- McGill University
Summary
Male song sparrows in southeastern Ontario have repertoires of five to 11 distinct song types. The singer repeats each song type a variable number of times before switching to another type. Analysis of territorial singing suggested that rate of switching song types is positively correlated with intensity of agonistic stimulation, where ‘agonistic’ signifies conflict-related. The song-switching hypothesis was tested with playback experiments which varied stimulus intensity in five different ways: (1) presence or absence of song stimuli; (2) addition of visual to auditory stimuli; (3) location of the song stimulus 20 m inside or outside territory; (4) nesting phases of the subjects; and (5) switching rate and rate of novel song type production of the playback (i.e. stimulus versatility). The subjects' switching rates and flight rates consistently increased with greater stimulation, although song rate tended to remain unaffected. The signalling of response intensity through adjustments in switching rate exemplifies how song repertoires function in agonistic communication.
Methodology
Experimental subjects were territorial male song sparrows sampled within a 24-kin radius of the Queen's University Biological Station, Chaffey's Locks, Ontario. The population was fully migratory, arriving between late March and early April, and departing in mid- to late October. Each subject was exposed to a single playback presented between 0600 and 1000 hours, between 1 May and 14 July in 1979 and 1980. A total of 114 playbacks were performed. Subjects were not retested within a breeding season.
Experimental Apparatus
Playbacks of conspecific songs were broadcast through an Atlas loudspeaker connected with a Fanon 12-W amplifier and a Uher 4200 tape recorder. Playback songs were recorded in 1978 from the full repertoire of one song sparrow which lived within the area covered by the present experiments, but which was not tested. One set of playbacks presented a stuffed, mounted song sparrow in addition to broadcast songs. The mount was placed inside a box which had a hinged lid and was supported by a 1.5-m stake that stood next to the loudspeaker. By pulling a string from 20 m away, one of the two observers could raise the lid and expose the stuffed bird during trial periods.
Marking of Territories
Two to five days before a playback experiment, one side of a subject's territory was marked with a unique combination of coloured tapes. The marking procedure entailed playing a recording of song sparrow song (other than the playback tape) from a movable tape recorder. The human observer moved away from the tape recorder after placing it beside a potential song post, which was marked if the subject alighted there and gave three or more songs. The procedure ended when one side of the 'singing territory' was delimited from the territory boundaries of contiguous neighbours, or from areas not defended by other song sparrows.
Playback Design Table I shows the nine categories of playback experiments that were presented to song sparrows. All playbacks contained a pre-trial period in which subjects were only observed; a trial in which we broadcast 36 songs with periods of 15 s (the average song period in natural contexts for the same population was 14.5 s; Kramer & Lemon 1983); and a post-trial period in which no stimuli were presented. Pre-trial periods lasted 2 min in 1979 and 5 min in 1980; trial and post-trial periods each lasted 9 min in both years. To correct for this discrepancy, frequency counts of pre-trial responses that were recorded in 1979 were multiplied by 4.5, and those recorded in 1980 were multiplied by 1.8. As seen in Table I, in eight of nine playback categories songs were broadcast only, and in one category the mounted bird was also presented. The main body of experiments, that is the song-only playbacks, are classified in two ways: (1) by the placement of the loudspeaker 20 m inside or outside a marked territorial boundary, and (2) by the versatility of the broadcast, i.e. number of switches and song types per 36 playback songs. The eight categories of song playbacks were actually four playback tapes (experiments I-IV), each performed inside and outside the territory. These categories will henceforth be designated as experiments I-IN, II OUT, etc. Song versatility was increased from experiment I through experiment IV. The first song of a playback was considered to be a switch. Thus experiment I presented one switch and one song type (type A), repeated 36 times. Experiment II presented three switches and three song types, each repeated 12 times. Experiment lII also presented three song types, but bout lengths were shortened to four songs; the set of bouts A, B and C was repeated three times, producing nine switches. Experiment IV simply had nine switches and nine song types, each repeated four times, The ninth category of playback was identical to experiment I-IN, but the mounted bird was also exposed during the trial period.