Anim. Behav., 1988, 36, 648-652
Two-choice phonotaxis in Hyperolius marmoratus (Anura: Hyperoliidae): the effect of temporal variation in presented stimuli M I R A N D A L. D Y S O N & N E V I L L E I. P A S S M O R E Department of Zoology, University of the Witwatersrand, Johannesburg, South Africa
Abstract. Female painted reed frogs, Hyperolius marmoratus, were subjected to two-choice discrimination experiments to determine whether temporal overlap in the presented stimuli affects frequency preferences. The results showed that females preferred low frequency calls when the stimuli were presented alternately. Simultaneous presentation of stimuli resulted in a random response by females. When presented stimuli partially overlapped or abutted each other, the females responded significantly more often to the leading call irrespective o fits frequency. These findings provide some insight into the possible effects of male vocal interactions on the ability of females actively to choose mates based on call frequency in natural breeding assemblages.
Although adaptive female choice occurs in species of anurans where males differ in the quality of resources they defend (e.g. Howard 1978), the role of the male phenotype in mate selection remains obscure. Several authors have proposed that differential mating success is generated through active female choice of optimally fit males, facilitated through size-related variation in the advertisement call (e.g. Ryan 1980, 1982; Fairchild 1981; Gatz 1981; Sullivan 1983; Forester & Czarnowsky 1985; Robertson 1986). This explanation for the apparent occurrence of non-random mating in many species ofanurans is based on results of studies that indicate that (1) large males (assumed to be fitter since size is indicative of age and age of survival ability) produce lower frequency calls than small males (e.g. Ryan 1980, 1982; Passmore & Telford 1983; Forester & Czarnowsky 1985; Robertson 1986) and (2) in two-choice phonotaxis experiments females respond preferentially to low frequency calls associated with larger males (Ryan 1980, 1982; Forester & Czarnowsky 1985; Robertson 1986; Telford & Passmore, unpublished data). Size-dependent mating success has in fact been convincingly demonstrated in only two species of anurans where males do not defend resources (Ryan 1980, 1982; Robertson 1986). The bias in mating success towards large male Physalaemus pustulosus in natural breeding assemblages, together with female preferences for low frequency calls associated with large males in two-choice phonotaxis experiments, led Ryan (1980) to claim that he had provided an experimental paradigm to demonstrate the role of female choice as a determi-
nant of male mating success in a natural anuran population. More recently, however, the generality of Ryan's (1980) claim has been questioned with the discovery that in some anuran species the frequency preferences displayed by females in phonotaxis experiments are not reflected in any bias in mating success towards certain males in natural populations. Gerhardt (1982) suggested that the apparent inconsistency evident in the discriminatory behaviour of female Hyla cinerea in two-choice phonotaxis experiments and in natural breeding asSemblages could be explained in terms of differences in the complexity of the acoustic environment to which females are exposed. The acoustic environment in natural breeding assemblages is far more complex than in two-choice phonotaxis experiments and the ability of females to discern slight differences in call frequency under these conditions may be severely hampered (see also Arak 1983; Halliday 1983). Previous studies concerning mate choice in female 11. marmoratus have shown their behaviour to be similar to that of H. cinerea with respect to frequency discrimination under experimental and natural conditions. Females show strong size-dependent frequency preferences when calls are broadcast alternately in two-choice phonotaxis experiments (Telford & Passmore, unpublished data) yet mating is random with respect to male body size in natural populations of this species (Passmore & Telford 1983; Telford & Dyson, unpublished data). Several factors may influence the level of acoustic complexity (and thus the discriminatory ability
Dyson & Passmore: Two-choice phonotaxis of females) in natural breeding assemblages. One of the more obvious is the number of conspecific males calling at any one time. The more males, the more possible choices a responding female has. By increasing the number and variety of calls from which a female must choose, one may effectively reduce the female's ability to discriminate accurately between them. Results of four-choice phonotaxis using frequency (Gerhardt 1982) and intensity (Bishop & Passmore, unpublished data) appear to support this argument. The ability of females to discern slight differences in frequency or intensity is dramatically reduced in four-choice compared with two-choice situations. Additional evidence of the effect of male density on female discriminatory ability comes from results of a cage study conducted on H. rnarmoratus (Dyson 1985). At low male densities there was a significant bias in mating success towards large males. However, as the density of males increased mating became random with respect to body size. This suggests that as the number of males calling in the vicinity of the female increases, variations in call characteristics become obscured and the female's discriminatory abilities are reduced. A factor that is related to male density and may influence the degree of complexity of the acoustic environment is the extent to which there is temporal overlap in the signals of participating males in a chorus. Call overlap would be common in species where males do not call in an organized fashion, and, although less common in species where males actively avoid call overlap, it may still occur under conditions of high male density. The possible effect of call overlap on female discriminatory ability has never been tested in either twochoice or four-choice phonotaxis experiments and may be largely responsible for the random mating patterns observed in natural populations. To determine whether frequency preferences of females may be influenced by the temporal relationship between the calls of neighbouring males we conducted a series of two-choice phonotaxis experiments in which the temporal relationship of the acoustic stimuli was systematically varied. By presenting females with two stimuli that were (1) simultaneous, (2) overlapping by half a call duration (40 ms), and (3) abutting, we were able to determine whether female frequency preferences were influenced by variation in the temporal arrangement of presented stimuli.
METHODS Amplexing pairs were collected from a shallow coastal pond near Mtunzini, Zululand (28°51'S, 31°46'E). The pairs were transported to a clearing about 6 km from the breeding area. Females were separated from the males immediately before testing and placed at the centre of the testing arena (2-0 x 1.5 m) beneath a perforated plastic release container. After an acclimatization period (3 min for the first trial, 2 min for subsequent trials), during which the acoustic stimuli were broadcast, the release container was removed. A positive response was scored if the females touched the loudspeaker or approached to within 5 cm of it. Calls were broadcast using a Nagra IV-S stereo tape-recorder, two Portamon amplifiers and two Philips AD 50600 5-inch diameter loudspeakers. The speakers were placed facing one another 1 m from the release container at either end of the arena. The arena was dimly but evenly illuminated by three 8-W red fluorescent striplights suspended 1.5 m above it. The two synthetic stimuli used in these experiments were identical with respect to the extent of the frequency sweep (500 Hz), duration (80 ms), rise time (10 ms) and fall time (25 ms). Start and stop frequencies of the sweeps were 2650-3150 Hz and 3050-3550 Hz which represented male snoutvent lengths of approximately 29 mm and 23 mm respectively (Dyson 1985). The centre frequency of each sweep is shown in Table I. Each synthetic stimulus was broadcast at 104 dB peak measured at 50 cm which is equivalent to that of natural calls at the same distance. A Bruel & Kjaer 2230 sound level meter and a Bruel & Kjaer type 4165 microphone were used to make these measurements. The experimental procedure is outlined in Table I. Figure 1 shows the temporal arrangement of the stimuli used in each trial. In experiment 1 we tested the responses of females to calls of different frequency varying in their temporal relationship to one another (i.e. simultaneous, 50% overlap, abutting). A total of 17 females were tested, each female being tested once in each of the tests of alternatives (a total of six trials each). Experiment 2 served as a control in which we tested female responses to identical stimuli using the same temporal arrangement of stimuli as in experiment 1. Each of 16 females was tested in four trials. In some cases a
Animal Behaviour, 36, 3
Table I. Acoustic stimuli used in the two-choice discrimi-
nation trials Stimulus 1 Stimulus 2 Temporal centre frequency centre frequency arrangement (Hz) (Hz) of stimulit Experiment 1
Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Trial 6
2900 2900 2900* 3300* 2900* 3300*
3300 3300 3300 2900 3300 2900
Alternate Simultaneous 50% Overlap 50% Overlap Abutting Abutting
2900 2900 2900* 2900
2900 2900 2900 2900*
Alternate Simultaneous 50% Overlap Abutting
Trial 1 Trial 2 Trial 3 Trial 4
* Indicates the leading call. t See Fig. la-f. female failed to respond in all trials, resulting in slight variation in the total number of trials obtained for each stimulus pair. The order of trials in each experiment was random. In addition, we assumed that the experiences of a female in one test of two alternatives did not influence her choice in subsequent tests. The ambient temperature during the experiments ranged from 19.8 to 25.5°C.
RESULTS Results of both experiments are given in Tables II and IlL When two calls of different frequency were broadcast alternately, the females showed a significant preference for the low frequency call. However, when calls were presented simultaneously the females responded randomly to either call. Responses to calls broadcast so that they abutted or overlapped one another revealed a significant preference by females for the leading call irrespective of its frequency. Thus it appears that the temporal relationship of stimuli to one another and the occurrence of stimulus overlap have a substantial influence on female frequency preferences in two-choice phonotaxis experiments.
DISCUSSION The significant difference in female responses to
calls broadcast alternately and calls broadcast simultaneously revealed an important limitation of female discriminatory ability: simultaneous perception of calls by responding females effectively obscures frequency cues which are evident to females when calls are perceived alternately. Although females had no difficulty in localizing simultaneously presented calls (the time taken to reach the loudspeaker in these trials was not significantly longer than that in other trials) they were unable to discern any difference in the frequency of the two calls. This inability was reflected in their random response to either the high or low frequency call in these trials. It appears, therefore, that simultaneous calling by neighbouring males of this species would negate any potential for female mating preferences based on call frequency. In trials where calls were presented so that the second call began halfway through the first, females responded almost exclusively to the leading call regardless of its frequency. A similar result was obtained when females were presented with calls that abutted one another: all preferred the leading call regardless of its frequency. There are several possible explanations for these results. First, partial call overlap or the production of abutting calls may eliminate frequency cues used by females to select optimally fit males and the next best option is to choose the male that calls first over the male that follows. A second alternative is that although frequency cues are not obscured when calls overlap or abutt, females prefer the leading call because call leadership is a better indicator of male fitness than is call frequency. A final alternative is that females do not prefer either low frequency calls or leading calls. Instead, the probability that a female responds to either is entirely dependent on which one acts as a better releaser of the phonotactic response. This explanation suggests that female responses are not adaptive but rather fall into Parker's (1982) category of passive attraction. Which of the above alternatives is the most likely explanation is a controversial issue and its resolution is beyond the scope of this study. However, one important consequence emerges from the results of these experiments: the production of overlapping or abutting calls by neighbouring males in a natural chorus would effectively eliminate any bias in mating success towards large males based on call frequency.
Dyson & Passrnore: Two-choice phonotaxis
5 J¢ 4 >-
,~ 5 Z
6.o8~. Figure 1. The temporal arrangement of stimuli used in the two-choice discrimination trials. (a) Alternate; (b) simultaneous; (c) 50% overlap, 3300 Hz call leading; (d) 50% overlap, 2900 Hz call leading; (e) abutting, 3300 Hz call leading; (f) abutting, 2900 Hz call leading.
In view of this finding, we conducted a preliminary study of the calling behaviour of pairs of H. marmoratus males in an attempt to ascertain whether call overlap occurs in natural choruses of this species. Confirmation of the occurrence of a high degree of call overlap between males could indicate that the random mating patterns observed in natural choruses are largely a result of the lack of chorus organization among males of this species. Our preliminary investigation revealed that H. marmoratus males do in fact call independently of one another and that 30% of the calls of pairs of neighbouring males overlap. This percentage would be expected to increase as the n u m b e r of
near neighbours increases. It appears therefore that female choice of certain males based on call frequency is unlikely in this species. In addition, because of the lack of strict antiphonal calling and the tendency of males to display no fixed chorus order, it is highly unlikely that mate choice based on call order occurs in this species. The question of how relevant the results of this study are to the reproductive behaviour of other species o f a n u r a n s depends largely on the degree to which males in a chorus interact vocally with one another. This in turn may depend to a large extent on the effects of chorus density on the ability of males to alternate their calls with their near
Animal Behaviour, 36, 3
Table II. Phonotactic responses of females in experiment
1 (figures in parentheses indicate the P-value of the twotailed binomial test) Choices made by females Temporal arrangement of stimuli Alternate Simultaneous 50% Overlap 50% Overlap Abutting Abutting
14 5 14" 0 14" 3
3 (0.05) 9 (Ns) 0 (0-01) 14" (0.01) 1 (0.01) 13" (0-05)
* Indicates the leading call. t See Fig. la-f.
Table !1I. Phonotactic responses of females in experiment
2 (control; figures in parentheses indicate the P-value of the two-tailed binomial test) Choices made by females Temporal arrangement of stimuli Alternate Simultaneous 50% Overlap Abutting
Council for Scientific and Industrial Research and the Communication Biology Research Group, Department of Zoology, University of the Witwatersrand. We thank D. Yager (Section of Neurobiology and Behavior, Cornell University), for preparing the synthetic stimuli used in these experiments and Philip Bishop for preparing the stimulus tapes. Our thanks also go to Alan Thornley for reading the manuscript and to Carl Gerhardt for his helpful comments on an earlier draft of this manuscript.
9 9 14" 15"
6 (NS) 6 (NS) 1 (0-01) 1 (0.01)
* Indicates the leading call. t See Fig. I a-f.
neighbours thereby avoiding temporal overlap in their signals. In summary our results are a further indication that two-choice discrimination experiments cannot be simply extrapolated to natural populations since they represent an unnaturally simple choice. U n d e r natural conditions it is doubtful that females will always be presented with calls that alternate precisely: a necessary requirement for accurate frequency discrimination in phonotaxis experiments. Call overlap, together with a number of other factors which may contribute to the complexity of the acoustic environment characteristic of natural breeding assemblages (see Arak 1983) would make mate choice based on call frequency exceedingly difficult for females.
Arak, A. 1983. Male--male competition and mate choice in anuran amphibians. In: Mate Choice (Ed. by P. Bateson), pp. 181-262. Cambridge: Cambridge University Press. Dyson, M. L. 1985. Aspects of social behaviour and communication in a caged population of painted reed frogs, Hyperolius marmoratus. M.Sc. thesis, University of the Witwatersrand. Fairchild, L. 1981. Mate selection and behavioral thermoregulation in Fowler's toads. Science, N.Y., 212, 950-951. Forester, D. C. & Czarnowsky, R. 1985. Sexual selection in the spring peeper, Hyla crucifer (Amphibia, Anura): role of the advertisement call. Behaviour, 92, 112-128. Gatz, J. A. 1981. Size selective mating in Hyla versicolor and Hyla crucifer. J. Herpetol., 15, 114-116. Gerhardt, H. C. 1982. Sound pattern recognition in some North American treefrogs (Anura: Hylidae): implications for mate choice. Am. Zool., 22, 581-595. Halliday, T. R. 1983. Do frogs and toads choose their mates? Nature, Lond., 306, 226-227. Howard, R. D. 1978. The evolution of mating strategies in bullfrogs, Rana catesbeiana. Evolution, 32, 850-871. Parker, G. A. 1982. Phenotype-limited evolutionary stable strategies. In: Current Problems in Sociobiology (Ed. by King's College Sociobiology Group), pp. 173201. Cambridge: Cambridge University Press. Passmore, N. I. & Telford, S. R. 1983. Random mating by size and age of males in the painted reed frog, Hyperolius marmoratus. S. Afr. J. Sci., 79, 353 355. Robertson, J. G. M. 1986. Female choice, male strategies and the role of vocalizations in the Australian frog Uperoleia rugosa. Anirn. Behav., 34, 773-784. Ryan, M. J. 1980. Female mate choice in a neotropical frog. Science, AT.Y., 209, 523 525. Ryan, M. J. 1982. Sexual selection and communication in a neotropical frog, Physalaemus pustulosus. Evolution, 37, 261-272. Sullivan, B. K. 1983. Sexual selection in Woodhouse's toad (Bufo woodhouseO. II. Female choice. Anim. Behav., 31, 1011-1017.
ACKNOWLEDGMENTS This research was supported by grants from the
(Received 26 February 1987; revised 13 June 1987; MS, number: 2971)