The rally call recognition in males of two hybridizing partridge species, red-legged (Alectoris rufa) and rock (A. graeca) partridges

Behavioural Processes 55 (2001) 1 – 12 www.elsevier.com/locate/behavproc

The rally call recognition in males of two hybridizing partridge species, red-legged (Alectoris rufa) and rock (A. graeca) partridges M. Ceugniet *, T. Aubin Uni6ersite´ Paris-Sud XI, Equipe Communications Acoustiques, NAMC CNRS UMR 8620, Bat. 446, F-91405 Orsay cedex, France Received 24 July 2000; received in revised form 3 November 2000; accepted 31 January 2001

Abstract The red-legged (Alectoris rufa) and rock (A. graeca) partridges hybridize and produce fertile offspring along a contact zone in French Southern Alps. The rally call emitted during pair formation, could play an important role in species recognition, acting as a behavioral reproductive isolating mechanism between males and females. In the present study, the coding system of the rally call was investigated from captive males of the two species and from F1 hybrids. By playing-back natural signals, we found that the two species as well as hybrid males responded to Alectoris signals but not to another species belonging to the Phasianidae family (Colinus 6irginianus). Results also indicate that red-legged and rock partridges responded stronger to conspecific calls than to heterospecific ones. However, they reacted similarly to conspecific and hybrid calls. F1 hybrids responded stronger to hybrids calls than to the two species ones. They did not distinguish the two parental species signals from each other. Although the two species showed the ability to discriminate the conspecific from the heterospecific signal, they clearly responded to the other species. This behaviour may play a role in the hybridization phenomenon. © 2001 Published by Elsevier Science B.V. All rights reserved. Keywords: Acoustic discrimination; Hybridization; Partridges; Species-specific barrier

1. Introduction Several phenomena may occur when two species encounter each other. First, reproductive differences between them may be accentuated in the zone of sympatry by means of reinforcement of premating isolation (Butlin, 1989), which corresponds, when the reinforcement is complete, to * Corresponding author.

the phenomenon called ‘Character displacement’ (Brown and Wilson, 1956). That is the case in the chaffinches Fringilla coelebs and F. teydea, which come into contact on Tenerife island (Lynch and Baker, 1990). Second, discriminatory behaviour may be reinforced (Becker, 1982) as for species of birds such as blue-winged (Vermi6ora pinus) and golden-winged (V. chrysoptera) warblers (Gill and Murray, 1972), and the firecrest Regulus ignicapillus and goldcrest R. regulus (Becker, 1977). The

0376-6357/01/$ - see front matter © 2001 Published by Elsevier Science B.V. All rights reserved. PII: S0376-6357(01)00141-3

2

M. Ceugniet, T. Aubin / Beha6ioural Processes 55 (2001) 1–12

third possibility is the interspecific convergence of characters, a phenomenon termed ‘Character convergence’ (Cody, 1973), which results in interspecific territoriality between males which represent an advantage for ecologically similar species. In this case, interbreeding is prevented by the retention of species-specific recognition cues that females use to select mates of their own species. Indeed, mate selection is usually the female’s choice in birds (Orians, 1969). The fourth case is the recognition of the other geographic species as a possible mate, or lack of recognition that the other species is different, which leads to hybridization (Rhymer and Simberloff, 1996). This case occurs in birds between blue-winged and golden winged warblers (Gill and Murray, 1972), between pied (Ficedula hypoleuca) and collared (F. albicollis) flycatchers (Alatalo et al., 1990) and between indigo (Passerina cyanea) and lazuli (P. amoena) buntings (Rising, 1983). Hybridization has been shown to be more common in birds than in any other group of vertebrates (Short, 1969; Pierotti, 1987). This phenomenon occurs in two partridge species (geographic species) of the genus Alectoris, the redlegged (A. rufa) and rock partridges (A. graeca). These two species are morphologically (Goodwin, 1986) and genetically (Randi, 1996) differentiated. Their breeding cycles strongly overlap (Birkan, 1990; Bernard-Laurent and Leonard, 1995). They inhabit different environments, but hybridize in the area where their distributions overlap (Bernard-Laurent, 1984). Red-legged partridges (RLP) inhabit the Iberian Peninsula, France and north-western Italy. Rock partridges (RP) are distributed throughout the Alps, Apennines, Sicily and the Balkans, reaching Albania and Greece. Hybridization exists along the border of the Southern French Alps between 800 and 2100 m altitude (Bernard-Laurent, 1984). This region corresponds to a change in climatic conditions and an interface between distinct floristic assemblages, which have been suggested to be responsible for the stability of hybridization zone (Moore, 1977; Grant and Grant, 1992, 1993). Although hybrid fitness is usually decreased in the habitats of the parental species, it has been shown to be superior to that of parental species in such mosaic environ-

ments. Thus, climatic and floristic transitional zones represent a possible cause for the maintenance of the hybrid zone. Contrary to cases of hybridization where hybrids are sterile (flour beetles Tribolium castaneum and T. freemani: Wade and Johnson, 1993; mallard Anas platyrhynchos and muskovy ducks Cairina moschata: Tchelycheva et al., 1993), hybrids between the two species of Alectoris partridges are viable and fertile (Bernard-Laurent, 1990). Acoustic communication often plays a major role in the species-specific recognition process of birds. Concerning acoustic, a sexual dimorphism exists between the two sexes since songs are mainly used by males (Becker, 1982). Male territorial songs present two main functions (Tinbergen, 1939), that are the repulsion of other males and the attraction of a mate. In the present study, we investigated mainly the first aspect of these twin functions. The ‘rally call’ is the most used call in partridges. During the spring, male partridges emit rally calls which main function is the establishment and the defence of a territory (Menzdorf, 1977). In a previous study, we have analysed the characteristics of the partridge rally call (Ceugniet et al., 1999). There were similarities and differences between acoustic parameters of both species. RLP present two types of calls: short (B 80 ms) and long (\ 80 ms), whereas RP emit only short ones. For the short calls, similarities occured in the inter-call silence and in the rhythm (sound/silence ratios). Spectral characteristics of short calls, however, differed between the two species (fundamental frequency, frequency with the highest level of energy and the frequency band containing 80% of the call energy). The present study examined the coding system of the rally call of each species and of hybrid F1 males, in order to investigate the territorial behaviour of the two partridge species. For that purpose, natural signals were played-back to partridges, in order to examine the recognition process. We asked whether the two species discriminated one another’s calls, and whether there was preference for the conspecific signal over the heterospecific one or the hybrid one. The call coding system of hybrids was also investi-

M. Ceugniet, T. Aubin / Beha6ioural Processes 55 (2001) 1–12

gated. We asked whether hybrids exhibited a preference for the signal of one of the parental species.

3

signal followed by 1 min of silence. Bird behaviour was observed during this 5 min period. The order of the different broadcast signals was randomised in order to exclude a bias due to the playback presentation.

2. Materials and methods

2.3. Broadcast signals 2.1. Experimental subjects and periods We tested 13 RLP and 10 RP males for the Section 2.3.1, 10 RLP and 3 RP males for the Section 2.3.2, and 6 HP (hybrid) males for Section 2.3.3. All hybrid males studied were from the F1 generation, resulting from a cross between a female RLP and a male RP, obtained in captivity. Experiments were conducted on birds reared in captivity at the ‘Office National de la Chasse’ center of St Benoist near Rambouillet (France). RLP males came from the center of France and RP males originated from the Alps and Abruzzes. Experiments were conducted during the springs 1998 and 1999. Playbacks were carried out between 6:00 and 10:30 h and between 17:00 and 22:30 h, periods corresponding to the maximal sensitiveness of partridges to the calls (BernardLaurent and Laurent, 1984; Pe´ pin and Fouquet, 1992). Partridges were tested in their rearing cages (dimensions 1.48× 0.47 ×0.36 m). Males were tested in presence of their female mate in order to stimulate territorial behaviour. Indeed, it has been shown that the agonistic behaviour of a paired male is stronger than that of an unpaired male (Green, 1983).

2.2. Playback procedure Signals were broadcast using a Sony stereo TCS-420 tape recorder connected to an autonomous 10 W amplifier manufactured in the laboratory, and an Audax loudspeaker. The intensity of signals was about 62 dB SPL measured at 1 m with a 4176 Brue¨ l and Kjear 0.5%% sonometer. This intensity level was equivalent to that produced in natural conditions. Each playback session consisted in broadcasting three or four stimuli separated by a 15 min duration, in order to avoid habituation. Each stimulus lasted 5 min, with 4 min of broadcasted

2.3.1. Experiment I Three natural signals were played back to RLP and RP males. Signals were natural rally calls of a RLP male, natural rally calls of a RP male, and natural calls of a northern bobwhite Colinus 6irginianus (another species belonging to the family of Phasianidae, but of a different genus). One call for each species is represented on Fig. 1. Signals correspond to call sequences of 5.9692.96 s for RLP, 10.059 3.58 s for RP and 1.569 0.34 s for the northern bobwhite (each sequence always contained only two or three calls) followed by a silence period of 3 s for RLP, 6 s for RP and 15 s for the northern bobwhite, repeated during 4 min. These emission rates correspond to natural ones for each species. Calls of the northern bobwhite constituted the ‘control’ signal. Although this bird belongs to the same family as the studied partridges, it is a different species that partridges might ignore due to different distributions. RLP and RP natural rally calls were chosen according to a previous analysis of the signals characteristics of the two species (Ceugniet et al., 1999). Temporal and frequency parameters had been studied using a Principal Components Analysis (PCA), in order to emphasize species characteristics. The two signals (one for each species) were chosen to be as different as possible, so that they occurred at the two ends of the first component in the PCA. This component was responsible for 82% of the variance among individual calls. 2.3.2. Experiment II In this experiment, RLP and RP males are tested with the two previous RLP and RP signals and with a hybrid F1 signal. This signal, recorded at the ‘Office National de la Chasse’, corresponds to a hybrid male resulting from the crossing female RLP-male RP. This hybrid signal corresponds to call sequences of 11.629 3.68 s

4

M. Ceugniet, T. Aubin / Beha6ioural Processes 55 (2001) 1–12

followed by a 8 s silence period, the whole repeated during 4 min.

2.3.3. Experiment III The four previous natural signals (northern bobwhite signal and RLP, RP and hybrid signals) were played back to hybrid males. 2.4. Classification of responses In natural encounters between males, two behaviours occur: the bird replies vocally to the intruder and then, most often, moves (by walking or flying) towards it (Goodwin, 1953; pers. obs.). In our experiments, it was impossible to take in account the distance of approach since cages were too small. That is the reason why we have chosen to analyse only the acoustic parameters. The following parameters were recorded from the test subjects: (1) the proportion of individuals that responded to the broadcast signal; (2) the latency of the first acoustic response (i.e. the time period between the beginning of the signal broadcast and the first response of the tested partridge); (3) the total duration of the acoustic response; and (4) the total number of rally calls emitted in

reply to the playback session. A maximal latency time of 600 s was attributed to individuals that did not respond during the 5 min test duration.

2.5. Statistical analysis Non parametric tests were used, because the studied values were not normally distributed. The proportions of individuals that responded to the broadcast signal were compared two by two, by the one-tailed Fisher exact test (Scherrer, 1984). The Friedman ANOVA test was used to compare acoustic responses between the three or four signals, and the one-tailed Wilcoxon matched pairs test (Scherrer, 1984) was used for comparisons of responses between pairs.

3. Results

3.1. Experiment I: do the males of the two species disciminate between RLP and RP rally calls? 3.1.1. Red-legged partridges Responses to the three natural stimuli differed significantly for all parameters analysed (Fig. 2,

Fig. 1. Oscillograms (down) and sonagrams (up) of red-legged (left), rock (middle) and hybrid (right) partridge calls.

M. Ceugniet, T. Aubin / Beha6ioural Processes 55 (2001) 1–12

5

Fig. 2. Means 9 SE of latencies, duration and total call number of red-legged partridges (N= 13) (left) and rock partridges (N= 10) (right) responses to natural red-legged ( ) and rock ( ) partridges and to northern bobwhite ( ) broadcast signals. *P B0.05, **PB 0.01

P B0.001 for the response latency, the response duration and the total number of calls: Friedman ANOVA test). The proportion of RLP individuals which responded to the conspecific signal (12/13) is greater than that responded to the northern bobwhite signal (0/13; P B 0.001: Fisher exact

test). Obviously, they responded more rapidly (PB 0.01 for the response latency: Wilcoxon matched pairs test) and longer (PB 0.01 for the response duration and number of calls on reply: Wilcoxon matched pairs test) to the conspecific signal than to the one of the northern bobwhite (Fig. 2).

6

M. Ceugniet, T. Aubin / Beha6ioural Processes 55 (2001) 1–12

More individuals responded to the RP signal compared to the northern bobwhite signal (6/13 versus 0/13; P B 0.01: Fisher exact test). Obviously, RLP responded more rapidly (P B 0.05 for the response latency: Wilcoxon matched pairs test) and longer (P B 0.05 for the response duration and number of calls on reply: Wilcoxon matched pairs test) to the RP signal than to the one of the northern bobwhite (Fig. 2). The proportion of RLP individuals which responded to the conspecific signal (12/13) is greater than that responded to the RP signal (6/13, P B0.05: Fisher exact test). However, the difference was not significant for the latency of the response (Fig. 2, P \0.05: Wilcoxon matched pairs test). Nevertheless, RLP males responded significantly longer (P B0.05: Wilcoxon matched pairs test) and emitted more calls (PB 0.05: Wilcoxon matched pairs test), when the conspecific signal was broadcast (Fig. 2). RLP did not respond to the call of the northern bobwhite. They discriminated between the conspecific signal and the RP one.

3.1.2. Rock partridges Responses of RP to the three stimuli differed significantly for all the analysed parameters (Fig. 2, P B 0.01 for the response latency and P B 0.05 for the duration and the number of calls of the response: Friedman ANOVA test). The proportion of RP individuals which responded to the conspecific signal (9/10) is significantly greater than that responded to the northern bobwhite signal (0/10, P B 0.001: Fisher exact test). Obviously, RP responded more rapidly (PB0.01 for the response latency: Wilcoxon matched pairs test) and longer (P B 0.01 for the response duration and number of calls on reply: Wilcoxon matched pairs test) to the conspecific signal than to the one of the northern bobwhite (Fig. 2). More RP individuals responded to the RLP signal compared to the northern bobwhite signal (4/10 vs. 0/10, P B 0.05: Fisher exact test). Obviously, RP responded more rapidly (P B0.05 for the response latency: Wilcoxon matched pairs

test) and longer (PB 0.05 for the response duration and number of calls on reply: Wilcoxon matched pairs test) to the RLP signal than to the one of the northern bobwhite (Fig. 2). The proportion of RP individuals which responded to the conspecific signal (9/10) is significantly greater than that responded to the RLP signal (4/10, PB 0.05: Fisher exact test). Moreover, RP responded faster to the conspecific signal than to the RLP one (Fig. 2, P B 0.05: Wilcoxon matched pairs test). However, responses of RP males to the conspecific signal and to the RLP signal did not differ significantly, either for the duration of the response or for the number of calls (Fig. 2, P\ 0.05 for the two parameters: Wilcoxon matched pairs test). RP did not respond to the call of the northern bobwhite. They discriminated to some extent the conspecific signal from the RLP one.

3.2. Experiment II: do the males of the two species disciminate between conspecific and hybrid rally calls? 3.2.1. Red-legged partridges Results are presented in Fig. 3. Acoustic responses by RLP to the three stimuli did not differ significantly for all the analysed parameters (P\ 0.05 for latencies, durations and number of calls of the responses: Friedman ANOVA test). The proportion of RLP individuals which responded to the conspecific signal (9/10) is similar to the one that responded to the hybrid signal (9/10, P= 0.76: Fisher exact test) and to the one that responded to the RP signal (7/10, P= 0.29: Fisher exact test). Similarly, the proportion of RLP individuals responding to RP and hybrid signals did not differ (P= 0.29: Fisher exact test). For RLP males, the hybrid signal is as reactive as the conspecific one. 3.2.2. Rock partridges Results are presented in Fig. 3. Acoustic responses by RP to the three stimuli did not differ significantly for all the analysed

M. Ceugniet, T. Aubin / Beha6ioural Processes 55 (2001) 1–12

parameters (P \ 0.05 for latencies, durations and number of calls of the responses: Friedman ANOVA test). The proportion of RP individuals which responded to the conspecific signal (2/3) is similar to the one that responded to the hybrid signal (1/3, P =0.5: Fisher exact test) and to the one that responded to the RLP signal (2/3, P = 0.8: Fisher exact test). Similarly, the proportion of RP indi-

7

viduals responding to RLP and hybrid signals did not differ (P= 0.5: Fisher exact test). For RP males, the hybrid signal is as reactive as the conspecific one.

3.3. Experiment III: how hybrid males percei6e the two parental and hybrid rally calls? Hybrid responses to the four stimuli differ signifi-

Fig. 3. Means 9SE of latencies, duration and total call number of red-legged partridges (N= 10) (left) and rock partridges (N= 3) (right) responses to natural red-legged ( ), rock ( ) and hybrid ( ) partridge broadcast signals.

8

M. Ceugniet, T. Aubin / Beha6ioural Processes 55 (2001) 1–12

Fig. 4. Means 9 SE of latencies, duration and total call number of hybrid (N =6) responses to natural red-legged ( ) and rock ( ) partridges, northern bobwhite ( ) and hybrid partridge ( ) broadcast signals. *P B0.05.

antly (Fig. 4, PB 0.01 for the response latency and PB 0.05 for the duration and the number of calls of the response: Friedman ANOVA test).

The proportion of HP individuals which responded to the hybrid signal (6/6) is significantly greater than that responded to the northern bobwhite signal (0/6, PB 0.01: Fisher exact test). Obviously, they responded more rapidly (PB 0.05 for the response latency: Wilcoxon matched pairs test) and longer (PB0.05 for the response duration and number of calls on reply: Wilcoxon matched pairs test) to the hybrid signal than to the one of the northern bobwhite. The proportion of HP individuals which responded to the RP signal (3/6) did not differ from that responded to the RLP signal (3/6, P\ 0.05: Fisher exact test). Similarly, responses of HP males to the two parental signals did not differ, either for the latency, the duration of the response, or for the number of calls (P\ 0.05 for the three analysed parameters: Wilcoxon matched pairs test). For each parental species signal (RLP and RP), three of six HP males responded compared to zero of six for the northern bobwhite signal (P\ 0.05 for the comparison with each species: Fisher exact test). Similarly, neither latency of the response nor duration and call number of the response differed in hybrids between each parental species signal and the northern bobwhite signal (P\ 0.05 for the three parameters: Wilcoxon matched pairs test). For each parental species signal (RLP and RP), three of six total HP males responded compared to 6 of 6 for the hybrid signal (P\ 0.05 for the comparison with each species: Fisher exact test). However, HP responded more rapidly (PB 0.05 for the response latency: Wilcoxon matched pairs test) and longer (PB 0.05 for the duration and number of calls of the response: Wilcoxon matched pairs test) to the hybrid signal than to the RLP one. Moreover they also responded more rapidly (PB 0.05 for the response latency: Wilcoxon matched pairs test) and longer (PB 0.05 for the duration and number of calls of the response: Wilcoxon matched pairs test) to the hybrid signal than to the RP one. Hybrid males did not respond to the call of the northern bobwhite. They responded equally to the two parental species, but clearly prefered hybrid calls over RLP and RP ones.

M. Ceugniet, T. Aubin / Beha6ioural Processes 55 (2001) 1–12

4. Discussion

4.1. Males discriminate conspecific rally calls Concerning call recognition, we showed that the two studied species of partridges did not respond to the northern bobwhite signal but responded to the two partridge ones. Thus, the partridges showed a specificity for the Alectoris genus calls. Moreover, RLP males discriminated the conspecific signal from the heterospecific one since, not only more individuals responded to the conspecific signal, but also their acoustic response was longer for the RLP broadcast signal. Similarly, RP males also reacted more aggressively to the conspecific signal than to the RLP one (perhaps to a lesser degree than RLP males), since more individuals responded to the conspecific signal. Other birds have been shown to respond more to the conspecific signal than to the heterospecific one in allopatric populations: lazuli and indigo buntings Passerina cyanea and P. amonea (Baker, 1991), nightingales Luscinia luscinia and L. megarhynchos (Sorjonen, 1986), warblers Acrocephalus schonobaenus with A. scirpaceus and with A. palustris (Catchpole, 1978). More, the Darwin’s finches, G. fortis, responded more to the conspecific song than to the one of G. fuliginosa on the Galapagos Island of Plaza Sur, and G. fuliginosa and G. difficilis responded weakly to each other on the Island of Pinta (Ratcliffe and Grant, 1985). The discrimination of the conspecific signal by the two species of partridges may be due to frequency parameters, often implicated in species recognition of signals of numerous species (Robin, Erithacus rubecula, Bre´ mond, 1967, 1968; Collared Dove, Streptopelia decaocto, Gu¨ rtler, 1973; Nighjar, Caprimulgus europaeus, Abs, 1963; Spotted Sandpiper, Actitis macularia, Heidemann and Oring, 1976). In a previous study (Ceugniet et al., 1999), we demonstrated that the frequency of the maximal amplitude and the frequency bandwidth were the factors that differ the must between the two species of partridges. Concerning the hybrid signal, the two species do not seem to distinguish it from the conspecific signal.

9

Hybrids did not respond to the northern bobwhite signal, whereas they reacted to both Alectoris signals. Moreover, they did not discriminate one parental species signal from the other. Such results have also been found for hybrid males between indigo and lazuli buntings which reacted equally to songs of the two parent species (Baker, 1991). Concerning the hybrid stimulus, HP exhibited an interesting pattern of response: they clearly reacted stronger to hybrid calls than to RLP and RP calls. If we compare the coding-decoding systems of the rally call of both the partridge species and their hybrids, differences appear. For RLP, the coding characteristics, according to the principal components analysis in Ceugniet et al. (1999) were homogeneous. In the same way, according to our playback experiments, the decoding process of calls of this species were relatively precise: RLP reacted stronger to conspecific calls. Thus, the coding-decoding processes are ‘symmetrical’. On the other hand, although RP exhibited great variability in their call characteristics (Ceugniet et al., 1999), they present the ability to discriminate the two partridge calls since more individuals responded to the conspecific signal. For RP, the decoding process seem more precise than the broad coding one. On the contrary, it has been shown for the cirl bunting, Emberiza cirlus, that decoding ability was greater than coding ability, (i.e. that artificial rythms faster and slower than natural ones were recognized by the bird; Kreutzer, 1983). For this author, a larger decoding system compared to the coding one may constitute a barrier against speciation.

4.2. Acoustic male beha6iour and e6olution of the hybrid zone Currently, the natural hybrid zone between RLP and RP stretches on about 15 km along the border of the French Southern Alps (BernardLaurent, 1984). The proportion of each type of partridge in the hybrid zone is: 86% hybrids, 10% RP and 4% RLP (Bernard-Laurent, 1984). Hybrids contain F1 individuals, backcrosses and other recombinants (Randi and Bernard-Laurent, 1999), which seems to indicate introgression at the

10

M. Ceugniet, T. Aubin / Beha6ioural Processes 55 (2001) 1–12

altitude borders both up and down stream of the hybrid zone. Consequently, RLP and RP seem to encounter each other infrequently as hybrid populations are spread out between them. Our study focused mainly on the territorial behaviour of males. Our results showed that RLP males reacted similarly to conspecific and hybrid calls. This suggests that they defend their territories against other RLP as strongly as against hybrid males. Hybrid males will consequently be stopped in their progress towards low altitudes, i.e. in RLP territories. On the other side of the hybrid zone, the situation seem to be similar as we showed that RP males responded to conspecific individuals as strongly as to hybrid ones. Thus, RP males will prevent hybrids from moving forward in their territories, i.e. towards high altitudes. Our results from male hybrids indicated that they responded more to hybrid calls compared to the two parental species calls. Consequently, they will more likely allow RLP and RP than other hybrid individuals to enter their territories. Moreover, studies on comparative reproductive biology of RP and hybrids, showed that hybrid productivity was higher than that of RP (Bernard-Laurent, 1987). The greater number of young per adult comes from a bigger brood size and from a greater number of broods per adult. Considering these data, we can suggest that the hybrid zone will maintain its width without extensing neither towards high altitudes nor towards low ones, with RLP and RP entering the hybrid zone and supplying RLP and RP gene flow towards the center of the hybrid zone.

4.3. Acoustic and non acoustic causes of hybridization In birds, the mate choice is generally made by the female (Orians, 1969). However, it is the male that establishes the territory. Concerning male behaviour, although the discrimination of the two Alectoris calls is performed by the two species, response to the heterospecific calls is not negligible for RLP and RP individuals. The recognition of heterospecific calls by the two species of partridges, might be due to the intercall silence between short calls which has previ-

ously been found to be similar in the two species (Ceugniet et al., 1999). However, in order to have a more precise view of hybridization mechanisms, female responses to male calls remain to be studied. Other causes, outside the acoustic range, could explain the hybridization phenomenon between the two species. First, the different timing of pair formation in the two species may play a role in hybridization. Indeed, since RLP begin to mate one month before RP (Green, 1984; Bernard-Laurent and Leonard, 1995), a great majority of RLP individuals are mated when RP individuals begin to pair. Moreover, previous studies suggested that the critical threshold of mate acceptability declines as the cost of searching increases (Alatalo, 1988; Real, 1990). The remaining RLP at the end of the pair formation period may then choose to mate with RP rather than remaining unpaired and failing reproduction. A second cause of hybridization may result from male dominance. Some females may prefer mating with greater and dominant heterospecific males rather than with conspecific ones (female black ducks, Anas rubripes with male mallards, A. platyrhynchos: Brodsky and Weatherhead, 1984; Brodsky et al., 1988). Since RP males are larger than RLP ones (550–650 versus 400–480 g : Bernard-Laurent and Leonard, 1995), they may dominate them in encounters. Consequently, RLP females might exhibit a preference for RP males rather than for conspecific males. To conclude, we demonstrated in this study (1) that RLP and RP males reacted both to conspecific and heterospecific calls (the other partridge species and hybrids); but (2) that the two parental species reacted stronger to conspecific calls over the other species ones but not over hybrid ones; and (3) that hybrid males responded more to hybrid calls over RLP and RP ones. Those acoustic behaviours may play a role in the hybridization phenomenon existing between these species. Nevertheless, to specify this acoustic investigation, female responses to playbacks of conspecific and heterospecific calls of males also need to be studied.

M. Ceugniet, T. Aubin / Beha6ioural Processes 55 (2001) 1–12

Acknowledgements We are grateful to the ‘Office National de la Chasse’ and to D. Soyez for the logistic support. We would like to thank F.S. Dobson for helpful comments and improvement of the English and J. Ge´ nermont for statistical help. We thank also V. Bre´ tagnolle and two anonymous referees for their helpful criticisms of the manuscript. Financial support was provided by the CNRS.

References Abs, M., 1963. Field tests on the essential components of the European Nightjar’s song. Proc. Int. Ornithol. Congr. 13, 202– 205. Alatalo, R.V., 1988. The search cost in mate choice of the pied flycatcher. Anim. Behav. 36, 289 –291. Alatalo, R.V., Eriksson, D., Gustafsson, L., Lundberg, A., 1990. Hybridization between pied and collared flycatchers — sexual selection and speciation theory. J. Evol. Biol. 3, 375– 389. Baker, M.C., 1991. Response of male indigo and lazuli buntings and their hybrids to song playback in allopatric and sympatric populations. Behaviour 119, 225 – 242. Becker, P.H., 1977. Verhalten auf Lauta¨ usserungen der Zwillingsart, interspezifische Territorialita¨ t und Habitatanspru¨ che von Winter- und Sommergoldha¨ hnchen (Regulus regulus, R. ignicapillus). J. Ornithol. 118, 233 – 260. Becker, P.H., 1982. The coding of species-specific characteristics in birds sounds. In: Kroodsma, D.E., Miller, E.H. (Eds.), Acoustic Communication in Birds, vol. 1. Academic Press, New York, pp. 213 –252. Bernard-Laurent, A., 1984. Hybridation naturelle entre perdrix bartavelle (Alectoris graeca saxatilis) et perdrix rouge (Alectoris rufa rufa) dans les Alpes-Maritimes. Gibier Faune Sauvage 2, 79 –96. Bernard-Laurent, A., 1987. De´ mographie compare´ e d’une population de perdrix bartavelle (Alectoris graeca saxatilis) et d’une population d’hybrides (A. G. saxatilis × A. rufa rufa) dans les Alpes-Maritimes. Rev. Ecol. (Terre Vie) (suppl) 4, 189 –197. Bernard-Laurent, A., 1990. Biologie de reproduction d’une population de perdrix rochassie`res Alectoris graeca saxatilis × Alectoris rufa rufa dans les Alpes me´ ridionales. Rev. Ecol. 45, 321 –344. Bernard-Laurent, A., Laurent, J.L., 1984. Me´ thode de recensement des perdrix bartavelles Alectoris graeca saxatilis au printemps; applications dans les Alpes-Maritimes. Gibier Faune Sauvage 4, 69 –85. Bernard-Laurent A. and Leonard P. 1995. La Perdrix Bartavelle. Brochure technique no 23. Off. Natl. Chasse, Paris.

11

Birkan M. 1990. La Perdrix Rouge. Brochure technique no 10. Off. Natl. Chasse, Paris. Bre´ mond, J.-C., 1967. Reconnaissance de sche´ mas re´ actoge`nes lie´ s a` l’information contenue dans le chant territorial du Rouge-gorge (Erithacus rubecula). Proc. Int. Ornithol. Congr. 14, 217 – 229. Bre´ mond, J.-C., 1968. Recherches sur la se´ mantique et les e´ le´ ments vecteurs d’information dans les signaux acoustiques du Rouge-Gorge. TerreVie 68, 109 – 220. Brodsky, L.M., Weatherhead, P.J., 1984. Behavioural and ecological factors contributing to american black duckmallard hybridisation. J. Wildl. Manage. 48, 846 – 852. Brodsky, L.M., Ankley, C.D., Dennis, D.G., 1988. The influence of male dominance on social interactions in black ducks and mallards. Anim. Behav. 36, 1371 – 1378. Brown, W.L. Jr., Wilson, E.O., 1956. Character displacement. Syst. Zool. 5, 49 – 64. Butlin, R., 1989. Reinforcement of premating isolation. In: Otte, D., Endler, J.A. (Eds.), Speciation and its Consequences. Sunderland, Massachusetts, pp. 158 – 179. Catchpole, C.K., 1978. Interspecific territorialism and competition in Acrocephalus warblers as revealed by playback experiments in areas of sympatry and allopatry. Anim. Behav. 26, 1072 – 1080. Ceugniet, M., Aubin, T., Bernard-Laurent, A., Soyez, D., 1999. Vocal signatures of the rally call of red-legged and rock partridges and of their hybrids. C.R. Acad. Sci. 322, 887 – 895. Cody, M.L., 1973. Character convergence. Annu. Rev. Ecol. Syst. 4, 189 – 211. Gill, F.B., Murray, B.G., 1972. Discrimination behavior and hybridization in sympatric Blue-winged and Goldenwinged Warblers. Evolution 26, 282 – 293. Goodwin, D., 1953. Observations on voice and behaviour of the red-legged partridge Alectoris rufa. Ibis 95, 581 – 614. Goodwin, D., 1986. The partridges of the genus Alectoris. Avic. Mag. 92, 23 – 36. Grant, P.R., Grant, B.R., 1992. Hybridization of bird species. Science 256, 193 – 197. Grant, B.R., Grant, P.R., 1993. Evolution of Darwin’s finches caused by a rare climatic event. Proc. R. Soc. Lond. B 251, 111 – 117. Green, R.E., 1983. Spring dispersal and agonistic behaviour of the Red-legged partridge (Alectoris rufa). J. Zool. Lond. 201, 541 – 555. Green, R.E., 1984. Double nesting of the red-legged partridge Alectoris rufa. Ibis 126, 332 – 346. Gu¨ rtler, W., 1973. Artisolierende Parameter des Revierrufs der Tu¨ rkentaube (Streptopelia decaocto). J. Ornithol. 114, 305 – 316. Heidemann, M.K., Oring, L.W., 1976. Functional analysis of Spotted Sandpiper (Actitis macularia) song. Behaviour 56, 181 – 193. Kreutzer, M., 1983. L’asyme´ trie entre le codage et le de´ codage du tempo, facteur d’e´ volution des chants chez le bruant zizi (Emberiza cirlus). C. R. Acad. Sc. Paris, Life Sciences 297, 71 – 74.

12

M. Ceugniet, T. Aubin / Beha6ioural Processes 55 (2001) 1–12

Lynch, A., Baker, A.J., 1990. Increased vocal discrimination by learning in sympatry in two species of chaffinches. Behaviour 116, 109 – 126. Menzdorf, A., 1977. Beitrag zur kenntnis der lauta¨ ußerungen von steinhu¨ hnern. Beitr. Vogelkd. 23, 85 –100. Moore, W.S., 1977. An evaluation of narrow hybrid zones in vertebrates. Q. Rev. Biol. 52, 263 – 277. Orians, G.H., 1969. On the evolution of mating systems in birds and mammals. Am. Nat. 103, 589 – 603. Pe´ pin, D., Fouquet, M., 1992. Factors affecting the incidence of dawn calling in red-legged and grey partridges. Behav. Process. 26, 167 – 176. Pierotti, R., 1987. Isolating mechanisms in seabirds. Evolution 41, 559 – 570. Randi, E., 1996. A mitochondrial cytochrome b phylogeny of the Alectoris partridges. Mol. Phylogenet. Evol. 6, 214 – 227. Randi, E., Bernard-Laurent, A., 1999. Population genetics of a hybrid zone between the red-legged partridge and rock partridge. Auk. 116, 324 –337. Ratcliffe, L.M., Grant, P.R., 1985. Species recognition in Darwin’s finches (Geospiza, Gould). III. Male responses to playback of different song types, dialects and heterospecific songs. Anim. Behav. 33, 290 –307.

.

Real, L., 1990. Search theory and mate choice. I. Models of single-sex discrimination. Am. Nat. 136, 376 – 405. Rhymer, J.M., Simberloff, D., 1996. Extinction by hybridization and introgression. Annu. Rev. Ecol. Syst. 27, 83 – 109. Rising, J.D., 1983. The great plains hybrid zones. Curr. Ornithol. 1, 131 – 157. Scherrer, B., 1984. Biostatistique. Gae¨ tan Morin, Que´ bec. Short, L.L., 1969. Taxonomic aspects of avian hybridisation. Auk 86, 84 – 105. Sorjonen, J., 1986. Mixed singing and interspecific territoriality — consequences of secondary contact of two ecologically and morphologically similar nightingale species in Europe. Ornis Scand. 17, 53 – 67. Tchelycheva, L.A., Delpech, P., Popescu, C.P., Rodionov, A.V., 1993. Arreˆ t de la condensation chromosomique chez l’hybride interspe´ cifique entre cane commune et canard de Barbarie, le mulard. Ann. Ge´ ne´ t. 36, 190 – 193. Tinbergen, N., 1939. The behaviour of the snow bunting in spring. Trans. Linn. Soc. NY 5, 1 – 94. Wade, M.J., Johnson, N.A., 1993. Reproductive isolation between two species of flour beetles, Tribolium castaneum and T. freemani: variation within and among geographical populations of T castaneum. Heredity 72, 155 – 162.