Sounds

Vocalizations: Song

Vocal Array

Song of pure whistled notes without detectable harmonics, arranged in pattern, typically including at least 1 major pitch change after first or second note. Borror and Gunn (1965) described 9 note types and 15 song patterns, 2 of which comprised 80% of their sample (711 birds). Most common pattern (ascending) begins with long, steady note followed by series of notes of higher pitch (see Fig. 3); second note, often a 3-part triplet, and third, often a steady note upslurred at beginning, are followed by variable number of similar triplets (occasionally doublets). Second most common song (descending) begins with 2 similar steady notes, followed by a third lower-pitched steady note and a series of triplets similar in pitch to third note (Fig. 3).

Songs of this species often paraphrased as Old Sam Peabody or Oh Sweet Canada, implying pattern beginning with two steady notes followed by triplets, unlike common songs just described. Borror and Gunn (1965) found only 3 such songs and suggest this previously common pattern has almost disappeared. Similarly Tuttle (1993) found few birds with this pattern in Adirondack Park, NY. Note length, mean 0.46 s, range 0.02–1.17; internote intervals, about 0.1–0.2 s; song length, mean 2.87 s, range 1.24–6.06 (n = 483; Borror and Gunn 1965). Most songs have 5 notes; variation in length among and within individuals depends on number of terminal triplets (0–16) (JBF). Triplets are replaced by doublets in central British Columbia (Scott Ramsay, pers. comm.). Frequency range 1.5–6.6 kHz, much narrower for individual songs. Major pitch changes more constant among birds if expressed in relative rather than absolute frequency (Hurly et al. 1991). Thus patterns have characteristic pitch intervals. There are also relationships between durations of some notes and internote intervals (Weary and Weisman 1992).

Individuals usually sing 1 stereotyped pattern (about 10% have 2, with 1 sung infrequently). Pitch less variable than timing (Borror and Gunn 1965); loudness, length of songs, and rate of delivery vary greatly (Falls 1969). Differences among individuals are chiefly in pattern and absolute pitch; 2 birds sharing a pattern may sing at different pitches. Some patterns widespread, others occur locally; proportions vary spatially and over time. No clear dialects described (Lemon and Harris 1974). Neighbors’ songs usually no more alike than those of nonadjacent birds (JBF). Song pitch tends to be lower in dense coniferous and mixed forest than in more open habitats (Wasserman 1979, Waas 1988).

Males of both morphs and WS females sing; TS females sing very rarely. WS and TS males share same patterns; songs of TS males, which tend to occupy denser forests, tend to be lower in pitch (JBF). WS females sing same patterns as males, but songs are shorter and notes less steady in pitch (Lowther 1962).

Development

Similar to related White-crowned Sparrow (Marler and Tamura 1964) but less fully documented. Piper (1995) described subsong as soft gurgling whistles given with the bill closed during normal activities and plastic song as loud whistles with truncated notes and phrases delivered with the bill open, often from high perches. Plastic song was often given during or after dominance interactions among hand-raised birds. Subsong and plastic song were both given during the first 9 months, as were alarm and contact calls.

The following is based on Thorneycroft (1967). Subsong (wide-frequency twitter and other sounds) was heard in captive young in spring and fall, in wild young captured in fall, before full song in birds injected with testosterone propionate in winter, and in spring adults (both sexes in all cases). Captive young produced typical whistles, and in some cases full song; most sang the following spring. Those isolated from nestling stage through full song development produced single patterns of normal length with at least 1 typical whistle, some repetition, at least 1 pitch change, no triplets, and many abnormal sounds. Isolates exposed to tape-recorded songs produced fewer abnormal sounds; some developed imitations of tutor tapes. Although birds housed together usually develop different songs, some copy others. Sensitive period for song learning about 30–100 d. Later changes in songs of captives occur in their second autumn. Thus, several features of song appear to be innate, whereas tutoring is needed for production of normal patterns.

Which birds act as tutors in nature is unknown; neighbors at the time territories are established usually have different songs and so are unlikely candidates. Songs of territorial adults remain the same year to year (JBF). Songs of a wild-caught Dark-eyed Junco x White-throated Sparrow hybrid included notes of both species (Jung et al. 1994).

Phenology

Song is occasional by mid-Jan, more regular by late Feb (Connecticut; Saunders 1947) and during spring migration. On breeding grounds (Algonquin Provincial Park, Ontario and Adirondack Park, New York), song is frequent at arrival in early May, with peaks in early Jun and early Jul (JBF, Tuttle 1993). Wasserman (1980; New Hampshire) found that song peaked just before females arrive. The smaller peak in Jul is associated with initiation of second broods (Tuttle 1993). After near absence in Aug, song is heard occasionally in fall migration and winter, although often imperfect (Saunders 1948, Lowther and Falls 1968) and associated with gonadal recrudescence (Quenzel and Helms 1974). In other species fall song is given largely by young of the year (Rost 1992, D. Kroodsma pers. comm.). This has not been tested in White-throated Sparrows. Singing is reduced by inclement weather and cold temperature (Lainevool 1966).

Timing of song in relation to the breeding season varies with sex and morph. For WS males in Algonquin Park, song is frequent early in the season, less so in incubation and nestling periods; increases again when young of the first brood fledge (Falls 1969). TS males sing much less than WS males, but their singing increases as season progresses. WS females sing infrequently before incubation begins (about as often as TS males but more often if stimulated by song playback) and give quiet, partial songs as late as nestling stage. TS females rarely sing (D. J. Loncke unpubl. data).

Daily Pattern Of Singing

In Algonquin Park in May and Jun, steady singing begins about 0330 Eastern Standard Time (EST), increases to a peak about 0400 (dawn chorus), declines irregularly to a low level in afternoon, followed by a short burst of singing about 2000, after which it decreases rapidly. By late Jul and Aug, only a few songs at 0430 and 1900. Both in Algonquin Park and in Churchill, Manitoba, regular singing begins and ends when light intensity starts to change rapidly (Falls 1969). Occasional songs at night during breeding season. Lowther noted night song more frequent with full moon than on dark nights (Lowther and Falls 1968). Song type does not vary throughout day. No data available on differences in daily pattern between sexes or morphs.

Effects of time of day and distance on transmission quality were investigated in open fields and dense forest. Mean quality deteriorates with distance but does not vary with time of day. However variability of quality is greater at midday than at dawn in both habitats. The dawn chorus may optimize consistency, important in conveying identity (Brown and Handford 2003).

Places Of Singing

Males deliver bouts of song mainly from a few song posts, usually coniferous trees taller than surrounding vegetation. Singers typically perch in peripheral branches about two-thirds of the way up a tree and above surrounding trees, but may sing higher or lower, even occasionally on ground. Song-post height in Algonquin Provincial Park: 10.0 m ± 5.3, n = 125; singing height: 6.7 m ± 4.5, n = 178. Total songs and rate of singing are proportional to song-post height. Females sing from lower perches. Attenuation of recorded song is less with playback at 3 m than at lower levels (Jones 1987). Singing tends to diminish from territory centre toward boundary. Singing area smaller during incubation and nestling stages than earlier or later in season (Wasserman 1982).

Repertoire And Delivery Of Song

Most birds have 1 song pattern; those with 2 sing 1 less often. Songs given in bouts with intersong intervals ≥ 10–15 s. Longest songs occur after a break in singing, and song length (number of terminal triplets) tends to decrease during a bout (Falls 1969). Close countersinging males alternate their songs (Wasserman 1977a); temporal overlap of close playback with a male’s songs stops its singing (JBF). Song duetting between pair members not observed. For singing posture see Agonistic Behavior: Visual Threat Displays.

Social Context

Song in the breeding season is a territorial advertisement given only in defended area. Vigorous countersinging often occurs before and after chasing or fighting in establishment of territory boundaries. WS birds may use song in intensive aggressive interactions even in winter (Schlinger 1990). Males (and sometimes WS females) respond to song playback by approach, chases, and countersinging. Intruders rarely sing in presence of a singing male but do so frequently if territory holder removed (singing by neighbors also increases). If a WS male is removed and replaced by recordings of its song, much less intrusion is apparent, especially near speakers, than if no song is played (Falls 1988).

Song may also function in mate attraction and certainly influences females. Male singing decreases with arrival of a mate; unmated males continue frequent singing. Removal of females increases singing by former mates, pairing again decreases it (Wasserman 1977b, JBF). Both wild and captive females respond to male song with trills and precopulatory displays (Thorneycroft 1967, JBF). Females implanted with estradiol respond more strongly to longer or more frequent songs but not to a greater variety of song patterns (Searcy and Marler 1984, Wasserman and Cigliano 1991). Neural effects of song stimulation in the auditory forebrain suggest that elevated levels of estradiol are necessary for females to distinguish conspecific song from a sequence of 1000Hz synthetic tones (Maney et al. 2006). Hearing courtship song has both short and long term effects on the neurendocrine system. White-throat females photostimulated with long days responded within an hour to 42 min of conspecific song by increasing levels of luteinizing hormone and concomitant neural changes. Exposure to synthetic tones had less effect (Maney et al. 2007).

Quiet or short songs (quiet songs usually short) are often given by (1) males close to a territory boundary, (2) close to a rival or an observer, (3) when drawn into a neighbor’s territory by playback, (4) when presented with a WS model, or (5) as a first response to playback (JBF). Collins and Houtman (1999) observed the latter effect only in TS males. Such songs may indicate ambivalence of singer and elicit weak responses from other males. Wasserman (1980) reported an increase in Whisper Song (may be different from quiet songs) during incubation period, especially if male and female were separated. Rate of singing increases in agonistic encounters or in response to playback but may decrease with prolonged playback (habituation); infrequent singing elicits weak responses (Falls 1969). All these variations may signal motivation of a singer; they also provide variety in singing of a bird with a single song pattern. (See also Behavior: Spacing, Territoriality and Dominance).

Species Recognition

White-throated Sparrows typically respond only to conspecific song but may respond weakly to congenors (White-crowned and Harris’ sparrows), which have somewhat similar songs (Rees 1973, Pleszczynska 1980). Responses to playback of altered songs show that several features (pure tones, normal pitch patterns, temporal patterns) are important for species recognition, whereas others (triplets, patterns of loudness) are unnecessary (Falls 1963, Hurly et al. 1992).

Individual Recognition

Recognition of mates by song is likely but unproven; males recognize and respond less strongly to neighbors’ as opposed to strangers’ songs, depending on location. Responses to modified songs indicate that individual recognition is based mainly on absolute pitch of song (Falls 1969, Lemon and Harris 1974, Brooks and Falls 1975a, b, Falls and Brooks 1975). White-throats display accurate pitch perception in laboratory trials; this ability is impaired in isolates (Weisman et al. 2004).

Control Mechanisms

Neural control is presumed similar to that in other song birds (De Voogt 1994). Neuroanatomical studies of White-throated Sparrows taken in the pre-laying period showed brain nuclei associated with song much larger in males than in females. HVC volume, associated with song complexity, did not differ between color morphs of males. However, several other brain structures associated with song production were larger in WS than in TS birds. Neither sexes nor morphs differed in brain weight. Results were consistent with similarity of songs and differences in amount of song between morphs (De Voogd et al. 1995).

Song output is primarily from the left side of the syrinx (Lemon 1973). Song is normally associated with gonadal growth in spring (e.g. Wolfson 1953) and coincides with gonadal recrudesence in fall (Kuenzel and Helms 1974), yet dominant birds may sing in winter when gonads are regressed and circulating androgen levels low (Schlinger 1990, Wiley et al. 1993). Testosterone implants induce song out of season, but dominant birds show the greatest effect. Androgen levels and social status apparently interact to influence song output (Wiley et al. 1993). Differences in amount of song by WS and TS males early in the breeding season are associated with differences in circulating testosterone levels, which are higher in WS birds (Swett and Bruener 2003, 2008, Spinney et al. 2006). Modulation of song production by sex steroid hormones may occur at sites of catecholaminergic cell groups in the brain stem. White-throated Sparrows (and other song birds) do not have additional cell groups beyond those in non-songbirds (Balthazart and Ball 1996)

Configuration of the vocal tract in song production involves beak movement. During song independent contractions of upper and lower jaw muscles control beak motion (Hoese and Westneat 1996). Degree of beak opening is positively correlated with frequency of sound produced (Westneat et al. 1993). Immobilizing the beak in a partly open position results in attenuation of relative amplitude of higher frequency sounds with associated loss of tonal quality (Hoese et al. 2000).

Vocalizations: Calls

Begging Call

Young produce Begging Call (from hatching to shortly after fledging) which changes with age. First is short, high-pitched, quiet squeak accompanying gaping. At 3–4 d, a louder, wide-frequency rasp about 0.14 s long develops; heard most often when young are disturbed. At 8 d, a call starting with an upslur and continuing with frequency modulation from about 4.0–6.5 kHz (about 0.2 s long) begins to replace rasp. By 12 d this call is shorter (0.1 s) and modulation rate slower. Call continues to change, becoming shorter while frequency range narrows; beginning upslur is lost in final form of Begging Call (Thorneycroft 1967).

Location Call

Tseet (or seep) (Fig. 4A) is a location or contact call given by flock or pair members when feeding, often when birds are visually isolated from one another. Captives in unfamiliar surroundings use it; also heard after copulation (Thorneycroft 1967). Varies in intensity; loud seeps may precede chip-up calls (see below and Fig. 4B). Tseet begins with a downslur and continues with a fairly narrow frequency spread in 5–8 kHz range (length 0.1–0.3 s); this call is difficult for humans to localize.

Alarm Calls

An abrupt, sharp chip (chink or pink) (Fig. 4C) is loud and easy to locate. Caller is usually conspicuous and alert; calling may be accompanied by a raised crest and tail-flicking. It is “contagious” and appears to be an alarm call that varies in intensity. It is given by adults in response to a predator or observer, especially near the nest or if young are nearby. Response is greater after the eggs hatch (Tuttle 1993). This call also occurs in agonistic encounters and in response to playback of songs or distress calls. It is often heard when a flock is going to roost. Common Yellowthroats (Geothlypis trichas) and Song Sparrows (Melospiza melodia) may respond giving their own alarm calls (Tuttle 1993). The sharp chip is about 0.07 s long, with a sudden rise in frequency from about 2.0 to 5.5 kHz, followed by a slow decline to about 4.0 kHz.

Fine chip (tip) is given when caller is hidden and a predator or person approaches young. It is quiet and difficult to locate; may occur before or after sharp chip. Wild birds in captivity give this call when approached, but hand-raised birds do not (Thorneycroft 1967). Fine chip is short (about 0.02 s) and has a narrow range of frequency at about 0.7 kHz. A call described as eeek or eeee is said to express fear (J. Hailman pers. comm.).

Trills

Trills (Fig. 4D) often given by females early in breeding cycle (courtship, nest-building, and egg-laying) and are accompanied by precopulatory display. WS females in particular trill spontaneously or in response to song or song playback; may elicit approach and copulation from their TS mates even when a stranger’s song is being played (Lowther 1962, Lowther and Falls 1968, Tuttle 1993, D. J. Loncke unpubl. data). Males also use similar call in early May, often as first response to song playback; it occurs in agonistic encounters, with male’s head forward and wings fluttered (D. J. Loncke unpubl. data). Trills are of variable length (1–2 s) and consist of a series of similar notes 0.02–0.07 s in duration separated by intervals of about 0.02 s. Individual notes range from about 4 to 8 kHz and appear to be split into 2 sounds overlapping in frequency. Some individual variation apparent.

Chip-Up

Chip-up (Fig. 4B), another repetitious call accompanied by a head-forward display, often precedes fighting. Commonly heard after female arrives in spring; may be given by males and especially by females in response to song playback. Captives give this call when displacing other individuals. In wild pairs, males or females often give chip-up when arriving at nest with food, apparently causing mate to vacate nest (JGK). Chip-up begins with a typical sharp chip followed by 2–10 up sounds (sounds are about 0.05 s and intervals about 0.12 s) ranging from 1 to 3.5 kHz and shaped like inverted "Us" on a sonogram.

Distress Calls

Distress calls, given when young or adults are restrained (by predator, mist net, handling), elicit approach, sharp chips, distraction displays, or attack on a predator, most strongly in late nestling and fledgling stages of breeding (Stefanski and Falls 1972a). Neighboring pairs may respond. Weaker interspecific responses to playbacks of each other’s distress calls occur among White-throated, Song, and Swamp sparrows (Melospiza georgiana). Possibly derived from rasp of young, White-throat distress calls are loud, wide-band sounds showing frequency modulation (about 34 Hz) of 2 carriers (2 voices) centered at about 6.6 kHz. Overall frequency ranges from about 5.5 to 7.8 kHz (for analysis and illustration see Stefanski and Falls 1972b).  Given in bursts of sounds, each sound about 0.2 s, separated by shorter intervals.

All calls described above are given by both sexes and occur in hand-raised and wild birds. Sharp chips and chip-ups are a little different in captives, suggesting that some learning takes place. Some captives use learned abnormal sounds as calls (Thorneycroft 1967).

Calls as well as song are used in territorial encounters. Collins and Houtman (1999) reported that WS and TS males did not differ in their use of song or pink calls in response to song playback. TS males gave more seeps and trills which these authors interpreted as less aggressive signals. More research is needed to elucidate the functions of calls.

Nonvocal Sounds

Apart from wing-fluttering in some displays, none known.