From studies of captive-reared individual males (n = 5), begins to sing formless songs within 2–3 wk after hatching, as is typical in many passerines; begins singing loudly and frequently at about 2 mo old, when learns quickly to sing species-typical songs (Ficken 1962a). By 5 mo, sings adult songs, including accented-ending type (see below; Ficken 1962a). No information on early song development in wild. Yearling male in first spring sings incompletely crystallized songs (Ficken 1962a, Lemon et al. 1994; C. A. Staicer and V. Ingalls, unpublished data), indicating that song maturation in the wild is relatively prolonged compared to that of some other species (see Breeding: Young Birds). Song maturation appears to occur independently and more rapidly than plumage maturation (which is also delayed), suggesting that delayed development is not generalized in all traits (Germain et al. 2012).
Song consists of 2–11 repeated, high-frequency (3–8 kHz, average 6.2 kHz) phrases or notes, thin and often sibilant in quality; some songs end with distinctive, accented phrase (Saunders 1951a, Bent 1953b, Griscom and Sprunt 1957, Curson et al. 1994, C. A. Staicer and V. Ingalls, personal communication). Male typically sings several song types. Both sexes utter variety of vocalizations in diverse contexts.
Typically male alone sings, primarily during breeding season and in spring migration (Bent 1953b, TWS), but also occasionally in overwintering areas (TWS, RTH; see Phenology). Song consists of repeated, evenly spaced notes or phrases, which vary both within most individuals over time, and among individuals within local population. Different songs may be transcribed as see see see see, tsit tsit tsit tsit, tsee tsee tsee tsee, tsee-bit tsee-bit tsee-bit tsee-bit, tsita tsita tsita tsita, wee-see wee-see wee-see wee-see, tsee tsee tsee tseeo, or tsee tsee tsee tsiee, etc. (given here arbitrarily as 4-note songs). Songs vary by number of repeated notes (or phrases), by whether or not they end on distinctive note(s), and by how an individual with > 1 song type patterns them sequentially—all depending on context, as described below.
Two categories of male songs include those with either an accented-ending or an unaccented-ending (Ficken 1962a). Accented-ending songs are usually created by rapid upward, then downward, frequency sweep (resembling inverted chevron in sonograms; see Figure 2A and B; see also Figure 1 in Lemon et al. 1985). Occasionally, accented song ends with just an upward or downward frequency sweep; sometimes penultimate syllable or note differs from all others (Figure 2B; Saunders 1951a, Bent 1953b, Lemon et al. 1985, Curson et al. 1994). Unaccented-ending songs (Figure 2C–F; see also Figure 2 in Lemon et al. 1985) lack accented terminal note, and generally comprise 1–2 repeated phrases or notes that are usually more complicated in terms of frequency sweeps than are accented-ending song phrases (C. A. Staicer, personal communication). Occasionally, song notes increase progressively in pitch within a song. Song typically lasts 1–1.5 s, and phrases are delivered in rapid succession.
Lemon et al. 1985 noted that some males never sing accented song, and reclassified songs into 2 “modes” based on pattern of song repetition: In “repeat mode,” delivers only 1 song type, which for 85% of individuals in this study consisted of accented-ending notes; in “serial mode,” 2–8 (average 4.3) different unaccented songs are given in random order (Lemon et al. 1987). The 15% of males with no accented song repeated 1 unaccented song type without progressing to serial mode. In Hubbard Brook Experimental Forest, New Hampshire, only 70% of males had accented-ending notes in their repeat-mode song; 25% sang a repeat-mode song that was indistinguishable from one of the serial-mode songs used by other males in the population; 5% either lacked chevron accent note or had distinctive introductory notes from other accented-ending songs (n = 100; C. A. Staicer and V. Ingalls, unpublished data). In both studies, proportion of birds with accented-ending songs did not vary with age of male. Yearling compared with older male songs tend to be shorter and more predictable in duration, have a higher frequency at maximum amplitude, have a larger first category syllable bandwidth, and a smaller high-frequency sweep bandwidth (Germain et al. 2012).
Male often gives muted or whispered songs of both accented and unaccented song types (Ficken 1962a, Ficken 1963a, TWS), typically audible only about 5–10 m from singer, and usually when male is near mate or when 2 males have been involved in territory dispute. Rarely, females have been observed to utter a few serial-mode song phrases, often weakly and near nest (TWS; see also Bent 1953b). Cause or function of muted songs and of female song not known.
Various other vocalizations have been described. Two usually distinctive Chip call notes (Ficken 1962a) are Metallic Chip, given at rates of > 1/s, often in context of retreat motivation—e.g., while mobbing a potential predator—and Harsh Chip, most often given by reproductively active female toward her mate. Redstart Chip notes may be described as sharp, and sweet-sounding—slightly higher-pitched and sweeter than Yellow Warbler (Setophaga petechia) chips (Getty 1993). Both sexes also produce almost inaudible Tsip call, which Ficken (Ficken 1962a) termed a “sibilant zeep,” from nest or upon encountering mate. Metallic Chip and Tsip also produced frequently during winter season (TWS). Male sometimes gives long series of high-pitched calls transcribed as Titi (Ficken 1962a), in context of territorial disputes with other males, and nesting female occasionally gives similar call when threatened by human intruder (TWS). Late nestling-stage chick and fledgling up to age of about 1 wk out of nest also give Titi calls, which probably provide directional and motivational information to food-provisioning adult (TWS). High-pitched Alarm Call with short duration (rapid onset, rapid termination) given in presence of Sharp-shinned Hawk (Accipiter striatus; TWS). Additional calls include Screech, given by fledglings in contexts of threatened individual distance and or potential predators; Screech calls grade with age of bird into Snarl, characterized by a hissing quality, which also serves as threat display, on basis of reaction it elicits from other birds (Ficken 1962a); Twitter, given in predator Distraction Display (see Behavior: Predation), and Migration/Flight Note, described as “a high, squeaky or piercing, rising sweet” (Getty 1993: 165).
In flight also gives “usually distinctive, bisyllabic, sibilant” note ( “v-shaped” in a sonogram), which is species-specific, but fairly similar to flight calls of Black-and-white Warbler (Mniotilta varia) and Chipping Sparrow (Spizella passerina); http://pjdeye.blogspot.com/2009/02/eastern-warblers-various-calls.html, Evans and O'Brien 2002, Evans 2005). Evans and O’Brien further describe this flight call as 63.7 ms (range: 49.3-78.8) in duration, and 5.5-10 kHz in frequency. Griffiths et al. (Griffiths et al. 2016) used flight note playbacks of heterospecific flight calls to elicit flight calls of temporarily caged redstarts captured during Fall migration to study intraspecific variation in flight calls for the first time in any songbird. Their results show that redstart flight calls vary more among individuals than within (five variant call note types identified; see their Figure 1), and differ significantly by sex, but not by age, although call note types overlap considerably in all these comparisons. It is not known how this variability arises, or what function(s) it may serve, but the authors argue that social functions of these calls may be more complex than previously suspected, and this variation can be used to monitor patterns of migration.
In spring, upon arrival on breeding territory, male sings in repeat mode, most commonly with accented-ending song type (Ficken 1962a, Lemon et al. 1985). After attracting mate, male shifts gradually to singing proportionately more unaccented-ending songs (Ficken 1962a) in serial mode (Macnally and Lemon 1985, Lemon et al. 1987, Weary et al. 1994a, Staicer et al. 2006). Rate of songs produced by yearling male increases to maximum just at time that nest-building activity grades into incubation stage; rate increases twice in older male, once during egg-laying and again while feeding nestlings and fledglings (Lemon et al. 1987). Unmated males, typically yearlings, continue to sing in repeat mode for 2–5 wk while they remain unmated (Staicer et al. 2006), often moving to new sites of territorial display after days or weeks. Unmated males cease singing earlier in season than nesting birds, and subsequently depart earlier from territory (Lemon et al. 1987).
After fledging young, male often sings mixtures of serial- and repeat-mode songs, which presumably mark gradual breakdown of the more stereotyped serial-mode songs (C. A. Staicer, personal communication; TWS). Male occasionally maintains territory, and countersings with serial-mode songs as late as August in New England. Serial-mode song occurs sporadically as late as first week of September, at time of departure from breeding range (TWS). Song is infrequent and weak in late summer (late July–early August), especially during period of molt.
In winter, male occasionally sings a few weak serial-mode phrases spontaneously (TWS; contra Ficken 1962a). In Jamaica, male sings typically unaccented-ending songs in serial mode in response to stimulation from song playbacks, which in these studies were repeat-mode, accented-ending songs coupled with intense bouts of Chip call notes (Sliwa and Sherry 1992); female often responds to tape-recorded songs (P. P. Marra, personal communication). Sings spontaneously much less on overwintering grounds than species such as Black-and-white Warbler or Northern Parula (Setophaga americana) (TWS, RTH). Combines Chip call notes with various ritualized visual displays to establish and maintain winter territories (Holmes et al. 1989, Sliwa and Sherry 1992). Male begins to sing actively 2–3 wk before departure from overwintering site (P. P. Marra, personal communication) and sings frequently during spring migration (TWS).
Daily Pattern Of Vocalizing
Mated male typically sings serial-mode song continuously for about 40 min at dawn, ending at about sunrise (Lemon et al. 1993, Staicer et al. 2006). This dawn song is followed by short period of repeat-mode, usually accented-ending song, then by sporadic song of both repeat- and serial-mode throughout day (Staicer et al. 2006). Unmated male typically sings nearly continuously, especially in morning, when he takes only brief breaks; and can be distinguished from mated male by relatively high song rate of exclusively repeat mode songs (Staicer et al. 2006). This difference in singing rates of mated and unmated males at different times of day potentially biases censuses of abundance based on auditory cues (TWS).
Neither nocturnal nor aerial song display reported in this species.
In winter, rate of Chip call note delivery greatest in early morning and late afternoon, just before sunset, when some individuals can call continuously for tens of minutes at a time (TWS). Responds to vocalization playback significantly less in midday than in morning or afternoon–evening (Sliwa and Sherry 1992).
Places Of Vocalizing
Typically vocalizes from within territory. Male tends to sing in serial mode near edge of territory, particularly in prolonged bouts of countersinging with adjacent territory holder (Lemon et al. 1987). Male then returns periodically to vicinity of female or nest to deliver accented-ending songs (Ficken 1962a, TWS).
Before leaf expansion in mid- to late May at Hubbard Brook, New Hampshire, male often sings from high in bare trees (TWS). During courtship, male sings while leading and following female, from all heights in forest (ground to top of tree canopy). Male expands range of singing sites to ground or near ground while following female involved in nesting activity or during cold and windy spring days when food may be sparse higher in bare trees. Especially during incubation phase of reproductive period, male often sings from favored perches with clear spaces beneath (e.g., dead branch in treefall gap or beneath continuous canopy; C. A. Staicer, personal communication; TWS).
Repertoire And Delivery Of Songs
Moderately large vocal repertoire for a parulid warbler, averaging 4.3–4.4 different songs per older male (Lemon et al. 1985, Lemon et al. 1987). Yearling males have smaller repertoire (average 2.5 songs), in part because of the many unmated young males that sing in repeat mode (a single song type; Lemon et al. 1985). In one study, from first breeding season to second, male invariably changed (usually added) songs, whereas only 40% of older males changed songs from one year to next (Lemon et al. 1994). At Hubbard Brook, individuals used up to 5 distinctive serial-mode songs, with almost 40 distinctive serial-mode song phrases recognizable within the contiguous local population studied (C. A. Staicer and V. Ingalls, unpublished data). Over 100 serial-mode song types were documented in a New Brunswick study population, with some individuals having up to 8 serial-mode song phrases (Lemon et al. 1994).
Social Context And Presumed Function Of Vocalizations
Two song modes are functionally distinct in redstart, as in many other parulids (Spector 1992). Repeat-mode, typically accented-ending songs attract mates, both upon initial arrival at breeding area and later while seeking secondary mates (polygyny; Secunda and Sherry 1991, Lemon et al. 1994, Weary et al. 1994a). This song mode is generally the only type used by unmated males (Staicer et al. 2006). Males with more variable, ‘adult-like’ repeat-mode songs tend to pair more successfully, breed earlier, and sire more offspring (Germain et al. 2012). Within a day or so of the time a particular male and female begin to associate with each other, he begins to shift to serial-mode song, which he uses predominantly through remainder of breeding season. Serial-mode songs communicate primarily to neighboring territorial males (Spector 1992, Lemon et al. 1994, Weary et al. 1994a). For example, males punctuate bouts of repeat-mode song (directed to female) early in pair-bond formation period with short bouts of serial-mode song immediately after fight with another male (TWS). Polygynous males may also change song modes within minutes as they shuttle between multiple territories containing mates at different stages of nesting cycle (Secunda and Sherry 1991).
Distinguishes stranger from neighbor songs (Weary et al. 1992); humans can also distinguish large numbers of individual birds (R. Secunda, personal communication; C. A. Staicer, personal communication), on basis of distinctive cadences, rare song types, peculiar syllables within a song, and combinations of songs within a repertoire. Ability of redstarts to learn songs from adjacent territory holders contributes to formation of song “neighborhoods” or local dialects, based on both serial-mode songs (Lemon et al. 1994) and repeat-mode songs (Shackell et al. 1988). Penultimate and ultimate syllables of repeat songs distinguish song neighborhoods better than initial syllables do (Shackell et al. 1988).
Song dialects may result from possible advantage gained by yearling males that imitate older, more reproductively successful individuals during extended sensitive period in this species (Lemon et al. 1994). Dialects are particularly noticeable with serial-mode songs, and older males (at least 2 yrs) learn new serial-mode songs compared to what they sing at an earlier age, but rarely change repeat-mode songs, suggesting dual modes of song acquisition for the 2 song modes (Lemon et al. 1994). Such dual acquisition in redstarts may be facilitated by the fact that males typically sing loud accented songs immediately following departure from nest site containing nestlings (TWS), which could have the effect of ensuring consistency in teaching this song type compared to serial songs which continue to be learned from multiple individuals and over a period of years (Lemon et al. 1994). Differences in redstart songs between neighborhoods (dialects) do not appear to be adapted to physical characteristics of habitat (Date and Lemon 1993), but instead appear to arise simply because birds learn song types locally, from parents or neighbors. Reproductive advantage to young birds of song imitation and other learning behaviors is not known.