ANIMALS often settle disputes by means of conventional displays. It has been suggested that this enables the contestants to assess each other's strength without resorting to a serious fight12,3. Often the outcome of a contest will depend simply on who is the larger, and many displays seem to involve assessment of body size1,4. In this paper we show experimentally that male toads, Bufo bufo, settle contests for the possession of females by means of vocalisations that give a reliable signal of body size and hence of fighting ability.
Toads spend most of the year on land but each spring they mass together in ponds to breed, all the spawning taking place within a week or two. The male clasps on to the female's back (amplexus) and may be carried around by her for several days before she eventually lays her eggs, whereupon he fertilises them externally. Our field observations were made in 1977 and 1978 at a pond near Oxford where we marked several hundred toads individually with numbered elastic waist bands5 or by unique combinations of toe-clipping.
The toads arrived at the pond by night and 84,4 % (n = 77) of the females that we found on land, on their way to the water, already had a male riding on their back. In the pond itself, all the females that we saw were paired (n = 150). Although all the females acquired mates, only quarter of the males did so. The excess of males came about partly because males spent all the breeding season in the pond, some mating more than once, whereas the females visited the pond for a shorter time. Thus, on any one day all the male population was present but only a fraction of the female population. However, this is not the entire story because even over the whole breeding season there was still more males than females.
(To view figures click on the yellow text, as for Fig. 1)
Fig. 1 Single male toads are more likely to persistently attack paired males who are smaller than themselves. A persistent attack is defined as repeated attempts by the single male to dislodge the paired male, involving attachment to the pair and prolonged struggles. Each point is based on 12 laboratory experiments in which a single male was placed in a tank together with a pair. Size classes of male toads: [circle], small (50-54 mm snout-vent length); [triangle], medium (56-60 mm); [square], large (62-66 mm).
The presence of all these unpaired males gave rise to intense competition for mates and we often saw two or more males tussling for the possession of a female both on the land and round the spawning sites in the pond. Single males attempted to dislodge paired males by pushing in between their rival and the female, either from the front or the back. The paired male sometimes managed to kick the attacker away, but if a single male got a hold a vigorous struggle ensued, which could last several hours as the attacker strove to achieve a takeover. Twenty out of 52 marked males (38.5%) were displaced from females in this manner before the female spawned. Males varied in size from 49-70 mm snout-vent length, probably largely a reflection of their different ages. In 11 out of the 14 cases in which the sizes of the competing males were known, the takeover was by a larger male. Similarly, in fights in the laboratory we found that large males could sometimes displace smaller ones (10 out of 23 cases) but not vice versa (none out of 18 cases) 6.
Because body size was such a good predictor of fighting success, we were interested to see whether a single male would ever attempt to engage in a long struggle with a paired male who was larger than himself. In the pond, 11 out of the 13 persistent fights that we observed between unmarked males were cases in which an attacker was attempting to dislodge a smaller male. In the laboratory we also found that males were only likely to persistently attack rivals who were smaller than themselves (Fig. 1). When the paired male was larger, the spare male usually gave up his attack after a brief encounter.
Fig. 2 Fundamental frequency of a defending male's cell is closely related to his body size; larger toads make deeper pitched croaks (y = 2258.75 - 17.83x,r = -0.884, P <0.001). All of the 20 toads were recorded in the laboratory at 15°C and the fundamental frequency was measured from sonograms.
What cues does the attacking male use to assess the size, and thus fighting ability, of his rival? Whenever a male attacked a pair, the defending male always called (n = 156). From sonograms, the structure of this call seems to be identical to that of the 'release call', which is given by a single male whenever he is clasped mistakenly by another male in search of a female7-8. Because the attacker hears the same call in the two contexts, he obviously cannot use it to inform him as to whether the caller is single or paired. Some other cues must be used because an attacker always immediately releases his grip of a single male but may persistently fight with a paired male.
However, once the attacker has detected a pair, he may use the call of the defender to influence his decision whether to attack and attempt a takeover. As reported for other anurans9-12, the frequency of the call is closely related to body size; larger males have deeper pitched croaks (Fig. 2). The fundamental frequency of the call is determined by the mass, length and tension of the vocal cords8,13. Larger individuals have a larger larynx14 and are therefore able to make a deeper croak. Thus, because call frequency is anatomically con¬strained, it could potentially give an attacker reliable information as to the size of his adversary.
To test this hypothesis we collected 24 medium-sized males (56-60 mm snout-vent length) from the pond and carried out an experiment in the laboratory. Twelve of these males were used as attackers against small paired males (50-54 mm) and the other 12 were used as attackers against large paired males (62-66 mm). The paired males were silenced by fitting them with small rubber bands placed behind their arms and passing through their mouths like a horse's bit. This prevented them from croaking, but it did not seem to affect their behaviour in any other way; they still gripped the female tightly in amplexus and kicked out vigorously at any attacker. Females of the same size were used in all the trials and no pairs were used for more than three attackers.
Each attacker was used in two experiments with the same silenced, paired male. In both trials he was placed in a small tank, and above the paired male we fixed a small loudspeaker through which we played tape-recorded croaks of either a small male (high pitched croaks) or of a large male (deep croaks). Previously we had found no significant relationship between body size and the rate of calling (r = 0.084, n= 19) nor the amplitude of the calls (r = 0.401, n = 22), so the croaks were broadcast at the same rate and at the same amplitude in each trial, the only difference being the pitch of the call. Each experiment was carried out at about 15°C and lasted 30 min. Whenever the spare male touched the pair, we broadcast croaks for 5 s and recorded the attacker's behaviour. The size of the tank restricted the movement of the pair, who usually remained quite still and did not behave differently between treatments. Half the attackers heard the high pitched croaks first and half heard the deep croaks first, with about 3 h between each trial.
Fig. 3 Medium-sized males were given the opportunity to attack small or large paired males who were prevented from croaking by rubber bands passing through their mouths. Each attacker was used in the experiments. In one he heard tape-recorded croaks of a small male and in the other the croaks of a large male. One set of 12 males attacked small defenders and another 12 attacked large defenders. Two measures of attacking behaviour were used, the number of attacks and the percentage time attacking. Histograms represent mean scores, with vertical lines indicating s.e.m. See also Table 1.
The results show that an attacker's behaviour can be manipulated by the frequency of the call that we broadcast to him (Fig. 3, Table 1). He was much more likely to attack the paired male when we played a high pitched croak than when we played a deep croak. If the call was the only cue that influenced the attacker's behaviour, then we would predict that when the broadcast call was the same frequency, small and large defenders would be attacked with equal vigour. However, fewer attacks were delivered to the large paired males. So although the playing of a deep croak inhibited attacks against small defenders, we could not induce a male to persistently attack a large defender by broadcasting high pitched croaks. Other cues, such as tactile cues or the strength of the defender's kick presumably gave the attacker information as to the real size of his adversary.
The advantage of calling to a large defender is clear; his deep croaks give other males a reliable indication of his body size and deter potential attackers. The signal is reliable in the sense that the pitch of a large male's croak is resistant to bluff by a smaller male. But why do small defenders call? Most other males are larger than them and high pitched croaks would seem to invite attacks more often than not. Perhaps the reason is that any small defender who remained silent, and was not prepared to give an honest demonstration of his size by croaking, would be immediately under suspicion and would be attacked anyway. We must assume that a small male croaks because it is better than not croaking, partly because it deters those attackers that are smaller than he and perhaps also because the call has other functions such as signalling to the female. Even against large males, small males have a 50% chance of success¬fully resisting a takeover6, so it will always be worth their while to try to maintain amplexus.
We suggest that male anuran calls may often act as a signal of body size, either to other males in agonistic encounters or to females in those species that have well developed mating calls. In mammals, low pitched sounds have also become ritualised as threat signals which may communicate body size or fighting ability and thus permit competing males to assess each other without recourse to costly fights15,16.
We thank H. Bennet-Clark, C. Dawkins, C. Halliday, M. Hunter, J. Krebs, J. Parr and D. Thompson for their help and encouragement.
N. B. DAVIES
Edward Grey Institute,
Department of Zoology,
University of Oxford,
South Parks Road, Oxford, UK
T. R. HALLIDAY
Department of Biology,
The Open University,
Milton Keynes, UK
1 Parker, G. A. J. theor. Biol. 47, 223-243 (1974).
2 Maynard Smith, J. & Price, G. R. Nature 246, 15-18 (1973).
3 Maynard Smith, J. & Parker G. A. Amin. Behav. 24, 159-175 (1976).
4 Dawkins, R. & Krebs, J. R. in Behavioural Ecology: An Evolutionary Approach (ed. Krebs, J. R. & Davies, N. B.) 282-309 (Blackwells, Oxford, 1978).
5 Emlen, S. T. Hepetologica 24, 172-173 (1968).
6 Davies, N. B. & Halliday, T. R. Nature 269, 56-58 (1977).
7 Heusse, H. von. Z. Tierpsychol. 27, 894-898 (1970).
8 Martin, W. F. & Gans, C. J. Morph. 137, 1-27 (1972).
9 Blair, W. F. Q. Rev. Biol. 39, 334-344 (1964).
10 Porter, K. R. Am. Midl. Nat. 71, 232-245 (1964).
11 Zweifel, R. G. Copeia 1967, 269-284 (1968).
12 Nevo, E. & Schneider, H. J. Zool. 178, 133-145 (1976).
13 Martin, W. F. J. exp. Zool. 176, 273-294 (1971).
14 Martin, W. F. in Evolution in the Genus Bufo (ed. Blair, W. F.) 279-309 (University of Texas Press, Austin and London, 1972).
15 Morton, E. S. Am. Nat. 111, 855-869 (1977).
16 Clutton-Brock, T. H. & Albon, S. D. (in preperation).