Search

The canyon treefrog, Hyla arenicolor, is a wide-ranging hylid found from southwestern US into southern Mexico. Recent studies have shown this species to have a complex evolutionary history, with several phylogeographically distinct lineages, a probable cryptic species, and multiple episodes of mitochondrial introgression with the sister group, the H. eximia complex. We aimed to use genome wide AFLP markers to better resolve relationships within this group. As in other studies, our inferred phylogeny not only provides evidence for repeated mitochondrial introgression between H. arenicolor lineages and H. eximia/H. wrightorum, but it also affords more resolution within the main H. arenicolor clade than was previously achieved with sequence data. However, as with a previous study, the placement of a lineage of H. arenicolor whose distribution is centered in the Balsas Basin of Mexico remains poorly resolved, perhaps due to past hybridization with the H. eximia complex. Furthermore, the AFLP data set shows no differentiation among lineages from the Grand Canyon and Colorado Plateau despite their large mitochondrial sequence divergence. Finally, our results infer a well-supported sister relationship between this combined Colorado Plateau/Grand Canyon lineage and the Sonoran Desert lineage, a relationship that strongly contradicts conclusions drawn from the mtDNA evidence. Our study provides a basis for further behavioral and ecological speciation studies of this system and highlights the importance of multi-taxon (species) sampling in phylogenetic and phylogeographic studies.

The temporal relationship between signals often has strong and repeatable influences on receiver behaviour. While several studies have shown that receivers prefer temporally leading signals, we show that the relative timing of signal elements within overlapping signals can also have repeatable influences on receiver responses. Female grey treefrogs, Hyla versicolor, preferred overlapping conspecific advertisement call alternatives in which pulses were in the leading position relative to pulses in an alternative. The preference was maintained even when the first pulse of the stimulus with leading pulses began after that of the call with following pulses. To rule out the possibility of masking interference of the pulse pattern, we used a split-pulse design in which the playback of two nonoverlapping pulse elements were synchronized from spatially separated speakers. Females were attracted to the source of the short (6 ms) leading pulse element, which did not attract females in isolation, even though its amplitude was 24 dB lower than the long (24 ms) following element, which did attract females in isolation. Taken together, our results fall within a range of phenomena that have been classified as precedence effects. However, to our knowledge, showing localization based on successive leading pulses rather than the very first-arriving pulse is a novel discovery for nonhuman animals.

We investigated the effects of overlap between the advertisement calls of two closely related and often-sympatric species of grey treefrogs on female phonotactic behaviour and signal selectivity. The tetraploid H. versicolor and the diploid H. chrysoscelis have advertisement calls consisting of trains of stereotyped pulses that are similar in their spectral composition. All females of H. versicolor from a population sympatric with H. chrysoscelis approached a speaker broadcasting synthetic advertisement calls modelled after conspecific calls when the calls of the two species alternated from speakers with a 90° angular separation with respect to the release point. Surprisingly, the same females almost exclusively approached the source of synthetic calls modelled after the heterospecific advertisement calls when these calls completely overlapped the conspecific stimuli. Female H. chrysoscelis from the same sympatric population approached the source of synthetic conspecific calls even when these calls were overlapped by the heterospecific stimuli, indicating that females in these closely related species may differ in their susceptibility to the negative consequences of call overlap. When female H. versicolor were presented with conspecific advertisement calls that were completely overlapped by those of H. chrysoscelis, many females failed to approach either signal source. Response times of females that did respond were also longer, and the direction of phonotaxis was shifted in the direction of the heterospecific signal source. These negative influences of acoustic interference were reduced or eliminated as the degree of overlap between the calls of the two species was reduced. The results demonstrate that the effects of signal overlap on female phonotactic behaviour may be complex.

Females are expected to use spectral properties of acoustic signals for mate choice because dominant frequency often provides reliable information about male size. The advertisement calls of grey treefrogs ( Hyla chrysoscelis and H. versicolor ) have two main frequency bands: the amplitude of the dominant, high-frequency band averages 7–10 dB greater than that of the low-frequency band. The frequencies of these peaks were negatively correlated with male body size in both species. I used two-speaker playbacks of synthetic calls to generate preference functions for both species based on simultaneous variation in the frequencies of the two bands and, for H. versicolor , on variation in single-component calls; frequencies of most of these stimuli fell within the range of the dominant, high-frequency band. When two-component sounds were used, preference functions were bimodal and broad, with peaks close to ( H. versicolor ) or slightly higher than ( H. chrysoscelis ) mean frequencies in conspecific calls. Because preferences were not strong, I predict that females exert only weak stabilizing or directional selection on male call frequencies. When single-component calls were used in tests of H. versicolor , the preference-function peak was shifted to a lower-than-average frequency within the high-frequency range. I hypothesize that such a frequency may better stimulate both of the inner ear organs, whose tuning roughly matches frequencies in the two bands in conspecific calls, than a single component of average or much lower-than-average frequency. Females of both species preferred calls with both spectral peaks to calls with a single, high-frequency peak, even when the low-frequency peak was attenuated by 24–30 dB. This result serves to emphasize that typically subdominant frequency components can have seemingly disproportionate effects on signal attractiveness.

Diencephalic and midbrain auditory nuclei are involved in the processing of auditory communication signals in anurans [Comparative Hearing: Fish and Amphibians, Springer-Verlag, New York, 1999, p. 218], but their exact roles in acoustically guided behavior, such as female phonotaxis, are unclear. To address this question, behavioral experiments were combined with lesions of dorsal thalamic nuclei and the midbrain torus semicircularis. Females were tested in two-alternative-forced-choice phonotactic experiments before and after a defined brain area was lesioned. During phonotactic tests, females had to choose between a “standard” synthetic call and one of three different variants, each of which had a single acoustic property (pulse rate, pulse rise-time, sound spectrum) that differed from the standard synthetic call. Results showed that dorsomedial thalamus lesions produced little or no effect on phonotaxis. In contrast, superficial and deep thalamus lesions, as well as lesions of the torus semicircularis, significantly decreased the number of phonotactic responses and increased the response time. Superficial thalamus lesions also abolished or reversed preferences for the standard call in the rise-time and sound spectrum tests. This effect is likely to have been caused by an imbalance in the stimulation of the thalamus by the low- and high-frequency pathways because these preferences were not affected in animals with more extensive lesions that included the superficial thalamus. Our data suggest that the torus semicircularis, but not the dorsal thalamus is crucial for phonotaxis in gravid, reproductively active females. Although dorsal thalamic nuclei seem to play a role in spectral sensitivity, they may additionally have motivational or attentional functions that contribute to achieving a state of phonotactic readiness.

Studies of mate attraction have traditionally employed one of two experimental methods: choice tests in which female preferences from among two or more signals are tabulated, or single-stimulus tests in which the female response to one signal is quantified. Choice tests have long been preferred for examining mate attraction in anurans. Inspired by Wagner's (1998, Animal Behaviour,55, 1029–1042) discussion of the strengths and weaknesses of the two experimental approaches, we used a single-stimulus design to examine mate attraction in two sibling species of treefrogs (Hyla versicolor and H. chrysoscelis). We quantified female responses based upon the relative time required to approach signals varying in pulse rate, pulse rise time and pulse number. The data were used to generate response functions providing a quantitative measure of female attraction to the stimuli. Comparisons with data from choice experiments reveal broad similarities, as well as properties of female responses that had not been detected with choice tests. The results are discussed with regard to female selectivity for call parameters that are likely to mediate sexual selection and homospecific pairing.

Some territorial animals discriminate among neighbours and strangers based on individual differences in acoustic signals. Male North American bullfrogs, Rana catesbeiana, display this form of discrimination based on individual variation in advertisement calls. In this study, we investigated the acoustic basis of neighbour–stranger discrimination to determine how individual identity might be encoded by particular properties of bullfrog advertisement calls. We analysed patterns of within-male and between-male variability in 1078 bullfrog advertisement calls recorded from 27 territorial males. All call properties that we examined varied significantly among males. However, fundamental frequency and dominant frequency showed the lowest within-male variation and the highest repeatability between two recording sessions, and both properties were highly correlated with the first canonical root from discriminant function analyses, which typically accounted for 70–80% of the variability between males. We suggest that neighbour–stranger discrimination in bullfrogs is partially mediated by between-male differences in the spectral or fine temporal properties of advertisement calls.

In nearly all female choice experiments with anurans, the values of acoustic properties of experimental stimuli have been fixed. In nature, however, a female usually experiences an array of calls in which acoustic properties, especially dynamic ones, differ between males in variability as well as in mean values. High variability per se does not make a synthetic call more or less attractive to female grey treefrogs, Hyla versicolor, than a non-variable or low-variability call with about the same mean values of pulse number per call, call period, or both. Females did prefer a low-variability alternative with a slightly higher number of sound pulses per unit time than the variable stimulus, suggesting that females can resolve small differences in mean values between the calls of different males despite considerable within-male variability.

Abstract. Because male treefrogs indiscriminately try to mate with other frogs of comparable size that move close to them, a major cause of heterospecific matings is thought to be inadvertent contact between a female of one species and a vocalizing male of another species. It follows that the probability of such mismatings would be reduced if females recognized the calls of sympatric, heterospecific males and avoided moving near those males during phonotactic approaches to vocalizing conspecific males. Gravid female treefrogs of three species were tested in situations in which they could avoid a source of heterospecific calls as they responded to conspecific calls. The results of these experiments provide no evidence that females avoid sources of heterospecific calls. Because females of all three species sometimes show positive phonotactic responses in 'no-choice' situations to the same kinds of heterospecific calls used in this study, these results suggest that the mechanisms responsible for homospecific pairing are qualitatively no different than those mediating intraspecific mate choice. The implications of these results for current debates about processes underlying mate choice are discussed.

Abstract. Females of the grey treefrog, Hyla chrysoscelis, from three areas where a genetically incompatible, sibling species, H. versicolor, also occurred, were much more likely to prefer synthetic sounds with fine-temporal properties typical of conspecific males than were females from two areas of remote allopatry. Specifically, more than 90% of the females from sympatric populations weighted a species-specific cue (pulse-repetition rate = pulse rate) more strongly than they did a cue (call duration) that mediates intraspecific mate choice, whereas only about 50% of the females from remote allopatric populations did so. These results suggest that selection for the avoidance of mating mistakes has acted on females in sympatric populations. If this is a general phenomenon, then the apparent rarity of reproductive character displacement may be explained at least in part by the fact that nearly all previous studies have focused solely on geographical patterns of variation in male courtship signals.

Different acoustic properties of communication signals potentially encode different kinds of biologically significant information. Studies of frogs and toads have identified behaviorally relevant properties that correlate with a male's species, body size, energetic investment in calling and perhaps physical and evolutionary fitness. Behavioral experiments reveal that simultaneous variation of different acoustic properties can interact in unpredictable ways to determine the overall effectiveness of a signal.

Patterns of variation in the acoustic properties of advertisement calls of several species of treefrogs (family Hylidae) are related to patterns of female preferences for these same properties. Properties were unequivocally classified as static or dynamic based on within-male variability during bouts of calling. Static properties are highly stereotyped within males and between males in natural populations. Dynamic properties often change by as much as 100% during single bouts of calling; dynamic properties are usually much more variable among males in natural populations than are static properties. An analysis of the calling performance of individual males over the course of a breeding season indicated that at least one dynamic property, pulse number (or call duration) in Hyla versicolor, was sufficiently repeatable to differentiate between males. Playback experiments that employed synthetic calls showed that: (1) females exhibit patterns of preferences that should result in selection that is stabilizing or weakly directional on static properties, and highly directional on dynamic properties; and (2) the overall attractiveness of a signal is affected by variation in both static and dynamic properties. Similar results were found in laboratory and field studies of other taxa, primarily anurans and insects. Patterns of variability in call properties of males and in female preferences are discussed in terms of proximate mechanisms and theoretical models of the evolution of female choice.

Green treefrogs, Hyla cinerea , at a study site in eastern Georgia behaved as prolonged breeders: there were many more males present each night than females, the arrival of which was unpredictable. Individual males were usually consistent in their behaviour during a single night, and were most often observed calling. There was a weak, negative correlation between male size (snout-vent length, unless otherwise stated) and the frequency of satellite behaviour, defined as a non-calling male's being situated near a calling male. In 2 of 3 years, the mean size of males observed as satellites was less than that of the calling males with which they were associated on a given night; however, many satellites were the same size or larger than calling males. The distributions of matings per male during three seasons were not significantly different than Poisson distributions with the same means. Individual mating success was best predicted by attendance at the breeding site; mating success was negatively related to the frequency of satellite behaviour in some analyses, but very little of the variance was explained. An exhaustive series of analyses failed to show any pattern of size-dependent mating, including weight-assortative mating, even when males that were frequently satellites were excluded. The relative sizes of males and females in amplexus had little effect on fertilization success. On a seasonal basis, males were clumped in their spatial distribution, and matings tended to occur where male density was highest. Individual males were highly variable in their patterns of movement, which were not correlated with size or mating success. Mating success was also unrelated to the frequency (Hz) of the low-frequency spectral peak in the advertisement call of the male; male size and weight were negatively correlated with the frequency of the low-frequency peak.

Field measurements and playback experiments designed to assess mechanisms of spacing between calling males of the spring peeper were conducted in central Missouri. In 10 choruses of typical density, mean inter-male distances ranged from 21 to 124 cm and usually varied more than the sound pressure level (SPL in decibels (dB) re 2×10−5Pa) of nearest neighbours' calls (overall range: 88–106 dB SPL; range at mean inter-male distance: 92–98 dB SPL). In playback experiments, conducted in choruses of unusually low density, the lowest sound pressure levels at which males vocally responded to a synthetic advertisement call of 2875 Hz ranged from 48 to 71 dB SPL; the median of 63 dB SPL is a conservative estimate of the effective threshold for antiphonal calling. The sound pressure levels of the calls of each frog's nearest neighbour ranged from 0 to 36 dB SPL (median 11 dB SPL) above its estimated threshold. These values are conservative because background noise precluded testing one-fourth of the forgs at sound pressure levels below 64 dB SPL. Thus, over a wide range of chorus densities, males of the spring peeper usually spaced themselves at places where the call amplitude of their nearest neighbour was well above their auditory threshold.

Four field experiments were designed to study calling and satellite mating strategies in the green treefrog, Hyla cinerea. (1) The calling male was removed from 19 satellite associations and 11 of the 19 satellite males began calling. (2) After the calling male was removed from 10 satellite associations, a speaker broadcast synthetic mating calls. All of the satellite males oriented to the speaker. (3) Synthetic mating calls were played back to 14 calling males. Eleven males stopped calling and oriented to the speaker during at least one trial. (4) Sequences of synthetic mating calls and synthetic encounter calls were broadcast to 12 calling males. Nine males became satellites when the speaker emitted mating calls; none did so when encounter calls were presented.

An extensive series of two-speaker and four-speaker playback experiments using synthetic advertisement calls identified acoustic criteria that may be used by females of the green treefrog, Hyla cinerea, in choosing a mate. At a sound pressure level (SPL) of 75 dB (decibels re 20 μPa) females preferred signals which had a low-frequency spectral peak in the range 0·8–1·0 kHz and a high-frequency spectral peak in the range 2·4–3·6 kHz. Some frequency preferences were reduced or abolished by changing the absolute or relative SPLs of the experimental stimuli. Frequency selectivity in the low-frequency range decreased with playback level; selectivity in the high-frequency range increased with playback level. Females were less selective with respect to variation of the low-frequency peak in four-speaker experiments than in two-speaker experiments. Frequency preferences were independent of female body size. Females responded selectively to signals that differed in call duration and relative SPL; some preferences depended on absolute SPL. Females preferred signals with higher (15–100%) than average repetition rates to those with average rates, but did not respond at all to calls repeated at about four times the normal rate. These results are discussed in terms of current knowledge on auditory physiology and recent data on patterns of call variation and mating success among male green treefrogs.

Males of the spring peeper (Hyla crucifer) in central Missouri produce frequency-modulated, sinusoidal advertisement calls with a duration of 90–250 ms, and a mid-point frequency of 2800–3360 Hz. The frequency of the call is inversely correlated with body size. In playback experiments with synthetic stimuli, females did not prefer a frequency-modulated call to a call of constant frequency. Females preferred a call with a duration of 150 ms to sounds with durations of 40, 75 and 400 ms; a call of 300 ms was just as attractive as the 150-ms call. Females preferred a call of 2875 Hz to alternatives of 4000 Hz and 2600 Hz. The auditory system of H. crucifer is thus only roughly tuned to the temporal and spectral properties of the advertisement call. The female's specificity with respect to duration alone is adequate for species recognition, but intraspecific mate choice based on call frequency is extremely unlikely.