biol. 2005;19: 1441–1448.
- 14. Lips KR, Brem F, Brenes R, Reeve JD, Alford RA, Voyles J, et al. Emerging infectious disease and the loss of biodiversity in a Neotropical amphibian community. Proc Natl Acad Sci U S A. 2006;103: 3165–3170. pmid:16481617
- 15. Vredenburg VT, Knapp RA, Tunstall TS, Briggs CJ. Dynamics of an emerging disease drive large-scale amphibian population extinctions. Proc Natl Acad Sci U S A. 2010;107: 9689–94. pmid:20457913
- 16. 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: 9695–9700. pmid:20457916
- 17. Rodriguez D, Becker CG, Pupin NC, Haddad CFB, Zamudio KR. Long-term endemism of two highly divergent lineages of the amphibian-killing fungus in the Atlantic Forest of Brazil. Mol Ecol. 2014;23: 774–787. pmid:24471406
- 18. Talley BL, Muletz CR, Vredenburg VT, Fleischer RC, Lips KR. A century of Batrachochytrium dendrobatidis in Illinois amphibians (1888–1989). Biol Conserv. Elsevier Ltd; 2015;182: 254–261.
- 19. Muletz C, Caruso NM, Fleischer RC, McDiarmid RW, Lips KR. Unexpected rarity of the pathogen Batrachochytrium dendrobatidis in Appalachian Plethodon salamanders: 1957–2011. PLoS One. 2014;9: 1–7. pmid:25084159
- 20. Bales EK, Hyman OJ, Loudon AH, Harris RN, Lipps G, Chapman E, et al. Pathogenic Chytrid Fungus Batrachochytrium dendrobatidis, but not B. salamandrivorans, detected on Eastern Hellbenders. PLoS One. 2015;10: e0116405. pmid:25695636
- 21. Rothermel BB, Walls SC, Mitchell JC, Dodd CK, Irwin LK, Green DE, et al. Widespread occurrence of the amphibian chytrid fungus Batrachochytrium dendrobatidis in the southeastern USA. Dis Aquat Organ. 2008;82: 3–18. pmid:19062748
- 22. Oullet M, Mikaelian I, PAULI BD, Rorigue J, Green DM. Historical evidence of widespread chytrid infection in North American amphibian populations. Conserv Biol. 2005;19: 1431–1440.
- 23. Rachowicz LJ, Knapp RA, Morgan JA, Stice MJ, Vredenburg VT, Parker JM, et al. Emerging infectious disease as a proximate cause of amphibian mass mortality. Ecology. 2006;87: 1671–1683. pmid:16922318
- 24. Bradley GA, Rosen PC, Sredl MJ, Jones TR, Longcore JE. Chytridiomycosis in native Arizona frogs. J Wildl Dis. 2002;38: 206–12. pmid:11838218
- 25. Savage AE, Sredl MJ, Zamudio KR. Disease dynamics vary spatially and temporally in a North American amphibian. Biol Conserv. Elsevier Ltd; 2011;144: 1910–1915.
- 26. Muths E, Corn PS, Pessier AP, Green DE. Evidence for disease-related amphibian decline in Colorado. Biol Conserv. 2003;110: 357–365.
- 27. Cheng TL, Rovito SM, Wake DB, Vredenburg VT. Coincident mass extirpation of neotropical amphibians with the emergence of the infectious fungal pathogen Batrachochytrium dendrobatidis. Proc Natl Acad Sci U S A. 2011;108: 9502–9507. pmid:21543713
- 28. Padgett-Flohr GE. Pathogenicity of Batrachochytrium dendrobatidis in two threatened California amphibians: Rana draytonii and Ambystoma californiense. Herp. 2008;3: 182–191.
- 29. Sette CM, Vredenburg VT, Zink AG. Reconstructing historical and contemporary disease dynamics: A case study using the California slender salamander. Biol Conserv. Elsevier B.V.; 2015;192: 20–29.
- 30. Yap TA, Gillespie L, Ellison S, Flechas S V., Koo MS, Martinez AE, et al. Invasion of the Fungal Pathogen Batrachochytrium dendrobatidis on California Islands. Ecohealth. Springer US; 2016;13: 145–150. pmid:26493624
- 31. Yuan ZY, Zhou WW, Chen X, Poyarkov NA, Chen HM, Jang-Liaw NH, et al. Spatiotemporal Diversification of the True Frogs (Genus Rana): A Historical Framework for a Widely Studied Group of Model Organisms. Syst Biol. 2016;65: 824–842. pmid:27288482
- 32. Daszak P, Cunningham A a., Hyatt AD. Emerging infectious diseases of wildlife: threats to biodiversity and human health. Science (80-). 2000;287: 443–449. pmid:10642539
- 33. Garner TW., Perkins MW, Govindarajulu P, Seglie D, Walker S, Cunningham a. a, et al. The emerging amphibian pathogen Batrachochytrium dendrobatidis globally infects introduced populations of the North American bullfrog, Rana catesbeiana. Biol Lett. 2006;2: 455–459. pmid:17148429
- 34. Huss M, Huntley L, Vredenburg V, Johns J, Green S. Prevalence of Batrachochytrium dendrobatidis in 120 archived specimens of Lithobates catesbeianus (American bullfrog) collected in California, 1924–2007. Ecohealth. 2013;10: 339–343. pmid:24419668
- 35. Rödder D, Schulte U, Toledo LF. High environmental niche overlap between the fungus Batrachochytrium dendrobatidis and invasive bullfrogs (Lithobates catesbeianus) enhance the potential of disease transmission in the Americas. North West J Zool. 2013;9: 178.
- 36.
USGS. National Hydrography Dataset [Internet]. 2017. Available: http://nhd.usgs.gov/data.html
- 37.
AmphibiaWeb. Amphibian Disease Portal [Internet]. 2017. doi:ark:/21547/AxB2
- 38.
Bd-Maps. Global Bd-Mapping Project [Internet]. 2014. Available: bd-maps.net
- 39.
VertNet. VertNet: Distributed Databases with Backbone [Internet]. 2017. Available: http://www.vertnet.org
- 40.
GBIF. Global Biodiversity Information Facility [Internet]. 2017. Available: https://www.gbif.org
- 41. Phillips SJ, Dudík M, Schapire RE. A maximum entropy approach to species distribution modeling. Proc twenty-first Int Conf Mach Learn. 2004; 655–662.
- 42. Boria RA, Olson LE, Goodman SM, Anderson RP. Spatial filtering to reduce sampling bias can improve the performance of ecological niche models. Ecol Modell. Elsevier B.V.; 2014;275: 73–77.
- 43.
Hijmans RJ, Phillips S, Leathwick J, Elith J. dismo: Species distribution modeling [Internet]. R package version 0.9–3; 2013. Available: http://cran.r-project.org/package=dismo
- 44.
Bivand R, Lewin-Koh N. maptools: Tools for reading and handling spatial objects [Internet]. R package version 0.8–30; 2014. Available: http://cran.r-project.org/package=maptools
- 45. Rödder D, Kielgast J, Bielby J, Schmidtlein S, Bosch J, Garner TWJ, et al. Global amphibian extinction risk assessment for the panzootic chytrid fungus. Diversity. 2009;1: 52–66.
- 46. Phillips SJ, Elith J. On estimating probability of presence from use–availability or presence–background data. Ecology. 2013;94: 1409–1419. pmid:23923504
- 47. Thorne LH, Johnston DW, Urban DL, Tyne J, Bejder L, Baird RW, et al. Predictive modeling of spinner dolphin (Stenella longirostris) resting habitat in the main Hawaiian Islands. PLoS One. 2012;7: e43167. pmid:22937022
- 48. Mainali KP, Warren DL, Dhileepan K, McConnachie A, Strathie L, Hassan G, et al. Projecting future expansion of invasive species: Comparing and improving methodologies for species distribution modeling. Glob Chang Biol. 2015;21: 4464–4480. pmid:26185104
- 49. Merow C, Smith MJ, Silander JA. A practical guide to MaxEnt for modeling species ‘ distributions: what it does, and why inputs and settings matter. 2013; 1058–1069.
- 50. Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A. Very high resolution interpolated climate surfaces for global land areas. Int J Climatol. 2005;25: 1965–1978.
- 51. Venter O, Sanderson EW, Magrach A, Allan JR, Beher J, Jones KR, et al. Sixteen years of change in the global terrestrial human footprint and implications for biodiversity conservation. Nat Commun. 2016;7: 1–11. pmid:27552116
- 52. Rödder D, Kielgast J, Bielby J, Schmidtlein S, Bosch J, Garner TWJ, et al. Global amphibian extinction risk assessment for the panzootic chytrid fungus. Diversity. 2009;1: 52–66.
- 53. Murray K a., Retallick RWR, Puschendorf R, Skerratt LF, Rosauer D, McCallum HI, et al. Assessing spatial patterns of disease risk to biodiversity: implications for the management of the amphibian pathogen, Batrachochytrium dendrobatidis. J Appl Ecol. 2011;48: 163–173.
- 54. Ron SR. Predicting the Distribution of the Amphibian Pathogen Batrachochytrium dendrobatidis in the New World. Biotropica. 2005;37: 209–221.
- 55. Puschendorf R, Carnaval AC, VanDerWal J, Zumbado-Ulate H, Chaves G, Bolaños F, et al. Distribution models for the amphibian chytrid Batrachochytrium dendrobatidis in Costa Rica: proposing climatic refuges as a conservation tool. Divers Distrib. 2009;15: 401–408.
- 56. Araújo MB, New M. Ensemble forecasting of species distributions. Trends Ecol Evol. 2007;22: 42–47. pmid:17011070
- 57. Lloyd-Smith JO, Schreiber SJ, Kopp PE, Getz WM. Superspreading and the effect of individual variation on disease emergence. Nature. 2005;438: 355–359. pmid:16292310
- 58. Power AG, Mitchell CE. Pathogen spillover in disease epidemics. Am Nat. 2004;164: S79–S89. pmid:15540144
- 59. Keesing F, Holt RD, Ostfeld RS. Effects of species diversity on disease risk. Ecol Lett. 2006;9: 485–498. pmid:16623733
- 60. Scheele BC, Hunter DA, Brannelly LA, Skerratt LF, Driscoll DA. Reservoir-host amplification of disease impact in an endangered amphibian. Conserv Biol. 2016; pmid:27594575
- 61. DiRenzo G V., Langhammer PF, Zamudio KR, Lips KR. Fungal infection intensity and zoospore output of Atelopus zeteki, a potential acute chytrid supershedder. PLoS One. Public Library of Science; 2014;9: e93356. pmid:24675899
- 62. Becker CG, Zamudio KR. Tropical amphibian populations experience higher disease risk in natural habitats. Proc Natl Acad Sci U S A. 2011;108: 9893–9898. pmid:21628560
- 63.
IUCN. The IUCN red list of threatened species [Internet]. World Conservation Union. 2016. Available: www.iucnredlist.org
- 64.
AmphibiaWeb. Information on amphibian biology and conservation [web application] [Internet]. 2016. Available: http://amphibiaweb.org/
- 65.
Wickham H, Francois R. dplyr: A grammar of data manipulation [Internet]. R package version 0.3.0.2; 2014. Available: http://cran.r-project.org/package=dplyr
- 66.
Hijmans RJ. raster: Geographic data analysis and modeling [Internet]. R package version 2.3–12; 2014. Available: http://cran.r-project.org/package=raster
- 67. Phillips SJ, Anderson RP, Schapire RE. Maximum entropy modeling of species geographic distributions. Ecol Modell. 2006;190: 231–259.
- 68. Xie GY, Olson DH, Blaustein AR. Projecting the Global Distribution of the Emerging Amphibian Fungal Pathogen, Batrachochytrium dendrobatidis, Based on IPCC Climate Futures. PLoS One. 2016;11: e0160746. pmid:27513565
- 69. Grant EHC, Miller DAW, Schmidt BR, Michael J, Amburgey SM, Chambert T, et al. Quantitative evidence for the effects of multiple drivers on continental-scale amphibian declines. Scientific. 2016;6: 1–9. pmid:27212145
- 70. Schloegel LM, Picco AM, Kilpatrick a. M, Davies AJ, Hyatt AD, Daszak P. Magnitude of the US trade in amphibians and presence of Batrachochytrium dendrobatidis and ranavirus infection in imported North American bullfrogs (Rana catesbeiana). Biol Conserv. Elsevier Ltd; 2009;142: 1420–1426.
- 71. Bai C, Garner TWJ, Li Y. First evidence of batrachochytrium dendrobatidis in China: Discovery of chytridiomycosis in introduced American bullfrogs and native amphibians in the Yunnan Province, China. Ecohealth. 2010;7: 127–134. pmid:20372969
- 72. Borzée A, Kosch TA, Kim M, Jang Y. Introduced bullfrogs are associated with increased Batrachochytrium dendrobatidis prevalence and reduced occurrence of Korean treefrogs. PLoS One. 2017;12: 1–14. pmid:28562628
- 73. Fisher MC, Garner TWJ, Walker SF. Global emergence of Batrachochytrium dendrobatidis and amphibian chytridiomycosis in space, time, and host. Annu Rev Microbiol. 2009;63: 291–310. pmid:19575560
- 74. Ghirardi R, López JA, Sanabria EA, Quiroga LB, Levy MG. Pathogenic fungus in feral populations of the invasive North American bullfrog in Argentina. Belgian J Zool. 2017;147: 81–86. https://doi.org/10.26496/bjz.2017.7
- 75. Reeder NMM, Pessier AP, Vredenburg VT. A reservoir species for the emerging amphibian pathogen Batrachochytrium dendrobatidis thrives in a landscape decimated by disease. PLoS One. 2012;7: e33567. pmid:22428071
- 76. De Leon ME, Vredenburg VT, Piovia-Scott J. Recent Emergence of a Chytrid Fungal Pathogen in California Cascades Frogs (Rana cascadae). Ecohealth. 2017;14: 155–161. pmid:27957606
- 77. Blaustein AR, Grant Hokit D, O’Hara RK, Holt RA. Pathogenic fungus contributes to amphibian losses in the pacific northwest. Biol Conserv. 1994;67: 251–254.
- 78.
Vredenburg VT, Koo MS, Wake DB. Declines of amphibians in California. In: Hoffman M, editor. Threatened Amphibians of the World. Barcelona, Spain: Lynx Edicions; 2008. p. 126.
- 79. Grant EHC, Muths E, Katz RA, Canessa S, Adams MJ, Ballard JR, et al. Using decision analysis to support proactive management of emerging infectious wildlife diseases. Front Ecol Environ. 2017;15: 214–221.
- 80. Yap TA, Koo MS, Ambrose RF, Wake DB, Vredenburg VT. Averting a North American biodiversity crisis. Science (80-). 2015;349: 481–482. pmid:26228132
- 81. Richgels KLD, Russell RE, Adams J, White CL, Campbell EH. Spatial variation in risk and consequence of Batrachochytrium salamandrivorans introduction in the USA. R Soc Open Sci. 2016;3: 150616. pmid:26998331
- 82. Parris MJ, Cornelius TO. Fungal pathogen causes competitive and developmental stress in larval amphibian communities. Ecology. 2004;85: 3385–3395.
- 83. Parris MJ, Beaudoin JG. Chytridiomycosis impacts predator-prey interactions in larval amphibian communities. Oecologia. 2004;140: 626–632. pmid:15235903
- 84. Davidson C, Benard MF, Shaffer HB, Parker JM, O’Leary C, Conlon JM, et al. Effects of chytrid and carbaryl exposure on survival, growth and skin peptide defenses in foothill yellow-legged frogs. Environ Sci Technol. 2007;41: 1771–1776. pmid:17396672
- 85. Bielby J, Fisher MC, Clare FC, Rosa GM, Garner TWJ. Host species vary in infection probability, sub-lethal effects, and costs of immune response when exposed to an amphibian parasite. Sci Rep. Nature Publishing Group; 2015;5: 10828. pmid:26022346
- 86. Fites JS, Ramsey JP, Holden WM, Collier SP, Sutherland DM, Reinert LK, et al. The invasive chytrid fungus of amphibians paralyses lymphocyte responses. Science (80-). 2013;342: 366–369. pmid:24136969
- 87. Martel A, Spitzen-van der Slulis A, Blooi M, Bert W, Ducatelle R, C.Fisher M, et al. Batrachochytrium salamandrivorans sp. nov. causes chytridiomicosis in amphibians. Proc Natl Acad Sci U S A. 2013;110: 15325–15329. pmid:24003137
- 88. Spitzen-van der Sluijs A, Martel A, Hallmann CA, Bosman W, Garner TWJ, Van Rooij P, et al. Environmental determinants of recent endemism of batrachochytrium dendrobatidis infections in amphibian assemblages in the absence of disease outbreaks. Conserv Biol. 2014;28: 1302–1311. pmid:24641583
- 89. McMahon T a., Sears BF, Venesky MD, Bessler SM, Brown JM, Deutsch K, et al. Amphibians acquire resistance to live and dead fungus overcoming fungal immunosuppression. Nature. Nature Publishing Group; 2014;511: 224–227. pmid:25008531
- 90. Martel A, Blooi M, Adriaensen C, Van Rooij P, Beukema W, Fisher MC, et al. Recent introduction of a chytrid fungus endangers Western Palearctic salamanders. Science (80-). 2014;346: 630–631. pmid:25359973