Wilson Bull., 11 l(3), 1999, pp. 368-375

USE OF SONG TYPES BY MOUNTAIN CHICKADEES (POECZLE GAMBELZ)

MYRA 0. WIEBE’s2 AND M. ROSS LEIN1,3

ABSTRACT.-We investigated the composition and function of individual song repertoires in Mountain Chickadees ( gambeli) in Alberta, Canada. Individual males had repertoires of 4-7 song types, but three types made up 90% of all songs. We tested and rejected the hypothesis that all song types convey the same behavioral messages. Different song types were associated with different behavioral situations. Males used 3- note songs predominantly during undisturbed singing and 2-note songs predominantly during non-aggressive activity. Three-note songs with each successive note lower pitched were associated with male-male interactions. We suggest that different song types convey messages indicating different levels of aggression by the singer. The function of individual repertoires in Mountain Chickadees appears to be similar to that of other North American chickadees and titmice, with different song types having different communicative functions. Received 7 Aug. 1998, acceptedII Feb. 1999.

Although individuals of some species of contrast, other species, such as the Blue songbirds sing only one type of song (Seamy (Parus caeruleus), the North American tit- 1983), males of many species possess reper- mice (), and some of the North toires comprising a number of discrete cate- American chickadees (Poecile), seem to pos- gories of songs, or song types (Dodson and sess song types that convey different messag- Lemon 1975). The significance of song rep- es (Smith 1972, Dixon and Martin 1979, Gad- ertoires has been the subject of much research dis 1983, Schroeder and Wiley 1983, Bijnens and speculation (reviewed by Krebs and and Dhondt 1984, Johnson 1987). Kroodsma 1980, Kroodsma 1982, Kroodsma We examine the role of individual reper- and Byers 1991, MacDougall-Shackleton toires of Mountain Chickadees (Poecile gam- 1998). beli). Mountain Chickadees sing relatively Most species of the genera Parus, Poecile, simple songs consisting of 2-6 whistled notes, and Baeolophus have individual song reper- with any number of these notes shifted to fre- toires (Hailman 1989), although the role of quencies lower than the others (Hill 1987). repertoires appears to differ among species Song types are defined easily by variation in [these closely-related genera were formerly number and pitch of notes; individual reper- lumped as Parus (American Ornithologists ’ toires consist of 3-5 song types (Hill and Lein Union 1997)]. Individual male Great Tits (Pa- 1989). Although no studies have investigated rus major) may use different song types in whether Mountain Chickadees can differenti- sequence with no apparent change in the ex- ate among song types, this seems probable be- ternal situation (Hinde 1952), suggesting that cause closely related Black-capped Chicka- all song types convey the same messages. dees (Poecile atricapillus) distinguish among Consequently, Great Tits have been used to songs varying in note number and pitch (Rat- test many of the hypotheses that suggest that cliffe and Weisman 1986; Weisman and Rat- overall repertoire size is important (e.g., Krebs cliffe 1989). 1976, 1977; McGregor et al. 1981; Baker et We documented song variation and singing al. 1986; Lambrechts and Dhondt 1988). In behavior of male Mountain Chickadees during the breeding season. Our null hypothesis was ’ ’ Division of Ecology (Behavioral Ecology Group), that all song types of the Dept. of Biological Sciences, Univ. of Calgary, Cal- convey the same behavioral messages, and gary, Alberta T2N lN4, Canada; thus the song type that a male sings would be E-mail: [email protected] independent of the situation in which it is L Present address: Canadian Wildlife Service, Suite used. The alternate hypothesis was that song 301, 5204 50th Avenue, Yellowknife, Northwest Ter- types convey different messages and thus cer- ritories XlA lE2, Canada; E-mail: [email protected]. tain song types would have a higher proba- 3 Corresponding author. bility of being sung in specific situations.

368 Wiebe and L&n l SONG TYPES OF MOUNTAIN CHICKADEES 369

METHODS gaged in other activities); b. engaged in non-aggressive activities (e.g., foraging, preening, feeding nestlings, The study was conducted at the Barrier Lake site etc.); c. engaged in aggressive activities (i.e. counter- (51” 00’ N, 115” 00’ W) of the University of Calgary’s singing with or chasing territorial neighbors). (2) Po- Kananaskis Field Stations in the Kananaskis Valley of sition on territory: a. within 50 m of the nest site; b. the Rocky Mountains of southwestern Alberta. The away from the nest site. (3) Location of mate: a. pres- forest inhabited by Mountain Chickadees is dominated ent in the vicinity of the singer; b. absent from the by trembling aspen (Po,ouZus tremuloides), white vicinity. Each song was assigned to a particular situ- spruce (Picea glnuca), and lodgepole pine (Pinus con- ation. If the situation changed while the male was sing- tortu). Mountain Chickadees are secondary cavity- ing, then songs recorded before the change were as- nesters, using pre-existing cavities such as natural signed to the first situation, and songs recorded after crevices or deserted nests of other cavity-nesting the change assigned to the second situation. Nests of (Hill 1987, pers. obs.). In southwestern Alberta, Moun- six pairs were found, allowing us to also assign songs tain Chickadees start to search for suitable nesting cav- for these males to particular breeding stages (nest ities during April (Hill 1987, pers. obs.). Nest-building search/build, egg-laying, incubation, or nestling). occurs in early May. Incubation starts near the end of Mountain Chickadees often were out of sight when May and lasts about 14 days. The nestling period gen- in the tops of coniferous trees. We assumed that songs erally lasts 18-21 days (Dahlsten and Copper 1979) given during such intervals were in the same situation but may be as short as 14 days (pers. obs.). Nestlings as the last song given prior to disappearance. However, fledge during the last week in June and early July. if a subject was out of sight for more than 5 minutes, Mountain Chickadee songs, and information on the we recorded its behavior for that interval as unknown. situations of song use, were recorded from April to We categorized different song types by variation in July 1993. Data were collected from 11 males, of number of notes and relative pitch of notes within a which 7 were marked with a unique combination of song. These two features showed the most obvious colored plastic leg bands. We could identify unbanded variation among songs and were relatively easy to dis- males because males in adjacent territories were band- tinguish by ear. We confirmed these classifications by ed; we also used territorial location, use of favorite examining audiospectrographs of many songs using singing sites, or association with a particular nest cav- SIGNAL bioacoustical analysis software (Engineering ity to identify unbanded males. All males except one Design, Belmont, Massachusetts). Transcriptions of re- foraged frequently with one other Mountain Chickadee cordings were made using OBSERVER software (Nol- and therefore were presumed to be mated. Because dus Information Technology, Wageningen, The Neth- Mountain Chickadees defend their entire home range erlands). (Hill 1987), we determined the extent of a male’s ter- We examined separately the relative importance of ritory by noting areas in which he was found regularly. number of notes in a song and the pitch of notes within Observations of aggressive interactions between males a song. To determine the influence of number of notes, helped to confirm the locations of territorial boundar- we combined all song types with the same number of ies. notes regardless of the pitch of the notes within the It was impossible to sample songs of all individuals song. Statistical tests were performed only for 2-note in one day, but usually each individual was observed and 3-note songs because sample sizes for songs with at least once in a three-day period. Observations oc- 1 note and 4 notes or more were too small. To examine curred between 04:OO and 12:00 (MST), the period the influence of pitch, statistical tests were performed when chickadees are most active. An observation pe- only for “common 3-note songs” and “descending 3- riod began when a male started to sing and lasted from note songs” (see descriptions in Results). Sample sizes a few minutes to over an hour, depending on how long of other song types with variations in pitch were too the individual sang. Songs and verbal descriptions of small for statistical testing. the different elements of the situation during singing If different song types are not used in different sit- were recorded onto Sony C-9OHF cassette tapes using uations, one would predict that a particular song type AKG D190E or Sony ECM-33P microphones, Sony should occur in a specific situation at the frequency PBR-330 parabolic reflectors, and Sony TCM-5000 or expected if song types are used at random. To examine Sony TCM-5000EV tape recorders. the influence of note number in songs, we calculated All recordings were made by either MOW or a field x2 values for 2 X 2 contingency tables of either the assistant. Before working independently, the field as- relative number of 2-note songs and all other song sistant accompanied MOW during 15 observation pe- types compared between specific situations of use, or riods to ensure that both observers were making com- the relative number of 3-note songs and all other song parable observations. We did not notice chickadees en- types compared between specific situations of use. To gaging in behaviors directed at the observer during any examine the influence of pitch in songs, we calculated of the observation periods, so we believe that the effect x2 values for 2 X 2 contingency tables of the number of our presence was minimal. of common 3-note songs and descending 3-note songs We recorded three elements of the situation while compared between specific situations of use. The in- males were singing. (1) Behavior of the singer: a. un- dependent event in all contingency tables was a single disturbed (i.e., singing while stationary and not en- song. 370 THE WILSON BULLETIN * Vol. 111, No. 3, September 1999

We conducted a separate contingency analysis for each individual for every situation of use. We were not able to record all individuals in different situations. Consequently, the number of individuals used in each comparison varied from four to seven. We combined the results of individual contingency analyses using a test described by Cochran (1971:lSl). This test ac- counts for differences in direction of response among individuals and can be used even if there is a wide range of sample sizes and probabilities among individ- uals. The test calculates a z-value that can be compared to the normal distribution to determine significance, with non-significant results indicating that the song types are independent of the situation of use. The sign of the z-value indicates the direction of deviation from the expected. Most statistical tests were performed using STATIS- TIX 3.5 for Windows (Analytical Software, St. Paul, Minnesota). Differences at an 01 5 0.05 were consid- ered to be significant.

1.2 2.0 .4 1.2 2.0 RESULTS Time (seconds) Composition of song repertoires.-We re- FIG. 1. Audiospectrograms of song types most corded an average of 1,501 songs (2 435 SE; commonly used by Mountain Chickadees on the study range = 222-4,372) from each of the 11 focal area. A. Common 3-note song. B. Common 2-note males, with an individual male being recorded song. C. Descending 3-note song. D. 2-note song with an average of 15.1 days (+ 2.2; range = 6- both notes the same pitch. E. Three-note song with all 28). Average total repertoire size was 7.4 song notes the same pitch. E Typical 4-note song. All high- quality recordings of songs had the short, upswept note types (2 0.3; range = 6-9). However, song at the beginning of the song that is present in the com- types used rarely by an individual may have mon 3-note, the common 2-note song, and the de- been “accidental” productions rather than scending 3-note song shown in this figure. regular elements in the repertoire. Excluding song types that represented less than 1% of the total songs for an individual, the average dividuals < 4% of total songs recorded were repertoire size was 5.1 song types (2 0.3; descending 3-note songs. All other variants range = 4-7). There was no relationship be- had frequencies of < 12% of an individuals’ tween the number of songs recorded from an repertoire, including 2-note and 3-note songs individual and the estimated size of his rep- with all notes of the same pitch (i.e., within ertoire (all song types: Spearman r = 0.36, P 200 Hz of one another; Figs. lD, 1E) and > 0.05, n = 11; song types > 1%: Spearman songs with four or more notes (Fig. 1F). Wie- r = -0.27, P > 0.05, n = ll), indicating that be (1995) gives details of frequencies of use all repertoires were sampled adequately. of song types by each focal male and the use “Common 3-note songs” (3-note songs of 2-note songs, common 3-note songs and with the last two notes lower in pitch than the descending 3-note songs, in different situa- first note, Fig. lA), “common 2-note songs” tions by individual males. These data are sum- (2-note songs with the second note lower- marized in the following sections. pitched, Fig. lB), and “descending 3-note Variation in song use among breeding stag- songs” (3-note songs with each successive es.-Two-note and 3-note songs were not note at least 200 Hz lower in pitch than the used at random during many breeding stages, previous note, Fig. 1C) were the prevalent although there was much individual variation song types. The most-frequent song type was in the relative use of these song types within common 2-note song for 5 of the 11 focal all stages (Tables 1, 2). Likewise, common males, common 3-note song for 4 males, and and descending 3-note songs were not used at descending 3-note song for 2 males. All males random during different stages and there was sang these three song types, but for three in- also much individual variation (Table 3). Wiebe and L&n l SONG TYPES OF MOUNTAIN CHICKADEES 371

TABLE 1. The use of 2-note songs by male Mountain Chickadees in various situations, compared to all other song types. Columns headed “All males” refer to combined analysis of contingency tables for individual males. Columns headed “Individual males” indicate numbers of contingency tables for individual males with significant departures from expectations,

Individual males All males Preferred song type

Situation n Preferred so@ Z-WlW P ¬e Other NOW

Nest searching/building stage Other -6.77 co.01 2 3 0 Egg-laying stage +1.86 0.06 3 1 2 Incubation stage 2-note +5.99 co.01 3 1 2 Nestling stage Other -2.30 0.02 1 1 2 Near nest 2-note +3.94 CO.01 4 0 2 Mate present Other -2.45 0.01 1 3 3 Undisturbed singing Other -9.11 co.01 0 5 2 Male-male interaction -0.22 0.83 0 0 6

a Indicates that 2.note songs were used significantly more frequently (2.note) or less frequently (Other) than expected m that situation. A blank indicates no sigmficant deviation from expected frequencies.

Variation in song use among behavioral sit- individuals using common 3-note songs pref- uations.-When singing near the nest site, erentially (Table 3). males used 2-note songs significantly more When engaged in undisturbed singing, in frequently (Table 1) and 3-note songs signifi- comparison to singing while engaged in an- cantly less frequently (Table 2) than expected. other non-aggressive activity such as foraging, This pattern also is reflected in the song types 2-note songs were used significantly less fre- used preferentially by individuals. Common quently (Table l), and 3-note songs signifi- 3-note songs were less frequent than expected cantly more frequently (Table 2) than expect- (and descending 3-note songs more frequent) ed. Song type preferences of individuals are when males were singing near the nest, al- entirely concordant with this overall pattern. though there was much individual variation Frequencies of common and descending 3- (Table 3). note songs did not differ from expectations Males sang 2-note songs significantly less during undisturbed singing (Table 3). frequently than expected in the presence of During male-male interactions, common 3- their mates, as indicated by the combined note songs were used significantly less fre- analysis and by the preferred song types of quently, and descending 3-note songs signifi- individuals (Table 1). Common 3-note songs cantly more frequently, than expected (Table were used significantly more frequently than 3). Three of five males showed significant expected in this situation, with five of seven preferential use of descending 3-note songs

TABLE 2. The use of 3-note songs by male Mountain Chickadees in various situations, compared to all other song types. See Table 1 for an explanation of the format.

Individual males All males Preferred song type

Situation n Preferred songa z-value P 3.note Other None

Nest searching/building stage 3-note f5.71 co.01 3 2 0 Egg-laying stage -0.76 0.45 1 3 2 Incubation stage Other -5.55 co.01 0 3 3 Nestling stage +1.55 0.12 1 1 2 Near nest Other -4.82 co.01 0 4 2 Mate present +1.48 0.14 3 2 2 Undisturbed singing 3-note + 10.34 co.01 5 0 2 Male-male interaction +0.59 0.56 0 1 5

a Indxates that 3.note songs were used significantly more frequently (3.note) or less frequently (Other) than expected in that situation. A blank mdicates no significant deviation from expected frequenctes. 372 THE WILSON BULLETIN l Vol. 111, No. 3, September 1999

TABLE 3. The use of common 3-note songs by male Mountain Chickadees compared to the use of descend- ing 3-note songs in various situations. See Table 1 for an explanation of the format.

Individual males Preferred song type

All males Descend- Common ing Situation n Preferred songa z-value P 3-note 3.note None

Nest searching/building stage Common +8.73

a Indicates the 3-note song type that was used slgnificantlv m”re frequently than expected m that situation. A blank indicates no significant deviation from expected frequenaes.

during this situation whereas none used com- the song length analyses should be revealed mon 3-note songs preferentially. in the song pitch analyses. Different song types had significant asso- DISCUSSION ciations with different breeding stages. How- Individual song repertoires.-Excluding in- ever, there was also much individual variation frequent song types, males had individual rep- in all stages, with different males using dif- ertoires of 4-7 song types. Hill and Lein ferent songs preferentially during the same (1989) estimated a slightly smaller repertoire breeding stage. This suggests that although the size of 3-5 song types. However, they sam- z-values were significant statistically, the re- pled songs less intensively and may not have sults may not be meaningful biologically. Dif- recorded enough songs to obtain complete ferent song types are probably not signaling song repertoires for all individuals. messages about the breeding stage of the sing- Other researchers imply that the most com- er. Instead, different breeding stages may be mon song type of Mountain Chickadees is a associated with other situations that are more 3-note song with all notes of the same pitch relevant biologically to the messages of the (Gaddis 1985, Hailman 1989, Hill and Lein song type. 1989). In contrast, we found that males in our The number of notes in songs of Mountain study rarely sang songs of this type. We are Chickadees was associated significantly with unable to explain this difference, but it may location relative to the singers’ nest. All in- support Gaddis ’ (1985) suggestion of geo- dividuals showing significant preferences ei- graphical variation in Mountain Chickadee ther used more 2-note songs near the nest than songs. away from it, or more 3-notes away from the Use of song types in difSerent situations.- nest than near it. Some other studies on song- Our analyses of song length grouped different birds also have found that males sang different song types with the same numbers of notes song types depending on their territorial lo- into a single category, possibly obscuring cation (e.g., Lein 1978, Weary et al. 1994). some patterns in the use of different song Although pitch of the last note in 3-note songs types. However, because almost all 2-note was associated with location relative to the songs recorded were common 2-note songs, nest, this association is weak because of the the results from analyses of all 2-note songs large degree of individual variation. would be almost identical to results using only In Mountain Chickadees, the tendency to common 2-note songs. Almost all 3-note sing 2-note songs near the nest and 3-note songs recorded were either common or de- songs away from the nest may be influenced scending 3-note songs. Any differences be- by factors other than just the singers’ location. tween these song types that were obscured in Males were more likely to engage in non-ag- Wiehe and Lein l SONG TYPES OF MOUNTAIN CHICKADEES 373 gressive activities while singing near the nest could convey more aggressive messages than than when singing away from the nest, sug- 2-note songs. gesting that location relative to the nest and Male-male interactions in Mountain Chick- male behavior were related. Although note adees were associated with changes in pitch number may communicate some information of notes in the song. Because most males were about the singers’ location, these factors may more likely to sing common 3-note songs be related only indirectly and the association when apparently unprovoked by another birds’ between male behavior and note number may activities than during interactions with rival be more important biologically. males, this song type may function in spon- Different song types were associated with taneous advertisement of the territory. De- the presence or absence of the singers’ mate. scending 3-note songs were associated with Other researchers (e.g., Temrin 1986, Staicer male-male interactions and it is probable that 1989) have claimed that song types used pref- lowering the pitch of the last note in 3-note erentially in the presence of females have a songs may convey some message to the rival. greater intersexual function. One methodolog- Other researchers have suggested that song ical problem in our study is that a female may types associated with male-male interactions have been recorded as absent when she actu- probably convey more aggressive messages ally was nearby in the nest hole, but not vis- than do other song types (e.g., Nelson and ible to the observer. Two-note songs and de- Croner 1991). Interactions between males are scending 3-note songs, which were positively situations of high levels of agonistic stimula- tion for Mountain Chickadees and so they associated with the absence of the female, may be more likely to use songs that convey were also positively associated with the nest stronger aggressive tendencies at this time. site. Furthermore, in most cases singers did Thus, descending 3-note songs may convey not seem to be directing song specifically at more aggressive messages than common 3- females. Males sometimes sang low-volume, note songs. “quiet” songs, usually of 1 or 2 notes, when Comparison to closely-related species.- approaching nests to feed incubating mates. Our findings suggest that the function of in- Although such songs may be directed specif- dividual repertoires in Mountain Chickadees ically at females, we never observed normal is similar to that of other North American volume songs used in this manner. Therefore, chickadees and titmice, with different song we are hesitant, without further experimental types used in different situations and appear- study, to suggest that common 3-note songs ing to have different communicative func- have a greater intersexual function than do 2- tions. Three species of North American tit- note songs and descending 3-note songs. mice have certain song types that are used The behavior of singing males was associ- predominantly in male-male interactions ated with the number of notes in the song, but (Gaddis 1983, Schroeder and Wiley 1983, not with pitch. Two-note songs were positive- Johnson 1987). These are probably similar in ly associated with singing while engaged in function to the descending 3-note song of the non-aggressive activity and 3-note songs were Mountain Chickadee, which is also used in positively associated with undisturbed sing- male-male interactions. The Bridled Titmouse ing. Although not all males had individual re- (Baeolophus wollweberi) has one song type sults that were significant, all males showed used predominantly in spontaneous advertise- this trend. ment of territory (Gaddis 1983), and we found We suggest that 3-note songs signal a high- that the common 3-note song of the Mountain er motivation level of the singer to sing than Chickadee is used predominantly in undis- do 2-note songs. If so, 3-note songs might in- turbed singing. Schroeder and Wiley (1983) dicate that the singer is more willing to en- suggested that different song types of the Tuft- gage in some of the agonistic actions associ- ed Titmouse (B. bicolor) convey different lev- ated with singing in Mountain Chickadees, els of aggression by the singer. This corre- such as countersinging bouts or interacting sponds to our suggestion that, in Mountain with other males at the edge of the territory Chickadees, descending 3-note songs, com- to confirm boundaries. Thus, 3-note songs mon 3-note songs, and 2-note songs indicate 374 THE WILSON BULLETIN * Vol. 1 I I, No. 3, September I999 high, intermediate, and low levels of aggres- ation of territorial advertisement. In both spe- sion, respectively. cies, pitch seems to be lowered during more There are also some differences in the man- aggressive situations. Morton (1977) suggest- ner in which titmice and Mountain Chicka- ed that calls of low pitch indicate higher ag- dees use songs. For instance, the Bridled Tit- gressiveness by the signaler than do calls of mouse has a song type used predominantly in higher pitch. This idea seems to be applicable long-distance countersinging (Gaddis 1983). to song in Mountain Chickadees and Black- We did not find any song type in Mountain capped chickadees. Chickadees that was used in this manner, al- though it is possible that some songs that we ACKNOWLEDGMENTS categorized as “undisturbed” singing may S. R. M. Shima provided able assistance in collec- have actually been in response to far away tion of the data. P Rodriguez de la Vega, M. R. Evans, song that was inaudible to the observer. John- J. Bolstad, J. Goddard, and K. Olson also assisted with son (1987) noted that the Plain Titmouse field work. We thank G. Chilton for his help and ad- vice during various stages of the project. The staff at (Baeolophus inornatus) was more likely to the Barrier Lake site of the University of Calgary’s use some song types in situations related to Kananaskis Field Stations provided support and ac- nesting activities. Although 2-note songs were commodation during the field season. R. M. R. Bat- associated with close proximity to the nest, clay, A. I? Russell, J? Asquith, and J. J? Hailman gave there was no indication that Mountain Chick- constructive comments on earlier drafts of the manu- adees were using these songs in any way that script. This research was supported by a graduate scholarship from Ellis Farm Limited and research was related specifically to nesting activity. assistantships from the University of Calgary to MOW, Carolina Chickadees (Poecile carolinensis) and a research grant from the Natural Sciences and have one song type associated with counter- Engineering Research Council of Canada to MRL. singing that is thought to be a more aggressive song type (Smith 1972). This song type could LITERATURE CITED be similar in function to descending 3-note AMERICAN ORNITHOLOGISTS’ UNION. 1997. Forty-first songs given by Mountain Chickadees during supplement to the American Ornithologists’ male-male interactions. However, Smith Union Check-list of North American Birds. Auk (1972) also observed that Carolina Chicka- 114:542-552. dees were more likely to use this aggressive BAKER, M. C., T. H. BJERKE, H. LAMPE, AND Y. ESP- MARK. 1986. Sexual response of female Great Tits song type while patrolling territorial bound- to variation in size of males’ song repertoires. Am. aries whereas we did not note any strong as- Nat. 128:49 l-498. sociation between descending 3-note songs BIJNENS,L. AND A. A. DHONDT. 1984. Vocalizations in and territorial boundaries. a Belgian Blue Tit, Purrs c. caeruleus, popula- Pitch may be an important cue in coding tion. Le Gerfaut 74:243-269. information in the songs of both Mountain COCHRAN, W. G. 1971. Some methods for strengthen- ing the common x2 test. Pp. 132-156 in Readings and Black-capped chickadees, but the two in statistics for the behavioral scientist (J. A. Ste- species differ in how they vary the pitch of ger, Ed.). Holt, Rinehart, and Winston, New York. their songs. Unlike Black-capped Chickadees, DAHLSTEN, D. L. AND W. A. COPPER. 1979. The use of which shift the entire song downward in pitch, nesting boxes to study the biology of the Moun- Mountain Chickadees shift individual notes in tain Chickadee (Parus gambeli) and its impact on a song to a lower pitch. Black-capped Chick- selected forest insects. Pp. 217-260 in The role of adee songs shifted downward in pitch were insectivorous birds in forest ecosystems (J. G. Dickson, R. N. Conner, R. R. Fleet, J. C. Kroll, observed during countersinging between and J. A. Jackson, Eds.). Academic Press, New males in the field and in response to playback York. in both wild and captive birds (Ratcliffe and DIXON, K. L. AND D. J. MARTIN. 1979. Notes on the Weisman 1985, Hill and Lein 1987). Moun- vocalization of the Mexican Chickadee. Condor tain Chickadees 3-note songs, with the last 81:421-423. note lower in pitch (descending 3-note songs), DODSON, D. W. AND R. E. LEMON. 1975. Re-examina- tion of monotony threshold hypothesis in bird were associated with male-male interactions song. Nature 257:126-128. whereas 3-note songs with the last two notes GADDIS, l? K. 1983. Differential usage of song types of the same pitch (common 3-note songs) by Plain, Bridled and Tufted titmice. Omis Stand. were associated with the less-aggressive situ- 14:16-23. Wiehe and Lein * SONG TYPES OF MOUNTAIN CHICKADEES 375

GADDIS, I! K. 1985. Structure and variability in the MACDOUGALL-SHACKLETON, S. A. 1998. Sexual selec- vocal repertoire of the Mountain Chickadee. Wil- tion and the evolution of song repertoires. Cum son Bull. 97:30-46. Ornithol. 14:81-124. HAILMAN, J. P 1989. The organization of major vocal- MCGREGOR, P. K., J. R. KREBS, AND C. M. PERRINS. izations in the Paridae. Wilson Bull. 101:305-343. 198 1. Song repertoires and lifetime reproductive HILL, B. G. 1987. Territorial behaviour and ecological success in the (Parus major). Am. Nat. relations of sympatric Black-capped (Parus atri- 188:149-159. capillus) and Mountain chickadees (Parus gam- MORTON, E. S. 1977. On the occurrence and signifi- be/i) in southwestern Alberta. M.Sc. thesis, Univ. cance of motivation-structural rules in some bird of Calgary, Calgary, Alberta. and mammal sounds. Am. Nat. 111:855-869. HILL, B. G. AND M. R. LEIN. 1987. Function of fre- NELSON, D. A. AND L. J. CRONER. 1991. Song catego- quency-shifted songs of Black-capped Chicka- ries and their functions in the Field Sparrow (Spi- dees. Condor 89:914-915. zella pusilla). Auk 108:42-52. HILL, B. G. AND M. R. LEIN. 1989. Natural and sim- RATCLIFFE, L. AND R. G. WEISMAN. 1985. Frequency ulated encounters between sympatric Black- shift in the fee bee song of the Black-capped capped Chickadees and Mountain chickadees. Chickadee (Parus atricapillus). Condor 87:555- Auk 106:645-652. 556. HINDE, R. A. 1952. The behaviour of the Great Tit RATCLIFFE, L. AND R. G. WEISMAN. 1986. Song se- quence discrimination in the Black-capped Chick- (Parus major) and some related species. Behav- adee (Pants atricapillus). J. Comp. Psychol. 100: iour, Suppl. 2: l-201. 361-367. JOHNSON, L. S. 1987. Pattern of song types use for SCHROEDER,D. J. AND R. H. WILEY. 1983. Communi- territorial defense in the Plain Titmouse Parus in- cation with repertoires of song themes in Tufted ornatus. Ornis Stand. 18:24-32. Titmice. Anim. Behav. 31:1128-l 138. KREBS, J. R. 1976. Habituation and song repertoires in SEARCY, W. A. 1983. Response to multiple song types the Great Tit. Behav. Ecol. Sociobiol. 1:297-303. in male Song Sparrows and Field Sparrows. KREBS, J. R. 1977. The significance of song reper- Anim. Behav. 31:948-949. toires: the Beau Geste hypothesis. Anim. Behav. SMITH, S. T. 1972. Communication and other social 25~475-478. behavior in Pm-us carolinensis. Publ. Nuttall Or- KREBS, J. R. AND D. E. KROODSMA. 1980. Repertoires nithol. Club 1 l:l-125. and geographical variation in bird song. Adv. STAICER, C. A. 1989. Characteristics, use, and signifi- Study Behav. 11:143-177. cance of two singing behaviors in Grace’s Warbler KROODSMA, D. E. 1982. Song repertoires: problems in (Dendroica graciae). Auk 106:49-63. their definition and use. Pp. 125-146 irz Acoustic TEMRIN, H. 1986. Singing behaviour in relation to po- communication in birds. Vol. 2. Song learning and lyterritorial polygyny in the Wood Warbler (Phyl- its consequences (D. E. Kroodsma and E. H. Mill- loscopus sibilatrix). Anim. Behav. 34: 146-152. er, Eds.). Academic Press, New York. WEARY, D. M., R. E. LEMON, AND S. PERREAULT.1994. KROODSMA, D. E. AND B. E. BYERS. 1991. The func- Male Yellow Warblers vary use of song types de- tion(s) of bird song. Am. Zool. 31:3 18-328. pending on pairing status and distance from the LAMBRECHTS,M. AND A. A. DHONDT. 1988. The anti- nest. Auk 111:727-729. exhaustion hypothesis: a new hypothesis to ex- WEISMAN, R. G. AND L. RATCLIFFE. 1989. Absolute and plain song performance and song switching in the relative pitch processing in Black-capped Chick- Great Tit. Anim. Behav. 36:327-334. adees, Parus atricapillus. Anim. Behav. 38:685- LEIN, M. R. 1978. Song variation in a population of 692. Chestnut-sided Warblers (Dendroica pensylvani- WIEBE, M. 0. 1995. The function of song types in the ca): its nature and suggested significance. Can. J. Mountain Chickadee (Parus gambeli). M.Sc. the- Zool. 56:1266-1283. sis, Univ. of Calgary, Calgary, Alberta.