Evolutionary divergence in behavioural traits related to mating may represent the initial stage of speciation. Direct selective forces are usually invoked to explain divergence in mate-recognition traits, often neglecting a role for neutral processes or concomitant differentiation in ecological traits. We adopted a multi-trait approach to obtain a deeper understanding of the mechanisms behind allopatric divergence in the Amazonian frog, Allobates femoralis. We tested the null hypothesis that geographic distance between populations correlates with genetic and phenotypic divergence, and compared divergence between mate-recognition (acoustic) and ecological (coloration, body-shape) traits. We quantified geographic variation in 39 phenotypic traits and a mitochondrial DNA marker among 125 individuals representing eight populations. Geographic variation in acoustic traits was pronounced and tracked the spatial genetic variation, which appeared to be neutral. Thus, the evolution of acoustic traits tracked the shared history of the populations, which is unexpected for pan-Amazonian taxa or for mate-recognition traits. Divergence in coloration appeared uncorrelated with genetic distance, and might be partly attributed to local selective pressures, and perhaps to Batesian mimicry. Divergence in body-shape traits was low. The results obtained depict a complex evolutionary scenario and emphasize the importance of considering multiple traits when disentangling the forces behind allopatric divergence.

The Neotropical poison frog genus Ranitomeya is revised, resulting in one new genus, one new species, five synonymies and one species classified as nomen dubium. We present an expanded molecular phylogeny that contains 235 terminals, 104 of which are new to this study. Notable additions to this phylogeny include seven of the 12 species in the minuta group, 15 Ranitomeya amazonica, 20 R. lamasi, two R. sirensis, 30 R. ventrimaculata and seven R. uakarii. Previous researchers have long recognized two distinct, reciprocally monophyletic species groups contained within Ranitomeya, sensu Grant et al. 2006: the ventrimaculata group, which is distributed throughout much of the Amazon, and the minuta group of the northern Andes and Central America. We restrict Ranitomeya to the former group and erect a new genus, Andinobates Twomey, Brown, Amézquita & Mejía-Vargas gen. nov., for members of the minuta group. Other major taxonomic results of the current revision include the following: (i) A new species, Ranitomeya toraro Brown, Caldwell, Twomey, Melo-Sampaio & Souza sp. nov., is described from western Brazil. This species has long been referred to as R. ventrimaculata but new morphological and phylogenetic data place it sister to R. defleri. (ii) Examination of the holotype of R. ventrimaculata revealed that this specimen is in fact a member of what is currently referred to as R. duellmani, therefore, Dendrobates duellmani Schulte 1999 is considered herein a junior synonym of D. ventrimaculatus Shreve 1935 (= R. ventrimaculata). (iii) For the frogs that were being called R. ventrimaculata prior to this revision, the oldest available and therefore applicable name is R. variabilis. Whereas previous definitions of R. variabilis were restricted to spotted highland frogs near Tarapoto, Peru, our data suggest that this color morph is conspecific with lowland striped counterparts. Therefore, the definition of R. variabilis is greatly expanded to include most frogs which were (prior to this revision) referred to as R. ventrimaculata. (iv) Phylogenetic and bioacoustic evidence support the retention of R. amazonica as a valid species related to R. variabilis as defined in this paper. Based on phylogenetic data, R. amazonica appears to be distributed throughout much of the lower Amazon, as far east as French Guiana and the Amazon Delta and as far west as Iquitos, Peru. (v) Behavioral and morphological data, as well as phylogenetic data which includes topotypic material of R. sirensis and numerous samples of R. lamasi, suggest that the names sirensis, lamasi and biolat are applicable to a single, widespread species that displays considerable morphological variation throughout its range. The oldest available name for this group is sirensis Aichinger; therefore, we expand the definition of R. sirensis. (vi) Ranitomeya ignea and R. intermedia, elevated to the species status in a previous revision, are placed as junior synonyms of R. reticulata and R. imitator, respectively. (vii) Ranitomeya rubrocephala is designated as nomen dubium. In addition to taxonomic changes, this revision includes the following: (i) Explicit definitions of species groups that are consistent with our proposed taxonomy. (ii) A comprehensive dichotomous key for identification of ‘small’ aposematic poison frogs of South and Central America. (iii) Detailed distribution maps of all Ranitomeya species, including unpublished localities for most species. In some cases, these records result in substantial range extensions (e.g., R. uakarii, R. fantastica). (iv) Tadpole descriptions for R. amazonica, R. flavovittata, R. imitator, R. toraro sp. nov., R. uakarii and R. variabilis; plus a summary of tadpole morphological data for Andinobates and Ranitomeya species. (v) A summary of call data on most members of Andinobates and Ranitomeya, including call data of several species that have not been published before. (vi) A discussion on the continued impacts of the pet trade on poison frogs (vii) A discussion on several cases of potential Müllerian mimicry within the genus Ranitomeya. We also give opinions regarding the current debate on recent taxonomic changes and the use of the name Ranitomeya.

In the Neotropics, almost every species of the stream-dwelling harlequin toads (genus Atelopus) have experienced catastrophic declines. The persistence of lowland species of Atelopus could be explained by the lower growth rate of Batrachochytrium dendrobatidis (Bd) at temperatures above 25°C. We tested the complementary hypothesis that the toads' skin bacterial microbiota acts as a protective barrier against the pathogen, perhaps delaying or impeding the symptomatic phase of chytridiomycosis. We isolated 148 cultivable bacterial strains from three lowland Atelopus species and quantified the anti-Bd activity through antagonism assays. Twenty-six percent (38 strains representing 12 species) of the bacteria inhibited Bd growth and just two of them were shared among the toad species sampled in different localities. Interestingly, the strongest anti-Bd activity was measured in bacteria isolated from A. elegans, the only species that tested positive for the pathogen. The cutaneous bacterial microbiota is thus likely a fitness-enhancing trait that may (adaptation) or not (exaptation) have appeared because of natural selection mediated by chytridiomycosis. Our findings reveal bacterial strains for development of local probiotic treatments against chytridiomycosis and also shed light on the mechanisms behind the frog-bacteria-pathogen interaction.

In species-rich assemblages of acoustically communicating animals, heterospecific sounds may constrain not only the evolution of signal traits but also the much less-studied signal-processing mechanisms that define the recognition space of a signal. To test the hypothesis that the recognition space is optimally designed, i.e., that it is narrower toward the species that represent the higher potential for acoustic interference, we studied an acoustic assemblage of 10 diurnally active frog species. We characterized their calls, estimated pairwise correlations in calling activity, and, to model the recognition spaces of five species, conducted playback experiments with 577 synthetic signals on 531 males. Acoustic cooccurrence was not related to multivariate distance in call parameters, suggesting a minor role for spectral or temporal segregation among species uttering similar calls. In most cases, the recognition space overlapped but was greater than the signal space, indicating that signal-processing traits do not act as strictly matched filters against sounds other than homospecific calls. Indeed, the range of the recognition space was strongly predicted by the acoustic distance to neighboring species in the signal space. Thus, our data provide compelling evidence of a role of heterospecific calls in evolutionarily shaping the frogs' recognition space within a complex acoustic assemblage without obvious concomitant effects on the signal.

The efficacy of communication relies on detection of species-specific signals against the background noise. Features affecting signal detection are thus expected to evolve under selective pressures represented by masking noise. Spectral partitioning between the auditory signals of co-occurring species has been interpreted as the outcome of the selective effects of masking interference. However, masking interference depends not only on signal's frequency but on receiver's range of frequency sensitivity; moreover, selection on signal frequency can be confounded by selection on body size, because these traits are often correlated. To know whether geographic variation in communication traits agrees with predictions about masking interference effects, we tested the hypothesis that variation in the male-male communication system of the Amazonian frog, Allobates femoralis, is correlated with the occurrence of a single species calling within an overlapping frequency range, Epipedobates trivittatus. We studied frogs at eight sites, four where both species co-occur and four where A. femoralis occurs but E. trivittatus does not. To study the sender component of the communication system of A. femoralis and to describe the use of the spectral range, we analyzed the signal's spectral features of all coactive species at each site. To study the receiver component, we derived frequency-response curves from playback experiments conducted on territorial males of A. femoralis under natural conditions. Most geographic variation in studied traits was correlated with either call frequency or with response frequency range. The occurrence of E. trivittatus significantly predicted narrower and asymmetric frequency-response curves in A. femoralis, without concomitant differences in the call or in body size. The number of acoustically coactive species did not significantly predict variation in any of the studied traits. Our results strongly support that the receiver but not the sender component of the communication system changed due to masking interference by a single species.

No abstract

Many territorial species respond less aggressively to familiar neighbours than to unfamiliar floating strangers based on individual differences in acoustic signals. This form of social recognition, termed neighbour–stranger discrimination (NSD) or dear‐enemy phenomenon has been reported so far from three anuran species. To investigate the potential of auditory signal features to convey information on sender’s identity, we determined patterns of within‐male and between‐male variability in the advertisement call of the aromobatid frog Allobates femoralis. We examined 285 calls from 19 males to assess those call properties showing sufficient and reliable inter‐individual differences to function as possible recognition cues. Beside calls per call bout and call rate, all other examined call properties were more variable among males than within males. Generally, temporal call features showed higher between‐ and within‐male variability ratios than spectral properties and contributed mostly to distinguish individual males in the discriminant‐function analysis. Mean classification success of 64.9% correctly assigned calls to individual males is mainly attributable to three temporal call properties (duration of note 1 and 4, note repetition rate). Altogether, our results suggest that there is sufficient variation in the advertisement call to discriminate statistically among individual males. However, assessed call differences between A. femoralis males were rather small, suggesting that potential NSD might be based either on a combination of call features or even on the whole pattern of individual call variation instead on single call properties. Habituation–discrimination experiments in the field using modified playback signals to test for differential behavioural responses are required to confirm this hypothesis.

We engineered a machine learning approach, MSHub, to enable auto-deconvolution of gas chromatography-mass spectrometry (GC-MS) data. We then designed workflows to enable the community to store, process, share, annotate, compare and perform molecular networking of GC-MS data within the Global Natural Product Social (GNPS) Molecular Networking analysis platform. MSHub/GNPS performs auto-deconvolution of compound fragmentation patterns via unsupervised non-negative matrix factorization and quantifies the reproducibility of fragmentation patterns across samples.Given its ease of use and low operational cost, GC-MS has applications with broad societal effect, such as detection of metabolic disease in newborns, toxicology, doping, forensics, food science and clinical testing. The predominant ionization technique in GC-MS is electron ionization (EI), in which all compounds are ionized by high-energy (70-eV) electrons. Because fragmentation occurs with ionization, EI GC-MS data are subjected to spectral deconvolution, a process that separates fragmentation ion patterns for each eluting molecule into a composite mass spectrum.The 70 eV for ionizing electrons in GC-MS has been the standard, making it possible to use decades-old EI reference spectra for annotation 1 . There are ~1.2 million reference spectra that have been accumulated and curated over a period of more than 50 years 2 . Many tools and repositories for GC-MS data have been introduced [3][4][5][6][7][8][9][10][11][12][13][14][15] ; however, much of GC-MS data processing is restricted to vendor-specific formats and software 8 . Currently, deconvolution requires setting multiple parameters manually [3][4][5] or posessing computational skills to run the software 7 . Also, the lack of data sharing in a uniform format precludes data comparison between laboratories and prevents taking advantage of repository-scale information and community knowledge, resulting in infrequent reuse of GC-MS data 8,[11][12][13][14][15] .Although batch modes exist, deconvolution quality is currently not enhanced by using information from all other files. To leverage across-file information, improve scalability of spectral deconvolution and eliminate the need for manually setting the deconvolution parameters (m/z error correction of the ions and peak shapeslopes of raising and trailing edges, peak RT shifts and noise/intensity thresholds), we developed an algorithmic learning strategy for auto-deconvolution (Fig. 1a-f). We deployed this functionality within GNPS/MassIVE (https://gnps.ucsd.edu) 16 (Fig. 1f-i). To promote analysis reproducibility, all GNPS jobs performed are retained in the 'My User' space and can be shared as hyperlinks.This user-independent 'automatic' parameter optimization is accomplished via fast Fourier transform (FFT), multiplication and inverse Fourier transform for each ion across an entire data set, followed by an unsupervised non-negative matrix factorization (NMF) (one-layer neural network). Then, the compositional consistency of spectral patterns for each spec...