Introduced species have the potential to become invasive and jeopardize entire ecosystems. The success of species establishing viable populations outside their original extent depends primarily on favorable climatic conditions in the invasive ranges. Species distribution modeling (SDM) can thus be used to estimate potential habitat suitability for populations of invasive species. Here we review the status of six amphibian species with invasive populations in Brazil (four domestic species and two imported species). We (i) modeled the current habitat suitability and future potential distribution of these six focal species, (ii) reported on the disease status of Eleutherodactylus johnstonei and Phyllodytes luteolus, and (iii) quantified the acoustic overlap of P. luteolus and Leptodactylus labyrinthicus with three co-occurring native species. Our models indicated that all six invasive species could potentially expand their ranges in Brazil within the next few decades. In addition, our SDMs predicted important expansions in available habitat for 2 out of 6 invasive species under future (2100) climatic conditions. We detected high acoustic niche overlap between invasive and native amphibian species, underscoring that acoustic interference might reduce mating success in local frogs. Despite the American bullfrog Lithobates catesbeianus being recognized as a potential reservoir for the frog-killing fungus Batrachochytrium dendrobatidis (Bd) in Brazil, we did not detect Bd in the recently introduced population of E. johnstonei and P. luteolus in the State of São Paulo. We emphasize that the number of invasive amphibian species in Brazil is increasing exponentially, highlighting the urgent need to monitor and control these populations and decrease potential impacts on the locally biodiverse wildlife.
Citation: Forti LR, Becker CG, Tacioli L, Pereira VR, Santos ACFA, Oliveira I, et al. (2017) Perspectives on invasive amphibians in Brazil. PLoS ONE 12(9): e0184703. https://doi.org/10.1371/journal.pone.0184703
Editor: Stefan Lötters, Universitat Trier, GERMANY
Received: November 4, 2016; Accepted: August 29, 2017; Published: September 22, 2017
Copyright: © 2017 Forti 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: The audio files used in this study are publicly available from the Zenodo repository at the following URL: https://zenodo.org/record/883621#.Waq5hciGOMo. All other relevant data are available from the Figshare repository at the following URL: https://doi.org/10.6084/m9.figshare.5393977.
Funding: The São Paulo Research Foundation (FAPESP) provided financial support (grant #2011/51694-7, #2011/52070-7, #2014/23388-7) to LFT. The National Council for Scientific and Technological Development (CNPq) provided support for LFT (#302589/2013-9; #405285/2013-2), to CGB (#312895/2014-3), to ISO (#161812/2011-2), and to LRF (#438675/2016-9). LRF and LT were supported by grants #2013/21519-4 and #2013/02219-0, São Paulo Research Foundation (FAPESP). CFBH was supported by grants #2013/50741-7 and #2014/50342-8, São Paulo Research Foundation (FAPESP) and CNPq (302518/2013-4) for a research fellowship. Coordination for the Improvement of Higher Education Personnel (CAPES) provided support to ISO (#3855/13-9). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Invasive species represent a new element within a long-term naturally established biological community [1, 2]. This often causes ecological imbalances, as native predators, competitors, or endemic pathogens and parasites may not have a strong negative impact on invasive species . Therefore, biological invasions have the potential to cause changes in the distribution of pathogens , promote gene invasion , and modify the interactions between native species and the structure of natural landscapes [6, 7, 8, 9, 10]. As a consequence, there may be loss of genetic variability and high mortality among native populations. Beyond damaging the environment, populations of invasive species may also pose a threat to human economy and health [11, 12, 13]. As a result, invasion of alien species can cause multiple hazards, which are usually difficult to manage.
The success of invasive species in a new environment depends on a variety of factors, such as their ability to establish and complete their life cycle in the new habitat without their original higher genetic variability [14, 15, 16, 17]. Invasive species may also displace native species out of their original niches, occupy niches that are not used by native species, or sometimes create their own niches . Although a myriad of abiotic and biotic factors dictate the likelihood of biological invasions [18, 19, 20], macro-environmental conditions are usually the main filter for establishment [21, 22]. Environmental modeling is thus a useful tool to predict the likelihood of invasions and also predict current and future invasive population trends [23, 24, 25].
Several amphibians are known for their success as invasive species . Examples include Eleutherodactylus coqui and E. planirostris, originally from the Caribbean and accidentally introduced to Hawaii with the horticultural trade [26, 27]; the cane toad (Rhinella marina), from Central and South America, introduced to Australia and in some of the Pacific and Atlantic Islands for pest control [28, 29, 30, 31]; the American bullfrog (Lithobates catesbeianus), introduced to Europe, Asia, and South America, mostly for frog farming [32, 31]; and Xenopus laevis, an African species introduced in many countries as an animal model for lab research [33, 34]. These species are linked to evolutionary changes in morphology and behavior of native species , disease spread [36, 37, 38], and above all, to a variety of ecosystem imbalances and biodiversity loss [2, 39].
In Brazil, the only invasive amphibian species with confirmed impacts on biodiversity is the American bullfrog, L. catesbeianus [40, 41]. Originally from North America, this species was brought to South America for farming both to supply local restaurants and for the international trade [42, 36, 32]. Among the main impacts observed in the local fauna, L. catesbeianus negatively impacts native populations [43, 44, 45, 39], causes acoustic interference , and competes for resources . In addition, this species is linked to disease spread in several continents, including South America [36, 37].
Three other frog species were recently added to the list of introduced species in Brazil (translocated by human action): E. johnstonei in the city of São Paulo , Phyllodytes luteolus in the city of Rio de Janeiro , and Leptodactylus labyrinthicus in central Amazonia . There are also records of two invasive amphibian species in the archipelago of Fernando de Noronha  as well as unconfirmed records of introductions of Xenopus laevis and Pipa carvalhoi [52, 53, 2, 54]. The potential impacts of these species are yet to be revealed.
Here we review the status of the six amphibian species with confirmed invasive populations in Brazil. To assess potential negative impacts of these invasive species we (1) modeled current and potential future habitat suitability using species distribution modeling (SDM), identifying suitable habitats for population establishment and potential invasion routes. We also (2) analyzed acoustic niche overlap between local and invasive species to infer about call interference that likely reduces mating success. Finally, we (3) screened individuals of Eleutherodactylus johnstonei and Phyllodytes luteolus for the frog-killing fungus Batrachochytrium dendrobatidis (Bd) to identify potential Bd reservoirs. Combined, these goals provide additional ecological information to guide conservation efforts and manage biological invasions that may aid in mitigation programs to safeguard the local fauna.
Materials and methods
We used a combination of field surveys, literature review, personal communications, free database assessment (Sistema de Informação Distribuído para Coleções Biológicas <http://splink.cria.org.br/> and Global Biodiversity Information Facility <http://www.gbif.org/>) to gather spatial and biological information of invasive species in Brazil (further details in S1 and S2 Tables). Although analyses were focused on potential environmental niches in Brazil for discussions on invasion processes, we gathered available geographical information of the whole distributional range for each species, including localities beyond Brazil’s borders (e.g., occurrence records of Scinax x-signatus in Colombia and Guyana, the original distribution of Eleutherodactylus johnstonei in the Caribbean region, and current available records for Lithobates catesbeianus in North America, Europe and Asia). This procedure was done to capture the species’ full climatic niche, and then project it to the Brazilian territory for interpretation.
To reduce potential taxonomic bias or misidentification we only selected species clearly named in the original database accessed, excluding vague identifications, such as Genus plus sp., cf., or aff. We considered ‘introduced species’ those translocated by human action (accidental or intentional) to areas outside the species’ original distribution . We recognized three categories of introduced species: (1) imported invader–a species that occurs originally abroad with stable population(s) in Brazil; (2) domestic invader–a species that is originally from Brazil with known stable populations outside its original occurrence area; and (3) domestic alien–a species with the same condition of domestic invader, but with unknown population status (i.e., demography).
We sampled individuals of the only known invasive population of Eleutherodactylus johnstonei in Brazil: Brooklin, São Paulo city (23.633° S, 46.682° W). We collected 80 individuals of E. johnstonei from three locations ~80m apart in January 2015. We sampled individuals of Phyllodytes luteolus in Serra do Guararu, municipality of Guarujá (a coastal island), state of São Paulo, southeastern Brazil (23.883° S, 46.166°W), between December 2013 and February 2014. We collected individuals of P. luteolus in a garden with 320 bromeliads at a private neighborhood. All collected tadpoles and post-metamorphs were deposited at Museu de Zoologia “Prof. Adão José Cardoso”, Universidade Estadual de Campinas (Unicamp), Campinas, São Paulo, Brazil (ZUEC 20931–57, 21245). The following permits were approved for sample collection (Ministério do Meio Ambiente–ICMBio—SISBIO permit number: 42817–4, and Comissão de Ética no Uso do Animal (CEUA)–UNESP #39/2015).
Although southern Brazil has been pointed to as a suitable environment for the establishment of Xenopus laevis , there are no official records of established populations of this species in the wild . Due to lack of scientific information supporting the introduction of X. laevis and Pipa carvalhoi in the wild (as speculated by [52, 53, 2]), we did not consider these species as invasive in Brazil.
Species distribution modeling
Occurrence records were inspected and processed in the ArcGIS platform (ESRI, 2010) to avoid data clustering (e.g., many points in the same locality). To build the SDMs we used all available occurrence records of both native and introduced populations, as well as records of species’ original distribution to capture the climatic niche closest to the species’ realized niche (e.g. ). Determination of native and introduced populations was done by crossing occurrence records and distributional maps available in AmphibiaWeb <http://amphibiaweb.org/> and IUCN Red List of Threatened Species <http://www.iucnredlist.org/>. Models were generated for the complete geographic extent of species occurrences (native and introduced) and then projected within Brazil’s political boundaries.
We downloaded interpolated bioclimatic rasters for the current years (1950 to 2000) from WorldClim <http://www.worldclim.org> at a grid cell resolution of 10 arc-minutes. We also downloaded rasters of projected future climatic conditions downscaled from global climate model (GCM) for representative concentration pathways (RCPs) to the year 2100 from WorldClim, obtained from CMIP5 (IPCC fifth assessment; ). We used the Hadley Global Environment Model 2 –Atmosphere Ocean (HADGEM2 –AO), RCP4.5 scenario, which predicts a global surface temperature likely increasing by more than 1.5°C . We chose RCP4.5 scenario because it is parsimonious and likely occurring at the end of the 21st century. Also, we avoided extreme scenarios as RCP8.5, which predicts an increase of global mean surface temperature by the order of almost 5°C . To avoid bias due to multi-collinearity among environmental rasters, we used principal component analysis (PCA) to consolidate cross-correlated bioclimatic variables (we recovered at least 80% of variation with 4 principal components for all species, using all 19 bioclimatic variables). These analyses were performed in R  by the use of the following packages: alphashape3d, version 1.2 , raster version 2.5–8 , and rgdal version 1.1–10 .
Models were run with a calibration test including 10% of total occurrence records for each species by random sub-sampling, with 15 replicates and maximum interactions of 5000 in attempt to maximize model performance. Furthermore, SDMs were statistically evaluated by the inspection of Area Under the Receiver Operating Characteristic (ROC) Curve value (AUC). AUC values close to 0.5 indicate model performance no better than random, while AUC values close to 1.0 indicate good model performance .
We built species distribution models (SDMs)  for six species, namely the domestic invaders: Leptodactylus labyrinthicus, Phyllodytes luteolus, Rhinella jimi, and Scinax x-signatus; and also for the imported invaders: Eleutherodactylus johnstonei and Lithobates catesbeianus. To assess the potential habitat suitability at both current and future climatic conditions, and to investigate potential habitats for population establishment and/or potential invasion routes, we built SDMs using the low-memory multinomial logistic regression, also known as maximum entropy modeling method , implemented in the software MaxEnt (version 3.3.3.k). MaxEnt is a largely used machine learning algorithm that builds models of environmental suitability by the use of presence-only species records to project probability densities in covariate space . All models were calibrated with the native range of the species (e.g.). Then, the common resulting output, a continuous map provided by MaxEnt, was transformed into binary maps (0/1) to better discriminate suitable from non-suitable habitats and also to diminish uncertainty and underestimation of suitable extents (e.g.[64, 55, 65]). We used max SSS approach (which is based on sensitivity-sensibility sum maximization) as a threshold selection method [66, 67, 68]. In addition, using the binary categorization of the study grid for both models (present and future), we quantified percentage retraction or expansion for each species occurrence range using the reclass function in the Spatial Analyst tool available in Esri/ArcMap platform.
The advertisement calls (sensu ) of three adult males of Phyllodytes luteolus from Guarujá were recorded to confirm species identity and to assess the acoustical overlap with a sympatric and supposedly competitive native species (Ischnocnema sp.). Both species were recorded calling at the same time and location. We obtained vocalization data of two males of P. luteolus (with an Ischnocnema male calling in the background) using a Marantz PMD660 digital recorder at a sampling rate of 48 kHz and 16 bit resolution, and using a microphone YOGA 9600EM positioned about 50 cm from the calling males. We recorded a third male of P. luteolus using a Tascam DR-680 recorder with sampling rate of 48 kHz and 24 bit resolution, using a Sennheiser ME67 microphone positioned about 50 cm from the calling male. Vocalizations obtained in the field were archived at Fonoteca Neotropical Jacques Vielliard (FNJV), Unicamp, Campinas, São Paulo (FNJV 30962–64). Additional recordings of Leptodactylus pentadactylus (FNJV 11042) and L. labyrinthicus (FNJV 13048) were obtained from the FNJV archives. We also downloaded the recordings of Leptodactylus knudseni (CENBAM 0645) from the website CENBAM database (http://ppbio.inpa.gov.br). We compared the bioacoustic profile of these three Leptodactylus species. With the invasive population of L. labyrinthicus, these three species are now sympatric in Manaus, state of Amazonas. We generated sonograms in Raven 1.4 (Bioacoustics Research Program, 2011), using settings of FFT (Fast Fourier Transformation) of 512 samples and a window overlap of 50 percent.
To obtain spectral similarity among calls of native and invasive species we assessed the power spectra and the energy distribution among frequencies transformed with a FFT of 512 samples using the software WASIS β . We conducted a spectral filtering by selecting the power bands within the species frequency range. Then, we evaluated the spectral overlap among species by means of Pearson correlations using WASIS β. The correlation coefficient was considered as a proxy of spectral overlap (SO), from 0 (no correlation) to 1 (identical call). Additionally, we obtained dominant frequency and call duration with the software Raven 1.4 (Bioacoustics Research Program, 2011).
We swabbed 80 individuals of Eleutherodactylus johnstonei and 26 individuals of Phyllodytes luteolus immediately after capture for later detection of the frog-killing fungus Batrachochytrium dendrobatidis . We tested swabs for Bd in singlicate using Taqman qPCR [72, 71] with standards of 0.1, 1, 10, 100, and 1000 zoospore genomic equivalents (GE) to detect Bd in each individual frog. We used strain CLFT 023 as a quantitative standard for all qPCR reactions . We considered samples positive for Bd when GE was >1 .
We synthesized information of invasive amphibian species in Brazil, providing new records for expanding populations for five of the six invasive species (Table 1). Distribution records for all species are in S2 Table and S1 Fig. We found that the reported number of frog species with invasive populations in Brazil is increasing exponentially (Fig 1).
Asterisks denote novel data from the present study.
Species distribution modeling
All SDMs showed good discrimination capacity for training with averaged AUC values of 0.81 (Lithobates catesbeianus), 0.93 (Scinax x-signatus), 0.96 (Leptodactylus labyrinthicus), and 0.97 (Phyllodytes luteolus, Rhinella jimi, and Eleutherodactylus johnstonei). AUC equally weights omission and commission errors , and can be considered as a good measure of model performance , providing added reliability to the models. Therefore, none of the models were discarded during habitat suitability analyses. Overall, SDMs for future climatic habitat suitability showed contrasting patterns across species (Figs 2 and 3).
Leptodactylus labyrinthicus (A), Phyllodytes luteolus (D), Rhinella jimi (G), and Scinax x-signatus (J). Species distribution models projected to 2100 as binary maps showing current (B, E, H, and K) and future (C, F, I, and L) potential suitable habitats (red spots) (RCP4.5 scenario provided by IPCC fifth assessment; ).
Eleutherodactylus johnstonei (A), and Lithobates catesbeianus (D). Species Distribution Models projected to 2100 as binary maps showing current (B and E) and future (C and F) potential suitable habitats (red spots) (RCP4.5 scenario provided by IPCC fifth assessment; ).
Current SDM projections for L. labyrinthicus indicated high habitat suitability in the southern and eastern Brazilian Amazon, most of Brazil’s Atlantic forest, and a fair amount of the Brazilian Cerrado. However, future predictions showed a trend of decreasing habitat suitability in small portions of southern Amazonia and eastern Atlantic forest, with remaining available habitats concentrated in Cerrado and Caatinga. It is important to notice that our prediction also suggested high habitat unsuitability in the southern section of the Atlantic forest. Our models indicated that L. labyrinthicus might suffer a retraction of 8% in its total geographical range until 2100. Models built for P. luteolus indicated widespread suitable habitats throughout Amazonia, the Atlantic forest, and part of the Cerrado, while future predictions showed an increase in potential suitable habitats in the Cerrado. However, we found that the total geographic range for P. luteolus is projected to reduce by 16% in 2100. SDMs for the toad R. jimi predicted large habitat suitability in the Amazon, Cerrado, Atlantic forest, and in the archipelago of Fernando de Noronha, where this species was introduced. Nevertheless, SDMs indicated expansions in habitat suitability toward southwestern Amazon through the Cerrado and the Atlantic forest. In the future (2100), our models indicate that this species may expand its geographic range by 22%. Finally, models pointed to current suitable habitats for S. x-signatus through the whole country, matching its available occurrence points. However, our models predicted that one third of its suitable habitats will be gone by 2100, with most habitat retraction predicted for the Amazon and southern Atlantic forest.
Our models for current suitable habitats available for E. johnstonei showed a fragmented pattern for potential distribution. Suitable areas for this species were concentrated in the southern and coastal Atlantic forest and portions of the Amazon forest. Nevertheless, future potential distribution indicated a dramatic range shift, with a 94% increase in distributional range. Under this scenario, we found that the southern Atlantic forest may become climatically unsuitable for E. johnstonei while the Cerrado at the same time that some areas of the Amazon forest may become suitable for this species. SDMs for L. catesbeianus showed current habitat suitability mostly in the southern Atlantic forest, with some pockets of highly suitable habitats in northeastern Brazil. Under future climatic conditions, our models pointed to an increase in habitat suitability for L. catesbeianus in the northeastern Atlantic forest, very small areas of suitable habitats in the southeast, and unsuitable habitats throughout the south. Our spatial analysis projected a slight range retraction (0.5%) for L. catesbeianus in 2100.
Bioacoustics and disease assessment
We found a high spectral overlap between the calls of the invasive species Phyllodytes luteolus and the native Ischnocnema sp. (SO = 0.88;Fig 4a). Despite this spectral overlap, the calls of these species showed distinct temporal structures. Phyllodytes luteolus showed more spaced pulses and longer call duration than Ischnocnema sp. (Fig 5c and 5d). The spectral overlap index between the calls of Leptodactylus labyrinthicus (invasive) and L. knudseni (native) was 0.88 and between those of L. labyrinthicus and L. pentadactylus was even higher (SO = 0.93; Fig 4b). The advertisement call of L. knudseni had the same duration and dominant frequency when compared to the call of the invasive species (about 0.2 s and 375 Hz). However, the dominant frequency of the call of L. pentadactylus was higher than that of the invasive species (about 470 Hz) and also differed in call duration (about 0.5 s) (Fig 5a). Despite the high spectral overlap, the calls of these three species had substantial structural differences. Calls of Leptodactylus knudseni and L. pentadactylus are trills, while calls of L. labyrinthicus have a tonal structure. Pulses in the calls of L. pentadactylus were also more fused than those of L. knudseni (Fig 5b).
Audiospectogram (A) and oscillogram (B) of the advertisement calls of three congeneric Leptodactylus spp. that are synchronopatric in Manaus, Amazonas, northern Brazil. Audiospectogram (C) and oscillogram (D) of the advertisement calls of sympatric Ischnocnema sp. and Phyllodytes luteolus synchronopatric at Serra do Guararu, Guarujá, São Paulo, in southeastern Brazil.
None of the sampled individuals of Eleutherodactylus johnstonei and Phyllodytes luteolus tested positive for the chytrid fungus Batrachochytrium dendrobatidis.
Our study indicates that the number of invasive amphibian species in Brazil is increasing exponentially, with at least two of these species potentially expanding their ranges in the near future. We also found that invasive populations of two species did not test positive for the lethal Bd fungus, which reduces the likelihood of introductions of novel pathogen strains. Our combined results highlight the need for further monitoring of invasive amphibians in Brazil and indicate that stricter regulations regarding the international plant trade are needed to prevent further introductions such as the recent case of Eleutherodactylus johnstonei in the city of São Paulo.
Habitat suitability and potential distribution
Leptodactylus labyrinthicus was the only of our focal invasive species showing an isolated occurrence point outside our model predictions. Specifically, some individuals of L. labyrinthicus were recorded in Manaus, Brazilian Amazon, though our SDMs did not indicate suitable habitats for the species in that particular region. Conversely, our SDMs indicated large suitable habitats for L. labyrinthicus in the southern Amazon basin, which did not match its likely current distribution (see The IUCN Red List of Threatened Species for further details), but may represent a potential invasion route (Fig 2). Records of L. labyrinthicus from Manaus may characterize recently introduced populations and, therefore, monitoring the area where these individuals were captured is needed to gather detailed information about the status of this potential invasion. Also, when building SDMs we assumed that occurrence localities represent source habitats, not sink ones, which may harbor species without the minimum conditions to maintain populations without immigration flux . Such assumption lacks ecological realism when dealing with highly vagile species  such as L. labyrinthicus . Under current climate conditions, southern parts of the Amazon may be particularly suitable for this species not only due to favorable climate, but also due to recent deforestation practices . Conversely, future climate itself may require species’ distributional shifts and also may interfere in the persistence of L. labyrinthicus within its original distribution, potentially driving much of its range towards northeastern Brazil (Fig 2).
The second species we modeled, Phyllodytes luteolus, is originally restricted to a narrow strip along the coastal Atlantic forest from the states of Espírito Santo to Pernambuco . Conversely, SDM outputs for this species provide evidence of wider suitable habitats under current climates throughout the southern Atlantic forest, Cerrado, and the Amazon forest (Fig 2), which may explain punctual and recent success for this species in the states of Rio de Janeiro  and São Paulo (current study). This pattern may suggest that other factors such as interspecific competition and predation may constrain the expansion of this species for such distributional range. Additionally, populations of P. luteolus may be declining according to expert opinion . Avoiding dispersal of P. luteolus individuals by the local and the international bromeliad trade may also contribute to the preservation of its native populations.
Rhinella jimi is another non-threatened domestic invader species according to IUCN. R. jimi is endemic to inland Brazil; it was introduced in the main island of the archipelago of Fernando de Noronha between the decades of 1890 and 1970 . Amerigo Vespucci visited the island in 1503 and in a letter to Piero Soderini he mentioned the presence of skinks (Trachylepis atlantica) and worm lizards (Amphisbaena ridleyi), but did not refer to amphibians . Charles Darwin sampled terrestrial vertebrates there in 1832, but did not report any amphibian . Likewise, Henry Nicholas Ridley did not report any frog or toad living in the archipelago towards the end of the 1880’s , nor Alfredo Tito dos Santos, which only reported invasive tegu lizards (Salvator merianae) and tortoises (probably Chelonoidis sp., but currently absent in the island) . Only in the 1980’s other authors reported the presence of amphibians in Fernando de Noronha [76, 51, 90]. Besides that, some specimens housed at the Smithsonian National Museum of Natural History (USNM 196575; 346812–17) were collected in 1973. During this 85-year time period (between 1888 –last trustable animal-oriented study and 1973 –first collected individual), the native island inhabitants speculate that the priest Francisco Adelino de Brito Dantas, who lived there during the end of the 19th century , might have brought some toads (R. jimi) from the mainland to control insect herbivores that were attacking his crops. Alternatively, during World War II (1939–1945), the American army established a military base in Fernando de Noronha, and there are also speculations that they brought some toads as a way of mosquito control ; that was a common practice of the North American government at that time (e.g., ). Both possibilities fall within our confirmed temporal window (1888–1973). Thus, measurements of the impacts caused by R. jimi on native organisms of Fernando de Noronha are needed before implementation of any management plan. Moreover, our SDMs suggest suitable current habitats for R. jimi in the Cerrado, and future suitability, with a large range expansion, in southern Amazonia, which may translate to future range shifts westward.
Similarly to R. jimi and in the same period of time, the treefrog Scinax x-signatus was also introduced to Fernando de Noronha, though likely unintentionally. The same recommendations made for R. jimi are applicable in this case. S. x-signatus has very wide inland distribution in Brazil, virtually occurring in all of the existing biomes within the country . Our SDMs suggest current suitable habitats that are even larger than the known distribution of this species, probably due to some occurrence records in Colombia, which may have enlarged the interpretation of the niche range by the algorithm we used. A significant decrease of habitat suitability can be expected in future climate (Fig 2), with large areas of Amazonia becoming unsuitable. Although it is unlikely that climate change or habitat alteration could threaten this wide-ranging species, our models indicate that localized population declines of S. x-signatus may happen in the foreseeable future.
The imported invader Eleutherodactylus johnstonei originally occurs in the Lesser Antilles, but it has been introduced to Jamaica and Venezuela and it is currently being considered as a species in the process of expansion . The SDMs we built for this species show a fragmented mosaic of suitable habitats under current conditions. The established population in the Brooklin neighborhood in São Paulo may represent a source population capable of spreading to southern areas of the Atlantic forest (Fig 3). The fact that there are no records of E. johnstonei outside urban centers in Brazil (the same is observed in other invaded areas outside its native range; [93,94, 95]) definitely bodes well for the local fauna. Because this species shows extremely high population densities in Brooklin, mitigation actions may be needed to halt the spread of E. johnstonei to other nearby locations . Furthermore, similar inspection protocols to those suggested for P. luteolus could be implemented to prevent the introduction of E. johnstonei to other locations in Brazil [96, 48].
Finally, SDMs built for Lithobates catesbeianus showed a restricted distributional pattern of suitable habitats when compared to other studies [24, 40, 41, 55]. Aside from using distinct analytical tools, data sources, and the choice of climatic variables used, our study agrees in terms of prediction to the current SDMs and the occurrence records of L. catesbeianus [see 40]. Nevertheless, Both and colleagues  reported some individuals outside their optimum climatic conditions, as indicated by our models for current time. These individuals could be in the process of invading new sites, either by range expansions or due to recent translocations, not necessarily corresponding to established populations. To investigate this case, local monitoring or molecular analyses are necessary. Furthermore, our models for the current bullfrog distribution corroborate most of the expected future distribution for this invasive species in Brazil . Our models mostly match those provided by Ficetola et al.  and partially match the models provided by Nori et al. . However, we cannot discard the possibility that our models were influenced by the high number of occurrence records available for bullfrogs in North America, and also by the threshold method we used (e.g., ), which seems to be slightly high. To investigate this potential bias, further studies focused on multiple models based on forecast techniques (e.g., ) would be fruitful. Nonetheless, all cited studies (including ours) investigating potential bullfrog distribution show high statistical confidence and indicate matching patterns. In our current study, only habitats highly favorable for bullfrog occurrence in terms of suitable climates were projected into binary maps, ignoring low suitable habitats that could be used as routes for population establishments or migrations. Additionally, when dealing with highly vagile species such as L. catesbeianus, models might not be able to properly interpret potential sink habitats in a realistic fashion [82, 61]. Moreover, as previously discussed, SDM assumes only environmental variables, excluding biotic interactions . Nonetheless, our SDMs suggest suitable habitats mainly in restricted areas of the southern Atlantic forest and the Pampas under current climates, and future favorable habitats in areas of the Cerrado and Caatinga (Fig 3). Although these results do not sound overly pessimistic, we suggest caution while using predicted bullfrog distributions.
When an invasive amphibian species joins native species in any reproductive habitat, its vocalization may generate enough noise to disrupt the communication between native conspecifics that use similar frequency channels [46, 98]. For instance, Both et al.  reported a spectral displacement on calls of Hypsiboas albomarginatus when that species shared a breeding site with the invasive Lithobates catesbeianus in southern Brazil. Likewise, it is possible that such interference is also occurring between Ischnocnema sp. and Phyllodytes luteolus, because these species are synchronopatric in Guarujá and have advertisement calls with spectral overlap. Furthermore, invasions can also lead to mistakes in acoustic recognition between highly related species (see ) due to the likely similarity of calls between sister species [100, 101, 102]. The establishment of heterospecific pairs could be deleterious for those species of the genus Leptodactylus in Manaus because the invasive L. labyrinthicus is synchronopatric with the two other congeners . Besides morphological similarities and the high phylogenetic relatedness, the high spectral overlap coupled with similar ascending modulation observed in their advertisement calls could lead to hybridization (e.g., ). Whether or not slight differences in temporal structure among the calls of these species are enough for species-specific recognition is a matter for further investigation (see [104, 105, 106]).
The calls of species such as E. johnstonei and E. coqui are often very loud, especially when males are in chorus [27, 48]. Therefore, invasive species may lead to a sharp drop in real estate assets; property value is expected to depreciate in areas surrounded by calling sites such as those from Brooklin neighborhood in the city of São Paulo .
Invasive species could pose negative impacts on local food webs by competition with and predation upon native species. Leptodactylus labyrinthicus, Rhinella jimi, and Lithobates catesbeianus have the potential to alter food web dynamics because they show large ontogenetic variation, feeding on algae (as tadpoles), invertebrates, and small vertebrates, including other native amphibians, as post metamorphs (e.g., [107, 108, 109, 110, 111, 112, 43, 113, 44, 114, 45]). Species with a relatively small body size could also impact ecosystem functioning. For instance, two amphibian species introduced in Hawaii (Eleutherodactylus coqui and E. planirostris from Puerto Rico and Cuba, respectively) live in dense populations without native predators and feed on a massive amount of invertebrate biomass [26, 2]. These two species are in turn dramatically altering the local biogeochemical cycles [2, 96]. These findings highlight the need for detailed research on how Scinax x-signatus and Phyllodytes luteolus change ecosystem functioning. Shift in trophic networks could be an additional environmental impact caused by Eleutherodactylus johnstonei in case it reaches undisturbed natural habitats.
Island ecosystems are generally more vulnerable and respond more rapidly to the impacts of invasive species, because islands tend to harbor less complex communities than the mainland . Thus, R. jimi and S. x-signatus may be also affecting the terrestrial ecosystem of Fernando de Noronha. Islands usually show high levels of endemicity [116, 117], which makes the invasion of both R. jimi and S. x-signatus even more problematic in terms of conservation. Rhinella jimi is closely related to the cane-toad, Rhinella marina ; the latter is one of the top 100 most successful invasive species globally . Cane-toads were introduced to Australia as a biological control agent against crop pests and disease vector insects [120, 28, 2]; the same likely reasons for the introduction of R. jimi in Fernando de Noronha. R. marina is able to avoid several types of predators  and is promoting the evolution of morphological and behavioral traits in some native species in Australia . Thus far, we have no data on the potential impacts of R. jimi on the fauna and flora of Fernando de Noronha, but these impacts are likely similar to those observed in Australia.
Invasive populations of Phyllodytes luteolus in the states of Rio de Janeiro and São Paulo are possibly affecting native species by competing for limited resources for breeding (i.e., bromeliads) [122, 123]. In the southeast, invasive P. luteolus coexists with other native species that are bromeligenous, such as Scinax perpusillus in Guarujá. Bromeliads can represent a limited and fundamental resource for the reproduction of these species [122, 123, 124]. Consequently, the coexistence of these frogs may be a source of spatial conflict. P. luteolus has the ability to reproduce using different species of bromeliads [122, 123] and this trait may facilitate its spread and success over other species. Moreover, the larger body size of P. luteolus [123, 125] may contribute to its success while defending its breeding site against species with smaller body size such as S. perpusillus, Dendrophryniscus brevipollicatus, and other bromeligenous anuran species. Furthermore, because different types of stressors could reduce amphibian immunity [126, 127], it is possible that stress caused by invasive species increases the susceptibility of native species to parasites and diseases.
Diseases and parasites
In addition to competing and displacing local species, invasive amphibians may also disseminate parasites and diseases. For instance, the American bullfrog was recently identified as a competent carrier of the frog-killing Bd fungus in Brazil [128, 36, 37]. Another example is Ranavirus, which may be also spread with the international trade of bullfrogs . Furthermore, some authors  found the parasite Rhabdias fuelleborni on in individuals of R. jimi from Fernando de Noronha. This parasite is not amphibian exclusive and, therefore, could infect other vertebrates from the archipelago. Biological invasions facilitating the spread of these and other pathogenic agents may explain a lot of amphibian populations declines globally [130, 131]. Although we did not screen invasive frogs for Ranavirus, the absence of Bd in the invasive populations of Phyllodytes luteolus and Eleutherodactylus johnstonei bodes well for the Brazilian herpetofauna. Eleutherodactylus johnstonei, in particular, could have brought a novel hypervirulent Bd strain from Central America or the Caribbean, potentially leading to epizootics and population declines such as the ones observed in Panama and Costa Rica .
Biological invasions in Brazilian islands
The introduction of Rhinella jimi to the archipelago of Fernando de Noronha is a peculiar case of biological invasion. This population presents high rates of morphological anomalies, both in tadpoles  and post-metamorphics . Rhinella jimi may be suffering either from sub-lethal environmental condition (e.g., chemical pollutants) or intrinsic handicaps such as inbreeding depression (see [78, 79]). Rhinella jimi has been established in Fernando de Noronha for about 100 years and these toads are thriving despite showing high prevalence of multiple types of abnormalities [78, 133]. This is another example of how invasive species can benefit from the lack of local competitors and predators in oceanic islands [78, 79].
Our SDMs suggest that at least two invasive amphibians may expand their ranges in Brazil in the near future, following the accelerated pace of deforestation and global warming [134, 135, 136, 137]. These results underscore the need to further monitor populations at the edge of their invasive ranges in eastern Amazonia and around several urban areas of the Atlantic forest. To date, only the American bullfrog has been included in action plans by the Brazilian government to control expanding populations . Further studies focused on acoustic interference are needed to evaluate the direct impact of the invasive species Eleutherodactylus johnstonei and Phyllodytes luteolus on native anurans breeding in southeastern Brazil, and Leptodactylus labyrinthicus on native species from northern Brazil. Our work underscores that biological invasions of amphibians are becoming more complex and widespread, requiring immediate attention from the government to safeguard Brazil’s biodiversity.
S1 Table. Invasive species analyzed and sources of occurrence records.
S2 Table. Geographical locations used for the species distribution models.
S1 Fig. Maps with the invasive species distribution in Brazil; black dots as natural populations and red dots as invasive populations.
(A) Eleutherodactylus johnstonei; (B) Lithobates catesbeianus; (C) Scinax x-signatus; (D) Rhinella jimi; (E) Leptodactylus labyrinthicus; (F) Phyllodytes luteolus.
Denis O. V. Andrade provided museum records for sampling sites in Rio Claro, Brazil; Roseli Maria Foratto provided field assistance; The administration of Residential Iporanga allowed us to collect data in Guarujá, Brazil. The São Paulo Research Foundation (FAPESP) provided financial support (grant #2011/51694-7, #2011/52070-7, #2014/23388-7) to LFT. The National Council for Scientific and Technological Development (CNPq) provided support for LFT (#302589/2013-9; #405285/2013-2), to CGB (#312895/2014-3), to ISO (#161812/2011-2), and to LRF (#438675/2016-9). LRF and LT were supported by grants #2013/21519-4 and #2013/02219-0, São Paulo Research Foundation (FAPESP). CFBH was supported by grants #2013/50741-7 and #2014/50342-8, São Paulo Research Foundation (FAPESP) and CNPq (302518/2013-4) for a research fellowship. Coordination for the Improvement of Higher Education Personnel (CAPES) provided support to ISO (#3855/13-9).
- 1. Williamson M. Biological invasions. London: Chapman & Hall. 1996.
- 2. Kraus F. Alien reptiles and amphibians: a scientific compendium and analysis. Springer Science and Business Media B.V. 2009.
- 3. Magnusson WE. Homogeneização biótica. In: Rocha CFD, Bergalla HG, Sluys MV, Alves MAS, editors. Biologia da conservação: Essências. Rio de Janeiro: Rima. 2006. pp. 211–231.
- 4. Crowl TA, Crist TO, Parmenter RR, Belovsky G, Lugo AE. The spread of invasive species and infectious disease as drivers of ecosystem change. Front. Ecol. Environ. 2008;6: 238–246.
- 5. Hughes J, Goudkamp K, Hurwood D, Hancock M, Bunn S. Translocation causes extinction of a local population of the freshwater shrimp Paratya australiensis. Conserv. Biol. 2003;17(4): 1007–1012.
- 6. Lodge DM. Biological Invasions: Lessons for ecology. Tree. 1993;8(4): 133–136. pmid:21236129
- 7. Wilson EO. The diversity of life. WW Norton & Company. 1999.
- 8. Sanders NJ, Gotelli NJ, Heller NE, Gordon DM. Community disassembly by an invasive species. PNAS. 2003;100(5): 2474–2477. pmid:12604772
- 9. Macdougall AS, Turkington R. Are invasive species the drivers or passengers of change in degraded ecosystems? Ecology. 2005;86(1): 42–55.
- 10. Oliveira IS, Ribeiro VM, Pereira ER, Vitule JRS. Predation on native anurans by invasive vertebrates in the Atlantic Rain Forest, Brazil. Oecologia Australis. 2016;20(3): 70–74.
- 11. Ruiz GM, Carlton JT, Grosholz ED, Hines AH. Global invasions of marine and estuarine habitats by non-indigenous species: mechanisms, extent, and consequences. Am. Zool. 1997;37: 621–32.
- 12. Gurevitch J, Padilla DK. Are invasive species a major cause of extinctions? Trends Ecol. Evol. 2004;19 (9): 470–474. pmid:16701309
- 13. Molnar JL, Gamboa RL, Revenga C, Spalding MD. Assessing the global threat of invasive species to marine biodiversity. Front. Ecol. Environ. 2008;6(9): 485–492.
- 14. Tsutsui ND, Suarez AV, Holway DA, Case TJ. Reduced genetic variation and the success of an invasive species. PNAS. 2000;97(11): 5948–5953. pmid:10811892
- 15. Colautti RI, MacIsaac HJ. A neutral terminology to define ‘invasive’ species. Divers. Distrib. 2004;10: 135–141.
- 16. Grapputo A, Boman S, Lindstrom L, Lyytinem A, Mappes J. The voyage of an invasive species across continents: genetic diversity of North American and European Colorado potato beetle populations. Mol. Ecol. 2005;14: 4207–4219. pmid:16313587
- 17. Dlugosch KM, Parker IM. Founding events in species invasions: genetic variation, adaptive evolution, and the role of multiple introductions. Mol. Ecol. 2008;17: 431–449. pmid:17908213
- 18. Holway DA, Suarez AV, Case TJ. Role of abiotic factors in governing susceptibility to invasion: a test with Argentine ants. Ecology. 2002;83(6): 1610–1619.
- 19. Shea K, Chesson P. Community ecology theory as a framework for biological invasions. Trends in Ecology & Evolution. 2002;17(4): 170–176.
- 20. Holle BV, Simberloff D. Ecological resistance to biological invasion overwhelmed by propagule pressure. Ecology. 2005;86(12): 3212–3218.
- 21. Pearson RG, Dawson TP. Predicting the impacts of climate change on the distribution of species: are bioclimate envelope models useful? Global Ecol. Biogeogr. 2003;12: 361–371.
- 22. Soberón J, Peterson AT. Interpretation of models of fundamental ecological niches and species’ distributional areas. Biodiversity Informatics. 2005;2: 1–10.
- 23. Jeschke JM, Strayer DL. Usefulness of bioclimatic models for studying climate change and invasive species. Ann. NY Acad. Sci. 2008;1134: 1–24. pmid:18566088
- 24. Ficetola GF, Thuiller W, Miaud C. Prediction and validation of the potential global distribution of a problematic alien invasive species–the American Bullfrog. Divers. Distrib. 2007;13 (4): 476–485.
- 25. Lobos G, Cattan P, Estades C, Jaksic FM. Invasive African clawed frog Xenopus laevis in southern South America: key factors and predictions. Stud. Neotrop. Fauna E. 2013;48 (1): 1–12.
- 26. Kraus F, Campbell EW, Allison A, Pratt T. Eleutherodactylus frog introductions to Hawaii. Herpetol. Rev. 1999;30: 21–25.
- 27. Kraus F, Campbell EW. Human-mediated escalation of a formerly eradicable problem: the invasion of Caribbean frogs in the Hawaiian Islands. Biol. Invasions. 2002;4: 327–332.
- 28. Lever C. The cane toad: the history and ecology of a successful colonist. Westbury. 2001. pp. 230.
- 29. Phillips BL, Shine R. Adapting to an invasive species: toxic cane toads induce morphological change in Australian snakes. Proc. Natl. Acad. Sci. USA. 2004;101(49): 17150–17155. pmid:15569943
- 30. Fukuda Y, Tingley R, Crase B, Webb G, Saalfeld K. Long-term monitoring reveals declines in an endemic predator following invasion by an exotic prey species. Animal Conservation. 2015;19: 75–87.
- 31. Frost DR. Amphibian Species of the World: an Online Reference. Version 6.0; 2016 [cited 11 Jan 2016]. Electronic Database: American Museum of Natural History, New York, USA. Accessible from: http://research.amnh.org/herpetology/amphibia/index.html.
- 32. Cribb AY, Afonso AM, Ferreira CM. Manual técnico de ranicultura. Embrapa, Brasília. 2013;73.
- 33. Gurdon JB, Hopwood N. The introduction of Xenopus laevis into developmental biology: of empire, pregnancy testing and ribosomal genes. Int. J. Dev. Biol. 2000;44: 43–50. pmid:10761846
- 34. Measey GJ, Rödder D, Green SL, Kobayashi R, Lillo F, Lobos G, et al. Ongoing invasions of the African clawed frog, Xenopus laevis: a global review. Biol. Invasions. 2012;14: 2255–2270.
- 35. Shine R. Invasive species as drivers of evolutionary change: cane toads in tropical Australia. Evol. Appl. 2012;5: 107–116. pmid:25568034
- 36. Schloegel LM, Ferreira CM, James TY, Hipolito M, Longcore JE, Hyatt AD, et al. The North American Bullfrog as a reservoir for the spread of Batrachochytrium dendrobatidis in Brazil. Anim. Conserv. 2009;13(1): 53–61.
- 37. 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
- 38. Vredenburg VT, Felt SA, Morgan EC, McNally SVG, Wilson S, Green SL. Prevalence of Batrachochytrium dendrobatidis in Xenopus collected in Africa (1871–2000) and in California (2001–2010). PLOS ONE. 2013;8(5): e63791. pmid:23691097
- 39. Both C, Madalozzo B, Lingnau R, Grant T. Amphibian richness patterns in Atlantic Forest areas invaded by American Bullfrogs. Austral Ecology. 2014;39(7): 864–874.
- 40. Giovanelli JGR, Haddad CFB, Alexandrino J. Predicting the potential distribution of the alien invasive American Bullfrog (Lithobates catesbeianus) in Brazil. Biol. Invasions. 2008;10: 585–590.
- 41. Both C, Lingnau R, Santos JA, Madalozzo B, Lima LP, Grant T. Widespread occurrence of the American Bullfrog, Lithobates catesbeianus (Shaw,1802) (Anura: Ranidae), in Brazil. South Am. J. Herpetol. 2011;6: 127–134.
- 42. Ferreira CM, Pimenta AGC, Paiva Neto JS. Introdução à ranicultura. Boletim Técnico do Instituto de Pesca, São Paulo. 2002;33: 1–15.
- 43. Silva ET, Reis EP, Feio RE, Ribeiro Filho OP. Diet of the invasive frog Lithobates catesbeianus (Shaw, 1802) (Anura: Ranidae) in Viçosa, Minas Gerais state, Brazil. South Am. J. Herpetol. 2009;4: 286–294.
- 44. Silva ET, Ribeiro Filho OP, Feio RN. Predation of native anurans by invasive Bullfrogs in southeastern Brazil: spatial variation and effect of microhabitat use by prey. South Am. J. Herpetol. 2011;6: 1–10.
- 45. Leivas PT, Savaris M, Lampert S, Lucas EM. Predation of Odontophrynus americanus (Anura: Odontophrynidae) by the invasive species Lithobates catesbeianus (Anura: Ranidae) in an Araucaria forest remnant in Southern Brazil. Herpetology Notes. 2013;6: 603–606.
- 46. Both C, Grant T. Biological invasions and the acoustic niche: the effect of Bullfrog calls on the acoustic signals of white-banded tree frogs. Biol. Letters. 2012;1–12.
- 47. Batista CG. Rana catesbeiana (Bullfrog). Effects on native anuran community. Herpetol. Rev. 2002;33 (2): 131.
- 48. Melo MA, Lyra ML, Brischi AM, Geraldi VC, Haddad CFB. First record of the invasive frog, Eleutherodactylus johnstonei Barbour, 1914 (Anura: Eleutherodactylidae) in São Paulo, Brazil. Salamandra. 2014;50(3): 177–180.
- 49. Salles ROL, Silva-Soares T. Phyllodytes luteolus (Anura, Hylidae) as an alien species in the Rio de Janeiro municipality, State of Rio de Janeiro, Southeastern Brazil. Herpetology Notes. 2010;3: 257–258.
- 50. Carvalho VT, Fraga R, Silva ALF, Vogt RC. Introduction of Leptodactylus labyrinthicus (Spix, 1824) (Anura: Leptodactylidae) in central Amazonia, Brazil. Check List. 2013;9(4): 849–850.
- 51. Oren DC. Resultados de uma nova expedição zoológica a Fernando de Noronha. Boletim do Museu Paraense Emilio Goeldi, Zoologia. 1984;1: 19–44.
- 52. Lever C. Naturalized reptiles and amphibians of the world. New York: Oxford University Press. 2003.
- 53. Measey GJ, Royero R. An examination of Pipa parva (Anura: Pipidae) from native and invasive populations in Venezuela. Herpetological Journal. 2005;15: 291–294.
- 54. AmphibiaWeb. Information on amphibian biology and conservation, Berkeley, California: AmphibiaWeb; 2014 [cited 4 Aug 2014]. http://amphibiaweb.org/.
- 55. Nori J, Urbina-Cardona JN, Loyola RD, Lescano JN, Leynaud GC. Climate change and American Bullfrog invasion: what could we expect in South America? PLoS One. 2011;6: e25718. pmid:21991339
- 56. IPCC. Climate change 2014 and impacts, adaptation, and vulnerability (AR5). Cambridge and New York: Cambridge University Press. IUCN. (2014): The IUCN Red List of Threatened Species. Version 2014.1; 2014 [cited 21 Jan 2014]. http://www.iucnredlist.org. 2014.
- 57. R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria; 2016. [cited 11 Jan 2016] https://www.R-project.org/.
- 58. Lafarge T, Pateiro-López B, Possolo A, Dunkers JP. Implementation of the 3D alpha-shape for the reconstruction of 3D sets from a point cloud. 2016. Available from: https://cran.r-project.org/package=alphashape3d.
- 59. Hijmans RJ, Raster. Geographic data analysis and modeling, Davis CA. 2013 [cited 02 Sept 2016]. http://cran.r-project.org/web/packages/raster/raster.pdf.
- 60. Bivand R, Keitt T, Rowlingson B. Rgdal: Bindings for the Geospatial Data Abstraction Library. R Package Version 0.9–1. 2014 [cited 11 Febr 2014] http://CRAN.R-project.org/package=rgda.
- 61. Phillips SJ, Anderson RP, Schapire RE. Maximum entropy modeling of species geographic distributions. Ecol. Model. 2006;190: 231–259.
- 62. Peterson AT, Soberón J. Species distribution modeling and Ecological niche modeling: getting the concepts right. Nat. Conservação. 2012;10(2): 102–107.
- 63. Elith J, Phillips SJ, Hastie T, Dudík M, En Chee Y, Yates CJ. A statistical explanation of MaxEnt for ecologists. Divers Distrib. 2011;17: 43–57.
- 64. Rodda GH, Jarnevich CS, Reed RN. Challenges in identifying sites climatically matched to the native ranges of animal invaders. PLoS ONE. 2011;6(2): e14670. pmid:21347411
- 65. Oliveira S I, Rödder D, Capinha C, Ahmadzadeh F, Oliveira AKC, Toledo LF. Assessing Future Habitat Availability for Coastal Lowland Anurans in the Brazilian Atlantic Rainforest. Studies of Neotropical Fauna and Environment. 2016;51(1): 45–55.
- 66. Cantor SB, Sun CC, Tortolero-Luna G, Richards-Kortum R, Follen M. A comparison of C/B ratios from studies using receiver operating characteristic curve analysis.–J. Clin. Epidemiol. 1999;52: 885–892. pmid:10529029
- 67. Liu C, Berry PM, Dawson TP, Pearson RG. Selecting thresholds of occurrence in the prediction of species distributions. Ecography. 2005;28: 385–393.
- 68. Liu C, White M, Newell G. Selecting thresholds for the prediction of species occurrence with presence‐only data. J. Biogeogr. 2013;40: 778–789.
- 69. Wells KD. The social behavior of anuran amphibians. Animal Behavior. 1977;25:666–693.
- 70. Tacioli L, Medeiros CB, Toledo LF. WASIS: Wildlife Animal Sound Identification System (Version β). Fonoteca Neotropical Jacques Vielliard, Unicamp, Brazil. 2014. http://www.naturalhistory.com.br/wasis.html.
- 71. 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. Org. 2007;73: 175–192. pmid:17330737
- 72. 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. Org. 2004;60: 141–148. pmid:15460858
- 73. Longo AV, Rodriguez D, Leite DS, Toledo LF, Almeralla CM, Burrowes PA, et al. ITS1 copy number varies among Batrachochytrium dendrobatidis strains: implications for qPCR estimates of infection intensity from field-collected amphibian skin swabs. PLoS One. 2013;8: 1–10
- 74. Kriger KM, Ashton KJ, Hines HB, Hero JM. On the biological relevance of a single Batrachochytrium dendrobatidis zoospore: a reply to Smith. Dis. Aquat. Org. 2007;73: 257–260. pmid:17330746
- 75. Neckel-Oliveira S, Magnusson WE, Lima AP, Albernaz ALK. Diversity and distribution of frogs in an Amazonian savanna in Brazil. Amphibia-Reptilia. 2000;21: 317–326.
- 76. Olson SL. Natural history of vertebrates on the Brazilian islands of the mid South Atlantic. Natl. Geogr. Society Research Reports. 1981;13: 481–492.
- 77. Vicente JJ, Rodrigues HO, Gomes DC, Pinto RM. Nematóides do Brasil: 2ª parte, nematóides de anfíbios. Rev. Bras. Zool. 1990;7(4): 549–626.
- 78. Toledo LF, Ribeiro RS. The Archipelago of Fernando de Noronha: an intriguing malformed toad hotspot in South America. EcoHealth. 2009;6: 351–357. pmid:20112047
- 79. Tolledo J, Silva ET, Nunes-de-Almeida CH, Toledo LF. Anomalous tadpoles in a Brazilian oceanic archipelago: implications of oral anomalies on foraging behaviour, food intake and metamorphosis. Herpetol J. 2014;24: 237–243.
- 80. Lobo JM, Jiménez-Valverde A, Real R. AUC: a misleading measure of the performance of predictive distribution models. Global Ecology and Biogeography. 2008;17: 145–151.
- 81. 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(8): e0160746. pmid:27513565
- 82. Pulliam HR. On the relationship between niche and distribution. Ecol. Lett. 2000;3: 349–361.
- 83. Tozetti AM, Toledo LF. Short-term movement and retreat sites of Leptodactylus labyrinthicus (Anura: Leptodactylidae) during the breeding season: a spool-and-line tracking study. J. Herpetol. 2005;39(4): 640–644.
- 84. Ferreira LV, Venticinque E, Almeida S. O desmatamento na Amazônia e a importância das áreas protegidas. Estudos Avançados. 2005;19(53): 157–166.
- 85. IUCN. The IUCN Red List of Threatened Species. Version 2016–2; 2016. [cited 16 June 2016]. http://www.iucnredlist.org/
- 86. Vespucci A. Vespucci (Amerigo) Letter concerning the Isles newly discovered in his four voyages. London: Bernard Quaritch, 15 Piccadilly. 1885.
- 87. Darwin C. Into the natural history and geology of the countries visited during the voyage of H.M.S. Beagle round the world, under the command of Capt. Roy Fitz, N R.. New Edition. New York: D. Appleton and Company, Broadway. 1871. pp. 549–551.
- 88. Ridley HN. Notes on the zoology of Fernando de Noronha. Journal of the Linnean Society of London. 1890;20: 473–56.
- 89. Santos AT. Fernando de Noronha: ilha de cenários múltiplos. Rio de Janeiro. 1950. pp. 154.
- 90. Malagodi M, Kerr WE, Soares AEE. Introdução de abelhas na Ilha de Fernando de Noronha. 2. População de Apis mellifera ligustica. Cien. Cult. 1986;38(10): 1700–1704.
- 91. Silva MBL. Fernando de Noronha: cinco séculos de história, First Edition. Recife, Gráfica Santa Marta. 2007.
- 92. Pemberton. As New Guinea parasite. Hawaii. 1933.
- 93. Pough HF, Stewart MM, Thomas RG. Physiological basis of habitat partitioning in Jamaican Eleutherodactylus. Oecologia. 1977;27: 285–293. pmid:28308955
- 94. Kaiser H. Origins and introductions of the Caribbean frog, Eleutherodactylus johnstonei (Leptodactylidae): management and conservation concerns. Biodivers. Conserv. 1997;6: 1391–1407.
- 95. Ortega JE, Serrano VH, Ramírez-Pinilla MP. Reproduction of an introduced population of Eleutherodactylus johnstonei at Bucaramanga, Colombia. Copeia. 2005;3: 642–648.
- 96. Ernst R, Massemim D, Kowarik I. Non-invasive invaders from the Caribbean: the status of Johnstone’s Whistling frog (Eleutherodactylus johnstonei) ten years after its introduction to Western French Guiana. Biol. Invasions. 2011;13: 1767–1777.
- 97. Araújo MB,New M. Ensemble forecasting of species distributions. Trends in Ecology & Evolution. 2007;22(1): 42–47.
- 98. Medeiros CI, Both C, Grant T, Hartz SM. Invasion of the acoustic niche: variable responses by native species to invasive American bullfrog calls. Biological Invasions. 2017;19: 675–690.
- 99. Haddad CFB, Pombal JP Jr, Batistic RF. Natural hybridization between diploid and tetraploid species of Leaf-Frogs, genus Phyllomedusa (Amphibia). J. Herpetol. 1994;28(4): 425.
- 100. Hödl W. Call differences and calling site segregation in anuran species from Central Amazonian floating meadows. Oecologia. 1977;28: 351–363. pmid:28308940
- 101. Wells KD, Schwartz JJ. The behavioral ecology of anuran communication. In Hearing and sound. Narins P.M., Feng A.S., Fay R.R., Popper A.N., Ed., Amphibians, Springer. 2007. pp. 44–86
- 102. Toledo LF, Llusia D, Vieira CA, Corbo M, Marquez R. Neither convergence nor divergence in the advertisement call of sympatric congeneric Neotropical treefrogs. Bioacoustics. 2014. pp.1–17.
- 103. Lima AP, Magnusson WE, Menin M, Erdtmann LK, Rodrigues DJ, Keller C, et al. Guia de sapos da Reserva Adolpho Ducke, Amazônia Central. Manaus, Attema Design Editorial Ltda. 2008.
- 104. Gerhardt HC, Doherty JA. Acoustic communication in the gray treefrog, Hyla versicolor: evolutionary and neurobiological implications. J. Comp. Physiol. A 1988;162: 261–278.
- 105. Ryan MJ, Rand AS. Species recognition and sexual selection as a unitary problem in animal communication. Evolution. 1993;42: 647–657.
- 106. Marshall VT, Schwartz JJ, Gerhardt HC. The effects of heterospecific call overlap on the phonotactic behaviour of grey treefrogs. Anim. Behav. 2006;72: 449–459.
- 107. Boelter RA, Kaefer IL, Both C, Cechin S. Invasive bullfrogs as predators in a neotropical assemblage: what frog species do they eat? Animal Biology 2012:62:397–408.
- 108. Cardoso AJ, Sazima I. Batracofagia na fase adulta e larvária da rã-pimenta, Leptodactylus labyrinthicus (Spix, 1824)—Anura, Leptodactylidae. Ciênc. e Cult. (SBPC). 1977;29(10): 1130–1132.
- 109. Duellman WE, Trueb L. Biology of Amphibians. Baltimore, Johns Hopkins University Press. 1994.
- 110. Vaz-Silva W, Silva HLR, Silva Júnior NJ. Leptodactylus labyrinthicus (Labyrinth Frog). Diet. Herpetol. Ver. 2003;34: 359.
- 111. Toledo LF, Ribeiro RS, Haddad CFB. Anurans as prey: an exploratory analysis and size relationships between predators and their prey. J. Zool. 2007;271: 170–177.
- 112. Silva ET, Ribeiro Filho OP. Predation on juveniles of the invasive American Bullfrog Lithobates catesbeianus (Anura, Ranidae) by native frog and snake species in southeastern Brazil. Herpetology Notes. 2009;2: 215–218.
- 113. Silva ET, Reis EP, Santos OS, Feio RN. Lithobates catesbeianus (American Bullfrog). Diet. Herpetol. Rev. 2010;41: 475–476.
- 114. Fonseca E, Lanna F, Carvalho R, Gehara M. Predation on Sibynomorphus neuwiedi (Serpentes: Dipsadidae) by Leptodactylus labyrinthicus (Anura: Leptodactylidae) in southeastern Brazil. Herpetology Notes. 2012;5: 167–168.
- 115. Vitousek PM, Loope LL, Adsersen H, D’Antonio CM. Island ecosystem: do they represent “natural experiments” in biological diversity and ecosystem function? In Functional roles of biodiversity: a global perspective. Mooney H.A., Cushman J.H., Medina E., Sala O.E., Schulze E.D., Ed., Wiley. 1996. pp. 245–259
- 116. Whittaker RJ, Fernández-Palacios JM. Island biogeography 2nd Edition. Oxford, Oxford University Press. 2007.
- 117. Kier G, Kreft H, Lee TM, Jetz W, Ibisch PI, Nowicki C, et al. A global assessment of endemism and species richness across island and mainland regions. PNAS. 2009;106: 9322–9327. pmid:19470638
- 118. Maciel NM, Collevatti RG, Colli GR, Schwartz EF. Late Miocene diversification and phylogenetic relationships of the huge toads in the Rhinella marina (Linnaeus, 1758) species group (Anura: Bufonidae). Mol. Phylogenet. Evol. 2010;57: 787–797. pmid:20813190
- 119. Lowe S, Browne M, Boudjelas S, De Poorter M. 100 of the world’s worst invasive alien apecies. A selection from the global invasive species database. Published by The Invasive Species Specialist Group (ISSG) a specialist group of the Species Survival Commission (SSC) of the World Conservation Union (IUCN). 2000. pp. 12.
- 120. Easteal S. The history of introductions of Bufo marinus (Amphibia: Anura): a natural experiment in evolution. Biol. J. Linn. Soc. 1981;16: 93–115.
- 121. Shine R. The ecological impact of invasive cane toads (Bufo marinus) in Australia. Q. Rev. Biol. 2010;85: 253–291. pmid:20919631
- 122. Eterovick PC. Use and sharing of calling and retreat sites by Phyllodytes luteolus in modified environment. J. Herpetol. 1999;33: 17–22.
- 123. Ferreira RB, Schineider JAP, Teixeira RL. Diet, fecundity, and use of bromeliads by Phyllodytes luteolus (Anura: Hylidae) in Southeastern Brazil. J. Herpetol. 2012;46(1): 19–24.
- 124. Alves-Silva R, Silva HR. Life in bromeliads: reproductive behaviour and the monophyly of the Scinax perpusillus species group (Anura: Hylidae). J. Nat. Hist. 2009;43(3–4): 205–217.
- 125. Izecksohn E, Carvalho-e-Silva SP. Anfíbios do município do Rio de Janeiro, 1st Edition. Rio de Janeiro, Editora UFRJ. 2001.
- 126. Carey C, Cohen N, Rollins-Smith L. Amphibian declines: an immunological perspective. Dev. Comp. Immunol. 1999;23: 459–472. pmid:10512457
- 127. Gahl MK, Calhoun AJK. The role of multiple stressors in ranavirus-caused amphibian mortalities in Acadia National Park wetlands. Can. J. Zoolog. 2010;88: 108–121.
- 128. Garner TWJ, 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. Letters. 2006;2: 455–459.
- 129. Lesbarréres D, Balseiro A, Brunner J, Chinchar VG, Duffus A, Kerby J, et al. Ranavirus: past, present and future. Biol. Letters. 2011. pp. 1–3.
- 130. 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. 2014;346: 630–631. pmid:25359973
- 131. 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. USA. 2006;103: 3165–3170. pmid:16481617
- 132. Lips KR, Diffendorfer J, Mendelson JR III, Sears MW. Riding the wave: reconciling the roles of disease and climate change in amphibian declines. PLoS Biol. 2008;6(3): e72. pmid:18366257
- 133. Tolledo J, Toledo LF. Blind toads in paradise: the cascading effect of vision loss on a tropical archipelago. Journal of Zoology. 2015;296: 167–176.
- 134. Fearnside PM. Deforestation in Brazilian Amazonia: history, rates and consequences. Conservation Biology. 2005;19(3): 680–688.
- 135. Didham RK, Tylianakis JM, Gemmell NJ, Rand TA, Ewers RM. Interactive effects of habitat modification and species invasion on native species decline. Trends in Ecology & Evolution. 2007;22: 489–496.
- 136. Costa MH, Pires GF. Effecs of Amazon and Central Brazil deforestation scenarios on the duration of the dry season in the arc of deforestation. Int. J. Climatol. 2009;30: 1970–1979.
- 137. Alho CJR, Mamede S, Bitencourt K, Benites M. Introduced species in the Pantanal: implications for conservation. Brazilian Journal of Biology. 2011;71(1 Suppl.): 321–325.
- 138. Verdade VK, Valdujo PH, Carnaval ACOQ, Schiesari L, Toledo LF, Mott T, et al. A leap further: the Brazilian amphibian conservation action plan. Alytes. 2012;29: 27–42.