http://orcid.org/0000-0002-4347-6629
Figures
Abstract
The amphibian disease chytridiomycosis in amphibians is caused by the chytrid fungus Batrachochytrium dendrobatidis (Bd) and has resulted in dramatic declines and extinctions of amphibian populations worldwide. A hypervirulent, globally-dispersed pandemic lineage (Bd-GPL) is thought to be largely responsible for population declines and extinctions, although numerous endemic lineages have also been found. Recent reports of amphibian declines have been linked to the emergence of Bd in Cameroon, a major hotspot of African amphibian diversity. However, it is not known whether Bd-GPL or other lineages have been found in this region. This study therefore aims to examine Bd lineage diversity in the region and predict the distribution of this pathogen under current and future climate conditions using data from this study and from historical records. Almost 15% (52/360) of individuals tested positive for Bd using a standard quantitative PCR diagnostic. Infected amphibians were found at all eight sites sampled in this study. Species distribution models generated in BIOMOD2 indicate that areas with highest predicted environmental suitability occur in the Cameroon highlands and several protected areas throughout the country. These areas of high environmental suitability for Bd are projected to shift or decrease in size under future climate change. However, montane regions with high amphibian diversity are predicted to remain highly suitable. Phylogenetic analysis of the ITS sequences obtained from a set of positive Bd samples indicate that most fall within the Bd-GPL lineage while the remainder group with isolates from either Brazil or South Korea. Although more in depth phylogenetic analyses are needed, identification of Bd-GPL lineages in areas of high amphibian diversity emphasizes the need to continue to monitor for Bd and develop appropriate conservation strategies to prevent its further spread.
Citation: Miller CA, Tasse Taboue GC, Ekane MMP, Robak M, Sesink Clee PR, Richards-Zawacki C, et al. (2018) Distribution modeling and lineage diversity of the chytrid fungus Batrachochytrium dendrobatidis (Bd) in a central African amphibian hotspot. PLoS ONE 13(6): e0199288. https://doi.org/10.1371/journal.pone.0199288
Editor: Stefan Lötters, Universitat Trier, GERMANY
Received: March 31, 2017; Accepted: June 5, 2018; Published: June 20, 2018
Copyright: © 2018 Miller 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: ITS nucleotide sequence data are available on GenBank (https://www.ncbi.nlm.nih.gov/genbank/) through the National Center for Biotechnology Information (NCBI) (accession numbers MG601117-MG601130).
Funding: The laboratory procedures and materials for data analysis in the current project was funded by the International Herpetological Symposium (CAM), http://www.internationalherpetologicalsymposium.com, and the Tulane Graduate Student fund. Field work and sample collection was supported by a separate ongoing project funded under the US National Science Foundation Partnerships for International Research and Education (PIRE) Grant (OISE 1243524), http://www.nsf.gov. The funders had no role in study design, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The infectious disease chytridiomycosis is a leading cause of global amphibian population declines. Clinical symptoms of the disease most commonly include excessive shedding of the skin, hyperkeratosis, and skin redness or discoloration. In general, the disease is diagnosed by the presence of maturing zoosporangia of the amphibian chytrid fungus, Batrachochytrium dendrobatidis (Bd) which infects the keratin-containing layers of amphibian skin [1]. According to a recent global assessment, Bd has been detected in over 500 amphibian species [2] and is established on every continent where amphibians are found. In Africa, Bd has been documented in South Africa [3], the Albertine Rift in the Democratic Republic of Congo [4] and Uganda [5], Ethiopia [6], Kenya [7], Tanzania [8], Malawi [9], Morocco [10], and recently Mozambique [11]. Across the rainforest belt of central equatorial Africa, Bd presence has been reported in Gabon [12], Nigeria [13,14], the island of Sao Tomé [15], and Cameroon [8,16,17]. Most of these reports from Africa consist of presence/absence assessments and estimations of Bd prevalence, thus there is little known about lineage diversity within Bd in this region.
Within the Central African region, the Cameroon highlands and flanking lowland forests represent a globally-important biodiversity hotspot [18,19]. With approximately 200 anuran species described [20], Cameroon has the third highest amphibian species richness in Africa, only exceeded by Madagascar and the Democratic Republic of Congo [21]. It also has the highest percentage of endemic amphibians in mainland Africa, 88% of which occur in the highlands [22]. Studies reporting Bd in the region have not found evidence for the effects of chytridiomycosis. However, recent analyses of frogs in montane areas of Cameroon reported that community-wide amphibian declines were correlated with Bd prevalence rather than habitat loss or land use change [23]. Increasing our understanding of Bd presence and the likelihood of this pathogen spreading under future climate change throughout this amphibian hotspot is essential for proactive conservation planning.
Species distribution models (SDMs) have proven to be useful tools for predicting Bd distribution [24–26] as several environmental factors are thought to limit its range. In general, Bd presence is associated with relatively cool to moderate temperatures of between 17 and 25°C [27] although it can survive temperatures between 4 and 28°C [28]. Bd infection prevalence has been shown to increase with cooler conditions and decrease with greater temperature [26]. Population declines and extinctions in amphibian hosts are also more likely in cooler, wetter, more thermally stable regions [29] and deaths from chytridiomycosis are common in high-altitude [30] and tropical montane regions [24,25]. Bd prevalence has also been shown to vary significantly by season [31] with the highest prevalence reported for cooler months [31–33]. Bd reproduction requires moist environments and several measures of precipitation have been demonstrated as significant predictors of Bd suitability [3,26,34].
Phylogenetic analyses indicate far greater endemic diversity in Bd than has been previously recognized [35,36]. To date, diverged lineages include BdCAPE (South Africa), BdASIA-1 (which includes Korea and a single BdCH isolate from Switzerland), BdASIA-2/BdBRAZIL (Korea and Brazil) and the global panzootic lineage, Bd-GPL [36]. Bd-GPL is the most globally widespread lineage and primarily associated with amphibian die-offs, leading some researchers to suggest that it has replaced rarer, less-virulent enzootic lineages [37]. Korea has been recently confirmed as the geographic origin of Bd and its distribution linked to the global commercial trade of amphibian [36]. Several Bd samples from Gabon and the islands Bioko, São Tomé, and Príncipe were recently characterized as Bd-GPL [38]. However, our knowledge of the distribution of Bd-GPL elsewhere in the rainforests of Equatorial Africa is poor.
To better understand predicted Bd distributions in a known biodiversity hotspot in the central African country of Cameroon, we combined data from Bd surveys carried out at eight sites in Cameroon with other available regional records to model environmental suitability for Bd under current environmental conditions and future climate change scenarios. We also sequenced the ITS1-5.8S-ITS2 region of the Bd genome from a subset of samples to estimate phylogenetic relationships of Cameroonian samples to determine if Bd in Cameroon identify as endemic lineage(s) or are characteristic of the globally dispersed pandemic lineage (Bd-GPL).
Materials and methods
Ethics statement
Protocols for this study were approved by the University of New Orleans Institutional Animal Care and Use Committee (IACUC) under protocol number 12–008. Field work was approved by the Cameroon Ministry of Forestry and Wildlife (MINFOF), under research permit number 0984/PRS/MINFOF/SG/DF/SDVEF/SC. All researchers involved in field work were issued research permits to enter the national parks and wildlife reserves by the Cameroonian Ministry of Scientific Research and Innovation (MINRESI).
Field surveys
A total of 360 amphibians from eight sites (average of 46 specimens per site) across Cameroon were swabbed for Bd during nocturnal surveys carried out in May–July in 2014 and March-April in 2015 (Fig 1, S1 Table). Post-metamorphic terrestrial and arboreal frogs were collected by hand using sterile nitrile gloves. To avoid potential cross-contamination, frogs were placed in individual plastic re-sealable bags and gloves were changed between individuals. Frogs were swabbed for Bd with sterile rayon swabs (Medical Wire & Equipment Co., Corsham Wilshire UK) by stroking the ventral surfaces of their thighs, abdomen, and foot webbing approximately five times each [39]. Swabs were placed in Eppendorf tubes containing 95% ethanol and stored at room temperature prior to their transfer to the laboratory. Frogs were identified to genus or species level and then released at their point of capture.
Locations of the study sites, Campo Ma’an National Park (CM), Ebo Forest (EF), Lobéké National Park (LB), Mbam Djerem National Park (MD), Mount Cameroon village (MCV), Mount Cameroon (MC), Ndikiniméki (ND), and Takamanda (TM) are marked with yellow circles. Previously published Bd sample sites are marked with black circles. The sites sampled in this study have all been previously sampled with the exception of Mbam Djerem National Park and Takamanda. This figure was created with Natural Earth vector and raster map data (naturalearthdata.com).
Molecular diagnostics for Bd
Genomic DNA was extracted from swab samples using a Qiagen DNAeasy Blood and Tissue kit, following the manufacturer’s protocol, with a final elution volume of 200μl. A quantitative, real-time TaqMan quantitative PCR (qPCR) assay (Applied Biosystems, CA) was used to diagnose the presence and quantity of Bd from the extracted DNA following Boyle et al. [40]. Prior to amplification, each sample was diluted 1:10 with double-distilled water and 0.7μl of bovine serum albumin was added to each reaction. An internal positive control (VICTM IPC, Applied Biosystems) was added to each qPCR reaction to confirm that PCR inhibition was not affecting our results. A negative control and a series of plasmid dilution standards, with concentrations ranging from 2.1 x 100–2.1 x 106 copies per μL (Pisces Molecular) were included in each qPCR run. While qPCRs are extremely sensitive, samples collected from swabs can contain relatively low levels of DNA, hence samples were run up to three times to screen for Bd. Samples were considered positive if the qPCR assay detected > 0.9 copies of the Bd target region in a 5 μL sample of extracted DNA. Whole swab infection intensity for individual amphibians was quantified as plasmid equivalents (PE) [41–43], or DNA copies, by multiplying qPCR results by 40 to account for the portion of the 200μl extract that was not included in the qPCR, then multiplying by the dilution factor of 10. This gives an estimate of the number of DNA copies on the whole swab. Upper and lower confidence intervals (95%) for Bd prevalence were calculated using the R package PropCIs [44].
Species distribution modeling
Current and projected future SDMs of Bd in Cameroon were created using the BIOMOD2 package [45] implemented in the statistical program R, version 3.3.0 [46]. BIOMOD2 is a framework for ensemble forecasting of species distributions using species occurrence data and environmental predictor variables. It can incorporate up to 11 different modeling algorithms and generate ensemble projections that are weighted by model performance. Eight modeling algorithms in BIOMOD2 were used to assess environmental suitability for Bd: generalized linear models (GLM, [47]), generalized additive models (GAM, [48]), generalized boosting models (GBM, [49]), classification tree analysis (CTA, [50]), flexible discriminant analysis (FDA, [51]), multivariate adaptive regression splines (MARS, [52]), random forest (RF, [53]), and the maximum entropy approach modeled in Maxent [54]. Model calibration was performed using random sampling (70%) of the data as implemented in BIOMOD2. Model evaluation was carried out with the true skill statistic (TSS) [55], using the remaining 30% of the data over the 10 model replicate runs. Only models with a TSS > 0.6 were incorporated into the final ensemble model. True skill statistic scores are not affected by prevalence and range from -1 to 1, with 0 indicating no predictive ability and 1 indicating a perfect ability. Species occurrence data were comprised of Bd presence and absence points from the current study as well as from the primary literature [6,16,56]. Supplementary Bd occurrence records (187 presence points and 739 absence points) were collected from Cameroon and neighboring regions (southern Nigeria and Gabon). Only sources that used qPCR as the detection method and involved specimens collected in the field were included (S2 Table).
Nineteen bioclimatic variables [57] and elevation [58] were considered as predictors in species distribution models. Highly correlated variables (R > 0.8) were eliminated based on Pearson’s correlation coefficients using the R package virtualspecies [59]. After comparing model performance for sets of uncorrelated variables, the following predictive environmental variables were selected to model Bd distributions under present conditions: annual temperature, isothermality (temperature evenness), minimum temperature of the coldest month, mean temperature of the coldest quarter, annual precipitation, and elevation. Future projections of the bioclimatic variables [60] were created by aggregating 20 available global climate models for the four primary potential climate change scenarios, termed representative concentration pathways (RCP) 2.6, 4.5, 6.0, and 8.5, for years 2030, 2050, and 2080 based on the Intergovernmental Panel on Climate Change (IPCC) 5th assessment report at a spatial resolution of 30 arc-seconds (approximately 1km2) [61,62].
Phylogenetic analysis of Bd sequences
A 300bp region encompassing the 5.8S rRNA gene and portions of both flanking internal transcribed spacers (ITS1, ITS2) was amplified from 14 Bd positive samples using ITS primers Bd1a (5’-CAGTGTGCCATATGTCACG-3’) and Bd2a (5’-CATGGTTCATATCTGTCCAG- 3’) [63,64]. Polymerase chain reaction (PCR) assays were conducted with 2 μL of each template DNA in a total reaction volume of 12 μL. The PCR reaction mix contained 1x PCR Buffer (Invitrogen, CA), 1.5 mM MgCl2, 0.2 mM of each dNTP, 0.5 mM of each primer, and 1.25 units of Taq DNA polymerase (Invitrogen, CA).
PCR amplification was performed using an initial denaturation at 95°C for 5 min followed by 44 cycles of denaturation at 93°C for 45 s, an annealing step at 60°C for 45 s, an extension step at 72°C for 1 min, and a final extension at 72°C for 10 min. Excess primers and unincorporated nucleotides were removed with ExoSAP-IT (Affymetrix) and PCR products were Sanger sequenced in both the forward and reverse directions (Hitachi 3130xl, Applied Biosystems). Sequences were trimmed in Geneious software [65] and aligned with an additional 189 sequences from previous studies (S3 Table) using the MUSCLE alignment implemented in MEGA7 [66]. Another genus of Chytridiomycota, Kappamyces laurelensis (ITA2582), was used as the outgroup [64]. Bayesian phylogenetic inference was performed with MrBayes v3.2.6 [67] under a TIM1+G model of evolution determined as the best model using jModelTest v.2.1.1 [68,69]. Bayesian analyses were run with four MCMC chains for 10 million generations and sampled every 1000 generations. The phylogenetic tree and posterior probabilities for individual nodes was visualized using FigTree v1.4.3 [70].
Results
Spatial distribution, prevalence and infection intensity of Bd and its host associations
In total, 52 frogs representing 5 families, 11 genera, and 23 species tested positive for Bd, with an average prevalence of 14.4% (Table 1). Bd was found at all eight sites sampled across Cameroon. The Mt. Cameroon Village site in the highlands had the highest prevalence (33.3%, N = 15), whereas Ebo Forest in central Cameroon had the lowest (9.2%, N = 130). Average infection intensity was 7023 ±24109.1 PE (N = 52) (S4 Table). The sites with the highest infection intensity were Ndikiniméki (27567 ± 53910.6 PE, N = 4) and Ebo Forest (17943.6 ± 38734.5 PE, N = 12) (Table 1). The individuals with the highest infection intensities were Alexteroon obstetricians (PE = 117508), Xenopus andrei (PE = 80512) from Ebo Forest, and Arthroleptis aff. poecilonotus (PE = 108432) from Ndikiniméki. The two genera with the highest prevalence were Xenopus (42.9%, N = 7) and Arthroleptis (30.8%, N = 39) (S5 Table). However, none of the amphibians collected exhibited clinical signs of chytridiomycosis in the form of skin or behavioral abnormalities.
Prevalence (with 95% confidence intervals) and infection intensity, mean and standard deviation (SD), is reported for Campo Ma’an (CM), Ebo Forest (EF), Lobéké National Park (LB), Mbam Djerem National Park (MD), Mount Cameroon (MC), Mount Cameroon Village (MCV), Ndikiniméki (ND), and Takamanda (TM).
Species distribution modeling
Of the 80 individual models (8 algorithms * 10 evaluation runs) generated by BIOMOD2 for present environmental conditions, 17 models scored a TSS >0.6 (TSSaverage = 0.63; Kappaaverage = 0.58, ROCaverage = 0.88). These 17 models incorporated weighted runs from GBM, GAM, CTA, and RF algorithms to create a final ensemble model. In this ensemble model, annual precipitation had the highest overall contribution (17.2%), followed by annual mean temperature (16.9%), elevation (16.1%), precipitation of the warmest quarter (14.1%), minimum temperature of the coldest month (13.4%), isothermality (11.9%), and mean temperature of the coldest quarter (10.5%). Under present climate conditions, the ensemble model indicated high environmental suitability for Bd throughout the Cameroon Volcanic Line, along the coast reaching into Equatorial Guinea and Nigeria, and towards central Cameroon including Mbam Djerem National Park (Fig 2). There are also pockets of high environmental suitability in other forested regions, such as in Lobéké National Park in the southeast and in Campo Ma’an National Park in the southwest of the country.
Future models include years 2030, 2050, and 2080 under the 2.6, 4.5, 6.0, and 8.5 RCPs. The scale indicates less suitable environment (cooler colors) and most suitable environment (warmer colors).
Fig 2 shows the projection models under future climate scenarios. Projection for RCP 2.6, which assumes that global annual greenhouse gas emissions peak between 2010–2020 and decline substantially thereafter, shows slight decreases in the range of highly suitable environment for the montane region and central Cameroon, but increases in the southern regions surrounding Campo Ma’an and Lobéké National Parks from years 2030 to 2080. Projection for RCP 4.5, which assumes emissions peak at 2040 and then decline, predicts similar patterns for the montane region and areas around Campo Ma’an. However, the Mbam Djerem region is predicted to decrease in its degree of environmental suitability but increase in area, while the Lobéké region is likely to see little change. Projection for RCP 6.0, which assumes emissions peak around 2080 and then decline, again predicts an increase in suitability for the Campo Ma’an region as well as a slight reduction in highly suitable area around the highlands over time. It also shows patchy changes in high environmental suitability in the Mbam Djerem region. The projection for RCP 8.0, which assumes emissions continue to rise, shows a significant reduction in suitable area in the northeastern region of the highlands for year 2030. However, this is followed by an increase in environmental suitability in this region as well as in the area near Campo Ma’an National Park by 2050. In year 2080, suitable area around Campo Ma’an decreases dramatically, while areas in northern Cameroon increase in environmental suitability. Notably, all future projection models predict high environmental suitability for Bd in areas surrounding the highland region.
Phylogenetic analysis
The phylogenetic tree of ITS rRNA haplotypes from worldwide Bd samples recovered similar relationships as reported in previous studies [71]. Basal haplogroups are dominated by haplotypes from South Korea and Brazil. Most of the Cameroon samples (n = 11) group with the Bd-GPL lineage whereas three fall within the basal clade (Fig 3 & S1 Fig). Samples from Campo Ma’an, Ebo Forest, Lobéké, and Ndikiniméki were identified as Bd-GPL, while additional samples from Ebo Forest cluster with the more basal clade. The Bd-GPL clade has high posterior probability (PP = 0.80) and contains haplotypes originating from Japan, China, South Africa, Italy, United States, and Ecuador.
Isolates are color coded according to the country where they were collected. Samples from Cameroon are highlighted in yellow.
Discussion
Spatial distribution, prevalence and infection intensity of Bd and its host associations
Considering the threat Bd poses to amphibians, it is essential to better understand potential patterns and drivers of distribution and prevalence. This requires surveying unsampled areas as well as monitoring previously sampled areas whenever possible. The overall low prevalence across the surveyed sites in Cameroon is consistent with recent assessments for Cameroon and other African regions. Surveys from 11 sites in Cameroon between 2007 and 2011 found an average Bd prevalence of 10.9%, whereas the average prevalence in the present study was 14.4% [17]. However, Bd had not been detected in Campo Ma’an [17], while the current study detected Bd at this site with a prevalence of 15.6%. Bd prevalence is associated with strong seasonal variation in temperature [72,73] and precipitation regimes [74]. Thus, it is possible that Bd prevalence is under-estimated in this study because samples were collected between the end of the dry season and beginning of the rainy season in Cameroon (March–July), when conditions are less suitable for infection.
Bd has been detected consistently at high elevations in Cameroon and neighboring regions. Here, Bd was detected at 23.5% prevalence at the higher elevation site on Mt. Cameroon. This is similar to the levels of prevalence estimated from more extensive field work on Mt. Oku and Mt. Manengouba in Cameroon [23] as well as in the Albertine Rift [75]. However, only one species (Wolterstorffina parvipalmata) was found at comparatively high abundance. This semi-arboreal species lays its eggs in water filled tree holes [76]. This behavior could facilitate Bd persistence in an area with low amphibian host diversity.
Species distribution modeling
Areas of environmental suitability observed from the SDMs are those with increased annual precipitation, overall warmer temperatures, and at higher elevation. These environmental conditions are found in several regions of Cameroon and are highly suitable for Bd, causing potential concern for the diverse amphibian communities inhabiting these regions [77]. Under most climate change projections, Campo Ma’an, Mbam Djerem, and Lobéké National Parks are predicted to retain areas with high environmental suitability. Prevalence is currently relatively low for these areas and amphibian declines have not been reported, making the apparent risk of Bd outbreaks low. However, considering these are protected areas, we should pay close attention to amphibian communities in light of the current Bd presence and the future changing environments.
The Cameroon highlands, a hotspot of African amphibian diversity and a region experiencing amphibian declines, consistently make up a large proportion of suitable environment for Bd. Further, our models of future climate scenarios suggest that high environmental suitability in the montane regions is likely to persist into the future, likely driven by increasing average annual temperatures. For montane frogs, research suggests that the incidence of disease depends heavily on the effect of temperature on the host’s resistance to chytridiomycosis, rather than an effect of the fungus alone [78]. Many amphibians have narrow elevational ranges [79], and may become more vulnerable to disease outbreak due to shifts in temperature and precipitation that are favorable to the pathogen. This effect may be further compounded by the simultaneously decrease in available habitat for its montane amphibian hosts. Further work is needed to increase knowledge of amphibian communities and their distributions along elevational gradients in Cameroon to better understand patterns of infection at different elevations.
The SDMs presented here predict that Cameroon is likely to experience considerable shifts in the overall area and degree of environmental suitability for Bd under future climate conditions. While the climate-linked epidemic hypothesis has not been supported in areas of Central and South America [80], climate change has been linked to infection susceptibility in other areas where Bd has been introduced and causing amphibian declines, such as western Europe[81]. Environmental factors may be more reliable predictors of Bd in areas where is not alien and thus can help predict environmental suitability. In addition to understanding environmental factors limiting Bd’s distribution, future research in this region should explore host-related factors, such as variation in infection susceptibility, that may influence Bd’s success in an area and incorporate those factors into predictive models.
Phylogenetic analysis
Molecular analyses show that the hypervirulent Bd-GPL was found at four of the eight sampling sites. These are all lower elevation forested sites in the center, south and east of the country. Two of these sites where the hypervirulent lineage is found (Ebo Forest and Ndikiniméki) could potentially be linked to populations declines in montane areas reported by Hirschfeld et al [23]. On the other hand, since Bd-GPL is considered to have arrived prior to recent amphibian population declines [35], amphibians in this region might offer insight into mechanisms of Bd tolerance or infection avoidance if current declines are not the result of Bd-GPL.
Other Bd samples in this study cluster with the lineages from Brazil and South Korea. While neither Brazilian nor South Korean haplotypes have been associated with amphibian declines [71], understanding how these haplotypes spread to or from Cameroon could offer insight into Bd dispersal. The presence of these global lineages could also be explained by the Cameroon wildlife and pet trade. Historical trade in native Xenopus laevis has been suggested as potential mechanism for the global spread of Bd [82]. Bd has also been detected in African caecilians from Cameroon involved in the amphibian pet trade [8] and infected Cameroon consignments of amphibians are considered very likely to have introduced of Bd into the United Kingdom [83]. Considering Bd is extremely resilient and can grow on a variety of potential carriers, such as dead amphibian and snake skin, dead algae, insect exoskeletons and feathers [84], its transmission from and into wild populations through trade or scientific facilities is not unlikely. It is also possible that the similarities observed between haplotypes from Cameroon, Brazil and South Korea are due to the retention of shared ancestral variation that has not been revealed due to the lack of sampling from many areas across the world.
Although existing primers have been shown to be highly specific to the Bd ITS1-5.8S-ITS2 region [63,64,71,85,86], the derived phylogenies assembled with the ITS region should be observed with caution because of this region’s high copy number, short length and lack of phylogenetic informativeness [86]. Future work should subject samples from Central Africa to a more rigorous genome-wide sequencing to better define the phylogenetic relationships of global samples.
Conclusion
While there is still much to be discovered regarding Bd diversity and host-pathogen-environment interactions, the present study provides further insight into genetic diversity and environmental suitability of Bd in an important amphibian biodiversity hotspot in Equatorial Africa. The Cameroon highlands, and specifically Mt. Cameroon, are designated as high conservation priority areas [87]. Presence of the hypervirulent Bd-GPL in this region, while not detected at the montane site, could be potentially linked to the community declines observed in the highlands. With ongoing habitat destruction and degradation, as well as reductions of species climate envelope, amphibians could be pushed into smaller pockets of habitat that could potentially exacerbate Bd transmission due to over-crowding. Much progress has been made monitoring amphibian populations in the light of Bd emergence, but more work is needed to fully understand the environmental drivers of habitat suitability and map more precisely the global distribution of Bd lineages. To further understand the level of threat that Bd poses to Central African amphibians it is important for future research to further characterize Bd lineages in this region and determine how susceptible local amphibians are to different strains. Infectious diseases, primarily chytridiomycosis, are considered a major threat to amphibian populations so understanding which factors contribute to or limit infection across the globe is essential for amphibian conservation.
Supporting information
S1 Table. Species sampled at eight sites in Cameroon.
https://doi.org/10.1371/journal.pone.0199288.s001
(XLSX)
S2 Table. Current and supplementary Bd occurrence locations used in BIOMOD2 analyses.
https://doi.org/10.1371/journal.pone.0199288.s002
(XLSX)
S3 Table. Additional Bd sequences used in phylogenetic analysis.
https://doi.org/10.1371/journal.pone.0199288.s003
(XLSX)
S4 Table. Infection intensities of Bd positive samples from Cameroon.
https://doi.org/10.1371/journal.pone.0199288.s004
(XLSX)
S5 Table. Bd prevalence within amphibian groups from eight sites in Cameroon.
https://doi.org/10.1371/journal.pone.0199288.s005
(XLSX)
Acknowledgments
We would like to acknowledge the Cameroonian Ministry of Forestry and Wildlife (MINFOF) and the Ministry of Scientific Research and Innovation (MINRESI) for issuing research, collection and export permits, and helping coordinate field expeditions. We also thank the village chiefs and community elders who permitted us to work in their community forests. We are grateful to our many field guides and assistants for their momentous help. Finally, we would like to thank two of the authors of BIOMOD2, Wilfried Thuiller and Damien Georges, for their technical assistance and troubleshooting.
References
- 1. Van Rooij P, Martel A, Haesebrouck F, Pasmans F. Amphibian chytridiomycosis: A review with focus on fungus-host interactions. Vet Res. 2015;46(1):1–22.
- 2. Berger L, Roberts AA, Voyles J, Longcore JE, Murray KA, Skerratt LF. History and recent progress on chytridiomycosis in amphibians. Fungal Ecol. 2016;19:89–99.
- 3. Tarrant J, Cilliers D, du Preez LH, Weldon C. Spatial Assessment of Amphibian Chytrid Fungus (Batrachochytrium dendrobatidis) in South Africa Confirms Endemic and Widespread Infection. PLoS One. 2013;8(7).
- 4. Greenbaum E, Meece J, Reed KD, Kusamba C. Extensive occurence of the amphibian chytrid fungus in the Albertine, a Central African amphibian hotspot. Herptetological J. 2015;25:91–100.
- 5. Goldberg TL, Readel AM, Lee MH. Chytrid fungus in frogs from an equatorial African montane forest in western Uganda. J Wildl Dis. 2007;43(3):521–4. pmid:17699093
- 6. Gower DJ, Doherty-Bone TM, Aberra RK, Mengistu A, Schwaller S, Menegon M, et al. High prevalence of the amphibian chytrid fungus (Batrachochytrium dendrobatidis) across multiple taxa and localities in the highlands of Ethiopia. Herpetol Notes. 2012;22:225–33.
- 7. Kielgast J, Röder D, Veith M, Lötters S. Widespread occurrence of the amphibian chytrid fungus in Kenya. Anim Conserv. 2010;13:36–43.
- 8. Gower DJ, Doherty-Bone T, Loader SP, Wilkinson M, Kouete MT, Tapley B, et al. Batrachochytrium dendrobatidis Infection and Lethal Chytridiomycosis in Caecilian Amphibians (Gymnophiona). Ecohealth. 2013;10:173–83. pmid:23677560
- 9. Conradie W, Harvey J, Kotzé A, Dalton DL, Cunningham MJ. Confirmed amphibian chytrid in mount Mulanje Area, Malawi. Herpetol Rev. 2011;42(3):369–71.
- 10. El Mouden EH, Slimani T, Donaire D, Fernández-Beaskoetxea S, Fisher MC, Bosch J. First record of the Chytrid fungus Batrachochytrium dendrobatidis in North Africa. Herpetol Rev. 2011;42(1):71–5.
- 11. Conradie W, Bittencourt-Silva GB, Loader SP, Dalton DL, Tolley KA. Batrachochytrium dendrobatidis Survey of Amphibians in the Northern Mozambique “Sky Islands” and Low-lying Areas. Herpetol Rev. 2016;47(1):42–6.
- 12. Bell RC, Gata Garcia AV., Stuart BL, Zamudio KR. High prevalence of the amphibian chytrid pathogen in Gabon. Ecohealth. 2011;8(1):116–20. pmid:21210295
- 13. Imasuen AA, Aisien MSO, Weldon C, Dalton DL, Kotze A, Du Preez LH. Occurrence of batrachochytrium dendrobatidis in amphibian populations of Okomu national park, Nigeria. Herpetol Rev. 2011;42(3):379–82.
- 14. Reeder NMM, Cheng TL, Vredenburg VT, Blackburn DC. Survey of the chytrid fungus Batrachochytrium dendrobatidis from montane and lowland frogs in eastern Nigeria. Herpetol Notes. 2011;4(1):83–6.
- 15. Hydeman ME, Bell RC, Drewes RC, Zamudio KR. Amphibian chytrid fungus confirmed in endemic frogs and caecilians on the Island of São Tomé, Africa. Herpetol Rev. 2013;44(2):254–7.
- 16. Baláz V, Kopecky O, Gvozdik V. Presence of the amphibian chytrid pathogen confirmed in Cameroon. Herptetological J. 2012;22:191–4.
- 17. Doherty-Bone TM, Gonwouo NL, Hirschfeld M, Ohst T, Weldon C, Perkins M, et al. Batrachochytrium dendrobatidis in amphibians of Cameroon, including first records for caecilians. Dis Aquat Organ. 2013;102(3):187–94. pmid:23446968
- 18. Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GA, Kent J. Biodiversity hotspots for conservation priorities. Nature. 2000;403(6772):853–8. pmid:10706275
- 19. Mittermeier RA, Mittermeier CG, Brooks TM, Pilgrim JD, Konstant WR, Fonseca GAB, et al. Wilderness and biodiversity conservation. Proc Natl Acad Sci U S A. 2003;100(18):10309–13. pmid:12930898
- 20.
Amiet J. Frog Diversity in Cameroon. In: Stuart S, Hoffmann M, Chanson J, Cox N, Berridge R, Ramani P, et al., editors. Threatened amphibians of the world Lynx Edicions with IUCN—The World Conservation Union, Conservation International and NatureServe. Barcelona; 2008. p. 16.
- 21.
IUCN 2017. The IUCN Red List of Threatened Species. Version 2017–2. Available from: http://www.iucnredlist.org
- 22. Bergl RA, Oates JF, Fotso R. Distribution and protected area coverage of endemic taxa in West Africa’ s Biafran forests and highlands. Biol Conserv. 2016;134:195–208.
- 23. Hirschfeld M, Blackburn DC, Doherty-Bone TM, Gonwouo LN, Ghose S, Rödel M-O. Dramatic Declines of Montane Frogs in a Central African Biodiversity Hotspot. PLoS One. 2016;11(5):1–15.
- 24. Rödder D, Kielgast J, Lötters S. Future potential distribution of the emerging amphibian chytrid fungus under anthropogenic climate change. Dis Aquat Organ. 2010;92(2–3):201–7. pmid:21268982
- 25. 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.
- 26. 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(2).
- 27. Voyles J, Johnson LR, Briggs CJ, Cashins SD, Alford RA, Berger L, et al. Experimental evolution alters the rate and temporal pattern of population growth in Batrachochytrium dendrobatidis, a lethal fungal pathogen of amphibians. Ecol Evol. 2014;4(18):3633–41. pmid:25478154
- 28. Piotrowski JS, Annis SL, Longcore JE. Physiology of Batrachochytrium dendrobatidis, a Chytrid Pathogen of Amphibians. Mycologia. 2004;96(1):9. pmid:21148822
- 29. Murray KA, 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. 201;48:163–73.
- 30. Lips KR. Decline of a Tropical Montane Amphibian Fauna. Conserv Biol. 1998;12(1):106–17.
- 31. Kriger KM, Hero JM. Large-scale seasonal variation in the prevalence and severity of chytridiomycosis. J Zool. 2007;271(3):352–9.
- 32. Thomson DM. Matrix models as a tool for understanding invasive plant and native plant interactions. Conserv Biol. 2005;19(3):917–28.
- 33. Phillott AD, Grogan LF, Cashins SD, Mcdonald KR, Berger L, Skerratt LF. Chytridiomycosis and seasonal mortality of tropical stream-associated frogs 15 years after introduction of batrachochytrium dendrobatidis. Conserv Biol. 2013;27(5):1058–68. pmid:23678872
- 34. Penner J, Adum GB, McElroy MT, Doherty-Bone T, Hirschfeld M, Sandberger L, et al. West Africa—A Safe Haven for Frogs? A Sub-Continental Assessment of the Chytrid Fungus (Batrachochytrium dendrobatidis). PLoS One. 2013;8(2).
- 35. 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 USA. 2013;110(23):9385–90. pmid:23650365
- 36. 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;627:621–7.
- 37. 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(18):4079–97. pmid:26445660
- 38. Hydeman ME, Longo AV, Velo-Antón G, Rodriguez D, Zamudio KR, Bell RC. Prevalence and genetic diversity of Batrachochytrium dendrobatidis in Central African island and continental amphibian communities. Ecol Evol. 2017;7(19):7729–38. pmid:29043029
- 39. Hyatt AD, Boyle DG, Olsen V, Boyle DB, Berger L, Obendorf D, et al. Diagnostic assays and sampling protocols for the detection of Batrachochytrium dendrobatidis. Dis Aquat Organ. 2007;73(3):175–92. pmid:17330737
- 40. 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(2):141–8. pmid:15460858
- 41. Robak MJ, Richards-Zawacki CL. Temperature-dependent effects of cutaneous bacteria on a frog’s tolerance of fungal infection. Front Microbiol. 2018;9:1–12.
- 42. Voyles J, Woodhams DC, Saenz V, Bryne AQ, Perez R, Rios-Sotelo G, et al. Shifts in disease dynamics in a tropical amphibian assemblage are not due to pathogen attenuation. Science. 2018;359:1517–9. pmid:29599242
- 43. Sonn JM, Berman S, Richards-Zawacki CL. The Influence of Temperature on Chytridiomycosis In Vivo. Ecohealth. 2017;14(4):762–70. pmid:28879516
- 44.
Scherer R. PropCIs: Various confidence interval methods for proportions. R package version 0.2–5; 2016.
- 45. Thuiller W. BIOMOD—Optimizing predictions of species distributions and projecting potential future shifts under global change. Glob Chang Biol. 2014;9(10):1353–62.
- 46.
R Core Team. R: A Language and Environment for Statistical Computing. Vienna: R Foundation for Statistical Computing; 2016.
- 47.
McCullagh P, Nelder JA. Generalized Linear Models. Chapman and Hall; 1989.
- 48.
Hastie TJ, Tibshirani RJ. Generalized Additive Models. Chapman and Hall; 1990.
- 49. Ridgeway G. The state of boosting. Comput Sci Stat. 1999;31:172–81.
- 50.
Breiman L, Friedman J, Stone CJ, Olshen RA. Classification and regression trees. Chapman and Hall; 1984.
- 51. Hastie T, Tibshirani R, Buja A. Flexible Discriminant Analysis by Optimal Scoring. J Am Stat Assoc. 1994;89(428).
- 52. Friedman J. Multivariate Adaptive Regression Splines. Ann Stat. 1991;19(1):1–141.
- 53. Breiman LEO. Random Forests. Mach Learn. 2001;45:5–32.
- 54. Phillips SJ, Anderson RP, Schapire RE. Maximum entropy modeling of species geographic distributions. Ecol Modell. 2006;190:231–59.
- 55. Allouche O, Tsoar A, Kadmon R. Assessing the accuracy of species distribution models: prevalence, kappa and the true skill statistic (TSS). J Appl Ecol. 2006;43:1223–32.
- 56. Doherty-Bone TM, Bielby J, Lebreton M, Cunningham AA. In a vulnerable position? Preliminary survey work fails to detect the amphibian chytrid pathogen in the highlands of Cameroon, an amphibian hotspot. Herptetological J. 2008;18:115–8.
- 57. 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(15):1965–78.
- 58. Jarvis A, Reuter A H.I., Nelson A, Guevara E. Hole-filled SRTM for the globe Version 4, available from the CGIAR-CSI SRTM 90m Database. CGIAR CSI Consort Spat Inf. 2008;1–9.
- 59. Leroy B, Meynard CN, Bellard C, Courchamp F. virtualspecies, an R package to generate virtual species distributions. Ecography. 2016;39(6):599–607.
- 60.
Sesink Clee PR. Distributions, Drivers, and Risks of Wildlife Infectious Diseases Across Africa: Using Geospatial Analyses to Elucidate Disease Occurrence in Biodiversity Hotspots (Doctoral dissertation). Philadelphia: Drexel University; 2017.
- 61.
IPCC. Climate change 2014–impacts, adaptation and vulnerability: Regional aspects. Cambridge University Press; 2014.
- 62.
Ramirez J, Jarvis A. High resolution statistically downscaled future climate surfaces. International Center for Tropical Agriculture (CIAT); CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS). Cali, Colombia. 2008.
- 63. Annis SL, Dastoor FP, Ziel H, Daszak P, Longcore JE. A DNA-based assay identifies Batrachochytrium dendrobatidis in amphibians. J Wildl Dis. 2004;40(3):420–8. pmid:15465708
- 64. Goka K, Yokoyama J, Une Y, Kuroki T, Suzuki K, Nakahara M, et al. Amphibian chytridiomycosis in Japan: Distribution, haplotypes and possible route of entry into Japan. Mol Ecol. 2009;18(23):4757–74. pmid:19840263
- 65. Kearse M, Moir R, Wilson A, Stones-Havas S, Cheung M, Sturrock S, et al. Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics. 2012;28(12):1647–9. pmid:22543367
- 66. Kumar S, Stecher G, Tamura K. MEGA7: Molecular Evolutionary Genetics Analysis Version 7.0 for Bigger Datasets. Mol Biol Evol. 2016;33(7):1870–4. pmid:27004904
- 67. Ronquist F, Huelsenbeck JP. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics. 2003;19(12):1572–4. pmid:12912839
- 68. Guindon S, Gascuel O. A Simple, Fast, and Accurate Method to Estimate Large Phylogenies by Maximum Likelihood. Syst Biol. 2003;52(5):696–704. pmid:14530136
- 69. Darriba D, Taboada GL, Doallo R, Posada D. jModelTest 2: more models, new heuristics and parallel computing. Nat Methods. 2012;9(722).
- 70.
FigTree. Available from: http://tree.bio.ed.ac.uk/software/figtree.
- 71. 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(16):4196–209. pmid:23802586
- 72. Kriger KM, Hero J-M. Large-scale seasonal variation in the prevalence and severity of chytridiomycosis. J Zool. 2006;271:352–9.
- 73. Murray KA, Skerratt LF, Garland S, Kriticos D, McCallum H. Whether the Weather Drives Patterns of Endemic Amphibian Chytridiomycosis: A Pathogen Proliferation Approach. PLoS One. 2013;8(4).
- 74. Ron SR. Predicting the Distribution of the Amphibian Pathogen Batrachochytrium dendrobatidis in the New World. Biotropica. 2005;37(2):209–21.
- 75. Seimon TA, Ayebare S, Sekisambu R, Muhindo E, Mitamba G, Greenbaum E, et al. Assessing the threat of amphibian chytrid fungus in the Albertine Rift: Past, present and future. PLoS One. 2015;10(12):1–24.
- 76. Hofer U, Bersier L-F, Borcard D. Ecotones and Gradient as Determinants of Herpetofaunal Community Structure in the Primary Forest of Mount Kupe, Cameroon. J Trop Ecol. 2000;16(4):517–33.
- 77. Brooks T, Balmford A, Burgess N, Fjeldså J, Hansen LA, Moore J, et al. Toward a Blueprint for Conservation in Africa. Bioscience. 2001;51(8):613.
- 78. Andre SE, Parker J, Briggs CJ. Effect of temperature on host response to Batrachochytrium dendrobatidis infection in the mountain yellow-legged frog (Rana muscosa). J Wildl Dis. 2008;44(3):716–20. pmid:18689660
- 79. Wake DB. Facing Extinction in Real Time. Ecology. 2012;335:1052–4.
- 80. Lips KR, Diffendorfer J, Mendelson JR, Sears MW. Riding the wave: Reconciling the roles of disease and climate change in amphibian declines. PLoS Biol. 2008;6(3):0441–54.
- 81. Clare FC, Halder JB, Daniel O, Bielby J, Semenov MA, Jombart T, et al. Climate forcing of an emerging pathogenic fungus across a montane multi-host community. Philos Trans R Soc B Biol Sci. 2016;371(1709):20150454.
- 82. Weldon C, Preez LH, Hyatt AD, Spearet R. Origin of the Amphibian Fungus. Emerg Infect Dis. 2004;10(12):2100–6. pmid:15663845
- 83.
Wombwell, Louise E. Emerging Infectious Disease and the Trade in Amphibians. Doctor of Philosophy (PhD) thesis, University of Kent, Durrell Institute of Conservation and Ecology. 2014.
- 84. Johnson ML, Speare R. Possible modes of dissemination of the amphibian chytrid Batrachochytrium dendrobatidis in the environment. Dis Aquat Organ. 2005;65(3):181–6. pmid:16119886
- 85. Gaertner JP, Forstner MRJ, O’Donnell L, Hahn D. Detection of batrachochytrium dendrobatidis in endemic salamander species from central texas. Ecohealth. 2009;6(1):20–6. pmid:19424755
- 86. 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(21):5162–77. pmid:22857789
- 87.
World Wildlife Fund. A Vision for Biodiversity Conservation in Central Africa: Biological Priorities for Conservation in the Guinean-Congolian Forest and Freshwater Region. Toham AK, D’Amico J, Olson D, Blom A, Neil B, Leann Trowbridge, et al., editors. 2006.