Ornis Fennica 83:153161. 2006
Heterospecific rival recognition in the Black Redstart (Phoenicurus ochruros)
Ondøej Sedláèek*, Blanka Cikánová & Roman Fuchs
Sedláèek, O., Charles University, Faculty of Science, Department of Ecology, Vinièná 7, CZ-128 44 Praha 2, Czech Republic. [email protected] (* Corresponding author) Cikánová, B., University of South Bohemia, Faculty of Biological Sciences, Braniovská 34, CZ-370 05 Èeské Budìjovice, Czech Republic. [email protected] Fuchs, R., University of South Bohemia, Faculty of Biological Sciences, Braniovská 34, CZ-370 05 Èeské Budìjovice, Czech Republic. [email protected]
Submitted 23 September 2005, revised 3 May 2006, accepted 25 May 2006
It has long been hypothesized that birds learn to recognize heterospecific competitors and for that reason males behave more aggressivelyagainst intruders in areas of local sym- patry. Recently, it has been confirmed that individuals of one species are able to learn spe- cies-specific visual and acoustic characteristics of heterospecific rivals and that theyare able to remember these associations. Here, we test hypotheses about the importance of learning in interspecific interactions, long-term memoryof species-specific traits and earlytiming of this learning to the post-fledging period in the Black Redstart (Phoenicurus ochruros). We show that Black Redstart males respond to playbacks of Redstart (P. phoenicurus) onlyin the areas where the two s pecies come into close contact. This result demonstrates that Black Redstart males learn individuallyto recognize the ri - val species in areas of local coexistence. Black Redstart males responded in heterospeci- fic experiments before the arrival of Redstarts from wintering grounds. This means that birds are able to retain the memoryof specific traits of heterospecific competitors to the next breeding season. Moreover, yearling and older males, which differ in plumage colouration, responded in heterospecific experiments with the same frequencyand inten - sity. This result indicates that young males learn to recognize Redstarts in the post-fledg- ing period before a first contact with heterospecific rivals in a territorial context.
1. Introduction that some interactions mayresult from misdirected intraspecific aggression, which is a consequence Birds often defend resources against individuals of of a failure to discriminate a different species with another species with similar ecologyliving in local similar appearance and/or song. sympatry(Cody1969). The original view of inter - Catchpole (1978) predicted that if a general specific aggression (Simmons 1951, Orians & failure to recognize different species is responsible Wilson 1964, Cody1969) supposes that it is an for interspecific aggression (Murray1981), indi - adaptive behaviour, which reduces interspecific viduals should show the same response to hearing competition between sympatric species. On the a playback of another species in a population other hand, Murray(1971, 1976, 1981) suggests where this other species also occurs (sympatric po- 154 ORNIS FENNICA Vol. 83, 2006 pulation) as in a population where this other spe- birds are able to learn species-specific signal traits cies does not occur (allopatric population). Many (song, plumage, behaviour etc.) (Gill & Murray experimental playback studies addressed this 1972, Irwin & Price 1999), there was no direct question about the origin and ecological relevance evaluation of whether individuals of one species of interspecific aggression. Lower levels of ag- are able to learn species-specific characteristics of gression in sympatry were found for Dendroica heterospecific intruders and if theyare able to re - warblers (Murray& Gill 1976), Vermivora war- member these associations. blers (Crook 1984), Acrocephalus warblers (Leis- Matyjasiak (2005) tested in his dual-choice ex- ler 1988) and Fringilla finches (Slater & Catch- periments the heterospecific recognition ability pole 1990, Lynch & Baker 1991). Although the and long-term memoryof Blackcaps (Sylvia atri- occurrence of stronger heterospecific response in capilla) that behave aggressivelytowards Garden allopatric populations implies misidentification Warblers (S. borin). He showed that male Black- (Murray1981), this result could be interpreted al - caps can associate species-specific songs with spe- ternativelyas an adap tive abilityto distinguish cies-specific plumage type. As he carried out the species which poses no ecological threat (Slater & playback experiments on Blackcap males before Catchpole 1990). the arrival of Garden Warblers, he showed that Stronger interspecific aggression in sympatry birds retain the memoryof these associations from was seen in Passerina buntings (Emlen et al. year to year. He also asked when Blackcap males 1975), Empidonax flycatchers (Prescott 1987), learn to recognize heterospecific intruders and as- Luscinia warblers (Lille 1988, Sorjonen 1986) and sumed that this abilitydevelops before the first Certhia treecreepers (Gil 1997). This aggressive breeding attempt in young Blackcaps. However, behaviour has been interpreted as being adaptive he was unable to estimate the age of tested males in these cases as it results in interspecific terri- and he documents his result onlyindirectlyusing a toriality, which prevents exploitation competition. proportion of particular age categories of males Martin and Martin (2001a, b) confirmed that be- living in the studyarea. havioural responses of Vermivora warblers to The Black Redstart (Phoenicurus ochruros) heterospecifics reflect ecological interactions of and Redstart (P. phoenicurus) provide a unique the species. opportunityto test hypotheses about the origin of Catchpole (1978) and Catchpole and Leisler interspecific aggression, learning of heterospe- (1986), who studied the interspecific aggression of cific recognition and long-term memoryof this Acrocephalus warblers, confirmed stronger hete- learned identification ability. rospecific response in the area of coexistence with New molecular data support the view that the the opposite warbler species. As in previous cases, Black Redstart and the Redstart are not very theysuggest that this behaviour serves as the terri - closelyrelated (Ertan 2002). There are major dif - torial exclusion of competing species. Moreover, ferences in the song structure (for spectrograms theyfirst pointed out that birds individuallylearn see e.g. Cramp 1988) and male colouration be- to recognize and respond to the songs of conge- tween them. Both species have a slate-greyhead neric competitors onlyin areas of the species and upperparts, orange-chestnut tail and rump. close contact. However, the Redstart male has black sides of the A number of recent studies have shown the im- face and throat, white forehead and orange-chest- portance of learning as a source of heterospecific nut flanks, whereas the Black Redstart has black recognition abilityin other contexts such as brood flanks and white wing-patches. parasitism (Payne et al. 2000, Slagsvold & Hansen The two species inhabit different types of habi- 2001), sexual rival and mate recognition (Hansen tats under natural conditions. Black Redstarts pri- & Slagsvold 2003, Slagsvold et al. 2002), the es- marilybreed in open rockyterrain while Redstarts tablishment of mating preferences during speci- favour open woods (Cramp 1988). Both redstart ation and premating reproductive isolation of species can associate with human beings and they sympatric species (Grant & Grant 1996, 1998, maycome into close contact in a mosaic urban en - Price 1998, Irwin et al. 2001). vironment. In such a situation, theyoccupyexclu - Although it has long been hypothesized that sive territories of different habitat compositions. Sedláèek et al.: Heterospecific rival recognition in Black Redstart 155
Each of the two species occupies habitat similar to 2. Material and methods that preferred under natural conditions (Sedláèek et al. 2004). Nevertheless, spontaneous interspeci- 2.1. Study sites fic conflicts were frequentlyobserved in areas where territories adjoin and the interspecific ag- Experiments were conducted in two towns (50 km gression was reliablyconfirmed in playbackex - apart) located in the south of the Czech Republic periments. The Black Redstart is dominant in both Bøeznice (3500 inhabitants; 13E56E, 49E34N) natural and simulated encounters (Sedláèek et al. and Èeské Budìjovice (100,000 inhabitants; 2004). Both redstart species are migratory. Red- 14E29E, 49E00N). starts migrate to sub-Saharan Africa, whereas The studyarea in Bøeznice is veryheteroge - Black Redstarts winter in the Mediterranean area neous, consisting of a mosaic of diverse urban (Cramp 1988). Black Redstarts arrive at breeding habitats industrial sites, several types of built-up areas about 3 weeks before Redstarts and leave areas (urban villas, terraced houses and apartment about 1 month later. Thus the two species cannot houses), sports fields, gardens, tree alleys, three interact with each other during migration and in small and one larger park. Populations of both spe- their winter quarters. cies are mixed and heterospecific individuals may In Black Redstarts, males exhibit delayed frequentlybe immediate neighbours, thanks to the plumage maturation during their first breeding mosaic nature of diverse urban habitats. We label season. The young males entering their second this site as sympatric. year of life are female-like dull grey, whereas In the larger town, Èeské Budìjovice, the two older males are black with a white wing-patch redstart species are present, but theyare spatially (Cramp 1988). well separated due to the much coarser mosaic of In this study, we investigated the response of urban habitats. Black Redstart males were tested the Black Redstart to playback of Redstart songs in within two large housing estates where onlythis a larger citywhere Black Redstarts occur sepa- species is present. These housing estates consist of rately(Èeské Budìjovice) and in a small town large blocks of flats alternating with open, grassy (Bøeznice) where the two species come into close areas with interspersed trees and shrubs. We label contact. The objective was to address three ques- this site as allopatric. tions: (1) Is the interspecific aggression of Black Redstarts stronger in the area where territories of the two species are mixed and heterospecific indi- 2.2. Mapping of redstart territories viduals are immediate neighbours? (2) Do Black Redstart males respond in heterospecific experi- In 1999, we investigated the number of territories ments before the arrival of Redstart males from of both redstart species in Bøeznice and Èeské their wintering grounds? (3) Do yearling males re- Budìjovice. We carried out three surveys during spond to heterospecific playback before the arrival the breeding season at the end of April, middle of of Redstarts? Mayand middle of June. If a male was seen in the Stronger responses in sympatry would imply samegeneralareainatleasttwoofthreecensuses that Black Redstarts learn heterospecific recogni- a territorywas recorded at the location. The map - tion and aggressive response to Redstarts. Positive ping covered the whole area of Bøeznice (about answers to the other questions would demonstrate 200 ha) and the northern part of Èeské Budìjovice that Black Redstarts remember once-learned he- (about 770 ha), representing all urban habitats in- terospecific intruders from year to year and that cluding the largest park. young Black Redstart males learn to recognize a heterospecific competitor in the year of their birth, before theyuse this abilityin a territorial context. 2.3. Experiments
We carried out the playback experiments between the 1st and 20th of April in the years 2002 and 2003, before the arrival of Redstarts from spring migra- 156 ORNIS FENNICA Vol. 83, 2006 tion. The experiments consisted of song playback ments belonged to the same Redstart song dialect with a fixed artificial stuffed dummy. We used two at onlyone of the studied localities, which are only differentlycoloured dummies for conspecific 50 km apart, and that this might stronglyinfluence playback because of the occurrence of delayed- the responses of Black Redstart males in hetero- plumage-maturation in the Black Redstart (Land- specific experiments. mann & Kollinsky1995) and the possible weaker The sample was 25 males (15 in 2002 and 10 in response of 1-year-old males (in their second cal- 2003; 11 1-year-old and 14 older males) in the endar year of life) to older males (Cucco & Mala- sympatric and 33 (20 in 2002 and 13 in 2003; 12 1- carne 1999). Males were presented with dummies year-old and 21 older males) in the allopatric area. of an age class identical to their own age class We located territoryboundaries before experi - (graydummypresented to 1-year-oldmales, black ments and we placed the sound-producing dummy dummyto older males). For auditoryplaybackwe at distances of 240 m from a singing or foraging used a CD-player (Sony D-EJ611) with two loud- male inside the territoryat a site where each male speakers (Panasonic RP-SP70). We used the same could definitelysee the dummy.The experiment equipment at both localities, the amplitude was duration was 10 min. This time was measured standardized and it was similar to that of natural from the first response (approach, threat displayor singing. Loudspeakers were placed 1.515.0m pass attack) of the male. If we did not record any above the ground. We did not perform control response (no approach to the dummy, no threat tests, because we did not record anyaggressive re - displays, or the male flew away), the experiment sponses to control playback (Robin Erithacus was finished after 10 min from the start of song rubecula) in our previous study(Sedláèek et al. playback. Behaviour and position of a focal male 2004) and pilot experiments. was continuouslytape-recorded during the whole Playback tapes were prepared from recordings trial period. of Pelz (1993). These recordings consist of a typi- Each male was tested first with conspecific and cal song of one male of each species. The Black second with heterospecific playback. We used this Redstart recording was 73 s long with 6 songs; the sequence to be sure of accurate location of the ex- Redstart recording was 58 s long with 8 songs. perimental apparatus inside a territoryand that These recordings were repeated continuallydur- Black Redstart males were prepared to respond to ing experiments. Although the use of a single song playback, at least to the conspecific song. As all sample for each playback treatment could be con- birds tested gave some initial response in con- sidered as pseudoreplication (Kroodsma et al. specific experiments, all of them are included in 2001), there are reasons to justifyit. We suppose the analysis. The two trials in a given territory were that the problem of intraspecific song variation is separated byat least 2 hours and the speaker was not that severe in our case. Firstly, songs from two placed in an identical location. different species with verydistinct songs are used. The observed response of the bird was re- It is likelythat the specific traits outweigh local or corded continuouslyduring the experiment and individual differences in the song structure that dictated into a dictaphone. We recorded locality, might affect the response. Secondly, we did not playback presented to a male, date and time of the find anydifferences in intraspecific responses be - trial and age of the experimental bird. The follow- tween the two localities suggesting that males at ing variables were measured: 1. latencyof re - both localities recognized species-specific play- sponse, 2. nearest distance to the loudspeaker, 3. back. Thirdly, we cannot definitely exclude an in- time spent within a 15 m circular plot around the fluence of dialectal differences of Redstart songs loudspeaker, 4. time spent singing during the ex- between the two studied localities in heterospeci- periment, 5. number of pass flights over the fic experiments. However, the recording of dummy(flying back and forth over the dummy Redstart comes from Pruánkyvillage (16 E59E, within 10 m), 6. time spent in threat display(ruf - 48E50N), South Moravia, Czech Republic. This fled feathers, frequent tail-shivering and squatting, localityis ca 220 km from È. Budìjovice and 260 shaking) (Landmann & Kollinsky, 1995) and 7. km from Bøeznice. It is highlyimprobable that the number of physical attacks (kicking the dummy, song recording of Redstart used in our experi- pecking). Sedláèek et al.: Heterospecific rival recognition in Black Redstart 157
Conspecific 2.4. Statistical analyses 100 sympatry; n = 25 allopatry; n = 33 We compared the proportions of males that re- 80 sponded by approach, threat displays, pass flying and physical contacts using Fisher Exact Tests; P- 60 % values were adjusted for multiple testing 40 (Bonferroni adjustment). Significant levels refer to a = 0.0125 in these tests. 20 Principal Component Analysis (PCA) was 0 performed for visualizing the differences and the approach aggressive display pass attack physical contact Heterospecific variabilityin response of Black Redstart males in 100 sympatry; n = 25 conspecific and heterospecific trials at the two lo- ** allopatry; n = 33 ** calities. PCA was computed using CANOCO 4.5 80 ** and visualized in CANODRAW 4.0 (ter Braak & ** milauer 2002). PCA is a noncanonical multi- 60 % variate statistical technique that enables one to re- 40 duce the dimensionalityof the original set data by * extracting a few complex orthogonal variables, 20 and it describes variation with a minimal loss of in- 0 approach aggressive display pass attack physical contact formation (ter Braak & milauer 2002). strength of aggression
Fig. 1. Responses of Black Redstart to conspecific 3. Results and heterospecific playback and dummies in popula- tions allopatric and sympatric with Redstart. Histo- grams show proportions of males responding by ap- 3.1. Population size of the two Redstart proach within 15 m, aggressive display, pass-flight at- species in Bøeznice and Èeské Budìjovice tack or physical attack. Fisher’s Exact Tests; a = 0.0125 after Bonferroni adjustment for multiple test- In Bøeznice, we located 74 territories of Black ing: * p = 0.0298, NS; ** p <